Towards enzyme-like enantioselectivity in the gas phase: Conformational control of selectivity in chiral macrocyclic dimers

Article abstract of DOI:10.1039/b909017g

Conformational factors in self-assembled chiral tetraamide macrocycles control their gas-phase enantioselectivity towards the ethyl ester of naphthylalanine to levels typical of enzymes. The Royal Society of Chemistry 2009.

Full text of DOI:10.1039/b909017g

View Article Online / Journal Homepage / Table of Contents for this issue
COMMUNICATION
www.rsc.org/chemcomm | ChemComm
Towards enzyme-like enantioselectivity in the gas phase: conformational
control of selectivity in chiral macrocyclic dimersw
Caterina Fraschetti, Marco Pierini, Claudio Villani,* Francesco Gasparrini,
Stefano Levi Mortera and Maurizio Speranza*
Received (in Cambridge, UK) 6th May 2009, Accepted 13th July 2009
First published as an Advance Article on the web 31st July 2009
DOI: 10.1039/b909017g
Conformational factors in self-assembled chiral tetraamide
macrocycles control their gas-phase enantioselectivity towards
the ethyl ester of naphthylalanine to levels typical of enzymes.
Enzymes are proteic macromolecules that enable and support
life functions. A key feature of enzymes is their capability for
‘‘molecular recognition’’. Size- and shape-specific non-covalent
interactions, assisted by extensive desolvation phenomena,1
allow the selective absorption of the target molecule into the
enzyme structure and its modification at the ‘‘active site’’.2–5
Chart 1 Formulae of the chiral tetraamidic receptors MR/S and of the
ethyl ester of (S)-naphthylalanine AS.
xhomo = kRhomo/kS
when referred to the configuration of the amine B (eqn (1a,b)).
and xhetero = kRhetero/kS
ratios,
hetero
ð1aÞ
homo
Reactions (1a,b) (n 1), involving the two-body
=
[MR/SꢀHꢀAS]+ complexes, exhibit enantioselectivities (r =
ca. 0.05) among the largest ever measured in the gas phase,
that were attributed to differences in the structure and the
relative stability of the [MR/SꢀHꢀAS]+ diastereoisomers.8 Indeed,
molecular mechanics (MM) calculations indicate that macro-
cycles MR/S may assume diverse stable conformations, as
illustrated in Chart 2. In both eq–eq and ax–ax geometries,
macrocycles MR/S have C2-symmetric folded structures with a
concave face F1 and a convex one F2. The eq–eq conformers,
by far the most stable ones in the isolated MR/S molecules,
acquire in the gas phase a different conformation by induced
fit on complexation with some amino acid derivatives, like AS.
This leads to the co-existence in the gas phase of stable
eq–eq and ax–ax structures, in proportions depending on the
configuration of MR/S and characterized by different stability
and reactivity towards the 2-aminobutane enantiomers. In this
case, reactions (1a,b) (n = 1) obey bi-exponential kinetics.
Despite its unusually large r value, the gas-phase enantio-
selectivity of [MR/SꢀHꢀAS]+ towards BR/S is still well below
ð1bÞ
The amazing selectivity and catalytic proficiency of enzymes
have provided chemists with a stimulus to design ‘‘synthetic
enzymes’’ for understanding the mechanism of action of
enzymes and for attempting to reproduce them for practical
applications. However, these synthetic receptors exhibit
selectivities which are orders of magnitude lower than those
of enzymes, even when acting in the gas phase, i.e. under
conditions reproducing the extensive desolvation undergone
by the target molecule when entering the enzyme cavity.6
The chiral macrocycles MR/S of Chart 1 are synthetic
receptors that gained some attention in recent gas-phase
studies.7,8 Their gas-phase enantioselectivity towards the
several amino acid derivatives, including the ethyl ester of
(S)-naphthylalanine (AS), was evaluated by generating the
proton-bonded [MR/SꢀHꢀAS]+ adducts in the cell of
a
Fourier-transform ion cyclotron resonance mass spectrometer
(FT-ICR-MS), equipped with an electrospray ionization
source (ESI), and by measuring their reaction kinetics
toward either (R)-(ꢁ)-(BR) or (S)-(+)-2-butylamine (BS)
QJ;(see ESIw). The reaction enantioselectivity is defined
by the rR = kRhomo/kR
and rS = kShomo/kS
hetero
hetero
ratios, when referred to the configuration of M, or by the
`
Dipartimento di Chimica e Tecnologie del Farmaco, Universita di Roma
‘‘La Sapienza’’, Rome, Italy. E-mail: maurizio.speranza@ uniroma1.it;
Fax: +39-06-49913497; Tel: +39-06-49913602
w Electronic supplementary information (ESI) available: ESI-FT-ICR
experimental and MM computational details. Kinetic plots. See
DOI:10.1039/b909017g
Chart 2 Relevant minimum energy structures of the MR host found
by conformational search.
ꢂc
This journal is The Royal Society of Chemistry 2009
5430 | Chem. Commun., 2009, 5430–5432

View Article Online
Table 1 Rate constants for the B-induced loss of AS from the diastereomeric [(MR/S)2ꢀHꢀAS]+ complexes
BR = (R)-(ꢁ)-C4H9NH2
BS = (S)-(+)-C4H9NH2
eff =
100kR/kC
rR
=
kRhomo/kR
eff =
100kS/kC
rS
=
kShomo/kS
kR a
kS a
b
b
Complex
%
xc
hetero
hetero
[(MS)2ꢀHꢀAS]+
[(MR)2ꢀHꢀAS]+
[(MS)2ꢀHꢀAS]+fast
—
13 ꢃ 1 40.7 ꢃ 1.6 37
0.62 ꢃ 0.03 0.6
0.015 ꢃ 0.002
0.17 ꢃ 0.01
0.73 ꢃ 0.02 0.7
41
0.73 ꢃ 0.02 0.7
3.20 ꢃ 0.18 2.9
0.016 ꢃ 0.001
0.23 ꢃ 0.01
xhomo = 0.86 ꢃ 0.06
xhetero = 0.89 ꢃ 0.05
xhomo = 0.86 ꢃ 0.06
xhetero = 1.13 ꢃ 0.08
45.5 ꢃ 2.2
— 0.62 ꢃ 0.03 0.6
87 ꢃ 1 3.62 ꢃ 0.15 3.3
[(MR)2ꢀHꢀAS]+
slow
k ꢄ 1011 cm3 per molecule per second. Reaction efficiency expressed as the ratio between the measured rate constants k and the corresponding
a
b
collision constant kC, calculated using the trajectory calculation method.11 c See text.
that typical of enzymes. In this communication, kinetic results
are presented which demonstrate how conformational effects
in self-assembled hosts MR/S may control their gas-phase
enantioselectivity towards AS to levels comparable to, if not
higher than, those of many enzymes.9
most representative structures of the homochiral [(MS)2ꢀHꢀAS]+
and the heterochiral [(MR)2ꢀHꢀAS]+ aggregates can be
conveniently clustered (within a 3 kcal molꢁ1 window) in two
geometric typologies (I and II in Fig. 1). Both Ihetero and Ihomo
structures are characterized by several H-bonds between the
–NH3+ group of the guest and the converging pairs of carbonyls
on the convex sides F2 of two facing host molecules (Fig. S5,
ESIw). In contrast, the IIhomo and IIhetero aggregates are
characterized by a minor number of H-bonds among the
–NH3+ group of the guest and the converging pair of carbonyls
placed on the convex sides F2 of only one host molecule (in blue
in Fig. 1), with ax–ax geometry in IIhetero and eq–eq one in
IIhomo. However, while in IIhomo, the second host molecule (in
grey in Fig. 1) cannot directly interact with the guest, in IIhetero
the second host molecule is in direct contact with ester A through
a H-bond and an efficient C–Hꢀꢀꢀp interaction (Fig. S5, ESIw).
Interestingly, the dimer (M)2 assumes a Y-shaped disposition
with a C2 symmetry in both Ihomo and Ihetero complexes. In
IIhomo, it shows a D2 symmetry,9 whereas in IIhetero, it does not
show any symmetry. The topological features of the most stable
Ihomo and Ihetero structures are reported in Fig. 2.
Table 1 reports the second-order rate constants of reactions
(1a,b) (n = 2) between the three-body [(MR/S)2ꢀHꢀAS]+
complexes and BR/S, as well as the relevant enantioselectivity
factors rR, rS, xhomo, and xhetero. Likewise the two-body
[MR/SꢀHꢀAS]+ complexes, their three-body [(MR)2ꢀHꢀAS]+
homologue exhibits bi-exponential kinetics pointing to the
occurrence of two stable structures, [(MR)2ꢀHꢀAS]+
and
fast
[(MR)2ꢀHꢀAS]+
,
characterized by different reactivity
slow
towards BR/S. In contrast, the homochiral [(MS)2ꢀHꢀAS]+
complex displays mono-exponential kinetics which suggests
the occurrence of a single structure or, alternatively, of several
ones but with comparable reactivity towards BR/S. Relative to
the homochiral [(MS)2ꢀHꢀAS]+complex, the heterochiral
[(MR)2ꢀHꢀAS]+slowstructure reacts from 4.4 to 5.8 times faster
(rR = 0.17 ꢃ 0.01; rS = 0.23 ꢃ 0.01), whereas the
[(MR)2ꢀHꢀAS]+fastone reacts from 62 to 66 times faster
(rR = 0.015 ꢃ 0.002; rS = 0.016 ꢃ 0.001). The latter r values
clearly highlight a gas-phase enantioselectivity (corresponding
to an ee of ca. 97% in a single reactive event) comparable to
that of many enzymes. A comparison of the r values of the
two-body [MR/SꢀHꢀAS]+ complexes and their three-body
[(MR/S)2ꢀHꢀAS]+ counterparts is given in Table 2. Its
inspection reveals that self-assembling of hosts MR/S does
produce a significant effect on the reaction enantioselectivity
which decreases with the slow components of the reaction and
increases with the fast ones.
The energy of the simulated [(MR/S)2ꢀHꢀAS]+ complexes can
be conveniently analyzed to attempt a rationalization of
the relevant kinetic results of Table 2. On the grounds of the
computed relative energy of the Ihetero and IIhetero isomers, the
first structure is attributed to the less reactive [(MR)2ꢀHꢀAS]+
slow
isomer and the latter to the more reactive [(MR)2ꢀHꢀAS]+fast one.
However, the experimental distribution of these two isomers
(fast: 13 ꢃ 1%; slow: 87 ꢃ 1%, corresponding to a conforma-
tional stability difference of about 1.1 kcal molꢁ1 at 300 K;
Table 1) is only in semiquantitative agreement with the calculated
energy gap of 3.0 kcal molꢁ1 (Fig. 1). Furthermore, the observed
mono-exponential kinetics of the reaction with the homochiral
[(MS)2ꢀHꢀAS]+ complex is inconsistent with the 1.2 kcal molꢁ1
To understand the factors determining such a behavior, an
in-depth theoretical study has been undertaken by performing
extensive ‘‘quasi-flexible’’ multiconformational molecular dockings
based on molecular mechanics calculations (see ESI,w p. S7).10 The
difference between the calculated energies of Ihomo and IIhomo
,
Table 2 Comparison of the gas-phase enantioselectivities of the reaction of [MR/SꢀHꢀAS]+ a and [(MR/S)2ꢀHꢀAS]+ towards B
b
eff = 100k/kC
rR = kRhomo/kR
rS = kShomo/kS
DDG*expc/kcal molꢁ1
DDH1thd/kcal molꢁ1
homo
hetero
Complex
hetero
hetero
[MR/SꢀHꢀAS]+
[(MR/S)2ꢀHꢀAS]+
fast
slow
fast
1
21
1
39
3.1
0.046 ꢃ 0.004
0.052 ꢃ 0.007
0.015 ꢃ 0.002
0.17 ꢃ 0.01
0.050 ꢃ 0.009
0.048 ꢃ 0.009
0.016 ꢃ 0.001
0.23 ꢃ 0.01
1.8 ꢃ 0.1
1.8 ꢃ 0.1
2.5 ꢃ 0.1
1.0 ꢃ 0.1
1.0
1.3
0.05
0.6
0.6
2.7–3.9e
0.9
slow
a
b
c
d
Ref. 8. See footnote b in Table 1, k = (kR + kS)/2. DDG*exp = DG*homo ꢁ DG*hetero = ꢁRTln (rR + rS)/2, T = 300 K. DDH1th
=
e
DH1homo ꢁ DH1hetero, calculated at T = 300 K. Energy differences calculated between the stability of the single structures Ihomo and IIhomo and
that of IIhetero
.
ꢂc
This journal is The Royal Society of Chemistry 2009
Chem. Commun., 2009, 5430–5432 | 5431

View Article Online
As shown for the two-body [MR/SꢀHꢀAS]+ complexes,8 the
comparable enantioselectivity exhibited by their fast and slow
isomers (rfast and rslow ca. 0.050; Table 2) is attributed to the fact
that the host assumes the same conformation, i.e. ax–ax in the
most stable slow isomer and eq–eq in the less stable fast one.
This means that the observed enantioselectivities are essentially
determined by conformational factors, i.e. by changes in the
intensity of non-covalent host–guest interactions as the host (or
guest) configuration is inverted. A similar conclusion can be
reached for the most stable three-body Ihomo and Ihetero isomers
(Fig. 1), wherein the two host molecules acquire the same eq–eq
geometry. However, the differently distorted Y-shaped disposition
of the eq–eq host molecules in Ihomo and Ihetero (Fig. 2) decreases
the reaction efficiency in general, but much less for the first
(eff = 0.6 (n = 2); 1 (n = 1)) than for the latter (eff = 3.1
(n = 2); 21 (n = 1); Table 2). This obviously reduces the
enantioselectivity of the reactions (1a,b) with Ihomo and Ihetero
(r’s of ca. 0.2) relative to that displayed by [(MR/SꢀHꢀAS]+ (r’s of
ca. 0.05). In IIhetero (Fig. 1), instead, no tridimensional regularity
is observed since one host molecule in the eq–eq conformation is
facing the other in the ax–ax one. With this fully asymmetric
disposition, the stabilizing p–p interactions in slow ax–ax
[MRꢀHꢀAS]+ isomer, between the naphthyl ring of AS and the
isophthalic rings of the host,8 are partially destroyed in IIhetero by
the presence of the second eq–eq host. The consequence is that
the reaction efficiency increases 39 times (Table 2). This obviously
enhances the enantioselectivity of the fast component of reactions
(1a,b) relative to that displayed by [(MR/SꢀHꢀAS]+. It is
concluded that the proton-bound [(MR/S)2ꢀH]+ dimer in the
diastereomeric [(MR/S)2ꢀHꢀAS]+ complexes may act as gaseous
enzyme mimics. Depending on the configuration of the amino
acidic guest, the [(MR/S)2ꢀH]+ moiety adapts itself to maximize
non-covalent interactions with the guest. Significant conforma-
tional changes take place in [(MR/S)2ꢀH]+ which confer to the
diastereomeric [(MR/S)2ꢀHꢀAS]+ complexes a relatively large
stability difference and, hence, an unprecedented thermodynamic
enantioselectivity in their gas-phase reactions (1a,b) (n = 2).
Fig. 1 Structures and relative energies (kcal molꢁ1) of the most stable
homochiral [(MR)2ꢀHꢀAR]+ (Ihomo and IIhomo
) and heterochiral
[(MR)2ꢀHꢀAS]+ (Ihetero and IIhetero) complexes. For the sake of clarity,
hydrogen atoms are omitted with the exception of those bonded to
nitrogen atoms.
Fig. 2 Topological representation of Ihomo (left) and Ihetero (right).
The represented C2 symmetry axes refer to the host dimer alone. The
two spiral bars represent the major axis m of the two host molecules
(Chart 1) in the corresponding structures. The (M)2 supramolecules
have a twisted structure, and the guest configuration controls their
absolute sense of twist. Thus, the guest induces a left (ꢁ371) and right
(+261) handed twist between the major axes m in the homo- and
heterochiral three-body structures, respectively.
Notes and references
which would allow their co-existence in the gas phase. In
this frame, the mono-exponential kinetics exhibited by
[(MS)2ꢀHꢀAS]+ can only be explained by admitting that Ihomo
and IIhomo are endowed with a comparable reactivity towards B.
Interestingly, this latter hypothesis finds a very persuasive
support from the kinetic behavior already evidenced in the
reactions (1a,b) (n = 1) with the two-body homochiral
[MSꢀHꢀAS]+ complex.8 Indeed, a strict structural relationship
1 F. Hollfelder, A. J. Kirby and D. S. Tawfik, Nature, 1996, 383, 60–63.
2 J. M. Lehn, in Supramolecular Chemistry. Concepts and
Perspectives, Wiley-VCH, Weinheim, 1995.
3 P. D. Beer, P. A. Gale and D. K. Smith, in Supramolecular
Chemistry (Oxford Chemistry Primers, 74), Oxford University
Press, Oxford, 1999.
4 H. J. Schneider and A. Yatsimirsky, Chem. Soc. Rev., 2008, 37,
263–277.
5 J. W. Steed and J. L. Atwood, in Supramolecular Chemistry, Wiley,
New York, 2000.
6 See, for instance: (a) M. Speranza, F. Gasparrini, B. Botta,
C. Villani, D. Subissati, C. Fraschetti and F. Subrizi, Chirality,
2009, 21, 69–86; (b) M. Speranza, Int. J. Mass Spectrom., 2004,
232, 277–317; (c) M. Speranza, Adv. Phys. Org. Chem., 2004, 39,
147–281; (d) C. Fraschetti, M. Aschi, A. Filippi, A. Giardini and
M. Speranza, Chem. Commun., 2008, 2544–2546.
7 A. Filippi, F. Gasparrini, M. Pierini, M. Speranza and C. Villani,
J. Am. Chem. Soc., 2005, 127, 11912–11913.
8 F. Gasparrini, M. Pierini, C. Villani, A. Filippi and M. Speranza,
J. Am. Chem. Soc., 2008, 130, 522–534.
exists between IIhomo and its two-body [MSꢀHꢀAS]+
fast
counterpart (Fig. S6, ESIw). Thus, the quite similar reaction
efficiency exhibited by Ihomo (average value: 0.6%, Table 2) and
[MSꢀHꢀAS]+
(average value: 1.0%, Table 2) leaves one to
fast
retain that the same virtual equivalence must exist between the
rate constants for the displacement processes involving Ihomo and
IIhomo. As reported in Table 2, an excellent correlation does exist
between the experimental DDG*exp (1.0 ꢃ 0.1 kcal molꢁ1) and the
calculated DDH1th difference (0.9 kcal molꢁ1) of the low-energy
9 P. F. Mugford, U. G. Wagner, Y. Jiang, K. Faber and
R. J. Kazlauskas, Angew. Chem., Int. Ed., 2008, 47, 8782–8793.
10 C. Fraschetti, M. Pierini, C. Villani, F. Gasparrini, A. Filippi and
M. Speranza, Collect. Czech. Chem. Commun., 2009, 74, 275–297.
11 T. Su and W. J. Chesnavitch, J. Chem. Phys., 1982, 76, 5183.
I
hetero and Ihomo structures. This correspondence suggests that the
measured enantioselectivities are determined more by the relative
stability of the diastereomeric [(MR/S)2ꢀHꢀAS]+ complexes than
by that of the corresponding eqn (1a,b) transition structures.
ꢂc
This journal is The Royal Society of Chemistry 2009
5432 | Chem. Commun., 2009, 5430–5432