Linear free energy substitutent effect on flavin redox chemistry

Article abstract of DOI:10.1021/ja972992n

A systematic study on the effect of various substituents at the 7- and/or 8-position on the redox properties of isoalloxazines (flavins) is reported. The redox properties of these flavin derivatives were studied by cyclic voltammetry in 100 mM, pH 7.4 HEPES and 200 mM, pH 10 borate buffers. The magnitude and direction of the effect was dependent on the nature and location of the substituent. The redox potentials of the substituted flavins were correlated with the Hammett σ value of the substituents.

Full text of DOI:10.1021/ja972992n

J. Am. Chem. Soc. 1998, 120, 2251-2255
2251
Linear Free Energy Substituent Effect on Flavin Redox Chemistry
Justin J. Hasford and Carmelo J. Rizzo*
Contribution from the Department of Chemistry, Box 1822, Station B, Vanderbilt UniVersity,
NashVille, Tennessee 37235
ReceiVed August 26, 1997
Abstract: A systematic study on the effect of various substituents at the 7- and/or 8-position on the redox
properties of isoalloxazines (flavins) is reported. The redox properties of these flavin derivatives were studied
by cyclic voltammetry in 100 mM, pH 7.4 HEPES and 200 mM, pH 10 borate buffers. The magnitude and
direction of the effect was dependent on the nature and location of the substituent. The redox potentials of the
substituted flavins were correlated with the Hammett σ value of the substituents.
Flavoenzymes are an ubiquitous and diverse class of biologi-
cal redox catalysts.1 The redox active centers of these enzymes
are derivatives of riboflavin (vitamin B2, 1a (see Figure 1)),
which consists of an isoalloxazine ring system substituted with
methyl groups at C7 and C8 and a ribityl group at N10; the
biologically active forms are derivatized at the 5′-hydroxyl to
give FMN (1b) or FAD (1c). A number of naturally occurring
flavin analogues have been isolated from a variety sources,
including roseoflavin (2) and 8-demethyl-8-hydroxyriboflavin
(3) in which the C8 methyl group is replaced with a dimethyl-
amino and hydroxyl group.2 Riboflavin derivatives with sub-
stitutents at C6 have also been isolated. Flavoprotein FP390
isolated from luminescent bacteria possesses a covalent linkage
from C6 of the isoalloxazine ring to C3 of myristic acid.3
Flavoenzymes in which the cofactor is covalently bound to the
protein have also been discovered.2 The covalent attachment
is usually to the C8-methyl group (4) via a thioether linkage to
Figure 1.
probes to better understand the redox properties of the holoen-
zyme,8 we have studied the effect of substituents on the redox
properties of flavins. A series of 7- and/or 8-substituted flavins
has been synthesized and their redox chemistry studied by cyclic
voltammetry (CV). The reduction potential of flavin analogues
is linearly correlated with the Hammett σ value.9
cysteine or ether linkage to tyrosine or to either imidazole
nitrogens of histidine. Trimethylamine dehydrogenase is co-
valently bound through a thioether linkage directly to C6 of
FMN.4
It is well-known that substituents on the benzene subnucleus
can change the redox chemistry of flavins with the magnitude
and direction of the effect being dependent on the nature and
position of the substituent.1 A systematic study, however, has
not been reported. Because of the interest in the preparation
of flavin models5-7 and analogues as structural and mechanistic
A series of flavins substituted at C7 and/or C8 was synthe-
sized by the condensation of alloxan with the properly substi-
(5) (a) Breinlinger, E.; Niemz, A.; Rotello, V. M. J. Am. Chem. Soc.
1995, 117, 5379. (b) Shinkai, S.; Honda, N.; Ishikawa, Y.; Manabe, O. J.
Am. Chem. Soc. 1985, 107, 6286. (c) Akiyama, T.; Simeno, F.; Murakami,
M.; Yoneda, F. J. Am. Chem. Soc. 1992, 114, 6613.
(6) (a) Hasford, J. J.; Kemnitzer, W.; Rizzo, C. J. J. Org. Chem. 1997,
62, 5244. (b) Hasford, J. J.; Rizzo, C. J. Tetrahedron Lett. 1998, 39, in
press. (c) Shokat, K. M.; Leumann, C. J.; Sugasawara, R.; Schultz, P. G.
Angew. Chem., Int. Ed. Engl. 1988, 27, 1172. (d) Seward, E. M.; Hopkins,
R. B.; Sauerer, W.; Tam, S.-W.; Diederich, F. J. Am. Chem. Soc. 1990,
112, 1783. (e) Shinkai, S.; Kawase, A.; Yamaguchi, T.; Manabe, O.; Wada,
Y.; Yoneda, F.; Ohta, T.; Nishimoto, K. J. Am. Chem. Soc. 1989, 111,
4928.
(7) (a) Breinlinger, E. C.; Rotello, V. M. J. Am. Chem. Soc. 1997, 119,
1165. (b) Zipplies, M. F.; Staab, H. A. Tetrahedron Lett. 1985, 25, 1035.
(c) Zipplies, M. F.; Krieger, C.; Staab, H. A. Tetrahedron Lett. 1983, 24,
1925.
(8) Walsh, C.; Fisher, J.; Spencer, R.; Graham, D. W.; Ashton, W. T.;
Brown, J.; Brown, R. D.; Rogers, E. F. Biochemistry 1978, 17, 1942.
(9) (a) Hansch, C.; Leo, A.; Unger, S. H.; Kim, K. H.; Nikaitani, D.;
Lien E. J. J. Med. Chem. 1973, 16, 1207. (b) Hine, J. Structural Effects on
Equilibria in Organic Chemistry; R. E. Krieger Publishing Co.: Huntington,
NY, 1981. (c) Murav, S. L.; Carmichael, I.; Hug, G. L. Handbook of
Photochemistry, 2nd ed.; Marcel Dekker, Inc.: New York, 1993; pp 345-
348.
(1) (a) Muller, F. In Chemistry and Biochemistry of FlaVoenzymes;
Muller, F., Ed.; CRC Press: Boca Raton, FL, 1991; Vol. 1, pp 1-71. (b)
Stankovich, M. T. In Chemistry and Biochemistry of FlaVoenzymes; Muller,
F., Ed.; CRC Press: Boca Raton, FL, 1991; Vol. 1, pp 401-425
(2) (a) Edmondson, D. E.; De Francesco, R. In Chemistry and Biochem-
istry of FlaVoenzymes; Muller, F., Ed.; CRC Press: Boca Raton, FL, 1991;
Vol. 1, pp 73-103. (b) Matsui, K.; Kasai, S. In Chemistry and Biochemistry
of FlaVoenzymes; Muller, F., Ed.; CRC Press: Boca Raton, FL, 1991; Vol.
1, pp 105-120. (c) Decker, K. In Chemistry and Biochemistry of
FlaVoenzymes; Muller, F., Ed.; CRC Press: Boca Raton, FL, 1991; Vol. 2,
pp 343-375. (d) Decker, K. F. Annu. ReV. Nutr. 1993, 13, 17. (e) Decker,
K.; Brandsch, R. Methods Enzymol. 1997, 280, 413.
(3) (a) Kita, A.; Kasai, S.; Miki, K. J. Biochem. 1995, 117, 575. (b)
Moore, S. A.; James, M. N. G.; O’Kane, D. J. O.; Lee, J. EMBO J. 1993,
12, 1767.
(4) (a) Barber, M. J.; Neame, P. J.; Lim, L. W.; White, S.; Mathews, F.
S. J. Biol. Chem. 1992, 267, 6611. (b) Lim, L. W.; Shamala, N.; Mathews,
F. S.; Steenkamp, D. J.; Hamlin, R.; Xuong, N. J. Mol. Biol. 1986, 261,
15140. (c) Steenkamp, D. J.; Mathews, F. S. In Chemistry and Biochemistry
of FlaVoenzymes; Muller, F., Ed.; CRC Press: Boca Raton, FL, 1991; Vol.
2, pp 395-424.
S0002-7863(97)02992-2 CCC: $15.00 © 1998 American Chemical Society
Published on Web 02/27/1998

2252 J. Am. Chem. Soc., Vol. 120, No. 10, 1998
Hasford and Rizzo
Scheme 1
Scheme 2
tuted N-methyl-o-phenylenediamines, which in turn were ob-
tained from the reduction of the corresponding N-methyl-o-
nitroaniline.10 In general, the nitroanilines were reduced by
catalytic hydrogenation over palladium on carbon except for
those bearing chlorine substitutents. In these cases, hydrogena-
tion resulted in partial reduction of the aryl chloride; thus
reduction of the nitro group was accomplished with tin(II)
chloride.11 The N-methyl-o-phenylenediamines were used
without further purification, and condensation with alloxan in
10% aqueous HCl gave the desired flavins in 25-40% overall
yield after recrystallization from formic acid/water (Scheme 1).
The redox chemistry of flavins is shown in Scheme 2.1 One-
electron reduction of flavins to give a delocalized radical anion
is the rate-limiting step. In protic media, this radical anion is
protonated to give the semiquinone which undergoes a fast,
second one-electron reduction to give the reduced flavin anion.
The pKa of the N1 proton of reduced flavin is approximately
6.5 and is largely ionized under the conditions we employed.
In protic solvent, the entire electrochemical-chemical-
electrochemical (ECE) reaction is faster than the CV time scale
and is observed as a single wave, and is electrochemically
reversible. For the flavins used in this study, the peak
separations were between 40 and 48 mV, indicative of a two-
electron process. Brockman and Pearson reported that the
polaragraphic half-wave potential (E1/2) of a series of substituted
benzophenone derivatives could be correlated to the Hammett
substituent constant.12a Since this initial observation, the E1/2
of a number of substituted benzenes carrying reducible groups
(carbonyls, thionocarbonyls, and nitroso and azo groups) have
been correlated with the Hammett σ and/or σ- value of the
substitutent.12b Of course, the Hammett σ value is dependent
upon the location of the substituent relative to the reducible
group. For flavins, since the initial radical anion intermediate
is coupled to the benzene subnucleus through the N5 position,
we reasoned that the position of substituents relative to N5 was
an important consideration in determining their influence on
redox potential. We thus defined C7 as the meta position and
C8 as para. The redox properties of flavins 9-25 were studied
Figure 2. (a) Plot of Hammett σ versus -log(E°′/E0). (b) Plot of
Hammett σ- versus -log(E°′/E0°′) for 7- and/or 8-substituted flavins.
E0°′ is the reduction potential of 9 (-407 mV vs Ag/AgCl).
by cyclic voltammetry using a standard three-electrode cell with
a glassy carbon working electrode, Ag/AgCl reference electrode
(+197 mV vs SHE), and platinum wire counter electrode. The
reduction potentials were measured in 100 mM, pH 7.4 HEPES
and 200 mM, pH 10 borate buffers. The observed reduction
potentials of the flavin analogues along with the Hammett σ
and σ- values are summarized in Table 1. The unsubstituted
flavin 9 is taken as the base value (E0°) and had a reduction
potential of -407 and -503 mV at pH 7.4 and 10, respectively.
When the log of the normalized reduction potentials were plotted
against the Hammett σ value, a linear correlation was observed
(Figure 2). For 7,8-disubstituted flavins, the σ values of the
substituents were simply added. The influence of the substituent
is relatively small as indicated by the F value of 0.17. This
value is similar to that obtained by Brockman and Pearson for
substituted benzophenones (F ) 0.25).12a
While the linear free energy correlation between flavin
reduction potential and the Hammett σ is acceptable (R ) 0.97
and 0.96 at pH 7.4 and 10.0, respectively), given the mechanism
shown in Scheme 2, it might be expected that σ- would give
a better correlation since the C8 substituent could interact
directly with the negative charge of the flavin radical anion
through the N5 position. When the reduction potentials were
plotted against σ-, the correlation was slightly worse at pH
7.4 (R ) 0.89); however, it should be noted that for most of
our substituents the σ and σ- values are nearly the same. There
were three substituents in which the σ and σ- values are
significantly different, the p-NMe2 (23), p-CN (24), and m-CN
(25). For the electron-rich p-NMe2, there was particularly large
(10) (a) Lambooy, J. P. Heterocycl. Compd. 1967, 9, 118. (b) Brown, S.
A.; Rizzo, C. J. Synth. Commun. 1996, 26, 4065.
(11) Bellamy, F. D.; Ou, K. Tetrahedron Lett. 1984, 25, 839.
(12) (a) Brockman, R. W.; Pearson, D. E. J. Am. Chem. Soc. 1952, 74,
4128. (b) Zuman, P. Substituent Effects in Organic Polarography; Plenum
Press: New York, 1967; pp 43-130.
(13) Bard, A. J.; Faulkner, L. R. Electrochemical Methods. Fundamentals
and Applications; J. Wiley & Sons: New York, 1980.

Substituent Effect on FlaVic Redox Chemistry
J. Am. Chem. Soc., Vol. 120, No. 10, 1998 2253
Table 1. Reduction Potentials (mV) of Flavins 9-25 in 100 mM, pH 7.4 HEPES and 200 mM, pH 10 Borate Buffers vs Ag/AgCl (+197 mV
vs SHE)a
pH 7.4
pH 10.0
flavin
Rm
Rp
E°′
∆E°′
E°′
∆E°′
σ
σ-
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
H
CH3
H
CH3
Cl
H
Cl
CH3
Cl
F
H
F
F
OMe
H
H
H
H
CH3
CH3
H
Cl
Cl
Cl
CH3
H
F
CH3
Cl
-407
-419
-439
-456
-371
-385
-338
-390
-402
-366
-399
-398
-344
-396
-580
-283
-321
0
-503
-517
-535
-548
-460
-469
-417
-485
-489
-454
-485
-488
-434
-483
0
0.00
-0.07
-0.17
-0.24
0.37
0.23
0.60
0.16
0.20
0.34
0.06
0.17
0.57
0.12
-0.83
0.66
0.56
0.00
-0.03
-0.15
-0.18
-12
-32
-49
+36
+22
+69
+17
+5
+41
+8
+9
+63
+11
-173
+124
+86
-14
-32
-45
+43
+34
+86
+18
+14
+49
+18
+15
+69
+20
0.27
0.24
0.05
H
0.13
-0.12
1.00
NMe2
CN
H
-369
-408
+134
+95
CN
0.68
a The formal standard potential of each substrate was determined by the relationship E°′ ) (Ep + Ep )/2.13
c
a
deviation when σ- was used; however, for the cyano-substituted
flavins 25 and 26,14 the Hammett σ- value appears to give better
correlation. Flavin 23 was not sufficently soluble at pH 10 for
electrochemical studies. When this substrate was omitted, the
σ- gave better correlation (R ) 0.99). It is established that
the spin of flavin radical anion and semiquinone is delocalized
into the aromatic ring.15 Our observation that the Hammett σ
gave better correlation than σ- would suggests that the charge
of the radical anion is probably not extensivly delocalized into
the aromatic ring except when the substituent is a strong
electron-withdrawing group such as cyano.
It has been pointed out that using the log of the redox potential
instead of E°′, is theoretically unjustified.12b When we plotted
the midpoint potentials of flavins 9-25 versus the Hammett σ,
a nearly identical correlation was observed, and when σ- was
used, the correlation was slightly worse. An advantage of using
the log of the redox potentials is that both pH conditions could
be plotted together to give a linear Hammett correlation so long
as the proper E0°′ was used (Figure 3). This would suggest
that the effect of benzo substitution on redox potential is largely
independent of the buffer microenvironment and pH in the pH
range used in our study.
Figure 3. Plot of Hammett σ versus -log(E°′/E0) in 100 mM, pH 7.4
HEPES (4, E0 ) 407 mV vs Ag/AgCl) and 200 mM, pH 10 borate
(1, E0 ) 503 mV vs Ag/AgCl) buffers.
Ortho substituents on aromatic rings often show nonlinear
Hammett relationships due to steric and dipole effects. In our
analysis, the 6-position of flavins is regarded as the ortho
position. We have previously examined the redox properties
of 9-substituted flavins which show significant steric effects.6a,b
The conformation of fully reduced flavin is bent along the N5,-
N10 axis; in free flavin, the bent conformation places the N10
substituent in a pseudoaxial position above the plane of the
isoalloxazine ring system. For 9,10-dimethylflavin (26, Figure
4), a better steric arrangement of the methyl groups is achieved
upon reduction. As such, we found that the reduction potential
of 26 was shifted +50 mV compared to unsubstituted flavin 9.
Thus, Hammett σ values should be regarded as unreliable for
predicting the redox potential of 9-substituted flavins. The
Figure 4.
reduction potential of 7,9,10-trimethylflavin (27) was determined
to be -370 mV at pH 7.4 (∆E°′) +37 mV) and shows an
additive effect for the C7 (10, ∆E°′) -12 mV) and C9 (26,
∆E°′) +50 mV) methyl groups. For 8,9,10-trimethylflavin,
however, the observed redox potential is -368 mV, which
deviates significantly from that predicted by simply adding the
individual effects of the 8- and 9-methyl groups (-389 mV at
(14) Bruice, T. C.; Chan, T. W.; Taulane, J. P.; Yokoe, I.; Elliot, D. L.;
Williams, R. F.; Novak, M. J. Am. Chem. Soc. 1977, 99, 6713.
(15) Zheng, Y.-J.; Ornstein, R. L. J. Am. Chem. Soc. 1996, 118, 9402.

2254 J. Am. Chem. Soc., Vol. 120, No. 10, 1998
Hasford and Rizzo
Table 2. Reduction Potentials (mV) of Flavins 26-28 in 100 mM,
pH 7.4 HEPES and 200 mM, pH 10 Borate Buffers vs Ag/AgCl
(+197 mV vs SHE)
hertz. NMR samples were prepared with 99.5 atom % D CF3CO2D
after first evaporating the sample from exchange grade CF3CO2D (99
atom % D). TMS (δ ) 0) or residual CF3CO2H (δ ) 11.5 for 1H and
δ ) 116.6 and 164.4 for 13C) was used as an internal reference. UV
spectra were recorded on a Shimadzu 2101 UV-vis spectrophotometer
in 0.1 M NaOH solution. High-resolution mass spectra were obtained
from the University of Illinois Mass Spectrometry Center. FAB mass
spectra were obtained using “magic bullet” as the matrix. Melting
points were collected on a Mel-Temp 3.0 apparatus and are uncorrected.
Flavins tend to be high melting compounds and often decompose at
their melting temperature. Most of our flavins decomposed without
melting, in which case the melting points listed are the temperature at
which they decomposed.
pH 7.4
pH 10.0
flavin
E°′
∆E°′
E°′
∆E°′
26
27
28
-357
-370
-368
+50
+37
-39
-458
-470
-463
+45
+33
+40
Scheme 3
The required N-methyl-o-nitroanilines were prepared as previously
described.10,17 N-Methyl-o-nitroanilines were purchased from Aldrich
and used without further purification. Lumiflavin (12) was purchased
from Sigma and used as received. Flavin 23 was prepared from 19
under conditions previously described for structurally related flavins.8
Flavins 24 and 25 were synthesized as described by Bruice et al.14
Synthesis of Flavins. The required N-methyl-o-nitroanilines were
reduced to the corresponding N-alkyl-o-phenylenediamine via catalytic
hydrogenation over palladium on carbon in methanol and 1 atm of
hydrogen or with SnCl2 in ethyl acetate at reflux.11 The crude
N-methyl-o-phenylenediamine was dissolved in 10% aqueous HCl, and
1.25 equiv of alloxan monohydrate added. The reaction was heated to
60 °C with stirring for 12 h, during which time a bright yellow
precipitate formed. The yellow precipitate was collected by filtration
and recrystallized from formic acid/water to give the desired flavin in
25-40% overall yield from the N-methyl-o-nitroaniline.
pH 7.4) (see Table 2). We attribute this nonlinear behavior to
an enhanced steric effect due to steric crowding of the 8,9,10-
methyl groups.
10-Methylisoalloxazine (9): decomposed without melting at 337
There have been a number of studies aimed at determining
how the protein modulates the redox and catalytic properties
of flavin cofactors.5-7 A combination of hydrogen bonding,
conformational, π-stacking, and desolvation effects contribute
to this phenomena. Our results show that benzo substitution
shifts the redox potential in a predictable manner and provides
insight into how modified flavins such as roseoflavin (2) and
8-demethyl-8-hydroxyriboflavin (3) could modulate redox po-
tential. For instance, the Hammett analysis predicts that
protonation of the C8 dimethylamino group of roseoflavin would
significantly alter the redox potential of this cofactor. The
Hammett σ value for p-NH2 is -0.66, while that for p-NH3+ is
+0.60, representing a shift of nearly 200 mV to more positive
values; it should be pointed out that the biological function of
roseoflavin is largely unexplored. Similarly, deprotonation of
the phenolic group of 8-demethyl-8-hydroxyriboflavin (3) is
predicted to shift the reduction potential by -80 mV based on
the Hammett σ values of p-OH (-0.37) versus p-O- (-0.81)
(see Scheme 3).
1
°C; H NMR (CF3CO2D) δ 8.64 (d, J ) 8.3, 1H), 8.54 (t, J ) 7.4,
1H), 8.42 (d, J ) 8.3, 1H), 8.34 (t, J ) 7.4, 1H), 4.67 (s, 3H); 13C
NMR (CF3CO2D) δ 161.0, 151.1, 144.4, 143.8, 142.4, 136.1, 135.0,
134.3, 133.9, 119.3, 38.7; UV (0.1 M NaOH) λ 437 (ꢀ ) 7800), 340
(ꢀ ) 8400), 264 (ꢀ ) 27 100); HRMS (FAB) calcd for C12H11N4O2
(M + H) 229.0726, found 229.0724.
7,10-Dimethylisoalloxazine (10): decomposed without melting at
1
336 °C (lit.18 mp > 280 °C; H NMR (CF3CO2D) δ 8.43-8.30 (m,
3H), 4.67 (s, 3H), 2.81 (s, 3H); 13C NMR (CF3CO2D) δ 161.1, 151.1,
148.2, 146.2, 143.5, 142.8, 134.6, 133.9, 123.2, 122.3, 118.9, 38.7,
21.8; UV (0.1 M NaOH) λ 448 (ꢀ ) 11 900), 341 (ꢀ ) 12 300), 269
(ꢀ ) 40 000); HRMS (FAB) calcd for C12H11N4O2 (M + H) 243.0882,
found 243.0883.
8,10-Dimethylisoalloxazine (11): decomposed without melting at
329 °C; 1H NMR (CF3CO2D) δ 8.50 (d, J ) 8.7 Hz, 1H), 8.22 (s, 1H),
8.13 (d, J ) 8.7 Hz, 1H), 4.65 (s, 3H), 2.93 (s, 3H); 13C NMR (CF3-
CO2D) δ 161.3, 160.1, 151.1, 144.1, 141.5, 137.3, 135.6, 134.4, 132.5,
118.6, 38.5, 24.5; UV (0.1 M NaOH) λ 437 (ꢀ ) 10 200), 352 (ꢀ )
9900), 265 (ꢀ ) 28 100); HRMS (FAB) calcd for C12H11N4O2 (M +
H) 243.0882, found 243.0883.
The linear relationship between the rate of flavin-catalyzed
reactions with the polaragaphic half-wave or redox potential
has been well appreciated.6b,8,16 We have demonstrated that the
redox potential for 7- and/or 8-substituted flavins is linearly
correlated with the Hammett σ value. This relationship allows
for the accurate prediction of the redox potential for substituted
flavins and should aid in the design of flavin analogues as
mechanistic probes and potential therapeutic agents as well as
mechanistic interpretation of flavin-mediated reactions.
7-Chloro-10-methylisoalloxazine (13): decomposed without melt-
1
ing at 330-332 °C; H NMR (CF3CO2D) δ 8.53 (s, 1H), 8.37-8.29
(m, 2H), 4.60 (s, 3H); 13C NMR (CF3CO2D) δ 160.8, 151.3, 144.6,
143.5, 142.3, 142.2, 135.7, 134.2, 132.5, 120.4, 38.8; UV (0.1 M NaOH)
λ 447 (ꢀ ) 10 900), 333 (ꢀ ) 10 000), 270 (ꢀ ) 38 400).
8-Chloro-10-methylisoalloxazine (14): mp 323-324 °C (dec) (lit.17
1
mp 315-325 °C); H NMR (CF3CO2D) δ 8.70 (d, J ) 8.8, 1H), 858
(s, 1H), 8.35, (d, J ) 8.8, 1H), 4.78 (s, 3H); 13C NMR (CF3CO2D) δ
160.7, 152.3, 151.3, 145.1, 140.5, 136.6, 135.8, 134.4, 133.9, 119.1,
38.4; UV (0.1 M NaOH); λ 435 (ꢀ ) 11 500), 350 (ꢀ ) 10 800), 268
(ꢀ ) 31 900); HRMS (FAB) calcd for C11H8ClN4O2 (M + H) 263.0336,
found 263.0337.
Experimental Section
Proton and carbon-13 NMR data were recorded at 300 and 75 MHz,
respectively, in CF3CO2D. Chemical shifts are reported in parts per
million downfield from TMS (δ ) 0); coupling constants are given in
7, 8-Dichloro-10-methylisoalloxazine (15): decomposed without
1
melting at 335-336 °C (lit.19 346 °C (dec)); H NMR (CF3CO2D) δ
(16) (a) Muller, F.; Massey, V. J. Biol. Chem. 1969, 244, 4007. (b)
Bruice, T. C.; Main, L.; Smith, S.; Bruice, P. Y. J. Am. Chem. Soc. 1971,
93, 7327. (c) Gumbley, S. J.; Main, L. Tetrahedron Lett. 1976, 3209. (d)
Gascoigne, I. M.; Radda, G. K. Biochim. Biophys. Acta 1967, 131, 498. (e)
Loechler, E. L.; Hollocher, T. C. J. Am. Chem. Soc. 1975, 97, 3235.
(17) Kumar, V.; Woode, K. A.; Bryan, R. F.; Averill, B. A. J. Am. Chem.
Soc. 1986, 108, 490.
(18) Hall, D. M.; Turner, E. E. J. Chem. Soc. 1945, 699.
(19) King, F. E.; Clark-Lewis, J. W. J. Chem. Soc. 1953, 172.

Substituent Effect on FlaVic Redox Chemistry
J. Am. Chem. Soc., Vol. 120, No. 10, 1998 2255
8.59 (s, 1H), 8.47 (s, 2H), 4.53 (s, 3H); 13C NMR (CF3CO2D) δ 160.1,
150.2, 145.7, 141.2, 140.5, 139.9, 135.6, 135.4, 132.9, 120.7, 38.6;
UV (0.1 M NaOH) λ 445 (ꢀ ) 8300), 340 (ꢀ ) 7200), 273 (ꢀ ) 32 100).
8-Chloro-7,10-dimethylisoalloxazine (16): mp 323-324 °C (dec);
1H NMR (CF3CO2D) δ 8.46 (s, 1H), 8.44 (s, 1H), 4.61 (s, 3H), 2.79,
(s, 3H); 13C NMR (CF3CO2D) δ 161.0, 153.9, 151.2, 146.6, 144.2,
141.1, 135.9, 133.7, 132.8, 119.4, 38.7, 20.9; UV (0.1 M NaOH) λ
445 (ꢀ ) 8700), 349 (ꢀ ) 8100), 269 (ꢀ ) 25 900).
9,10-Dimethylisoalloxazine (26): mp 315 °C; 1H NMR (CF3CO2D)
δ 8.45 (d, J ) 8.2 Hz, 1H); 8.33 (d, J ) 7.2 Hz, 1H); 8.14 (t, J ) 7.8
Hz, 1H); 4.75 (s, 3H); 3.10 (s, 3H); 13C NMR (CF3CO2D) δ 161.0,
151.0, 148.1, 145.7, 144.0, 135.3, 134.6, 134.4, 133.6, 131.9, 45.2,
24.2; UV (0.1 M NaOH) λ 443 (ꢀ ) 6400), 354 (ꢀ ) 8600), 267 (ꢀ )
22 300); HRMS (FAB) calcd for C11H9N4O2 (M + H) 243.0882, found
243.0881.
7,9,10-Trimethylisoalloxazine (27): mp 314-6 °C (dec); 1H NMR
(CF3CO2D) δ 8.23 (s, 1H), 7.49 3H), 2.55 (s, 3H); 13C NMR (CF3-
CO2D) δ 161.2, 151.0, 150.4, 147.4, 144.8, 144.3, 133.6, 133.2, 133.1,
131.2, 44.9, 24.1, 21.4; UV (0.1 M NaOH) λ 452 (ꢀ ) 8600), 355 (ꢀ
) 11 000), 273 (ꢀ ) 31 500); HRMS (FAB) calcd for C13H13N4O2 (M
+ H) 257.1039, found 257.1038.
8,9,10-Trimethylisoalloxazine (28): mp 283-5 °C (dec); 1H NMR
(CF3CO2D) δ 8.20 (d, J ) 8.6 Hz, 1H), 8.10 (d, J ) 8.5 Hz, 1H), 4.67
(s, 3H), 2.88 (s, 3H), 2.84 (s, 3H); 13C NMR (CF3CO2D) δ 161.6, 160.4,
151.3, 145.9, 143.2, 138.0, 136.5, 133.2, 131.9, 129.8, 46.8, 23.6, 20.1;
UV (0.1 M NaOH) λ 443 (ꢀ ) 14 200), 374 (ꢀ ) 19 000), 269 (ꢀ )
44 800); HRMS (FAB) calcd for C13H13N4O2 (M + H) 257.1039, found
257.1038.
7-Chloro-8,10-dimethylisoalloxazine (17): decomposed without
1
melting at 341-3 °C; H NMR (CF3CO2D) δ 8.53 (s, 1H), 8.27 (s,
1H), 4.58 (s, 3H), 2.86, (s, 3H); 13C NMR (CF3CO2D) δ 161.0, 156.9,
151.2, 144.2, 143.6, 141.4, 134.2, 133.9, 132.7, 120.1, 38.7, 23.4; UV
(0.1 M NaOH) λ 446 (ꢀ ) 7400), 344 (ꢀ ) 6700), 270 (ꢀ ) 26 000).
7-Fluoro-10-methylisoalloxazine (18): decomposed without melting
1
at 303-305 °C; H NMR (CF3CO2D) δ 8.39 (dd, J ) 9.2, 4.2, 1H),
8.16 (d, J ) 7.1, 1H), 8.13 (m, 1H), 4.59 (s, 3H); 13C NMR (CF3-
CO2D) δ 166.1 (JC-F ) 263 Hz), 160.6, 151.2, 144.3, 143.3 (JC-F
)
13 Hz), 135.5, 132.7 (JC-F ) 27 Hz), 130.9, 121.7 (JC-F ) 10 Hz),
119.7 (JC-F ) 23 Hz), 39.0; UV (0.1 M NaOH) λ 446 (ꢀ ) 10 200),
330 (ꢀ ) 9900), 266 (ꢀ ) 26 500); HRMS (FAB) calcd for C11H7-
FN4O2 (M + H) 247.0063, found 247.0063.
8-Fluoro-10-methylisoalloxazine (19): decomposed without melting
8-(Dimethylamino)-10-methylisoalloxazine (23). 8-Fluoro-10-
methylisoalloxazine (19) (45 mg, 0.18 mmol), dimethylamine hydro-
chloride (58 mg, 0.72 mmol), and sodium acetate (88 mg, 1.08 mmol)
in N,N-dimethylformamide (3.6 mL) were heated at 70 °C for 4 h.
During this time, the yellow suspension was converted into a brick red
solid. The solid was collected by filtration and washed with water
then methanol. Recrystallization from formic acid/water gave 23 (32
mg, 66%) as a brick red powder which decomposed without melting
at 389-90 °C (lit.20 mp > 360 °C): 1H NMR (CF3CO2D) δ 8.09 (d, J
) 9.9 Hz, 1H), 7.78 (d, J ) 9.9 Hz, 1H), 6.90 (s, 1H), 4.25 (s, 3H),
3.63 (s, 6H); 13C NMR (CF3CO2D) δ 161.3, 152.1, 144.4, 141.6, 139.3,
137.6, 125.0, 120.0, 96.5, 43.3, 36.9; UV (0.1 M NaOH) λ 487 (ꢀ )
14 900), 308 (ꢀ ) 4600), 254 (ꢀ ) 18 600); HRMS (FAB) calcd for
C13H14N5O2 (M + H) 272.1148, found 272.1147.
Cyclic Voltammetry. A 0.5 mM solution of the flavin in 100 mM,
pH 7.4 HEPES (using HEPES hemisodium salt) or 200 mM, pH 10
borate buffer was prepared. The flavin was dissolved by sonication,
and any undissolved flavin was removed by filtration through a
kimwipe; the solution was degassed with argon. Electrochemical
measurements were performed on an EG&G model 263 potentiostat/
galvanostat using a standard three-electrode cell with a glassy carbon
working electrode (BAS No. MF-2012), a Ag/AgCl reference electrode
(BAS No. MF-2063), and a platium wire counter electrode (BAS No.
MW-1032). The sweep rate was 20 mV/s.
1
at 374-376 °C; H NMR (CF3CO2D) δ 8.60 (dd, J ) 9.3, 5.5, 1H),
8.01 (dd, J ) 8.8, 2.1, 1H), 7.92 (dt, J ) 7.3, 2.2, 1H), 4.53 (s, 3H);
13C NMR (CF3CO2D) δ 172.8 (JC-F ) 275 Hz), 161.1, 151.5, 145.5,
139.6 (JC-F ) 25 Hz), 139.5, 136.4 (JC-F ) 13.6 Hz), 133.4, 125.1
(JC-F ) 26 Hz), 106.2 (JC-F ) 30 Hz), 38.9; UV (0.1 M NaOH); λ
430 (ꢀ ) 10 000), 343 (ꢀ ) 8800), 263 (ꢀ ) 46 900); HRMS (FAB)
calcd for C11H7FN4O2 (M + H) 247.0063, found 247.0631.
7-Fluoro-8,10-dimethylisoalloxazine (20): mp 331-332 °C (dec);
1H NMR (CF3CO2D) δ 8.23 (d, J ) 5.8, 1H), 8.08 (d, J ) 7.7, 1H),
4.56 (s, 3H), 2.73 (s, 3H); 13C NMR (CF3CO2D) δ 167.2, 160.9, 151.1,
148.7 (JC-F ) 22 Hz), 143.7, 142.7 (JC-F ) 13 Hz), 133.9, 131.2, 121.3
(JC-F ) 6 Hz), 118.7, 39.0, 17.8 (JC-F ) 4 Hz); UV (0.1 M NaOH) λ
446 (ꢀ ) 12 700), 344 (ꢀ ) 12 000), 266 (ꢀ ) 30 000); HRMS (FAB)
calcd for C12H10FN4O2 (M + H) 261.0788, found 261.0790.
8-Chloro-7-fluoro-10-dimethylisoalloxazine (21): decomposed
1
without melting at 274-276 °C; H NMR (CF3CO2D) δ 9.08 (d, J )
6.0 Hz, 1H), 8.31 (d, J ) 9.4 Hz, 1H), 4.76 (s, 3H); 13C NMR (CF3-
CO2D) δ 161.3 (d, JC-F ) 255 Hz) 160.7, 152.1 (d, JC-F ) 21 Hz),
146.0, 143.0 (d, JC-F ) 14 Hz), 139.7 (d, JC-F ) 18 Hz), 137.7, 131.5,
121.8, 120.0 (JC-F ) 24 Hz), 39.0; UV (0.1 M NaOH) λ 448 (ꢀ )
4800), 336 (ꢀ ) 4400), 268 (ꢀ ) 15 700).
7-Methoxy-10-methylisoalloxazine (22): mp 331-334 °C (dec);
1H NMR (CF3CO2D) δ 8.36 (d, J ) 9.7 Hz, 1H), 8.18 (dd, J ) 9.7,
2.7, 1H), 7.90 (d, J ) 2.7 Hz, 1H), 4.67 (s, 3H), 4.18 (s, 3H); UV (0.1
M NaOH) λ 467 (ꢀ ) 7100), 339 (ꢀ ) 5700), 277 (ꢀ ) 19 700); HRMS
(FAB) calcd for C12H11N4O3 (M + H) 259.0831, found 259.0830.
8-Cyano-10-methylisoalloxazine (24): prepared as previously de-
scribed by Bruice;14 mp 335-9 °C (lit.14 mp 341-342 (dec)); 1H NMR
(CF3CO2D) δ 8.79 (d, J ) 0.9 Hz, 1H), 8.70 (d, J ) 8.6 Hz, 1H), 8.33
(dd, J ) 8.6, 0.9 Hz, 1 H), 4.60 (s, 3H); UV (0.1 M NaOH) λ 445 (ꢀ
) 8000), 328 (ꢀ ) 8300), 278 (ꢀ ) 20 800).
7-Cyano-10-methylisoalloxazine (25): prepared as previously de-
scribed by Bruice;14 mp 368-70 °C (lit.14 mp 360-362 °C (dec)); 1H
NMR (CF3CO2D) δ 8.91 (d, J ) 1.7 Hz, 1H), 8.53 (dd, J ) 9.2, 1.7
Hz, 1H), 8.42 (d, J ) 9.2 Hz, 1H), 4.55 (s, 3H); 13C NMR (CF3CO2D)
δ 161.8; 149.8, 141.8, 140.5, 139.6, 138.8, 137.5, 121.0, 116.3, 115.5,
37.3; UV (0.1 M NaOH) λ 433 (ꢀ ) 8100), 335 (ꢀ ) 7000), 281 (ꢀ )
35 700).
Acknowledgment. Acknowledgment is made to the donors
of the Petroleum Research Fund, administered by the American
Chemical Society, and the Vanderbilt University Research
Council for support of this work. J.J.H. acknowledges support
from the Vanderbilt Undergraduate Summer Research Program.
JA972992N
(20) (a) Yoneda, F.; Shinozuka, K.; Hiromatsu, K.; Matsushita, R.;
Sakuma, Y.; Hamana, M. Chem. Pharm. Bull. 1980, 28, 3576. (b) King, F.
E.; Acheson, R. M.; Yorke-Long, A. B. J. Chem. Soc. 1948, 1926.