Development of an enzyme mimic using self-selection

Article abstract of DOI:10.1002/ijch.201200080

The development of a serine protease model using a self-selection protocol is described. The developed system mimics the regeneration step of an enzyme involved in covalent enzyme catalysis. A transition-state analogue of a transesterification reaction is used to self-select functional groups able to accelerate ester cleavage. It is shown that the insertion of a tertiary amine substituent flanking the reaction center reinforces transition-state stabilization by directing the reactive center towards the self-selected functionality. In addition, the tertiary amine activates a bland (solvent) nucleophile for attack on an ester bond similar to what occurs in a serine protease. A quantitative correspondence is observed between the amplification factors and catalytic activity, illustrating the potential of the dynamic covalent capture strategy to precisely detect and quantify weak noncovalent interactions.

Full text of DOI:10.1002/ijch.201200080

Full Paper
L. J. Prins et al.
DOI: 10.1002/ijch.201200080
Development of an Enzyme Mimic Using Self-Selection
Giulio Gasparini,[a] Marta Dal Molin,[a] Stefano Corrꢀ,[a] Patrizia Galzerano,[a] Paolo Scrimin,[a] and
Leonard J. Prins*[a]
Abstract: The development of a serine protease model
using a self-selection protocol is described. The developed
system mimics the regeneration step of an enzyme involved
in covalent enzyme catalysis. A transition-state analogue of
a transesterification reaction is used to self-select functional
groups able to accelerate ester cleavage. It is shown that
the insertion of a tertiary amine substituent flanking the re-
action center reinforces transition-state stabilization by di-
recting the reactive center towards the self-selected func-
tionality. In addition, the tertiary amine activates a bland
(solvent) nucleophile for attack on an ester bond similar to
what occurs in a serine protease. A quantitative correspond-
ence is observed between the amplification factors and cata-
lytic activity, illustrating the potential of the dynamic cova-
lent capture strategy to precisely detect and quantify weak
noncovalent interactions.
Keywords: dynamic combinatorial chemistry · enzyme catalysis · supramolecular chemistry · thermodynamic control · transition states
1. Introduction
Catalyst discovery is a challenging task because of the
subtle energy differences that determine the efficiency
and selectivity of a catalyst.[1] The search for new catalysts
is facilitated by combinatorial methodologies, which
permit a high-throughput synthesis and screening of cata-
lyst libraries.[2–8] In recent years, supramolecular ap-
proaches towards catalyst discovery are rapidly emerging,
mainly driven by the advantage of self-assembly as a tool
to generate structurally diverse catalyst libraries with
minimal synthetic effort.[9–11] An alternative approach
relies on the use of self-selection protocols for the sponta-
neous identification of a catalyst from a dynamic combi-
natorial library of potential candidates.[12] This relies on
the shift in the thermodynamic equilibrium towards the
optimal catalyst upon exposure of the library to a transi-
tion-state analogue (TSA) of the reaction under scruti-
ny.[13–17] The attractiveness of this approach is twofold.
First, it has the potential of identifying catalysts from
a large virtual library, which eliminates the need of ac-
tually synthesizing and screening all of them. Second,
self-selection is by definition the result of all energetic in-
teractions that occur between catalyst and transition
state, including those that would be difficult to anticipate
in an approach based on rational design. Nonetheless, the
fact that the application of dynamic combinatorial
chemistry for catalyst discovery has so far received much
less attention compared to the development of receptors,
sensors, and materials, indicates that the translation from
transition-state recognition to reactivity is far from trivi-
al.[18] Recently, we have introduced a dynamic covalent
capture strategy that facilitates this process.[19] It relies on
the covalent anchoring of the transition-state analogue to
a scaffold (1b) equipped with an aldehyde group for the
reversible covalent capture of hydrazides, which are able
to develop stabilizing interactions with the transition-
state analogue.[15] The major advantage is that self-selec-
tion is driven by intramolecular interactions, which per-
mits the detection of weak energetic contributions that
would go unobserved in a classical dynamic combinatorial
library.[20] Additionally, the covalent attachment of the
target precisely defines the position where the reaction
has to occur. A similar mode of action is employed by en-
zymes that mechanistically rely on covalent enzyme catal-
ysis.[21] Here, a transient covalent enzyme-substrate inter-
mediate is formed, which is subsequently hydrolyzed to
regenerate the enzyme (Figure 1a). Previously, we have
shown that the phosphonate target (acting as a TSA for
the hydrolysis of a carboxylic ester) in scaffold 1b self-se-
lects hydrazides B and C from a library of hydrazides
(Figure 2).[15] It was indeed shown that a correspondence
exists between the amplification factors and the ability of
the selected hydrazides to accelerate the cleavage of
a neighboring carboxylic ester through transition-state
stabilization. Mechanistically, this resembles the regenera-
tion step of an enzyme involved in covalent enzyme catal-
[a] G. Gasparini, M. Dal Molin, S. Corrꢀ, P. Galzerano, P. Scrimin,
L. J. Prins
Department of Chemical Sciences
University of Padova, Via Marzolo 1
35131 Padova (Italy)
fax: +390498275050
e-mail: leonard.prins@unipd.it
Supporting information for this article is available on the WWW
under http://dx.doi.org/10.1002/ijch.201200080.
122
ꢁ 2013 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Isr. J. Chem. 2013, 53, 122 – 126

Development of an Enzyme Mimic Using Self-Selection
ysis. Here, we show an evolution of our system towards
a model that mechanistically operates in a way that is
more similar to that of an enzyme. It is demonstrated that
decoration of the scaffold molecule with an additional
(tertiary) amine functionality renders the system suscepti-
ble to nucleophilic attack by bland solvent molecules.
Furthermore, the presence of an additional functionality
on the scaffold enhances the catalytic efficiency by direct-
ing the reaction center towards the self-selected function-
ality, thus reinforcing transition-state stabilization. Finally,
a quantitative correspondence is observed between the
amplification factors and catalytic activity.
Figure 1. a) Regeneration of a serine protease upon solvolysis of
the covalent enzyme-substrate intermediate, and b) the same pro-
cess in a synthetic enzyme mimic.
2. Results and Discussion
Our previous system relied on the use of the highly active
methoxide anion as a nucleophile for deacylation. This is
much different from serine proteases, which rely on a his-
tidine residue to activate water for nucleophilic attack on
the acylated serine residue in the covalent enzyme inter-
mediate (Figure 1a).[21] Similar observations have also
been made in synthetic systems, in which neighboring
amines activate alcohols for nucleophilic attack on car-
boxylic esters.[22–25] We argued that, in a similar manner,
the insertion of a flanking tertiary amine near the carbox-
ylic ester group would enable deacylation in methanol,
thus eliminating the necessity of using the methoxide
anion (Figure 1b). In order to test this hypothesis we pre-
pared phenyl acetate derivative 2a, containing a dimethy-
laminomethyl substituent in the ortho position to the
ester. On the other ortho position an aldehyde group was
inserted for reversible covalent bond formation with hy-
drazides. The stability of the ester moiety in reference hy-
drazone 2a-A was studied by following the changes in the
absorbance of a solution of 2a-A in CH3OH at 258. Trans-
esterification occurred highly efficiently, with a pseudo-
Figure 2. Scaffold molecules and hydrazide library.
Figure 3. Changes in the absorbance at 340 nm as a function of time for the deacylation of 2a-A (&) and 2a-B (&) in CH3OH (0.6 mM) at
258. Under the same conditions, deacylation of 2b-A (*, 0.6 mM) in the presence of one equivalent of Et3N could not be detected in the
same interval. The addition of one equivalent of trifluoroacetic acid completely inhibited deacylation of 2a-A (*). Rate constants for the
latter two deacylations were calculated from the initial rates of prolonged measurements (see Supporting Information).
Isr. J. Chem. 2013, 53, 122 – 126
ꢁ 2013 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.ijc.wiley-vch.de
123

Full Paper
L. J. Prins et al.
first order rate constant of 1.55ꢁ10À3 sÀ1 (&, Figure 3).
Deacylation was quantitative after one hour, as evidenced
by the fact that ESI-MS showed exclusively the presence
of the corresponding phenol. The importance of the pres-
ence of the tertiary amine was evident from the fact that
the analogous compound 2b-A, which lacks the tertiary
amine, showed deacylation with a pseudo-first order rate
constant of 1.44ꢁ10À6 sÀ1 even in the presence of one
equivalent of triethylamine (*, Figure 3). Furthermore,
the addition of one equivalent of trifluoroacetic acid to
2a-A completely inhibited transesterification (kobs =7.07ꢁ
10À7 sÀ1; *, Figure 3). The absence of large chemical
1
shifts for the methyl groups in the H NMR spectrum of
2a-A and the absence of charge transfer bands in the UV/
Vis spectrum indicated that intramolecular acyl transfer
to the amine did not appear to take place, suggesting that
the amine is indeed involved in nucleophile activation.
Thus, the mere presence of the o-dimethylaminomethyl
substituent in 2a-A increases the efficiency of the acyl
transfer to methanol by more than three orders of magni-
tude.
Figure 4. Amplification factors for hydrazides B-H obtained from
competition experiments with reference hydrazide A using scaffold
1a (dark gray). Samples were kept at 508 until the thermodynamic
equilibrium was reached (typically 12 hours), which was confirmed
by the absence of changes in the peaks of the UPLC chromato-
gram. The previously reported amplification factors obtained using
scaffold 1b (light gray) have been added for comparison. All meas-
urements were performed in CH3OH at 508 using five equivalents
of each hydrazide (25 mM) compared to the scaffold (5 mM).
Motivated by the positive effect of the tertiary amine
on the deacylation rate of 2a-A, we then focused on the
effect of the dimethylaminomethyl substituent on the
self-selection experiments. Recent studies on this system
had shown that substituents on the aromatic ring of the
phosphonate scaffold play a large role in determining the
strength of the intramolecular interaction between the
phosphonate group and the functional group present in
the hydrazide.[26] This behavior is very similar to the way
in which even remote groups in enzymes influence
enzyme activity by changing the orientation of functional
groups in the active site. Therefore, we repeated the
screening of the previously studied hydrazides B-H
against reference hydrazide A using the new phospho-
nate-containing scaffold 1a and a reference scaffold 3
with a neutral methoxy group. The concentration of hy-
drazones at thermodynamic equilibrium was determined
by UPLC. To each hydrazide B-H, an amplification factor
was assigned, corresponding to the ratio of the equilibri-
um constants using scaffolds 1a and 3. The obtained am-
plification factors are given in Figure 4, together with the
values that we had obtained earlier under the same condi-
tions for scaffold 1b lacking the dimethylaminomethyl
substituent.[15] The data reveal two important features.
First, the obtained amplification profile is identical for
both phosphonate scaffolds 1a and 1b and shows a signifi-
cant amplification for hydrazides B and C, which contain
positively charged ammonium or pyridinium groups, re-
spectively. Second, much higher amplification factors are
observed for scaffold 1a compared to 1b. In fact, for the
most strongly amplified hydrazide B, the amplification
factor increases from 1.8 to 4.2, the latter corresponding
to a DDG0 between 1a-A and 1a-B of 3.8 kJmolÀ1. We as-
cribe this to the substituent-induced orientation of the
phosphonate group in the direction of the positive charge.
Interestingly, these data point to a double role for the di-
methylaminomethyl substituent both as a nucleophile ac-
tivator and as a steering group. The observed stabilizing
interaction between phosphonate and ammonium groups
in hydrazone 1a-B should lead to an enhanced rate of de-
acylation of phenyl acetate derivative 2a-B compared to
the reference compound 2a-A, because of transition-state
stabilization by the ammonium group. In particular, as il-
lustrated in Figure 5, in the case of a quantitative corre-
spondence between TSA recognition in the self-selection
experiments and transition-state stabilization in catalysis,
the expected rate acceleration should correspond directly
to the amplification factor (4.2). For that reason, we were
very excited to find out that a threefold rate enhancement
was measured for the deacylation of 2a-B compared to
2a-A (kobs,2a-B =4.64ꢁ10À3 sÀ1) in CH3OH at 258 (&,
Figure 3). Compared to the reference compound 2b-A,
which lacks the dimethylaminomethyl substituent, this
implies a total rate acceleration of more than 3200 times.
3. Conclusions
In conclusion, we have developed a system that mimics
some key features of the regeneration step of an enzyme
involved in covalent enzyme catalysis. This system con-
tains functional groups involved in transition-state stabili-
zation and nucleophile activation. Furthermore, the terti-
ary amine substituent has a steering function by directing
the reaction center towards the functional group that sta-
bilizes the transition state. The result is an enhanced cata-
lytic efficiency. From a more general point of view, this
work illustrates the integration of a self-selection proce-
124
www.ijc.wiley-vch.de
ꢁ 2013 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Isr. J. Chem. 2013, 53, 122 – 126

Development of an Enzyme Mimic Using Self-Selection
Figure 5. Correlation between the amplification factors in the self-selection experiments and rate acceleration in catalysis. S=solvent.
[3] C. Gennari, U. Piarulli, Chem. Rev. 2003, 103, 3071.
[4] J. G. de Vries, A. H. M. de Vries, Eur. J. Org. Chem. 2003,
799.
[5] M. T. Reetz, Angew. Chem., Int. Ed. 2008, 47, 2556.
[6] M. J. Wilkinson, P. van Leeuwen, J. N. H. Reek, Org.
Biomol. Chem. 2005, 3, 2371.
[7] B. Breit, Angew. Chem., Int. Ed. 2005, 44, 6816.
[8] A. J. Sandee, J. N. H. Reek, Dalton Trans. 2006, 3385.
[9] Supramolecular Catalysis (Ed.: P. W. N. M. van Leeuwen),
1st edition, Wiley-VCH, Weinheim, 2008.
[10] J. Meeuwissen, J. N. H. Reek, Nat. Chem. 2010, 2, 615.
[11] M. J. Wiester, P. A. Ulmann, C. A. Mirkin, Angew. Chem.,
Int. Ed. 2011, 50, 114.
dure in catalyst development. In particular, the results
show that the application of a dynamic covalent capture
strategy gives an unprecedented access to the experimen-
tal quantification of very small energy variations at the
transition state. Importantly, a neat correspondence is ob-
served between amplification factor and rate enhance-
ment. In our opinion, the advantage of the dynamic cova-
lent capture strategy is twofold. First, the covalent attach-
ment of the TSA to the scaffold molecule locates it pre-
cisely at the position where the actual transition state is
formed. Consequently, during the self-selection and catal-
ysis the functional groups are positioned at identical posi-
tions, which permits a quantitative correspondence be-
tween amplification factor and rate enhancement.
Second, the fact that amplification is a result of intramo-
lecular stabilization gives an unprecedented sensitivity to
the selection process. This is perfectly in line with related
approaches aimed at quantifying weak energetic contribu-
tions in (bio)molecular recognition events.[19,27,28]
[12] G. Gasparini, M. Dal Molin, L. J. Prins, Eur. J. Org. Chem.
2010, 2429.
[13] B. Brisig, J. K. M. Sanders, S. Otto, Angew. Chem., Int. Ed.
2003, 42, 1270.
[14] L. Vial, J. K. M. Sanders, S. Otto, New J. Chem. 2005, 29,
1001.
[15] G. Gasparini, L. J. Prins, P. Scrimin, Angew. Chem., Int. Ed.
2008, 47, 2475.
[16] M. Matsumoto, D. Estes, K. M. Nicholas, Eur. J. Inorg.
Chem. 2010, 1847.
[17] J. Wassenaar, E. Jansen, W. J. van Zeist, F. M. Bickelhaupt,
M. A. Siegler, A. L. Spek, J. N. H. Reek, Nat. Chem. 2010,
2, 417.
Acknowledgements
[18] P. T. Corbett, J. Leclaire, L. Vial, K. R. West, J. L. Wietor,
J. K. M. Sanders, S. Otto, Chem. Rev. 2006, 106, 3652.
[19] L. J. Prins, P. Scrimin, Angew. Chem., Int. Ed. 2009, 48,
2288.
[20] M. Martin, G. Gasparini, M. Graziani, L. J. Prins, P. Scrimin,
Eur. J. Org. Chem. 2010, 3858.
Financial support from the European Research Council
under the European Communityꢂs Seventh Framework
Programme (FP7/2007-2013) and ERC Starting Grant
Agreement (No. 239898) and COST (CM0703 and
CM0905) is acknowledged.
[21] A. Fersht, Structure and Mechanism in Protein Science, 1st
edition, W. H. Freeman and Company, 1998.
[22] J. Hine, M. N. Khan, J. Am. Chem. Soc. 1977, 99, 3847.
[23] K. A. Wayman, T. Sammakia, Org. Lett. 2003, 5, 4105.
[24] G. T. Notte, T. Sammakia, P. J. Steel, J. Am. Chem. Soc.
2005, 127, 13502.
References
[1] Comprehensive Asymmetric Catalysis, Vol. I–III (Eds.: E. N.
Jacobsen, A. Pfaltz, H. Yamamoto), Springer, New York,
1999.
[25] L. A. Wessjohann, M. Z. Zhu, Adv. Synth. Catal. 2008, 350,
107.
[2] M. T. Reetz, Angew. Chem., Int. Ed. 2001, 40, 284.
Isr. J. Chem. 2013, 53, 122 – 126
ꢁ 2013 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.ijc.wiley-vch.de
125

Full Paper
L. J. Prins et al.
[26] M. Dal Molin, G. Gasparini, P. Scrimin, F. Rastrelli, L. J.
Prins, Chem. Commun. 2011, 47, 12476.
[28] M. G. Woll, S. H. Gellman, J. Am. Chem. Soc. 2004, 126,
11172.
[27] D. A. Erlanson, A. C. Braisted, D. R. Raphael, M. Randal,
R. M. Stroud, E. M. Gordon, J. A. Wells, Proc. Natl. Acad.
Sci. U.S.A. 2000, 97, 9367.
Received: October 30, 2012
Accepted: December 5, 2012
Published online: February 19, 2013
126
www.ijc.wiley-vch.de
ꢁ 2013 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Isr. J. Chem. 2013, 53, 122 – 126