Atropisomeric flavoenzyme models with a modified pyrimidine ring: Syntheses, physical properties, and stereochemistry, in the reactions with NAD(P)H analogs

Article abstract of DOI:10.1021/jo961799t

Optically active 5-deazaflavin derivatives (3-aryl-10-(4-tert-butylphenyl)pyrimido[4,5-b]quinoline-2,4(3H,10H)-di one) with an axial chirality at the pyrimidine ring have been synthesized, and the kinetics of enantiomerization have been measured for some

Full text of DOI:10.1021/jo961799t

9344
J . Org. Chem. 1996, 61, 9344-9355
Atr op isom er ic F la voen zym e Mod els w ith a Mod ified P yr im id in e
Rin g: Syn th eses, P h ysica l P r op er ties, a n d Ster eoch em istr y in th e
Rea ction s w ith NAD(P )H An a logs
Atsuyoshi Ohno,*,† J un Kunitomo,† Yasushi Kawai,† Tetsuji Kawamoto,‡,1
Masaki Tomishima,‡,2 and Fumio Yoneda‡,3
Institute for Chemical Research, Kyoto University, Uji, Kyoto 611, J apan, and Faculty of Pharmaceutical
Sciences, Kyoto University, Sakyo-ku, Kyoto 606-01, J apan
Received September 20, 1996X
Optically active 5-deazaflavin derivatives (3-aryl-10-(4-tert-butylphenyl)pyrimido[4,5-b]quinoline-
2,4(3H,10H)-dione) with an axial chirality at the pyrimidine ring have been synthesized, and the
kinetics of enantiomerization have been measured for some of them. The absolute configurations
of these compounds have been determined by X-ray crystallographic analysis and chemical reactions
for the first time in atropisomeric flavoenzyme models. Enantioface-differentiating (net) hydride-
transfer reactions with 1-benzyl-1,4-dihydronicotinamide (BNAH) have revealed that the selectivity
of the reacting face of the 3-[2-(hydroxymethyl)phenyl] derivative 1 changes depending on the
presence or absence of Mg2+; the hydroxymethyl group of 1 exerts steric inhibition in the absence
of Mg2+, whereas it facilitates the approach of BNAH in the presence of Mg2+. Asymmetric (net)
hydride-transfer reactions with chiral 1,4-dihydro-2,4-dimethyl-N-(R-methylbenzyl)-1-propylnico-
tinamide (Me2PNPH) predict that the most favorable intermolecular arrangement of these two
molecules at the transition state is the one in which the pyrimidine ring of 1 and the carbamoyl
group of Me2PNPH tend to face each other and the maximum overlap of their molecular planes is
achieved regardless of the presence or absence of Mg2+. The arrangement mimics that of FAD and
NADPH in the active site of a flavoenzyme. The present result indicates an energetically favorable
overlap of the molecular planes of a flavin and an NAD(P)H coenzyme, as well as a significant
influence of functional groups from an apoenzyme in proximity to a flavin coenzyme on the
stereochemistry of biological redox reactions.
In tr od u ction
investigated widely. For example, Shinkai et al.15-19
synthesized optically active flaviophanes and 5-deazafla-
vinophanes, which oxidized optically active thiols and
NAD(P)H analogs in an asymmetric manner. Yoneda et
al.20-22 synthesized an optically active dFl with axial and
planar chirality, where one face of the molecule was
blocked by an alkyl chain at the N(10) position, and
determined its absolute configuration.22 It has been
Flavoenzymes require flavin mononucleotide (FMN) or
flavin adenine dinucleotide (FAD) as a coenzyme and
catalyze oxidation-reduction reactions in biological sys-
tems, in which the flavin group is the oxidizing or
reducing agent.4-7 At the active sites of flavoenzymes,
flavin coenzymes are covalently bound or tightly held to
apoproteins to form chiral environments, and the reactiv-
ity of a flavin coenzyme and stereoselectivity in the
reactions with substrates are markedly enhanced.8-14
To clarify the reaction mechanism of flavoenzymes,
optically active flavins (Fl) and 5-deazaflavins (dFl) with
a function as apoproteins have been synthesized and
suggested that a ternary complex composed of dFl, Mg2+
,
and NAD(P)H analog is involved in the mimetic inter-
coenzyme reactions. These flavoenzyme models oxidized
NAD(P)H analogs asymmetrically, revealing that asym-
metry in the face of the flavin significantly influences the
stereochemistry of (net) hydride-transfer reactions with
an NAD(P)H analog.
† Institute for Chemical Research, Kyoto University.
‡ Faculty of Pharmaceutical Sciences, Kyoto University.
X Abstract published in Advance ACS Abstracts, December 1, 1996.
(1) Present address: Takeda Chemical Industries, Ltd., Yodogawa-
ku, Osaka 532, J apan.
Recently, Yoneda et al.23-25 synthesized a series of dFls
with an axial chirality at the N(3) position and observed
diastereofacial deactivation through steric hindrance23,24
(2) Present address: Fujisawa Pharmaceutical Co., Ltd., Yodogawa-
ku, Osaka 532, J apan.
(3) Present address: Fujimoto Pharmaceutical Co., Ltd., Matsubara,
Osaka 580, J apan.
(15) Shinkai, S.; Yamaguchi, T.; Nakao, H.; Manabe, O. Tetrahedron
Lett. 1986, 27, 1611-1614.
(16) Shinkai, S.; Yamaguchi, T.; Kawase, A.; Kitamura, A.; Manabe,
O. J . Chem. Soc., Chem. Commun. 1987, 1506-1508.
(17) Shinkai, S.; Kawase, A.; Yamaguchi, T.; Manabe, O. J . Chem.
Soc., Chem. Commun. 1988, 457-458.
(4) Walsh, C. Enzymatic Reaction Mechanism; W. H. Freeman: San
Francisco, 1979; pp 358-431.
(5) Massey, V.; Hemmerich, P. Biochem. Soc. Trans. 1980, 8, 246-
(18) Shinkai, S.; Kawase, A.; Yamaguchi, T.; Manabe, O.; Wada, Y.;
Yoneda, F.; Ohta, Y.; Nishimoto, K. J . Am. Chem. Soc. 1989, 111,
4928-4935.
257.
(6) Walsh, C. Acc. Chem. Res. 1980, 13, 148-155.
(7) Bruice, T. C. Acc. Chem. Res. 1980, 13, 256-262.
(8) Fisher, J .; Walsh, C. J . Am. Chem. Soc. 1974, 96, 4345-4346.
(9) Yamazaki, S.; Tsai, L.; Stadtman, T. C.; J acobson, F. S.; Walsh,
C. J . Biol. Chem. 1980, 255, 9025-9027.
(19) Shinkai, S.; Yamaguchi, T.; Kawase, A.; Manabe, O.; Kellogg,
R. M. J . Am. Chem. Soc. 1989, 111, 4935-4940.
(20) Kawamoto, T.; Tanaka, K.; Yoneda, F.; Hayami, J . Tetrahedron
Lett. 1989, 30, 7431-7434.
(10) Pai, E. F.; Schulz, G. E. J . Biol. Chem. 1983, 258, 1752-1757.
(11) J acobson, F.; Walsh, C. Biochemistry 1984, 23, 979-988.
(12) Yamazaki, S.; Tsai, L.; Stadtman, T. C.; Teshima, T.; Nakaji,
A.; Shiba, T. Proc. Natl. Acad. Sci. U.S.A. 1985, 82, 1364-1366.
(13) Manstein, D. J .; Pai, E. F.; Schopfer, L. M.; Massey, V.
Biochemistry 1986, 25, 6807-6816.
(21) Kawamoto, T.; Tanaka, K.; Kuroda, Y.; Yoneda, F. Chem. Lett.
1990, 1197-1200.
(22) Kawamoto, T.; Taga, T.; Bessho, K.; Yoneda, F.; Hayami, J .
Tetrahedron Lett. 1994, 35, 8631-8634.
(23) Kawamoto, T.; Tomishima, M.; Yoneda, F.; Hayami, J . Tetra-
hedron Lett. 1992, 33, 3169-3172.
(14) Manstein, D. J .; Massey, V.; Ghisla, S.; Pai, E. F. Biochemistry
1988, 27, 2300-2305.
(24) Kawamoto, T.; Tomishima, M.; Yoneda, F.; Hayami, J . Tetra-
hedron Lett. 1992, 33, 3173-3176.
S0022-3263(96)01799-9 CCC: $12.00 © 1996 American Chemical Society

Atropisomeric Flavoenzyme Models
J . Org. Chem., Vol. 61, No. 26, 1996 9345
Sch em e 127
Ta ble 1. Sp ecific Rota tion s of th e En a n tiom er s of 1,
2, a n d 19a
[R]D (abs config)
dFl (c)
T, °C
1st fraction
2nd fraction
1 (1.00)
2 (1.01)
19 (1.00)
16
25
23
+59.1° (S)
+25.1° (R)
+7.5° (R)
-57.3° (R)
-25.9° (S)
-8.1° (S)
a
All dFls were obtained in >99% ee.
Ta ble 2. Ra te Con sta n ts for En a n tiom er iza tion (kr ot.) of
1 a n d 2 in DMF
as well as diastereofacial activation through intramo-
lecular acid catalysis.25 The results suggest that there
is diastereoselective interaction between the pyrimidine
ring of a dFl and the carbamoyl group of an NAD(P)H
analog.
krot. × 106, s-1
T, °C
1
2
30
40
50
60
70
1.12 ( 0.01
4.37 ( 0.06
15.1 ( 0.2
46.7 ( 0.7
152 ( 3
1.07 ( 0.01
4.08 ( 0.04
14.2 ( 0.1
55.1 ( 0.6
170 ( 3
Here, we designed and synthesized an optically active
dFl with a 2-(hydroxymethyl)phenyl group at the N(3)
position, 1, as a novel flavoenzyme model (Scheme 1).26
The hydroxymethyl group that occupies one face of the
pyrimidine ring has the ability to coordinate onto a metal
ion through an electronegative oxygen atom that is
absent in an alkyl group. The hydroxymethyl group is
expected to mimic a potential functional group of an
apoprotein in proximity to a flavin coenzyme. By com-
paring the physical properties and stereochemical reac-
tivity of 1 with those of analogous dFl derivatives,
especially those of 2, which has a methyl group instead
of a hydroxymethyl group, we will discuss the effects of
the hydroxymethyl group on the stereochemistry in the
reactions with NAD(P)H analogs. The intermolecular
arrangement between a flavin model and an NAD(P)H
analog at the transition state of (net) hydride-transfer
reaction will also be discussed.
Ta ble 3. Activa tion P a r a m eter s for En a n tiom er iza tion
of 1 a n d 2
activation param
1
2
Ea, kcal mol-1
25.0 ( 0.2
25.9 ( 0.02
24.4 ( 0.2
26.3 ( 0.4
26.0 ( 0.03
25.6 ( 0.4
∆Gq, kcal mol-1
∆Hq, kcal mol-1
∆Sq, cal mol-1 deg-1
-5.26 ( 0.63
-1.31 ( 1.17
omers of 1 and 2 were obtained as the first fractions,
respectively. The specific rotations of the enantiomers
of 1 and 2 are shown in Table 1.
Th er m a l En a n tiom er iza tion . In order to estimate
the conformational stability of 1 and 2, the rate constants
(krot.) and activation parameters for the rotation around
the N(3)-C(aryl) bond in these compounds have been
studied (Scheme 3). The rate constants are listed in
Table 2. The Arrhenius and Eyring plots have excellent
linear relationships (r > 0.9996). The activation param-
eters obtained from the plots are listed in Table 3. The
results indicate that the difference between a hydroxym-
ethyl group and a methyl group has little influence on
the energy barrier for rotation around the N(3)-C(aryl)
bond and confirm that the racemization of 1 and 2 can
be disregarded under the present reaction conditions
(within 3 h at 298 K).
Resu lts a n d Discu ssion
Syn th eses a n d Op tica l Resolu tion s of d F l Der iva -
tives. dFl derivatives were prepared according to the
procedure shown in Scheme 2. The synthesis of the dFl
skeleton was followed by the procedure reported by
Yoneda et al.28 5-Deuterated derivatives of dFl, 2-5-d and
8-5-d, were synthesized by the reactions of 7 and 13 with
2-fluoro[R-2H]benzaldehyde, respectively. The 5-deuter-
ated counterpart of 1, 1-5-d was prepared from 8-5-d
according to the same synthetic route as that shown in
Scheme 2. Since both the 5-deazaisoalloxazine ring and
the carboxyl group were reduced in the reduction of 8 to
X-r a y Cr ysta llogr a p h ic An a lyses a n d Deter m in a -
tion s of th e Absolu te Con figu r a tion s. Structures of
crystals of 1 and 2 obtained by recrystallization from
ethanol and toluene, respectively, have been revealed by
X-ray crystallography. The ORTEP drawings are il-
lustrated in Figures 1 and 2, and the crystallographic
data are summarized in Table 4. There is an intermo-
lecular hydrogen bond between the hydroxy group and
the carbonyl group at the C(2) position of an adjacent
molecule (Figure 1b).
1
r ed, sodium borodeuteride was employed for this reaction
instead of sodium borohydride in order to substitute two
deuterium atoms at the C(5) position of 1r ed
.
Although not only the C(5) position but also the
methylene part of the hydroxymethyl group in the
substituent at the N(3) position of 1-5-d is deuterated,
the deuteration may not affect stereochemical results of
the present reactions.
In order to determine the absolute configuration of 1
by X-ray crystallographic analysis, we first prepared
several chiral dFls from 1 (Scheme 4)31 and attempted
to confirm the absolute configurations of these enanti-
omers. However, since these dFls are conformationally
unstable or partially racemize before crystals suitable for
The optical resolutions of 1 and 2 were carried out by
HPLC on a chiral stationary phase.29,30 The (+)-enanti-
(25) Kawamoto, T.; Tomishima, M.; Kunitomo, J .; Yoneda, F.;
Hayami, J . Tetrahedron Lett. 1992, 33, 7173-7176.
(26) Ohno, A.; Kunitomo, J .; Kawamoto, T.; Tomishima, M.; Bessho,
K.; Yoneda, F. Tetrahedron Lett. 1994, 35, 9729-9732.
(27) Only one enantiomer is shown in the scheme for convenience.
(28) Nagamatsu, T.; Hashiguchi, Y.; Yoneda, F. J . Chem. Soc.,
Perkin Trans. 1 1984, 561-565.
(30) Conditions for preparative chromatography of 2: column,
CHIRALCEL OD (2 cm ø × 25 cm); eluent, ethanol; flow rate, 4.0 mL/
min; detection, UV 254 nm; retention times, 78.5 and 156 min:
conditions for analytical chromatography of 2: column, CHIRALCEL
OD (0.46 cm ø × 5 cm); eluent, ethanol; flow rate, 0.5 mL/min;
detection, UV 254 nm; retention times, 7.6 and 14.1 min.
(31) Structures and purities of all dFl derivatives were confirmed
by 1H NMR.
(29) Conditions for preparative chromatography of 1: column,
CHIRALCEL OD (2 cm ø × 25 cm); eluent, ethanol; flow rate, 3.5 mL/
min; detection, UV 254 nm; retention times, 55.1 and 83.0 min:
conditions for analytical chromatography of 1: column, CHIRALCEL
OD (0.46 cm ø × 5 cm); eluent, ethanol; flow rate, 0.2 mL/min;
detection, UV 254 nm; retention times, 11.3 and 16.7 min.

9346 J . Org. Chem., Vol. 61, No. 26, 1996
Ohno et al.
Sch em e 227
Sch em e 3
From the X-ray crystallographic analysis of (-)-19,
which was recrystallized from methanol, by means of
anomalous dispersion effect of the bromine atoms, it has
been confirmed that (-)-19 has the S configuration.33
This is the first successful determination of the absolute
configuration of an optically active Fl or dFl derivative
that has no chiral substituent in the molecule. The
circular dichroism (CD) spectra of the enantiomers of 19
in acetonitrile are shown in Figure 3.
Next, (S)-(-)-19 was debrominated by catalytic hydro-
genation, and the resultant 2 was subjected to HPLC
from which its conformation was determined to be (R)-
(+) (Scheme 6).
X-ray crystallographic analyses are obtained, we failed
to determine the absolute configuration of 1 by this
method.
Therefore, we designed 19, which was expected to
maintain its conformation for a long time even at room
temperature. The synthesis of 19 was carried out ac-
cording to Scheme 5, and the optical resolution was
accomplished by HPLC.32 The (+)-enantiomer was ob-
tained from the first fraction. The specific rotations of
the enantiomers of 19 are also listed in Table 1. It has
been confirmed that the conformation of 19 is stable for
over 1 month at room temperature.
(32) Conditions for preparative chromatography of 19: column,
CHIRALCEL OD (2 cm ø × 25 cm); eluent, ethanol; flow rate, 1.5 mL/
min; detection, UV 254 nm; retention times, 89.5 and 95.6 min:
conditions for analytical chromatography of 19: column, CHIRALCEL
OD (0.46 cm ø × 5 cm + 0.46 cm ø × 25 cm); eluent, ethanol; flow
rate, 0.5 mL/min; detection, UV 254 nm; retention times, 23.7 and 29.1
min. The second fraction was obtained in >99% ee by repetition of the
optical resolution.
(33) Kawai, Y.; Kunitomo, J .; Ohno, A. Acta Crystallogr., Sect. C,
in press.

Atropisomeric Flavoenzyme Models
J . Org. Chem., Vol. 61, No. 26, 1996 9347
F igu r e 2. ORTEP drawing of 2 with displacement ellipsoids
at 30% probability level. Only the (S)-(-)-enantiomer is
illustrated here.
Ta ble 4. Su m m a r y of Cr ysta llogr a p h ic Da ta a
dFl
formula
formula wt
cryst color, habit
cryst dimens, mm
cryst syst
space grp
a, Å
(()-1
(()-2
C28H25N3O3
451.52
C28H25N3O2
435.52
yellow, prismatic
0.85 × 0.20 × 0.10
orthorhombic
Pbca (#61)
22.340(6)
20.03(1)
10.534(5)
yellow, prismatic
0.64 × 0.20 × 0.20
monoclinic
P21/n (#14)
22.406(1)
13.200(4)
7.858(2)
94.31(1)
2317(1)
b, Å
c, Å
â, deg
V, Å3
4713(6)
8
Z value
4
D
calc, g/cm3
1.294
120.2
3918
3621
1.227
120.1
3967
3967
914
2θmax, deg
no. total reflns
no. unique reflns
no. observns
(I > 3.00σ(I))
no. variables
R
2047
408
298
0.042
0.043
1.64
0.061
0.072
1.96
Rw
S
a
Common to all structures: diffractometer, Rigaku AFC7R;
radiation, Cu KR (λ ) 1.541 78 Å); temperature, 20.0 °C.
Ta ble 5. En a n tiofa ce-Differ en tia tin g (Net)
Hyd r id e-Tr a n sfer Rea ction betw een Ra cem ic d F l-5-d
a n d BNAHa
F igu r e 1. ORTEP drawings of (a) 1 in the front view and (b)
two molecules of 1 linked by a hydrogen bond (indicated by a
solid line) with displacement ellipsoids at 50% probability level.
Only the (S)-(+)-enantiomer is illustrated in both drawings.
catalyst
(equiv)b
reaction
time, min
ratio of reacting
dFl-5-d
faces, syn:antic
1
Mg(ClO4)2 (5)
CCl3CO2H (10)
Mg(ClO4)2 (5)
Mg(ClO4)2 (5)
Mg(ClO4)2 (5)
10
60
180
180
120
78:22
31:69
30:70
20:80
29:71
Finally, the (+)-enantiomer of 1 was converted into 2
according to the procedure shown in Scheme 6. The
resultant 2 was subjected to HPLC, which confirmed that
the compound was the (-)-enantiomer. Consequently,
the absolute configuration of (+)-1 has been assigned as
S.
2
3
8
[dFl-5-d] ) 4.0 × 10-2 M, [BNAH] ) 2.0 × 10-1 M. In the
a
dark under Ar at 298 K. Equivalency to [dFl-5-d]. c Estimated
b
En a n tiofa ce-Differ en tia tin g (Net) Hyd r id e-Tr a n s-
fer Rea ction s. In order to investigate the selectivity of
the faces in which a (net) hydride is transferred, reduc-
tions of various dFl-5-ds with 1-benzyl-1,4-dihydronico-
tinamide (BNAH) were studied (Scheme 7). The reac-
tions were carried out in acetonitrile at 298 K in the dark
under argon atmosphere and monitored by TLC. Syn/
anti ratios of the reacting faces were determined by H
NMR spectroscopy based on the area ratio of two dia-
stereotopic protons at the C(5) position of the reduced
errors are within (2 for all observed values.
dFl-5-d thus obtained. Syn and anti hydrogen atoms
were assigned in a similar manner as described previ-
ously.24,25 The results are summarized in Table 5.
In order to distinguish diastereomeric hydrogens at the
C(5) position, either the dFl derivative or an NAD(P)H
analog must be substituted by deuteration at the C(5) or
C(4) position, respectively. We employed dFls deuterated
at their C(5) positions (1-5-d, 2-5-d, 3-5-d, and 8-5-d) for
1

9348 J . Org. Chem., Vol. 61, No. 26, 1996
Ohno et al.
Sch em e 427
Sch em e 527
F igu r e 3. CD spectra of 19 in acetonitrile. The solid and
broken lines represent the spectra of (S)- and (R)-enantiomers,
respectively.
Mg2+. Since this type of reaction does not proceed with
a weak acid,25,34 a strong acid is required for this purpose.
Indeed, the reaction took place under the acid catalysis
of an excess amount of trichloroacetic acid. The result
obtained is also listed in Table 5. Although dichloroacetic
and chloroacetic acids are too weak to complete the
reaction and the reactions are associated by the acid-
catalyzed decomposition of BNAH, the syn/anti ratios in
the product obtained, 1r ed -5-d, after 60 min of the
reactions were the same, within experimental error, as
that listed in Table 5.35 The syn/anti selectivity in the
Brønsted acid-catalyzed reaction is reversed from that
observed under the Lewis acid (Mg2+)-catalyzed reaction
and the same as those observed with 2, 3, or 8 in the
presence of Mg2+. Thus, there remains no doubt that the
reversed stereoselectivity observed for 1 in the presence
of Mg2+ stems entirely from the interaction between 1
and Mg2+. The interaction does not exist in the reaction
with 2, 3, or 8.
this purpose, and these deuterated dFls were reduced by
an undeuterated NAD(P)H analog. However, for sim-
plicity, we will denote the deuterated dFls without the
term “-5-d” hereafter. The deuterated compounds are
indicated appropriately in Tables 5 and 8 and the
Experimental Section.
In the presence of Mg2+, the (net) hydride transfer from
BNAH to 2, 3, or 8 takes place predominantly in the anti
face, which is reasonable from the viewpoint of steric
interference by the substituent on the N(3) phenyl ring.
On the other hand, the selectivity observed in the
reaction of 1 is the opposite of that of 2, 3, or 8. It should
also be noted that the reaction of 1 proceeds much faster
than that of 2, 3, or 8.
In order to clarify the origin of this interaction, the
association constants (K) of 1 and 2 with Mg2+ in
anhydrous acetonitrile at 293 K were measured.36 Linear
plots shown in Figure 4 indicate that a 1:1 complex is
In the absence of the catalyst, however, the reaction
of 1 did not proceed at all, which reveals that the proton
in the hydroxy group does not play the role of an acid
catalyst or the proton in this hydroxy group is not
dissociated in contrast to the one in a hydroxynaphthyl
group.25 Therefore, we studied the (net) hydride-transfer
reaction in the presence of an acid catalyst in place of
(34) Shinkai, S.; Kawanabe, S.; Kawase, A.; Yamaguchi, T.; Manabe,
O. Bull. Chem. Soc. J pn. 1988, 61, 2095-2102.
(35) Although trifluoroacetic acid was also used as an acid catalyst,
syn/anti ratio was unable to be determined, because BNAH was
decomposed by the acid very quickly.
(36) The association constants of several dFl derivatives with Mg2+
were reported previously. Fukuzumi, S.; Kuroda, S.; Tanaka, T. Chem.
Lett. 1984, 417-420.

Atropisomeric Flavoenzyme Models
J . Org. Chem., Vol. 61, No. 26, 1996 9349
Sch em e 6
Sch em e 727
Ta ble 6. Associa tion Con sta n ts (K) of 1 a n d 2 w ith
Ma gn esiu m Ion a in Aceton itr ileb a t 293 K
dFl + Mg2+ y
\
Kz dFl‚Mg2+
dFl
R
K, M-1
1c
2d
CH2OH
CH3
668 ( 3
374 ( 2
a
b
Mg(ClO4)2 was used as the source of Mg2+
.
Water content in
acetonitrile used for this measurement was <0.030% v/v (1H
NMR). c [1] ) 5.11 × 10-4 M. [2] ) 5.96 × 10-4 M.
d
formed between dFl and Mg2+. As listed in Table 6, the
association constant of 1 is about twice as large as that
of 2. Although the factor of 2 of the association constant
observed here seems small, the value is acceptable
because the isoalloxazine ring has the ability to coordi-
nate onto a metal ion in an organic solvent, as exempli-
fied by 2,36 and the coordinating ability of an oxygen atom
in an undissociated hydroxy group (vide supra) might be
small similarly to that of an ethereal oxygen. Thus, on
the basis of the variation in enantioface differentiation
in the presence and absence of Mg2+, the factor of 2 can
F igu r e 4. Plots of (R-1 - 1)-1 against 103([Mg2+]0 - R[dFl]0).
Mg(ClO4)2 was used as the source of Mg2+. Correlation coef-
ficients are >0.9999 for both plots.
be accepted as the indication for the crucial role of the
hydroxymethyl group.

9350 J . Org. Chem., Vol. 61, No. 26, 1996
Ohno et al.
Ta ble 7. Dia ster eofa ce-Differ en tia tin g (Net)
Hyd r id e-Tr a n sfer Rea ction betw een Ra cem ic 1 a n d
Ch ir a l Me2P NP Ha
Ta ble 8. Asym m etr ic (Net) Hyd r id e-Tr a n sfer Rea ction
betw een Ch ir a l 1-5-d a n d Ch ir a l Me2P NP Ha
confign
confign of
reaction
ratio of reacting
confign of
Me2PNPH
equiv of
Mg(ClO4)2
ratio of 1
of 1-5-d Me2PNPH Mg(ClO4)2 time, min
faces, syn:antib
reacted S:Rc,d
b
S
4S,9S
4R,9R
4R,9R
4S,9S
yes
no
yes
no
yes
no
yes
no
<1
15
<1
7
<1
15
<1
7
62:38
49:51
44:56
29:71
60:40
48:52
45:55
30:70
4S,9S
0
10
50
0
10
50
47:53
62:38
74:26
52:48
36:64
28:72
4R,9R
R
[1] ) 1.0 × 10-2 M, [Me2PNPH] ) 5.0 × 10-4 M. In the dark
a
under Ar at 298 K. Equivalency to [Me2PNPH]. c Estimated
[1-5-d] ) 2.0 × 10-2 M, [Me2PNPH] ) 5.0 × 10-2 M,
b
a
errors are within (1 for all observed values. The ratio was kept
[Mg(ClO4)2]) 5.0 × 10-4 M. In the dark under Ar at 298 K.
d
b
constant meaningfully throughout the reaction.
Estimated errors are within (2 for all observed values.
Taking into account the fact that an NAD(P)H analog
mixture is not large enough to allow detailed discussion.
Even for the Mg2+-catalyzed reactions, about 10 times
excess Mg2+ is required for obtaining a result similar to
that obtained from the reaction with BNAH.
The observation stems not only from the fact that Me2-
PNPH exhibits lower selectivity due to its higher reactiv-
ity than BNAH but also from the fact that the reacting
face in BNAH is enantiotopic with respect to its molecular
plane, whereas that in Me2PNPH is diastereotopic and
only one of the faces can contribute to the (net) hydride-
transfer reaction.
can form a complex with Mg2+ 37
,
it is reasonable to
propose the formation of a ternary complex (1‚Mg2+
‚
BNAH)20,23,38 preceding the transition state of the reac-
tion. The contribution of the hydroxy oxygen for com-
plexation inevitably predicts that the syn face has more
chance than the anti face to form a complex, and then to
react, with BNAH under the Mg2+ catalysis. Thus, the
hydroxymethyl group of 1 has not only a catalytic
function in the presence of Mg2+ but also a stereocon-
trolling function, showing the change in stereochemistry
from that without Mg2+
.
Previously, we proposed, on the basis of the results of
enantioface-differentiating (net) hydride-transfer reac-
tions, that the pyrimidine ring of dFl and the carbamoyl
group of an NAD(P)H analog must face each other at the
The results predict that the hydroxymethyl group of 1
plays a significant role in coordinating onto Mg2+ to form
a ternary complex with BNAH rigidly and that stabiliza-
tion of the system through the formation of a complex in
the syn face is much more feasible than destabilization
through steric repulsion in this face.
On the other hand, the hydroxymethyl group in the
absence of Mg2+ is not different from other substituents
such as methyl, trifluoromethyl, and [(tert-butyldimeth-
ylsilyl)oxy]methyl groups in terms of interaction with
BNAH in the sense that it is nothing but a sterically
interfering group. Consequently, these substituents
result in the deactivative anti preference rather than a
syn face reaction.
transition state in the presence of Mg2+ 24
When the
.
same concept is applied to the present reaction between
(S)-1 and (4R,9R)-Me2PNPH, for example, the faces of
the two reacting molecules cannot overlap each other in
order to transfer the (net) hydride in the syn face of 1.
The unfavorable molecular arrangement (vide infra) may
reduce the relative reactivity of 1 in the syn face.
Consequently, the syn/anti selectivity of 1 becomes lower
in the reaction with Me2PNPH than that with BNAH.
Asym m etr ic (Net) Hyd r id e-Tr a n sfer Rea ction s.
Since the diastereoface-differentiating selectivity in the
reaction of 1 and Me2PNPH is low, especially in the
absence of Mg2+, there exists a limitation to the detailed
discussion of the intermolecular arrangement of 1 and
an NAD(P)H analog at the transition state when a
racemic mixture of 1 is employed for the reaction.
Therefore, we studied asymmetric (net) hydride-transfer
reactions between chiral 1 and chiral Me2PNPH. Each
enantiomer of 1, which was optically resolved by HPLC,
was reduced by a chiral Me2PNPH under conditions
similar to those described above, and the syn/anti selec-
tivity was measured. The results are summarized in
Table 8.
Dia st er eofa ce-Differ en t ia t in g (Net ) H yd r id e-
Tr a n sfer Rea ction s. To elucidate the intermolecular
arrangement between 1 and an NAD(P)H analog at the
transition state of (net) hydride-transfer reactions, a
diastereoface-differentiating reaction between 1 and a
chiral NAD(P)H analog, 1,4-dihydro-2,4-dimethyl-N-(R-
methylbenzyl)-1-propylnicotinamide (Me2PNPH),39 was
studied. Racemic 1 was reduced with (4S,9S)- or (4R,9R)-
Me2PNPH in acetonitrile at 298 K in the dark under
argon atmosphere, and 1r ed obtained after 30 min was
then oxidized to 1 with 2,3-dichloro-5,6-dicyano-1,4-
benzoquinone (DDQ). The oxidized product was sub-
jected to HPLC to measure the S/R ratio in 1. The results
are listed in Table 7.
In the presence of Mg2+, the reduction of 1 with Me2-
PNPH proceeds rapidly. It is noteworthy, however, that
the reduction of (S)-1 with (4R,9R)-Me2PNPH takes place
more than twice as fast as that with (4S,9S)-Me2PNPH
Although this reaction proceeds without Mg2+ because
the reactivity of Me2PNPH is much higher than that of
BNAH, the S/R ratio in 1 obtained from the reaction
in the absence of Mg2+
.
Here, we categorize the reduction of (S)-1, for example,
into four depending on the presence or absence of Mg2+
and whether the reducing agent is (4S,9S)- or (4R,9R)-
Me2PNPH, and each case will be discussed below.
(1) In the reaction with (4S,9S)-Me2PNPH in the
presence of Mg2+, two arrangements of Me2PNPH, a and
c shown in Figure 5a, are possible under the assumption
that the pyrimidine ring of dFl interacts with the
carbamoyl group of an NAD(P)H analog at the transition
state.24 It is plausible to expect that a (net) hydride
(37) It has been reported that the association constant between 1,4-
dihydro-N-(R-methylbenzyl)-1-propylnicotinamide, an analog of BNAH,
and Mg2+ in acetonitrile at 273 K is 1.9 × 103 M-1
: Ohno, A.;
Yamamoto, H.; Okamoto, T.; Oka, S.; Ohnishi, Y. Bull. Chem. Soc. J pn.
1977, 50, 2385-2386.
(38) For the contribution of a ternary complex in the reaction, see:
Ohno, A.; Yasui, S.; Nakamura, K.; Oka, S. Bull. Chem. Soc. J pn. 1978,
51, 290-293.
(39) It has been confirmed that the conformation at the side-chain
carbamoyl group does not affect stereochemistry of the reaction
largely: Ohno, A.; Ikeguchi, M.; Kimura, T.; Oka, S. J . Am. Chem.
Soc. 1979, 101, 7036-7040.

Atropisomeric Flavoenzyme Models
J . Org. Chem., Vol. 61, No. 26, 1996 9351
transfer in the syn face is favored over the one in the
anti face, because the coordination of the hydroxymethyl
group of 1 onto Mg2+ overwhelms the steric hindrance
in energy due to matched arrangement. However, the
syn/anti selectivity becomes lower than expected with this
combination because Me2PNPH exhibits such a high
reactivity in the presence of Mg2+ that the reaction
proceeds even in an unfavorable conformation as depicted
in c.
(2) In the reaction with (4S,9S)-Me2PNPH in the
absence of Mg2+, four arrangements of Me2PNPH, a-d
shown in Figure 5a, are possible. If the arrangements b
and d contributed mainly to the (net) hydride transfer,
the results observed in the reaction with (4S,9S)- and
(4R,9R)-Me2PNPH (vide infra) might be similar in con-
trast to the observed results. Instead, the arrangements
a and c, in which the pyrimidine ring of 1 and the
carbamoyl group of Me2PNPH face each other, are more
favorable than the arrangements b and d. In addition,
it is recognized that the syn attack becomes more
important in the reaction with Me2PNPH (syn/anti ) 49/
51) than that with BNAH (syn/anti ) 31/69) in spite of
the absence of coordination influence by Mg2+ in these
reaction systems. This result supports the presence of a
certain effect that makes the reaction in the syn face more
favorable than that in the anti face. Thus, it seems likely
that conformation a, where the maximum overlap of
molecular planes of 1 and Me2PNPH is achieved, is the
most favorable arrangement for the formation of a binary
complex (1‚Me2PNPH),40 and stabilization through this
conformation is large enough to compensate for energeti-
cally unfavorable steric hindrance.
(3) In the reaction with (4R,9R)-Me2PNPH in the
presence of Mg2+, two arrangements, a′ and c′ shown in
Figure 5b, are possible similarly to the system mentioned
in case 1. Here, however, the anti selectivity predomi-
nates slightly over the syn selectivity, or the syn selectiv-
ity is lowered in this combination by the presence of Mg2+
.
The result strongly suggests that large overlapping of
molecular planes of 1 and Me2PNPH is important in
determining the molecular arrangement at the transition
state of the reaction, and this effect overwhelms the
favorable electronic interaction through the coordination
by a hydroxy group.
(4) In the reaction with (4R,9R)-Me2PNPH in the
absence of Mg2+, four arrangements, a′-d′ shown in
Figure 5b, are possible. For the same reason as described
in case 2, arrangements a′ and c′ are more plausible than
the arrangements b′ and d′. Since the (net) hydride
transfer in the anti face is predicted to be more favorable
than the syn face in this reaction system from the
viewpoint of steric hindrance, together with the fact that
the reaction with (4R,9R)-Me2PNPH proceeds faster than
that with (4S,9S)-Me2PNPH, arrangement c′ is nomi-
nated as the most favorable arrangement at the transi-
tion state.
Thus, it is concluded that the most suitable intermo-
lecular arrangement between 1 and NAD(P)H analog at
the transition state of (net) hydride-transfer reactions is
the one in which two molecules are arranged with
maximum overlap of their molecular planes and the
pyrimidine ring of 1 is set in front of the carbamoyl group
of the analog, regardless of the presence or absence of
Mg2+ as portrayed in Figure 6.
F igu r e 5. Intermolecular arrangements between (a) (S)-(+)-1
and (4S,9S)-Me2PNPH and (b) (S)-(+)-1 and (4R,9R)-Me2-
PNPH. Magnesium ion is omitted from the drawing for
simplicity.
(40) Cf. Ohno, A.; Goto, T.; Nakai, J .; Oka, S. Bull. Chem. Soc. J pn.
1981, 54, 3478-3481.

9352 J . Org. Chem., Vol. 61, No. 26, 1996
Ohno et al.
Exp er im en ta l Section
In str u m en ts. Melting points (mp) were obtained using a
Yanagimoto micromelting point apparatus and were uncor-
rected. UV-vis spectra were recorded on a Hitachi U-3210
spectrophotometer equipped with a Hitachi SDR-30 temper-
ature controller. Infrared (IR) spectra were recorded on a
J ASCO FT/IR-5300 spectrometer. Proton nuclear magnetic
resonance (1H NMR) spectra were recorded on a Varian VXR
200 FT-NMR spectrometer with tetramethylsilane as an
internal reference, and all shifts are indicated in ppm. High-
resolution mass (HRMS) spectra were recorded on a J EOL
J MS-01SG-2 mass spectrometer. Elemental analyses were
performed using a Yanaco MT-3 elemental analyzer. Optical
rotations were measured on a J ASCO DIP-181 digital pola-
rimeter. Circular dichroism (CD) spectra were recorded on a
J ASCO J -720W spectropolarimeter. Column chromatography
was performed using Nacalai Tesque silica gel 60 (70-230
mesh). Preparative TLC was run on 20 cm × 20 cm plates
with a 0.5 mm layer of Nacalai Tesque silica gel 60 PF254
.
Ma ter ia ls. Organic reagents and solvents were purchased
from Wako Pure Chemical Industries, Ltd., Nacalai Tesque,
Inc., Tokyo Kasei Kogyo Co., Ltd., and Aldrich Chemical Co.,
Inc. Methanol and THF were dried by reflux and distillation
on sodium. Chloroform, dichloromethane, and DMF were
dried by reflux and distillation on calcium hydride.
1-(2-Methylphenyl)urea (10),44 2,4-dibromo-6-methylaniline
(14),45 1-benzyl-1,4-dihydronicotinamide (BNAH),46 and 1,4-
dihydro-2,4-dimethyl-N-(R-methylbenzyl)-1-propylnicotina-
mide (Me2PNPH)39 were synthesized as reported previously.
Their mp, IR, 1H NMR, and elemental analysis data were
consistent with the structures. 2-Fluoro[R-2H]benzaldehyde
was prepared by the oxidation of 2-fluoro[R,R-2H2]benzyl
alcohol with pyridinium chlorochromate in dry dichloromethane.
2-Fluoro[R,R-2H2]benzyl alcohol was prepared by the reduction
of ethyl 2-fluorobenzoate with lithium aluminum deuteride (98
atom % D) in dry ether. Both deuterated compounds exhibited
F igu r e 6. Most predominant intermolecular arrangements
between (S)-(+)-1 and an NAD(P)H analog at the transition
states of (net) hydride-transfer reactions in the presence (syn
face) and absence (anti face) of magnesium ion, respectively.
The conformation of the side-chain carbamoyl group of NAD-
(P)H analog is drawn arbitrarily.
It should be noted that the hydroxymethyl group of 1
controls both reactivity and stereochemistry in the (net)
hydride-transfer reaction; in the presence of Mg2+, the
(net) hydride transfer takes place predominantly in the
syn face of 1 due to the coordination of the hydroxy group
onto Mg2+ to form a ternary complex, whereas in its
absence, the reaction proceeds predominantly in the anti
face because of the predominance of steric interference
by this substituent.
1
IR, H NMR, and HRMS data consistent with the structures.
1-[2-(Tr iflu or om et h yl)p h en yl]u r ea (4). 2-(Trifluoro-
methyl)aniline (62.0 mL, 500 mmol) was dissolved in acetic
acid/water (1:1, 400 mL) and stirred at room temperature. To
the solution was slowly added a suspension of sodium cyanate
(65.0 g, 1000 mmol) in water (500 mL). A white precipitate of
the product appeared quickly. The suspension was stirred for
2.5 h at room temperature and poured into ice-water. The
precipitate was collected by filtration, washed with water, and
air dried. Recrystallization from ethyl acetate gave 4 (34.0 g,
33% yield) as white plates: mp 201-202 °C; IR (KBr) 3439,
The intermolecular arrangement postulated here is
similar to that reported for FAD and NADPH in the
active site of glutathione reductase:10,41 the flavin moiety
of FAD is stacked onto the nicotinamide ring of NADPH
and the pyrimidine ring of the flavin and the carbamoyl
group of the nicotinamide face each other. It is of great
interest that the intermolecular arrangements such as
those shown in Figure 6 can be seen in a model system
even though no steric compulsion exists to arrange them
in this order.
3341, 3218, 1657, 1618, 1524, 1321, 760 cm-1
;
1H NMR
(DMSO-d6) δ 6.42 (brs, 2H), 7.18 (t, J ) 7.6 Hz, 1H), 7.53-
7.63 (m, 2H), 7.82 (brs, 1H), 7.96 (d, J ) 8.4 Hz, 1H). Anal.
Calcd for C8H7F3N2O: C, 47.07; H, 3.46; N, 13.72. Found: C,
47.28; H, 3.43; N, 13.88.
1-[2-(Tr iflu or om eth yl)p h en yl]ba r bitu r ic Acid (5). So-
dium (2.8 g, 120 mmol) was dissolved in dry methanol (100
mL) at 0 °C. To the solution were added 4 (20.4 g, 100 mmol)
and diethyl malonate (18.2 mL, 120 mmol), and the mixture
was refluxed with stirring for 16 h under argon atmosphere.
After being cooled, the solution was poured into ice-water and
made strongly acidic with 2 M hydrochloric acid. The pre-
cipitate was collected by filtration, washed with water, and
air dried. Recrystallization from hot methanol gave 5 (24.5
g, 90% yield) as white plates: mp 270-272 °C; IR (KBr) 3233,
3160, 3102, 2992, 2893, 1686, 1437, 1346, 1318, 775 cm-1; 1H
NMR (DMSO-d6) δ 3.87 (dd, J ) 21.2, 75.4 Hz, 2H), 7.52-
7.85 (m, 4H), 11.72 (brs, 1H). Anal. Calcd for C11H7F3N2O3:
C, 48.54; H, 2.59; N, 10.29. Found: C, 48.77; H, 2.47; N, 10.36.
The present result strongly indicates not only a pos-
sibility that there might exist stabilizing effects due to
the overlap of molecular planes of a flavin and an NAD-
(P)H coenzyme but also a possibility that functional
groups in an apoprotein in proximity to a flavin coenzyme
in the active site of a flavoenzyme have significant
influence on the stereoselective interaction with a sub-
strate.
The effect of stereochemical control by a polar func-
tional group elucidated in the present study may suggest
that particular amino acid residues in the flavoprotein
might have contributed to establish the stereochemistry
of modern and sophisticated biochemical redox reactions
during chemical evolution of the enzyme.42,43
(42) Pai, E. F.; Karplus, P. A.; Schultz, G. E. Biochemistry 1988,
27, 4465-4474.
(43) Ohno, A.; Tsutsumi, A.; Kawai, Y.; Yamazaki, N.; Mikata, Y.;
Okamura, M. J . Am. Chem. Soc. 1994, 116, 8133-8137.
(44) Kurzer, F. Organic Syntheses; Wiley: New York, 1963; Vol. IV,
pp 49-51.
(45) Nevile, R. H. C.; Winther, A. Ber. Dtsch. Chem. Ges. 1880, 13,
962-973.
(46) Mauzerall, D.; Westheimer, F. H. J . Am. Chem. Soc. 1955, 77,
2261-2264.
(41) Karplus, P. A.; Schulz, G. E. J . Mol. Biol. 1989, 210, 163-180.

Atropisomeric Flavoenzyme Models
J . Org. Chem., Vol. 61, No. 26, 1996 9353
6-Ch lor o-3-[2-(tr iflu or om eth yl)p h en yl]u r a cil (6). To a
mixture of 5 (9.53 g, 35 mmol) in phosphorus oxychloride (65.2
mL, 700 mmol) was added water (3.15 mL, 175 mmol)
portionwise. The resulting mixture was heated with stirring
at 60 °C for 28 h. Phosphorus oxychloride was removed under
reduced pressure, and the residue was poured into ice-water.
The mixture was neutralized with 2 M aqueous sodium
hydroxide, and the precipitate was collected by filtration,
washed with water, and air dried. Recrystallization from ethyl
acetate gave 6 (7.33 g, 72% yield) as white needles: mp 227-
228 °C; IR (KBr) 3133, 2951, 2799, 1744, 1653, 1487, 1456,
1410, 1318, 812, 770 cm-1; 1H NMR (DMSO-d6) δ 6.05 (s, 1H),
7.56 (d, J ) 7.8 Hz, 1H), 7.67 (t, J ) 7.4 Hz, 1H), 7.77-7.87
(m, 2H), 12.68 (brs, 1H). Anal. Calcd for C11H6ClF3N2O2: C,
45.46; H, 2.08; N, 9.64. Found: C, 45.57; H, 2.06; N, 9.60.
6-(4-ter t-Bu t yla n ilin o)-3-[2-(t r iflu or om et h yl)p h en yl]-
u r a cil (7). A solution of 6 (7.27 g, 25 mmol) and 4-tert-
butylaniline (8.76 mL, 55 mmol) in butanol (50 mL) was
refluxed with stirring for 3 h under argon atmosphere. After
the mixture was cooled, the precipitate was collected by
filtration and washed with methanol. Recrystallization from
hot methanol gave 7 (7.70 g, 76% yield) as white needles: mp
292-293 °C; IR (KBr) 3329, 3081, 2965, 1734, 1599, 1559,
1422, 1318, 766 cm-1; 1H NMR (DMSO-d6) δ 1.29 (s, 9H), 4.83
(s, 1H), 7.20 (d, J ) 8.6 Hz, 2H), 7.43 (d, 3H), 7,63 (t, J ) 7.4
Hz, 1H), 7.73-7.83 (m, 2H), 8.34 (brs, 1H), 10.76 (brs, 1H).
Anal. Calcd for C21H20F3N3O2: C, 62.53; H, 5.00; N, 10.42.
Found: C, 62.74; H, 5.00; N, 10.44.
sodium sulfate, and evaporated to dryness under reduced
pressure. Crude 10-(4-tert-butylphenyl)-1,5-dihydro-3-[2-(hy-
droxymethyl)phenyl]pyrimido[4,5-b]quinoline-2,4(3H,10H)-di-
one (1r ed ), which contained a small amount of 1 due to air-
oxidation, was obtained as a pale yellow solid.
The solid was dissolved in chloroform (50 mL), and 2,3-
dichloro-5,6-dicyano-1,4-benzoquinone (1.82 g, 8 mmol) was
added to the solution with stirring at room temperature. After
15 min, the reaction mixture was washed with saturated
aqueous sodium bicarbonate twice as well as with brine once
and dried over anhydrous sodium sulfate, and the solvent was
evaporated to dryness under reduced pressure. The residue
was subjected to column chromatography on silica gel with
chloroform/methanol (50:1) as eluent, and the eluted material
was recrystallized from chloroform/methanol to give 1 (3.32
g, 92% yield) as yellow plates. 10-(4-tert-Butylphenyl)-3-[2-
(hydroxy[2H2]methyl)phenyl][5-2H]pyrimido[4,5-b]quinoline-
2,4(3H,10H)-dione (1-5-d, 98 atom % D) was prepared similarly
from 9-5-d and sodium borodeuteride (98 atom % D) as starting
materials: mp 300 °C; UV (CH3CN) λmax 267 (ꢀ 37 800), 319 (ꢀ
10 700), 401 nm (ꢀ 12 000); IR (KBr) 3493, 3042, 2967, 2874,
1709, 1649, 1612, 1564, 1530, 1487, 1453, 1412, 797, 752 cm-1
;
1H NMR (CDCl3) δ 1.43 (s, 9H), 2.60 (t, J ) 6.4 Hz, 1H), 4.47-
4.51 (m, 2H), 6.95 (d, J ) 8.4 Hz, 1H), 7.17-7.31 (m, 3H),
7.43-7.52 (m, 3H), 7.59-7.73 (m, 4H), 7.97 (dd, J ) 1.4, 8.0
Hz, 1H), 9.08 (s, 1H); HRMS (EI) m/z calcd for C28H25N3O3
(M•+) 451.1897, found 451.1899. Anal. Calcd for C28H25N3-
O3: C, 74.48; H, 5.58; N, 9.31. Found: C, 74.28; H, 5.67; N,
9.14.
10-(4-ter t-Bu tylp h en yl)-3-[2-(tr iflu or om eth yl)p h en yl]-
p yr im id o[4,5-b]qu in olin e-2,4(3H,10H)-d ion e (8). 2-Fluo-
robenzaldehyde (1.26 mL, 12 mmol) and 7 (4.03 g, 10 mmol)
were dissolved in DMF (8 mL), and the solution was heated
with stirring at 100 °C for 6 h. After the mixture was cooled,
the precipitate was collected by filtration and washed with
methanol. Recrystallization from hexane/chloroform gave 8
(3.18 g, 65% yield) as yellow needles. 10-(4-tert-Butylphenyl)-
3-[2-(trifluoromethyl)phenyl][5-2H]pyrimido[4,5-b]quinoline-
2,4(3H,10H)-dione (8-5-d, 98 atom % D) was prepared from 7
and 2-fluoro[R-2H]benzaldehyde similarly: mp >300 °C; IR
(KBr) 3474, 3081, 2967, 2874, 1717, 1669, 1620, 1566, 1534,
1491, 1454, 1414, 1318, 797, 764 cm-1; 1H NMR (CDCl3) δ 1.42
(s, 9H), 6.94 (d, J ) 8.0 Hz, 1H), 7.21-7.35 (m, 3H), 7.46 (t, J
) 7.0 Hz, 1H), 7.56 (t, J ) 7.8 Hz, 1H), 7.62-7.73 (m, 4H),
7.80 (d, J ) 7.8 Hz, 1H), 7.95 (dd, J ) 1.6, 7.8 Hz, 1H), 9.07
(s, 1H). Anal. Calcd for C28H22F3N3O2‚0.5CHCl3: C, 62.33;
H, 4.13; N, 7.65. Found: C, 62.62; H, 4.13; N, 7.70.
1-(2-Meth ylp h en yl)ba r bitu r ic Acid (11). Sodium (2.8 g,
120 mmol) was dissolved in dry methanol (100 mL) at 0 °C.
To the solution were added 10 (15.0 g, 100 mmol) and diethyl
malonate (18.2 mL, 120 mmol), and the solution was refluxed
with stirring for 16 h under argon atmosphere. After being
cooled, the solution was poured into ice-water and made
strongly acidic with 2 M hydrochloric acid. The precipitate
was collected by filtration, washed with water, and air dried.
Recrystallization from hot methanol gave 11 (17.7 g, 81% yield)
as white plates: mp 250-252 °C; IR (KBr) 3219, 3162, 3098,
2992, 2892, 1680, 1497, 1435, 820, 774, 725 cm-1
;
1H NMR
(DMSO-d6) δ 2.09 (s, 3H), 3.78 (dd, J ) 21.2, 42.0 Hz, 2H),
7.14-7.31 (m, 4H), 11.53 (brs, 1H). Anal. Calcd for C11H10
-
N2O3: C, 60.55; H, 4.62; N, 12.84. Found: C, 60.32; H, 4.66;
N, 12.79.
6-Ch lor o-3-(2-m eth ylp h en yl)u r a cil (12). To a mixture
of 11 (10.9 g, 50 mmol) in phosphorus oxychloride (46.6 mL,
500 mmol) was added water (2.25 mL, 125 mmol) portionwise.
The resulting mixture was heated with stirring at 60 °C for
24 h. Phosphorus oxychloride was removed from the mixture
under reduced pressure, and the residue was poured into ice-
water. The mixture was neutralized with 2 M aqueous sodium
hydroxide, and the precipitate was collected by filtration,
washed with water, and air dried. Recrystallization from ethyl
acetate gave 12 (6.20 g, 52% yield) as white plates: mp 228-
229 °C; IR (KBr) 3102, 2888, 2799, 1736, 1649, 1495, 1408,
10-(4-ter t-Bu tylp h en yl)-3-(2-ca r boxyp h en yl)p yr im id o-
[4,5-b]qu in olin e-2,4(3H,10H)-d ion e (9). A solution of 8
(4.90 g, 10 mmol) in concentrated sulfuric acid (15 mL) was
heated at 130 °C with stirring for 5 h. After being cooled, the
solution was poured into ice-water, and the yellow precipitate
was extracted with chloroform. The organic layer was washed
with brine and dried over anhydrous sodium sulfate, and the
solvent was evaporated to dryness under reduced pressure.
The residue was recrystallized from chloroform/methanol to
give 9 (4.14 g, 89% yield) as a yellow powder. 10-(4-tert-
Bu t ylph en yl)-3-(2-ca r boxyph en yl)[5-2H ]pyr im ido[4,5-b]-
quinoline-2,4(3H,10H)-dione (9-5-d, 98 atom % D) was pre-
pared from 8-5-d similarly: dec 278-280 °C; IR (KBr) 3410,
3185, 3050, 2965, 2868, 2616, 2504, 1715, 1642, 1618, 1532,
1
1323, 826, 764, 720 cm-1; H NMR (DMSO-d6) δ 2.04 (s, 3H),
6.00 (s, 1H), 7.13-7.34 (m, 4H), 12.56 (brs, 1H). Anal. Calcd
for C11H9ClN2O2: C, 55.83; H, 3.83; N, 11.84. Found: C, 55.88;
H, 3.74; N, 11.88.
6-(4-ter t-Bu tyla n ilin o)-3-(2-m eth ylp h en yl)u r a cil (13).
A solution of 12 (3.55 g, 15 mmol) and 4-tert-butylaniline (5.57
mL, 35 mmol) in butanol (25 mL) was refluxed with stirring
for 3 h under argon atmosphere. After the mixture was cooled,
the precipitate was collected by filtration and washed with
methanol. Recrystallization from hot methanol gave 13 (3.58
g, 68% yield) as white needles: mp 300 °C; IR (KBr) 3310,
1
1491, 1453, 1416, 797, 754 cm-1; H NMR (DMSO-d6) δ 1.42
(s, 9H), 6.79 (d, J ) 8.4 Hz, 1H), 7.28-7.42 (m, 3H), 7.52 (m,
2H), 7.64-7.82 (m, 4H), 8.03 (dd, J ) 1.6, 7.8 Hz, 1H), 8.27
(dd, J ) 1.2, 7.8 Hz, 1H), 9.23 (s, 1H), 12.79 (brs, 1H). Anal.
Calcd for C28H23N3O4‚CH3OH: C, 70.01; H, 5.47; N, 8.45.
Found: C, 69.77; H, 5.34; N, 8.45.
3079, 2963, 1726, 1597, 1516, 1300, 725 cm-1 1H NMR
;
10-(4-ter t-Bu tylp h en yl)-3-[2-(h yd r oxym eth yl)p h en yl]-
p yr im id o[4,5-b]qu in olin e-2,4(3H,10H)-d ion e (1). To a
suspension of 9 (3.72 g, 8 mmol) in dry THF (30 mL) was added
sodium borohydride (0.91 g, 24 mmol), and the mixture was
stirred at 0 °C. After the generation of hydrogen had ceased,
a solution of boron trifluoride diethyl etherate (ca. 47%, 3 mL)
in dry THF (15 mL) was added to the suspension, and the
mixture was stirred for 1 h at room temperature. Then, the
mixture was poured into ice-water, and organic materials
were extracted with chloroform three times. The combined
organic solution was washed with water, dried over anhydrous
(DMSO-d6) δ 1.30 (s, 9H), 2.06 (s, 3H), 4.86 (s, 1H), 7.07-7.29
(m, 6H), 7.42 (d, J ) 8.6 Hz, 2H), 8.26 (brs, 1H), 10.60 (brs,
1H). Anal. Calcd for C21H23N3O2: C, 72.18; H, 6.63; N, 12.03.
Found: C, 72.21; H, 6.59; N, 12.11.
10-(4-ter t-Bu t ylp h en yl)-3-(2-m et h ylp h en yl)p yr im id o-
[4,5-b]qu in olin e-2,4(3H,10H)-d ion e (2). 2-Fluorobenzalde-
hyde (0.63 mL, 6 mmol) and 13 (1.75 g, 5 mmol) were dissolved
in DMF (5 mL), and the solution was heated with stirring at
100 °C for 6 h. After the mixture was cooled, the precipitate
was collected by filtration and washed with methanol. Re-

9354 J . Org. Chem., Vol. 61, No. 26, 1996
Ohno et al.
crystallization from hexane/chloroform gave 2 (1.31 g, 60%
yield) as yellow needles. 10-(4-tert-Butylphenyl)-3-(2-meth-
ylphenyl)[5-2H]pyrimido[4,5-b]quinoline-2,4(3H,10H)-dione (2-
5-d, 98 atom % D) was prepared from 13 and 2-fluoro[R-
2H]benzaldehyde similarly: mp 293-294 °C; UV (CH3CN) λmax
266 (ꢀ 40 200), 318 (ꢀ 11 400), 402 nm (ꢀ 12 800); IR (KBr) 3461,
3061, 2967, 2868, 1711, 1661, 1618, 1566, 1534, 1491, 1454,
was recrystallized from ethyl acetate to give 17 (1.18 g, 75%
yield) as a white powder: mp 237-238 °C; IR (KBr) 3179,
3108, 2942, 2801, 1730, 1665, 1462, 1410, 1323, 802, 775 cm-1
;
1H NMR (DMSO-d6) δ 2.12 (s, 3H), 6.12 (s, 1H), 7.65 (d, J )
1.6 Hz, 1H), 7.86 (d, J ) 1.6 Hz, 1H), 12.86 (brs, 1H). Anal.
Calcd for C11H7Br2ClN2O2: C, 33.50; H, 1.79; N, 7.10. Found:
C, 33.76; H, 1.86; N, 7.05.
1
1414, 797, 748 cm-1; H NMR (CDCl3) δ 1.42 (s, 9H), 2.18 (s,
3-(2,4-Dibr om o-6-m eth ylph en yl)-6-(4-ter t-bu tylan ilin o)-
u r a cil (18). A mixture of 17 (1.97 g, 5 mmol) and 4-tert-
butylaniline (1.75 mL, 11 mmol) in butanol (12.5 mL) was
refluxed with stirring for 3 h under argon atmosphere. After
the mixture was cooled, the precipitate was collected by
filtration, washed with methanol, and dried to give 18 (2.22
g, 88% yield) as white plates: mp >300 °C; IR (KBr) 3316,
3108, 2961, 2868, 1723, 1601, 1516, 1462, 1416, 1300, 783
cm-1; 1H NMR (DMSO-d6) δ 1.29 (s, 9H), 2.11 (s, 3H), 4.85 (s,
1H), 7.20 (d, J ) 8.6 Hz, 2H), 7.43 (d, J ) 8.6 Hz, 2H), 7.61 (d,
J ) 2.2 Hz, 1H), 7.81 (d, J ) 2.2 Hz, 1H), 8.44 (brs, 1H), 10.86
(brs, 1H). Anal. Calcd for C21H21Br2N3O2: C, 49.73; H, 4.17;
N, 8.28. Found: C, 49.75; H, 4.10; N, 8.24.
3H), 6.93 (d, J ) 8.6 Hz, 1H), 7.12-7.17 (m, 1H), 7.22-7.34
(m, 5H), 7.45 (t, J ) 7.4 Hz, 1H), 7.61-7.70 (m, 3H), 7.94 (dd,
J ) 1.6, 7.8 Hz, 1H), 9.06 (s, 1H); HRMS (EI) m/z calcd for
C28H25N3O2 (M•+) 435.1948, found 435.1950. Anal. Calcd for
C28H25N3O2‚CHCl3: C, 62.77; H, 4.72; N, 7.57. Found: C,
62.71; H, 4.76; N, 7.63.
3-[2-[[(ter t-Bu tyld im eth ylsilyl)oxy]m eth yl]p h en yl]-10-
(4-ter t-bu tylph en yl)pyr im ido[4,5-b]qu in olin e-2,4(3H,10H)-
d ion e (3). tert-Butyldimethylsilyl chloride (226 mg, 1.5
mmol), 1 (226 mg, 0.5 mmol), and imidazole (204 mg, 3.0 mmol)
were dissolved in dry DMF (8 mL), and the solution was stirred
for 1 h at room temperature under argon atmosphere. The
solution was poured into ice-water, and organic materials
were extracted with ethyl acetate. The organic solution was
washed with water four times as well as with brine once and
dried over anhydrous sodium sulfate, and the solvent was
evaporated to dryness under reduced pressure. The yellow
residue was recrystallized from hexane/chloroform to give 3
(207 mg, 73% yield) as a yellow powder. 3-[2-[[(tert-Butyldi-
methylsilyl)oxy][2H2]methyl]phenyl]-10-(4-tert-butylphenyl)[5-
2H]pyrimido[4,5-b]quinoline-2,4(3H,10H)-dione (3-5-d, 98 atom
% D) was prepared similarly from 1-5-d: mp 225-226 °C; IR
(KBr) 3451, 3040, 2957, 2859, 1715, 1663, 1618, 1566, 1535,
3-(2,4-Dibr om o-6-m eth ylph en yl)-10-(4-ter t-bu tylph en yl)-
p yr im id o[4,5-b]qu in olin e-2,4(3H,10H)-d ion e (19). 2-Fluo-
robenzaldehyde (0.63 mL, 6 mmol) and 18 (2.54 g, 5 mmol)
were dissolved in DMF (15 mL), and the solution was heated
with stirring at 120 °C for 6 h. After DMF was removed from
the solution under reduced pressure, methanol (50 mL) was
added to the residue, and insoluble compounds were removed
by filtration. The methanol was removed from the filtrate
under reduced pressure, the residue was subjected to column
chromatography on silica gel with hexane/ethyl acetate (1:1)
as eluent, and the product was recrystallized from chloroform/
methanol to give 19 (1.78 g, 60% yield) as yellow plates: mp
224-225 °C; UV (CH3CN) λmax 266 (ꢀ 35 300), 319 (ꢀ 10 900),
402 nm (ꢀ 11 900); IR (KBr) 3476, 3050, 2965, 1707, 1659,
1
1491, 841, 754 cm-1; H NMR (CDCl3) δ 0.01 (s, 6H), 0.88 (s,
9H), 1.43 (s, 9H), 4.64 (s, 2H), 6.91 (d, J ) 8.8 Hz, 1H), 7.13
(dd, J ) 1.6, 7.4 Hz, 1H), 7.21-7.48 (m, 5H), 7.61-7.69 (m,
4H), 7.94 (dd, J ) 1.4, 7.8 Hz, 1H), 9.06 (s, 1H). Anal. Calcd
for C34H39N3O3Si: C, 72.18; H, 6.95; N, 7.43. Found: C, 72.02;
H, 6.87; N, 7.41.
1
1613, 1566, 1530, 1489, 1453, 1412, 795, 762 cm-1; H NMR
(CDCl3) δ 1.42 (s, 9H), 2.20 (s, 3H), 6.95 (d, J ) 8.8 Hz, 1H),
7.23 (t, J ) 8.8 Hz, 1H), 7.29 (s, 1H), 7.42 (s, 1H), 7.47 (t, J )
8.0 Hz, 1H), 7.63-7.72 (m, 4H), 7.96 (dd, J ) 1.6, 8.0 Hz, 1H),
9.11 (s, 1H). Anal. Calcd for C28H23Br2N3O2‚CH3OH: C, 55.70;
H, 4.35; N, 6.72. Found: C, 55.52; H, 4.30 N, 6.65.
(S)-(-)-19: CD (CH3CN) λmax 265 (∆ꢀ -10.43), 291 (∆ꢀ
+0.415), 324 (∆ꢀ -0.723), 403 nm (∆ꢀ +0.560).
1-(2,4-Dibr om o-6-m eth ylp h en yl)u r ea (15). To acetic
acid (45 mL) was dissolved 14 (11.9 g, 45 mmol), and the
solution was stirred at room temperature. A suspension of
sodium cyanate (5.85 g, 90 mmol) in water (45 mL) was added
to the solution dropwise until a white precipitate appeared.
Then, the rest of the suspension and water (45 mL) were added
quickly with vigorous stirring. After 3 min, the precipitate
that appeared was collected by filtration, washed with water,
and air dried. Recrystallization from ethanol gave 15 (1.85 g,
13% yield) as white needles: mp >300 °C; IR (KBr) 3434, 3281,
1659, 1599, 1535, 1352, 852 cm-1; 1H NMR (DMSO-d6) δ 2.22
(s, 3H), 5.95 (brs, 2H), 7.45 (d, J ) 2.2 Hz, 1H), 7.66 (d, J )
2.2 Hz, 1H), 7.78 (brs, 1H). Anal. Calcd for C8H8Br2N2O: C,
31.20; H, 2.62; N, 9.10. Found: C, 31.26; H, 2.57; N, 9.13.
1-(2,4-Dibr om o-6-m eth ylp h en yl)ba r bitu r ic Acid (16).
Sodium (0.34 g, 15 mmol) was dissolved in dry methanol (20
mL) at 0 °C. To the solution were added 15 (3.08 g, 10 mmol)
and diethyl malonate (1.82 mL, 12 mmol), and the mixture
was refluxed with stirring for 16 h under argon atmosphere.
After being cooled, the solution was poured into ice-water and
made strongly acidic with 2 M hydrochloric acid. The pre-
cipitate was collected by filtration, washed with water, and
air dried. Recrystallization from hot methanol to gave 16 (2.48
g, 66% yield) as white plates: mp 249-250 °C; IR (KBr) 3544,
3233, 3127, 2886, 1709, 1582, 1557, 1377, 1348, 779 cm-1; 1H
NMR (DMSO-d6) δ 2.16 (s, 3H), 3.97 (brs, 2H), 7.63 (d, J )
1.8 Hz, 1H), 7.84 (d, J ) 1.8 Hz, 1H), 11.83 (brs, 1H). Anal.
Calcd for C11H8Br2N2O3: C, 35.14; H, 2.14; N, 7.45. Found:
C, 35.04; H, 2.18; N, 7.34.
(R)-(+)-19: CD (CH3CN) λmax 265 (∆ꢀ +9.93), 291 (∆ꢀ
-0.228), 319 (∆ꢀ +0.723), 410 nm (∆ꢀ -0.800).
Kin etic Mea su r em en ts for Th er m a l En a n tiom er iza -
tion . Thermal enantiomerization of 1 or 2 was carried out
by immersing a solution of chiral 1 or 2 in DMF (ca. 1.0 ×
10-3 M) in a thermostated oil bath at 30-70 °C. At appropri-
ate intervals, aliquots were withdrawn and subjected to HPLC
analysis to measure ee.29,30
Cr ysta llogr a p h ic Stu d ies. The lattice parameters and
intensity data were measured on a Rigaku AFC7R diffracto-
meter and an 18 kW rotating anode generator with 8 kW Cu
KR radiation. The structures were solved by the direct
method, and non-hydrogen atoms were refined anisotropically.
All calculations were performed using a Texsan crystal-
lographic software package developed by Molecular Structure
Corporation (1985 and 1992). The crystallographic parameters
are listed in Table 4.
Deter m in a tion of th e Absolu te Con figu r a tion of 2. (S)-
(-)-19 (>99% ee, 20.2 mg, 0.034 mmol) was dissolved in
dichloromethane (0.2 mL). To the solution were added 10%
palladium on carbon (50 mg), triethylamine (0.05 mL), and
ethanol (5 mL), and the resulting mixture was stirred for 1 h
at room temperature under hydrogen atmosphere. After
palladium on carbon was removed by filtration, the solvent
was removed from the filtrate under reduced pressure. The
residue was subjected to preparative TLC with chloroform/
methanol (50:1) as a developing solvent to give (R)-2 (5.7 mg,
28% yield), which was subjected to HPLC to confirm that this
is the (R)-(+)-enantiomer (98% ee).
Deter m in a tion of th e Absolu te Con figu r a tion of 1.
(+)-1 (>99% ee, 16.3 mg, 0.036 mmol), p-toluenesulfonyl
chloride (19.1 mg, 0.10 mmol), and 4-(dimethylamino)pyridine
(12.2 mg, 0.10 mmol) were dissolved in dry chloroform (5 mL),
and the solution was stirred for 1 h at 0 °C under argon
atmosphere. After the solvent was evaporated from the
6-Ch lor o-3-(2,4-d ib r om o-6-m et h ylp h en yl)u r a cil (17).
To a mixture of 16 (1.50 g, 4 mmol) in phosphorus oxychloride
(3.73 mL, 40 mmol) was added water (0.22 mL, 12 mmol)
portionwise. The resulting mixture was heated with stirring
at 60 °C. After 72 h, the mixture was poured into ice-water,
neutralized with 2 M aqueous sodium hydroxide, and extracted
with ethyl acetate twice. The organic solution was washed
with brine and dried over anhydrous sodium sulfate, and the
solvent was evaporated to dryness under reduced pressure.
The residue was subjected to column chromatography on silica
gel with hexane/ethyl acetate (1:1) as eluent, and the product

Atropisomeric Flavoenzyme Models
J . Org. Chem., Vol. 61, No. 26, 1996 9355
solution to dryness under reduced pressure, the residue was
subjected to column chromatography on silica gel with chlo-
roform as eluent to give 10-(4-tert-butylphenyl)-3-[2-[[(p-tolu-
enesulufonyl)oxy]methyl]phenyl]pyrimido[4,5-b]quinoline-2,4-
(3H,10H)-dione. Dry THF (5 mL) was added to the compound,
and the suspension was stirred at 0 °C. Lithium aluminum
hydride (3.8 mg, 0.10 mmol) was added portionwise to the
suspension, and the resulting mixture was stirred for 1 h at 0
°C. After excess lithium aluminum hydride was decomposed
with ethyl acetate, the mixture was poured into water and
organic materials were extracted with chloroform twice. The
combined organic layers were washed with water and dried
over anhydrous magnesium sulfate, and the solvent was
evaporated to dryness under reduced pressure to give crude
2r ed . To the residue were added chloroform (1 mL) and 2,3-
dichloro-5,6-dicyano-1,4-benzoquinone (11.4 mg, 0.050 mmol),
and the mixture was stirred for 5 min at room temperature.
After the chloroform was removed from the solution under
reduced pressure, the residue was subjected to column chro-
matography on silica gel with hexane/ethyl acetate (1:2) as
eluent to give 2 (12.7 mg, 81% yield), which was subjected to
HPLC to confirm that this is the (S)-(-)-enantiomer (94% ee).
Consequently, the absolute configuration of (+)-1 has been
determined to be S.
E n a n t iofa ce-Differ en t ia t in g (Net ) Hyd r id e-Tr a n sfer
Rea ction betw een Ra cem ic d F l-5-d a n d BNAH. To a
solution of BNAH (214 mg, 1.0 mmol) and a catalyst (magne-
sium perchlorate (223 mg, 1.0 mmol) or trichloroacetic acid
(327 mg, 2.0 mmol)) in acetonitrile (5 mL) was added racemic
dFl-5-d (0.2 mmol), and the mixture was stirred in the dark
at 298 K under argon atmosphere. The reaction was moni-
tored by TLC. The time required for completion of the reaction
is shown in Table 5. After the reaction was complete, the
solvent was removed under reduced pressure and chloroform
(ca. 50 mL) was added to the residue. The suspension was
washed with saturated aqueous ammonium chloride and brine
and dried over anhydrous magnesium sulfate, and the solvent
was evaporated from the solution to dryness under reduced
pressure. The residue was subjected to short column chro-
matography on silica gel with chloroform as eluent to give a
mixture of diastereomers of dFlred-5-d. The syn/anti ratios of
the reacting faces of 2, 3, and 8 were determined with 2r ed -5-
d, 3r ed -5-d, and 8r ed -5-d thus obtained, respectively. The ratio
in 1r ed-5-d was determined through 3r ed-5-d, which was derived
from 1r ed -5-d as follows: 1r ed -5-d, tert-butyldimethylsilyl chlo-
ride (90 mg, 0.6 mmol), and imidazole (82 mg, 1.2 mmol) were
dissolved in dry DMF (3 mL), and the solution was stirred in
the dark at room temperature under argon atmosphere. After
1 h, the solution was poured into ice-water, and organic
materials were extracted with ethyl acetate. The organic layer
was washed with water four times as well as with brine once
and dried over anhydrous magnesium sulfate, and the solvent
was evaporated to dryness under reduced pressure. The
residue was subjected to preparative TLC with hexane/ethyl
acetate (2:1) as a developing solvent in the dark to give
diastereomers of 3r ed -5-d. In the same manner as described
in previous papers,24,25 the syn/anti ratio of the reacting faces
was determined from the ratio of peak areas of the two
diastereomeric C(5) protons in the 1H NMR spectrum. The
C(5) protons of the diastereomers could be discriminated by
the difference in 1H NMR chemical shifts in the presence of
Eu(fod)3-d27 (0.10 mol equiv to dFlred-5-d).
Mea su r em en ts of Associa tion Con sta n ts. The forma-
tion of a complex between dFl and Mg2+ was monitored
spectrophotometrically by following the disappearance of the
absorbance at 439 nm of dFl. The linearity of the plot shown
in Figure 4 reveals that the molar ratio in the complex is 1:1.
The association constants were obtained from the slopes of the
plots.
Dia ster eofa ce-Differ en tia tin g (Net) Hyd r id e-Tr a n sfer
Rea ction betw een Ra cem ic 1 a n d Ch ir a l Me2P NP H.
Racemic 1 (45 mg, 1.0 × 10-1 mmol) was added to a solution
of chiral Me2PNPH (1.5 mg, 5.0 × 10-3 mmol) and magnesium
perchlorate (0-50 mol equiv to Me2PNPH) in acetonitrile (10
mL), and the solution was stirred in the dark for 30 min at
298 K under argon atmosphere. No products other than 1r ed
were detected on TLC and HPLC. The solution was evapo-
rated to dryness under reduced pressure, and the residue was
subjected to preparative TLC with hexane/ethyl acetate (1:3)
as a developing solvent in the dark to give 1r ed . To a solution
of the 1r ed in chloroform (3 mL) was added 2,3-dichloro-5,6-
dicyano-1,4-benzoquinone (23 mg, 1.0 × 10-1 mmol), and the
mixture was stirred at room temperature. After 5 min, the
solvent was evaporated from the suspension to dryness under
reduced pressure, and the residue was subjected to preparative
TLC with ethyl acetate as a developing solvent to give 1. The
S/R ratio in 1 reduced by Me2PNPH was then determined by
HPLC.
Asym m et r ic (Net ) H yd r id e-Tr a n sfer R ea ct ion b e-
tw een Ch ir a l 1-5-d a n d Ch ir a l Me2P NP H. To a solution
of chiral Me2PNPH (75 mg, 0.25 mmol) and magnesium
perchlorate (0 or 2.5 mol equiv to 1-5-d) in acetonitrile (5 mL)
was added chiral 1-5-d (45 mg, 0.1 mmol), and the mixture
was stirred in the dark at 298 K under argon atmosphere. The
reaction was monitored by TLC, and the results are sum-
marized in Table 8 together with the time required for
completion of the reaction. Then, the same procedures as those
described for the enantioface-differentiating (net) hydride-
transfer reactions were carried out to determine the syn/anti
ratio of the reacting faces.
Ack n ow led gm en t. We are deeply indebted to Daicel
Chemical Co., Ltd. for providing us with a CHIRALCEL
OD column for HPLC. A part of this work was sup-
ported by a Grant-in-Aid for Scientific Research No.
07454194 from the Ministry of Education, Science and
Culture of J apan.
J O961799T