The Oxidation of Indole Derivatives Catalyzed by Horseradish ...

Archives of Biochemistry and Biophysics Vol. 387, No. 2, March 15, pp. 173–179, 2001 doi:10.1006/abbi.2000.2228, available online at http://www.idealibrary.com on

The Oxidation of Indole Derivatives Catalyzed by Horseradish Peroxidase Is Highly Chemiluminescent Valdecir F. Ximenes,* Ana Campa,† and Luiz H. Catalani* ,1 *Instituto de Quı´mica, Universidade de Saõ Paulo, CP 26.077, 05513-970 Saõ Paulo, Brazil; and †Faculdade de Cieˆncias Farmaceûticas, Universidade de Saõ Paulo, CP 66.083, 05389-970 Saõ Paulo, Brazil

Received May 8, 2000, and in revised form October 31, 2000; published online February 16, 2001

The indole moeity is present in many substances of biological occurrence. Its metabolism, in most cases, involves an oxidative pathway. This study reports the oxidation of a series of indole derivatives, including several of biological origin, catalyzed by horseradish peroxidase in the presence of H 2O 2. Chemiluminescence emission was observed in most cases and its intensity and spectral characteristics were correlated with structural features of the substrates. The structures of the main products were determined. The participation of molecular oxygen and superoxide ion in the reaction was demonstrated and a general mechanism for product formation proposed. Since the oxidation of 2-methylindole proved to be highly chemiluminescent, its potentiality as a developing system for peroxidase-based assays was tested and showed to be very effective. © 2001 Academic Press Key Words: indole; chemiluminescence; horseradish peroxidase (HRP); superoxide anion.

The first reports identifying chemiluminescent properties of naturally occurring indoles emerged in the sixties. In 1965, Philbrook and coworkers (1) reported a study of the oxidation of 24 indoles by K 2S 2O 4/OH ⫺ in DMSO or aqueous media and showed that the chemiluminescence of 3-methylindole/DMSO is as intense as that of the luminol/H 2O 2 system. Two years later, Sugiyama and coworkers (2) showed that the autoxidation of 2,3-dimethylindole in alkaline DMSO developed chemiluminescence dependent on the formation of an intermediate hydroperoxide. The identification of N-acetyl-o-amine-acetophenone as the decomposition product led these authors to propose a dioxetane as the 1

To whom correspondence and reprint requests should be addressed. Fax: (⫹5511) 3815 5579. E-mail: [email protected]. 0003-9861/01 $35.00 Copyright © 2001 by Academic Press All rights of reproduction in any form reserved.

key intermediate responsible for the chemiexcitation step. Later, the same cleavage product was described in the reaction of naturally occurring indoles with dioxygenases (3, 4). Indole structures are present in a great number of compounds of biological importance, e.g., the plant growth hormone indoleacetic acid (IAA), 2 the pineal gland hormone melatonin, serotonin and tryptophan. The horseradish peroxidase (HRP) catalyzed aerobic oxidation of IAA and its ethyl ester, and the chemiluminescence generated therefrom, have been well studied (5, 6). Among the suggested possibilities, the presence of a tertiary radical at 3-position detected by EPR pointed to the cleavage of the pyrrole ring as the key step for chemiexcitation (7). Again, the authors proposed the cleavage of a dioxetane, leading to a compound of the N-formyl-kynurenine type in an excited state. On the other hand, the oxidative cleavage of the pyrrole ring, resulting in kynuric type products, is well established for tryptophan (8), melatonine (9), and related compounds. This reaction is enzymatically catalyzed by dioxygenases and non-enzymatically by O 2 in the presence of hemin (10). The heme-protein HRP is an oxidative system capable of acting on a wide range of possible substrates. With H 2O 2 as cosubstrate, intermediates like Compounds I, II, and III, which are formally higher oxidation states of the iron complex, are generated. Aldehydes and imines containing ␣-hydrogens are examples of substrates readily oxidized by this system, with the peculiar characteristic that excited carbonyls are produced, most probably via a high energy dioxetane 2 Abbreviations: AMP, 2-amino-2-methyl-1-propanol; HRP, horseradish peroxidase; IAA, indolacetic acid; 2-MI, 2-methylindole; SOD, superoxide dismutase.

173

174

XIMENES, CAMPA, AND CATALANI

intermediate (11, 12). In some cases, HRP may present dioxygenase-like activity (13). In this work, we explore the chemiluminescent features of the reaction of indole derivatives submitted to the HRP/H 2O 2/O 2 system. In particular, we present mechanistic aspects of the oxidation of 2-methylindol, which showed the highest light production efficiency compared to the other indole structures and led us to evaluate its use as substrate in chemiluminescent applications for detection of HRP.

MATERIALS AND METHODS Reagents. Catalase (EC 1.11.1.6; from bovine liver), superoxide dismutase (SOD; EC 1.15.1.1; from bovine erythrocytes), horseradish peroxidase (HRP; EC 1.11.1.7; type VI), 2-amino-2-methyl-1-propanol (AMP) and all indole derivatives were from Sigma, except for N-methylindole, 2,5-dimethylindole and 2,3-dimethylindole, which were from ACROS; all were used as supplied. Hydrogen peroxide 60%, from INTEROX, was diluted to the appropriate stock concentration. Instrumentation. To follow the time course of chemiluminescence we used either an EG&G Berthold LB96V Microplate Luminometer or a “photon-counting” system consisting of an EG&G PAR 1121A Amplifier-discriminator and a Thorn EMI 9658AM photomultiplier cooled to ⫺12°C by a Thorn EMI FACT-50 MKIII thermoelectric cooler. The latter system permits the use of cut-off filters in front of the PMT tube. Fluorescence and chemiluminescence spectra were obtained in a SPEX-FLUOROLOG 1681 fluorometer with a cooled photomultiplier. Oxygen consumption was monitored with a YSI-5300 Biological Oxygen Monitor. Spectrophotometric determinations were performed with a Hitachi U-2000 spectrophotometer and NMR analyses with a Brucker DPX-300 spectrometer. The reaction mixtures were analyzed by high performance liquid chromatography using a SHIMADZU LC-10A system coupled to SPD-10A UV-Vis and RF535 as fluorescence detectors. The analyses were carried out on a SUPERCOSIL LC-18 reversed phase column in isocratic mode (25 ⫻ 4.6 mm, 5 ␮m) using water/acetonitrile 1:2 as mobile phase at a constant flux of 1 ml/mim. The mass spectra were obtained employing a Hewlett Packard 5988 quadrupole mass spectrometer attached to a 5890 gas chromatograph using a HP1 (12 m ⫻ 0.25 mm ⫻ 0.25 ␮m) column. Methods. Unless otherwise stated, the standard reaction mixture was: [HRP] ⫽ 0.2 ␮mol/L, [H 2 O 2 ] ⫽ 5 mmol/L, [indole derivative] ⫽ 1 mmol/L, and 0.05 mol/L AMP as buffer at pH 9.3, at 37°C. Final volumes were 0.3 ml for the luminometer and 3 ml for the photon-counter and oxygen monitor. Usually, the reaction was initiated by addition of substrate (indole derivative) after a 2-min preincubation of all other components. Larger batch reactions (300 ml) were run using standard conditions. After solvent evaporation, the products were isolated by rotary preparative thin-layer chromatography (Chromatotron) using a 2-mm silica gel layer (Merck, Cat. 1.07749) and pure ethyl acetate as eluent. o-Acetamidoacetophenone was prepared by acylation of o-aminoacetophenone in 85% yield, according to literature (14). 1 HNMR (CDCl 3 , ␦ in ppm): 2.23 (s, 3H), 2.67 (s, 3H), 7.08 (ddd, 1H), 7.57 (ddd, 1H), 7.90 (dd, 1H), 8.74 (dd, 1H), 11.70 (b, 1H). MS (m/e): 177(23)P, 162(6), 144(2), 135(45), 134(22), 120(100)B, 92(21), 77(6).

FIG. 1. Light emission profile of selected indole derivatives. Standard reaction conditions using [indole derivative] ⫽ 1 mM. (A) 2-methylindole; (B) 2,5-dimethylindole; (C) 3-methylindole; (D) 2,3dimethylindole.

RESULTS

Chemiluminescence Efficiency within a Series of Indole Derivatives The main criterion employed in choosing the indole derivatives to be tested was their potential for elucidating the structural features responsible for the observed chemiluminescence. In addition several biological indoles were also included. Figure 1 shows representative examples of light emission profiles for derivatives 3, 4, 6, and 7, while Table I lists the integrated light intensities for the series of indoles studied in the present work. The first aspect evident in Table I is that most of the compounds tested exhibit some chemiluminescence, demonstrating that the HRP/H 2O 2 system is indeed capable of oxidizing indole structures, leading to the formation of emissive excited states. In all cases, control experiments without addition of either enzyme or H 2O 2 showed no emission. Moreover, inactivation of the enzyme by heating at 80°C for 3 h also eliminated chemiluminescence, confirming the enzymatic catalysis of the process. Second, there is a pronounced substitution effect on the chemiluminescence intensity, the optimal case being alkyl substitution at position 2 and hydrogen at position 3. Thus, the chemiluminescence of 2-methylindole (2-MI), 3, and 2,5-dimethylindole, 7, is two to three orders of magnitude higher than that of indole, 1. Electronic effects do not seem be important since neither electron-releasing groups (e.g., compare 2-phenylindole, 8, and methoxyindoles 9 –12 to indole, 1) nor electron-withdrawing groups (e.g., compare 6-chloromelatonin, 17, and 2-iodomelatonin, 18, to melatonin, 16) produced effects. On the other hand, a hydrogen at the 3-position seems to be critical: 3-methylin-

ENZYMATIC CHEMILUMINESCENT OXIDATION OF INDOLES TABLE I

Comparison of Light Emission Intensities for the Series of Indole Derivatives

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 a

Indole derivative

Integrated emission a over 20 min (⫻10 5 counts)

Indole N-Methylindole 2-Methylindole 3-Methylindole 5-Methylindole 2,3-Dimethylindole 2,5-Dimethylindole 2-Phenylindole 4-Methoxyindole 5-Methoxyindole 6-Methoxyindole 7-Methoxyindole 3-Indole acetic acid 5-Methoxy-3-indole acetic acid 2-Methyl-5-methoxy-3-indole acetic acid Melatonin 6-Chloromelatonin 2-Iodomelatonin 6-Hydroxymelatonin 5-Hydroxytriptamine (serotonin) 5-Methoxytriptamine Tryptophan

4.9 1.5 8194 25.1 2.3 11.1 1082 0.2 4.7 0.4 4.8 1.0 0.5 1.5 1.0 6.9 5.2 3.3 0 0 0 0

Standard conditions using [indole derivative] ⫽ 1 mmol/L.

dole, 4, 3-indoleacetic acid, 13, and 2,3-dimethylindole, 6, presented low emission, close to indole, 1. Product Characterization The products formed from compounds 1, 3, 4, 6, 7, and 16 were analyzed by HPLC using UV-Vis and fluorescence detectors. In all cases, one major product with fluorescence emission in the 400 – 460 nm spectral region and similar elution volume was observed. For compounds 3 and 6, a 300-ml batch reaction was run, concentrated by evaporation, and the products isolated by prep-TLC (see Materials and Methods). Their characterization by CG-MS revealed the formation of oacetamidobenzaldehyde, 23, and o-acetamidoacetophenone, 24, two kynuric-type products with structures of the type expected from the decomposition of an intermediate dioxetane (MS of 23, m/e 163 (13)P, 135(14), 121(12), 93(100)B, 66(15); MS of 24, m/e 177(18)P, 162(5), 135(40), 134(18), 120(100)B, 92(19), 77(4)). The characterization of 24 was further confirmed by coinjection of an authentic sample prepared by conventional methodology (see Materials and Methods). The similarity of the products of the indoles analyzed with respect to (i) elution volume, (ii) absorption spectra, and (iii) fluorescence spectra points to indole-ring cleavage as the common pathway for this oxidation (Scheme 1).

175

The product analysis of 2-methylindole, 3, and 2,5dimethylindole, 7, by HPLC revealed, only for these two compounds, a second major product with common features, including retention time and absorption and fluorescence spectra. A preparative scale oxidation of 3 was run and product isolation and analysis led to the identification of 2,2⬘-dimethyl-2,2⬘-diindoxyl, 25 ( 1HNMR 300 MHz, CDCl 3 (␦ in ppm): 1.56 (s, 6H), 6.80 – 7.62 (m, 8H), 4.92 (broad s, 2H); MS: m/e 290(53), 261(16), 235(16), 185(14), 162(35), 146(100)B, 117(33), 77(26)). According to the literature (15), diindoxyl 25 is the expected product from the dimerization of 2-methyl-3-indolinone, a species never isolated due its instability. Chemiluminescence versus Product Fluorescence Spectra Dioxetane thermolysis is known to produce excited states (16). Assuming that a dioxetane is involved in the oxidation of 2-methylindole, 3, and 2,3-dimethylindole, 6, their thermolysis would result in electronically excited o-acetamidobenzaldehyde, 23, or o-acetamidoacetophenone, 24, respectively. Surprisingly, the chemiluminescence spectrum observed for the oxidation of 2-methylindole, 3, does not match the fluorescence spectra of the kynuric product 23, but instead that of diindoxyl 25. Figure 2 compares the chemiluminescence spectrum of 3 against the fluorescence spectra of these two products. Similar results (not shown) were observed for the oxidation of 2,5-dimethylindole, 7. Considering that the formation of excited diindoxyl 25 as a primary product cannot be readily rationalized on the basis of known chemiexcitation steps, its role simply as an enhancer of the emission of another primary excited species (i.e., as an excitation energy acceptor) was investigated. For this purpose, a semiprep HPLC fraction containing pure diindoxyl was added to the oxidation reactions of various indole derivatives.

SCHEME 1.

Products of 2-methylindole oxidation.

176

XIMENES, CAMPA, AND CATALANI TABLE II

pH Effect on Light Emission, Oxygen Consumption, and Product Formation on 2-Methylindole Oxidation

pH

Integrated emission over 20 min (rel. units)

Rate of O 2 uptake (rel. units)

Fluorescence ratio a I 25 /I 23

4.8 7.4 9.3 11.6

1 3 1214 1428

471 64 17 1

1 1.1 2.8 4.3

a

Relative intensity of HPLC peak of compounds 23 and 25, as analyzed by the fluorescence detector. FIG. 2. Chemiluminescence of 2-methylindole oxidation versus fluorescence spectra of products. Fluorescence spectra were obtained from the fluorescence detector by halting the elution of (A) compound 23 (␭ exc ⫽ 320 nm) and (B) compound 25 (␭ exc ⫽ 400 nm) during HPLC analysis. Chemiluminescence spectrum (C) of oxidation of 3 was obtained on a SPEX fluorometer.

For indole, 1, 5-methylindole, 5, 5-methoxyindole, 10, the addition of 20 ␮l of this solution resulted in an emission enhancement of one order of magnitude, as shown for 1 in Fig. 3. On the other hand, the same addition failed to cause changes in the intensity of 3-methylindole, 2,3-dimethylindole and melatonin and, for 2-methylindole itself, only a 10% increase was observed. The Oxidation of 2-Methylindole: A Case Study 2-Methylindole, 3, deserves special attention since its oxidation produces 2–3 orders of magnitude more light than most of the other indole derivatives studied. The following phenomena were observed for the reaction of 3/HRP/H 2O 2:

FIG. 3. Effect of diindoxyl addition on the oxidation of indole. Standard reaction conditions without (A) and with (B) addition of diindoxyl 25 HPLC fraction.

pH effect. The data for light emission efficiency and the rate of oxygen consumption for this system at different pHs, shown in Table II, indicate that basic medium favors light emission and slows down O 2 uptake. In addition, the yield of diindoxyl 25 increases relative to that of the kynuric product 23. Formation vs production of H 2O 2, O 2 and O 2⫺•. Because the stability of the superoxide anion is greatly augmented in basic media, its intermediacy was verified by adding superoxide dismutase (SOD). This enzyme is responsible for the in vivo control of O 2⫺• levels via catalysis of its dismutation into O 2 and H 2O 2. Indeed, addition of SOD to the standard reaction mixture resulted in suppression of the emission at pH 9.3 (Fig. 4, right). Conversely, addition of 5 mM KO 2 to the standard mixture at pH 7.4 increased the emission 13-fold (data not shown). In both cases, the emission temporal profile remained unchanged. Superoxide anion also exerts an effect on the product distribution, favoring diindoxyl production. The addition of 30

FIG. 4. Effect of addition of SOD and catalase on 2-methylindole oxidation. Figure at right: 1 ␮mol/L HRP, 5 mmol/L H 2O 2, 0.1 mmol/L 2-MI, 0.05 mol/L AMP as buffer at pH 9.3, and (A) control, (B) 0.05, (C) 0.5, (D) 1.2, (E) 2.5, and (F) 5.1 SOD units/ml. Figure at left: 0.1 ␮mol/L HRP, 5 mmol/L H 2O 2, 1 mmol/L 2-MI, 0.05 mol/L AMP as buffer at pH 9.3, and (A) control, (B) 12, (C) 24, (D) 36, (E) 72, and (F) 108 Catalase units/ml.

ENZYMATIC CHEMILUMINESCENT OXIDATION OF INDOLES

177

blotting techniques. The potential of employing this simple, inexpensive, off-the-shelf substrate in chemiluminescence assays where HRP activity is the ultimate signal was evaluated. Figure 6 shows that the integrated light emission correlates linearly with total HRP added. Hence, the detection limit for determination of free HRP by this method should be of the order of tens of fentomols. If a HRP-conjugate is employed, the same panorama is observed, with a detection limit on the order of 0.01 units/ml. DISCUSSION FIG. 5. Molecular oxygen concentration during 2-methylindole oxidation. Reaction conditions: 1 ␮mol/L HRP, 5 mmol/L H 2O 2, 0.05 mol/L AMP as buffer at pH 9.3, and without (A) and with addition of 0.1 mmol/L (B) or 1 mmol/L (C) 2-methylindole.

units/ml of SOD (pH 9.3) decreases the ratio of HPLC fluorescence signal of 25/23 to 60% of its original value, while addition of 5 mM of KO 2 (pH 7.4) increases by a factor of 100%. It is known that, at high H 2O 2 concentrations, many peroxidases may act as catalases, dismuting this substrate into H 2O and O 2 (17, 18). In the case of HRP, a “suicide” pathway is activated in the absence of reductant. HPR Compound I, a very powerful oxidant, reacts with H 2O 2 again, giving the native enzyme and O 2 or decomposing to an inactive form. In the presence of a reductant, the normal cycle, going through Compound II, is preferred (19). Molecular oxygen consumption is expected judging by similar oxygenase and dioxygenease systems that produces kynuric products. Indeed, oxygen consumption is very fast at acid pHs (see Table II). Addition of small amounts of catalase leads to an increase of light emission by enriching the medium with O 2 (Fig. 4, left). However, higher catalase contents led to a decrease of this emission, confirming the need for H 2O 2 as the primary oxidant in the HRP cycle. As supporting evidence for O 2 consumption, light emission is almost completely suppressed by pre-purging the mixture with N 2 and is tripled by bubbling with pure O 2 (data not shown). Under our standard conditions, the H 2O 2 concentration is relatively high, hence consumption of oxygen was expected to be competitive with its production via HRP-catalytic activity. Figure 5 shows the relative O 2 concentration in the presence and absence of reductant, confirming this hypothesis. These combined activities of HRP may, in part, be responsible for the high quantum yields observed. Potential of the 2-MI/HRP/H 2O 2 System Horseradish peroxidase is one of the most popular labels used in immunoassays, including avidin– biotin-

As already mentioned, it is well established that chemical and biological oxidation of several indole derivatives leads to kynurenine-like products, most probably through a 2,3-dioxetane-type intermediate. Indeed, the formation of electronically excited products has been reported in some cases. The main goal of this account is an understanding of the primary features of HRP/H 2O 2 oxidation of indole derivatives for the purpose of (i) obtaining a comparative framework for the study of biologically related systems and (ii) exploring its potential as a chemiluminescent reaction for development of HRP-based assays. Heme-based proteins are responsible for several oxidative reactions in animal and plants. The most fully understood are: HRP, myeloperoxidase, cytochrome c oxidase, cytochrome P-450, and catalase. The most prevalent aspect among these is their basic redox cycle: a two electron reduction step of H 2O 2 (or other peroxide), followed by two single electron oxidation steps of specific substrates. However, other reactions may also be possible (or even prominent), reaching a different electronic configuration or oxidation state of the iron

FIG. 6. Light emission kinetics from HRP/H 2O 2/2-methylindole reaction. Figure at left: 5 mmol/L H 2O 2, 1 mmol/L 2-MI, 0.05 mol/L AMP as buffer at pH 9.3, and (A) 30, (B) 3, (C) 0.3, (D) 0.03, and (E) 0 picomols of HRP per assay. Figure at left: HRP concentration dependence of the integrated light emission.

178

XIMENES, CAMPA, AND CATALANI

SCHEME 2. Proposed mechanism for 2-methylindole product formation.

porphyrin center. Generally, the medium and oxidant concentration will define the operative cycle (20). The basic HRP cycle involves the components (20): native HRP (ferric) 3 Compound I 3 Compound II. Excess H 2O 2 generates Compound III, which may be regarded as a complex of native HRP and superoxide ion. It is formed via a biomolecular reaction of native HRP or HRP-II with H 2O 2. Ferrous HRP and molecular oxygen also produces Compound III, though this pathway requires the presence of a strong reductant to generate the former and is, therefore, less important. Singlet molecular oxygen could also be formed, but in other peroxidases halide ion is generally required to form hypohalous acid as an intermediate (21, 22). A conspicuous pathway for 2-methylindole oxidation is its reaction with HRP-I or HRP-II producing an indolyl radical. The fate of this radical is the determinant of the product type, including chemiluminescence. Our results establish unequivocally the role of superoxide ion and molecular oxygen, since both may add to the indolyl radical, producing peroxide or peroxide radical (Scheme 2). The kynuric product does not seem to be specific of either oxygen species. On the contrary, addition of O 2⫺• does lead to an increase of diindoxyl product, while SOD quenches it. The simplest conclusion would therefore be that reaction of indoxyl radical with O 2 produces kynuric products, while reaction with O 2⫺• produces diindoxyl. The expected thermal pathways initiated by these two possibilities are somewhat clearer for the addition of superoxide ion. In Scheme 2, the hydroperoxide thus formed may cyclize to a dioxetane intermediate by an intramo-

lecular nucleophilic ring closure. The decomposition of this dioxetane leads to the kynuric product. Alternatively, the hydroperoxide may dehydrate to the indolynone, a reaction known to occur for secondary hydroperoxides (23). Very little is known about indolynone, but literature attempts to synthesize it always report the production of dimers, including diindoxyl (15, 24). Its reduction to a indolynone radical, followed by dimerization would be a plausible step. Here, superoxide ion would be a likely candidate for this reduction, with production of O 2. The addition of molecular oxygen leads to an alkylperoxyl radical; therefore, the cyclization step leading to the kynuric product is the exclusive possibility. Timmins et al. (25) report conclusive reasoning for a very similar case, in which the cyclization of an alkylperoxyl radical to a dioxetane radical intermediate was detected by EPR experiments. Likewise, chemiluminescence was observed as a product of dioxetane cleavage. The observed indole structural effects on the emission are: (i) hydroxyl substitution on the phenyl ring and free amino substitution on the side chain lead to non-chemiluminescent oxidation; (ii) kynuric products were observed in most cases; (iii) the 2-alkyl-3-hydrogen-substitution pattern leads to an additional diindoxyl product and the reaction is highly chemiluminescent, while different 2,3-substitution patterns show low levels of chemiluminescence. The fact that compounds 19 –22 totally lack emission has a simple rationale: while phenols are known to react with HRP oxidative systems, producing quinones and semiquinones in a dark pathway (26), tryptophan

ENZYMATIC CHEMILUMINESCENT OXIDATION OF INDOLES

and serotonin have a free amino group in their structure. Since amines are known quenchers of excited states, an intramolecular process would suffice to quench all emission. The spectral distributions indicate that, in the absence of diindoxyl, electronically excited kynuric product, formed during the thermolysis of a dioxetane intermediate, is the primary emitter. In the cases where diindoxyl is formed, its likely function is as an “enhancer” of the emission. This was clearly shown by addition of diindoxyl during the oxidation of other indole derivatives. Its chemiexcitation presumably occurs by energy transfer from the primary electronic excited state. Unfortunately, all attempts to prepare this new sensitizer failed and, hence, this must be regarded as a hypothesis only. Concluding Remarks Peroxidases are known to be involved in the oxidative metabolism of several indole derivatives. In this study, we clearly show that, with horseradish peroxidase, electronically excited states are involved as a product. Molecular oxygen and superoxide ion are definitely involved in such reactions. These findings point to a general mechanism that should be considered in all oxidative metabolic reactions of indole compounds of biological occurrence. Hence, like reactive free radicals, electronically excited states should be regarded as ordinary participants in biological reactions within the context of “photobiochemistry in the dark” (27). Finally, the HRP catalyzed oxidation of 2-methylindole by H 2O 2 exhibited an especially intense chemiluminescence. Since HRP is widely used as a label in immunoassays in general, the effectiveness of this ready-to-use substrate as a chemiluminescent detector of HRP was evaluated. ACKNOWLEDGMENTS The authors are indebted to the Fundaçaõ de Amparo a Pesquisa do Estado de Saõ Paulo, FAPESP, Saõ Paulo, Brazil, the Conselho Nacional de Desenvolvimento Cientı´fico e Tecnolo´gico, CNPq, Brası´lia, Brazil, for grant support, and Professor Frank Quina (Universidade de Saõ Paulo), for reviewing this manuscript.

REFERENCES 1. Philbrook, G. E., Ayers, J. B., Garst, J. F., and Totter, J. R. (1965) Photocem. Photobiol. 4, 869 – 876. 2. Sugiama, N., Akutagama, M., and Yamamoto, H. (1968) Bull. Chem. Soc. Japan 41, 937–941.

179

3. Hayaishi, O., Rothberg, S., Mehler, A. H., and Saito, Y. (1957) J. Biol. Chem. 299, 889 – 896. 4. Sono, M., Taniguchi, T., Watanable, Y., and Hayaishi, O. (1980) J. Biol. Chem. 255, 1339 –1345. 5. Metodiewa, D., Melo, M. P., Escobar, J. A., Cilento, G., and Dunford, H. B. (1992) Arch. Biochem. Biophys. 296, 27–33. 6. Melo, M. P., Escobar, J. A., Metodiewa, D., Dunford, H. B., and Cilento, G. (1992) Arch. Biochem. Biophys. 296, 34 –39. 7. Escobar, J. A., Vasquez-Vivar, J., and Cilento, G. (1992) Photochem. Photobiol. 55, 895–902. 8. Hirata, F., and Hayaishi, O. (1975) J. Biol. Chem. 10, 5960 – 5966. 9. Hardeland, R., Balzer, I., Poeggeler, B., Fuhrberg, B., Urıá, H., Behrmann, G., Wolf, R., Meyer, T. J., and Reiter, R. J. (1995) J. Pineal Res. 18, 104 –111. 10. Uemura, T., and Kadota, K. (1984) in Progress in Tryptophan and Seronin Research (Schlossburger, H. G., Kochen, W., Linzen, B., and Steinhart, H., Eds.), pp. 673– 676, de Gruyter, Berlin. 11. Adam, W., Baader, W. J., and Cilento, G. (1986) Biochim. Biophys. Acta 881, 330 –336. 12. Medeiros, M. H. G., and Bechara, E. J. H. (1984) in Oxygen Radicals in Chemistry and Biology (Bors, W., Saran, M., and Tait, D., Eds.), pp. 539 –542, de Gruyter, Berlin. 13. Kobayashi, K., Hayashi, K., and Sono, M. (1989) J. Biol. Chem. 264, 15280 –15283. 14. Marvel, C. S., and Lazier, W. A. (1941) Org. Synth. 1, 99 –101 (1941). 15. Hassner, A., and Haddadin, M. J. (1963) J. Org. Chem. 28, 224 –225. 16. Adam, W., and Cilento, G. (1983) Angew. Chem. Int. Ed. Engl. 22, 529 –542. 17. Iwamoto, H., Kobayashi, T., Hasegawa, E., and Morita, Y. (1987) J. Biochem. 101, 1407–1412. 18. Sun, W., and Dunford, H. B. (1994) Biochem. Cell Biol. 72, 321–331. 19. Arnao, M. B., Acosta, M., Del Rio, J. A., Varon, R., and GarciaCańovas, F. (1990) Biochim. Biophys. Acta 1041, 43– 47. 20. Dunford, H. B. (1991) in Peroxidases in Chemistry and Biology (Everse, J., Everse, K. E., and Grisham, M. B., Eds.), Vol. II, pp. 1–24, CRC Press, Boca Raton. 21. Kanofsky, J. R. (1983) J. Biol. Chem. 258, 5991–5993. 22. Khan, A. U. (1983) J. Am. Chem. Soc. 105, 7195–7197. 23. Kornblum, N., and DeLaMare, H. E. (1951) J. Am. Chem. Soc. 73, 880 – 881. 24. Piozzi, F., and Langella, M. R. (1963) Gazz. Chim. Ital. 93, 1373–1381. 25. Timmins, G. S., Santos, R. E., Whitwood, A. C., Catalani, L. H., Di Mascio, P., Gilbert, B. C., and Bechara, E. J. H. (1997) Chem. Res. Toxicol. 10, 1090 –1096. 26. Job, D., and Dunford, H. B. (1976) Eur. J. Biochem. 66, 607– 614. 27. Cilento, G., and Adam, W. (1994) Free Rad. Biol. Med. 19, 103– 114.