C O MMU N I C A T I O N S
Table 2. Kinetic Parameters for R-Methylbenzylamine Binding to Mb Mutants
(M- s-1
1
)
k
-1 (s-1
)
K (M )
-1
k
1
R
S
R
S
R
S
4
3
4
3
2
2
2
4
4
H64D/V68A
H64D/V68S
H64D
(1.3 ( 0.1) × 10
(2.2 ( 0.3) × 10
30 ( 10
(1.3 ( 0.1) × 10
(2.7 ( 0.1) × 10
4 ( 2
16
22
0.096
0.59
0.24
0.057
8.1 × 10
1.0 × 10
3.1 × 10
2.2 × 10
1.1 × 10
70
In summary, we have engineered the distal pocket of Mb to
convert the enantioselective cavity for the sulfoxidation of thio-
anisole. The H64D/V68A and H64D/V68S mutants are found to
oxidize thioanisole with high enantioselectivity and reactivity. These
mutants are also capable of enantioselective binding of R-methyl-
benzylamine, which mimics an expected sulfoxidation intermediate.
The kinetic study of the amine binding suggests that the Fe-O
bond cleavage in the intermediate is the chiral discrimination step
of the sulfoxidation. We are undertaking the crystal structural
analysis to elucidate the details of the substrate binding.
Acknowledgment. This work was supported by Grant-in-Aid
for Scientific Research 11490036 and 1122828 (to Y.W.), 13740384
(to T.U.), and 13780496 (to O.S.).
References
Figure 2. Proposed reaction coordinate for the thioanisole sulfoxidation
and the amine binding in H64D/V68A and H64D/V68S Mb.
(1) Adachi, S.; Nagano, S.; Ishimori, K.; Watanabe, Y.; Morishima, I.
Biochemistry 1993, 32, 241-252.
(
2) Matsui, T.; Nagano, S.; Ishimori, K.; Watanabe, Y.; Morishima, I.
Biochemistry 1996, 35, 13118-13124.
V68S (Table 1), the binding constant (K) values of (S)-R-
methylbenzylamine with H64D/V68A and H64D/V68S are 27-fold
and 112-fold larger than those of the corresponding (R)-amine,
respectively (Table 2).19 In the case of H64D Mb, the binding
constant for the (S)-amine is significantly weaker than the others,
whereas those for (R)-amine are about the same (Table 2).
To determine the chiral discrimination step in the amine binding,
(3) Ortiz de Montellano, P. R.; Catalano, C. E. J. Biol. Chem. 1985, 260,
9265-9271.
(4) Rao, S. I.; Wilks, A.; Ortiz de Montellano, P. R. J. Biol. Chem. 1993,
268, 803-809.
(5) Colonna, S.; Gaggero, N.; Casella, L.; Carrea, G.; Pasta, P. Tetrahedron:
Asymmetry 1992, 3, 95-106.
(6) Ozaki, S.; Ortiz de Montellano, P. R. J. Am. Chem. Soc. 1995, 117, 7056-
7064.
(
(
(
7) Savenkova, M. I.; Newmyer, S. L.; Ortiz de Montellano, P. R. J. Biol.
Chem. 1996, 271, 24598-24603.
8) Velde, F. v. d.; Rantwijk, F. v.; Sheldon, R. A. Trends Biotechnol. 2001,
we have measured the on-rate (k
binding to the Mb mutants by stopped-flow experiments. The
results are also summarized in Table 2. The on-rates (k ) of (R)-
and (S)-R-methylbenzylamine to H64D/V68A and H64D/V68S are
1
) and the off-rate (k-1) of amine
19, 73-80.
1
9
9) Matsui, T.; Ozaki, S.; Liong, E.; Phillips, G. N. J.; Watanabe, Y. J. Biol.
Chem. 1999, 247, 2838-2844.
1
(
10) Matsui, T.; Ozaki, S.; Watanabe, Y. J. Am. Chem. Soc. 1999, 121, 9952-
9957.
4
4
-1 -1
almost identical, 1.3 × 10 and 2.2-2.7 × 10 M
s
, respec-
(11) Ozaki, S.; Yang, H.-J.; Matsui, T.; Goto, Y.; Watanabe, Y. Tetrahedron:
Asymmetry 1999, 10, 183-192.
tively. In contrast, a tremendous difference is seen for the off-rate.
This indicates that the chiral discrimination of the (S)-amine ligation
over the (R)-amine by H64D/V68A and H64D/V68S is exclusively
caused by a very small off-rate of the (S)-amine relative to the
(
12) Phillips, G. N. J.; Arduini, R. M.; Springer, B. A.; Sligar, S. G. Proteins:
Struct., Funct., Genet. 1990, 7, 358-365.
(
13) Carver, T. E.; Rohlfs, R. J.; Olson, J. S.; Gibson, Q. H.; Blackmore, R.
S.; Springer, B. A.; Sligar, S. G. J. Biol. Chem. 1990, 265, 20007-20020.
(14) Egeberg, K. D.; Springer, B. A.; Sligar, S. G.; Carver, T. E.; Rohlfs, R.
J.; Olson, J. S. J. Biol. Chem. 1990, 265, 11788-11795.
(R)-amine, 1:27 for H64D/V68A and 1:92 for H64D/V68S. These
(
15) Quillin, M. L.; Li, T.; Olson, J. S.; Phillips, G. N. J.; Dou, Y.; Ikeda-
Saito, M.; Regan, R.; Carlson, M.; Gibson, Q. H.; Li, H.; Elber, R. J.
Mol. Biol. 1995, 245, 416-436.
selectivities would correspond to 93 and 98% ee for the amine
binding, respectively.
(
16) For example, H64D/V68A Mb has the Soret maximum at 408 nm,
Judging from a much smaller off-rate of the (S)-isomer in the
amine binding than that of the (R)-isomer, the off-rate for the Fe-O
bond cleavage in the (S)-sulfoxide formation may also be much
smaller than that for the (R)-sulfoxide due to the higher activation
energy of the former (Figure 2). Thus, enantioselectivity in the
sulfoxidation of thioanisole by H64D/V68A and H64D/V68S Mb
may be determined by the off-rate of sulfoxide.
In the case of H64D Mb, the on-rate of the (S)-amine to H64D
Mb is 7.5-fold smaller than that of (R)-amine whereas the off-rate
is only 1.7-fold smaller (Table 2). In this case, the (R)-sulfoxide
formation is expected to be slightly faster than (S)-isomer formation.
In fact, H64D Mb shows small enantiomeric excess for the (R)-
sulfoxidation (Table 1).
indicating that the ferric heme iron is in a hexacoordinated high-spin state.17
Upon the addition of (S)-R-methylbenzylamine to H64D/V68A Mb, the
Soret band shifts to 413 nm with a decrease in the absorbance. The
absorption band at 632 nm of the ferric high-spin state disappears upon
addition of the (S)-amine. The absorption spectrum of the H64D/V68A
Mb-amine complex, having a â-band at 532 nm and an R-band shoulder
18
around 560 nm, is typical for the hexacoordinated ferric low-spin state.
(17) Takano, T. J. Mol. Biol. 1977, 110, 537-568.
(
(
18) Iizuka, T.; Yonetani, T. AdV. Biophys. 1970, 1, 157-182.
19) The reactions of Mb mutants with the amine were performed in 50 mM
Tris/HCl (pH 9.0) at 20 °C. The final concentration was 5.0 µM for Mbs
and 0.5-5 mM for (R)- or (S)-R-methylbenzylamine. The kinetic traces
at 408 nm were used for determining pseudo-first-order rates. The binding
1
(k ) and dissociation (k-1) rates were given by the slope and intercept of
a plot of the observed rates versus the (R)- and (S)-amine concentration.
1
The binding constants (K) were given by k /k-1.
JA0256414
J. AM. CHEM. SOC.
9
VOL. 124, NO. 29, 2002 8507