Rational creation of mutant enzyme showing remarkable enhancement of catalytic activity and enantioselectivity toward poor substrates

Article abstract of DOI:10.1039/c001561j

Catalytic activity and enantioselectivity of lipase toward poor substrates bearing bulky substituents on both sides have been dramatically improved by rational design; the E value for a poor substrate was increased from 5 (wild-type enzyme) to >200 (I287F

Full text of DOI:10.1039/c001561j

View Article Online / Journal Homepage / Table of Contents for this issue
This article is part of the
2010 ‘Enzymes & Proteins’
web themed issue
This issue showcases high quality research in the field of enzymes and
proteins.
Please visit the website to access the other papers in this issue:-
http://www.rsc.org/chemcomm/enzymesandproteins

COMMUNICATION
www.rsc.org/chemcomm | ChemComm
Rational creation of mutant enzyme showing remarkable enhancement
of catalytic activity and enantioselectivity toward poor substrateswz
Tadashi Ema,* Shusuke Kamata, Masahiro Takeda, Yasuko Nakano and Takashi Sakai*
Received 26th January 2010, Accepted 23rd March 2010
First published as an Advance Article on the web 9th April 2010
DOI: 10.1039/c001561j
Catalytic activity and enantioselectivity of lipase toward poor
substrates bearing bulky substituents on both sides have been
dramatically improved by rational design; the E value for a poor
substrate was increased from 5 (wild-type enzyme) to 4200
(I287F/I290A double mutant) with an acceleration of the
reaction rate.
will hinder the acylation of (S)-1a, as is the case for (S)-1-
phenylethanol.13
The recombinant lipases were prepared as reported
previously,13 and the lipase-catalyzed kinetic resolutions of 1
were conducted with vinyl acetate in dry i-Pr2O at 30 1C
(Scheme 1). Enantioselectivity was evaluated by the E value
(the ratio of the rate constants for both enantiomers).15 The
results are summarized in Table 1. Because of the poor
reactivity of 1a as compared with 1-phenylethanol,13 higher
catalyst loadings and longer reaction times were needed. To
our delight, the conversion and the E value for the I287F
mutant toward 1a were approximately 2-fold and 6-fold
higher, respectively, than those for the wild-type enzyme
(entries 1 and 2). A similar trend was observed for homolog
1b (entries 7 and 8). It should be noted that the I287F mutant
has previously shown a decrease in conversion (nearly half)
and only 2-fold enhancement in the E value for 1-phenylethanol
as compared with the wild-type enzyme.13 In contrast to the
I287F mutation, the I287A mutation, diminishing the steric
hindrance at position 287, did not improve the conversion of
1a (entry 3). These results suggest that the I287F mutant has a
Improvement of the catalytic function of enzymes is important
from scientific and industrial viewpoints. Although lipases are
widely recognized as synthetically useful biocatalysts,1 the kinetic
resolution of secondary alcohols bearing bulky substituents on
both sides remains difficult. This drawback may be overcome
by altering the enzyme structure.2 Whereas directed evolution
based on random mutagenesis and high-throughput screening
has become the most popular method for the alteration of
biocatalysts,3–6 rational approaches based on site-directed
mutagenesis are also growing steadily.7–13
We have previously succeeded in controlling (both increasing
and decreasing) the enantioselectivity of a Burkholderia cepacia
lipase toward 1-phenylethanol by a single mutation on the
basis of the mechanism.13 In that study, enantioselectivity was
controlled by modulation of the steric repulsion between the
enzyme and the slower-reacting (S)-enantiomer. To achieve a
more remarkable improvement, a mutation that enhances
the reactivity of the (R)-enantiomer and that suppresses the
reactivity of the (S)-enantiomer should be effective. Here
we report the rational creation of mutant enzymes showing
the remarkable enhancement of both catalytic activity and
enantioselectivity toward poor substrates bearing bulky
substituents on both sides.
We employed 1-phenyl-1-hexanol (1a) as a poor substrate
and performed manual docking with reference to the transition-
state model (Fig. 1a).14 We set up a working hypothesis that
the I287F mutation (Fig. 1b and c) would enable attractive/
repulsive dual-mode interactions with (R)/(S)-1a. As shown in
Fig. 1c, Phe287 appears to make favorable contact with the
alkyl chain of (R)-1a in the transition state, which will accelerate
the acylation of (R)-1a. On the other hand, Phe287 is expected
to serve as a steric barrier to the phenyl group of (S)-1a, which
Fig. 1 (a) The transition-state model to rationalize the enantioselectivity
in the lipase-catalyzed kinetic resolution of secondary alcohols
(residues 287 and 290 are added to the original version). The catalytic
triad residues, the ester being produced, residue 287, and residue 290 are
shown in green, blue, red, and magenta, respectively. Typically, the (R)-
enantiomer reacts faster because, in this favorable conformation shown in
blue, the larger substituent (R1) can be directed toward external solvent
without severe strain and/or steric hindrance. (b)–(d) The active sites of
(b) the wild-type enzyme, (c) the I287F mutant, and (d) the I287F/I290A
double mutant. (R)-1a connected to Ser87 in the tetrahedral intermediate
is shown in light blue.
Division of Chemistry and Biochemistry, Graduate School of Natural
Science and Technology, Okayama University, Tsushima,
Okayama 700-8530, Japan. E-mail: ema@cc.okayama-u.ac.jp;
Fax: +81-86-251-8092; Tel: +81-86-251-8091
w This article is part of the ‘Enzymes and Proteins’ web-theme issue for
ChemComm.
z Electronic supplementary information (ESI) available: Site-directed
mutagenesis, synthesis of 1, lipase-catalyzed kinetic resolution of 1,
determination of kinetic constants, design of mutants, and copies of
NMR spectra. See DOI: 10.1039/c001561j
ꢀc
This journal is The Royal Society of Chemistry 2010
5440 | Chem. Commun., 2010, 46, 5440–5442

introducing single or double mutations, which amounts to
an energetic difference of at least 3.0 kcal molꢁ1
.
The degree of catalytic activity can be estimated roughly
from the total turnover number (TTN) and the turnover
frequency (TOF). TTN is defined as the number of substrate
molecules converted by one enzyme molecule, and TOF is the
turnover number per unit time. One can calculate these values
from the data in Table 1. For example, the TTN value for the
wild-type enzyme toward 1a (entry 1) is calculated to be 1080
from TTN = 0.5 mmol ꢂ 0.23/(700 mg ꢂ 0.005/33 000), and
the TOF value is calculated to be 26 hꢁ1 where 33 000 is the
molecular weight of the enzyme. These values for the poor
substrate 1a are much smaller than those reported for the same
enzyme toward a good substrate, 1-phenylethanol (TTN = 7800
in 4.5 h; TOF = 1700 hꢁ1).13 On the other hand, Table 1
indicates that the I287F/I290A double mutant has much
higher values for 1a (entry 4) than the wild-type enzyme;
Scheme 1
mechanism accelerating (R)-1a, such as the attractive inter-
action between Phe287 and the alkyl chain of (R)-1a, which
was supported by the kinetic constants and the experiments
with partially fluorinated substrates as described later.
Next, we tried to create a second-generation biocatalyst.
Manual docking suggested that Ile290 comes into contact with
a part of the alkyl chain of (R)-1a (Fig. 1c). We expected
that the I290A mutation would make space to accommodate
(R)-1a nicely in the transition state (Fig. 1d), leading to higher
activity. As expected, the I287F/I290A double mutant exerted
much higher activity and enantioselectivity for 1a and 1b
(Table 1, entries 4 and 9). In both cases, the conversions
reached almost 50% within a few hours, and the E values
exceeded 200. In sharp contrast with the I287F/I290A double
mutant, the I287F/I290F double mutant showed very poor
activity and enantioselectivity for 1a (entry 5). The I290A
mutant had catalytic activity and enantioselectivity intermediate
between those of the wild-type enzyme and the I287F/I290A
double mutant, as is the case for the I287F mutant (entry 6). It
is therefore most likely that the second I290A mutation
eliminated the steric hindrance to enhance the reactivity of
(R)-1a or (R)-1b.
TTN = 2300 in 2.5 h and TOF = 940 hꢁ1
.
To gain a deeper insight, we next determined the kinetic
constants for the wild-type and mutant enzymes toward each
enantiomer of 1a. The kinetic constants determined are
summarized in Table 2, which are helpful for understanding
the effect of each mutation on the enzymatic activity. First, the
enantioselectivity resulted mainly from the difference in the
Vmax value (in other words, the transition state) as demon-
strated previously.14 Importantly, the mutations altered the
Vmax values (the transition state) rather than the Km values
(the substrate-binding step). Moreover, the I287F mutant had
a Vmax value for (R)-1a that was B3-fold higher than that of
the wild-type enzyme, while the I287F mutant had a Vmax
value for (S)-1a that was B2-fold lower than that of the wild-
type enzyme. These results clearly support our proposal on the
synergic effect that Phe287 contributes to both acceleration
of (R)-1a and deceleration of (S)-1a. Furthermore, the
I287F/I290A double mutant had a Vmax value for (R)-1a that
was B5-fold greater than that of the I287F mutant, which
supports our expectation that the second I290A mutation
would eliminate the steric hindrance to enhance the reactivity
of (R)-1a. The kinetic constants in Table 2 thus indicate that the
transition state was manipulated successfully as we expected.
Using these recombinant enzymes, we investigated the
above-mentioned attractive interaction between Phe287 and
the alkyl chain of (R)-1a in more detail. We synthesized
partially fluorinated analogs 1c and 1d and used them as
The difference in activation free energy between both
enantiomers (chiral recognition energy) can be calculated from
D
tion energy between enzymes A and B can be calculated from
RꢁSDGz = ꢁRTln E, and the difference in the chiral recogni-
D
AꢁBDRꢁSDGz = ꢁRTln EA/EB.5,13 For example, the chiral
recognition energies of the I287F mutant and the I287F/I290A
double mutant toward 1a are greater by 1.1 and at least
2.2 kcal molꢁ1, respectively, than that of the wild-type enzyme.
The E value for 1a was changed from 1.4 to 4200 by
Table 1 Kinetic resolution of 1 with wild-type and mutant enzymesa
Entry
1
Lipase
Time/h Conversion (%)b Ec
1
2
3
4
5
6
7
8
9
10
11
12
13
1a Wild-type
1a I287F
1a I287A
1a I287F/I290A
1a I287F/I290F
1a I290A
1b Wild-type
1b I287F
1b I287F/I290A
1c
1c
1c
41
41
41
23
46
20
50
10
41
39
47
47
34
47
50
19
5
32
1.4
Table 2 Kinetic constants for the lipase-catalyzed acylations of 1aa
2.5
4200
4
79
9
71
4200
14
55
1a Vmax/M minꢁ1 mg(lipase)ꢁ1 Km/M
41
41
41
22
4
41
22
4
Lipase
Wild-type
Wild-type
I287F
(R) (5.0 ꢃ 0.4) ꢂ 10ꢁ7
(S) (5.9 ꢃ 0.8) ꢂ 10ꢁ8
(R) (1.6 ꢃ 0.1) ꢂ 10ꢁ6
(S) (2.7 ꢃ 0.4) ꢂ 10ꢁ8
(1.9 ꢃ 0.4) ꢂ 10ꢁ1
(4.9 ꢃ 1.0) ꢂ 10ꢁ1
(3.5 ꢃ 0.3) ꢂ 10ꢁ1
(3.7 ꢃ 0.7) ꢂ 10ꢁ1
(1.1 ꢃ 0.2) ꢂ 10ꢁ1
(3.1 ꢃ 0.5) ꢂ 10ꢁ1
I287F
Wild-type
I287F
I287F/I290A
I287F/I290A (R) (7.7 ꢃ 0.5) ꢂ 10ꢁ6
I287F/I290A (S) (1.0 ꢃ 0.1) ꢂ 10ꢁ8
4200
10
1d I287F/I290A 75
a
For the detailed procedure, see ESI. Because of the heterogeneous
reaction, the nonlinear least-squares method was applied to the
Michaelis–Menten type of equation: v0 = Vmax (E)mg [S]0/(Km + [S]0).
The Vmax value, which is normalized by the weight of the immobilized
enzyme powder (E)mg, corresponds to the kcat value in homogeneous
enzymatic reactions.
a
Conditions: immobilized lipase (700 mg, 0.5% (w/w) enzyme/
Toyonite-200M), 1 (0.50 mmol), vinyl acetate (1.0 mmol), molecular
b
sieves 3 A (three pieces), dry i-Pr2O (5 mL), 30 1C. Conversion
calculated from c = ee(1)/(ee(1) + ee(2)). Calculated from E =
c
ln[1 ꢁ c(1 + ee(2))]/ln[1 ꢁ c(1 ꢁ ee(2))].
ꢀc
This journal is The Royal Society of Chemistry 2010
Chem. Commun., 2010, 46, 5440–5442 | 5441

the SC-NMR Laboratory of Okayama University for the
measurement of NMR spectra.
Notes and references
1 (a) Enzyme Catalysis in Organic Synthesis, ed. K. Drauz and
H. Waldmann, Wiley-VCH, Weinheim, 2nd edn, 2002;
(b) K. Faber, Biotransformations in Organic Chemistry, Springer-
Verlag, Berlin, 5th edn, 2004; (c) U. T. Bornscheuer and
R. J. Kazlauskas, Hydrolases in Organic Synthesis, Wiley-VCH,
Weinheim, 2nd edn, 2006; (d) Industrial Biotransformations,
ed. A. Liese, K. Seelbach and C. Wandrey, Wiley-VCH, Weinheim,
2nd edn, 2006; (e) Biocatalysis in the Pharmaceutical and Bio-
technology Industries, ed. R. N. Patel, CRC, Florida, 2007;
(f) Practical Methods for Biocatalysis and Biotransformations, ed.
J. Whittall and P. Sutton, John Wiley & Sons, West Sussex, 2010.
2 P. F. Mugford, U. G. Wagner, Y. Jiang, K. Faber and
R. J. Kazlauskas, Angew. Chem., Int. Ed., 2008, 47, 8782–8793.
3 (a) M. T. Reetz, L.-W. Wang and M. Bocola, Angew. Chem., Int.
Ed., 2006, 45, 1236–1241; (b) M. D. Mihovilovic, F. Rudroff,
A. Winninger, T. Schneider, F. Schulz and M. T. Reetz, Org.
Lett., 2006, 8, 1221–1224; (c) M. T. Reetz and S. Wu, J. Am. Chem.
Soc., 2009, 131, 15424–15432.
Fig. 2 (a) The CH/p interactions in the transition-state of the I287F/
I290A double mutant-catalyzed acylation of (R)-1a. (b) The steric
repulsion in the transition-state of the I287F/I290A double mutant-
catalyzed acylation of (R)-1d.
probes. For example, the CH/p interaction16 cannot take place
at the fluorinated moiety of the substrate, which may enable us
to specify the kind and position of the attractive interaction.
As shown in Table 1, the I287F mutant showed higher activity
and enantioselectivity for 1c than the wild-type enzyme
(entries 10 and 11), and the I287F/I290A double mutant achieved
much higher values (entry 12). This trend is quite similar to that
observed for the non-fluorinated counterpart 1a, which strongly
suggests that the terminal methyl group of 1a experienced no
specific, attractive interactions. In contrast, the catalytic activity
and enantioselectivity of the I287F/I290A double mutant toward
1d were found to be very low (entry 13), and a clear difference
was observed between 1d (entry 13) and 1a (entry 4) or 1c
(entry 12), which strongly suggests that the fluorinated methylene
moiety of (R)-1d underwent a severe steric repulsion (Fig. 2b).
Based on these results and computer representations, we
suppose that the hydrogen atoms at the o ꢁ 1 position (next
to the terminal methyl group) and more weakly at the o ꢁ 3
position of (R)-1a participate in the CH/p interaction with the
phenyl group of Phe287 as shown in Fig. 2a.
4 (a) S. Bartsch, R. Kourist and U. T. Bornscheuer, Angew. Chem.,
Int. Ed., 2008, 47, 1508–1511; (b) H. Jochens, K. Stiba, C. Savile,
R. Fujii, J.-G. Yu, T. Gerassenkov, R. J. Kazlauskas and
U. T. Bornscheuer, Angew. Chem., Int. Ed., 2009, 48, 3532–3535.
5 S. Park, K. L. Morley, G. P. Horsman, M. Holmquist, K. Hult and
R. J. Kazlauskas, Chem. Biol., 2005, 12, 45–54.
6 (a) R. Carr, M. Alexeeva, M. J. Dawson, V. Gotor-Fernandez,
´
C. E. Humphrey and N. J. Turner, ChemBioChem, 2005, 6,
637–639; (b) N. J. Turner, Nat. Chem. Biol., 2009, 5, 567–573.
7 (a) D. Rotticci, J. C. Rotticci-Mulder, S. Denman, T. Norin and
K. Hult, ChemBioChem, 2001, 2, 766–770; (b) A. O. Magnusson,
M. Takwa, A. Hamberg and K. Hult, Angew. Chem., Int. Ed.,
2005, 44, 4582–4585.
8 C. Gentner, R. D. Schmid and J. Pleiss, Colloids Surf., B, 2002,
26, 57–66.
9 R. Weis, R. Gaisberger, W. Skranc, K. Gruber and A. Glieder,
Angew. Chem., Int. Ed., 2005, 44, 4700–4704.
In summary, only two mutations (I287F/I290A) dramati-
cally enhanced both catalytic activity and enantioselectivity
toward poor substrates 1a–1c by the synergic effect: Phe287
contributed to both acceleration of the (R)-enantiomer and
deceleration of the (S)-enantiomer, while Ala290 made space
to facilitate the acylation of the (R)-enantiomer. To the best
of our knowledge, this is the first example of introducing the
CH/p interaction in the transition state to promote the enzymatic
reaction. Almost all the mutant enzymes that we have so far
prepared are presented here, and the enzyme structure has
not yet been optimized, for example, by a random approach
such as saturation mutagenesis.3–6 Nevertheless, the impact of
the mutations was found to be great enough. Therefore, the
present results clearly demonstrate the efficiency and power of
the rational design approach to the creation of an excellent
biocatalyst.
10 (a) A. E. Tripp, D. S. Burdette, J. G. Zeikus and R. S. Phillips,
J. Am. Chem. Soc., 1998, 120, 5137–5141; (b) C. Heiss,
M. Laivenieks, J. G. Zeikus and R. S. Phillips, J. Am. Chem.
Soc., 2001, 123, 345–346.
11 (a) F. P. Seebeck, A. Guainazzi, C. Amoreira, K. K. Baldridge and
D. Hilvert, Angew. Chem., Int. Ed., 2006, 45, 6824–6826;
(b) M. D. Toscano, M. M. Muller and D. Hilvert, Angew. Chem.,
Int. Ed., 2007, 46, 4468–4470.
¨
12 (a) Y. Ijima, K. Matoishi, Y. Terao, N. Doi, H. Yanagawa and
H. Ohta, Chem. Commun., 2005, 877–879; (b) Y. Terao,
K. Miyamoto and H. Ohta, Chem. Commun., 2006, 3600–3602.
13 T. Ema, T. Fujii, M. Ozaki, T. Korenaga and T. Sakai, Chem.
Commun., 2005, 4650–4651.
14 (a) T. Ema, J. Kobayashi, S. Maeno, T. Sakai and M. Utaka, Bull.
Chem. Soc. Jpn., 1998, 71, 443–453; (b) T. Ema, M. Jittani,
K. Furuie, M. Utaka and T. Sakai, J. Org. Chem., 2002, 67,
2144–2151; (c) T. Ema, K. Yamaguchi, Y. Wakasa, A. Yabe,
R. Okada, M. Fukumoto, F. Yano, T. Korenaga, M. Utaka and
T. Sakai, J. Mol. Catal. B: Enzym., 2003, 22, 181–192.
15 C.-S. Chen, Y. Fujimoto, G. Girdaukas and C. J. Sih, J. Am.
Chem. Soc., 1982, 104, 7294–7299.
This work was supported by a Grant-in-Aid for Scientific
Research from the Japan Society for the Promotion of Science
(JSPS) and by The Sumitomo Foundation. We are grateful to
16 M. Nishio, M. Hirota and Y. Umezawa, The CH/p Interaction:
Evidence, Nature, and Consequences, Wiley-VCH, New York, 1998.
ꢀc
This journal is The Royal Society of Chemistry 2010
5442 | Chem. Commun., 2010, 46, 5440–5442