Dynamic combinatorial evolution within self-replicating supramolecular assemblies

Article abstract of DOI:10.1002/anie.200804602

(Figure Presented) Survival of the fittest: Self-assemblies made of dynamic block copolymers (dynablocks) can self-replicate by catalyzing the formation of their own building blocks. Moreover, in competition experiments, the differential thermodynamic sta

Full text of DOI:10.1002/anie.200804602

Angewandte
Chemie
DOI: 10.1002/anie.200804602
Combinatorial Chemistry
Dynamic Combinatorial Evolution within Self-Replicating
Supramolecular Assemblies**
Rꢀmi Nguyen, Lionel Allouche, Eric Buhler, and Nicolas Giuseppone*
Dynamic combinatorial chemistry (DCC) rests on the design
and the study of libraries of species connected by reversible
constituents competing in a series of coupled thermodynamic
[10]
equilibria. Although this reported DCL has both kinetic
and thermodynamic biases that amplify the best duplicator
and decrease its competitors, it does not present a strong
autocatalytic behavior. Herein we describe another DCL
which avoids the drawback of product inhibition by taking
advantage of the growth/division cycles of micellar self-
assemblies, and which displays a particular case of autocatal-
ysis, namely autopoiesis. This concept appeared in the mid-
1970s, when Maturana, Varela, and Uribe proposed that living
systems are essentially characterized by their aptitude to
continuously organize the generation of their own compo-
nents, thus maintaining the very network process that
[
1,2]
(
supra)molecular bonds.
It represents a very attractive
domain of modern chemistry because it associates combina-
torial features together with the spontaneous self-organiza-
[
3–5]
tion of molecules.
Dynamic combinatorial libraries
(
DCLs) are governed by thermodynamics and are conse-
quently subjected to the influence of internal or external
parameters that can reversibly modify the expression of their
constituents through selection/adaptation. In bioinspired
chemistry, other efforts to understand molecular evolution
are focused on minimal autocatalytic and self-replicating
[
6,7]
systems that are governed by kinetics. Herein we show that
by coupling DCC with the autocatalytic formation of
specifically designed supramolecular assemblies, a self-repli-
cating selection can occur at two length scales with a sigmoid
concentration–time profile. Indeed, we have found that by
using a new kind of molecular objects, namely dynamic
amphiphilic block copolymers (dynablocks), in which a
hydrophobic block is reversibly linked to a hydrophilic one,
[
11]
produces them. The minimal criteria defining autopoiesis
should verify whether 1) the system has a semipermeable
boundary that is 2) produced within the system, and 3) that
encompasses reactions which regenerate the components of
[
8]
the system. The seminal work of Luisi et al. brought to light
the first examples of minimal chemical autopoietic systems
that produce surfactants inside micelles or vesicles built by
[
8,9]
[12,13]
the formation of micelles can have autopoietic
growth in
these very constituents.
Whereas the definition of life is
water. Moreover, when different hydrophilic blocks compete
for the same hydrophobic block in coupled equilibria, the
differential thermodynamic stabilities and autocatalytic effi-
ciencies of the resulting mesoscopic structures lead to the
selection of the most efficient self-replicator and to the
depletion of its competitors.
We have recently shown that DCC could be associated
with self-replicating systems to increase the concentration of a
single product, by duplication from a pool of reshuffling
controversial, and is more popularly defined by self-replica-
[14]
tion according to the prebiotic RNA world view, autopois-
[
9,15]
esis remains at least a complementary approach
and
defines a very interesting conceptual framework that encom-
passes collective properties, such as self-assembly, self-organ-
ization, and emergence. Thus, the possible association of
autopoiesis with selection processes—for instance, those
occurring through the network of coupled equilibria in a
DCL—constitutes a very attractive pathway in the field of
molecular evolution.
To set up our study, we first designed a new type of
amphiphilic molecular objects that, because of their rever-
sible connections and through molecular recombination,
allow the production of various types (in size and shape) of
micellar self-assemblies in water. These objects were con-
structed by using the reversible connection of a single imine
bond between hydrophilic and hydrophobic blocks, thus
[
*] R. Nguyen, Prof. Dr. N. Giuseppone
SAMS Laboratory—icFRC—Universitꢀ de Strasbourg—Institut
Charles Sadron, CNRS—UPR 22
23 rue du Loess, BP 84047, 67034 Strasbourg cedex 2 (France)
Fax: (+33)3-8841-4099
E-mail: giuseppone@ics.u-strasbg.fr
Dr. L. Allouche
[
16,17]
Institut de Chimie, Service de RMN, Universitꢀ Louis Pasteur
leading to dynamic amphiphilic blocks (dynablocks).
The
1
rue Blaise Pascal, 67008 Strasbourg cedex (France)
individual condensations of aliphatic, benzylic, aromatic, and
hydroxy amines 1–8 (having PEO units of different lengths)
with the p-substituted benzaldehyde A having a hydrophobic
tail of 8 carbons lead to the formation of dynablocks 1A–8A.
These compounds have different Hydrophilic/hydrophobic
Prof. Dr. E. Buhler
Matiꢁre et Systꢁmes Complexes Laboratory, Universitꢀ Paris-VII,
UMR 7057
10 rue Alice Domon et Lꢀonie Duquet, 75205 Paris cedex 13
(
France)
[18]
ratios (rH/h) that are related to surfactant shape parameters
Figure 1 and Supporting Information, Table S1 and Fig-
[
**] We thank the CNRS, the icFRC (RTRA), and the University Louis
Pasteur for financial support. This work was supported by a doctoral
fellowship of the Rꢀgion Alsace (R.N.). We also acknowledge ESF-
COST System Chemistry action.
(
ure S1). The equilibrium constants were determined by
1
H NMR spectroscopy in deuterated acetonitrile at 298 K
(
[aldehyde]init = [amine]init = 50 mm). These experiments show
Supporting information for this article is available on the WWW
under http://dx.doi.org/10.1002/anie.200804602.
that, as expected, the condensation of imines depends on the
Angew. Chem. Int. Ed. 2009, 48, 1093 –1096
ꢀ 2009 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
1093

Communications
[
21]
diffusion coefficients.
For example, when
studying dynablock 7A in water (Supporting
Information, Figure 3), the correlation between
the signals in the first dimension with the
diffusion values in the second dimension lead
to the following conclusions. The imine dyna-
2
À1
block diffuses with a rate of 28 mm s , corre-
sponding to a micellar object of 14.2 nm in
diameter that contains in its core all the remain-
ing free aldehyde whereas the free amine stays
outside of the micellar system with a diffusion of
Figure 1. Structures of dynablocks. The individual reactions of hydrophilic amines 1–8
with hydrophobic aldehyde A lead to imines 1A–8A. In D O, these dynablocks self-
2
2
À1
assemble into supramolecular micellar structures.
250 mm s . This location of the free aldehyde
within the boundary of the structure is the first
[
8]
requirement to give rise to autopoietic behavior.
nucleophilicity of the amine reacting groups (hydroxy
amine @ aliphatic ꢀ benzylic @ aromatic; Supporting Infor-
mation, Table S1). The dynablocks were then diluted in
deuterated water and the acetonitrile was evaporated, keep-
ing a final concentration of 50 mm (the residual quantity of
We then turned to the kinetic and thermodynamic studies
of the formation of compound 7A by mixing 7 and A directly
in deuterated water (equimolar ratio of 50 mm each). The
1
condensation of the product, measured by H NMR spectros-
copy, reveals a sigmoid concentration–time profile, which is
characteristic of an autocatalytic system (Figure 2a). To
determine the origin of the autocatalytic process, we set up
1
[19]
acetonitrile was not detectable by H NMR spectroscopy).
The clear solutions (except for 2A and 3A) were studied by
1
H NMR and DOSY NMR spectroscopy and by light and
1
neutron scattering. In most cases, H NMR spectroscopy
indicates that the signals for the imine groups are present in
water (d = 7.5–8.5 ppm) together with the free aldehyde (d =
9
.8 ppm). The equilibrium constants were measured, and
although the presence of water as a solvent should favor the
hydrolysis of the imine bonds, a high degree of condensation
was observed for compounds 4A–8A (Supporting Informa-
tion, Table S1). This increase of the equilibrium constant can
be attributed to the self-assembly of the dynablocks in highly
stable supramolecular structures, except for 2A and 3A,
which quickly hydrolyze and release the insoluble free
aldehyde. Only product 1A has an equilibrium constant that
is lower in water than in acetonitrile, which reveals a
relatively weak thermodynamic stability of its amphiphilic
self-assembled superstructure, probably because of the partial
lack of p–p stacking interactions. The micellar structures were
confirmed for 1A, 4A, 5A, 7A, and 8A by using DOSY
[20]
NMR spectroscopy with which the diffusion of the imine
self-assemblies in water can be correlated to their hydro-
dynamic radii (Supporting Information, Table S1). The results
indicate the formation of micellar self-assemblies having
hydrodynamic radii comprised between 5 and 7.1 nm, and
which vary inversely with rH/h. For instance, dynablock 4A
(
(
rH/h=2.1) has a hydrodynamic radius of 6.9 nm, whereas 5A
rH/h=2.9) shows a smaller hydrodynamic radius (6.5 nm). The
evolution from cylindrical to spherical micelles depending on
[
18]
rH/h
,
was shown by light- and neutron-scattering experi-
Figure 2. Autopoietic behavior of dynablock 7A. a) Concentration of
ments for compounds 6A and 8A (Supporting Information,
Figure S2a–b). For compound 6A (rH/h = 1.3), the size of the
structure is too large to be observed by DOSY experiments,
but neutron scattering reveals the presence of cylindrical
micellar imine 7A versus time starting from 7 and A in D O (50 mm
2
each), and as a function of the quantity of initially added micellar
À4
À4
À4
imine 7A:
&
0, * 5.1ꢂ10 , & 51ꢂ10 , * 102ꢂ10 mmol. The
micelle concentrations were determined by the integration of the
1
À1
micellar imine H NMR spectroscopy signal (8.12 ppm), which differs
micelles having a mass of 507000 gmol , a number of
from the free imine signal (8.23 ppm). Maximum standard deviation of
the data presented 5%, determined by setting up the reaction three
times under the same conditions. Dotted lines connecting the
experimental results drawn to guide the eye. b) Hydrodynamic radii Hr
of the micellar structure formed from 7A as a function of the course of
the condensation reaction from 7 and A.
aggregation of 783, and a size of 6 nm diameter and 34 nm
in length. For compound 8A (rH/h = 4), neutron scattering
data support the presence of spherical micelles. In this study,
another asset of the DOSY NMR spectroscopy technique is to
discriminate, within mixtures, the components with different
1
094 www.angewandte.org
ꢀ 2009 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Angew. Chem. Int. Ed. 2009, 48, 1093 –1096

Angewandte
Chemie
the same experiment but in the presence of increasing initial
amounts of preformed micelles of 7A. The progressive loss of
the sigmoid shape clearly indicates that the micelle catalyses
its own formation through the condensation of 7 and A with a
À1
À1
vmax of 72 10 mmolh (autocatalytic efficiency e ꢁ 80;
Supporting Information, Table S2). This maximum rate is
reached for a concentration of micellar imine of 5 mm, and
this saturation effect suggests the assistance of the micelle to
solubilize the hydrophobic aldehyde until attaining the
maximum rate of the templated imine condensation itself
(
Supporting Information, Figure S4a). For low quantities of
À4
À4
catalyst (5.1 ꢀ 10 mmol < x < 51 ꢀ 10 mmol), the plot of
log(V ) against the logarithm of the initial micellar concen-
0
tration shows a linear dependence of rate on catalyst
concentration, demonstrating that the rate of the uncatalyzed
reaction is comparatively negligible (Supporting Information,
Figure S4b). We also determined the size of the micellar
structures as a function of the advancement of the condensa-
tion reaction between 7 and A. After nucleation, we expected
to observe a constant average size of the micellar self-
assembly as the structure grows and divides constantly
because of the sheer forces leading to a thermodynamic
[22]
instability above a critical size.
However, DOSY NMR
studies show a more complex behavior, with the decrease of
the average hydrodynamic radii of the micelles from 16 nm
with a condensation of 36% to 7 nm with a condensation of
Figure 3. Molecular selection in coupled equilibria through the self-
replication of a specific mesostructure. a) Concentration of imines 1 A
and 7A versus time starting from an equimolar mixture of 1, 7, and A
7
0% (Figure 2b). The reason is that the size and shape of the
micelles are determined by both rH/h and the intramicellar free
aldehyde/imine ratio (ia/i). At the beginning of the reaction
the ia/i ratio is high because the micelles solubilize a high
quantity of aldehyde, thus producing larger objects with a
(
c=50 mm each) in CD CN and, after reaching the thermodynamic
3
equilibrium, by changing the solvent to pure D O. b) Concentration of
2
imines 1 A and 7 A versus time starting from an equimolar mixture of
[
18]
1
, 7, and A in D O (c=50 mm each).
(
probably) cylindrical structure. This effect on the micellar
2
dimensions was confirmed by the addition of free aldehyde A
(
(
up to 300%) to a preformed solution of micellar 7A
Supporting Information, Figure S5a). However, as the
50 mm. The concentration–time profile shows a dramatic
evolution of the selectivity in favor of 7A, owing to the
critical size of the spherical structure is reached, an expected
growth/division cycle with a constant average size of the
population takes place, as was shown by increasing the
concentration of both 7 and A from 2 mm to 75.5 mm
[23]
formation of the most stable micellar self-assembly. The
À1
formation of 7A (V = 15 mm h , and c = 32 mm) is ach-
0
eq
ieved by the destruction of its competitor 1A (V =
0
À1
(
Supporting Information, Figure S5b). Finally, the impor-
À15 mm h , and c = 4.5 mm). The second competition was
eq
tance of the micellar structure to the thermodynamic stability
of dynablock 7A was demonstrated by mixing acetonitrile
and water. For the lower molar fractions in water, the
dynablocks lose their supramolecular stabilization and start
hydrolyzing (Supporting Information, Figure S6a–b). The
combined results of the kinetic and thermodynamic studies
clearly demonstrate that the self-replication process of the
dynablocks described herein amounts to a minimal auto-
poietic system without any other reagent than the two
building blocks and water.
performed directly in deuterated water from 1, 7, and A
(50 mm each). The sigmoid concentration–time profile indi-
cates a highly selective self-replicating process in favor of 7A,
with 1A being formed in quantities that are always less than
5 mm and reaching a concentration at the equilibrium similar
to that shown in Figure 3a. Moreover, the half-time reaction
for this experiment (720 min, conversion of 16 mm) is twice
the half-time of the formation of neat 7A (340 min, con-
version of 17.5 mm), illustrating the competition between the
coupled equilibria. The final sizes, measured by DOSY NMR
spectroscopy, of the micellar self-assemblies produced from
these two competitions are similar to one another (7.2 nm)
but also near equal to the structure of neat 7A (7.1 nm;
Supporting Information, Table S1).
Finally, we set up two competition experiments between
1
A and 7A by mixing 1, 7, and A (50 mm each), and we
determined the corresponding concentration–time profiles by
1
H NMR spectroscopy (Figure 3). In the first experiment, the
beginning of the competition was performed in deuterated
This work describes a general concept for the synergistic
relationships that exist at two length scales within a self-
replicating DCL (Figure 4). The molecular constituents com-
pete at the subnanometer scale for the reversible production
of dynablocks having different rH/h. This ratio, together with
acetonitrile, showing a clear domination of 1A (V =
0
À1
1
2
5 mm h , and c = 31 mm at equilibrium) over 7A (V =
0
À1
.1 mm h , and c = 16 mm). The acetonitrile was then
exchanged for D O, keeping the concentration at a constant
2
Angew. Chem. Int. Ed. 2009, 48, 1093 –1096
ꢀ 2009 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.angewandte.org
1095

Communications
[
2] P. T. Corbett, J. Leclaire, J. Vial, K. R. West, J.-L. Wietor, J. K. M.
Sanders, S. Otto, Chem. Rev. 2006, 106, 3652 – 3711.
[
[
3] J.-M. Lehn, Proc. Natl. Acad. Sci. USA 2002, 99, 4763 – 4768.
4] O. Ramstrꢁm, J.-M. Lehn, Nat. Rev. Drug Discovery 2002, 1, 26 –
36.
[
[
[
[
[
5] J.-M. Lehn, Chem. Soc. Rev. 2007, 36, 151 – 160.
6] N. Paul, G. F. Joyce, Curr. Opin. Chem. Biol. 2004, 8, 634 – 639.
7] V. Patzke, G. von Kiedrowski, ARKIVOC 2007, 338, 293 – 310.
8] P. L. Luisi, Naturwissenschaften 2003, 90, 49 – 59.
9] J. W. Szostak, D. P. Bartel, P. L. Luisi, Nature 2001, 409, 387 – 390.
[
[
10] S. Xu, N. Giuseppone, J. Am. Chem. Soc. 2008, 130, 1826 – 1827.
11] F. G. Varela, H. R. Maturana, R. Uribe, Biosystems 1974, 5, 187 –
196.
[
[
12] P. A. Bachmann, P. L. Luisi, J. Lang, Nature 1992, 357, 57 – 59.
13] H. H. Zepik, E. Blꢁchliger, P. L. Luisi, Angew. Chem. 2001, 113,
205 – 208; Angew. Chem. Int. Ed. 2001, 40, 199 – 202.
[
14] G. F. Joyce, Nature 1989, 338, 217 – 224.
[
15] S. Mann, Angew. Chem. 2008, 120, 5386 – 5401; Angew. Chem.
Int. Ed. 2008, 47, 5306 – 5320.
[
16] Although another type of amphiphilic block copolymers con-
nected through terpyridine–ruthenium complexes is kinetically
labile, the exchange of one block to another occurs only with
Figure 4. a) Synergistic constitutional relationships observed at two
5
10 m excess of the exchanging block and at high temperatures
length scales within b) a model minimal self-replicating DCL. For
clarity, the growth/division cycles of micellar structures are not
represented.
for hours, thus not allowing practical applications in DCC. J.-F.
Gohy, G. G. Lohmeijer, U. S. Schubert, Chem. Eur. J. 2003, 9,
3472 – 3479.
[
[
17] The syntheses of the blocks will be published elsewhere.
18] J. N. Isrealachvili, Intermolecular & surface forces, 2nd ed.,
Academic Press, New York, 1992.
19] The experiment was carried out by adding first water to
acetonitrile (ratio 1:1) and then evaporating under reduced
the stacking effect, mainly determines the formation and the
thermodynamic stability of the bounded structures at the tens
of nanometer scale. Then, in a first autocatalytic loop with a
sigmoid concentration–time profile, these self-assemblies are
able to generate their own formation by increasing the rate of
the dynablock condensation and they entirely fulfill the
required characteristics of a minimal autopoietic process.
Moreover, in a second thermodynamic loop, the self-assem-
blies discriminate between the incorporated dynablocks and
thus favor the preferential synthesis of their own blocks. Such
a system, combining cooperative processes at different length
scales in networks of equilibria and displaying autocatalysis
within DCLs, is of interest for the understanding of the
emergence of self-organizing collective properties but also for
[
pressure about 75% of the solution. The volume of D O was
2
1
then increased to reach a concentration of 50 mm. H NMR
spectroscopy showed that there was no remaining acetonitrile
detectable. In addition, by forming the micelles directly in water,
we observed identical results to those observed by using
acetonitrile/water. Moreover, the effect of acetonitrile is negli-
gible for this range of concentration in water (see the Supporting
Information, Figure S6a).
[
20] Y. Cohen, L. Avram, L. Frish, Angew. Chem. 2005, 117, 524 –
560; Angew. Chem. Int. Ed. 2005, 44, 520 – 544.
[
21] N. Giuseppone, J.-L. Schmitt, L. Allouche, J.-M. Lehn, Angew.
Chem. 2008, 120, 2267 – 2271; Angew. Chem. Int. Ed. 2008, 47,
2235 – 2239.
[
24–26]
the design of responsive systems.
We are currently
[
22] R. Pool, P. G. Bolhuis, Phys. Rev. Lett. 2006, 97, 018302.
working on the development of more complex dynamic
combinatorial networks of competing self-replicating assem-
blies.
[23] A control competition experiment was carried out using similar
sizes of hydrophilic building blocks (i.e. similar micellar struc-
tures) between 7A and 4A. As expected, low selectivity was
observed in favor of 4A (1.8:1), which is mainly related to the
differential nucleophilities between the benzylic and aromatic
amines.
Received: September 18, 2008
Revised: November 10, 2008
Published online: December 29, 2008
[24] R. F. Ludlow, S. Otto, Chem. Soc. Rev. 2008, 37, 101 – 108.
25] D. H. Lee, K. Severin, M. R. Ghadiri, Curr. Opin. Chem. Biol.
[
1
997, 1, 491 – 496.
26] Z. Dadon, N. Wagner, G. Ashkenasy, Angew. Chem. 2008, 120,
221 – 6230; Angew. Chem. Int. Ed. 2008, 47, 6128 – 6136.
Keywords: block copolymers · combinatorial chemistry ·
micelles · self-assembly · self-replication
[
.
6
[
1] J.-M. Lehn, Chem. Eur. J. 1999, 5, 2455 – 2463.
1
096 www.angewandte.org
ꢀ 2009 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Angew. Chem. Int. Ed. 2009, 48, 1093 –1096