Generating System-Level Responses from a Network of Simple Synthetic Replicators

Article abstract of DOI:10.1021/jacs.7b09735

The creation of reaction networks capable of exhibiting responses that are properties of entire systems represents a significant challenge for the chemical sciences. The system-level behavior of a reaction network is linked intrinsically to its topology and the functional connections between its nodes. A simple network of chemical reactions constructed from four reagents, in which each reagent reacts with exactly two others, can exhibit up-regulation of two products even when only a single chemical reaction is addressed catalytically. We implement a system with this topology using two maleimides and two nitrones of different sizes - either short or long and each bearing complementary recognition sites - that react pairwise through 1,3-dipolar cycloaddition reactions to create a network of four length-segregated replicating templates. Comprehensive ¹H NMR spectroscopy experiments unravel the network topology, confirming that, in isolation, three out of four templates self-replicate, with the shortest template exhibiting the highest efficiency. The strongest template effects within the network are the mutually cross-catalytic relationships between the two templates of intermediate size. The network topology is such that the addition of different preformed templates as instructions to a mixture of all starting materials elicits system-level behavior. Instruction with a single template up-regulates the formation of two templates in a predictable manner. These results demonstrate that the rules governing system-level behavior can be unraveled through the application of wholly synthetic networks with well-defined chemistries and interactions.

Full text of DOI:10.1021/jacs.7b09735

Subscriber access provided by READING UNIV
Article
Generating system-level responses from
a network of simple synthetic replicators
Jan W. Sadownik, Tamara Kosikova, and Douglas Philp
J. Am. Chem. Soc., Just Accepted Manuscript • DOI: 10.1021/jacs.7b09735 • Publication Date (Web): 31 Oct 2017
Downloaded from http://pubs.acs.org on November 3, 2017
Just Accepted
“Just Accepted” manuscripts have been peer-reviewed and accepted for publication. They are posted
online prior to technical editing, formatting for publication and author proofing. The American Chemical
Society provides “Just Accepted” as a free service to the research community to expedite the
dissemination of scientific material as soon as possible after acceptance. “Just Accepted” manuscripts
appear in full in PDF format accompanied by an HTML abstract. “Just Accepted” manuscripts have been
fully peer reviewed, but should not be considered the official version of record. They are accessible to all
readers and citable by the Digital Object Identifier (DOI®). “Just Accepted” is an optional service offered
to authors. Therefore, the “Just Accepted” Web site may not include all articles that will be published
in the journal. After a manuscript is technically edited and formatted, it will be removed from the “Just
Accepted” Web site and published as an ASAP article. Note that technical editing may introduce minor
changes to the manuscript text and/or graphics which could affect content, and all legal disclaimers
and ethical guidelines that apply to the journal pertain. ACS cannot be held responsible for errors
or consequences arising from the use of information contained in these “Just Accepted” manuscripts.
Journal of the American Chemical Society is published by the American Chemical
Society. 1155 Sixteenth Street N.W., Washington, DC 20036
Published by American Chemical Society. Copyright © American Chemical Society.
However, no copyright claim is made to original U.S. Government works, or works
produced by employees of any Commonwealth realm Crown government in the course
of their duties.

Page 1 of 29
Journal of the American Chemical Society
1
2
3
4
Journal of the American Chemical Society
5
6
7
8
9
Generating system-level responses from a
network of simple synthetic replicators
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
Jan W. Sadownik, Tamara Kosikova and Douglas Philp*
School of Chemistry and EaStCHEM, University of St Andrews, North Haugh,
St Andrews, Fife KY16 9ST, United Kingdom
*Corresponding author e-mail: d.philp@st-andrews.ac.uk
ORCID IDs:
TK: orcid.org/0000-0001-7886-9660
DP: orcid.org/0000-0002-9198-4302
ACS Paragon Plus Environment

Journal of the American Chemical Society
Page 2 of 29
1
2
3
4
Abstract
The creation of reaction networks capable of exhibiting responses that are properties of
entire systems represents a significant challenge for the chemical sciences. The system-
level behavior of a reaction network is linked intrinsically to its topology and the
functional connections between its nodes. A simple network of chemical reactions
constructed from four reagents, in which each reagent reacts with exactly two others, can
exhibit upregulation of two products even when only a single chemical reaction is
addressed catalytically. We implement a system with this topology using two maleimides
and two nitrones of different sizes—either short or long and each bearing complementary
recognition sites—that react pairwise through 1,3-dipolar cycloaddition reactions to create
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
1
a network of four length-segregated replicating templates. Comprehensive H NMR
spectroscopy experiments unravel the network topology, confirming that, in isolation,
three out of four templates self-replicate, with the shortest template exhibiting the highest
efficiency. The strongest template effects within the network are the mutually
crosscatalytic relationships between the two templates of intermediate size. The network
topology is such that the addition of different preformed templates as instructions to a
mixture of all starting materials elicits system-level behavior. Instruction with a single
template up-regulates the formation of two templates in a predictable manner. These
results demonstrate that the rules governing system-level behavior can be unraveled
through the application of wholly synthetic networks with well-defined chemistries and
interactions.
2
ACS Paragon Plus Environment

Page 3 of 29
Journal of the American Chemical Society
1
2
3
4
Introduction
Complex systems1 are constructed from intricate networks of interconnected components,
with their global functional and structural properties being determined by the nature of
the recognition and reaction processes embedded within them. However, these systems
often operate within environments that do not allow the individual interactions that give
rise to the system-level behavior to be decoupled readily. Systems chemistry2 attempts to
create complexity and emergent phenomena through the application of wholly synthetic
chemical platforms. In this approach, a collection of synthetic chemical entities, designed
to interact and react in programmed ways, can express complex dynamic phenomena.
This approach targets the creation of synthetic chemical systems whose properties are not
simply the linear sum of the attributes of the individual components. Central to this goal is
the design and synthesis of chemical structures that are capable of copying themselves or
complementary partners, i.e., molecular replicators. By examining the dynamic processes
that govern replication in synthetic chemical systems, it is possible to maintain structural
and interactional simplicity in network components and design systems with well-defined
chemistries and interactions. In this manner, a better understanding of the principles
governing the creation and function of systems that express emergent phenomena can be
developed, ultimately shedding light on the origins of biological complexity.
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
Currently, the general design principles2b,2c,3 required to create a self-replicating
system and engineer its behavior are well understood. In its simplest form, a self-
replicating system is constructed from two molecules, where the product of their reaction
acts a specific catalyst (template) for its own formation. Such templates allow the creation
of systems that are responsive to feedback, i.e., the addition of pre-formed template acts as
chemical input directing the synthesis of exact copies of that template. To date, self-
replication has been demonstrated experimentally in systems fabricated from a variety of
building blocks, ranging from more prebioticaly relevant oligonucleotide4 and peptide5,6
frameworks to completely synthetic, small organic7 molecules. With the advances in
analytic tools and methods, it is now possible to analyze and monitor the behavior of
complex chemical networks and systems where multiple catalytic pathways function
simultaneously. Despite the significant advances8 in the study of peptide replicator
3
ACS Paragon Plus Environment

Journal of the American Chemical Society
Page 4 of 29
1
2
3
4
5
6
7
8
9
networks, these systems are often very densely connected and susceptible to the reaction
environment, e.g., pH. We have therefore become interested in exploiting systems
constructed from synthetic replicators based on small organic molecules. Such systems9
can, in principle, offer tight control over network topology through careful manipulation
of chemical structure and, hence, catalytic relationships, thus allowing the fabrication of
networks with system-level behavior that moves beyond simple information transfer.
The emergent system-level2,3 behavior displayed by a network is linked intrinsically
to the topology of the network and, in particular, to the functional connections between
the network nodes. An example of this type of phenomenon can be displayed (Figure 1)
even by a simple chemical reaction network.
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
(a)
(b) 10
kAB = 1
8
6
4
2
0
AB
kAB
A
C
B
D
kAC
kBD
AC
BD
AB AC BD CD
Product
10
8
kAB = 2
kCD
6
CD
4
2
d[A]
k
k
= − AB[A][B] − AC[A][C]
0
dt
AB AC BD CD
Product
10
8
kAB = 10
d[AB]
= kAB[A][B]
dt
6
4
d[B]
= − kAB[A][B] − kBD[B][D]
2
dt
0
AB AC BD CD
Product
Figure 1 (a) A simple network of interconnected reactions can be created by four reagents A to D. Four
chemical transformations convert these starting materials into four products, AB, AC, BD, and CD.
Although the rate of formation of an individual product, e.g., AB, is described by a simple second order rate
law, the concentration terms in this differential equation are dependent intrinsically on two other reactions
in the network. (b) If the values of all four rate constants are the same (kAB = kAC = kBD = kCD = 1), the final
concentrations of all four products are identical. Increasing the value of one rate constant results in an
increase in the final concentrations of two of the products. Thus, increasing kAB from 1 to 2 to 10, while kAC
=
k
BD = kCD = 1, results ultimately in [AB] = [CD] = 8.1 and [AC] = [BD] = 1.9.
4
ACS Paragon Plus Environment

Page 5 of 29
Journal of the American Chemical Society
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
Within a network (Figure 1) constructed from four reagents that each react with
exactly two others in the network, the rate of formation of an individual product, for
example, AB, can be described (Figure 1a) by a second order rate law. However, each of
the concentration terms within this rate law depends explicitly (Figure 1a) on two
reactions in the network. For example, the change in concentration of A with respect to
time depends on two reactions: A + B → AB and A + C → AC. Consequently, the transient
concentrations of the reagents in any given reaction are connected either directly or
indirectly to the transient concentrations of all of the other reagents in the network.
Therefore, changes in the rate of consumption of one reagent must propagate throughout
the entire network. If the values of all four rate constants (Figure 1b) are the same (kAB = kAC
= kBD = kCD), these interconnections will not express themselves and the final concentrations
of all four products are identical. By contrast, increasing the value of one rate constant, e.g.,
kAB, results in an increase (Figure 1b) in the final concentrations of two of the products, in
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
this case AB and CD.
In order for this network to operate in the manner described, it must be possible to
address each of the four reactions individually and specifically in a catalytic manner. In
this context, self-replicators provide the ideal vehicle for the implementation of this
network topology, since they can act as specific catalysts for their own formation. We
therefore envisaged (Figure 2) a network of four replicating templates that could be
constructed through the combination of two pairs of reagents that each bear
complementary recognition and reactive sites. These pairs of reagents are distinguished by
the length of the spacer—either short (S) or long (L)—connecting the recognition site with
the reactive site. Thus, the shortest self-replicator SNSM can be constructed by
combination of SN and SM and would be expected to replicate itself through an
autocatalytic cycle (Figure 2, SNSM → SNSM). Similarly, the longest self-replicator
LNLM can be constructed by combination of LN and LM and would also be expected to
template its own formation through an autocatalytic cycle (Figure2, LNLM → LNLM). As
there is a large size difference between these two replicators, we envisaged that, although
each would be capable of catalyzing its own formation, crosscatalysis between SNSM and
LNLM would be impossible.
5
ACS Paragon Plus Environment

Journal of the American Chemical Society
Page 6 of 29
1
2
3
4
5
6
7
8
9
By contrast, combination of SN with LM and LN with SM will lead to two self-
complementary templates of intermediate size. Thus, whilst we would expect SNLM and
LNSM to function as autocatalytic templates (Figure 2, SNLM → SNLM and LNSM →
LNSM), there is also the possibility that crosscatalytic relationships could exist between
these templates (Figure 2, SNLM → LNSM and LNSM → SNLM).
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
Figure 2 Four building blocks, SN, LN, SM, and LM, which all bear a recognition site and a reactive site
connected by either a short (S) or a long (L) spacer, can react with each other in a pairwise fashion to afford
four templates, SNSM, SNLM, LNSM, and LNLM. As a result of their different sizes, SNSM and LNLM
can participate only in autocatalytic cycles (SNSM → SNSM and LNLM → LNLM) in which they catalyze
their own formation. By contrast, SNLM and LNSM are of similar sizes. Therefore, in addition to catalyzing
their own formation through autocatalytic cycles (SNLM → SNLM and LNSM → LNSM), they can
potentially catalyze the formation of each other through crosscatalytic cycles (SNLM → LNSM and LNSM
→ SNLM).
Molecular design
6
ACS Paragon Plus Environment

Page 7 of 29
Journal of the American Chemical Society
1
2
3
4
5
6
7
8
9
In order to create a functional version of the replicator network shown in Figure 2, it is
necessary to identify the correct chemical building blocks that are capable of reacting
together to form the four replicating templates, SNSM, SNLM, LNSM, and LNLM.
In previous work, the replicating platform derived from nitrone SN and maleimide SM
has been utilized7b,7e,9a,9e in a number of scenarios. Using this successful design as a basis
for this work, we envisaged that we could create the required library of templates by
simply incorporating an extended linker into the nitrone and maleimide components to
afford long nitrone LN and long maleimide LM. Thus, pairwise combination of nitrones
SN and LN with maleimides SM and LM affords two pairs of templates10—SNSM and
LNLM, which have matched linker lengths, and SNLM and LNSM, which have
mismatched linker lengths.
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
7
ACS Paragon Plus Environment

Journal of the American Chemical Society
Page 8 of 29
1
2
3
4
5
6
7
8
9
We envisaged that this design, incorporating the long diphenylacetylene linker in
both LN and LM would limit the number of crosscatalytic opportunities within this
network, described conceptually in Figure 2. Thus, only SNLM and LNSM can potentially
participate in mutually crosscatalytic pathways as the combination of one short and one
long linker renders these two templates approximately equivalent in size. By contrast,
SNSM and LNLM—bearing either two short (phenyl) or two long (diphenylacetylene)
linkers—are dramatically different in size. Therefore, we expect that they are
fundamentally incompatible with each other and also with the linker mismatched pair
(SNLM and LNSM).
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
Behavior of individual templates
Before exploring the network created by the four length-segregated templates in its
entirety, it is necessary to establish the replicating efficiency of each of the four templates
in isolation. To this end, we performed a series of kinetic experiments where the formation
of each template was examined in CDCl at 273 K from its constituent components at
3
starting concentrations of 10 mM. In addition, in each case, the same experiment was
repeated with the addition of pre-formed template in order to establish the autocatalytic
nature of each replicator. Finally, the baseline reactivities of the nitrone and maleimide
components in the absence of the recognition elements were established, as series of
reactions were performed on control maleimides that lacked the carboxylic acid
recognition sites required for replication (Supplementary Figure S1 to S4). The results of
these experiments are summarized in Figure 3.
8
ACS Paragon Plus Environment

Page 9 of 29
Journal of the American Chemical Society
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
Figure 3 (a) Summary of kinetic data for individual templates determined by 500 MHz 1H NMR
spectroscopy (273 K, CDCl , 10 mM starting concentrations of reagents). In each case, the identity of the
3
experiment is as follows: circles = uninstructed; triangles = autocatalytic template added; squares =
crosscatalytic template added. (b) Schematic representation of the catalytic inter-relationships between the
templates SNSM, SNLM, LNSM, and LNLM. An arrow indicates a catalytic relationship, e.g., LNSM →
SNLM indicates that LNSM catalyzes the formation of SNLM. Efficiency: + = low (EMkinetic < 5 M); ++ =
medium (5 M < EMkinetic < 15 M); +++ = high (EMkinetic > 15 M).
From the kinetic results shown in Figure 3a, it is clear that three of the templates,
SNSM, SNLM, and LNLM, show significant aptitudes for directing their own
formation—i.e., the addition of the appropriate autocatalytic template at the start of each
of the reactions results in significant enhancements of the rates of formation for each of
these templates (triangles, Figure 3a). By contrast, LNSM in isolation did not display any
9
ACS Paragon Plus Environment

Journal of the American Chemical Society
Page 10 of 29
1
2
3
4
5
6
7
8
9
obvious recognition-mediated reactivity and its rate of formation is close to that of the
bimolecular background reaction (Supporting Figure S3). The solubility of this template
limited the range of experiments that could be performed in order to establish the
autocatalytic behavior of this template. Although we cannot discount completely the
possibility that this template is capable of instructing its own formation, the efficiency of
this process is undoubtedly extremely low.
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
Despite the lack of obvious autocatalytic activity exhibited by LNSM, a powerful
crosscatalytic relationship exists between SNLM and LNSM. The addition of SNLM to the
reaction between LN and SM results in a dramatic increase in the rate of formation of
LNSM (Figure 3a, squares, left). Similarly, the addition of LNSM to the reaction between
SN and LM results in a significant increase in the rate of formation of SNLM (Figure 3,
squares, right)—close in magnitude to that observed for the template-directed
autocatalytic formation of SNLM (Figure 3a, triangles, bottom right). Simulation and
fitting (see Supporting Information for details) of the kinetic profiles shown in Figure 3a
(and Supporting Figs. S1 to S4) allowed us to extract key kinetic parameters for these
replicators (Figure 3a, data in boxes). A measure of the efficiency of each replicator comes
from its value of EMkinetic, i.e., the effective molarity for the cycloaddition reaction achieved
by the replicating template within the relevant catalytically-important ternary complex.
The catalytic relationships within this network are summarized in Figure 3b.
SNSM (EMkinetic = 10.8 M) and LNLM (EMkinetic = 3.6 M) replicate independently, with
SNSM being the more efficient of these two templates. As expected, the rate of formation
of the linker-matched template SNSM is unaffected by the addition of LNLM. Similarly,
the rate of formation of the linker-matched template LNLM is also unaffected by the
addition of SNSM, confirming the absence of crosscatalysis between these two templates.
Only one of the linker-mismatched pair of templates, SNLM (EMkinetic = 3.0 M), is
capable of self-replication and the mutual crosscatalytic relationships that exist between
SNLM (EMkinetic = 18.7 M for the formation of LNSM) and LNSM (EMkinetic = 31.1 M for the
formation of LNSM) are the most powerful catalytic connections within this network.
In order to gain an insight into the profound difference in the autocatalytic
performance of SNLM compared to LNSM, we performed a series of calculations at the
10
ACS Paragon Plus Environment

Page 11 of 29
Journal of the American Chemical Society
1
2
3
4
5
ωB97X/def2-SVP level of theory (see Supporting Information for details). The results of
these calculations are summarized in Figure 4.
6
7
8
(a)
(c)
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
B
ω
5
4
3
2
1
0
A
LNSM
(b)
SNLM
0
10
20
30
40
Angle ωin degrees
Figure 4 Stick representations of the calculated structures of the transition states accessed by (a)
[LN•SM•LNSM] and (b) [SN•LM•SNLM] leading to the corresponding template duplexes. Calculations
were performed at the ωB97X/def2-SVP level of theory using the PCM solvation model for chloroform.
Hydrogen bonds are represented by dashed lines. Colored shaded areas indicated by letters are discussed in
the main text. Carbon atoms are colored gray, nitrogen atoms blue, oxygen atoms red and hydrogen atoms
white. Most hydrogen atoms are omitted for clarity. (c) Computed relationship (dashed line) between
geometry and complex stability for the association of acetic acid and 2-acetoamido-6-methylpyridine. The
colored data points represent the values observed for the angle ω in the associations between the
amidopyridines and carboxylic acids in [LN•SM•LNSM]‡ (•) and [SN•LM•SNLM]‡ (•). Calculations were
performed at the ωB97X/def2-TZVP level of theory using the PCM solvation model for chloroform.
The calculations reveal that plausible transition states are accessible from both the
[LN•SM•LNSM] (Figure 4a) and [SN•LM•SNLM] (Figure 4b) ternary complexes,
suggesting that both SNLM and LNSM should be capable of acting as templates for their
own formation. Comparison of the transition state geometries accessed by the
[LN•SM•LNSM] and [SN•LM•SNLM] ternary complexes with the geometry transition
state calculated11 for the parent reaction between diphenyl nitrone and N-phenyl
maleimide reveals (see Supporting Information, Fig. S8) little evidence of significant
distortions that might result in the reaction within the [LN•SM•LNSM] ternary complex
being less efficient than the corresponding reaction within the [SN•LM•SNLM] complex.
11
ACS Paragon Plus Environment

Journal of the American Chemical Society
Page 12 of 29
1
2
3
4
5
6
7
8
9
However, there are three significant structural differences between the two transition
states accessed from these ternary complexes. Firstly, both alkynes present in the
transition state accessed from the [LN•SM•LNSM] ternary complex exhibit bending
away from linearity (4° and 8°, highlight A in Figure 4a). By contrast, in the transition state
accessed from the [SN•LM•SNLM] ternary complex, there is almost no bending (<2°) in
either of the alkyne spacers. Whilst the bending potential for alkynes is somewhat soft, the
presence of this bending could serve to disfavor reaction from [LN•SM•LNSM].
Secondly, the alkene protons present in the SM component of the transition state accessed
from [LN•SM•LNSM] are located within a pocket that contains other hydrogen atoms
that bear partial positive charges (highlight B in Figure 4a). These electrostatic interactions
will serve to destabilize [LN•SM•LNSM]‡. Finally, and perhaps most significantly, there
are significant differences between the geometries observed for the recognition elements
present in the transition states accessed from two ternary complexes. In the case of
[LN•SM•LNSM]‡, both carboxylic acid units are twisted by more than 30° with respect to
the mean plane of the pyridine ring to which they are hydrogen bonded. By contrast,
within [SN•LM•SNLM]‡, the twists are only 12° and 23°. In order to assess the likely
impact of these differences in geometry within the recognition elements of these two
complexes, we performed a series of calculations at the ωB97X/def2-TZVP level of theory
in which the angle ω (Figure 4c) subtended between mean plane of the COOH fragment of
acetic acid and the mean plane of the pyridine ring to which the acid is hydrogen bonded
is increased. These calculations reveal that, as expected, the lowest energy geometry is the
coplanar arrangement of the two recognition elements (ω = 0°), as this arrangement allows
the optimum geometry for both hydrogen bonds present in this complex. As ω increases,
the complex becomes significantly less stable. Based solely on the different geometries of
the recognition elements observed in the two transition state structures computed for
[LN•SM•LNSM]‡ and [SN•LM•SNLM]‡, the geometries observed in recognition
elements of [LN•SM•LNSM]‡ are expected, at this level of theory, to be 4.64 kJ mol–1 less
stable than those observed in [SN•LM•SNLM]‡. If this difference is expressed fully in the
kinetics of these reaction processes, the expected autocatalytic rate constant for the
formation of LNSM would be around 7 times smaller than that for SNLM, suggesting that
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
12
ACS Paragon Plus Environment

Page 13 of 29
Journal of the American Chemical Society
1
2
3
4
5
6
7
8
9
the EMkinetic for the formation of LNSM would be << 1 M. This expectation is consistent
with the observed data and it is therefore possible that the inability of LNSM to template
its own formation efficiently arises from destabilization of the [LN•SM•LNSM] ternary
complex as it enters the transition state for the formation of the self-complementary
template.
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
Network behavior
Having established the catalytic relationships between and kinetic parameters for the
individual templates, we were able to create a kinetic model capable of simulating the
behavior of the complete network shown in Figure 2. These simulations revealed (Figure
5, center) that, in the absence of any template instruction, simply mixing the two nitrones
and two maleimides at starting concentrations of 10 mM results in a population of
replicators after 16 hours in which all four replicators have significant representations—
the predicted ratio of SNSM : SNLM : LNSM : LNLM is 1.8 : 1.3 : 1.5 : 1.
In order to benchmark our simulation, we performed an experiment in which SN,
LN, SM and LM were dissolved at starting concentrations of 10 mM in CDCl
3
at 273 K.
1
The formation of the four corresponding templates was monitored by 500 MHz H NMR
spectroscopy over 16 hours (see Figure S5 for an example spectrum) and these data were
used to construct the concentration-time profiles for each replicator within this mixture.
Pleasingly, there is excellent agreement (Figure 5, center) between the experimental data
(the observed ratio of SNSM : SNLM : LNSM : LNLM is 1.7 : 1.5 : 1.4 : 1) and those
calculated using our kinetic simulation.
Next, we examined the behavior of the network in response to the addition of either
a single template or to two templates added simultaneously. For each scenario, we
calculated the upregulation or down-regulation of the formation of a given template
(Enhancement factor, %EF, Figure 5) relative to a baseline, either experimental or
simulated, where there was no template input (Figure 5, center). In each simulation, the
amount of the instructing template reflected the concentrations of the individual templates
in CDCl that were achievable experimentally.
3
13
ACS Paragon Plus Environment

Journal of the American Chemical Society
Page 14 of 29
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
Figure 5 A reagent pool containing two nitrones (SN and LN) and two maleimides (SM and LM) is
instructed by the addition of one (Single Input) or two (Dual Input) replicating templates. In all cases, the
system responds in specific and predictable ways to the template input and there is good agreement
between the behavior of the system predicted from kinetic simulation and experiment. Experiments: 273 K,
CDCl , 10 mM. For details of simulations, see Methods. Data points and bars are shaded according to their
3
association with a specific template—SNSM (white), LNSM (light gray), SNLM (dark gray), LNLM (black).
The enhancement factor (%EF) is calculated relative to the uninstructed experiment (No input) using the
equation shown for each of the four templates.
14
ACS Paragon Plus Environment

Page 15 of 29
Journal of the American Chemical Society
1
2
3
4
5
6
7
8
9
In the case of single template inputs, either SNSM or SNLM, (Figure 5, top half), it
is evident from the simulations that the instruction of the network by the appropriate
template results in the upregulation (%EF > 0, Figure 5) of not only the added template but
also its linker-length related partner. Thus, when SNSM provides the instruction, LNLM
is also up-regulated, since it is created from the two long components that are not required
for the formation of SNSM. In other words, the addition of a linker-length matched
template (SNSM) results in the upregulation of both templates in this group (SNSM and
LNLM). Similarly, when SNLM provides the instruction, LNSM is up-regulated as it is
created from the short and long components that are not required for the formation of
SNLM. In order to verify these predictions experimentally, we performed two
experiments in which SN, LN, SM, and LM were dissolved at starting concentrations of 10
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
mM in CDCl at 273 K. In one experiment, SNSM was added at a concentration of 3.7 mM
3
and in the second experiment, SNLM was added at a concentration of 3.2 mM. In both
1
cases, the formation of the four templates was monitored by 500 MHz H NMR
spectroscopy over 16 hours and these data were used to construct the concentration-time
profiles for each replicator within this mixture. Subsequently, these profiles were used to
compute the corresponding %EF for each replicator (Figure 5, top half). Pleasingly, there is
a good agreement between the behavior of the network predicted by the simulations and
that observed experimentally. In constructing the kinetic model, certain simplifications
were made with respect to single point recognition events and the rate constants
associated with bimolecular reactions (see Supporting Information for details) and, most
likely, it is these decisions that are responsible for the slight differences between the
experiments and simulations.
In the case of dual template inputs (Figure 5, bottom half), the situation is more
complex. Simulations predict that the simultaneous addition of both linker-matched
templates, SNSM and LNLM, will reinforce the instruction to the network to make only
linker-matched templates. In order to verify this prediction experimentally, we performed
an experiment in which SN, LN, SM, and LM were dissolved at starting concentrations of
10 mM in CDCl at 273 K and templates SNSM and LNLM were added at concentrations
3
of 1.4 mM and 3.4 mM, respectively. Once again, the formation of the four templates was
15
ACS Paragon Plus Environment

Journal of the American Chemical Society
Page 16 of 29
1
2
3
4
5
6
7
8
9
monitored by 500 MHz 1H NMR spectroscopy over 16 hours. The data acquired were used
to construct the concentration-time profiles for each replicator and to compute the
corresponding enhancements (Figure 5, bottom half). Once again, there is a good
agreement between the behavior of the network predicted by the simulations and that
observed experimentally. Finally, we examined the response of the network to two
contradictory instructions, namely SNLM and LNLM. The addition of SNLM should
instruct the network to form linker-mismatched templates, whereas the addition of LNLM
should instruct the network to form linker-matched templates. Simulations predict that
the addition of these two templates should result in the upregulation of SNLM, LNSM,
and LNLM and the down-regulation of SNSM. We verified this prediction experimentally
in the same type of experiment as described previously, in which SNLM and LNLM were
added at concentrations of 4.4 mM and 5.7 mM to a mixture of SN, LN, SM, and LM at
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
starting concentrations of 10 mM in CDCl at 273 K. The data acquired were used to
3
construct the concentration-time profiles for each replicator and to compute the
corresponding %EF (Figure 5, bottom half). As before, there is good agreement between
the behavior of the network predicted by the simulations and that observed
experimentally.
The system-level behavior expressed by this network of synthetic replicators in
response to template inputs is summarized in Figure 6. When a mixture of the building
blocks SN, LN, SM, and LM is instructed by a single template (Figure 6, entries 1 to 4), the
network topology determines the output, i.e., all templates of the same linker type,
matched or mismatched, are up-regulated. For example, addition of SNSM catalyzes the
formation of this template, removing SN and SM rapidly from solution. Since, SNLM and
LNSM both require one of the short components, these two templates are down-regulated
(%EF < 0, Figure 6). Since LNLM requires only the long components LN and LM, it
benefits from the template-directed formation of SNSM and is up-regulated. This
behavior is identical to that exhibited by the simple network shown in Figure 1.
In the case of dual inputs, the situation is more complex, but the behavior of the
system remains predictable. When two templates are added that are either both linker-
matched or both linker-mismatched (for example, Figure 6, addition of SNSM and LNLM
16
ACS Paragon Plus Environment

Page 17 of 29
Journal of the American Chemical Society
1
2
3
4
5
6
7
8
9
and Figure 6, entry 5) the template effects are reinforcing. When two templates are added
that belong to different linker classes (for example, the addition of SNLM and LNLM,
Figure 6, entry 6) the template effects are, in principle, in competition. As a result of the
crosscatalytic relationships between SNLM and LNSM, both of these templates are up-
regulated. The addition of LNLM mitigates the effect of this crosscatalytic connection by
providing a catalytic pathway for the formation of LNLM. Therefore, the only species in
solution whose formation is not enhanced directly in a catalytic manner is SNSM.
Consequently, this template is down-regulated. A similar outcome is observed when
SNSM and SNLM are added simultaneously at the start of the reaction (Figure 6, entry 7
and Figure S7). In this case LNLM is the only species in solution whose formation is not
enhanced directly in a catalytic manner and, consequently, it is this template that is down-
regulated.
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
Figure 6 The output of the replicator network shown in Figure 2 is determined by the nature of the
instructing template that is added. Entries 1 to 4 show single inputs where the network topology determines
that the input of a single template of a particular linker type, matched or mismatched, results in the up-
regulation of both templates of the same linker type. Entries 5 to 7 illustrate more complex cases where more
than one instructing template is added.
17
ACS Paragon Plus Environment

Journal of the American Chemical Society
Page 18 of 29
1
2
3
4
Conclusions
5
6
7
8
Networks of replicating templates that interact and react in well-defined ways can serve as
models for the kind of primitive metabolic pathways that may have arisen12 on the early
Earth. In this work, we have demonstrated that it is possible to create such a network,
whose behavior can be directed specifically by the introduction of instructional templates
that possess a very limited structural and interactional lexicon. By maintaining tight
control over the structural complexity of the network components, it is possible to
quantify the relative contributions of each replicator to the overall operation of the
network. Ultimately, the topology of this network is such that it is capable of expressing
system-level behavior through the introduction of instructional replicating templates and
the addition of a single input instruction elicits upregulation of more than one product in a
programmed manner. The programmability of the network also extends to situations
where multiple instructional templates are added simultaneously. This work creates a
foundation for the construction and exploitation of replicator networks under conditions
where dynamic covalent libraries9e,13 or diffusion phenomena7b are coupled with
interconnected replication processes. These studies are currently in progress in our
laboratory.
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
18
ACS Paragon Plus Environment

Page 19 of 29
Journal of the American Chemical Society
1
2
3
4
Acknowledgements
The financial support for this work was provided by EaStCHEM and the Engineering and
Physical Sciences Research Council (Grant EP/K503162/1).
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
Competing financial interests
The authors declare no competing financial interests.
Supporting Information
The Supporting information is available free of charge on the ACS Publications website:
General experimental procedures; synthetic procedures and compound characterization;
details of kinetic analyses, fitting (including fitted kinetic profiles), and simulations, and
details of DFT calculations.
19
ACS Paragon Plus Environment

Journal of the American Chemical Society
Page 20 of 29
1
2
3
4
Notes and References
5
6
7
8
(1)
(a) Mitchell, M. Complexity: A Guided Tour; Oxford University Press: New York, 2009.
(b) Corbett, P. T.; Sanders, J. K. M.; Otto, S. Angew. Chem. Int. Ed. 2007, 46, 8858. (c)
Boccaletti, S.; Latora, V.; Moreno, Y.; Chavez, M.; Hwang, D. U. Phys. Rep. 2006, 424,
175. (d) Severin, K. Chem. Eur. J. 2004, 10, 2565. (e) Newmann, M. E. J. SIAM Rev.
2003, 45, 167. (f) Barabási, A.-L. Linked: The New Science of Networks; Perseus Book
Group: New York, 2002. (g) Strogatz, S. H. Nature 2001, 410, 268.
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
(2)
(a) Miljanić, O. Š. Chem 2017, 2, 502. (b) Duim, H.; Otto, S. Beilstein J. Org. Chem. 2017,
13, 1189. (c) Ashkenasy, G.; Hermans, T. M.; Otto, S.; Taylor, A. F. Chem. Soc. Rev.
2017, 46, 2543. (d) De la Escosura, A.; Briones, C.; Ruiz-Mirazo, K. J. Theor. Biol. 2015,
381, 11. (e) Mattia, E.; Otto, S. Nat. Nanotechnol. 2015, 10, 111. (f) Ruiz-Mirazo, K.;
Briones, C.; de la Escosura, A. Chem. Rev. 2014, 114, 285. (g) von Kiedrowski, G.;
Otto, S.; Herdewijn, P. J. Syst. Chem. 2010, 1, 1. (h) Stankiewicz, J.; Eckardt, L. H.
Angew. Chem. Int. Ed. 2006, 45, 342.
(3)
(4)
(a) Kosikova, T.; Philp, D. Chem. Soc. Rev. 2017, DOI: 10.1039/c7cs00123a.
(b) Bissette, A. J.; Fletcher, S. P. Angew. Chem. Int. Ed. 2013, 52, 12800. (c) Huck, J.;
Philp, D. In Supramolecular Chemistry: From Molecules to Nanomaterials; John Wiley &
Sons, Ltd.: New York, 2012; pp 1415–1446. (d) Vidonne, A.; Philp, D. Eur. J. Org.
Chem 2009, 2009, 593. (e) von Kiedrowski, G. Bioorganic Chem. Front. 1993, 3, 113.
(a) Plöger, T. A.; von Kiedrowski, G. Org. Biomol. Chem. 2014, 12, 6908. (b) Lincoln, T.
a; Joyce, G. F. Science 2009, 323, 1229. (c) Paul, N.; Joyce, G. F. Proc. Natl. Acad. Sci.
USA 2002, 99, 12733. (d) Sievers, D.; Von Kiedrowski, G. Chem. Eur. J. 1998, 4, 629. (e)
Luther, A.; Brandsch, R.; von Kiedrowski, G. Nature 1998, 396, 245. (f) von
Kiedrowski, G.; Wlotzka, B.; Helbing, J.; Matzen, M.; Jordan, S. Angew. Chem. Int. Ed.
Engl. 1991, 30, 423. (g) von Kiedrowski, G. Angew. Chem. Int. Ed. Engl. 1986, 25, 932.
(a) Sadownik, J. W.; Mattia, E.; Nowak, P.; Otto, S. Nat. Chem. 2016, 8, 264. (b)
Colomb-Delsuc, M.; Mattia, E.; Sadownik, J. W.; Otto, S. Nat. Commun. 2015, 6, 7427.
(c) Carnall, J. M. A.; Waudby, C. A.; Belenguer, A. M.; Stuart, M. C. A.; Peyralans, J.
J.-P.; Otto, S. Science 2010, 327, 1502. (d) Rubinov, B.; Wagner, N.; Matmor, M.;
Regev, O.; Ashkenasy, N.; Ashkenasy, G. ACS Nano 2012, 6, 7893. (e) Rubinov, B.;
Wagner, N.; Rapaport, H.; Ashkenasy, G. Angew. Chem. Int. Ed. 2009, 48, 6683.
(a) Samiappan, M.; Dadon, Z.; Ashkenasy, G. Chem. Commun. 2011, 47, 710. (b) Li, X.;
Chmielewski, J. J. Am. Chem. Soc. 2003, 125, 11820. (c) Issac, R.; Chmielewski, J. J. Am.
Chem. Soc. 2002, 124, 6808. (d) Yao, S.; Ghosh, I.; Zutshi, R.; Chmielewski, J. Angew.
Chem. Int. Ed. Engl. 1998, 37, 478. (e) Yao, S.; Ghosh, I.; Zutshi, R.; Chmielewski, J. J.
Am. Chem. Soc. 1997, 119, 10559. (f) Severin, K.; Lee, D. H.; Kennan, A. J.; Ghadiri, M.
R. Nature 1997, 389, 706. (g) Lee, D. H.; Granja, J. R.; Martinez, J. A.; Severin, K.;
Ghadri, M. R. Nature 1996, 382, 525.
(5)
(6)
(7)
(a) Taylor, J. W.; Eghtesadi, S. A.; Points, L. J.; Liu, T.; Cronin, L. Nat. Commun., 2017,
8, 237. (b) Bottero, I.; Huck, J.; Kosikova, T.; Philp, D. J. Am. Chem. Soc. 2016, 138,
20
ACS Paragon Plus Environment

Page 21 of 29
Journal of the American Chemical Society
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
6723. (c) Dieckmann, A.; Beniken, S.; Lorenz, C. D.; Doltsinis, N. L.; Von Kiedrowski,
G. Chem. Eur. J. 2011, 17, 468. (d) Dieckmann, A.; Beniken, S.; Lorenz, C.; Doltsinis,
N. L.; von Kiedrowski, G. J. Syst. Chem. 2010, 1, 10. (e) Kassianidis, E.; Philp, D.
Angew. Chem. Int. Ed. 2006, 45, 6344. (f) Pearson, R. J.; Kassianidis, E.; Slawin, A. M.
Z.; Philp, D. Chem. Eur. J. 2006, 12, 6829. (g) Kindermann, M.; Stahl, I.; Reimold, M.;
Pankau, W. M.; von Kiedrowski, G. Angew. Chem. Int. Ed. 2005, 44, 6750. (h)Wang,
B.; Sutherland, I. O. Chem. Commun. 1997, 1495. (i) Rotello, V.; Hong, J.-I.; Rebek, J. J.
J. Am. Chem. Soc. 1991, 113, 9422. (j) Tjivikua, T.; Ballester, P.; Rebek, Jr., J. J. Am.
Chem. Soc. 1990, 112, 1249.
(8)
(a) Dadon, Z.; Samiappan, M.; Wagner, N.; Ashkenasy, G. Chem. Commun. 2012, 48,
1419. (b) Ashkenasy, G.; Dadon, Z.; Alesebi, S.; Wagner, N.; Ashkenasy, N. Isr. J.
Chem. 2011, 51, 106. (c) Dadon, Z.; Wagner, N.; Ashkenasy, G. Angew. Chem. Int. Ed.
2008, 47, 6128. (d) Ashkenasy, G.; Jagasia, R.; Yadav, M.; Ghadiri, M. R. Proc. Natl.
Acad. Sci. USA 2004, 101, 10872. (e) Saghatelian, A.; Yokobayashi, Y.; Soltani, K.;
Ghadiri, M. R. Nature 2001, 409, 797. (f) Severin, K.; Lee, D. H.; Martinez, J. A.; Vieth,
M.; Ghadiri, M. R. Angew. Chem. Int. Ed. Engl. 1998, 37, 126. (g) Yao, S.; Ghosh, I.;
Zutshi, R.; Chmielewski, J. Nature 1998, 396, 447. (h) Lee, D. H.; Severin, K.;
Yokobayashi, Y.; Ghadiri, M. R. Nature 1997, 390, 591.
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
(9)
(a) Kosikova, T.; Hassan, N. I.; Cordes, D. B.; Slawin, A. M. Z.; Philp, D. J. Am. Chem.
Soc. 2015, 137, 16074. (b) Allen, V. C.; Robertson, C. C.; Turega, S. M.; Philp, D. Org.
Lett. 2010, 12, 1920. (c) Kassianidis, E.; Pearson, R. J.; Wood, E. A.; Philp, D. Faraday
Discuss. 2010, 145, 235. (d) Xu, S.; Giuseppone, N. J. Am. Chem. Soc. 2008, 130, 1826.
(e) Sadownik, J. W.; Philp, D. Angew. Chem. Int. Ed. 2008, 47, 9965. (f) Kassianidis, E.;
Philp, D. Chem. Commun. 2006, 4072.
(10) The 1,3-dipolar cycloaddition reaction between a nitrone and maleimide can give
rise to two diastereoisomers. In the trans diastereoisomer, the proton originally
located on the nitrone is in a configuration on the isoxazolidine ring that is trans to
the two protons originally present on the maleimide dipolarophile. In the cis
diastereoisomer, this relative stereochemical relationship is cis. In this network, only
the trans diastereoisomer is capable of taking part in template-directed replication
processes, and, thus, the notation of the cycloadduct products capable of replication
will omit the trans descriptor throughout. In all recognition-mediated experiments
reported here, after 16 h, the ratio of [trans]:[cis] ranged from a minimum of 8:1 to >
100:1.
(11) Kosikova, T.; Philp, D. J. Am. Chem. Soc. 2017, 139, 12579.
(12) (a) Muchowska, K. B.; Varma, S. J.; Chevallot-Beroux, E.; Lethuillier-Karl, L.; Li, G.;
Moran, J. Nat. Ecol. Evol. 2017, 1, 1716. (b) Ariga, K. Mater. Chem. Front. 2017, 1, 208.
(c) Shapiro, R. Q. Rev. Biol. 2006, 81, 105.
(13) (a) Kosikova, T.; Mackenzie, H.; Philp, D. Chem. Eur. J. 2016, 22, 1831. (b) Sadownik,
J. W.; Philp, D. Org. Biomol. Chem. 2015, 13, 10392.
21
ACS Paragon Plus Environment

Journal of the American Chemical Society
Page 22 of 29
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
22
ACS Paragon Plus Environment

Page 23 of 29
Journal of the American Chemical Society
1
2
3
4
For Table of Contents Only:
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
23
ACS Paragon Plus Environment

10
(a)
(b)
k
= 1
Journal of the American ChemicPalaSgoec2i4etoyf 29
AB
8
AB
6
1
2
3
4
5
6
kAB
4
A
C
B
D
2
0
kAC
kBD
AC
BD
AB AC BD CD
Product
10
8
7
8
kAB = 2
kCD
9
10
11
6
CD
4
2
12
d[A]
13
= −kAB[A][B] − kAC[A][C]
0
dt
14
AB AC BD CD
Product
15
16
17
18
19
10
8
kAB = 10
d[AB]
= kAB[A][B]
dt
6
4
20
d[B]
21
= −kAB[A][B] − kBD[B][D]
ACS Paragon Plus Environm2ent
dt
22
0
23
24
AB AC BD CD
Product

SNSM
LNLM
Page 25 of 29
Journal of the American Chemical Society
SNSM
LNLM
1
2
SNSM
LNLM
3
4
5
SNSM → SNSM
Autocatalytic
Cycle
LNLM → LNLM
Autocatalytic
Cycle
LNLM
SNSM
6
7
SN
SM
LN
LM
8
9
SM
LN
LM
SN
SNSM
LNLM
10
11
12
13
14
15
16
17
18
SNLM
LNSM
Building
Blocks
SN
SM
LM
LN
SNLM → LNSM
Crosscatalytic
Cycle
LNSM → SNLM
Crosscatalytic
Cycle
LNSM
SNLM
SNLM
LNSM
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
LNSM
SNLM
SM
LN
LM
SN
LNSM
SNLM
LNSM → LNSM
Autocatalytic
Cycle
SNLM → SNLM
Autocatalytic
Cycle
SNLM
LNSM
SNLM
LNSM
SNLM
LNSM
ACS Paragon Plus Environment
LNSM
SNLM

10
8
10
Journal of the American Chemical Society
Page 26 of 29
(a)
LNLM
SNSM
8
6
4
2
0
1
2
3
4
5
6
7
8
1.29 × 10–4 s–1
kbi
1.73 × 10–4 s–1
6.20 × 10–4 M–1 s–1
3.58 M
kbi
6
kauto
kauto
13.9 × 10–4 M–1 s–1
4
EMkinetic
10.8 M
EMkinetic
Duplex
0.371 × 106 M–1
Duplex
0.510 × 106 M–1
Ka
Ka
2
0
0
4
8
12
16
0
4
8
12
16
9
Time / h
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
Time / h
SNLM → LNSM
Crosscatalytic
Cycle
LNSM → SNLM
Crosscatalytic
Cycle
10
8
10
8
SNSM → SNSM
Autocatalytic
Cycle
LNLM → LNLM
Autocatalytic
Cycle
6
6
Building
Blocks
SN
SM
LM
LN
4
4
2
2
LNSM → LNSM
Autocatalytic
Cycle
SNLM → SNLM
Autocatalytic
Cycle
0
0
0
0
4
8
12
16
4
8
12
16
Time / h
Time / h
10
8
10
8
LNSM
SNLM
1.12 × 10–4 s–1
kbi
1.87 × 10–4 s–1
6
6
kbi
5.66 × 10–4 M–1 s–1
—
—
kauto
kauto
4
4
3.03 M
EMkinetic
EMkinetic
1.02 × 103 M–1
Duplex
Duplex
Ka
Ka
5.04 × 103 M–1
2
2
58.2 × 10–4 M–1 s–1
kcross
18.3 × 10–4 M–1 s–1
kcross
0
0
18.7 M
31.1 M
EMkinetic
EMkinetic
0
4
8
12
16
0
4
8
12
16
Time / h
Time / h
37
(b)
38
+++
39
40
41
42
43
44
45
SNSM
++
+
LNLM
LNSM
SNLM
+
ACS Paragon Plus Environment
+++

(a)
(c)
Page 27 of 29
Journal of the American Chemical Society
1
2
3
4
5
6
B
ω
5
4
3
2
1
0
A
LNSM
7
(b)
8
9
10
11
12
13
14
15
16
17
ACS Paragon Plus Environment
SNLM
0
10
Angle ω in degrees
20
30
40

Simulation
Experiment
Journal of the American Chemical Society
Page 28 of 29
180
180
10
1
2
3
4
5
6
7
8
120
120
60
0
6
4
2
0
60
0
0
4
8
Time / h
12 16
8
9
SNSM
−60
−60
180
180
10
11
12
13
14
15
16
17
18
19
20
10
8
120
60
120
60
6
4
2
0
0
0
SNLM
0
4
8
Time / h
12 16
−60
−60
10
10
8
21
228
LN
SN
6
6
No Input
23
244
25
4
LM
SM
2
262
0
0
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
0
4
8
Time / h
12 16
0
4
8
Time / h
12 16
LNLM
180
120
60
180
120
60
10
8
SNSM
6
4
2
0
0
0
0
4
8
Time / h
12 16
−60
−60
SNLM
180
180
10
8
120
60
120
60
6
4
2
ACS Paragon Plus Environment
0
0
0
0
4
8
Time / h
12 16
−60
−60

SNSM
urnPaalgoef t2h9eoAfm29erican Chemical Soc
LNSM
1
2
3
SNLM
4
5
6
7
8
9
LNLM
Input
Output
1
10
2
11
12
3
13
4
14
15
5
16
ACS Paragon Plus Environment
17
6
18
7
19