A flavin analogue with improved solubility in organic solvents

Article abstract of DOI:10.1016/j.tetlet.2007.05.165

We report the synthesis and initial electrochemical characterization of a benzene-soluble flavin analogue: N(10)-2,2-dibenzylethyl-7,8-dimethylisoalloxazine (DBF, 1). This analogue, which has an unmodified flavin headgroup, is intended for use in the spectroscopic examination of the electronic effects of flavin hydrogen bonding in simple model systems in aprotic, non-hydrogen bonding solvents. With future spectroscopic studies in mind, we have developed a synthetic route, which allows the incorporation of isotopic labels using inexpensive starting materials.

Full text of DOI:10.1016/j.tetlet.2007.05.165

Tetrahedron Letters 48 (2007) 5517–5520
A flavin analogue with improved solubility in organic solvents
a
a
Ronald L. Koder,a,b, Bruce R. Lichtenstein, Jose F. Cerda,
*
Anne-Frances Millerc and P. Leslie Duttona
aThe Johnson Research Foundation and Department of Biochemistry and Biophysics, The University of Pennsylvania,
Philadelphia, A 19104, United States
bDepartment of Physics, The City College of New York, New York, NY 10031, United States
cThe University of Kentucky, Department of Chemistry, Lexington, KY 40506-0055, United States
Received 24 April 2007; revised 28 May 2007; accepted 29 May 2007
Available online 2 June 2007
Abstract—We report the synthesis and initial electrochemical characterization of a benzene-soluble flavin analogue: N(10)-2,2-di-
benzylethyl-7,8-dimethylisoalloxazine (DBF, 1). This analogue, which has an unmodified flavin headgroup, is intended for use in
the spectroscopic examination of the electronic effects of flavin hydrogen bonding in simple model systems in aprotic, non-hydrogen
bonding solvents. With future spectroscopic studies in mind, we have developed a synthetic route, which allows the incorporation of
isotopic labels using inexpensive starting materials.
Ó 2007 Elsevier Ltd. All rights reserved.
Flavins are versatile enzyme prosthetic groups or cofac-
tors essential to enzymes found in all five kingdoms of
life, where they catalyze a remarkably diverse chemistry
that draws on its rich oxidation–reduction chemistry
and strong coupling to proton release and uptake. This
includes their role in light signaling and transcription
activation, in light activated DNA repair, in respiratory
electron chains, and in dehydrogenation, dehalogen-
ation, hydroxylation and oxygenation reactions.1 It is
clear that for each enzyme the specific interactions
between the cofactor and the protein control both the
type and driving force behind specific types of flavin
chemistries. Thus, flavoenzymes vary widely with
respect to their one- and two-electron midpoint poten-
tials and associated pKa values in the multiplicity of
states that describe the redox-coupled acid–base chemis-
try of the flavin cofactor.1–4
teins,5–9 but global analysis of these data supports only
qualitative conclusions due to the complexity of the pro-
tein environment and the attendant difficulty in decon-
structing the simultaneous effects of the large number
of flavin–protein interactions.
The Rotello group and others10–14 have examined flavin
oxidation–reduction in simplifying aprotic solvents.
This work demonstrated that, in model hydrogen bond-
ing complexes between a chloroform-soluble flavin
analogue and mono- and diamidopyridines, the one-
electron reduction potentials (from neutral oxidized to
semiquinone radical anion) of the flavin are modulated
by over 4 kcal/mol. These hydrogen bonded complexes
offer the opportunity to examine, in the atomic detail
we are currently lacking, the changes brought in the
electronic structure of the flavin by individual flavin–
cosolute interactions previously connoted only indi-
rectly by the aforementioned changes in the flavin reduc-
tion potentials. Examination of these model systems
promises significant new insights into the manner in
which flavoproteins direct flavin reactivity.
Despite many years of research, this area remains poorly
understood. Powerful spectroscopic methods such as
infrared spectroscopy, resonance Raman (RR) spectro-
scopy, Stark spectroscopy and solution nuclear magnetic
resonance (NMR) spectroscopy have all been utilized in
the examination of both free flavins (i.e., FMN and
FAD in aqueous buffers) and a number of flavopro-
The flavin analogue used in Rotello’s pioneering exper-
iments bears an isobutyl side chain at the N(10) position
of the isoalloxazine moiety to increase the solubility in
organic solvents (see Fig. 1). The oxidized form was
1
examined by H and 13C solution NMR, and its reduc-
Keywords: Flavin model; Isotopic labeling; Electrochemistry.
*
Corresponding author. E-mail: koder@sci.ccny.cuny.edu
tion to the semiquinone was studied by cyclic voltammetry
0040-4039/$ - see front matter Ó 2007 Elsevier Ltd. All rights reserved.
doi:10.1016/j.tetlet.2007.05.165

5518
R. L. Koder et al. / Tetrahedron Letters 48 (2007) 5517–5520
riboflavin.22 While this synthesis was performed in aque-
R
ous acid, which would not be amenable to the synthesis
of more hydrophobic flavin analogues, the method does
have the advantage that the N(5) nitrogen can be readily
and inexpensively labeled using 15N-sodium nitrite, and
the uracil ring can be inexpensively labeled at many
positions by synthesizing barbituric acid from urea
and diethyl malonate.23 Thus, every atom predicted to
exhibit large changes in chemical shifts or vibrational
frequencies24 can be readily labeled using our strategy,
with the exception of the N(10) nitrogen.
10a
4a
9a
5a
N
N
O
H
9
6
10
1
2
8
7
3 N
5
4
N
O
Riboflavin
FMN
R = CH2(CHOH)3CH2OH
R = CH2(CHOH)3CH2OPO3-
FAD
R = CH2(CHOH)3CH2(OPO3)2-Adenine
R = CH2(CHOAc)3CH2OAc
R = CH2CH(CH2CH3)2
TARF
11
Rotello
Our approach is outlined in Scheme 1 (synthetic details,
including product characterization, are included in the
Supplementary data). 2,2-Dibenzyl-diethylmalonate
(2), synthesized using the method of Maslak,25 was
saponified with KOH and decarboxylated by heating
to form 2,2-dibenzylacetic acid 3.27 Coupling of 3 with
dimethylaniline using N-(3-dimethylaminopropyl)-N0-
ethylcarbodiimide hydrochloride and dimethylamino-
pyridine followed by LiAlH4 reduction28 yielded second-
ary amine 5, which was diazotinated with 15N-sodium
nitrite and p-toluidine using the method of Tishler,22
substituting THF for water and performing the reaction
at 0 °C to maximize the formation of the meta-substi-
tuted kinetic product. The resultant compound, 6, was
condensed with barbituric acid 7 in refluxing glacial ace-
tic acid for 16 h forming N(10)-(2,2-dibenzylpropyl) iso-
alloxazine (DBF) 1 in 69% yield. The moderate yield of
this final step is compensated for by the fact that this is
the step in which the positions most likely to bear iso-
topic labels, those on the uracil ring of the dimethyliso-
alloxazine, are attached. This greatly reduces label loss
in comparison to other possible approaches in which
there are several synthetic steps subsequent to uracil ring
formation, each of which would entail a geometrically
growing loss of expensive labeled material.
Figure 1. The flavin moiety and some common natural and synthetic
derivatives.
(CV), in conjunction with electron paramagnetic reso-
nance (EPR).4 The oxidized form of this molecule was
found to be soluble to <1 mM in chloroform, less so
in methylene chloride and almost insoluble in benzene
(J. D. Walsh and A.-F. Miller, unpublished observa-
tions). Moreover, the reduced anionic forms of the flavin
are considerably less soluble than the oxidized ones in
any of these solvents. Sub-millimolar concentrations
are insufficient for many forms of spectroscopic analysis.
Additionally, chloroform, the solvent in which the
magnetic resonance studies have by necessity been
performed, is a solvent which possesses considerable
hydrogen bond-donating ability.15 Hydrogen bonding
to solvents has been shown to significantly affect 15N
chemical shifts in other cases.16 Pyridine, for example,
has been demonstrated to exhibit a difference in chemi-
cal shift between vacuum and chloroform of over
30 ppm.17 As an ideal model system should examine
the effects of single or multiple hydrogen bonds in isola-
tion, nonpolar solvents such as benzene, toluene, carbon
tetrachloride or cyclohexane must be used. Therefore,
the non-redox-active N(10) side chain must be highly
soluble in these solvents.
DBF is soluble to 1.5 mM in benzene and 20 mM in
methylene chloride as determined optically. This is a
15-fold improvement over the benzene solubility of tetra-
acetylriboflavin (TARF), the molecule most often used
in spectroscopic investigations of free flavins.6 Thus,
DBF is sufficiently soluble for solution NMR, IR, RR
and Stark spectroscopies. Importantly, DBF lacks the
multiple ester functionalities of TARF, so it is less likely
to form intermolecular hydrogen bonds to the alloxazine
N(3) proton.26 DBF’s solubility in nonpolar solvents has
already proven useful in the generation of dry, pow-
dered, reduced flavin samples for solid state NMR
spectroscopy.18
It will prove informative to expand the examination of
these complexes to include 15N NMR (both in solution
and in the solid state18), IR, RR and Stark spectro-
scopies. In order to circumvent solubility problems, we
have undertaken the synthesis of novel flavin analogues
with improved solubility in aprotic, non-hydrogen bond-
ing solvents. We have chosen benzene as a model solvent
because its relatively high melting temperature is ideal
for solid state NMR spectroscopy on frozen flavin solu-
tions, although there remains a possibility that flavin
energetics in this solvent will be perturbed by p stacking
effects.19 As each of NMR, infrared and raman spectro-
scopy necessitates the synthesis of multiple flavins, each
isotopically labeled at different positions, our intent
was to find a synthetic approach based on the most inex-
pensive isotopically labeled starting materials.
As expected, DBF is redox-active as demonstrated by
CV, displaying a reversible single-electron oxidation/
reduction transition between neutral oxidized and the
anionic semiquinone radical at ꢀ1287 mV versus ferro-
cene (see Fig. 2). We observe a reduced peak area in
the reduction half-wave, presumably due to some degree
of protonation of reduced flavin anions by oxidized
flavins in the bulk solution.4 To our knowledge, this is
the first report of flavin electrochemistry in benzene.
Spectroscopic examination of DBF and its complexes
is currently underway.
Some high-yield synthetic approaches to flavin ana-
logues have been reported,20,21 but each has the disad-
vantage of expensive and/or synthetically challenging
isotopically labeled precursors. In 1947, Tishler et al.
reported the condensation of 1-(D-ribitylamino)-2-aryl-
azo-4,5-dimethylbenzene and barbituric acid to form

R. L. Koder et al. / Tetrahedron Letters 48 (2007) 5517–5520
5519
H
O
OH
N
O
O
4
5
O
O
1) KOH
2) Δ
EDC, DMAP
O
+
LiAlH3
3
H
N
NH2
6
p-toluidine, NaNO2,
HOAc, THF, 0 o
C
H
N
O
N
N
N
O
HOAc, reflux
NH
+
NH
N
N
7
O
HN
O
O
1
8
Scheme 1.
Supplementary data
Supplementary data associated with this article can be
found, in the online version, at doi:10.1016/j.tetlet.
2007.05.165.
References and notes
1. Hemmerich, P.; Massey, V. In Oxidations and Related
Redox Systems; Pergamon Press: Oxford, 1982; pp 379–
403.
2. Schopfer, L. M.; Ludwig, M. L.; Massey, V. In Flavins
and Flavoproteins; Curti, B., Ronchi, S., Zanetti, G., Eds.;
Walter de Gruyter: New York, 1990; p 990.
3. Koder, R. L.; Haynes, C. A.; Rodgers, M. E.; Rodgers, D.
W.; Miller, A. F. Biochemistry 2002, 41, 14197–14205.
4. Niemz, A.; Imbriglio, J.; Rotello, V. M. J. J. Am. Chem.
Soc. 1997, 119, 887–892.
Figure 2. Cyclic voltammogram of 37 lM DBF in benzene with 0.2 M
tetrahexylammonium perchlorate as a supporting electrolyte. A 1.6-
mm diameter platinum electrode was used for the measurements. The
voltammogram was recorded at a scan rate of 100 mV/s. The potential
is internally referenced versus ferrocene and the positive peak denotes
reduction of the flavin analogue.
5. Stanley, R. J. Antioxidants Redox Signal 2001, 3, 847–
866.
6. Muller, F. In Chemistry and Biochemistry of Flavoenzymes;
Muller, F., Ed.; CRC Press: Boca Raton, FL, 1991; pp
2–72.
7. Muller, F. In Chemistry and Biochemistry of Flavoenzymes;
Muller, F., Ed.; CRC Press: Boca Raton, Fl, 1991; pp
558–595.
8. Wille, G.; Ritter, M.; Friedemann, R.; Mantele, W.;
Hubner, G. Biochemistry 2003, 42, 14814–14821.
9. Altose, M. D.; Zheng, Y. G.; Dong, J.; Palfey, B. A.;
Carey, P. R. Proc. Natl. Acad. Sci. U.S.A. 2001, 9, 3006–
3011.
10. Breinlinger, E.; Niemz, A.; Rotello, V. M. J. Am. Chem.
Soc. 1995, 117, 5379–5380.
11. Deans, R.; Cooke, G.; Rotello, V. M. J. Org. Chem. 1997,
62, 836–839.
12. Cuello, A. O.; McIntosh, C. M.; Rotello, V. M. J. Am.
Chem. Soc. 2000, 122, 3517–3521.
Acknowledgements
The authors would like to thank Dr. Colin Wraight of
the Department of Biochemistry, the University of Illi-
nois at Urbana-Champaigne for the kind donation of
a sample of 2,2-dibenzylacetic acid to be used as a syn-
thetic standard. This work was supported by DOE
Grant DE-FG02-05ER46223 (P.L.D.), NIH grants
GM41048 (P.L.D.), GM063921 (A.F.M.), NSF-
MRSEC Grant DMR05-20020 (P.L.D.) and the
Petroleum Research Fund via PRF #44321-AC4
(A.F.M.). R.L.K. was an NIH Postdoctoral Fellow
(GM64090).

5520
R. L. Koder et al. / Tetrahedron Letters 48 (2007) 5517–5520
13. Goodman, A. J.; Breinlinger, E. C.; McIntosh, C. M.;
Grimaldi, L. N.; Rotello, V. M. Org. Lett. 2001, 3, 1531–
1534.
14. Hasford, J. J.; Kemnitzer, W.; Rizzo, C. J. J. Org. Chem.
1997, 62, 5244–5245.
21. Epple, R.; Wallenborn, E. U.; Carell, T. J. Am. Chem.
Soc. 1997, 119, 7440–7451.
22. Tishler, M.; Pfister, K.; Babson, R. D.; Ladenburg, K.;
Fleming, A. J. J. Am. Chem. Soc. 1947, 69, 1487–1492.
23. Murray, J. I. In Organic Synthesis, Collected Volume 4;
John Wiley & Sons: New York, NY, 1963; pp 744–
746.
15. Kamlet, M. J.; Taft, R. W. J. Chem. Soc., Perkin Trans. 2
1979, 349–356.
16. Kamlet, M. J.; Dickinson, C.; Taft, R. W. J. Chem. Soc.,
Perkin Trans. 2 1981, 353–355.
24. Walsh, J. D.; Miller, A. F. J. Phys. Chem. B 2003, 107,
854–863.
17. Witanowski, M.; Sicinska, W.; Biernat, S.; Webb, G. A.
J. Magn. Reson. 1991, 91, 289–300.
25. Maslak, P.; Varadarajan, S.; Burkey, J. D. J. Org. Chem.
1999, 64, 8201–8209.
18. Koder, R. L.; Walsh, J. D.; Pometum, M. S.; Dutton, P.
L.; Wittebort, R. J.; Miller, A. F. J. Am. Chem. Soc. 2006,
128, 15200–15208.
19. Breinlinger, E. C.; Rotello, V. M. J. Am. Chem. Soc. 1997,
119, 1165–1166.
26. Stanley, R. J.; Siddiqui, M. S. J. Phys. Chem. A 2001, 105,
11001–11008.
27. Euler, H. V.; Karrer, P.; Malmsberg, M.; Scho¨pp, F.;
Benz, F.; Becker, B.; Frei, P. Helv. Chim. Acta 1935, 18,
522–535.
20. Yoneda, F.; Sakuma, Y.; Ichiba, M.; Shinomura, K. J.
Am. Chem. Soc. 1976, 98, 830–835.
28. Bojic, U.; Elmazar, M. M. A.; Hauck, R. S.; Nau, H.
Chem. Res. Toxicol. 1996, 9, 866–870.