Chem. Res. Toxicol. 1992, 5, 346-355
346
Correlation Studies of Anodic Peak Potentials and Ionization Potentials for Polycyclic Aromatic Hydrocarbons Paolo Cremonesi, Eleanor Rogan, and Ercole Cavalieri*
in Cancer, University of Nebraska M e d i c a l Center, 600 South 42nd S t r e e t , Omaha, N e b r a s k a 68198-6805
E p p l e y Institute f o r Research
Received August 26, 1991
One-electron oxidation represents one of the major metabolic pathways of bioactivation of polycyclic aromatic hydrocarbons (PAH) to ultimate carcinogens capable of binding to cellular macromolecules, thereby initiating the cancer process. Since the ionization potential (IP) is related to the ease of removal of a T electron from an aromatic molecule, a low IP is a necessary condition for the PAH to undergo one-electron oxidation. The principal aim of this study was to provide a general and simple technique suitable for obtaining IP of PAH with satisfactory accuracy. Anodic peak potentials (E ) of 90 PAH were measured by cyclic voltammetry under irreversible oxidation conditions anrcorrelated with the corresponding IP. This allowed determination of a least-squares regression line. From the corresponding equation, IP = 1.70Eap 5.29, IP can be calculated with a narrow margin of error after a simple electrochemical measure. It was also found that PAH substituted with a methyl group on a position of appreciable electron density are best represented by a different line, corresponding to the equation IP = 1.65Eap 5.27. The calculated IP were also compared to other tabulated values, determined by different experimental techniques, and our set of IP values proved to yield the most satisfactory correlation. For some PAH, further studies under reversible voltammetric conditions allowed determination of two additional parameters: formal oxidation potentials (EO) and the number of electrons (n) involved in the redox process. IP is an important parameter in predicting the metabolic activation of carcinogenic PAH.
+
+
Introductlon Covalent binding of polycyclic aromatic hydrocarbons (PAH)’ to cellular macromolecules constitutes the first critical step in the tumor initiation process (1, 2). The chemical properties of PAH and the catalytic properties of cytochrome P-450 suggest that PAH are activated by two major mechanisms: one-electron oxidation with formation of radical cations and monooxygenation to produce bay-region diol epoxides (3-7). In fact, the one-electron mechanism of activation of PAH can be catalyzed not only by horseradish peroxidase (8-1 1 ) and prostaglandin H synthase (12) but also by cytochrome P-450 (8,9,13). A correlation of ionization potentials (IP) of PAH with their extent of binding to DNA mediated by horseradish peroxidase (10) and prostaglandin H synthase (12) has been established, suggesting that a low IP is necessary for binding to occur by one-electron oxidation. Thus, accurate determination of IP for these PAH represents the first critical parameter in predicting whether or not such chemical carcinogens can be effectively activated by oneelectron oxidation in biological systems. Formation of a charge-transfer complex between the PAH (the donor) and chloranil (the acceptor) has been used previously for measuring the IP of a large number of PAH (IO). Unfortunately, this technique is limited to alternant, unsubstituted, and halogeno- or alkyl-substituted PAH, whereas other nonalternant or other substituted PAH do not form a charge-transfer complex with chloranil. Therefore, IP cannot be determined for the latter compounds. On the basis of these premises, the major objective of this work was to develop a reliable and more general procedure to measure IP of PAH, as well as other chemicals. To achieve this, anodic peak potentials (Eap)were measured by cyclic voltammetry for a large number of ~~
* To whom correspondence should be addressed.
PAH (Figure l),and these potentials were then correlated with the IP previously determined by a charge-transfer complex energy (ECT) correlation method.2 The two parameters show a linear correlation, and a regression line can be determined. From the corresponding equation, of general applicability to PAH, IP can be calculated with accuracy after measuring E,, by cyclic voltammetry.
Materials and Methods Dimethylformamide and acetonitrile were distilled over calcium hydride and were stored under argon. Potassium perchlorate and tetra-n-butylammonium hexafluorophosphate were analytical grade (Aldrich Chemical Co., Milwaukee, WI) and were used without further purification. All PAH were available in our laboratories in analytically pure form. When required, recrystallization from benzene/hexane or acetone/methanol afforded pure compounds. Because some of the PAH are hazardous chemicals, all PAH were handled according to NIH guidelines for carcinogens (15). Electrochemical Apparatus. Cyclic voltammetry analyses were run on a voltammograph Model CV 27 (BAS Bioanalytical Systems, Inc., Lafayette, IN) equipped with a standard threeelectrode cell Model C1B (BAS) with glassy carbon (3-mm diameter) or platinum (1.5-mm diameter) working electrode, platinum wire auxiliary electrode, and silver/silver chloride electrode. Voltammograms were recorded on a XY recorder, Model 100 Omnigraphic (Houston Instruments, Houston, TX), and peak potentials were read on a digital multimeter, Model 8060 (Fluke Mfg. Co., Everett, VA), to three decimal places; alternatively, the signals were sent to an Apple IIe computer via a data Abbreviations: B[a]A, benz[a]anthracene;B[a]P, benzo[a]pyrene; B[e]P, benzo[e]pyrene; DBA, dibenzanthracene; DBP, dibenzopyrene; E,, anodic peak potential(s); E,,, cathodic peak potential(s); E o , oxidation potential(s); ECT,charge-transfercomplex energy; GCE, glassy carbon electrode; IP, ionization potential(s); PAH, polycyclic aromatic hydrocarbon(s); SCE,saturated calomel electrode; TBP, tribenzopyrene. The equation IP = 1 . 2 2 8 E c ~+ 5.038 (eV) (IO)was obtained from a least-squares plot of the IP of 15 PAH determined by polarographic
oxidation (14)against the experimentally determined ECT ( r > 0.99). In turn, the IP obtained by polarographic oxidation were correlated values from the equation IP = (1.433 & 0.027)E,p(oxo)+ (5.821 f 0.009) (14).
0093-220~/92/2705-0346$03.00/0 0 1992 American Chemical Society
Zonization Potentials of PAH from Anodic Peak Potentials
Chem. Res. Toxicol., Vol. 5, No. 3, 1992 347
2
67&2 5
10
10
3
4
6
I
II
CJp 12
Ill
5
IV
1
'000 S
1
w
VI
V
7
@ @ VI I
6
VI I I
5
XI I XI IX X Figure 1. Basic structures of PAH used for correlation studies: I, anthracene; 11, benz[a]anthracene; 111, benzo[c]phenanthrene; IV, pyrene; V, benzo[a]pyrene; VI, benzo[e]pyrene; VII, perylene; VIII, benzo[ghi]perylene; IX, dibenzo[a,l]pyrene; X, anthanthrene; XI, dibenzo[a,e]pyrene; XII, dibenzo[a,h]pyrene. acquisition interface Adalab (IMI Interactive Microware, Inc., State College, PA) equipped with a fast A/D converter, Model A113 (IMI), and MiCES electrochemical software (IMI) for postrun analpis. Normally, the output signal obtained in response to the potential excitation signal was monitored as output current. However, the semidifferential mode (semidifferential logarithm of the absolute value of the output current) was often used to help resolved closely overlying peaks and eliminate some noise. The glassy carbon electrode was polished as often as needed for good reproducibility over 0.5-pm aluminum oxide and electrochemically pretreated weekly by anodization a t +1.75 V for 5 min followed by cathodization at -1.0 V for 1 min, to increase the electrochemical activity (16). For fast-sweep cyclic voltammetry a different experimental setup was used. The potential wave form generated with the programmer Model 175 (EG & G Princeton Applied Research, Princeton, NJ) was used to control the potentiostat Model 173 (EG & G PAR) equipped with digital coulometer Model 179 (EG & G PAR) and electrometer probe Model 178 (EG & G PAR). The output signal from this unit was sent to a digital storage oscilloscope, Model 310 (Nicolet Instrument Co., Madison, WI). Stored voltammograms were then plotted on the XY recorder. For these experiments, the cell was composed of a 3-neck, 50-mL round-bottom flask, outfitted with self-assembled electrodes: 1-mm diameter Pt wire working electrode, coiled Pt wire auxiliary electrode, and saturated calomel reference electrode (SCE). The latter was housed in a Teflon bridge tube separated from the bulk solution by a Vycor tip and contained the same solvent/supporting electrolyte mixture. Such a cell geometry allowed positioning of the electrodes as close as possible to each other, so as to minimize the extent of ohmic drop. The cell was kept under a constant flow of argon. Standard Cyclic Voltammetry Analyses. The solvent/ supporting electrolyte mixture was composed of 0.5 M potassium perchlorate in dimethylformamide. The potential window for analysis was set between 0 and +1.60 V vs Ag/AgC1,3 and a scan rate of 200 mV/s was used in all experiments. The temperature Ag/AgCl is generally considered unstable in dimethylformamide,and SCE should be used instead. However, we have not encountered any stability problems associated with this electrode. When potentials appeared to no longer be reproducible from time to time, we simply replaced the reference electrode with a new one.
was always 22 i~ 1 "C. A standard procedure was followed for all analyses: a 1-2 mM solution of the PAH in the solvent/ electrolyte mixture was degassed under stirring by bubbling argon into it for 5 min. When the argon flow was stopped, a constant positive pressure of the gas was maintained over the solution to prevent contamination with oxygen and moisture. While monitoring the output current, the initial potential, always 0 V, was applied to the cell. When the resulting charging current had decreased to zero, the scan was started. The Eapthus determined represent averaged values from three to four different scans. Generally the variation in E,, was in the range f3-10 mV. Fast-Sweep Cyclic Voltammetry Analyses. These experiments were aimed to measure reversible voltammograms and determine oxidation potentials (E"). The solvent/electrolyte mixture was 0.3 M tetra-n-butylammonium hexduorophosphate in acetonitrile. The potential window was set, for each PAH, within a 0-2-V interval. For each analysis, the initial scan rate (200 mV/s) was raised to 500 mV/s, 1, 2, 5, 10, 20, 50, and 100 V/s, and, in some instances, to 200,500, and 1000 V/s until the ratio of cathodic and anodic peak currents was 0.5-0.75 or, in some instances, higher. For scan rates above 10-20 V/s, a smallerdiameter electrode (Pt wire, 0.4mm) was used to minimize the ohmic drop caused by cell resistance. When satisfactory conditions were achieved, the voltammogram was recorded and the potential values (anodic and cathodic peak potentials) were read on the oscilloscope screen, taking advantage of the 12-bit digitizer resolution and expansion capabilities of the scope. Determination of Ionization Potentials. I P values used in this study were taken from the 81 PAH set previously published, based on the formation of a charge-transfer complex between the PAH and chloranil (10). Values for additional PAH, previously not determined, were measured according to the procedure therein described. Computerized Data Analysis. Linear regression analyses were carried out with Sigma-Plot V 3.00 software (Jandel Scientific, Corte Madera, CA).
Results and Dlscusslon Introductory Remarks. Before discussing the results obtained, a few points deserve special comment. Choice of the Ionization Potential Values. IP of PAH have been experimentally measured, and calculated
348 Chem. Res. Toxicol., Vol. 5, No. 3, 1992 by a variety of techniques (14, 17-29). Values obtained by gas-phase measures (i.e., photoionization, electron impact, electronic absorption) are generally regarded as the most reliable experimental values, and several studies have correlated these values with spectroscopic and electrochemical parameters such as charge-transfer energies, polarographic half-wave potentials, voltammetric oxidation potentials, etc. (14, 19-26, 30-32). In our opinion, measurement of appropriate parameters, such as charge-transfer energies, in the liquid phase is preferable since it resembles more closely the biological environment in which the metabolic oxidative processes occur. From these experimental values, IP can then be calculated via correlation studies. Indeed, solvation energies for charged species, such as radical cations generated by abstraction of one electron, may be quite large even in aprotic solvents. This significant contribution of solvation energies to the experimental value of the IP is actually a part of the measurement in solution, whereas it can only be estimated, although rather accurately, in gas phase. Furthermore, considering the practical applications of this work, the experimental value of the IP for a particular PAH is most meaningful when compared to the values of other PAH, rather than used per se as an absolute value. We wish to emphasize that the scope of this study is not to obtain a compilation of quantum mechanical parameters of rather imaginary entities (gas-phase PAH), but to provide with ease and satisfactory accuracy values to be used as guidelines for predicting metabolic pathways. Due to the physical properties of PAH, another important factor is that gas phase-measured LP are available only for unsubstituted PAH (17,18,27-29), whereas the corresponding values extrapolated from correlation lines have been determined for a much larger set of PAH (IO, 13,including various halogeno- and alkyl-substituted PAH that represent a more complete set of carcinogenic PAH. All these facts justify the choice of a liquid-phase measurement for PAH to be used in carcinogenesis studies. Choice of Electroanalytical Conditions. Two aspects must be considered in the correlation of IP and E . First, it may seem awkward to correlate electrochemicaly measured parameters (E,,) with IP originally obtained from a correlation already involving electrochemical values (polarographic half-wave potentials) (10). However, as stated above, this set of data represents the only alternative to the use of IP measured in gas phase or derived from quantum mechanical calculations. We have included some of these data in our study, in order to compare these values with the ones selected as the main set for correlation studies. The second important aspect to consider is related to the irreversible nature of the voltammetric wave from which the Ea, were measured. It is generally best to use electrochemical potentials from reversible waves to obtain meaningful correlations with any other parameter of electron-donating ability. Very few PAH give rise to reversible voltammograms in dimethylformamide, used in all our experiments because of its excellent solvent power for PAH. Conversely, acetonitrile would be the solvent of choice for its ability to support reversible voltammetric waves, but its solvent power for PAH is limited to the extent that, for example, in the case of dibenzopyrenes, not enough material for analysis can be dissolved. Dichloromethane could be useful, having both good solvent power and ability to yield reversible voltammograms. However, once again the practical implications of this study must be considered. It is reasonable to assume that a dipolar aprotic solvent (like dimethylformamide) resem-
Cremonesi et al. bles the liquid phase in a biological environment more closely than conventional organic solvents such as dichloromethane. Furthermore, dimethylformamide is the solvent employed for preparative-scale electrochemical coupling of PAH to nucleosides. This procedure has been shown to be extremely useful in obtaining model adducts for the study of the binding of PAH to DNA (12,33). In these preparations neither acetonitrile nor dichloromethane can dissolve the required amounts of nucleoside and PAH. Thus, measuring E,, of these PAH in dimethylformamide has further significance in view of the application to electrochemical oxidations. In addition, while conventional instrumentation and fairly simple experimental conditions are required for measuring irreversible voltammetric waves (within the scan rate range of hundreds of millivolts to a few volts per second), fast-sweep analysis (up to thousands of volts per second) demands more sophisticated equipment and a more laborious experimental setup. We have determined the reversible oxidation potentials of a representative set of PAH. During this study an additional limitation to the use of reversible potentials was found: under these conditions several PAH (for instance, the dibenzopyrenes) undergo electrocrystallization at the electrode surface and consequently yield unreadable voltammograms, thus precluding even the measure of peak potentials. Since one of the goals of this study was to find a simple experimentai procedure for determining IP, the facts discussed above encouraged us to attempt to correlate irreversible E,, and IP extrapolated from charge-transfer energy measurements. The logical development of the work and the resulh obtained are discwed in detail below. Preliminary Correlation Studies. In a preliminary set of experiments, 12 selected unsubstituted and halogene and alkyl-substituted PAH were analyzed by cyclic voltammetry (34). In a plot (not shown) of E,, against IP, all PAH showed good linear correlation,with the exception of 6-methylB[a]P. The correlation coefficient, r = 0.987, dropped significantly when 6-methylB[a]P was included in the correlation. The idea that methyl substitution at a meso-anthracenic position or any other position of relatively high electron density could be related to the exceptional behavior of 6-methylB[a]P prompted us to study other methyl PAH of this type. Four additional meso-methyl PAH did not correlate well with the other PAH, but showed excellent correlation when considered by themselves. These findings led us to the conclusion that derivatives bearing a methyl group on a position of high electron density should be considered separately for these studies. Further evidence in support of this separation of PAH into two classes came from studying two additional sets: the 12 isomeric methylbenzo[a]pyrenes and methylbenz[a]anthracenes (Figure 1). The former set showed good correlation when the 6-, 1-,and 3-methyl isomers were disregarded, whereas the latter showed good correlation when the 7- and 12-methyl isomers were omitted. Therefore, for these correlation studies these particular methyl-substituted PAH (briefly referred to as "mesomethyl") were considered as a separate class with respect to the others ("normal" PAH). Irreversible Voltammetric Conditions. Correlation Studies. I. "Normal"PAH. E,, were measured for 90 PAH under irreversible conditions, i.e., in dimethylformamide containing potassium perchlorate as the supporting electrolyte at a GCE (Table I). Out of this set, 27 PAH were selected for linear regression analysis (Table IIA).
Ionization Potentials of PAH from Anodic Peak Potentials
~
Chem. Res. Toricol., Vol. 5, No. 3, 1992 349
Table I. Determination of Anodic Peak Potentials (Eap)by Cyclic VoltammetryO PAH E., IP PAH 0.912 6.74 IO-azaB[a]P 6J2-dimethylanthanthrene 0.944 6.85 benzo[ghi]perylene 6-methylanthanthrene 5-fluoro-7,12-dimethylB[a]A 0.945 6.96 naphthacene lO-fluoro-7,12-dimethylB[a]A 0.967 6.97 DB[a,h]P 0.990 6.96 4-fluoro-7,12-dimethylB [a]A anthanthrene 9-fluoro-7,12-dimethylB[a]A 1,2,3,4-tetrahydro-7,12-dimethylB[a]A 1.008 1.026 7.06 anthracene 1,3-dimethylB [a]P 1.034 6.98 6-OAcB[a]P 3,6-dimethylB [a]P DB[a,e]P 1.037 7.00 1,g-dimethylB[a]P 1.042 7.20 I-methylpyrene DB [a,i]P 1.056 7.19 7-methylB[a]A 2-fluoroDB[a,i]P 7.14 12-methylB[a]A 1.073 1 I-methylB[a]P 7.11 6-chloroB[a]P 1.074 3-methylB[a]P 1.077 7.10 6,8,12-trimethylB[a]A 4,5-dimethylB[a]P 7.27 12-ethylB [a]A 1.078 3-fluoroDB[a ,i]P 7.12 6-bromoB [a]P 1.079 I-methylB[a]P 2-methylpyrene 1.082 7.12 3-methylcholanthrene 7.14 8-methylB[a]A 1.086 12-methylB[a]P 7.23 2-methylB[a]A 1.090 2,lO-difluoroDB[a,i]P 5-methylB[a]A 1.093 7.06 perylene 7.11 g-methylB[a]A 1.093 7,IO-dimethylB[a]P 1.096 7.17 10-methylB[a]A 7-methylB[a]P 4-methylB[a]A 1.098 7.19 2-methylB [alp 7.14 1.099 9-methylB[a ]P WQIA 1.100 1I-methylB[a]A 7.08 6-methylB[a ]P 1.106 7.16 6-methylB[a]A 4-methylB[a]P 1.108 7.14 6,8-dimethylB[a]A 8-fluoro-3-methylcholanthrene 7.14 1.110 pyrene 9,10-dimethylanthracene 7.17 3-methylB[a]A 1.110 5-methylB[a]P 7.17 1.112 I-methylB[a]A 10-methylB[a]P 7.18 cyclopenta[cdlpyrene 1.115 8-methylB[a]P . . 7.17 1.118 DB [a,h]A l0-fluoro-3-methylcholanthrene 7.09 1.119 BlelP 6-ethylB[a]P DB[’a,c]A 7.23 1.121 B[aIP 7.31 1.133 DB[aJ]A TB[a,e,i]P 7.26 5-methylchrysene 1.140 9-fluoroB[a ]P 7.27 1.143 DB[e,l]P DB[a,l]P 7.20 4-methylchrysene 1.150 3,4-dihydrocyclopenta[cd]pyrene 7.23 3-methylchrysene 1.157 6-fluoroB[a]P 7.29 picene 1.158 lO-fluoroB[a]P 7.32 1.168 I-methylchrysene 8-fluoroB[a]P 7.31 chrysene 1.182 7-fluoroB[a]P 7.25 1.189 benzo [c]phenanthrene 7,8,9,10-tetrahydroB [a]P 7.22 1.198 naphthalene 7,12-dimethylB[a]A 7.24 triphenylene 1.205 2-fluoro-7,12-dimethylB[a]A
E..
~~
1.210 1.224 1.224 1.225 1.226 1.230 1.231 1.235 1.239 1.243 1.250 1.256 1.258 1.261 1.268 1.272 1.281 1.285 1.287 1.288 1.290 1.291 1.291 1.292 1.293 1.298 1.301 1.304 1.312 1.321 1.329 1.364 1.367 1.397 1.410 1.436 1.441 1.442 1.447 1.458 1.463 1.519 1.558 ndb nd
IP 7.40 7.32 7.26 7.27 7.29 7.29 7.43 7.29 7.35 7.36 7.37 7.38 7.26 7.30 7.43 7.30 7.45 7.46 7.46 7.46 7.46 7.54 7.48 7.50 7.45 7.50 7.52 7.50 7.62 (7.7) 7.82 (7.7)
(7.8) 7.93 8.13 8.15
Eapvalues are expressed in V vs Ag/AgCl. Corresponding ionization potentials (IP) were obtained from charge-transfer complex formation with chloranil (IO) and are expressed in eV.* IP values in parentheses are approximated to only the first decimal place. bnd, not detected.
The criteria for the choice were the following: first, a s m d set of data points was deemed more compatible with both ease of manipulation and accuracy; second, PAH with only approximate IP [Le., dibenz[a,c]anthracene, dibenz[a,h]anthracene, chrysene (IO)]were disregarded. The selected IP showed good linear correlation with the corresponding E,; coefficient of determination 0.970, coefficient of correlation, r, 0.985. Thus, a least-squares regression line was determined. IP = 1.70E,, + 5.29 (1) The corresponding plot is shown in Figure 2A. The following parameters were calculated for the fitted line: 95% confidence limits on slope, 1.58-1.81; on intercept, 5.15-5.43. The regression line is characterized by moderate values of standard deviation (0.0373) and standard errors (SE of slope, 0.0591; of intercept, 0.0711). Most important when considering the practical applications of this work, IP values calculated from eq 1in all cases approximate very closely the experimental values (visually,this translates into the fact that many data points in Figure 2A are directly intersected by the line). The 95% confidence limits of calculated IP values are found in Table IIIA: further evidence for the accuracy of the correlation
comes from the fact that this range is only 0.03 eV wide for the majority of points. Most of the data points used for this correlation have IP falling in the interval between 7.00 and 7.50 eV (Figure 2A). This choice has been made because PAH having IP within this range are more likely to be susceptible to bioactivation by one-electron oxidation. Therefore, the greatest accuracy for the fitted line is most desirable in this region. As a first application of these results, we calculated from eq 1IP that previously could not be precisely determined, due to intrinsic limitations of the charge-transferprocedure with chloranil (IO). These potentials are listed in Table IV, as well as, when available, corresponding values published in the literature. As can be seen, the calculated potentials are in good agreement with tabulated data. Irreversible Voltammetric Conditions. Correlation Studies. 11. “meso-Methyl”PAH. As a consequence of the preliminary work discussed above, a separate correlation was made for those PAH bearing the methyl group on a position of appreciable electron density. Following the same criteria already discussed, 14 PAH were selected (Table IIB). The plot of E, against IP showed excellent linear correlation for these PAH. The following parameters
Cremonesi et al.
350 Chem. Res. Toxicol., Vol. 5, No. 3, 1992 Table 11. Correlation Studies of Selected PAH: Reference 10 and Measured E,, PAH E., (A) ‘Normal” PAH DBla.hlP 0.967 0.990 Lthanthrene 1.077 4,5-dimethylB[a]P 1.093 perylene 1.093 7,lO-dimethylB [a]P 1.098 2-methylB[a]P 1.106 4-methylB [a]P 1.110 5-methylB[a]P 1.115 8-methylB[a]P 1.121 BbIP 1.143 DB[a,l]P 1.158 10-fluoroB[a]P 1.182 7-fluoroB[a]P 1.189 7,8,9,10-tetrahydroB[a]P 1.210 10-azaB[a]P 1.224 benzo[ghi]perylene 1.231 anthracene 1.239 DB[a,e]P 1.243 1-methylpyrene 1.281 2-methylpyrene 1.287 2-methylB[a]A 1.292 Blal-4 1.293 11-methylB[a]A 1.304 Pyrene 1.312 3-methylB[a]A 1.367 BkIP 1.558 benzo[c] phenanthrene
IP from
6.97 6.96 7.10 7.06 7.11 7.19 7.16 7.17 7.18 7.23 7.27 7.29 7.31 7.25 7.40 7.32 7.43 7.35 7.36 7.45 7.46 7.54 7.48 7.50 7.52 7.62 7.93
(B) ”meso-Methyl” PAH 0.912 6J2-dimethylanthrene 0.944 6-methylanthanthrene 1.034 3,6-dimethylB[a1P 1.037 1,6-dimethylB[a]P 1.100 6-methylB[alp 1.110 9,lO-dimethylanthracene 1.198 7,12-dimethylB[a]A 1.205 2-fluoro-7,12-dimethylB[a]A 1.224 5-fluoro-7,12-dimethylB[a]A 1.225 lO-fluoro-7,12-dimethylB[a]A 1.226 4-fluoro-7,12-dimethylB[a]A 1.230 9-fluoro-7,12-dimethylB[a]A 1.250 7-methylB[a]A 1.256 12-methylB[a]A
6.74 6.85 6.98 7.00 7.08 7.14 7.22 7.24 7.26 7.27 7.29 7.29 7.37 7.38
IP
~~
L
I
2-
Table 111. Correlation Studies of Selected PAH: IP from Reference 10 vs IP Calculated from E.. 95% conf PAH IP” calcd IPb limitsc (A) “Normal” PAH anthanthrene 6.96 6.97 6.94-7.00 DB[a,h]P 6.97 6.93 6.90-6.96 7.06 perylene 7.14 7.13-7.16 7.10 4,5-dimethylB[a]P 7.12 7.10-7.14 7.11 7.14 7.13-7.16 7,lO-dimethylB[alp 7.16 4-methylB [a]P 7.17 7.15-7.18 7.17 7.17 5-methylB[a]P 7.16-7.19 7.18 7.18 7.17-7.20 8-methylB[a]P 7.19 7.14-7.17 2-methylB [a]P 7.15 7.23 7.19 7.17-7.21 B[aIP 7.25 7,8,9,10-tetrahydroB[a]P 7.31 7.29-7.32 7.27 7.23 7.21-7.24 DB[a ,l]P 7.29 7.26 7.24-7.27 10-fluoroB[a]P 7.31 7-fluoroB[a]P 7.30 7.28-7.31 7.32 7.37 benzo[ghi]perylene 7.35-7.38 7.35 7.39 7.38-7.41 DB[a,e]P 7.36 7.40 7.38-7.4 1 1-methylpyrene 7.40 7.34 7.33-7.36 lO-azaB[a]P 7.43 anthracene 7.38 7.36-7.39 7.45 7.46 7.45-7.48 2-methylpyrene 7.46 2-methylB [a]A 7.47 7.46-7.49 7.48 11-methylB[a]A 7.48 7.47-7.50 7.50 7.48-7.52 7.50 PFene 7.52 7.52 3-methylB [a]A 7.50-7.54 7.54 7.48 7.46-7.50 B[aIA 7.62 7.61 7.59-7.63 WeIP 7.93 7.93 benzo[ c]phenanthrene 7.89-7.98 (B) ’meso-Methyl” PAH 6,12-dimethylanthanthrene 6.74 6.77 6-methylanthanthrene 6.85 6.83 3,6-dimethylB[a]P 6.98 6.98 1,6-dimethylB[a]P 7.00 6.98 6-methylB[a]P 7.08 7.09 9,lO-dimethylanthracene 7.14 7.10 7,12-dimethylB[a]A 7.22 7.25 2-fluoro-7,12-dimethylB[a]A 7.24 7.26 5-fluoro-7,12-dimethylB[a]A 7.26 7.29 lO-fluor0-7,12-dimethylB[a]A 7.27 7.29 4-fluoro-7,12-dimethylB[a]A 7.29 7.29 9-fluor0-7,12-dimethylB[a]A 7.29 7.30 7-methylB[a]A 7.37 7.33 12-methylB[a]A 7.38 7.34
6.74-6.81 6.80-6.86 6.96-7.00 6.96-7.00 7.07-7.10 7.09-7.11 7.23-7.26 7.24-7.28 7.27-7.31 7.27-7.31 7.28-7.31 7.28-7.32 7.31-7.35 7.32-7.36
were calculated: coefficient of determination 0.982, coefficient of correlation, r, 0.991. The least-squares regression line (eq 2) thus determined is shown in Figure 2B.
Values from ref 10. *Values calculated from eq 1. 95% confidence limits of the calculated IP.
IP = 1.65Ea, + 5.27
Table IV. IP Calculated from Equation 1 PAH E.,” calcd IPb lit. IP” 1.519 chrysene 7.87 7.83 1.463 1-methylchrysene 7.78 1.458 picene 7.77 7.75 1.441 DB[e,l]P 7.74 7.75 1.447 3-methylchr ysene 7.75 1.442 4-methylchrysene 7.74 1.436 7.73 5-methylchr ysene 1.397 DB[a,c]A 7.67 7.60 1.364 DB[o,h]A 7.61 7.58 1.329 cyclopenta[cd]pyrene 7.55 1.291 4-methylB[a]A 7.48 1.291 10-methylB[a]A 7.48 1,2,3,4-tetrahydro-7,12-dimethylB[a]A1.008 7.00
The following parameters were calculated for the fitted line: 95% confidence limits on slope, 1.53-1.78; on intercept, 5.12-5.41. In this instance, too, low values for standard deviation (0.0272) and standard errors were obtained (SE of slope, 0.0645; of intercept, 0.0739). Calculated IP values are listed in Table IIIB with their 95% confidence limits. Similarly to the previous set (Table IIIA), calculated IP approximate the corresponding experimental values very closely, with a narrow range of accuracy. When these data points are added to the plot previously obtained for “normal” PAH (Figure 3), it is evident that the former set deviates from the fitted line by a larger extent. This fact proves indeed that exceptional behavior is common to PAH with a methyl substituent on a position of appreciable electron density. E,, measured for these PAH are always higher than expected, on the basis of the corresponding IP (from charge-transfer complex formation). When a methyl group is adjacent to an aromatic carbon atom with high electron density, the hyperconjugation
(I
“Measured Eap. b I P calculated from eq 1. CThecorresponding values in the literature (20); additional I P values can be found in Table VA.
energy associated with the presence of the methyl group effectively exerts additional electronic stabilization. Only the electrochemical measurement seems to reflect this anomalous behavior. This can be rationalized as follows. Actual removal of one T electron occurs only during anodic oxidation, whereas in the formation of a charge-transfer
Cremonesi et al.
352 Chem. Res. Toxicol., Vol. 5, No. 3, 1992 Table V. Correlation of E,, with IP Values in the Literature (A) E,, vs IP IP" PAH E,, 1 2 3 4 5 0.945 6.96 6.64 6.95 6.96 7.00 naphthacene 0.967 6.97 6.75 7.04 DB[a,h]P 0.990 6.96 6.84 7.11 anthanthrene 1.042 7.20 7.06 7.30 DB[a,i]P 1.093 7.06 6.83 7.11 7.07 7.03 perylene 1.121 7.23 7.15 7.37 7.21 7.19 BbIP TB[a,e,i]P 1.133 7.31 7.17 7.38 7.03 7.27 7.51 1.143 7.27 DB[a,l]P 7.31 7.13 7.35 1.224 7.32 benzo[ghi]perylene 7.43 7.37 7.23 7.43 7.43 anthracene 1.231 7.31 7.27 7.20 7.41 1.239 7.35 DB[a,e]P 7.57 7.45 7.35 7.53 1.292 7.54 BbI-4 7.53 7.58 7.72 7.55 7.50 1.304 PFene DB[a,h]A 1.364 7.42 7.58 7.57 7.80 1.367 7.62 7.60 7.73 7.69 7.56 B[eIP 1.397 7.43 7.60 7.66 7.61 DB[a,c]A DB[aj]A 1.410 7.82 7.42 7.58 7.68 7.68 DB[e,l]P 1.441 7.62 7.75 picene 1.458 7.62 7.75 7.78 7.80 chrysene 1.519 7.72 7.83 7.81 7.80 benzo[c]phenanthrene 1.558 7.93 7.76 7.86 (B) Parameters of the Corresponding Fitted Lines data sets l b 1C 2 3 4 5 0.912 0.910 0.882 0.904 0.970 0.953 coefficient of determination 0.955 0.954 0.939 0.951 0.985 0.976 coefficient of correlation 0.091 0.091 0.100 0.082 0.065 standard deviation 0.037 1.71 1.39 1.55 1.66 1.70 1.67 slope 5.33 5.73 5.50 5.13 5.29 5.35 intercept
6 7.597
7.724 7.747 7.777 7.846 7.972 7.841 8.037 7.951 8.070 8.028 8.098 8.095 8.157 8.193
6 0.952 0.975 0.041 1.12 6.51
7 6.97 6.82
6.95 6.97 7.10 6.99 7.07 7.15 7.41 7.11
7.41 7.41 7.38 7.41 7.39 7.46 7.39 7.52 7.59 7.60
7 0.885 0.941 0.083 1.27 5.66
IP values obtained from the following: (1)charge-transfer complex with chloranil (10); (2) absorption spectra (19); (3) electronic absorption (20);(4) polarogrpahic oxidation (14); (5) charge-transfer complex with chloranil (21);(6) calculation by Koopmans' theorem (22); (7) high-resolution photoelectron spectra (27-29). b27-PAH set (Table 11). 16-PAH set (set 1 in part A of this table).
reversible voltammetric wave was obtained for some PAH, thus enabling a qualitative measurement of E" (Table VI). Still, many PAH failed to yield a reversible voltammogram, which can be interpreted in terms of rate of decay of radical cations and rate of heterogeneous electron transfer. In addition, in some instances a further complication arose in the tendency of some PAH to undergo electrocrystallization at the electrode surface, yielding unreadable voltammograms. It should be noted that a direct comparison between reversible and irreversible potentials is not appropriate, since the solvent/electrolyte mixture and reference electrode are different for each system. The two potential scales could be linked by using a recognized standard, such as ferrocene, in the two systems. However, this determination lies beyond the scope of this work, since we are interested only in a qualitative comparison between the two systems. As can be seen, Eo's show a high degree of correlation with IP values, and the points on the corresponding regression line are characterized by small standard deviations. On the basis of the improvement shown by the irreversible Eap'son switching from the limited 10-PAH to the original 27-PAH set, it seems reasonable to assume that a larger number of data points would also improve the quality of the relation between IP and E,, values. These results are not surprising, and similar studies carried out on alkylbenzenes (31) and other classes of compounds such as hydrazines (35), indoles, and indolizines (32) have shown the existence of a good correlation between E" and IP values. In spite of these encouraging results, the complexity of the instrumentation required for fast-scan voltammetry and, furthermore, the intrinsic tendency of some PAH to
Table VI. Correlation Studies under Reversible Voltammetric Conditions Part A" PAH IP E,, (irrev) E,, (rev) E" 6.96 0.990 0.940 0.920 anthanthrene pery1ene 7.06 1.093 1.052 1.031 1.106 4-methylB[a]P 7.16 1.129 1.073 6-fluoroB[a]P 7.23 1.157 1.218 1.193 6-chloroB(a]P 7.26 1.258 1.288 1.225 1.301 1.241 7.30 1.272 6-bromoB[a]P 1.168 1.228 1.196 8-fluoroB[a]P 7.32 anthracene 7.43 1.231 1.302 1.260 1.341 1.282 7.50 1.304 pyrene 1.292 1.406 1.375 w1.4 7.54 Part Bb , Ew Ew (irrevP (irrev)d E., (rev) E" coefficient of 0.970 0.816 0.905 0.917 determination 0.903 0.951 0.958 coefficient of correlation 0.985 0.084 0.061 0.056 standard deviation 0.037 1.32 1.63 1.24 1.70 slope 5.77 5.72 5.29 5.34 intercept Part A: Ionization potentials (IP, from ref lo), reversible and irreversible anodic peak potentials (Eap),and oxidation potentials (E"). Part B: Parameters of the corresponding fitted lines. 27PAH set (Table 11). dlO-PAH set (part A of this table).
yield irreversible voltammetric waves, in our opinion, pose a serious limitation to routine determination and utilization of E" values. Within the scope of this study these electrochemical parameters, and consequently the IP values, should be simply regarded as useful guidelines to explain and predict some aspects of the behavior of this class of chemicals in a biological environment. Accordingly, the qualitative nature of this study does not require a
Ionization Potentials of PAH from Anodic Peak Potentials Table VII. Determination of Anodic and Cathodic Peak Separation by Cyclic Voltammetry under Reversible Conditions PAH 6E, mVa electrodeb anthracene 105 GCE 9J0-dimethylanthracene 54 GCE 62 Pt BbIA GCE 7,12-dimethylB[a]A 50 pyrene 119 Pt 66 GCE, Pt WeIP GCE, Pt 50 WaIP Pt 6-fluoroB[a]P 50 6-chloroB[a]P 125 Pt 6-bromoB[a]P 120 Pt 6-OAcB[a]Pc GCE, Pt 8-fluoroB[a]P 65 Pt 4-methylB[a]P 112 Pt g-methylB[a]P 67 Pt perylene 53 GCE GCE, P t DB[a,e]Pd 31 GCE, Pt DB[a,i]P' GCE, Pt DB[a,h]PC 45 GCE, Pt DB [a,l]P 44 GCE anthanthrene GCE, glassy carbon eleca Peak separation, 6E = Ea4- ECv trode; Pt, platinum. CIrreversiblewave at any scan rate. dPeaks not resolved enough. e Electrocrystallization at the electrode surface.
rigorous approach and should not wander from a simple operational setup. Reversible Voltammetric Conditions. 11. Determination of n When the redox system is reversible, the anodic and cathodic peak separation, 6E (6E= E, - E ), in the voltammogram enables determination n, {Le number of electrons transferred in the redox process, since 6E is proportional to n [6E= (59 mV)/n]. Knowledge of the number of electrons transferred from the PAH to the electrode during the oxidation process is particularly important when we consider that oxidation of PAH can produce two distinct intermediates: radical cations from one-electron oxidation and dications from two-electron transfer. The latter can be composed, at least in some instances, of two sequential single-electron steps occurring so rapidly that, for practical reasons, the radical cation formed after the initial transfer is too short-lived to be considered as a real intermediate. Both radical cations and dications are electrophilic; however, they possess different properties that impart a particular reactivity to each species. Therefore, the formation and fate in a biological environment will be different for each intermediate. For these reasons, the ability to determine whether a particular PAH is oxidized by a one-electron or two-electron mechanism allows the possibility of predicting its fate in a biological system (i.e., the types of metabolites and macromolecular adducts formed through bioactivation). A rigorous determination of n through the equation 6E = (59 mV)/n requires a fully reversible system. Recently, the issue of the absolute determination of the number of electrons consumed at the electrode process was addressed (36),and the proposed method is based on a comparison between the current functions obtained by two different electrochemical techniques operated on an identical time scale. Again, our approach to this complex problem was to test whether a relatively simple measurement could give some useful indication as a comparative study, rather than to attempt a rigorous determination of absolute values. Accordingly, 20 representative PAH were analyzed under these conditions. Measured 6E values are listed in Table VII. Values close to 60 mV (theoretical 59 mV)
.
OF
Chem. Res. Toxicol., Vol. 5, No. 3, 1992 353 represent one-electron transfer; one-half of that value would correspond to a two-electron transfer. As described above, some of the dibenzopyrenes showed electrocrystallization phenomena at the platinum electrode, and this could not always be circumvented by the use of a GCE. For some PAH the anodic wave was irreversible at any scan rate. In both cases 6E could not be determined. The 6E values measured for several PAH (9,lO-dimethylanthracene, B[a]A, 7,12-dimethylB[a]A, perylene, B[e]P, B[a]P, 6-FB[a]P, 8-FB[a]P, and 6-methylB[a]P) seem to indicate single-electron electrode processes, whereas nothing can be said about anthracene, pyrene, 6-ClB[a]P, 6-BrB[a]P, and 4-methylB[a]P. In the case of the dibenzo[a]pyrenes, the measurement is complicated by the phenomena seen above, with the exception of DB[a,Z]P and anthanthrene, which give a 6E value corresponding to consumption of ca. 1.3 electrons. Previous studies have shown that anthracene and pyrene undergo two-electron transfer under electrochemcial oxidation conditions, with formation of the corresponding dication electrophilic intermediates (37), whereas the initial oxidation of B[a]P is always a one-electron process to form the radical cation (38,39). An important qualitative finding is that 6-FB[a]P shows one-electron transfer, whereas 6-ClB[a]P and 6BrB[a]P do not. It is reasonable to assume that the presence of bulky, highly polarizable heteroatoms in the latter two PAH considerably affects the a electron density. This result is particularly meaningful when considered in relation to their carcinogenic activity. In fact, the inertness to metabolism (40) and lack of carcinogenic activity (41) of 6-ClB[a]P and 6BrB[a]P compared to B[a]P and 6FB[a]P (both undergoing one-electron transfer) suggest the critical importance of single-electron oxidation in carcinogenesis.
Conclusions In conclusion, this study has demonstrated the existence of a linear correlation between E,, and IP of PAH. From this correlation, IP can be calculated with good accuracy after a simple electrochemicalanalysis. Good results have been obtained utilizing easily measurable experimental parameters: irreversible E , and IP obtained from liquid-phase measurement of ckarge-transfer energy. This procedure is much simpler than previously published similar studies utilizing reversible oxidation potentials accessible through a more complex experimental determination. We have found that PAH substituted with a methyl group at a position of appreciable electron density require a separate correlation line, since the hyperconjugation energy of the methyl group influences the electrochemical measure. In addition, a useful comparison has been made between the IP we determined and values previously measured under different experimental conditions. Lastly, from electrochemical studies under reversible conditions we have determined, for some representative PAH, the one- or two-electron nature of the redox process. Knowledge of this parameter is helpful in predicing metabolic pathways of bioactivation for PAH.
Acknowledgment. This research was supported by USPHS Grants R01 CA44686 and PO1 CA49210 from the National Cancer Institute. Institutional support to the Eppley Institute came from USPHS Grant P30 CA36727. Registry No. DB[a,h]P, 189-64-0; 1,3-dimethylB [a]P , 16757-86-1;3,6-dimethylB[a]P, 16757-91-8; 1,6-dimethylB[a]P, 16757-90-7; DB[a,i]P, 189-55-9; 2-fluoroDB[a,i]P, 73368-38-4;
364 Chem. Res. Toxicol., Vol. 5, No. 3, 1992 11-methylB[a]P, 16757-80-5; 3-methylB[a]P, 16757-81-6; 4,5dimethylB[a]P, 16757-89-4; 3-fluoroDB[a,i]P, 61735-77-1; 1methylB[a]P, 40568-90-9; 12-methylB[a]P, 4514-19-6; 2,lO-difluoroDB[a,i]P, 61735-78-2; 7,10-dimethylB[a]P, 63104-33-6; 7-methylB[a]P, 63041-77-0; 2-methylB[a]P, 16757-82-7; 9methylB[a]P, 70644-19-8; 6-methylB[a]P, 2381-39-7; 4-methylB[alp, 16757-83-8; 5-methylB[a]P, 31647-36-6; lO-methylB[a]P, 63104-32-5; &methylB[a]P, 63041-76-9;6-ethylB[a]P, 78694-66-3; B[a]P, 50-32-8; TB[a,e,i]P, 192-47-2; 9-fluoroB[a]P, 71171-93-2; DB[a,l]P, 191-30-0; 6-fluoroB[a]P, 59417-86-6; 10-fluoroB[a]P, 74018-58-9; bfluoroB[a]P, 71171-92-1; 7-fluoroB[a]P, 71511-38-1; 7,12-dimethylB[a]A, 57-97-6; lO-azaB[a]P, 189-92-4;6-OAcB[a]P, 53555-67-2; DB[a,e]P, 192-65-4; 7-methylB[a]A, 2541-69-7; 12methylB[a]A, 2422-79-9; 6-chloroB[a]P, 21248-01-1; 6,8,12-trimethylB[a]A, 20627-34-3; 12-ethylB[a]A, 18868-66-1; 6-bromoB[alp, 21248-00-0; 8-methylB[a]A, 2381-31-9; 2-methylB[a]A, 2498-76-2; 5-methylB[a]A, 2319-96-2; 9-methylB[a]A, 2381-16-0; 10-methylB[a]A, 2381-15-9; 4-methylB[a]A, 316-49-4; B[a]A, 56-55-3; 11-methylB[a]A, 6111-78-0; 6-methylB[a]A, 316-14-3; B,B-dimethylB[a]A, 317-64-6; 3-methylB[a]A, 2498-75-1; 1methylB[a]A, 2498-77-3; DB[a,h]A, 53-70-3; B[e]P, 192-97-2; DB[a,c]A, 215-58-7; DB[aj]A, 224-41-9; DB[e,l]P, 192-51-8; 6,12-dimethylanthanthrene,41217-05-4; 6-methylanthanthrene, 31927-64-7; naphthacene, 92-24-0; anthanthrene, 191-26-4; 1,2,3,4-tetrahydro-7,12-dimethylR[a]A, 67242-54-0; 3-methylcholanthrene, 56-49-5; perylene, 198-55-0; 8-fluoro-3-methylcholanthrene, 74924-89-3; g,lO-dimethylanthracene,781-43-1; l0-fluoro-3-methylcholanthrene, 74924-90-6; 3,4-dihydrocyclopenta[cd]pyrene, 25732-74-5; 7,8,9,10-tetrahydroB[a]P, 17750-93-5; 2-fluoro-7,12-dirnethylB[a]A, 68141-56-0; benzo[ghi]perylene, 191-24-2; 5-fluoro-7,12-dimethylB[a]A, 794-00-3; 10-fluoro-7,12dimethylB[a]A, 71172-13-9; 4-fluoro-7,12-dimethylB[a]A, 737-22-4; 9-fluoro-7,12-dimethylB[a]A, 71172-11-7; anthracene, 120-12-7; 1-methylpyrene, 2381-21-7; 2-methylpyrene, 3442-78-2; pyrene, 129-00-0; cyclopenta[cd]pyrene, 27208-37-3; &methylchrysene, 3697-24-3; 4-methylchrysene, 3351-30-2; 3-methylchrysene, 3351-31-3; picene, 213-46-7; 1-methylchrysene, 3351-28-8; chrysene, 218-01-9; benzo[c]phenanthrene, 195-19-7;naphthalene, 91-20-3; triphenylene, 217-59-4.
References (1) Miller, J. A. (1970) Carcinogenesis by chemicals: An overview. G. H. A. Clowes Memorial Lecture. Cancer Res. 30, 559-576. (2) Miller, E. C., and Miller, J. A. (1981) Searches for ultimate chemical carcinogens and their reactions with cellular macromolecules. Cancer 47, 2327-2345. (3) Cavalieri, E. L., and Rogan, E. G. (1984) One-electron and twoelectron oxidation in aromatic hydrocarbon carcinogenesis. In Free Radicals in Biology (Pryor, W. A., Ed.) Vol. VI, pp 323-369, Academic Press, New York, NY. (4) Cavalieri, E., and Rogan, E. (1985) Role of radical cations in aromatic hydrocarbon carcinogenesis. Enoiron. Health Perspect. 64, 69-84. (5) Cavalieri, E. L., and Rogan, E. G. (1990) Radical cations in aromatic hydrocarbon carcinogenesis. Free Radical Res. Commun. 11, 77-87. (6) Sims, P., and Grover, P. L. (1981) Involvement of dihydrodiols and diol epoxides in the metabolic activation of polycyclic hydrocarbons other than benzo[a]pyrene. In Polycyclic Hydrocarbons and Cancer (Gelboin, H. V., and Ts’o, P. 0. P., Eds.) pp 117-181, Academic Press, New York, NY. (7) Conney, A. H. (1982) Induction of microsomal enzymes by foreign chemicals and carcinogenesis by polycyclic aromatic hydrocarbons. G. H. A. Clowes Memorial Lecture. Cancer Res. 42, 4875-4917. (8) Devanesan, P. D., RamaKrishna, N. V. S., Todorovic, R., Rogan, E. G., Cavalieri, E. L., Jeong, H., Jankowiak, R., and Small, G. J. (1992) Identification and quantitation of benzo[a]pyrene-DNA adducts formed by rat liver microsomes in vitro. Chem. Res. T O X ~ C5,O302-309. ~. (9) RamaKrishna, N. V. S., Devanesan, P. D., Rogan, E. G., Cavalieri, E. L., Jeong, H., Jankowiak, R., and Small, G. J. (1992) Mechanism of metabolic activation of the potent carcinogen 7,12-dimethylbenz[a]anthracene.Chem. Res. Toxicol. 5,220-226. (10) Cavalieri, E. L., Rogan, E. G., Roth, R. W., Saugier, R. K., and Hakam, A. (1983) The relationship between ionization potential and horseradish peroxidase/ hydrogen peroxide-catalyzedbinding
Cremonesi et al. of aromatic hydrocarbons to DNA. Chem.-Biol. Interact. 47, 87-109. (11) Rogan, E. G., Cavalieri, E. L., Tibbels, S. R., Cremonesi, P., Warner, C. D., Nagel, D. L., Tomer, K. B., Cemy, R. L., and Gross, M. L. (1988) Synthesis and identification of benzo[a]pyreneguanine nucleoside adducta formed by electrochemical oxidation and by horseradish peroxidase catalyzed reaction of benzo[a]pyrene with DNA. J. Am. Chem. SOC. 110, 4023-4029. (12) Devanesan, P., Rogan, E., and Cavalieri, E. (1987) The relationship between ionization potential and prostaglandin H synthase-catalyzed binding of aromatic hydrocarbons to DNA. Chem.-Bid. Interact. 61, 89-95. (13) Cavalieri, E. L., Rogan, E. G., Devanesan, P. D., Cremonesi, P., Cerny, R. L., Gross, M. L., and Bodell, W. J. (1990) Binding of benzo[a]pyrene to DNA by cytochrome P-450-catalyzed oneelectron oxidation in rat liver microsomes and nuclei. Biochemistry 29, 482C-4827. (14) Pysh, E. S., and Yang, N. C. (1963) Polarographic oxidation 85, potentials of aromatic compounds. J. Am. Chem. SOC. 2 124-2 130. (15) NIH Guidelines for the Laboratory Use of Chemical Carcinogens (1981) NIH Publication No. 81-2385, US. Government Printing Office, Washington, DC. (16) Engstrom, R. C., and Strasser, V. A. (1984) Characterization of electrochemically pretreated glassy carbon electrodes. Anal. Chem. 56, 136-141. (17) Pullman, B., Pullman, A., Umans, R., and Maigret, B. (1968) In The Jerusalem Symposia on Quantum Chemistry and Biochemistry. Volume I. Physico-Chemical Mechanisms of Carcinogenesis. Proceedings of an International Symposium (Bergmann, E. D., and Pullman, B., Eds.) pp 326-338, Academic Press, New York, NY. (18) Wacks, M. E., and Dibeler, V. H. (1959) Electron impact studies of aromatic hydrocarbons. I. Benzene, Naphthalene, Anthracene, and Phenanthrene. J. Chem. Phys. 31, 1557-1562. (19) Matsen, F. A. (1956) Electron affinities, methyl affinities, and ionization energies of condensed ring aromatic hydrocarbons. J. Chem. Phys. 24, 602-606. (20) Briegleb, G. (1964) Electron affinity of organic molecules. Angew. Chem., Int. Ed. Engl. 3, 617-632. (21) Birks, J. B., and Slifkin, M. A. (1961) *-Electronic excitation and ionization energies of condensed ring aromatic hydrocarbons. Nature 191, 761-764. (22) Dewar, M. J. S., Hashmall, J. A., and Trinajstic, N. (1970) Ground states of conjugated molecules. XXII. Polarographic 92, reduction potentials of hydrocarbons. J. Am. Chem. SOC. 5555-5559. (23) Foster, R. (1959) Ionization potentials of electron donors. Nature 183, 1253. (24) Watanabe, K. (1957) Ionization potentials of some molecules. J . Chem. Phys. 26, 542-547. (25) Foster, R. (1960) The absorption spectra of molecular complexes. Tetrahedron 10, 96-101. (26) Bergman, I. (1954) The polarography of polycyclic aromatic hydrocarbons and the relationship between their half-wave po50,829-838. tentials and absorption spectra. Trans. Faraday SOC. (27) Clar, E., and Schmidt, W. (1977) Correlations between photoelectron and ultraviolet absorption spectra of polycyclic hydrocarbons. The perylene, coronene and bisanthene series. Tetrahedron 33, 2093-2097. (28) Schmidt, W. (1977) Photoelectron spectra of polynuclear aromatics. V. Correlations with ultraviolet absorption spectra in the catacondensed series. J. Chem. Phys. 66, 828-845. (29) Clar, E., and Schmidt, W. (1979) Correlations between photoelectron and UV absorption spectra of polycyclic hydrocarbons. The pyrene series. Tetrahedron 35, 1027-1032. (30) Parker, V. D. (1976) Energetics of electrode reactions. 11. The relationship between redox potentials, ionization potentials, electron affinities, and solvation energies of aromatic hydrocarbons. J. Am. Chem. SOC. 98, 98-103. (31) Howell, J. O., Goncalves, J. M., Amatore, C., Klasinc, L., Wightman, R. M., and Kochi, J. K. (1984) Electron transfer from aromatic hydrocarbons and their r-complexes with metals. Comparison of the standard oxidation potentials and vertical ionization potentials. J . Am. Chem. SOC. 106, 3968-3976. (32) Colonna, M., Greci, L., Poloni, M., Marrosu, G., Trazza, A., Colonna, F. P., and Distefano, G. (1986). A correlation between half-wave and ionization potentials for indoles and indolizines. J . Chem. Soc., Perkin Trans. 2, 1229-1231.
Chem. Res. Toxicol. 1992,5,355-360 (33) RamaKrishna, N.V. S., Cavalieri, E. L., Rogan, E. G., Dolnikowski, G., Cerny, R. L., Gross, M. L., Jeong, H., Jankowiak, R., and Small, G. J. (1992)Synthesis and structure determination of the adducts of the potent carcinogen 7,12-dimethylbenz[a]anthracene and deoxyribonucleosides formed by electrochemical oxidation: models for metabolic activation by one-electron oxidation. J. Am. Chem. SOC.114,1863-1874. (34) Cremonesi, P., and Cavalieri, E. (1985)Determination of oxidation potentials of aromatic hydrocarbons by cyclic voltammetry. In 1985 Electroanalytical Symposium, pp 67-70, BAS Press, West Lafayette, IN. (35) Nelsen, S. F., Blackstock, S. C., Petillo, P. A., Agmon, I., and Kaftory, M. (1987)One-electron oxidation equilibria for acylated 109,5724-5731. hydrazines. J. Am. Chem. SOC. (36)Amatore, C.,Azzabi, M., Calas, P., Jutand, A., Lefron, C., and Rollin, Y. (1990)Absolute determination of electron consumption in transient or steady state electrochemical techniques. J. Electroanal. Chem. 288,4543. (37) King, P. F., and O'Malley, R. F. (1984)Fluorination of poly-
355
cyclic aromatic hydrocarbons: charge vs. frontier orbital control in substitution reaction of radical cations. J . Org. Chem. 49, 2803-2807. (38) Jeftic, L., and Adams, R. N. (1970)Electrochemical oxidation pathways of benzo[a]pyrene. J. Am. Chem. SOC.92,1332-1337. (39) Podany, V., Benesova, M., and Bahna, L. (1979)Electrochemical properties of polycyclic compounds studied by the polarographic method in anhydrous systems. VII. Polarographic oxidation of carcinogenic hydrocarbons in dimethylformamide. Neoplasma 26,685-689. (40) Cavalieri, E. L.,Rogan, E. G., Cremonesi, P., and Devanesan, P. D. (1988)Radical cations as precursors in the metabolic formation of quinones from benzo[a]pyrene and 6-fluorobenzo[a]pyrene. Fluoro substitution as a probe for one-electron oxidation in aromatic substrates. Biochem. Pharmacol. 37, 2173-2182. (41) Cavalieri, E.,Rogan, E., Cremonesi, P., Higginbotham, S., and Salmasi, S. (1988)Tumorigenicity of 6-halogenated derivatives of benzo[a]pyrenein mouse skin and rat mammary gland. J . Cancer Res. Clin. Oncol. 114,10-15.
Antioxidant Activity of the Pyridoindole Stobadine. Pulse Radiolytic Characterization of One-Electron-Oxidized Stobadine and Quenching of Singlet Molecular Oxygen Steen Steenken,t Alfred R. Sundquist,i Slobodan V. Jovanovic,t Rowena Crockett,t and Helmut Sies*f* Institut fur Physiologische Chemie I, Uniuersitat Diisseldorf, Moorenstrasse 5, W-4000 Diisseldorf, Germany, and Max-Planck-Institut fur Strahlenchemie, Stiftstrasse 34-36, W-4330 Mulheim, Germany Received January 22, 1992 Antioxidant properties of stobadine, a pyridoindole derivative described to exhibit cardioprotective properties, were characterized. The radical scavenging potential of stobadine was evaluated using pulse radiolysis with optical detection, by which it is shown that one-electron oxidation of stobadine with radicals such as C6H50*,CC1302', Br2'-, and HO' (reaction rate constants 4X 108-1010 M-' s-l) leads to the radical cation (absorbance maxima a t 280 and 445 nm) which deprotonates from the indolic nitrogen (pK, = 5.0) to give a nitrogen-centered radical (absorbance maxima a t 275,335, and 410 nm), probably bearing a positive charge a t the pyrido nitrogen. The radical of stobadine reacts with Trolox (i.e., 6-hydroxy-2,5,7,8-tetramethylchromane-2-carboxylic acid) with a rate constant of 1.2 X lo7 M-' s-l at pH 7.0 by one-electron oxidation to yield the phenoxyl-type radical of Trolox. This reaction is reversible (12 = 2 X lo5 M-l s-l). The redox potential of stobadine at pH 7 is 0.58 V/NHE. Stobadine is also a quencher of singlet molecular oxygen (lo2) with an overall quenching rate constant of 1.3 X lo8 M-l s-l, determined with the endoperoxide of 3,3'-( 1,4-naphthylene)dipropionate(NDP02)as lo2source and by monitoring lo2photoemission with a germanium diode.
I ntroductlon Stobadine (Figure l),a novel drug of a pyridoindole structure, has been described to exhibit cardioprotective effects and to decrease functional damage to isolated rat hearts subjected to a period of ischemia followed by reperfusion (1, 2). Experiments to detect peroxidative damage to lipids suggested that these pharmacological effects arise from antioxidant properties of stobadine (3-6). In order to characterize the radical-scavenging properties of stobadine, and the spectroscopic and acidlbase prop-
* To whom correspondence should be addressed at the Institut fur Physiologische Chemie I, Universitiit Diisseldorf, Moorenstrasse 5, D-4000 Dusseldorf, Germany. + Max-Planck-Institut fiir Strahlenchemie. Universitiit Diisseldorf.
*
erties of its one-electron-oxidizedform, experiments with pulse radiolysis and time-resolved optical detection were performed. Further, the activity of stobadine as a was assessed using the quencher of singlet oxygen (lo2)' endo-peroxide of 3,3'-(1,4-naphthylene)dipropionateas a water-soluble '02 source and monitoring the infrared (1270-nm) emission of lo2(7,8). The antioxidant activity of dehydrostobadine (Figure 1) was also examined.
Materials and Methods Reagents. Stobadine, &-(-)-2,3,4,4a,5,9b-hexahydro-2,8-dimethyl-lH-pyrido[4,3-b]indole,and related compounds were synthesized at the Institute of Experimental Pharmacology, Slovak Abbreviations: lo2,singlet molecular oxygen; NDP02, the endoperoxide of 3,3'-(1,4-naphthylene)dipropionate.
0893-228x/92/2705-0355$03.00/00 1992 American Chemical Society
