Protein-protein communication and enzyme activation mediated by a synthetic chemical transducer

Article abstract of DOI:10.1021/jacs.5b01123

The design and function of a synthetic chemical transducer that can generate an unnatural communication channel between two proteins is described. Specifically, we show how this transducer enables platelet-derived growth factor to trigger (in vitro) the catalytic activity of glutathione-s-transferase (GST), which is not its natural enzyme partner. GST activity can be further controlled by adding specific oligonucleotides that switch the enzymatic reaction on and off. We also demonstrate that a molecular machine, which can regulate the function of an enzyme, could be used to change the way a prodrug is activated in a programmable manner.

Full text of DOI:10.1021/jacs.5b01123

Subscriber access provided by NEW YORK UNIV
Communication
Protein-Protein Communication and Enzyme Activation
Mediated by a Synthetic Chemical Transducer
Ronny Peri-Naor, Tal Ilani, Leila Motiei, and David Margulies
J. Am. Chem. Soc., Just Accepted Manuscript • DOI: 10.1021/jacs.5b01123 • Publication Date (Web): 08 May 2015
Downloaded from http://pubs.acs.org on May 26, 2015
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 11
Journal of the American Chemical Society
1
2
3
4
5
6
7
8
9
Protein-Protein Communication and Enzyme Activation Medi-
ated by a Synthetic Chemical Transducer
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
Ronny Peri-Naor,† Tal Ilani,‡ Leila Motiei,† and David Margulies*,†
Departments of †Organic Chemistry and ‡Structural Biology, The Weizmann Institute of Science, Rehovot 76100, Israel.
Supporting Information Placeholder
a
ABSTRACT: The design and function of a synthetic ‘chemical
transducer’ that can generate an unnatural communication channel
between two proteins is described. Specifically, we show how this
transducer enables platelet-derived growth factor (PDGF) to trig-
ger (in vitro) the catalytic activity of glutathione-s-transferase
(GST), which is not its natural enzyme partner. GST activity can
be further controlled by adding specific oligonucleotides (ODNs)
that switch the enzymatic reaction on and off. We also demon-
b
(i)
strate that a molecular machine, which can regulate the function
of an enzyme, could be used to change the way a prodrug is acti-
vated in a ‘programmable’ manner.
(v)
(iii)
(iv)
There is a growing interest in developing synthetic protein binders
based on oligonucleotide (ODN)-small molecule or ODN-peptide
conjugates that, in response to external stimuli, undergo a major
conformational change that enables them to modulate the activity
of their protein targets, akin to allosteric proteins.1 The use of
ODNs for scaffolding such binders not only facilitates projecting
the synthetic conjugates in specific orientations2 – it also enables
one to change the conformation of these constructs by adding
complementary strands. It has been shown that when this structur-
al change affects the affinity of such systems, they can operate as
allosteric switches that reversibly interact with different protein
partners.3
(ii)
(vi)
Figure 1. a) Structure of artificial ‘chemical transducer’ 1 inte-
grating a PDGF aptamer, a bivalent GST inhibitor, a fluorophore
(FAM), and a quencher (dabcyl). b) In the presence of 1 the activ-
ity of GST is inhibited, owing to the binding of the two inhibitor
units at the enzyme active sites (i→ii). The catalytic activity can
then be restored by adding a complementary strand ODN-2
(ii→iii) or PDGF (ii→v) that binds 1 and disrupts its interaction
with GST. The subsequent addition of ODN-3 (iii→ii) or a PDGF
aptamer (v→ii) liberates 1 from the 1-ODN-2 or 1-PDGF com-
plex, allowing it to inhibit the enzyme one more time.
One of the key roles of allosteric proteins in nature is to medi-
ate signal transduction pathways in which the rise and fall of one
protein remotely affects the activity of another protein. Such
communication networks become possible owing to the function
of various allosteric signaling proteins (e.g., adaptors, mediators,
amplifiers, and modulators) that reversibly interact with different
protein partners, and activate or inactivate them.4 It occurred to us
that by endowing ODN-synthetic molecule hybrids with the abil-
ity to bind different proteins, it might be possible to obtain allo-
steric signaling switches that can mediate unnatural signal trans-
duction steps. Herein, we present an artificial chemical transducer
that enables a platelet-derived growth factor (PDGF) to trigger (in
vitro) the catalytic activity of glutathione-s-transferase (GST),
which is not its natural enzyme partner. By adding specific ODNs
to the system, the chemical transducer-enzyme interaction can be
reversibly controlled, which allows one to switch the enzymatic
reaction on and off. We also show that the system can be used to
reconfigure the conditions needed for prodrug activation, in a way
that resembles the activation setting of an electronic logic circuit.
result, can mediate unnatural protein-protein communication, in
which a growth factor (PDGF) activates an unrelated enzyme
(GST). The structure of 1 integrates a DNA aptamer and a bis-
ethacrynic amide (EA) inhibitor, which serve as PDGF and GST
binders, respectively. The flexible scaffold, consisting of a DNA
backbone and elongated linkers, provides the system with alloster-
ic switching capabilities. A bivalent inhibitor was used for target-
ing the GST dimer because bis-EA derivatives have been shown
to be much better inhibitors than monovalent EA compounds.2a
We also incorporated a fluorophore (FAM) and a quencher
(dabcyl) in the vicinity of the 3’ and 5’ termini to allow monitor-
ing the binding of 1 to its targets by fluorescence spectroscopy.
The operation of the system is schematically illustrated in Fig-
ure 1b. In addition to depicting the principle underlying PDGF-
GST communication (state ii→v→i), this scheme also shows how
the function of 1 could be SET and RESET by adding specific
DNA strands. In the absence of an inhibitor, the dimeric enzyme
catalyzes the conversion of the substrate (S) into a chromophoric
product (P) (state i). In the presence of 1, however, the enzymatic
reaction is inhibited owing to the simultaneous binding of the two
EA units of 1 to the two active sites of GST (state ii). The activity
of the enzyme can then be regenerated by adding ODN-2, which
The activation of PDGFR kinase by its PDGF binding partner
is an important signal transduction step that is mediated by a re-
ceptor that connects the enzyme (i.e., kinase) to PDGF.5 Based on
this principle, we designed an artificial ‘chemical transducer’ 1
(Figure 1a) that can interact with both PDGF and GST and, as a
ACS Paragon Plus Environment

Journal of the American Chemical Society
Page 2 of 11
ODN-2 and PDGF, we took advantage of the fact that 1 equili-
1
2
3
4
5
6
7
8
brates between an open and closed form, which leads to a partial
quenching of fluorescence.6 Accordingly, the binding of 1 to
ODN-2 (Figure S3b) and to PDGF (Figure S3c) was followed by
monitoring the change in the emission signal, which results from
changes in the distance between FAM and dabcyl. A dissociation
constant (Kd) of 26 3 nM was determined for the PDGF-1 inter-
action, which is in agreement with previous reports.6 These bind-
ing studies also confirm two important principles in our design.
The enhanced fluorescence generated by the binding of ODN-2
(Figure S3b) reflects an increase in the distance between the two
EA groups, which is needed for obtaining a monovalent/bivalent
inhibitor switch (Figure 1b, states ii ↔ iii). The stronger binding
of 1 to PDGF, compared with its affinity to GST, is another im-
portant property that should enable the growth factor to compete
with the 1-GST interaction (Figure 1b, ii→v).
a
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
ODN-2
PDGF
ODN-2
PDGF
a
PDGF
aptamer
PDGF
aptamer
ODN-3
ODN-3
Figure 2. a) Enzymatic activity of GST (10 nM) before (—) and
after (—) the addition of 1 (500 nM), and after subsequent addi-
tions of increasing concentrations (250, 500, or 1000 nM) of
ODN-2 (---) (left) or PDGF (---) (right). Insets: Calculated (•) vs.
measured (•) V0 values. b) Left: Initial velocity (V0) measured
after sequential additions (II→V) of ODN-2 and ODN-3 to the 1-
GST complex: (I) none, (II) ODN-2 (2 µM), (III) ODN-3 (3 µM ),
(IV) ODN-2 (4.5 µM), and (V) ODN-3 (6 µM). Right: A similar
experiment performed with the addition of PDGF and PDGF ap-
tamer: (I) none, (II) PDGF (750 nM), (III) PDGF aptamer (4 µM),
(IV) PDGF (5 µM), and (V) PDGF aptamer (10 µM).
b
is complementary to the PDGF aptamer and hence, can form with
1 a rigid duplex that projects the two EA groups in opposite direc-
tions (state iii). This conformational change transforms the biva-
lent GST inhibitor into a much weaker monovalent one,3a which
leads to reactivation of the enzyme. ODN-2 was designed to con-
tain two terminal “toehold” sequences that are not complementary
to the PDGF aptamer. Therefore, with the addition of ODN-3,
which is fully complementary to ODN-2 (state iv), 1 can be dis-
placed from the duplex and can inhibit GST one more time.
Figure 3. a) Real-time enhancement of the enzymatic activity
observed in a solution containing GST (10 nM) and 1 (500 nM)
upon the addition of 1 µM ODN-2 (left) or 750 nM PDGF (right)
at t = 3.5 min. b) Left: Decrease in the enzymatic reaction rate
observed upon the addition of ODN-3 (3 µM) to a solution con-
taining GST (10 nM), 1 (500 nM), and ODN-2 (1 µM) at t = 1.5
min. Right: Addition of a PDGF aptamer (5 µM) to GST (10 nM),
1 (500 nM), and PDGF (750 nM) at t = 1.5 min. The black line
corresponds to the reactions observed without addition of inputs.
An even more challenging goal than using a synthetic input
signal (ODN-2) to induce an enzymatic reaction is to activate the
enzyme with another protein (PDGF), which would correspond to
an artificial signal transduction step. Our group has recently
shown that bringing a synthetic agent in the vicinity of a protein is
likely to promote interactions between the synthetic molecule and
the surface of this protein.2a We therefore expected that the strong
binding of PDGF to the aptamer unit of 1 would make the two EA
groups less available for binding (state v), as well as create streric
hindrance that would prevent 1 from inhibiting GST. Note that
PDGF does not necessarily need to interact with the GST-bound 1
(state ii), but rather, it can bind to the excess of free 1 in the solu-
tion, which would shift the equilibrium toward dissociation of the
1-GST complex. This process could also be reversed by adding an
unmodified PDGF aptamer that would displace the PDGF-bound
1 (state vi) and enable it to re-inhibit the enzymatic reaction.
Next, we investigated whether, in the presence of 1, the activity
of GST would be triggered by the synthetic ODN-2 and, most
importantly, by PDGF, which is not its natural binding partner
(Figure 2a). To this end, GST (10 nM) was incubated with 1 (500
nM) and the enzymatic activity was followed in the presence of
increasing concentrations of ODN-2 (Figure 2a, left). As ex-
pected, a dose-dependent response was observed, showing almost
full reactivation of the enzyme with 1 ꢀM of ODN-2. We then
performed a similar experiment with an incremental addition of
PDGF (Figure 2a, right). Remarkably, PDGF (1 ꢀM) successfully
restored the activity of GST, confirming the possibility of induc-
ing communication between two unrelated proteins. These meas-
urements were also used to confirm the operating mechanism of 1
by comparing the observed initial velocities (V0) with the theoret-
ical values calculated using the Michaelis-Menten model accord-
ing to the IC50 of 1 and the Kd of the PDGF-1 and ODN-2-1 inter-
actions (Supporting Information). As shown in Figure 2a, similar
values were obtained for the calculated and observed V0 values,
indicating that the enhanced activity of GST results from the
competitive binding of PDGF or ODN-2. The higher V0 value
observed upon the addition of 250 nM PDGF most likely results
We first confirmed that 1 individually binds each of its targets:
GST, PDGF, and ODN-2 (Supporting Information). An enzymatic
assay, which follows the conjugation of glutathione to 1-chloro-
2,4-dinitrobenzene (CDNB), was used to monitor the activity of
GST (Figures S1a) and to verify that 1 inhibits GST with an IC50
of 143.8 1 nM (Figure S3a). To follow the binding of 1 to
2
ACS Paragon Plus Environment

Page 3 of 11
Journal of the American Chemical Society
from the formation of the 12:PDGF complex (Figure 1b) when
there is an excess of 1 in the medium, which leads to a more sig-
nificant reduction in the concentration of the free inhibitor.
activity-based truth table corresponds to the logic of a digital cir-
cuit (Figure 4b, left). This Boolean logic representation8 shows
that only when GST was combined with PDGF or ODN-2, or with
both, a significant amount of drug was released, which indicates
that the metabolism of prodrugs could be altered either by ‘exter-
nal’ stimuli (e.g., ODN-2) or by specific protein biomarkers such
as PDGF,5 which is known to be secreted in high concentrations
by several cancer cells. The fact that some JS-K cleavage was also
observed in the presence of GST alone and that micromolar con-
centrations of inputs were used indicates that more potent ‘trans-
ducers’ should be generated before considering such systems for
therapeutic applications. What distinguishes this system from
related logic-based therapeutics that respond to several input sig-
nals9a,10 is that here, the medication can be used as is. Namely, the
drug does not have to be chemically modified10a-c or be loaded on
an auxiliary molecular computational device.10d-l Instead, the
‘chemical transducer’ ‘reprograms’ the natural regulation mecha-
nism of the activating enzyme (i.e., GST), which changes the
conditions needed for prodrug activation.
1
2
3
4
5
6
7
8
The reversibility of our system was also demonstrated by mon-
itoring the response of the 1-GST complex to the sequential addi-
tion of ODN-2 and ODN-3 (Figure 2b, left) or PDGF and its ap-
tamer (Figure 2b, right), which resulted in inhibition/activation
cycles. The gradual loss of on/off signals observed in the PDGF/
aptamer cycle results from the similar binding affinities of 1 and
the unmodified aptamer, which slow down the displacement pro-
cess. These changes in the reaction rate were also monitored in
real time by adding each input while measuring the enzymatic
activity (Figure 3). As shown in Figure 3a, an immediate en-
hancement of the reaction kinetics was observed when ODN-2
(left) or PDGF (right) was added to a solution containing the 1-
GST complex 3.5 minutes after adding the substrates. Similarly, a
rapid decrease in the reaction rate was observed (Figure 3b) upon
the addition of ODN-3 (left) or the PDGF aptamer (right), which
reversed the previous effect. Taken together, these experiments
(Figures 2 and 3) demonstrate that, in addition to inducing PDGF-
GST communication, 1 can operate as a molecular machine that
can be carefully controlled, namely, it is reversible and can rapid-
ly adapt to changes in the environment by responding to different
input signals in real time.
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
a
b
0, 1
a
b
1
0
Figure 5. a) Grey bars: viability of cancer cells incubated with
PBS buffer (i), JS-K (ii) or JS-K and GST (iii). Black bars: cells
incubated with 1 and JS-K and different combinations of PDGF
and ODN-2. b) Fluorescent images of representative cell samples
(000, vs. 110). The scale bar represents 20 µm.
Figure 4. a) JS-K activation reaction. b) A logic circuit (left) in
which the output signal corresponds to the release of NO (right) in
a solution containing JS-K (45 µM), GSH (750 µM), and 1 (750
nM), upon the addition of different combinations of inputs: GST
(10 nM), PDGF (2 µM), and ODN-2 (2 µM). *p < 0.001.
The ability to alter the kinetics of JS-K cleavage according to an
additional protein present in the solution (i.e., PDGF, Figure 4b)
is of particular importance, because activation of prodrugs by
specific enzymes is a common and effective tool for achieving
selective drug release. The activating enzymes could be recombi-
nant enzymes linked to antibodies that direct them to the outer
membrane of specific cells.11 Alternatively, they may be natural
enzymes that are overexpressed in cancer, which leads to high
enzyme concentrations within the cell and/or at the extracellular
space (ECS). Elevated levels of GST, in particular, have been
detected both within cancer cells, as well as in extracellular flu-
ids.12 To demonstrate how ‘chemical transducers’ could be used
to control the effect of prodrugs on cells according to the presence
of a specific protein in their immediate environment, breast cancer
cells (MDA-MB-231) that stably express a fluorescent Cherry-
Red protein were treated with the same concentrations of prodrug
(10 µM), GSH (200 ꢀM), and ‘chemical transducer’ (750 nM),
but with a different combination of GST (10 nM), PDGF (2 µM)
and ODN-2 (2 µM) (Figure 5). This model system was mainly
intended to demonstrate how communication between a growth
factor (PDGF) and an enzyme (GST) at the ECS can induce JS-K
cleavage outside the cell, which would prevent intracellular pro-
drug activation by cytosolic GST and consequently cell death.
After 3 hours of incubation, live cells were counted by a hemocy-
tometer (Figure 5a), which showed a decrease in cell viability in
the absence of inputs (000) (59 2 %) or when only PDGF (010)
The activation and deactivation of enzymes play an important
role in controlling the responses of cells to various environmental
signals.4 In the following proof-of-principle experiments (Figures
4 and 5), we demonstrate how ‘chemical transducers’ such as 1,
which can alter the natural regulation mechanisms of enzymes,
could be used to control the way environmental changes affect
cells. We tested the effect of 1 on the activation of JS-K, an anti-
cancer prodrug whose intracellular cleavage by GST induces the
release of toxic nitric oxide (NO) (Figure 4a).7 Initially, different
combinations of GST (10 nM), PDGF (2 µM), and ODN-2 (2
µM) were added to a PBS buffer solution containing 1 (750 nM),
JS-K (45 µM), and GSH (750 µM), and the release of the drug
(NO) was measured using two colorimetric assays that either
monitor JS-K metabolism or directly follow the production of NO
(Supporting Information). To elucidate how 1 affected the activa-
tion of JS-K, we applied principles of molecular logic,8 which
have been effectively used to describe the function of various
multi-stimuli responsive therapeutics.9 Accordingly, GST, PDGF,
and ODN-2 were denoted as digital inputs (0 or 1) and NO as a
digital output (0 or 1), which depends on the concentration of NO
with respect to the threshold line8 (Figure 4b, right). The resulting
3
ACS Paragon Plus Environment

Journal of the American Chemical Society
Page 4 of 11
2. (a) Motiei, L.; Pode, Z.; Koganitsky, A.; Margulies, D., Angew. Chem
(70% 5 %), ODN-2 (001) (74 12 %) or GST (100) (76 4 %)
was present in the medium. In contrast, in the presence of GST
and PDGF (110) or GST and ODN-2 (101) the viability remained
intact (113 25 % or 107 11 %, respectively). Namely, it is
similar to that of control cells, which were not treated with the
prodrug (Figure 5a, i).
Int. Ed. 2014, 53, 9289; (b) Rosenzweig, B. A.; Ross, N. T.; Tagore, D.
M.; Jayawickramarajah, J.; Saraogi, I.; Hamilton, A. D., J. Am. Chem.
Soc. 2009, 131, 5020; (c) Tagore, D. M.; Sprinz, K. I.; Fletcher, S.;
Jayawickramarajah, J.; Hamilton, A. D., Angew. Chem Int. Edit. 2007, 46,
223; (d) Abendroth, F.; Bujotzek, A.; Shan, M.; Haag, R.; Weber, M.;
Seitz, O., Angew. Chem Int. Ed. 2011, 50, 8592; (e) Eberhard, H.;
Diezmann, F.; Seitz, O., Angew. Chem Int. Ed. 2011, 50, 4146; (f)
Melkko, S.; Zhang, Y.; Dumelin, C. E.; Scheuermann, J.; Neri, D., Angew.
Chem Int. Ed. 2007, 46, 4671; (g) Kleiner, R. E.; Dumelin, C. E.; Tiu, G.
C.; Sakurai, K.; Liu, D. R., J. Am. Chem. Soc. 2010, 132, 11779; (h)
Williams, B. A. R.; Diehnelt, C. W.; Belcher, P.; Greving, M.; Woodbury,
N. W.; Johnston, S. A.; Chaput, J. C., J. Am. Chem. Soc. 2009, 131,
17233; (i) Liu, R.; Jiang, B.; Yu, H.; Chaput, J. C., ChemBioChem 2011,
12, 1813; (j) Matsuura, K.; Hibino, M.; Yamada, Y.; Kobayashi, K., J.
Am. Chem. Soc. 2000, 123, 357.
3. (a) Harris, D. C.; Chu, X.; Jayawickramarajah, J., J. Am. Chem. Soc.
2008, 130, 14950; (b) Harris, D. C.; Saks, B. R.; Jayawickramarajah, J., J.
Am. Chem. Soc. 2011, 133, 7676; (c) Röglin, L.; Ahmadian, M. R.; Seitz,
O., Angew. Chem Int. Ed. 2007, 46, 2704; (d) Röglin, L.; Altenbrunn, F.;
Seitz, O., ChemBioChem 2009, 10, 758; (e) Saghatelian, A.; Guckian, K.
M.; Thayer, D. A.; Ghadiri, M. R., J. Am. Chem. Soc. 2002, 125, 344; (f)
Picuri, J. M.; Frezza, B. M.; Ghadiri, M. R., J. Am. Chem. Soc. 2009, 131,
9368; (g) Kim, Y.; Cao, Z.; Tan, W., Proc. Nati. Acad. Sci. U.S.A 2008,
105, 5664; (h) Han, D.; Zhu, Z.; Wu, C.; Peng, L.; Zhou, L.; Gulbakan, B.;
Zhu, G.; Williams, K. R.; Tan, W., J. Am. Chem. Soc. 2012, 134, 20797;
(i) Zhou, C.; Yang, Z.; Liu, D., J. Am. Chem. Soc. 2012, 134, 1416; (j)
Dittmer, W. U.; Reuter, A.; Simmel, F. C., Angew. Chem Int. Ed. 2004,
43, 3550; (k) Portela, C.; Albericio, F.; Eritja, R.; Castedo, L.;
Mascareñas, J. L., ChemBioChem 2007, 8, 1110.
4. Dennis, E. A.; Bradshaw, R. A., Transduction Mechanisms in Cellular
Signaling. Academic Press: 2011.
5. Yu, J.; Ustach, C.; Kim, H. R., J. Biochem. Mol. Biol. 2003, 36, 49.
6. Fang, X.; Sen, A.; Vicens, M.; Tan, W., ChemBioChem 2003, 4, 829.
7. McMurtry, V.; Saavedra, J. E.; Nieves-Alicea, R.; Simeone, A.; Keefer,
L. K.; Tari, A. M., Int. J. Oncol. 2011, 38 963.
8. For recent reviews see: (a) de Silva, A. P.; Uchiyama, S., Nat.
Nanotechnol. 2007, 2, 399; (b) Andreasson, J.; Pischel, U., Chem. Soc.
Rev. 2015, 44, 1053; (c) Szaciłowski, K., Inforchemistry. Wiley:
Chichester, 2013; (d) Katz, E.; Privman, V., Chem. Soc. Rev. 2010, 39,
1835; (e) Stojanovic, M. N.; Stefanovic, D.; Rudchenko, S., Acc. Chem.
Res. 2014, 47, 1845; (f) Benenson, Y., Mol. BioSyst. 2009, 5, 675; (g) de
Ruiter, G.; van der Boom, M. E., Acc. Chem. Res. 2011, 44, 563.
9. For an in-depth review on molecular computation-based therapy, see:(a)
De-Silva, A. P. In Molecular Logic-based Computation, RSC: Cambridge,
UK, 2013; p 364; For a review that focuses on nucleic acid-based compu-
ting, see (b) Wu, C.; Wan, S.; Hou, W.; Zhang, L.; Xu, J.; Cui, C.; Wang,
Y.; Hu, J.; Tan, W., Chem. Commun. 2015, 51, 3723. For a review that
focuses on enzyme-assisted logic gate therapy, see (c) Wang, J.; Katz, E.,
Anal. Bioanal. Chem. 2010, 398, 1591.
1
2
3
4
5
6
7
8
These differences in cell viability can also be visualized using a
fluorescent microscope. As shown in Figures 5b and S5, cell
death (000 vs. 110) leads to changes in the morphology of the
cells, transforming them into smaller spherical shapes, as well as
to a reduction in the number of imaged cells owing to their de-
tachment from the surface. Thus, under these conditions the acti-
vation of GST by PDGF or ODN-2 induces an extracellular deg-
radation of JS-K, which prevents the release of NO inside the
cells. Thus, this model system indicates the feasibility of control-
ling the way prodrugs affect cells through an artificial regulatory
system that makes the activating enzyme responsive to the pres-
ence of specific proteins or synthetic stimuli in its surroundings.
This could be used, for example, to protect or damage specific
cells (Figure 5) upon treatment with broad-spectrum medications.
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
To summarize, the main concept highlighted in this work is the
ability to design synthetic agents that mimic the function of sig-
naling proteins and therefore, can generate de novo communica-
tion channels between proteins. Whereas in nature PDGF acti-
vates its PDGFR enzyme partner, we have shown that in the pres-
ence of a synthetic ‘chemical transducer” the same growth factor
can trigger the enzymatic activity of an unrelated enzyme (i.e.,
GST). Another important property of the system is the ability to
regulate it in real time by using specific ODN inputs. The strength
of a molecular machine, which can change the way an enzyme is
regulated, was further demonstrated by using it to induce differen-
tial cell death by ‘reprograming’ the conditions needed for pro-
drug activation. Although this ‘transducer’ prototype does not
fully inhibit the enzyme and is currently limited to controlling
prodrug activation outside the cell, it demonstrates a general ap-
proach that could potentially be applied to activate other classes
of prodrugs, as well as to generate more effective, cell-permeable
transducers that regulate the function of enzymes by mediating
intracellular protein-protein communication. Given that many of
the cell’s functions are mediated by signaling proteins that contin-
uously activate and deactivate enzymes, we believe that mimick-
ing the function of these proteins might open up new possibilities
for controlling biological processes.
ASSOCIATED CONTENT
Supporting Information
10. For small molecule-based logic gate therapy, see: (a) Erbas-Cakmak,
S.; Bozdemir, O. A.; Cakmak, Y.; Akkaya, E. U., Chem. Sci. 2013, 4, 858;
(b) Ozlem, S.; Akkaya, E. U., J. Am. Chem. Soc. 2008, 131, 48; (c) Amir,
R. J.; Popkov, M.; Lerner, R. A.; Barbas, C. F.; Shabat, D., Angew. Chem
Int. Ed. 2005, 44, 4378; For material-based logic gate therapy see: (d)
Zhou, M.; Zhou, N.; Kuralay, F.; Windmiller, J. R.; Parkhomovsky, S.;
Valdés-Ramírez, G.; Katz, E.; Wang, J., Angew. Chem Int. Ed. 2012, 51,
2686; (e) Bocharova, V.; Zavalov, O.; MacVittie, K.; Arugula, M. A.;
Guz, N. V.; Dokukin, M. E.; Halámek, J.; Sokolov, I.; Privman, V.; Katz,
E., J. Mater. Chem. 2012, 22, 19709; For DNA computing-based therapy,
see: (f) Douglas, S. M.; Bachelet, I.; Church, G. M., Science 2012, 335,
831; (g) You, M.; Peng, L.; Shao, N.; Zhang, L.; Qiu, L.; Cui, C.; Tan, W.,
J. Am. Chem. Soc. 2013, 136, 1256; (h) You, M.; Zhu, G.; Chen, T.;
Donovan, M.; Tan, W., J. Am. Chem. Soc. 2015, 137 667; (i) Elbaz, J.;
Lioubashevski, O.; Wang, F.; Remacle, F.; Levine, R. D.; Willner, I., Nat.
Nanotechnol. 2010, 5, 417; (j) Xie, Z.; Wroblewska, L.; Prochazka, L.;
Weiss, R.; Benenson, Y., Science 2011, 333, 1307; (k) Benenson, Y.; Gil,
B.; Ben-Dor, U.; Adar, R.; Shapiro, E., Nature 2004, 429, 423; (l)
Kolpashchikov, D. M.; Stojanovic, M. N., J. Am. Chem. Soc. 2005, 127,
11348.
Experimental procedures and serum stability test. This material is
available free of charge via the Internet at http://pubs.acs.org.
AUTHOR INFORMATION
Corresponding Author
david.margulies@weizmann.ac.il
Notes
The authors declare no competing financial interests.
ACKNOWLEDGMENT
This research was supported by the Minerva Foundation, the
HFSP Organization, and an European Research Council Grant.
REFERENCES
1. For recent reviews see: (a) Battle, C.; Chu, X.; Jayawickramarajah, J.,
Supramol. Chem. 2013, 25, 848; (b) Diezmann, F.; Seitz, O., Chem. Soc.
Rev. 2011, 40, 5789.
11. Bagshawe, K. D., Expert. Rev. Anticancer. Ther. 2006, 6, 1421.
12. Tsuchida, S.; Sekine, Y.; Shineha, R.; Nishihira, T.; Sato, K., Cancer.
Res. 1989 49, 5225.
4
ACS Paragon Plus Environment

Page 5 of 11
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
5
ACS Paragon Plus Environment

Journal of the American Chemical Society
Page 6 of 11
1
2
3
4
Figure for the table of content:
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
ACS Paragon Plus Environment

Page 7 of 11
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
(i)
(iii)
(iv)
(v)
(ii)
(vi)
Figure 1. a) Structure of artificial ‘chemical transducer’ 1 integrating a PDGF aptamer, a bivalent GST inhibitor, a
fluorophore (FAM), and a quencher (dabcyl). b) In the presence of 1 the activity of GST is inhibited, owing to the
binding of the two inhibitor units at the enzyme active sites (iii). The catalytic activity can then be restored by
adding a complementary strand ODN-2 (iiiii) or PDGF (iiv) that binds 1 and disrupts its interaction with GST.
The subsequent addition of ODN-3 (iiiii) or a PDGF aptamer (vii) liberates 1 from the 1-ODN-2 or 1-PDGF
complex, allowing it to inhibit the enzyme one more time.
ACS Paragon Plus Environment

Journal of the American Chemical Society
Page 8 of 11
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 2. a) Enzymatic activity of GST (10 nM) before (—) and after (—) the addition of 1 (500 nM), and after
subsequent additions of increasing concentrations (250, 500, or 1000 nM) of ODN-2 (---) (left) or PDGF (---)
(right). Insets: Calculated () vs. measured () V0 values. b) Left: Initial velocity (V0) measured after sequential
additions (IIV) of ODN-2 and ODN-3 to the 1-GST complex: (I) none, (II) ODN-2 (2 µM), (III) ODN-3 (3 µM
), (IV) ODN-2 (4.5 µM), and (V) ODN-3 (6 µM). Right: A similar experiment performed with the addition of
PDGF and PDGF aptamer: (I) none, (II) PDGF (750 nM), (III) PDGF aptamer (4 µM), (IV) PDGF (5 µM), and
(V) PDGF aptamer (10 µM).
ACS Paragon Plus Environment

Page 9 of 11
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) Real-time enhancement of the enzymatic activity observed in a solution containing GST (10 nM) and
1 (500 nM) upon the addition of 1 µM ODN-2 (left) or 750 nM PDGF (right) at t = 3.5 min. b) Left: Decrease in
the enzymatic reaction rate observed upon the addition of ODN-3 (3 µM) to a solution containing GST (10 nM), 1
(500 nM), and ODN-2 (1 µM) at t = 1.5 min. Right: Addition of a PDGF aptamer (5 µM) to a solution of GST (10
nM), 1 (500 nM), and PDGF (750 nM) at t = 1.5 min. The black line corresponds to the reactions observed without
the addition of inputs.
ACS Paragon Plus Environment

Journal of the American Chemical Society
Page 10 of 11
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 4. a) JS-K activation reaction. b) A logic circuit (left) in which the output signal corresponds to the release
of NO (right) in a solution containing JS-K (45 µM), GSH (750 µM), and 1 (750 nM), upon the addition of
different combinations of inputs: GST (10 nM), PDGF (2 µM), and ODN-2 (2 µM). *p < 0.001.
ACS Paragon Plus Environment

Page 11 of 11
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 5. a) Grey bars: viability of cancer cells incubated with PBS buffer (i), JS-K (ii) or JS-K and GST (iii).
Black bars: cells incubated with 1 and JS-K and different combinations of PDGF and ODN-2. b) Fluorescent
images of representative cell samples (000, vs. 110). The scale bar represents 20 µm.
ACS Paragon Plus Environment