How substrate solvation contributes to the enantioselectivity of subtilisin toward secondary alcohols

Article abstract of DOI:10.1021/ja0528937

The current rule to predict the enantiopreference of subtilisin toward secondary alcohols is based on the size of the substituents at the stereocenter and implies that the active site contains two differently sized pockets for these substituents. Several experiments are inconsistent with the current rule. First, the X-ray structures of subtilisin show there is only one pocket (the S₁- pocket) approximately the size of a phenyl group to bind secondary alcohols. Second, the rule often predicts the incorrect enantiomer for reactions in water. To resolve these contradictions, we refine the current rule to show that subtilisin binds only one substituent of a secondary alcohol and leaves the other in solvent. To test this refined empirical rule, we show that the enantioselectivity of a series of secondary alcohols in water varied linearly with the difference in hydrophobicity (log P/P₀) of the substituents. This hydrophobicity difference accounts for the solvation of one substituent in water. Copyright

Full text of DOI:10.1021/ja0528937

Published on Web 08/13/2005
How Substrate Solvation Contributes to the Enantioselectivity of Subtilisin
toward Secondary Alcohols
Christopher K. Savile and Romas J. Kazlauskas*
Department of Chemistry, McGill UniVersity, 801 Sherbrooke Street West, Montre´al, Que´bec, Canada, H3A 2K6,
and Department of Biochemistry, Molecular Biology & Biophysics and the Biotechnology Institute, UniVersity of
Minnesota, 1479 Gortner AVenue, St. Paul, Minnesota 55108
Received May 3, 2005; E-mail: rjk@umn.edu
Enantioselective enzymes, especially hydrolases, are useful cata-
lysts to make enantiomerically pure pharmaceuticals, agrochemi-
cals, and fine chemicals.1 Several empirical rules predict which sub-
strate/hydrolase combinations work best. For example, a rule to
predict the enantiopreference of subtilisin toward secondary alcohols
is based on the size of the substituents at the stereocenter (Figure
1a).2,3 This model implies that subtilisin has two differently sized
pockets for these substituents, but several experiments are incon-
sistent with this rule. First, the X-ray crystal structure shows only
Figure 1. Empirical rules that predict the enantiopreference of subtilisins
one pocket (the S1′ pocket) to bind secondary alcohols.4 Second,
toward secondary alcohols. (a) A rule based on relative substituent size,
where L is the large substituent and M is the medium substituent, is reliable
the rule often predicts the incorrect enantiomer for reactions in
water. In this communication, we resolve this contradiction with a
more general rule that shows subtilisin binds only one substituent
of a secondary alcohol and leaves the other in solvent. This refined
rule allows quantitative design of enantioselective reactions and
rationalizes why solvent alters the enantioselectivity.
in organic solvent. (b) A revised rule that is reliable in water as well. One
substituent (RSOLV) remains in solvent, while the other (RS1′) binds in a
hydrophobic pocket. (c) In water, the nonpolar aryl group of alcohol 3 favors
binding in the S1′ pocket, thus favoring the (R)-enantiomer. (d) An isosteric
substrate alcohol 5 contains a polar aryl group that favors the water-solvated
orientation, thus favoring the (S)-enantiomer.
X-ray crystal structures of subtilisin reveal one pocket (the S1′ poc-
ket) that binds the alcohol portion of an ester. Molecular modeling
of a tetrahedral intermediate for subtilisin E-catalyzed hydrolysis
of 1a reveals that (S)-1a places the methyl group in the S1′ pocket,
while (R)-1a places the phenyl group in this pocket (see SI Figure
S2). In both cases, the other substituent remains in the solvent. The
S1′ pocket is a shallow crevice large enough to accommodate para-
substituted aryl groups, but too small for multisubstituted aryl groups.
Although the rule in Figure 1a is reliable for reactions in organic
solvents,2,3 it is not reliable in water. In organic solvent, the subtili-
sin-catalyzed transesterification of secondary alcohols 1-13 with di-
hydrocinnamic acid vinyl ester favored the predicted (S)-enantiomer
for 26 out of 29 reactions with varying enantioselectivity (E ) 1.5
to 66; see SI Tables S2-S4). In water, however, subtilisin favored
hydrolysis of the opposite (R)-enantiomer in most cases: 20 out
of 33 reactions (Table 1).
To resolve these contradictions, we propose a revised rule for
the enantiopreference of subtilisins with secondary alcohols (Figure
1b). This rule places one substituent in solvent and limits the size
of the other substituent to approximately the size of a phenyl group.
This rule predicts that solvation of one substituent contributes to
the enantiopreference of subtilisin. In particular, placing a nonpolar
substituent in water is unfavorable. Reactions in water involving
methyl and nonpolar aryl substituents will favor the nonpolar aryl
substituent in the S1′ pocket, opposite to that predicted based on
size alone. Thus, the revised rule predicts that subtilisin favors the
(R)-enantiomer of 3a in water, but the (S)-enantiomer in organic
solvents. On the other hand, with a polar aryl group such as that in
5a (4-pyridine N-oxide), the (S)-enantiomer is favored both in water,
where solvation of the pyridine N-oxide is favorable, and in organic
solvent, where placing the pyridine N-oxide in the solvent avoids
steric interactions in the S1′ pocket.
Table 1. Enantioselectivity of Subtilisin BPN′-, Carlsberg-, and
E-catalyzed Hydrolysis of 1a-13aa
enantioselectivity, Eb
log P/P0
diff.c
subtilisin
E
subtilisin
subtilisin
BPN
entry
substrate
Carlsberg
′
1
1a
2a
3a
4a
5a
+1.1
-0.3
+1.6
+1.9
-2.2
+3.9
+2.6
+0.7
-3.1
7.0 (R)
1.5 (R)
16 (R)
7.7 (R)
4.5 (S)
>150 (R)e
110 (R)
2.8 (R)
5.5 (S)
20 (R)
17 (R)
1.8 (R)
n.r.h
1.2 (R)
1.7 (S)
1.1 (S)
2.2 (S)
3.1 (S)
11 (R)e
2.5 (R)
2.3 (S)
3.6 (S)
2.0 (R)
1.7 (S)
3.1 (S)
n.r.
15 (R)
2.6 (R)
37 (R)
9.9 (R)
2.5 (S)
50 (R)f
109 (R)
4.2 (R)
6.2 (S)
18 (R)
4.9 (R)
3.1 (S)
n.r.
2
3
4
5
6
N-HCinn-p-TSd
7
6a
8
7a
9
8a
10
11
12
13
14
9ag
10bg
11ag
12b
13a
n.r.
n.r.
n.r.
a See SI Tables S2-S5 for complete details. b Relative rate of the fast
vs slow enantiomer.5 c Substituent hydrophobicity difference (log P/P0RLarge
substituent - log P/P0RMedium substituent). d N-Dihydrocinnamoyl-p-toluenesulfi-
namide. This is a secondary alcohol ester isostere with the methine replaced
with sulfur and the methyl replaced with oxygen. e Reference 6. f Reference
7. g Not included in Figure 2 because one substituent is much large than
phenyl. h No reaction.
aryl groups (1a, 3a, 6a, 9a, 10a, and 11b) for 14 out of 18 reactions
and the (S)-enantiomer for substrates with hydrophilic aryl groups
(2a, 5a and 8a) for eight of nine reactions. It is difficult to predict
the favored enantiomer for moderately hydrophilic aryl groups (4a
and 7a), and indeed the enantioselectivity in these cases is low to
The revised rule in Figure 1b correctly predicted the (R)-
enantiomer for reactions in water for substrates with hydrophobic
9
12228
J. AM. CHEM. SOC. 2005, 127, 12228-12229
10.1021/ja0528937 CCC: $30.25 © 2005 American Chemical Society

C O M M U N I C A T I O N S
increases from E ) 3 (S) in acetonitrile to E ) 54 (S) in benzene,
likely due to better solvation of the solvent-exposed phenyl substit-
uent in benzene as compared to acetonitrile.2 Researchers previously
explained changes in enantioselectivity of subtilisins toward chiral
acids using a similar rationale for solvation of the solvent-exposed
groups,9,10 but our model is the first to use this approach for chiral
alcohols.
Unlike subtilisins, which bind substrates in an extended confor-
mation,11 lipases bind substrates in a folded conformation.12 This
folding and the deeper hydrophobic pockets in lipases place both
substituents of typical secondary alcohols in hydrophobic pockets
that substantially shield the substituents from the solvent.13 For this
reason, the enantioselectivity of lipase-catalyzed resolutions of
secondary alcohols shows less variation with changes in substituent
polarity14 or solvent.15 The SI shows that lipase from Burkholderia
cepacia (PCL) favors the (R)-enantiomer for all compounds in Table
1 and shows no reversal in enantiopreference upon changing from
water to organic solvent.
In conclusion, this revised model of the enantioselectivity of
subtilisins toward secondary alcohols is consistent with the structure
of subtilisin, rationalizes why enantioselectivity changes and even
reverses with changes in solvent, and provides a strategy to increase
enantioselectivity by modifying the substrate.
Acknowledgment. We thank McGill University and University
of Minnesota for financial support, Dr. F. Schendel and R.
Dillingham (University of Minnesota Biotechnology Institute) for
the large-scale fermentation and purification of subtilisins BPN′
and E and the Minnesota Supercomputing Institute for access and
support for molecular modeling computers and software.
Supporting Information Available: Synthesis of compounds 1a-
13a, determination of absolute configuration, preparation of subtilisin
E and BPN′, molecular modeling details for 1a, and enantioselectivity
data. This material is available free of charge via the Internet at http://
pubs.acs.org.
Figure 2. Differences in substituent hydrophobicity affect the enantiose-
lectivity subtilisins toward secondary alcohols. All reactions are in water.
This plot does not include substrates 9a-11a because their substituted aryl
groups are too large to fit in the S1′ pocket of subtilisins. (a) Enantiose-
lectivity data from Table 1 is given in energy using ∆∆Gq ) -RT ln E. (b)
Hydrophobicity partition coefficient (log P/P0).
moderate (E ) 2.2 to 9.9). With nonpolar substituents and nonpolar
solvents, the rule simplifies to the previous rule in Figure 1a.
The revised rule also suggests a quantitative link between enan-
tioselectivity and solvation of the substituents. For example, reaction
of dihydrocinnamoyl esters 1a-13a with subtilisin E showed that
the enantioselectivity toward secondary alcohol esters in water varied
linearly with the difference in hydrophobicity (log P/P0)8 between
the large aryl substituent and the methyl group (Figure 2). This
hydrophobicity difference accounts for the solvation of one sub-
stituent in water and the other in the hydrophobic S1′ pocket. In-
creases in hydrophobicity of the aryl group favored the (R)-enan-
tiomer, while decreases favored the (S)-enantiomer. For example,
subtilisin E-catalyzed hydrolysis of 6a containing the nonpolar
4-isopropylphenyl group gave (R)-6 with E ) 110, while 7a con-
taining the more polar, but similar sized 4-nitrophenyl group gave
(R)-7 with lower enantioselectivity (E ) 2.8), and 8a containing
the hydrophilic carboxylate group gave the opposite enantiomer
(S)-8 with E ) 5.5. Subtilisin BPN′ showed similar enantioselec-
tivity toward substrates 1a-13a consistent with the similar S1′ poc-
ket in both cases. The enantioselectivity of subtilisin Carlsberg was
lower, and the change in enantioselectivity (slope of the line in
Figure 2) varied less with changes in substituent hydrophobicity,
presumably due to weaker interaction between substrate and S1′
pocket.
This revised model also predicts that increasing the polarity
difference between the substituents will increase the enantioselec-
tivity of subtilisins. Consistent with this prediction, subtilisin shows
high enantioselectivity toward arylsulfinamides (entry 6).6 This
toluenesulfinamide is a polar isostere of 3a, where a polar oxygen
replaces the methyl group and thereby increases the difference in
polarity between the two substitutents (log P difference ) +1.6
for 3a and +3.9 for the sulfinamide). The enantioselectivity of the
subtilisin-E-catalyzed hydrolysis increases from E ) 16 for 3a to
E ) >150 for the sulfinamide.
References
(1) (a) Faber, K. Biotransformations in Organic Chemistry, 5th ed.; Springer-
Verlag: Berlin, 2004. (b) Drauz, K.; Waldman, H., Eds. Enzyme Catalysis
in Organic Synthesis; Wiley-VCH: Weinheim, Germany, 2002. (c)
Bornscheuer, U. T.; Kazlauskas, R. J. Hydrolases in Organic Synthesis:
Regio- and StereoselectiVe Biotransformations; Wiley-VCH: Weinheim,
Germany, 1999.
(2) Fitzpatrick, P. A.; Klibanov, A. M. J. Am. Chem. Soc. 1991, 113, 3166.
(3) Kazlauskas, R. J.; Weissfloch, A. N. E. J. Mol. Catal. B: Enzym. 1997,
3, 65.
(4) (a) Jain, S. C.; Shinde, U.; Li, Y.; Inouye, M.; Berman, H. M. J. Mol.
Biol. 1998, 284, 137. The crystal structure of subtilisin Carlsberg, where
crystals have been soaked in acetonitrile or dioxane to be nearly
indistinguishable from the structure determined in water: (b) Fitzpatrick,
P. A.; Steinmetz, A. C. U.; Ringe, D.; Klibanov, A. M. Proc. Natl. Acad.
Sci. U.S.A. 1993, 90, 8653. (c) Schmitke, J. L.; Stern, L. J.; Klibanov, A.
M. Proc. Natl. Acad. Sci. U.S.A. 1997, 94, 4250.
(5) Chen, C. S.; Fujimoto, Y.; Girdaukas, G.; Sih, C. J. J. Am. Chem. Soc.
1982, 104, 7294.
(6) Savile, C. K.; Magloire, V. M.; Kazlauskas, R. J. J. Am. Chem. Soc. 2005,
127, 2104.
(7) Mugford, P. F.; Magloire, V. M.; Kazlauskas, R. J. J. Am. Chem. Soc.
2005, 127, 6536.
(8) Log P/P0 is the substituent hydrophobicity partition coefficient relative
to hydrogen.
(9) (a) Margolin, A. L.; Tai, D.-F.; Klibanov, A. M. J. Am. Chem. Soc. 1987,
109, 7885. (b) Sakurai, T.; Margolin, A. L.; Russell, A. J.; Klibanov, A.
M. J. Am. Chem. Soc. 1988, 110, 7236. (c) Tawaki, S.; Klibanov, A. M.
J. Am. Chem. Soc. 1992, 114, 1882.
(10) Wangikar, P. P.; Rich, J. O.; Clark, D. S.; Dordick, J. S. Biochemistry
1995, 34, 12302.
(11) Tyndall, J. D. A.; Tessa Nall, T.; Fairlie, D. P. Chem. ReV. 2005, 105, 973.
(12) For example: Mezzetti, A.; Schrag, J.; Cheong, C. S.; Kazlauskas, R. J.
Chem. Biol. 2005, 12, 427.
(13) Cygler, M.; Grochulski, P.; Kazlauskas, R. J.; Schrag, J. D.; Bouthillier,
F.; Rubin, B.; Serreqi, A. N.; Gupta, A. K. J. Am. Chem. Soc. 1994, 116,
3180.
Increasing the hydrophobicity difference by adding nonpolar sub-
stituents to the aryl group is not a good strategy to increase enantio-
selectivity because it creates a substituent too large for the S1′
pocket. For example, compounds 9a-11a contain very large aryl
groups. The poor fit of this aryl group in the S1′ pocket destabilizes
reaction of the (R)-enantiomer. Subtilisins favor the (S)-enantiomer
in these cases, but the enantioselectivity is usually low.
This model also rationalizes how changing the organic solvent
can increase the enantioselectivity of subtilisins. The enantioselec-
tivity of subtilisin Carlsberg toward 1-phenethyl alcohol (1) increases
(14) Ho¨nig, H.; Shi, N.; Polanz, G. Biocatalysis 1994, 9, 61.
(15) (a) Parida, S.; Dordick, J. S. J. Am. Chem. Soc. 1991, 113, 2253. (b)
Cernia, E.; Palocci, C.; Soro, C. Chem. Phys. Lipids 1998, 93, 157.
JA0528937
9
J. AM. CHEM. SOC. VOL. 127, NO. 35, 2005 12229