Synthesis and characterization of 1-substituted 5-alkylphenazine derivatives carrying functional groups

Eur. J. Biochem. 179, 293 - 298 (1989) c) FEBS 1989 Synthesis and characterization of 1-substituted 5-alkylphenazine derivatives carrying functional groups Tetsuya YOMO, Haruyo SAWAI, Itaru URABE, Ydsuhiro YAMADA and Hirosuke OKADA Department of Fermentation Technology, Faculty of Engineering, Osaka University (Received March 28/August 23, 1988) - EJB 88 0362 The following 1-substituted derivatives of Smethylphenazine and 5-ethylphenazine were synthesized: 1-(3- carboxypropyloxy)-5-methylphenazine (lB), l-(3-carboxypropyloxy)-5-ethylphenazine (2B), 1-(3-ethoxycarbo- nylpropyloxy)-5-ethylphenazine (2C) and l-[N-(2-aminoethyl)carbamoylpropyloxy]-5-~thylphenazine (2D); their spectra, stability and reactivity as electron mediators were investigated, together with those of 5-methylphenazine (1A) and 5-ethylphenazine (2A). The 1-substituted derivatives are all insensitive to light and the derivatives of 5- ethylphenazine are more stable than those of 5-methylphenazine under neutral and alkaline conditions; 2 8 is the most stable of all the derivatives. The spectral properties of the decomposed compounds showed that photodecomposition of 1A and 2A is associated with hydroxylation at position 1, alkali decomposition of 1A and 1 B with elimination of the 5-methyl group and alkali decomposition of 2A, 2B, and 2D with a ring-opening reaction. The second-order rate constant kl for the reaction of the phenazine derivatives with NADH was measured under steady-state conditions. The k l values vary depending on the substituents at positions 1 and 5 : the values for lA, lB, 2A, 2B, 2C and 2D are 1.83 mM- ' s- ', 3.33 mM-' s-', 0.75 mM-' s-' , 1.42 mM-' s- ' , 1.68 mM-' s - ' and 2.03 mM-' s - I , respectively. The rate constants, k, and k3, for the reactions of the reduced form of 2B with oxygen and with 3-(4,5'-dimethyl~hiazole-2-yl)-2,5-diphenyltetrazolium ion, respectively, were k2 = 1.21 mM-' sC1 and k3 = 91 mM-' s-' . These phenazine derivatives have potential applications in the biochemical field. 5-Methylphenazine (Fig. 1, 1A) is a unique electron me- diator as it can accept electrons from NAD(P)H and is widely used in various redox reactions. However, its application is still limited, mainly because of its instability to light and to alkaline solution. 5-Ethylphenazine (Fig. 1, 2A) has been recommended as a more stable electron mediator in alkaline solutions [l, 21 and light-insensitive derivatives have also been prepared and used in the assay of dehydrogenases [3 - 61, but the mechanisms of these instabilities are still unclear. We have studied co-enzyme-recycling systems [7 - 91 and now we intend to use phenazine derivatives as NAD(P)+ regenerators. For this purpose, it is necessary to prepare stable phenazine derivatives with a functional group suitable for binding them to other molecules, such as water-soluble poly- mers, solid supports and enzymes. These derivatives are also useful for investigating the mechanisms of the electron-trans- fer reactions from NAD(P)H. Here, we describe the prep- aration of new phenazine derivatives carrying a carboxyl group (Fig. I , 1 B and 2B) and an amino group (Fig. 1, 2D) and also their stability and electron-transfer activity. The ki- netic constants obtained in this work are necessary for under- standing the coupled-reaction system of NAD(P)+ regener- ation. Correspondence to 1. Urabe, Department of Fermentation Tech- nology, Faculty of Engineering, Osaka Univcrsity, 2-1 Yamada-oka, Suita-shi, Osaka-ru, Japan 565 Abbreviation. MTT, 3-(4',5'-dimclhylth1azolc-2-yl)-2,5-diphenyl- tetrazolium bromide. Enzjrne. Pcroxidasc (EC 1 . 1 1.1.7). EXPERIMENTAL PROCEDURES Materials Ethyl 4-bromobutyrate, 18-crown-6 and 2,2'-azino-bis(3- ethylbenzothiazoline-6-sulfonate) were purchased from Wako Pure Chemical Industries Ltd (Osaka); 5-methylphenazinium methyl sulfate, 5-ethylphenazinium ethyl sulfate and 1 -ethyl- 3-(3-dimethylaminopropyl)carbodiimide . HC1 were from Nakarai Chemicals (Kyoto); NADH was from Oriental Yeast Co. Ltd (Tokyo); MTT was from Dojindo Laboratories (Kumamoto); horseradish peroxidase (grade 111) was from Toyobo Co. Ltd (Osaka); thin-layer-chromatography plates of silica gel 60 F254 were from Merck (Darmstadt); reversed- phase thin-layer-chromatography plates (MKC18F) were from W hatman, Analytical procedures Thin-layer chromatography was performed on silica gel in CH2CI2/CH30H (9:1, system A) and on MKC18F in CH30H/H20 (2:1, system B) or CH3CH/0.1 M NaCl (3:7, system C). Phenazine derivatives were detected using ultra- violet light and aliphatic amines by ninhydrin spray. Absorp- tion spectra were obtained with a Hitachi 220A spec- trophotometer. 'H-NMR spectra were obtaincd with a Jeol SP-100 NMR spectrometer operating at 100 MHz or with a Jeol GX-400 NMR spectrometer at 400 MHz. Nitrogen was measured by the method of Jacobs [lo]. Amino groups were measured using 2,4,6-trinitrobenzene sulfonic acid, with a millimolar absorption coefficient of 20 mM-' cm-' a t 420 nm for a trinitrophenylated amino group [ll]. Hydrogen 294 R P CH, 1 3 A: R = H 8: R = O(CH2)3COeH C: R = O(CHe)3C02CHeCH3 D: P = O(CHr)aCONH(CHe)eNHr E: R=OH F: R = 0CI-h Fig. 1. Plienmzino derivarivi~s peroxide was measured by the method of Putter and Becker [12]. The concentrations of NADH and phenazine derivatives were measured spectrophotometrically using the following millimolar absorption coefficients: NADH, 6.3 mM-' cm-' at 340 nm; derivatives 1A and lB, 26.3 mM-' cm-' at 386 nm [13]; derivatives 2A, 2B, 2C and 2D, 17 mM-' cm-' at 386 nm. Preparation of 1 - (3-ethoxycarhonylpropy1oxy)phenazine (3C) 1-Hydroxyphenazine (3E) was prepared by the method of Surrey [14], except that lead tetraacetate was used instead of lead dioxide, and the yield was about 30%. Thin-layer chromatography showed that the RF values of 3E are 0.81 in system A and 0.44 in system B. 3E (0.43 g) K2C03 (1.5 g), 18-crown-6 (0.5 g) and ethyl 4-bromobutyrate (2.1 g) were dissolved in acetone (50 ml) and the solution was refluxed for 1 h; the mixture was evaporated under reduced pressure. The residue was dissolved in benzene, extracted with water and the organic layer was dried over anhydrous Na2S04 and evaporated. The residue was put on a silica gel column (CH2C12) and 3C was obtained by elution with CH2C12 con- taining 1 % methanol. The yield from 3E was 90%. Its purity and structure werc confirmed by thin-layer chromatography, elemental analysis and 'H-NMR spectroscopy (in CDC13 at 100 MHz). RF valucs: 0.84 in system A and 0.29 in system B. Found: C, 69.32; H, 5.79; N, 8.95; CI8Hl8O3N2 requires C, 69.66; H, 5.85; N, 9.03. 'H-NMR spectrum showed the following characteristic signals for the 3-ethoxycarbonyl- propyloxy group in addition to thc aromatic proton signals (total 7H) at similar positions to those reported by Romer for 1-methoxyphenazine [15]: 6 = 1.24 ppm (3H, t, CH2 -CH,), 2.26 ppm (2H, m, CHI - CH, -CHI), 2.66 ppm (2H, t, CH2-CO), 4.16ppm (2H, q, COO-CH2) and 4.39ppm (2H, t, 0 - CH,). Preparation of' I - (3 -carho .~~~ropyloxy) -5-methylphenazine ( I B) 3C (0.1 g) was dissolved in dimethyl sulfate (0.3 ml) and the solution was heated at 300°C for 30min. The reaction mixture was cooled and l-(3-ethoxycarbonylpropyloxy)-5- methylphenazine (1C) was precipitated with diethyl ether and washed as described by Surry [14]. The residue was dissolved in CH2C12 and evaporated. The RF value of 1C in system C is 0.19. The crude product was hydrolyzed in 0.1 M HCl at 37°C for 3 days and then the pH of the solution was adjusted to 6.7. The derivative 1B thus produced was purified by pas- sage of the reaction mixture (after dilution with water to reduce the conductivity) through a DEAE-Sephadex A-25 column equilibrated with water (yield 63%). The RF value of 1B in system C is 0.60. Preparation of I - (3-carhoxypropyloxy) -5-ethyi)henazine (2B) 3C (1.0 g) was dissolved in diethyl sulfate (4.2 ml) and the solution was heated at 120°C for 3 h. 1-(3-Ethoxycarbo- nylpropyloxy)-5-ethylphenazine (2C) was obtained from the reaction mixture by the procedures described above. When 2C was used for kinetic studies, it was purified by chromatography on an Amberlite XAD-4 column in meth- anol. The RF value of 2C in system C is 0.17. 2C was hydrolyzed in 0.1 M HCl at 37°C for 6 days and 2B thus produced was purified by passage through a DEAE-Sephadex A-25 column as described above (yield 59%). Thc RF value of 2B in system C is 0.50. The 'H-NMR spectrum (400 MHz) of 2B shows the following characteristic proton signals: 6 = 1.77 ppm (3H, t, N-CH2-CFJ3), 2.30 ppm (2H, m, CH2- CH2-CHI), 2.61 ppm (2H, t, CH2-COOH), 4.47 ppm (2H, t, 0- CH2), 5.41 ppm (2H, q, N -CH2), 7.50 ppm (d, 1 H), 8.07 ppm (d, lH), 8.19 ppm (lH, t), 8.37 ppm (lH, t), 8.43 ppm (lH, t), 8.51 ppm (lH, d) and 8.58 ppm ( lH, d). Preparation of I-"- (2-aminoethyl) carbumoylpropyloxy J-5- ethylphenazine (2D) 2B was coupled to 1,2-diaminoethane at 4 "C in a reaction mixture (7 ml, pH was kept at 5.3) containing 6.1 mM 2B, 67 mM 1,2-diaminoethane, and 15 mM l-ethyl-3-(3-dimethyI- aminopropy1)carbodiimide . HC1. After 6 h the same amount of carbodiimide was added and the reaction was continued overnight. 2D was purified by chromatography on an SP- Sephadex C-25 column (4 x 35 cm, equilibrated with 10 mM sodium acetate buffer, pH 5.4) with a linear gradient of 0- 1 M NaCl (total 2 1). The molar ratio of the amino group and 2D in the fractions was 1 .O - 1 .l. Assay of the electron-transjer activity of phenasine derivatives Initial rates of reactions were measured at 30'C in 0.1 M sodium phosphate buffer, pH 7.5, using a Hitachi 220A spectrophotometer with a thermostatted cell compartment and a magnetic stirrer. The other components of the reaction mixtures are given in the legends to Table 1 and Figs 6 and 7. The reactions were recorded as the decrease in absorbance at 340 nm (NADH concentration) or the increase in absorbance at 570 nm (formazan concentration) and the reaction rates were calculated from the initial linear parts, after a short lag time. Blank tests were carried out without addition of NADH or phenazine derivatives. Assays were performed at least in triplicate with a reproducibility of less than 5% error. The concentration of the formazan produced from the reduction of MTT was calculated using a millimolar absorption coef- ficient of 13 mM-' cm-' at 570 nm [16]. RESULTS Spectral properties The structures of the phenazine derivatives are shown in Fig. 1. The derivatives lB, 2B, 2C and 2D have substituents at positions 1 and 5 and are all wine-colored. They have an identical absorption spectrum in water (pH 7) with absorp- tion maxima at 278nm (c = 47mM-I cm-'), 386nm ( E = 17 mM-' cm-') and 510 nm (Fig. 2); the millimolar absorption coefficients were estimated for derivatives 1 B and 2B on the basis of the amount of nitrogen they contained. 29 5 T " " " " ' m1 Wave Length ( n m 1 Fig. 2. Absorption spectru qf 2B ut p H 7 ( -), 3C at p H 7 (- . -) und3E ot p H I2 (----) These spectral properties are almost the same as those of 1- methoxy-5-methylphenazine (1 F) [3] and pyocyanine (lE, at pH 2) [13]. 1A and 2A, which have substituents only at pos- ition 5 , have an identical spectrum in water, with absorption maximum at 386 nm as reported by Ghosh and Quayle 121. 3C and 3E, which have substituents only at position 1, have an identical spectrum in water (pH 7) with an absorption maximum at 367 nm ( E = I2 mM-' cm-', Fig. 2). The spectra of 3C and 3E vary with pH: at pH 1,3C and 3E have an identical spectrum (A,,, = 273 nm, 382 nm and 490 nm), which is very similar to that of 2B; at pH 12, the spectrum of 3C is the same as that at pH 7, but that of 3E is different (A,,, = 373 nm, Fig. 2). These results show that the spectra of these phenazine derivatives are grouped into five patterns (see Discussion) corresponding to the following structural characteristics: the cationic charge at position 5, substitution at position 1, ionization of the hydroxyl group at position 1 and combinations of these. Stability The 1 -substituted derivatives we prepared are all insensi- tive to light, in contrast to 1A and 2A. For example, the electron-transfer activity and the spectra of 1 8 and 2B did not change when stored for one month at room temperature under scattered light in 0.2 M sodium acetate, pH 4.0. I-Methoxy- 5-methylphenazine is also reported to be photochemically stable 13, 41. The I-substitution seems to add photostability to 5-alkylphenazines. The effects of pH on the stability of the phenazine deriva- tives were investigated and the results for lB, 2B and 2D are shown in Fig. 3. The results for 1A and 2A (in the dark) are the same as those for 1B and 2B, respectively. The stability was monitored by measuring both the absorbance at 386 nm and the electron-transfer activity and both methods gave the same results. The 5-methylphenazine derivatives (1A and 1 B) are unstable above pH 5, while 5-ethylphenazine derivatives (2A, 2B and 2D) are much more stable in neutral and alkaline solutions: 2A and 2B are stable up to pH 9 and 2D up to pH 8. In addition, 2B was stable for one month at pH 9 under scattered light. Therefore, 2B is the most stable derivative, both under light and alkaline conditions. Spectral unalysis of decomposed compounds The spectral properties of the solution containing the alkali(pH 9)-decomposed compound of 1B are the same as those of 3C. These results suggest that the decomposition of 1B is due to the elimination of the methyl group at position 5 and the product is l-(3-carboxypropyloxy)phenazine. The alkali(pH 12)-decomposed compounds of 2B and 2D show similar spectral properties, but the properties are quite differ- ent from those of 3C: for the decomposed compound of 2B, I,,, at pH 12 is 287 nm and Amax at pH 1 is 275 nm; for that of 2D, A,,, at pH 12 is 292 nm and A,,, at pH 1 is 279 nm. Therefore, it seems that the ethyl group at position 5 of 2B is not eliminated, but the ring is opened, probably by a mecha- nism similar to that described by Johnson and Morrison for 1-substituted pyridinium cations [17, 181. The spectral properties of the solutions containing the photodecomposed (at pH 4) compounds of 1A and 2A are the same as those of pyocyanine (lE), reported by Zaugg [13], and 1E is reported to be produced under light from 1A [19- 221. Therefore, the photo-decomposed compounds of I A and 2A should be 1E and 2E, respectively. It is noteworthy that 1E and 2E are more stable in alkaline conditions than 1B and 2B, respectively. For example, 1E is stable at pH 9 and is decomposed at pH 12 into a compound whose spectral prop- erties are the same as those of 3E and 2E is stable even at pH 12. Steady-state kinetic studies When these phenazine derivatives are used as electron mediators in enzyme assays, they are reduced with NAD(P)H and the reduced derivatives are reoxidized with oxygen or other electron acceptors, such as MTT. We assume the scheme shown in Fig. 4 (which is illustrated for 2B) for these coupled reactions, where k l , k2 and k3 are the second-order rate con- stants of the indicated reactions and the value of k2 depends on the proton concentration, which was kept constant 296 - 1LI I NAD' M" *xCSkb $q $H2 cSh& &f$CM 0, CH3 MTT ly;Cstk H202 c 6 J N ~ ~ - w S CH3 f ormazan Fig. 4. Scheme of a coupled reaction system containing N A D H , u phenazine derivative. MTT and oxygen. k , , k z and k3 are the second- order rate constants of the indicated reactions. The phenazine deriva- tive shown is 2B throughout this study. The reactions in this scheme are ex- pressed by the following equations: d[NADH]/dt = -k,[NADH][P+] (1) d[P+]/dl= [PH](k,[02] + k,[MTT]) -kl[NADH][P+] (2) d[formazan]/dt = k3[PH][MTT] (3) (4) where [P'] and [PHI are the concentrations of the phenazine derivatives in their oxidized and reduced forms, respectively, and [ lo shows the initial concentration. At steady state, d[P+]/dt = 0 and Eqn (2) becomes: [P '10 = [P'] + [PHI [PH](kz[O,] + k,[MTT]) -kl[NADH][P '1 = 0. (5) From Eqns (3 - 5), the following equation is obtained: d[formazan]/dt = klk3[NADH][P'],,[MTT]/(kl[NADH] + k2[021 + k,[MTTI). (6) At first, we obtained the k , values for the phenazine de- rivatives by the following procedure. Under the experimental conditions of k3[MTT] 9 (k,[NADH] + k2[0,]), Eqn (6) is simplified to : d[formazan]/dt = kl [NADH][P+l0. (7) Therefore, if the concentration of MTT is large enough, the rate of formazan formation becomes independent of the change in MTT concentration. Under the experimental con- ditions shown in the legend to Table 1 the above conditions arc satisfied and, as the rate of the formazan formation can be measured within a 5% decrease of the initial NADH con- centration, Eqn (7) can be replaced by: d[formazan]/dt = k , [NADH],[P+],, . (8) Fig. 5 shows that the rate of formazan formation is pro- portional to the initial concentration of NADH and, using Eqn (8), we can calculate the k l value for derivative 2B to be , [NADHlo ( m M ) Fig. 5. Efects qf the initial concentration of' N A D H on the rate of' formazan formation in the system for 2B. The assay conditions are described in the legend to Table 1 Table 1 . Rate constants of N A D H oxidation by phenazine derivatives The rcaction mixture contained 1.5 mM MTT, phcnazine derivatives at the indicated concentrations and NADH; NADH concentration ranged over 0-2 mM for 1 A, 1 B and 2A and over 0-0.2 mM for 2B, 2C and 2D. Thc values of k l werc obtained from Eqn (8) as described in the text Compound Concentration k i CIM rnM- ' s - ' 1A 1.10 1.83 1 R 3.90 3.33 2A 0.190 0.75 2B 0.626 1.42 2c 0.61 3 1.6X 2 D 0.516 2.03 1.42 mM-' s-' from the slope of this line. It was confirmed that the rate is also proportional to the concentration of the phenazine derivatives. The kl values for the other derivatives were also obtained by the same method and are summarized in Table 1 . These results show that substitution at position 1 with a 3-carboxypropyloxy group increases the k , values about twofold and that the replacement of the 5-methyl group with the ethyl group decreases the kl values by about 60%. Next, we investigated the reaction of the reduced form of thc phenazine derivative (2B) with oxygen, using the same reaction system shown in Fig. 4, except for the omission of MTT. It was confirmed that 0.3 mM H202 was produced from 0.3 mM NADH during the incubation of NADH with 2B (30 pM) at 30°C for 45 min in 0.1 M phosphate buffer, pH 7.5; this means that the reducing equivalents of NADH convert O2 to Hz02. At steady state and without MTT, Eqn (2) becomes : kz[PH][Oz] - k,[NADH][P+] = 0 . (9) From Eqns (l), (4) and (9), the rate of NADH oxidation is expressed : -d[NADH]/dt = k i ~ ~ ~ ~ ~ ~ ~ I ~ ~ ' l o ~ ~ ~ l / ~ ~ ~ ~ ~ ~ ~ ~ l + kz[021). (10) Rearranging and inverting: k1[NADH][P']o/( -d[NADH]/df) = k,[NADH]/k,[O2] + 1. (1 1) Under the conditions of [NADH]x[NADH], and [O,] % [O2l0, the NADH-oxidation rates were measured at various 297 0 50 100 150 200 250 300 [ NADH 10 ( pM 1 Fig. 6. Plot of' Eqn ( 1 1 ) . The reaction mixture contained 3.3 pM 2B and the indicated concentrations of NADH. The reactions were monitored by the decrease in absorbancc at 340 nm 1.8 I I 0 2 4 6 8 1 0 1 2 I MTT 16' ( mM-' 1 Fig. 7. Plot uf Byn (12). The reaction mixture contained 62 pM 2B, 3.4 mM NADH and the indicakd concentrations of MTT. The reactions were monitorcd by the increase in absorbancc at 570 nm initial NADH concentrations. Fig. 6 shows the results obtained for 2B. From the slope of the straight line, the k2 value was calculated to be 1.21 mM-l s-', using Eqn (11), k l to be 1.42 mM ~ s ~ ' and the oxygen concentration to be 0.24 mM [22-241. Fig. 6 also shows that the value of the intercept on the ordinatc is 1.07, which agrees well with the theoretical value of 1 (Eqn 11). Finally, the value of k 3 for derivative 2B was measured by the following method. Eqn (6) is rearranged and inverted to: k ,[NADH][P ' ],,/(d[formazan]/dt) = (/cl[NADH] + k,[O2])/k3[MTT] + 1 . (12) Under conditions where [NADH] z [NADHIo, [02] z [ 0 2 ] 0 and [MTT] x [MTTIo, the formazan-formation rates were measurcd at various initial concentrations of MTT and the results are shown in Fig. 7. The data fit a straight line in- tercepting the ordinate at 1.02, which agrees well with the theoretical value of 1 (Eqn 12). From the slope of this line, thc k3 value is calculated to be 91 mM-' s - ' using the known values of k l , k , and [0210. DISCUSSION The spcctra of phenazine derivatives prepared in this case and those reported previously [3,13] are related to the follow- ing structural characteristics: alkylation (i.e. the cationic charge) at position 5, substitution at position 1 and ionization of the hydroxyl group at position 1. Therefore, we can group these spectra into five patterns which correspond to the fol- lowing compounds: (a) 1A and 2A; (b) 3C (at pH 7) and 3E (at pH 7);(c)3E(atpH 12);(d) lB, lE(atpH 2)[13], 1F[3], 2B, 2C, 2D, 3C (at pH 1) and 3E (at pH 1); (e) 1E (at pH 7) [I 31. On the basis of the above results, we identified the photo- and alkali-decomposed compounds of the phenazine derivd- tives and can indicate the structural factors affecting on the stability of the derivatives. The photo-decomposition of 5-alkylphenazines is associ- ated with hydroxylation at position 1, as has been reported for 1A [19 - 221, and 1-substitution makes them photostable, probably by preventing the formation of an adduct of 5-alkylphenazine with water [21]. The alkali-decomposition of 1 is due to elimination of the 5 methyl group and 1E is more resistant to this elimination than 1A and 1B. The ethyl group of 2 is not eliminated and the alkali-decomposition of 2 seems to be due to a ring- opening reaction [17,18]. As the ring-opening reaction occurs under more alkaline conditions than the elimination of the 5-methyl group of 1, 2 is more stable than 1. The pH depen- dence of the ring-opening reaction of 2 is affected by the kind of the substituent at position 1 : 2A and 2B are more stable than 2D and 2E is not decomposed, even at pH 12, probably due to the stabilizing effect of the ionized hydroxyl group. In previous reports [2, 221, photostability and alkali stability were not clearly distinguished, but now it is evident that their mechanisms are quite different. Kinetic studies of 5-methylphenazine have been reported by Ottaway [23] and Halaka et al. [22]. They used a more complicated reaction scheme than ours: Ottaway took the reaction of 1A with NADH to be reversible [23] and Halaka et al. added a step for inactive-complex formation of 1 A with the reduced form of 1A [22]. Their schemes, however, are not based on the chemical evidence, but are used for obtaining a better fit of a theoretical curve to the data of their transient kinctics. Therefore, we expressed the reactions of the phena- zine derivatives by Eqns (1-4) on thc basis of the simple scheme in Fig. 4. The results of the steady-statc reactions shown in Figs 5-7 are compatible with our equation5 and, therefore, under our experimental conditions, Eqns (1 -4) are valid for the reactions of the phenazine derivatives. It should also be pointed out that the rate constants can be obtained more directly and acculately by our steady-state method than by transient methods with complex curvc fitting. The k l value for 1A obtained in this work is about half of that reported by Halaka et al. [22]; this difference seem to be due to thc differences in the assay system and the method of analysis. Table 1 shows that the k , value depends on the substituents at positions 1 and 5. It has been demonstrated that the second- order rate constants for outer-sphere electron-transfer reac- tions are related to the difference in the standard oxidation- reduction potential (,!$) between the reactants [25]. The diffcr- ence in the k l values between the derivatives with the 5-methyl and 5-ethyl groups seems to be due to the difference in the E;; value because ,!$ values for 1A and 2A are 80 mV and 55 mV, respectively [26]; and the ratios of the k , values for 1A and 2A and the ratios of those for 3 B and 2B are almost the same (2.44 and 2.36, respectively). The effects of the substituents at position 1 on the k l values seem to include potential and charge effects; the analysis of these effects will be reported elsewhere. The phenazine derivatives lB, 2B and 2D synthesized in this work have a carboxyl or amino group at position 1 in the substituent. These functional groups can be used for coupling of these derivatives to other molecules such as water-soluble polymers, solid supports for immobilization and enzymes. In addition, 2B and 2D are stable under light and neutral conditions. Therefore, these derivatives have potential appli- cations in biochemical reactors and in analytical fields. In the following paper [27], we will describe a 5-ethylphenazine- poly(ethy1ene glycol) - NAD' conjugate, prepared by linking of 2B with a poly(ethy1ene glycol)-bound derivative of NAD' [7], as an example of the application of 2B; the conjugate works as a unique catalyst, having an intramolecular reaction step within its catalytic cycle, and provides a good model system for investigating the kinetics of the catalytic cycle of an enzyme. We are grateful to Dr Sasamoto (Dojindo Laboratories) for his advice on thc preparation of 1 -hydroxyphenazine. REFERENCES 1. Bernofsky, C. & Swan, M. (1973) Anal. Biochem. 53,452-458. 2. Ghosh, R. & Quayle, J. R. (1979) Anal. Biochem. 99, 112- 3. Hisada, R. & Yagi, T. (1977) J . Biochem. (Tokyo) 82, 1469- 4. Nakamura, S., Arimura, K., Ogawa, K. & Yagi, T. (1980) Clin. 5. Orsonneau, J. L., Meflah, K., Lustenberger, P., Cornu, G. & 6. Orsonneau, J . L., Meflah, K., Lustenberger, P., Cornu, G. & 7. Katayama, N., Urabe, I. & Okada, H. (1983) Eur. J . Biochem. 117. 1473. Chim. Acta 101,321 -326. Bernard, S . (1982) Clin. Chim. Acta 125, 177-184. Bernard, S. (1982) Clin. Chim. Acta 125, 185-192. 132,403-49. 8. Eguchi, T., Iizuka, T., Kagotani, T., Lee, J. H., Urabe, I . & 9. Nakamura, A., Urabe, I. & Okada, €1. (1986) J . Biol. Chem. 261, Okada, H. (1986) Eur. J . Biochem. 155, 415-421. 16792-16794. 10. Jacobs, S . (1959) Nature (Lund.) 183, 262. 11. Fields, R. (1971) Biochem. J . 124, 581 -590. 12. Putter, J. & Becker, R. (1983) in Methodr ofenzymatic analysis (Bergmeyer, H. U., ed.) 3rd edn, vol. 3, pp. 286-293, Verlag Chemie GmbH, Weinheim. 13. Zaugg, W. S. (1964) J . Biol. Chem. 239, 3964-3970. 14. Surrey, A. R. (1955) in Organic synthesb (Homing, E. C . , ed.), collective vol. 3, pp. 753-756, John Willey & Sons, New York. 15. Romer, A. (1982) Org. Mugn. Resonance 19, 66-68. 16. Eguchi, T., Kanzaki, N., Kagotani, T., Tdniguchi, T., Urabe, I. & 17. Johnson, S . L. & Rumon, K. A. (1970) Biochemistry 9,847-857. 18. Johnson, S. L. & Morrison, D. L. (1970) Biochemistry 9, 1460- 19. McIlwain, H. (1937) J . Chem Soc., 1704-1711. 20. Chew, V. S. F. & Bolton, J. R. (1980) J . Phys. Chem. 84, 1903- 21. Chew, V. S. F., Bolton, J. R., Brown, R . G. & Porter, G. (1980) 22. Halaka, F. G., Babcock, G. T. & Dye, J . L. (1982) J . Biol. Chem. 23. Ottaway, J. H. (1966) Biochem. J . 99, 253-256. 24. Reynafarje, B., Costa, L. E. & Lehninger, A. L. (1985) Anal. 25. Marcus, R. A. (1964) Annu. Rev. Phys. Chem. 15, 155-196. 26. Sober, H. A. (1970) Handbook of hiochemistry, 2nd edn, 5-36, The 27. Yomo, T., Sawai, H., Urabe, 1. & Okada, H. (1989) Eur. J . Okada, H. (1985) J . Ferment. Technol. 63, 563-565. 1470. 1908. J . Phys. Chem. 84, 1909-1916. 257,1458-1461. Biochem. 145,406-418. Chemical Rubber Co., Cleveland. Biochem. 179, 299 -305.