Potential for minimal self-replicating systems in a dynamic combinatorial library of equilibrating imines

Article abstract of DOI:10.1016/j.tet.2017.06.045

The presence of a self-replicator in a dynamic combinatorial library (DCL) offers function above and beyond libraries under thermodynamic control, moving towards out-of-equilibrium systems which mimic biological networks. In this work, we examine a previously reported DCL based on reversible imine formation to give amphiphilic structures. The amphiphilic imines were readily produced in organic solvents, and were found to aggregate to micelles in water as judged by diffusion-ordered NMR spectroscopy, dynamic light scattering and interferometric scattering microscopy. Unfortunately, the autocatalytic formation of products was not observed in water, and preformed imines slowly hydrolysed to aldehyde and amine components at neutral pD.

Full text of DOI:10.1016/j.tet.2017.06.045

Tetrahedron xxx (2017) 1e6
Contents lists available at ScienceDirect
Tetrahedron
journal homepage: www.elsevier.com/locate/tet
Potential for minimal self-replicating systems in a dynamic
combinatorial library of equilibrating imines
Sarah M. Morrow, Andrew J. Bissette, Stephen P. Fletcher*
Department of Chemistry, University of Oxford, 12 Mansfield Road, Oxford, OX1 3TA, UK
a r t i c l e i n f o
a b s t r a c t
Article history:
The presence of a self-replicator in a dynamic combinatorial library (DCL) offers function above and
beyond libraries under thermodynamic control, moving towards out-of-equilibrium systems which
mimic biological networks. In this work, we examine a previously reported DCL based on reversible
imine formation to give amphiphilic structures. The amphiphilic imines were readily produced in organic
solvents, and were found to aggregate to micelles in water as judged by diffusion-ordered NMR spec-
troscopy, dynamic light scattering and interferometric scattering microscopy. Unfortunately, the auto-
catalytic formation of products was not observed in water, and preformed imines slowly hydrolysed to
aldehyde and amine components at neutral pD.
Received 30 March 2017
Received in revised form
19 June 2017
Accepted 20 June 2017
Available online xxx
Keywords:
Self-replication
Self-reproduction
Dynamic combinatorial library
Imines
© 2017 Published by Elsevier Ltd.
Amphiphiles
1. Introduction
itself are, in theory, reversible.6e11 These systems have the potential
to operate far from equilibrium, like biological systems, where li-
Dynamic combinatorial libraries (DCLs) typically operate under
thermodynamic control. The resultant equilibrium distribution of
products from a complex mixture maximises the total thermody-
namic stability of the system, which can be a result of assembly into
stable interlocked structures,1 binding to a template,2 or the for-
mation of supramolecular aggregates3 amongst other factors.4
Coupling a DCL with the amplification of components based on
kinetic control gives rise to a more complex, emergent system, with
the promise for increased functionality in comparison to systems
under thermodynamic control.4 Biological systems operate far from
equilibrium e many biological structures persist under dissipative
conditions to perform function before their eventual decay into less
functional materials. The construction of DCLs whose components
can be transiently amplified by autocatalysis offers one approach
towards mimicking these systems (Fig.1).5 In the first report of self-
replication coupled to a dynamic combinatorial library, the repli-
cation step acted as an irreversible kinetic trap, selecting between
equilibrating components.5 Now, the autocatalytic behaviour of
template replicators within DCLs is known where both the ex-
change between equilibrating components and the self-replication
brary components persist under dissipative conditions.4,12
Aiming to overcome the strong product inhibition that can
accompany template replication and reduce self-replicative effi-
ciency, Nguyen et al. reported a system based on the self-repro-
duction of micellar assemblies (Fig. 1).13e15 The synthesis of the
amphiphilic components occurs by reversible imine bond forma-
tion between a hydrophobic aldehyde and a series of hydrophilic
amines, and has the potential to create a DCL of equilibrating imines
3 (Fig. 2). The amphiphilic imines are reported to be stable in water
on account of their aggregation into micellar structures and can
self-reproduce via physical autocatalysis.13,16,17 The kinetics of the
reaction between aldehyde 1 and amine 2c in isolation show
sigmoidal growth of imine 3c, and additional seeding experiments
also convincingly illustrate autocatalysis in this system.
A competition experiment between amines 2a and 2c for the
imine formation with aldehyde 1 illustrates self-reproduction
within a (small) DCL. In this case, imine 3a is thermodynamically
less stable than imine 3c in water, producing 3c autocatalytically at
the expense of imine 3a. It should be noted that the autocatalytic
replication of imine 3c does not push the system out-of-
equilibrium; the ratio of imine 3c to imine 3a at the end of the
autocatalytic period is equal to the thermodynamic distribution.
The high reported thermodynamic stability of the imine library
in water and the construction of self-reproducing amphiphiles
* Corresponding author.
E-mail address: stephen.fletcher@chem.ox.ac.uk (S.P. Fletcher).
http://dx.doi.org/10.1016/j.tet.2017.06.045
0040-4020/© 2017 Published by Elsevier Ltd.
Please cite this article in press as: Morrow SM, et al., Potential for minimal self-replicating systems in a dynamic combinatorial library of
equilibrating imines, Tetrahedron (2017), http://dx.doi.org/10.1016/j.tet.2017.06.045

2
S.M. Morrow et al. / Tetrahedron xxx (2017) 1e6
Fig. 1. A dynamic combinatorial library (DCL) is characterised by the equilibration between library members by the exchange of building blocks. If a member of the library is a self-
replicator, it may be amplified. Replication may occur via template replication or physical reproduction.13 Template replication: molecular recognition between product C and
substrates A and B catalyses their reaction to form another molecule C. Physical reproduction: hydrophobic and hydrophilic components react slowly at the interface to form
amphiphilic molecules. Above the critical aggregation concentration, aggregate structures are formed which allow increased mixing between phases and an increased rate of
formation of amphiphilic products.
Fig. 2. Reaction between aldehyde 1 and amines 2 in a DCL results in the formation of imines 3 which aggregate into supramolecular structures in D2O. In isolation, the reaction
between amine 2c and aldehyde 1 in D2O is shown to be autocatalytic; micelles of product 3c catalyse the condensation between the hydrophobic and hydrophilic components.
based on a reversible imine bond formation was of interest to our
group.3,14,18e26 We have previously studied physical autocatalysis
under kinetic control, in which the bond-forming step is an irre-
versible thiol-Michael reaction.16 The reaction was studied both in
the bulk and at the single-particle level using interferometric
scattering microscopy (iSCAT).27 At this level the autocatalytic self-
reproduction of micellar structures can be observed and quantified
at the nano-scale in real time, allowing direct observation of the
reactive interface and local variations in reactivity which are not
possible on observation of the bulk.28 We were therefore interested
in using the technique to observe an autocatalytic reaction under
thermodynamic control, and decided to examine components of
the imine-based system reported by Nguyen et al.14
Of the reported components we chose to examine amines 2a
and 2b (Fig. 2), allowing comparison between amines of different
nucleophilicity with the same polyethylene glycol chain length. The
amine components were synthesised in accordance with a later
report,24 and the aldehyde according to an independent
publication.29
We first performed a reaction to determine the thermodynamic
stability of the imines 3a and 3b in deuterated acetonitrile. In order
to reach equilibrium, each amine was stirred rapidly with an
equivalent of aldehyde at room temperature for 24 h before taking a
1H NMR spectrum (Scheme 1). Increased imine condensation was
observed for amine 2a as expected on account of the increased
nucleophilicity of the amine and concomitant increased stability of
the imine bond, as expected and as previously reported.14
Using diffusion-ordered spectroscopy (DOSY) we confirmed that
no aggregate structures were formed from imines 3a and 3b in d3-
MeCN, as expected, since all components are fully soluble and there
is therefore no phase separation. These experiments also showed
that the free aldehyde and amine components at equilibrium
diffuse independently of imines 3a and 3b (see SI).
Nguyen et al. subsequently measured the equilibrium compo-
sition of the system in aqueous conditions by dilution of the
experiment performed in d3-MeCN with D2O, before removal of d3-
MeCN in vacuo to reach a final concentration of 50 mM in D2O.14 It
is unclear whether this mixture was allowed to equilibrate after the
solvent exchange and hence whether it is truly an equilibrium
measurement. We performed this measurement as detailed in
Scheme 2 in D2O, analogous to the equilibration conditions in d3-
MeCN.
After rapid stirring of both systems for 48 h, we saw no evidence
of imine formation in D2O. This is in contrast to the results obtained
by the reported procedure, which involved a solvent switch of the
pre-equilibrated mixture from d3-MeCN to D2O followed by NMR
spectroscopy of the mixtures to determine an equilibrium ratio
under aqueous conditions. In the previous report, although rela-
tively little condensation between amine 2a and aldehyde 1 was
seen, for amine 2b nearly complete (>95%) formation of 3b from
aldehyde 1 was reported at equilibrium. The aqueous stability of 3b
in these experiments was attributed to its self-assembly into su-
pramolecular aggregates, which provided protection against
hydrolysis.
In order to test the thermodynamic stability of imines 3a and 3b
they were synthesised under anhydrous conditions before
Please cite this article in press as: Morrow SM, et al., Potential for minimal self-replicating systems in a dynamic combinatorial library of
equilibrating imines, Tetrahedron (2017), http://dx.doi.org/10.1016/j.tet.2017.06.045

S.M. Morrow et al. / Tetrahedron xxx (2017) 1e6
3
Scheme 1. Equilibrium formation of imines 3a and 3b in d3-MeCN. Both starting material components are fully soluble under these conditions. a Ratios determined by integration of
aldehyde and imine peaks in 1H NMR spectra of the equilibrated mixtures.
Scheme 2. Attempted formation of imines 3a and 3b in D2O. Aldehyde 1 is poorly soluble in D2O.
Table 1
dissolution in D2O. Both imines showed extensive hydrolysis in an
Hydrodynamic radii recorded in D2O, 50 mM.
NMR tube over approximately 15 h, and both showed complete
hydrolysis to the starting materials after several days. Imine 3b
showed more rapid decomposition to D2O-soluble amine 2b and
D2O-insoluble aldehyde 1 (Fig. 4, Fig. S1), whilst imine 3a decom-
posed more slowly, with an apparent solubilisation of the forming
aldehyde 1. It appeared clear that in our hands imines 3a and 3b
were thermodynamically unstable under neutral aqueous condi-
tions (Fig. 4), excluding the possibility for their formation in situ
from aldehyde and amine components.
Entry
Species
Hydrodynamic radii Rh/nm
This study
DOSY
Literature values
DLS
DOSY
DLS
SANSa
1
2
3a
3b
6.0
8.1
e
5
e
e
e
e
6.1
6 ꢀ 34b
a
Small angle neutron scattering.
Diameter ꢀ length: small angle neutron scattering found cylindrical micelles.
b
The previous publication reports the use of neutral aqueous
solutions only. However, later reports from the same group also
investigate the stability of these compounds over a range of pH/pD
values.24e26 We were therefore interested to see whether aqueous
solutions at higher pD offered enhanced stability.
Solutions of imines 3a and 3b were prepared in D2O at pH* 12
(pD 12.45)30 and their thermodynamic stability was monitored by
NMR, as above. Imine 3a showed no increase in stability, hydro-
lysing to aldehyde and amine components. However, imine 3b, in
contrast to the experiment at neutral pD, showed no hydrolysis
over at least 13 h.
alone in D2O (Table 2, entries 1 and 4; Fig. 3). The association of the
hydrophobic component of the reaction with the micelle structure
is a crucial requirement for the development of an autocatalytic
system.
DOSY of imine 3b was less facile on account of the relatively
broad peaks in the spectrum, but consistent with a hydrodynamic
radius of 8.1 nm (Table 1, entry 2). Fortunately, however, we were
also able to perform DLS experiments on this species, and observed
micelles with a hydrodynamic radius of 6.1 nm. The previous
publication reports aggregates too large to observe by DOSY spec-
troscopy, which may explain the broad peaks we observed; neutron
scattering experiments revealed cylindrical micelles of 6 nm in
diameter and 34 nm in length (Table 2).14 It is difficult to compare
Encouraged by this result, we attempted to observe the forma-
tion of imine 3b from aldehyde and amine components at pD 12.45
over 20 h in an experiment analogous to that in Scheme 2. How-
ever, only trace quantities of imine 3b were observed.
Failing to observe autocatalysis, it was important to determine
whether imines 3a and 3b did, in fact, aggregate in aqueous solu-
tion. The aggregation properties of 3a and 3b were studied using
DOSY as well as dynamic light scattering (DLS) in water. DOSY of
imine 3a was most facile since it was well solvated with sharp, well-
resolved peaks in its 1H NMR spectrum in D2O. Imine 3a was found
to diffuse with a hydrodynamic radius of 6.0 nm in D2O, compa-
rable to the value reported of 5 nm and consistent with the for-
mation of a supramolecular aggregate structure (Table 1, entry 1).
Fortuitously, a small quantity of aldehyde 1 was present in the
sample since imine 3a was not purified following synthesis. DOSY
revealed this aldehyde to be associated with the aggregate struc-
ture, diffusing at a rate of similar magnitude to the imine 3a and at a
much decreased rate to the control experiment of the aldehyde
Table 2
Diffusion coefficients established by DOSY NMR spectroscopy. All components were
measured at a concentration of 50 mM in D2O (assuming 100% purity). The presence
of a small quantity of aldehyde 1 and amines 2a and 2b in experiments on 3a and 3b
(entries 4 and 5) are a result of their synthesis without purification.
Entry
Species present
Diffusion coefficient D/10ꢁ10 m2sꢁ1
1
2a
2b
3a
3b
1
2
3
4
5
1
1.41
e
e
e
e
e
2a
2b
3a
3b
4.23
e
n.d.a
e
e
e
e
e
3.70
e
n.d.a
e
e
0.26
n.d.b
0.41
e
e
0.30
a
Could not determine D due to overlap of peaks.
Aldehyde peak too small for accurate integration.
b
Please cite this article in press as: Morrow SM, et al., Potential for minimal self-replicating systems in a dynamic combinatorial library of
equilibrating imines, Tetrahedron (2017), http://dx.doi.org/10.1016/j.tet.2017.06.045

4
S.M. Morrow et al. / Tetrahedron xxx (2017) 1e6
Fig. 3. 2D DOSY NMR showing uptake of aldehyde 1 into micelles of 3a (entry 4, Table 1). Imine 3a was dissolved in D2O at a concentration of 50 mM.
Fig. 4. a) Imine 3a ([imine]0 ¼ 50 mM) showed significant hydrolysis to amine and aldehyde components over 15 h, 1H NMR spectra recorded every 15 min b) Imine 3b
([imine]0 ¼ 50 mM) showed very rapid hydrolysis over 15 h to amine and aldehyde components, 1H NMR spectra recorded every 15 min.
our DLS data with this measurement since our methods cannot give
accurate information for non-spherical objects. We could be
reasonably confident, however, that we also observed aggregation
of imine 3b.
Finally, unable to observe any autocatalysis of these systems in
the bulk, it was of interest to examine them by iSCAT microscopy.
Our previous study of the thiol-Michael reaction found that per-
forming biphasic autocatalytic reactions on the microliter scale on a
glass coverslip significantly reduced the lag period before the onset
of autocatalysis and the production of product. We were unable to
observe aggregates of imine 3a by iSCAT, but experiments using
imine 3b allowed us to observe the presence of supramolecular
aggregates consistent with micelles binding to the coverslip, as
shown in Fig. 5. However, upon mixing aldehyde 1 and amine 2b on
the glass coverslip, no micelle formation appeared to occur (Fig. S2).
In conclusion, Nguyen et al. reported the physical self-
amplification of imine 3c, whose dominance in a DCL is based on
its high thermodynamic stability in water. We were interested in
this system as an autocatalytic reaction in which the bond-forming
step was reversible, and examined parts of their reported DCL
Please cite this article in press as: Morrow SM, et al., Potential for minimal self-replicating systems in a dynamic combinatorial library of
equilibrating imines, Tetrahedron (2017), http://dx.doi.org/10.1016/j.tet.2017.06.045

S.M. Morrow et al. / Tetrahedron xxx (2017) 1e6
5
Fig. 5. Images A and B are independent images of the binding of imine 3b (~1 mM in H2O) to the microscope coverslip as aggregates (individual aggregates numbered within each
image). Each image is the average of 100 frames. Scale bar ¼ 1
mm. The bar below the images represent the contrast range of each image; the individual contrast of each aggregate
provides a measure of its size.
system which we felt might be able to undergo physical
autocatalysis.
References
1. Ponnuswamy N, Cougnon FBL, Clough JM, Pantos¸ GD, Sanders JKM. Discovery
of an organic trefoil knot. Science. 2012;338:783e785.
2. Corbett PT, Sanders JKM, Otto S. Competition between receptors in dynamic
combinatorial libraries: amplification of the fittest? J Am Chem Soc. 2005;127:
9390e9392.
3. Minkenberg CB, Florusse L, Eelkema R, Koper GJM, Van Esch JH. Triggered self-
assembly of simple dynamic covalent surfactants. J Am Chem Soc. 2009;131:
11274e11275.
4. Li J, Nowak P, Otto S. Dynamic combinatorial libraries: from exploring molec-
ular recognition to systems chemistry. J Am Chem Soc. 2013;135:9222e9239.
5. Sadownik JW, Philp D. A simple synthetic replicator amplifies itself from a
dynamic reagent pool. Angew Chem - Int Ed. 2008;47:9965e9970.
6. Sadownik JW, Mattia E, Nowak P, Otto S. Diversification of self-replicating
molecules. Nat Chem. 2016;8:264e269.
7. Del Amo V, Slawin AMZ, Philp D. Manipulating replication processes within a
dynamic covalent framework. Org Lett. 2008;10:4589e4592.
8. Xu S, Giuseppone N. Self-duplicating amplification in a dynamic combinatorial
library. J Am Chem Soc. 2008;130:1826e1827.
9. Carnall JMA, Waudby CA, Belenguer AM, Stuart MCA, Peyralans JJ-P, Otto S.
Mechanosensitive self-replication. Science. 2010;327:1502e1506 (80-).
10. Pal A, Malakoutikhah M, Leonetti G, et al. Controlling the structure and length
of self-synthesizing supramolecular polymers through nucleated growth and
disassembly. Angew Chem - Int Ed. 2015;54:7852e7856.
11. Dadon Z, Samiappan M, Wagner N, Ashkenasy G. Chemical and light triggering
of peptide networks under partial thermodynamic control. Chem Commun.
2012;48:1419e1421.
12. Mattia E, Otto S. Supramolecular systems chemistry. Nat Nanotechnol. 2015;10:
111e119.
Though imines 3a and 3b formed micellar structures in water
according to a variety of techniques (DOSY, DLS, iSCAT), we found
that these structures were thermodynamically unstable and
hydrolysed fully to aldehyde and amine components in neutral
aqueous solution. The only apparent difference between this work
and the previous study of 3a and 3b is the way the aqueous samples
were prepared. Since it was not clear to us that the reported
solvent-switch from d3-MeCN to D2O would measure the true
equilibrium position, we instead allowed the aldehyde and amine
components to equilibrate in D2O, and, from the other direction,
dissolved pre-formed imines in D2O. Both procedures resulted in no
detectable imine at equilibrium.
The enhancement in stability for imine 3b at pD 12.45 did not
result in the observation of autocatalysis over the first 20 h of the
reaction. As the kinetics of formation of 3b were not reported by
Nguyen et al. it is difficult to draw a direct comparison. However the
strong dependence on pH/pD, variation in sample preparation, and
subsequent reports of the group in which other parameters are
varied such as stoichiometry,24e26 reveal the complex and
numerous factors which may influence imine stability and the
observation of autocatalysis.
13. Bissette AJ, Fletcher SP. Mechanisms of autocatalysis. Angew Chem Int Ed.
2013;52:12800e12826.
14. Nguyen R, Allouche L, Buhler E, Giuseppone N. Dynamic combinatorial evo-
lution within self-replicating supramolecular assemblies. Angew Chem Int Ed.
2009;48:1093e1096.
Financial interests
The authors declare no competing financial interests.
15. Bachmann PA, Luisi PL, Lang J. Autocatalytic self-replicating micelles as models
for prebiotic structures. Nature. 1992;357:57e59.
16. Bissette AJ, Odell B, Fletcher SP. Physical autocatalysis driven by a bond-
forming thioleene reaction. Nat Commun. 2014;5:4607.
17. Stano P, Luisi PL. Achievements and open questions in the self-reproduction of
vesicles and synthetic minimal cells. Chem Commun (Camb). 2010;46:
3639e3653.
18. Meguellati K, Fallah-Araghi A, Baret J-C, et al. Enhanced imine synthesis in
water: from surfactant-mediated catalysis to host-guest mechanisms. Chem
Commun (Camb). 2013;49:11332e11334.
19. Toyota T, Takakura K, Sugawara T. Temporal emergence of giant vesicles
accompanied by hydrolysis of ammonium amphiphiles with a Schiff-base
segment. Chem Lett. 2004;33:1442e1443.
Acknowledgements
SMM is supported by the EPSRC Centre for Doctoral Training in
Synthesis for Biology and Medicine (EP/L015838/1) and the Oxford-
Radcliffe Scholarship. SPF thanks the EPSRC (EP/M0025241/1) and
ERC (Consolidator Grant, Autocat, 681491) for funding.
Appendix A. Supplementary data
20. Minkenberg CB, Li F, van Rijn P, et al. Responsive vesicles from dynamic co-
valent surfactants. Angew Chem Int Ed. 2011;50:3421e3424.
21. Sheng L, Kurihara K. Generation of catalytic amphiphiles in a self-reproducing
giant vesicle. Chem Lett. 2016;2:4e6.
Supplementary data related to this article can be found at http://
dx.doi.org/10.1016/j.tet.2017.06.045.
Please cite this article in press as: Morrow SM, et al., Potential for minimal self-replicating systems in a dynamic combinatorial library of
equilibrating imines, Tetrahedron (2017), http://dx.doi.org/10.1016/j.tet.2017.06.045

6
S.M. Morrow et al. / Tetrahedron xxx (2017) 1e6
22. Takakura K, Toyota T, Yamada K. Morphological change of giant vesicles trig-
gered by dehydrocondensation reaction. Chem Lett. 2002;4:404e405.
23. Kurihara K, Tamura M, Shohda K, Toyota T, Suzuki K, Sugawara T. Self-repro-
duction of supramolecular giant vesicles combined with the amplification of
encapsulated DNA. Nat Chem. 2011;3:775e781.
24. Nguyen R, Jouault N, Zanirati S, et al. Core-shell inversion by pH modulation in
dynamic covalent micelles. Soft Matter. 2014;10:3926e3937.
25. Nguyen R, Buhler E, Giuseppone N. Dynablocks: structural modulation of
responsive combinatorial self-assemblies at mesoscale. Macromolecules.
2009;42:5913e5915.
Soft Matter. 2011;7:4787.
27. Ortega-Arroyo J, Kukura P. Interferometric scattering microscopy (iSCAT): new
frontiers in ultrafast and ultrasensitive optical microscopy. Phys Chem Chem
Phys. 2012;14:15625e15636.
28. Ortega-Arroyo J, Bissette AJ, Kukura P, Fletcher SP. Visualization of the spon-
taneous emergence of a complex, dynamic, and autocatalytic system. Proc Natl
Acad Sci U S A. 2016. http://dx.doi.org/10.1073/pnas.1602363113, 201602363.
29. Koide Y, Hasegawa T, Takahash A, et al. Development of novel EDG3 antago-
nists using a 3D database search and their structure-activity relationships.
J Med Chem. 2002;45:4629e4638.
26. Jouault N, Nguyen R, Rawiso M, Giuseppone N, Buhler E. SANS, SAXS, and light
scattering investigations of pH-responsive dynamic combinatorial mesophases.
30. Kre¸ zel A, Bal W. A formula for correlating pKa values determined in D 2O and
H2O. J Inorg Biochem. 2004;98:161e166.
Please cite this article in press as: Morrow SM, et al., Potential for minimal self-replicating systems in a dynamic combinatorial library of
equilibrating imines, Tetrahedron (2017), http://dx.doi.org/10.1016/j.tet.2017.06.045