- Email: [email protected]
Identification of the Colored Guaiacol Oxidation Product Produced by Peroxidases
ANALYTICAL BIOCHEMISTRY ARTICLE NO.
250, 10–17 (1997)
AB972191
Identification of the Colored Guaiacol Oxidation Product Produced by Peroxidases Daniel R. Doerge,1 Rao L. Divi, and Mona I. Churchwell Division of Chemistry, National Center for Toxicological Research, U.S. Food and Drug Administration, Jefferson, Arkansas 72079-9502
Received February 14, 1997
Oxidation of guaiacol by peroxidases in the presence of H2O2 is the basis for a widely used colorimetric assay. However, the nature of the assay product, which has an absorption maximum around 470 nm, had not been determined. In the present study, we combined HPLC with a rapid scanning uv-visible detector and observed a single product with a spectrum identical to the assay product from the reaction catalyzed by lactoperoxidase. Analysis of the reaction product using on-line HPLC with atmospheric pressure chemical ionization detection (LC-APCI/MS) yielded a mass spectrum consistent with 3,3*-dimethoxy-4,4*-biphenylquinone. A minor reaction product was observed with mass spectrum consistent with 3,3*-dimethoxy4,4*-dihydroxybiphenyl. The presence of a catechol impurity in guaiacol was previously shown to yield an additional product from peroxidase-mediated oxidation based on its visible absorption (Taurog et al., 1992 Anal. Biochem. 205, 271–277). When such an incubation mixture was analyzed using LC-APCI/MS, a product with mass spectrum consistent with 3-methoxy-2*,3*,4trihydroxybiphenyl was observed. Identification of such a heterodimeric product supports the previously proposed mechanism for catechol interference in the guaiacol assay as well as the radical nature of peroxidase-catalyzed oxidation of phenols. q 1997 Academic Press
(2), 2,2*-dihydroxy-3,3*-dimethoxybiphenyl (3), or 4,4*dihydroxy-3,3*-dimethoxybiphenyl (4). In these cases, the product was isolated (e.g., synthesis or thin-layer chromatography) prior to structure determination. It has also been proposed that the guaiacol oxidation product is 3,3*-dimethoxybiphenoquinone (5, 6). Taurog et al. observed interference in the formation of the guaiacol oxidation product when using a newly purchased guaiacol reagent in assays of thyroid peroxidase (TPO)2 and lactoperoxidase (LPO) activity (6). These investigators showed that the presence of small amounts of catechol in the guaiacol led to the formation of a blue color instead of the expected amber color. They 2 Abbreviations used: TPO, thyroid peroxidase; LPO, lactoperoxidase; APCI, atmospheric pressure chemical ionization; LC, liquid chromatography; CID, collision-induced dissociation; RT, retention time; TIC, total ion chromatogram.
Peroxidases are found in microorganisms, plants, and animals. A colorimetric assay of peroxidase activity using guaiacol is a common analytical method for quantifying enzymatic activity based on the change in absorbance at 470 nm (1, 2). The nature of the ambercolored product formed from guaiacol by peroxidasecatalyzed oxidation in the presence of H2O2 has been investigated previously and described as tetraguaiacol 1 To whom correspondence should be addressed. Fax: (501) 5437720. E-mail: [email protected].
SCHEME 1. Proposed mechanism for formation of the colored guaiacol oxidation product.
10
AID
0003-2697/97 $25.00 Copyright q 1997 by Academic Press All rights of reproduction in any form reserved.
AB 2191
/
6m3a$$$361
06-17-97 23:13:16
abal
GUAIACOL OXIDATION BY PEROXIDASES
FIG. 1. Ultraviolet-visible spectrophotometry of guaiacol oxidation products. Guaiacol (5 mM) was incubated with LPO (20 nM) as described under Materials and Methods and the reaction started by addition of H2O2 (200 mM). Spectra were recorded at the times indicated using a diode array spectrophotometer.
proposed that, in addition to the guaiacol oxidation product responsible for the 470-nm absorption, in the presence of catechol, the corresponding heterodimeric biphenoquinone oxidation product, containing both guaiacol and catechol units, was formed. They also observed that other phenols interfered in the guaiacol assay for peroxidase activity (e.g., pyrogallol and hydroquinone). This suggests a common reaction mechanism involving phenoxyl radical species derived from
11
one-electron oxidation by the peroxidase in the presence of H2O2 (see Scheme 1). Resorcinol and phloroglucinol blocked guaiacol oxidation, an observation consistent with our subsequent report on the mechanismbased inactivation of LPO and TPO by these compounds (7). The widespread use of the guaiacol assay, the uncertain chemical composition of the 470-nm absorbing product, and the susceptibility to interference by phenolic cosubstrates make structural characterization of oxidized guaiacol important. In the present study, we use an HPLC separation coupled with visible detection and atmospheric pressure chemical ionization mass spectrometry (LC-APCI/MS) to identify the 470-nm absorbing product formed under assay conditions using LPO and H2O2 . MATERIALS AND METHODS
Reagents Guaiacol and catechol were purchased from Aldrich Chemical Co. (Milwaukee, WI) and used as received. LPO and 30% hydrogen peroxide were obtained from Sigma Chemical Co. (St. Louis, MO). Solutions of guaiacol and H2O2 were prepared fresh daily. LPO concen-
FIG. 2. HPLC with visible detection of guaiacol oxidation products. A 30-s incubation mixture as described in the legend to Fig. 1 was sampled by injecting a 10-ml aliquot onto the HPLC column and collecting spectral data over the wavelength range of 300–600 nm using a rapid-scanning uv-visible detector as described under Materials and Methods. The extracted chromatogram corresponding to 414 nm is shown in (A). (B) Visible spectrum of the 3.98-min peak.
AID
AB 2191
/
6m3a$$$362
06-17-97 23:13:16
abal
12
DOERGE, DIVI, AND CHURCHWELL
FIG. 3. LC/MS and LC/uv chromatograms from analysis of guaiacol oxidation mixtures. A 30-s incubation mixture as described in the legend to Fig. 1 was sampled by injecting a 10-ml aliquot onto the HPLC column in-line with a uv detector and the mass spectrometer. Mass spectral data (m/z 100–600) and uv absorbance (270 nm) were acquired sequentially. Total ion chromatogram (TIC), reconstructed m/z 244 mass chromatogram, and the uv chromatogram are shown as relative intensities.
tration was determined spectrophotometrically (e Å 114 mM01 cm01 at 412 nm) and H2O2 concentration was determined using iodometric titration.
chol (0.5 mM) was added to some guaiacol oxidation reactions as indicated. Liquid Chromatography
Guaiacol Oxidation Assay Guaiacol (5 mM final concentration) was incubated with LPO (20 nM) at 25 { 0.017C and the reaction was initiated by addition of H2O2 (200 mM) in a total volume of 1.0 mL. The reaction was monitored spectrophotometrically in 1-s acquisitions from 280 to 820 nm at specified times using a Model 8452 diode array spectrophotometer (Hewlett–Packard, Palo Alto, CA). Cate-
AID
AB 2191
/
6m3a$$$362
06-17-97 23:13:16
The liquid chromatography (LC) separation in conjunction with uv-visible or APCI/MS detection was performed with a GPM or GP40 gradient pump, respectively (Dionex, Sunnyvale, CA), using a Waters RCM 8 1 10 module containing a NovaPak C18 cartridge (8 1 100 mm, 4-mm particles, Waters Associates, Milford, MA) and gradient elution consisting of 30 – 80% acetonitrile in water over 30 min at a flow
abal
GUAIACOL OXIDATION BY PEROXIDASES
13
FIG. 4. Mass spectra of the colored guaiacol product at different sampling cone-skimmer potentials. The mass spectrum of the 5.23-min peak in Fig. 3 was obtained at 15 V (bottom) and 60 V (top) using simultaneous full-scan functions (m/z 100–600) as described under Materials and Methods.
rate of 1.0 mL/min. Small differences in retention times were observed between the two HPLC systems, presumably from the different gradient mixing volumes and column performance during pressure changes caused by effluent diversion (see below). These differences did not affect chromatographic separation efficiency. A SpectraFocus rapid-scanning HPLC detector (ThermoSeparation Products, San Jose, CA) was used to generate spectra over either uv (190 – 350 nm) or visible (350 – 600 nm) ranges. Spectral data were collected over the desired range and individual wavelength chromatograms were extracted for display as needed. A Spectra 100 variable wavelength uv detector (ThermoSeparation Products) was placed in-line with the mass spectrometer to provide an alternative means of detecting peaks. Mass Spectrometry A Platform II single quadrupole mass spectrometer (Micromass, Altrincham, UK) equipped with an atmospheric pressure ionization source and APCI was used for the mass spectrometry analyses. The total LC col-
AID
AB 2191
/
6m3a$$$362
06-17-97 23:13:16
umn effluent was delivered into the ion source (1507C) through a heated nebulizer probe (5007C) using nitrogen as the bath and probe gas. Chromatographically unretained compounds (e.g., buffer salts) were diverted to waste during the first 4 min of each run by using an automated six-port switching valve (Model TPMV, Rheodyne Co., Cotati, CA). Negative ions were acquired using two full-scan functions (m/z 100–600 in 1-s cycle times with an interchannel delay time of 0.1 s). In concert with these scan functions, the sampling cone-skimmer potential was alternated between a value of 15 and 60 V to produce varying amounts of in-source collisioninduced dissociation (CID). At 15 V, mass spectra consisted primarily of deprotonated molecules (M-H0) and at 60 V, fragment ions were formed. Background-subtracted mass spectra were obtained by averaging spectra across the respective chromatographic peak and subtracting average background spectra immediately before and after this peak. The mass spectrometer was calibrated over the range m/z 85–1200 using a solution of polyethylene glycols [PEG 200 (25 mg/mL), 300 (50 mg/mL), 600 (75 mg/mL), and 1000 (250 mg/mL)] in 50% acetonitrile in aqueous ammonium acetate (5 mM).
abal
14
DOERGE, DIVI, AND CHURCHWELL
SCHEME 2. Proposed collision-induced dissociation reactions for 3,3*-dimethoxy-4,4*-biphenoquinone.
RESULTS AND DISCUSSION
Figure 1 shows the results of repetitive spectrophotometric scanning of a reaction mixture containing LPO and guaiacol in which the reaction was initiated by the addition of H2O2 at Time 0. The typical amber-colored product (lmax Å 412, 470 nm) was rapidly produced and under these conditions, product concentration was maximal at 30 s. No evidence for the formation of intermediates was observed at 10 s. When a similar sample was analyzed using gradient HPLC with visible detection, a prominent peak with retention time (RT) Å 4.0 min was observed (see Fig. 2A). Using a rapid scanning detector, the visible spectrum shown in Fig. 2B was obtained. This spectrum is virtually identical to that observed for the entire assay solution shown in Fig. 1. Small peaks were observed at 4.8 and 8.9 min with lmax Å 276 and 290 nm, respectively (not shown). The 4.8min peak was identified as guaiacol, based on retention time and lmax . All three components had absorbance at 270 nm, so this was chosen as an alternative means to detect peaks using a fixed wavelength uv detector in-line with the mass spectrometer. The products were further characterized by LCAPCI/MS analysis of a similar 30-s reaction mixture using identical chromatographic conditions. Small differences in retention times were observed because different HPLC conditions were used, but the separation efficiency was not affected (see Fig. 3). Guaiacol eluted at RT Å 6.77 min. The major product eluted at a reten-
AID
AB 2191
/
6m3a$$$362
06-17-97 23:13:16
tion time of 5.23 min and was observed using the mass spectrometer and the in-line uv detector. The mass spectrum of the product was determined using two scan functions to give either predominately (M-H)0 ions or fragment ions produced through in-source CID reactions (8). Figure 4 shows the mass spectra obtained at two different sampling cone-skimmer potentials that produced increasing amounts of fragment ions. At 15 V, the spectrum consists primarily of m/z 244, corresponding to the radical anion derived from a dimethoxybiphenoquinone. Measurable amounts of m/z 245, corresponding to the product of in-source reduction of the quinone to the corresponding hydroquinone, were observed. This is consistent with our previous observations on APCI/MS of quinones (9). At 15 V, the loss of a methyl group to yield the m/z 229 fragment was observed, and at 60 V, loss of two methyl groups was observed, in addition to further fragmentation of the ring system by sequential loss of two CO units (see Scheme 2). These results are consistent with the proposed reaction shown in Scheme 1 in which the phenoxyl radical derived from guaiacol undergoes substitution at the para position of guaiacol. After rearrangement of the intermediate, the product 3,3*dimethoxy-4,4*-dihydroxybiphenyl is formed. The colored 3,3*-dimethoxy-4,4*-biphenoquinone is formed as a result of the oxidation of the 3,3*-dimethoxy-4,4*-dihydroxybiphenyl. The visible absorbance maximum of the product is consistent with the extended conjugation present in the biphenoquinone. The mass spectral and uv-visible data cannot exclude the formation of the corresponding 3,3*-dimethoxybiphenyl-2,2*-quinone that results from ortho coupling of the phenoxyl radicals. However, ortho substitution is considered unlikely, based on steric considerations. Furthermore, ortho substitution intermediates could not assume the planar configuration required for biphenoquinone formation. The minor product peak observed in the total ion (TIC) and uv 270-nm chromatograms of Fig. 3 at retention time 10.9 min had a deprotonated molecule mass spectrum consistent with the reduced form of the guaiacol dimer (i.e., 3,3*-dimethoxy-4,4*-dihydroxybiphenyl, M-H0 Å m/z 245 at 15 V with fragment ions at m/ z 215, 187, and 159 at 60 V, not shown). This component had uv absorbance (lmax Å 268 and 290 nm) but was devoid of absorption bands in the visible range between 350 and 600 nm. Haruchi and Yoshizaki isolated and identified 3,3*-dimethoxy-4,4*-dihydroxybiphenyl from the oxidation of guaiacol by thyroid peroxidase using thin-layer chromatography and direct insertion electron impact mass spectrometry (4). The formation of this and at least two other unidentified cyclohexanesoluble fluorescent products permitted the sensitive fluorometric detection of peroxidase activity. The isolated product had a nearly identical uv spectrum to that described above (lmax Å 268 and 291 nm). The direct identification of minor amounts of this reduced
abal
GUAIACOL OXIDATION BY PEROXIDASES
15
FIG. 5. LC/MS and LC/uv analysis from guaiacol/catechol oxidation mixtures. Guaiacol (5 mM) and catechol (0.5 mM) were incubated with LPO and H2O2 as described under Materials and Methods. After 30 s incubation time, a 10-ml sample was injected into the LC/MS system equipped with a uv detector as described in the legend to Fig. 2. Total ion chromatogram (TIC), the reconstructed m/z 231 mass chromatogram, and uv chromatogram are shown as relative intensities.
product in guaiacol oxidation mixtures suggests that it may be present because of incomplete oxidation to the biphenoquinone (see Scheme 1). Furthermore, it does not contribute to the colorimetric determination of peroxidase activity. When catechol was included in the guaiacol assay solution, the formation of a blue color (lmax Å 680 and 408 nm) was observed by visible spectrophotometry as reported previously (6). Analysis of the reaction mixture by HPLC with uv-visible detection showed several product peaks in addition to the oxidation products observed from peroxidase-mediated oxidation of either guaiacol (see Fig. 3) or catechol (not shown) alone. The
AID
AB 2191
/
6m3a$$$362
06-17-97 23:13:16
maximum wavelength of the HPLC visible detector was 600 nm, which precluded direct observation of the component responsible for the 680-nm absorption. An identical incubation mixture was also analyzed using LC-APCI/MS as described above and the uv, TIC, and mass chromatogram for m/z 231 are shown in Fig. 5. Figure 6 shows the mass spectrum of the product peak with a retention time of 9.2 min. The deprotonated molecule (m/z 231) and fragment ions corresponding to M-H-CH3 (m/z 216), guaiacol-H (m/z 123), and catechol-H (m/z 109) are all consistent with the heterodimeric 3-methoxy-4,3*,4*-trihydroxybiphenyl but do not exclude the possibility of ortho substitution
abal
16
DOERGE, DIVI, AND CHURCHWELL
FIG. 6. Mass spectra of the guaiacol/catechol oxidation mixture at different sampling cone-skimmer potentials. The mass spectrum of the 9.21-min peak in Fig. 5 was obtained at 15 V (bottom) and 60 V (top) using simultaneous full-scan functions (m/z 100–600).
products. This product, presumably the result of incomplete oxidation to the biphenoquinone, is evidence for the previously proposed mechanism (6) that is also shown in Scheme 3. The smaller peaks observed in Fig. 5 had negative ion mass spectra consistent with dimeric guaiacol species (7.9 and 10.9 min), including 3,3*-dimethoxybiphenoquinone (5.2 min). Unreacted guaiacol (6.8 min) and catechol (4.3 min) were also observed. Oxidation products from catechol alone were observed in HPLC-uv-visible experiments, but the
SCHEME 3. Proposed mechanism for the formation of heterodimeric products from peroxidase-catalyzed oxidation of catechol and guaiacol.
AID
AB 2191
/
6m3a$$$362
06-17-97 23:13:16
products had retention times õ4 min so they eluted during the part of the chromatogram that was diverted in the MS experiment. It is likely, based on the short HPLC retention time observed for the guaiacol-derived biphenoquinone (5.2 min) and the more polar nature of the catechol moiety, that the heterodimeric biphenoquinone derived from guaiacol and catechol had a retention time õ4 min and was also not observed in the LC/MS experiment. An unretained peak eluting at 1.6 min was observed using HPLC with visible detection (not shown). The visible spectrum of this peak showed increasing absorbance starting from ca. 570 nm up past the maximum range of the detector, 600 nm. Although the lmax was not recorded, it is possible that this peak is responsible for the 680-nm absorbance maximum observed under assay conditions. These results provide evidence in support of the previous conclusion of Taurog and co-workers (6) that catechol interfered with the guaiacol oxidation reaction by diverting the reaction away from guaiacol dimer formation and toward the formation of heterodimeric product(s). However, the results of the present study cannot distinguish between the possible reaction products previously proposed as the 680-nm-absorbing component,
abal
GUAIACOL OXIDATION BY PEROXIDASES
the heterodimeric biphenoquinone, a semiquinone radical, and a charge transfer complex. In addition to characterizing the colored product formed by peroxidase-mediated oxidation of guaiacol, this study provides evidence for the putative formation of phenoxyl radicals during peroxidase-catalyzed oxidation of phenols. This mechanism is consistent with the observed effects on peroxidase-catalyzed guaiacol oxidation of redox active compounds that can either inhibit (9) or stimulate (10) the reaction, depending on the relative reactivity of the intermediate.
ACKNOWLEDGMENTS R.L.D. was supported by a fellowship from the Oak Ridge Institute for Science and Education, administered through an interagency agreement between the U.S. Department of Energy and the U.S. Food and Drug Administration. We thank Dr. Alvin Taurog, University of Texas Southwestern Medical Center, and Dr. Dwight Miller, Division of Chemistry, NCTR, for helpful discussions.
AID
AB 2191
/
6m3a$$$363
06-17-97 23:13:16
17
REFERENCES 1. Maehly, A. C., and Chance, B. (1954) in Methods of Biochemical Analysis (Glick, D., Ed.), Vol. 1, pp. 357–424, Interscience, New York. 2. Chance, B., and Maehly, A. C. (1964) in Methods in Enzymology (Colowick, S. P., and Kaplan, N. O., Eds.), Vol. 2, pp. 764–775, Academic Press, New York. 3. Booth, H., and Saunders, B. C. (1956) J. Chem. Soc., 940–948. 4. Harauchi, T., and Yoshizaki, T. (1982) Anal. Biochem. 126, 278– 284. 5. Saunders, B. C., Holmes-Seidle, A. G., and Stark, B. P. (1964) in Peroxidase, pp. 26–27, Butterworth, Washington, DC. 6. Taurog, A., Dorris, M. L., and Guziec, F. S., Jr. (1992) Anal. Biochem. 205, 271–277. 7. Divi, R. L., and Doerge, D. R. (1994) Biochemistry 33, 9668– 9674. 8. Doerge, D. R., Bajic, S., and Lowes, S. (1993) Rapid Commun. Mass Spectrom. 7, 462–464. 9. Doerge, D. R., Divi, R. L., and Taurog, A. (1997) Chem. Res. Toxicol. 10, 49–58. 10. Divi, R. L., and Doerge, D. R. (1996) Chem. Res. Toxicol. 9, 16– 23.
abal
250, 10–17 (1997)
AB972191
Identification of the Colored Guaiacol Oxidation Product Produced by Peroxidases Daniel R. Doerge,1 Rao L. Divi, and Mona I. Churchwell Division of Chemistry, National Center for Toxicological Research, U.S. Food and Drug Administration, Jefferson, Arkansas 72079-9502
Received February 14, 1997
Oxidation of guaiacol by peroxidases in the presence of H2O2 is the basis for a widely used colorimetric assay. However, the nature of the assay product, which has an absorption maximum around 470 nm, had not been determined. In the present study, we combined HPLC with a rapid scanning uv-visible detector and observed a single product with a spectrum identical to the assay product from the reaction catalyzed by lactoperoxidase. Analysis of the reaction product using on-line HPLC with atmospheric pressure chemical ionization detection (LC-APCI/MS) yielded a mass spectrum consistent with 3,3*-dimethoxy-4,4*-biphenylquinone. A minor reaction product was observed with mass spectrum consistent with 3,3*-dimethoxy4,4*-dihydroxybiphenyl. The presence of a catechol impurity in guaiacol was previously shown to yield an additional product from peroxidase-mediated oxidation based on its visible absorption (Taurog et al., 1992 Anal. Biochem. 205, 271–277). When such an incubation mixture was analyzed using LC-APCI/MS, a product with mass spectrum consistent with 3-methoxy-2*,3*,4trihydroxybiphenyl was observed. Identification of such a heterodimeric product supports the previously proposed mechanism for catechol interference in the guaiacol assay as well as the radical nature of peroxidase-catalyzed oxidation of phenols. q 1997 Academic Press
(2), 2,2*-dihydroxy-3,3*-dimethoxybiphenyl (3), or 4,4*dihydroxy-3,3*-dimethoxybiphenyl (4). In these cases, the product was isolated (e.g., synthesis or thin-layer chromatography) prior to structure determination. It has also been proposed that the guaiacol oxidation product is 3,3*-dimethoxybiphenoquinone (5, 6). Taurog et al. observed interference in the formation of the guaiacol oxidation product when using a newly purchased guaiacol reagent in assays of thyroid peroxidase (TPO)2 and lactoperoxidase (LPO) activity (6). These investigators showed that the presence of small amounts of catechol in the guaiacol led to the formation of a blue color instead of the expected amber color. They 2 Abbreviations used: TPO, thyroid peroxidase; LPO, lactoperoxidase; APCI, atmospheric pressure chemical ionization; LC, liquid chromatography; CID, collision-induced dissociation; RT, retention time; TIC, total ion chromatogram.
Peroxidases are found in microorganisms, plants, and animals. A colorimetric assay of peroxidase activity using guaiacol is a common analytical method for quantifying enzymatic activity based on the change in absorbance at 470 nm (1, 2). The nature of the ambercolored product formed from guaiacol by peroxidasecatalyzed oxidation in the presence of H2O2 has been investigated previously and described as tetraguaiacol 1 To whom correspondence should be addressed. Fax: (501) 5437720. E-mail: [email protected].
SCHEME 1. Proposed mechanism for formation of the colored guaiacol oxidation product.
10
AID
0003-2697/97 $25.00 Copyright q 1997 by Academic Press All rights of reproduction in any form reserved.
AB 2191
/
6m3a$$$361
06-17-97 23:13:16
abal
GUAIACOL OXIDATION BY PEROXIDASES
FIG. 1. Ultraviolet-visible spectrophotometry of guaiacol oxidation products. Guaiacol (5 mM) was incubated with LPO (20 nM) as described under Materials and Methods and the reaction started by addition of H2O2 (200 mM). Spectra were recorded at the times indicated using a diode array spectrophotometer.
proposed that, in addition to the guaiacol oxidation product responsible for the 470-nm absorption, in the presence of catechol, the corresponding heterodimeric biphenoquinone oxidation product, containing both guaiacol and catechol units, was formed. They also observed that other phenols interfered in the guaiacol assay for peroxidase activity (e.g., pyrogallol and hydroquinone). This suggests a common reaction mechanism involving phenoxyl radical species derived from
11
one-electron oxidation by the peroxidase in the presence of H2O2 (see Scheme 1). Resorcinol and phloroglucinol blocked guaiacol oxidation, an observation consistent with our subsequent report on the mechanismbased inactivation of LPO and TPO by these compounds (7). The widespread use of the guaiacol assay, the uncertain chemical composition of the 470-nm absorbing product, and the susceptibility to interference by phenolic cosubstrates make structural characterization of oxidized guaiacol important. In the present study, we use an HPLC separation coupled with visible detection and atmospheric pressure chemical ionization mass spectrometry (LC-APCI/MS) to identify the 470-nm absorbing product formed under assay conditions using LPO and H2O2 . MATERIALS AND METHODS
Reagents Guaiacol and catechol were purchased from Aldrich Chemical Co. (Milwaukee, WI) and used as received. LPO and 30% hydrogen peroxide were obtained from Sigma Chemical Co. (St. Louis, MO). Solutions of guaiacol and H2O2 were prepared fresh daily. LPO concen-
FIG. 2. HPLC with visible detection of guaiacol oxidation products. A 30-s incubation mixture as described in the legend to Fig. 1 was sampled by injecting a 10-ml aliquot onto the HPLC column and collecting spectral data over the wavelength range of 300–600 nm using a rapid-scanning uv-visible detector as described under Materials and Methods. The extracted chromatogram corresponding to 414 nm is shown in (A). (B) Visible spectrum of the 3.98-min peak.
AID
AB 2191
/
6m3a$$$362
06-17-97 23:13:16
abal
12
DOERGE, DIVI, AND CHURCHWELL
FIG. 3. LC/MS and LC/uv chromatograms from analysis of guaiacol oxidation mixtures. A 30-s incubation mixture as described in the legend to Fig. 1 was sampled by injecting a 10-ml aliquot onto the HPLC column in-line with a uv detector and the mass spectrometer. Mass spectral data (m/z 100–600) and uv absorbance (270 nm) were acquired sequentially. Total ion chromatogram (TIC), reconstructed m/z 244 mass chromatogram, and the uv chromatogram are shown as relative intensities.
tration was determined spectrophotometrically (e Å 114 mM01 cm01 at 412 nm) and H2O2 concentration was determined using iodometric titration.
chol (0.5 mM) was added to some guaiacol oxidation reactions as indicated. Liquid Chromatography
Guaiacol Oxidation Assay Guaiacol (5 mM final concentration) was incubated with LPO (20 nM) at 25 { 0.017C and the reaction was initiated by addition of H2O2 (200 mM) in a total volume of 1.0 mL. The reaction was monitored spectrophotometrically in 1-s acquisitions from 280 to 820 nm at specified times using a Model 8452 diode array spectrophotometer (Hewlett–Packard, Palo Alto, CA). Cate-
AID
AB 2191
/
6m3a$$$362
06-17-97 23:13:16
The liquid chromatography (LC) separation in conjunction with uv-visible or APCI/MS detection was performed with a GPM or GP40 gradient pump, respectively (Dionex, Sunnyvale, CA), using a Waters RCM 8 1 10 module containing a NovaPak C18 cartridge (8 1 100 mm, 4-mm particles, Waters Associates, Milford, MA) and gradient elution consisting of 30 – 80% acetonitrile in water over 30 min at a flow
abal
GUAIACOL OXIDATION BY PEROXIDASES
13
FIG. 4. Mass spectra of the colored guaiacol product at different sampling cone-skimmer potentials. The mass spectrum of the 5.23-min peak in Fig. 3 was obtained at 15 V (bottom) and 60 V (top) using simultaneous full-scan functions (m/z 100–600) as described under Materials and Methods.
rate of 1.0 mL/min. Small differences in retention times were observed between the two HPLC systems, presumably from the different gradient mixing volumes and column performance during pressure changes caused by effluent diversion (see below). These differences did not affect chromatographic separation efficiency. A SpectraFocus rapid-scanning HPLC detector (ThermoSeparation Products, San Jose, CA) was used to generate spectra over either uv (190 – 350 nm) or visible (350 – 600 nm) ranges. Spectral data were collected over the desired range and individual wavelength chromatograms were extracted for display as needed. A Spectra 100 variable wavelength uv detector (ThermoSeparation Products) was placed in-line with the mass spectrometer to provide an alternative means of detecting peaks. Mass Spectrometry A Platform II single quadrupole mass spectrometer (Micromass, Altrincham, UK) equipped with an atmospheric pressure ionization source and APCI was used for the mass spectrometry analyses. The total LC col-
AID
AB 2191
/
6m3a$$$362
06-17-97 23:13:16
umn effluent was delivered into the ion source (1507C) through a heated nebulizer probe (5007C) using nitrogen as the bath and probe gas. Chromatographically unretained compounds (e.g., buffer salts) were diverted to waste during the first 4 min of each run by using an automated six-port switching valve (Model TPMV, Rheodyne Co., Cotati, CA). Negative ions were acquired using two full-scan functions (m/z 100–600 in 1-s cycle times with an interchannel delay time of 0.1 s). In concert with these scan functions, the sampling cone-skimmer potential was alternated between a value of 15 and 60 V to produce varying amounts of in-source collisioninduced dissociation (CID). At 15 V, mass spectra consisted primarily of deprotonated molecules (M-H0) and at 60 V, fragment ions were formed. Background-subtracted mass spectra were obtained by averaging spectra across the respective chromatographic peak and subtracting average background spectra immediately before and after this peak. The mass spectrometer was calibrated over the range m/z 85–1200 using a solution of polyethylene glycols [PEG 200 (25 mg/mL), 300 (50 mg/mL), 600 (75 mg/mL), and 1000 (250 mg/mL)] in 50% acetonitrile in aqueous ammonium acetate (5 mM).
abal
14
DOERGE, DIVI, AND CHURCHWELL
SCHEME 2. Proposed collision-induced dissociation reactions for 3,3*-dimethoxy-4,4*-biphenoquinone.
RESULTS AND DISCUSSION
Figure 1 shows the results of repetitive spectrophotometric scanning of a reaction mixture containing LPO and guaiacol in which the reaction was initiated by the addition of H2O2 at Time 0. The typical amber-colored product (lmax Å 412, 470 nm) was rapidly produced and under these conditions, product concentration was maximal at 30 s. No evidence for the formation of intermediates was observed at 10 s. When a similar sample was analyzed using gradient HPLC with visible detection, a prominent peak with retention time (RT) Å 4.0 min was observed (see Fig. 2A). Using a rapid scanning detector, the visible spectrum shown in Fig. 2B was obtained. This spectrum is virtually identical to that observed for the entire assay solution shown in Fig. 1. Small peaks were observed at 4.8 and 8.9 min with lmax Å 276 and 290 nm, respectively (not shown). The 4.8min peak was identified as guaiacol, based on retention time and lmax . All three components had absorbance at 270 nm, so this was chosen as an alternative means to detect peaks using a fixed wavelength uv detector in-line with the mass spectrometer. The products were further characterized by LCAPCI/MS analysis of a similar 30-s reaction mixture using identical chromatographic conditions. Small differences in retention times were observed because different HPLC conditions were used, but the separation efficiency was not affected (see Fig. 3). Guaiacol eluted at RT Å 6.77 min. The major product eluted at a reten-
AID
AB 2191
/
6m3a$$$362
06-17-97 23:13:16
tion time of 5.23 min and was observed using the mass spectrometer and the in-line uv detector. The mass spectrum of the product was determined using two scan functions to give either predominately (M-H)0 ions or fragment ions produced through in-source CID reactions (8). Figure 4 shows the mass spectra obtained at two different sampling cone-skimmer potentials that produced increasing amounts of fragment ions. At 15 V, the spectrum consists primarily of m/z 244, corresponding to the radical anion derived from a dimethoxybiphenoquinone. Measurable amounts of m/z 245, corresponding to the product of in-source reduction of the quinone to the corresponding hydroquinone, were observed. This is consistent with our previous observations on APCI/MS of quinones (9). At 15 V, the loss of a methyl group to yield the m/z 229 fragment was observed, and at 60 V, loss of two methyl groups was observed, in addition to further fragmentation of the ring system by sequential loss of two CO units (see Scheme 2). These results are consistent with the proposed reaction shown in Scheme 1 in which the phenoxyl radical derived from guaiacol undergoes substitution at the para position of guaiacol. After rearrangement of the intermediate, the product 3,3*dimethoxy-4,4*-dihydroxybiphenyl is formed. The colored 3,3*-dimethoxy-4,4*-biphenoquinone is formed as a result of the oxidation of the 3,3*-dimethoxy-4,4*-dihydroxybiphenyl. The visible absorbance maximum of the product is consistent with the extended conjugation present in the biphenoquinone. The mass spectral and uv-visible data cannot exclude the formation of the corresponding 3,3*-dimethoxybiphenyl-2,2*-quinone that results from ortho coupling of the phenoxyl radicals. However, ortho substitution is considered unlikely, based on steric considerations. Furthermore, ortho substitution intermediates could not assume the planar configuration required for biphenoquinone formation. The minor product peak observed in the total ion (TIC) and uv 270-nm chromatograms of Fig. 3 at retention time 10.9 min had a deprotonated molecule mass spectrum consistent with the reduced form of the guaiacol dimer (i.e., 3,3*-dimethoxy-4,4*-dihydroxybiphenyl, M-H0 Å m/z 245 at 15 V with fragment ions at m/ z 215, 187, and 159 at 60 V, not shown). This component had uv absorbance (lmax Å 268 and 290 nm) but was devoid of absorption bands in the visible range between 350 and 600 nm. Haruchi and Yoshizaki isolated and identified 3,3*-dimethoxy-4,4*-dihydroxybiphenyl from the oxidation of guaiacol by thyroid peroxidase using thin-layer chromatography and direct insertion electron impact mass spectrometry (4). The formation of this and at least two other unidentified cyclohexanesoluble fluorescent products permitted the sensitive fluorometric detection of peroxidase activity. The isolated product had a nearly identical uv spectrum to that described above (lmax Å 268 and 291 nm). The direct identification of minor amounts of this reduced
abal
GUAIACOL OXIDATION BY PEROXIDASES
15
FIG. 5. LC/MS and LC/uv analysis from guaiacol/catechol oxidation mixtures. Guaiacol (5 mM) and catechol (0.5 mM) were incubated with LPO and H2O2 as described under Materials and Methods. After 30 s incubation time, a 10-ml sample was injected into the LC/MS system equipped with a uv detector as described in the legend to Fig. 2. Total ion chromatogram (TIC), the reconstructed m/z 231 mass chromatogram, and uv chromatogram are shown as relative intensities.
product in guaiacol oxidation mixtures suggests that it may be present because of incomplete oxidation to the biphenoquinone (see Scheme 1). Furthermore, it does not contribute to the colorimetric determination of peroxidase activity. When catechol was included in the guaiacol assay solution, the formation of a blue color (lmax Å 680 and 408 nm) was observed by visible spectrophotometry as reported previously (6). Analysis of the reaction mixture by HPLC with uv-visible detection showed several product peaks in addition to the oxidation products observed from peroxidase-mediated oxidation of either guaiacol (see Fig. 3) or catechol (not shown) alone. The
AID
AB 2191
/
6m3a$$$362
06-17-97 23:13:16
maximum wavelength of the HPLC visible detector was 600 nm, which precluded direct observation of the component responsible for the 680-nm absorption. An identical incubation mixture was also analyzed using LC-APCI/MS as described above and the uv, TIC, and mass chromatogram for m/z 231 are shown in Fig. 5. Figure 6 shows the mass spectrum of the product peak with a retention time of 9.2 min. The deprotonated molecule (m/z 231) and fragment ions corresponding to M-H-CH3 (m/z 216), guaiacol-H (m/z 123), and catechol-H (m/z 109) are all consistent with the heterodimeric 3-methoxy-4,3*,4*-trihydroxybiphenyl but do not exclude the possibility of ortho substitution
abal
16
DOERGE, DIVI, AND CHURCHWELL
FIG. 6. Mass spectra of the guaiacol/catechol oxidation mixture at different sampling cone-skimmer potentials. The mass spectrum of the 9.21-min peak in Fig. 5 was obtained at 15 V (bottom) and 60 V (top) using simultaneous full-scan functions (m/z 100–600).
products. This product, presumably the result of incomplete oxidation to the biphenoquinone, is evidence for the previously proposed mechanism (6) that is also shown in Scheme 3. The smaller peaks observed in Fig. 5 had negative ion mass spectra consistent with dimeric guaiacol species (7.9 and 10.9 min), including 3,3*-dimethoxybiphenoquinone (5.2 min). Unreacted guaiacol (6.8 min) and catechol (4.3 min) were also observed. Oxidation products from catechol alone were observed in HPLC-uv-visible experiments, but the
SCHEME 3. Proposed mechanism for the formation of heterodimeric products from peroxidase-catalyzed oxidation of catechol and guaiacol.
AID
AB 2191
/
6m3a$$$362
06-17-97 23:13:16
products had retention times õ4 min so they eluted during the part of the chromatogram that was diverted in the MS experiment. It is likely, based on the short HPLC retention time observed for the guaiacol-derived biphenoquinone (5.2 min) and the more polar nature of the catechol moiety, that the heterodimeric biphenoquinone derived from guaiacol and catechol had a retention time õ4 min and was also not observed in the LC/MS experiment. An unretained peak eluting at 1.6 min was observed using HPLC with visible detection (not shown). The visible spectrum of this peak showed increasing absorbance starting from ca. 570 nm up past the maximum range of the detector, 600 nm. Although the lmax was not recorded, it is possible that this peak is responsible for the 680-nm absorbance maximum observed under assay conditions. These results provide evidence in support of the previous conclusion of Taurog and co-workers (6) that catechol interfered with the guaiacol oxidation reaction by diverting the reaction away from guaiacol dimer formation and toward the formation of heterodimeric product(s). However, the results of the present study cannot distinguish between the possible reaction products previously proposed as the 680-nm-absorbing component,
abal
GUAIACOL OXIDATION BY PEROXIDASES
the heterodimeric biphenoquinone, a semiquinone radical, and a charge transfer complex. In addition to characterizing the colored product formed by peroxidase-mediated oxidation of guaiacol, this study provides evidence for the putative formation of phenoxyl radicals during peroxidase-catalyzed oxidation of phenols. This mechanism is consistent with the observed effects on peroxidase-catalyzed guaiacol oxidation of redox active compounds that can either inhibit (9) or stimulate (10) the reaction, depending on the relative reactivity of the intermediate.
ACKNOWLEDGMENTS R.L.D. was supported by a fellowship from the Oak Ridge Institute for Science and Education, administered through an interagency agreement between the U.S. Department of Energy and the U.S. Food and Drug Administration. We thank Dr. Alvin Taurog, University of Texas Southwestern Medical Center, and Dr. Dwight Miller, Division of Chemistry, NCTR, for helpful discussions.
AID
AB 2191
/
6m3a$$$363
06-17-97 23:13:16
17
REFERENCES 1. Maehly, A. C., and Chance, B. (1954) in Methods of Biochemical Analysis (Glick, D., Ed.), Vol. 1, pp. 357–424, Interscience, New York. 2. Chance, B., and Maehly, A. C. (1964) in Methods in Enzymology (Colowick, S. P., and Kaplan, N. O., Eds.), Vol. 2, pp. 764–775, Academic Press, New York. 3. Booth, H., and Saunders, B. C. (1956) J. Chem. Soc., 940–948. 4. Harauchi, T., and Yoshizaki, T. (1982) Anal. Biochem. 126, 278– 284. 5. Saunders, B. C., Holmes-Seidle, A. G., and Stark, B. P. (1964) in Peroxidase, pp. 26–27, Butterworth, Washington, DC. 6. Taurog, A., Dorris, M. L., and Guziec, F. S., Jr. (1992) Anal. Biochem. 205, 271–277. 7. Divi, R. L., and Doerge, D. R. (1994) Biochemistry 33, 9668– 9674. 8. Doerge, D. R., Bajic, S., and Lowes, S. (1993) Rapid Commun. Mass Spectrom. 7, 462–464. 9. Doerge, D. R., Divi, R. L., and Taurog, A. (1997) Chem. Res. Toxicol. 10, 49–58. 10. Divi, R. L., and Doerge, D. R. (1996) Chem. Res. Toxicol. 9, 16– 23.
abal