Temperature and solvent effects on enzyme stereoselectivity: Inversion temperature in kinetic resolutions with lipases

Article abstract of DOI:10.1039/b006086k

Eyring plots of InE vs. 1/T show inversion temperature in kinetic resolution by lipases, demonstrating that clustering effects in the substrate solvation manage the enzymatic selectivity.

Full text of DOI:10.1039/b006086k

Temperature and solvent effects on enzyme stereoselectivity: inversion
temperature in kinetic resolutions with lipases
Gianfranco Cainelli,* Valeria De Matteis, Paola Galletti, Daria Giacomini* and Paolo Orioli
Department of Chemistry ‘G. Ciamician’, University of Bologna Italy. E-mail: cainelli@ciam.unibo.it,
giacomin@ciam.unibo.it
Received (in Liverpool, UK) 24th July 2000, Accepted 16th October 2000
First published as an Advance Article on the web
Eyring plots of lnE vs. 1/T show inversion temperature in
kinetic resolution by lipases, demonstrating that clustering
effects in the substrate solvation manage the enzymatic
selectivity.
two independent experimental observations concerning the
same solvation-clustering phenomenon.
We report here the first experimental example of inversion
temperature in the field of stereoselective enzymatic reactions14
and also confirm the presence of interconversion of solute–
solvent supramolecules by means of VT 13C NMR analysis.
As part of an ongoing project on the total synthesis of new b-
lactam antibiotics,15 we became interested in the kinetic
resolution of racemic 1-N-(propen-1-yl)-3-N-(phenylacetoxy)-
aminoazetidin-2-one 1 by PGA (supported on Eupergit)
(Scheme 1). The experiments were carried out by incubating the
substrate in phosphate buffer (pH = 7.8) and acetone (2+1 ratio)
at different temperatures. Conversion of the reaction was
monitored by titration of the phenylacetic acid formed with
NaOH solution and the enantiomeric excesses of 3 and
unreacted 1 were determined by HPLC analysis on a chiral
column.16 With an increase in temperature, we observed an
initial linear lowering of the selectivity ratio E17 until it reached
a minimum value of 24 corresponding to T = 301 K (Fig. 1).
Upon further warming, E increased, to reach a maximum value
of 130 at T = 328 K. Least-squares analysis of these results
enabled us to determine that Tinv = 301 K. Notably, the highest
enantioselectivity is reached at the highest temperature.18
Usually, E = 100 is considered the minimum satisfactory value
for the industrial application of an enzymatic process. Thus, in
our case, a slight variation in temperature made an inadequate
enzymatic method very efficient.
Over the past few years, the use of enzymes in organic synthesis
has become increasingly important owing to the basic discovery
that enzymatic reactions can occur in organic solvents as well as
in water.1 Indeed, enzymatic selectivity is remarkably depend-
ent on the organic solvent2 and many attempts have been made
to explain this influence.3 However, less attention has been paid
to the effect of temperature on selectivity.4
We report here the results of our studies on the effect of
temperature on the kinetic resolution of racemic b-lactam 1 by
PGA (Penicillin G-Acylase). Furthermore, we also reanalyzed
another enzymatic process that has been reported previously,
involving the kinetic resolution of azirinol 2 by lipase PS.5
Over the past decade, the dependence of stereoselectivity on
temperature has been evaluated in several non-enzymatic
reactions; e.g. cis-dihydroxylation of olefins,6 reduction of
ketones7 and nucleophilic addition to imines8 and carbonyl
compounds.9 By plotting the logarithm of selectivity vs. the
reciprocal of T, according to the modified Eyring equation [eqn.
(1)], a peculiar behavior is sometimes observed: the existence of
two linear regions intersecting at a point defining a temperature
denoted the inversion temperature (Tinv), whose significance is
still a matter of debate.10
Scheme 1
ln S = 2DDH‡/RT + DDS‡/R
(1)
Next, we performed a series of VT 13C NMR analyses of 1 in
deuterated water and d6-acetone (2+1 ratio). The d value of the
phenylacetoxy CNO chemical shift decreases with an increase in
temperature between 278 and 323 K, and two linear regions may
be clearly recognized by least-squares analysis (Fig. 2). TNMR
occurs at T = 306 K and corresponds to Tinv, within the
experimental error. We looked for other studies on temperature
dependent selectivity in enzymatic processes in the literature. In
1997 Sakai et al.5 described the kinetic resolution of racemic
phenyl-2H-azirine-2-methanol 2 in diethyl ether. The selective
This break point leads to two sets of activation parameters:
one for T < Tinv and the other for T > Tinv. Our preceding
studies demonstrated that the presence of an inversion tem-
perature accounts for a particular effect of the reaction solvent
on the stereoselectivity. We have proposed that Tinv constitutes
a transition temperature between two different solute–solvent
clusters, which act as two distinct supramolecules with different
thermodynamic properties, reactivity and, therefore, different
trends of stereoselectivity with temperature.11
We have recently shown12 that in nucleophilic addition to
aldehydes, Tinv depends mainly on the aldehyde–solvent couple
and its value may be determined by 13C NMR analysis using
variable temperature (VT) experiments. NMR techniques
provide a powerful method for investigating the local electronic
environment of the molecular structure, and nuclear shielding is
a sensitive probe of intermolecular interactions and solvent
effects.13 We found that the variation of the CNO chemical shift
of the studied aldehydes with temperature presents two linear
trends, whose intersection (referred to here as TNMR) lies near
the Tinv of the same substrate–solvent system, within the
experimental error. Therefore, we considered that TNMR and
Tinv might have a common origin, and could in fact represent
Fig. 1 Eyring plot for the selectivity ratio E in the kinetic resolution of 1 by
PGA at different temperatures in H2O–acetone (2+1).
DOI: 10.1039/b006086k
Chem. Commun., 2000, 2351–2352
This journal is © The Royal Society of Chemistry 2000
2351

above and below the Tinv, it follows that those solute–solvent
clusters contribute to differentiate the Gibbs free energies of the
two diastereomeric substrate–enzyme complex transition states.
Thus our data provide evidence that solvent molecules are still
present in substrate–enzyme transition states.
Our experimental data clearly suggest that enzymatic reac-
tions are only a particular case of a more general phenomenon.
Solvation always plays a fundamental role in stereoselectivity
and the substrate cannot be considered a mere isolated species
in solution, but rather a part of a more complex and well defined
solute–solvent cluster.
This work was supported by MURST and by the University
of Bologna (funds for selected topics). Penicillin G Acylase was
kindly supplied by Recordati (Italy).
Fig. 2 Plot of the 13C chemical shift of the phenylacetoxy CNO vs. T for 1
in D2O–d6-acetone (2+1).
acetylation reaction promoted by lipase Amano PS in the
presence of vinyl acetate as an acylating agent, proceeded
smoothly until 233 K, with progressively higher selectivity (Fig.
3). At T = 233 K, the enantioselection reached a maximum (E
= 99) and then decreased. A least-squares analysis of the
tabulated data allowed us to determine a Tinv at T = 229 K. To
test for the presence of a TNMR, we prepared 2 by the reported
method19 and studied the evolution of d vs. T in d10-diethyl
ether. The results for the quaternary aromatic carbon20 are
reported in Fig. 4. Again, two linear regions exist and TNMR
occurs at T = 222 K, quite near to the corresponding Tinv. The
different rates of variation of the chemical shift d vs. T indicate
an abrupt change in solute–solvent interactions, which reflects a
change in substrate solvation.21 The same solvation change
causes the presence of the Tinv in an Eyring plot which regards
the same substrate in the same solvent of the NMR analysis.
Notes and references
1 For a review, see: C. R. Wescott and A. M. Klibanov, Biochim. Biophys.
Acta, 1994, 12, 1.
2 A. Zaks and A. M. Klibanov, J. Am. Chem. Soc., 1986, 108, 2767.
3 T. Ke, C. R. Wescott and A. M. Klibanov, J. Am. Chem. Soc., 1996, 118,
3366; T. Ke and A. M. Klibanov, J. Am. Chem. Soc., 1998, 120,
4259.
4 G. Talsky, Angew. Chem., Int. Ed. Engl., 1971, 10, 548; E. Keinan, E. K.
Hafeli, K. K. Seth and R. Lamed, J. Am. Chem. Soc., 1986, 108, 162;
Y.-Y. Huang, T. Hara, S. Sligar, M. J. Coon and T. Kimura,
Biochemistry, 1986, 25, 1390; Y.-Y. Chang, R. D. Scott and D. J.
Graves, Biochemistry, 1986, 25, 1932; I. H. Segel, Enzyme Kinetics:
behavior and analysis of rapid equilibration and steady-state enzyme
systems, Wiley, New York, 1993; R. S. Phillips, Trends Biotechnol.,
1996, 14, 13.
5 T. Sakai, I. Kawabata, T. Kishimoto, T. Ema and M. Utaka, J. Org.
Chem., 1997, 62, 4906.
6 T. Göbel and K. B. Sharpless, Angew. Chem., Int. Ed. Engl., 1993, 32,
1329.
7 G. B. Stone, Tetrahedron: Asymmetry, 1994, 5, 465.
8 G. Cainelli, D. Giacomini and M. Walzl, Angew. Chem., Int. Ed. Engl.,
1995, 34, 2150; G. Cainelli, D. Giacomini and P. Galletti, Eur. J. Org.
Chem., 1999, 61.
9 G. Cainelli, D. Giacomini, P. Galletti and A. Marini, Angew. Chem., Int.
Ed. Engl., 1996, 35, 2849.
10 H. Buschmann, H.-D. Scharf, N. Hoffmann and P. Esser, Angew. Chem.,
Int. Ed. Engl., 1991, 30, 477; K. J. Hale and J. H. Ridd, J. Chem. Soc.,
Perkin Trans. 2, 1995, 1601.
11 G. Cainelli, D. Giacomini and P. Galletti, Chem. Commun., 1999,
567.
Fig. 3 Eyring plot for the selectivity ratio E in the kinetic resolution of 2 by
lipase Amano PS at different temperatures in diethyl ether.
12 G. Cainelli, D. Giacomini, P. Galletti and P. Orioli, Angew. Chem., Int.
Ed., 2000, 39, 523.
13 For solvent effects on chemical shift see, for instance: T. Helgaker, M.
Jaszunski and K. Ruud, Chem. Rev., 1999, 99, 293; E. Y. Lau and J. T.
Gerig, J. Am. Chem. Soc., 1996, 118, 1194; E. Y. Lau and J. T. Gerig,
J. Chem. Phys., 1995, 103, 3341 and references therein.
14 Scharf in his review10 interpreted the Tinv on a different basis and
suggested its existence in enzymatic reactions without reporting any
experimental example.
15 G. Cainelli, D. Giacomini, P. Galletti and M. Da Col, Tetrahedron:
Asymmetry, 1997, 8, 3231; G. Cainelli, D. Giacomini and P. Galletti,
Synthesis, 2000, 289.
16 The chiral column S,S-DACH.DNB Lichosorb was furnished by
Professor Gasparrini, University of Rome, Italy.
Fig. 4 Plot of the 13C chemical shift of the quaternary aromatic carbon atom
vs. T for 2 in d10-diethyl ether.
17 The selectivity ratio E is defined as the ratio of the specificity constants
Vmax/KM: C.-S. Chen, Y. Fujimoto, G. Girdaukas and C. J. Sih, J. Am.
Chem. Soc., 1982, 104, 7294.
18 This behavior is more common than usually considered. For some
examples in diastereoselective reactions, see refs. 8 and 11.
19 G. Hortmann, D. A. Robertson and B. K. Gillard, J. Org. Chem., 1972,
37, 322.
Thus, once again the Tinv and the TNMR appear as two
independent experimental results due to the same phenomenon.
Actually, we think that a reorganization between two differently
ordered solvation clusters generates both the Tinv and the TNMR
.
The TNMR reveals the presence of dynamic phenomena on the
ground state of solute–solvent clusters which thus appear to be
much more structured than generally believed. The Tinv reveals
the same dynamic phenomena, which now act on diastereo-
20 The same conclusion is achieved even in the case of the other carbon
atoms.
21 Partial ordering of molecules is possible in solution, and a temperature-
dependent change in this order is already manifested in the nonlinearity
of some spectroscopic properties. See e.g.: J. B. Robert, Mol. Phys.,
1997, 90, 399; M. A. Wendt, J. Meiler, F. Weinhold and T. C. Farrar,
Mol. Phys., 1998, 93, 145.
22 An alternative explanation of Tinv could be a change of the enzyme
conformation induced by temperature. However, the presence of Tinv in
many non-enzymatic reactions6–12 renders our interpretation more
plausible (see for instance ref. 3).
meric transition states leading to a different stereoselectivity
22
below and above the Tinv
.
From the experimental data reported here, it results that even
enzymatic reactions show temperature dependent phenomena
of substrate solvation because of the presence of a Tinv and a
corresponding TNMR, so that even enzymatic reactions experi-
ence reorganization of different solute–solvent clusters depend-
ing on temperature. Because of the difference in selectivity
2352
Chem. Commun., 2000, 2351–2352