Design of small molecule reagents that enable signal amplification via an autocatalytic, base-mediated cascade elimination reaction

Article abstract of DOI:10.1039/c2cc17566e

This Communication describes three small molecule reagents that amplify the signal for a detection event via an autocatalytic reaction. Two signals are obtained from each reagent: (i) the dibenzofulvene chromophore and (ii) piperidine, which can be visualized using a pH indicator dye. The reagents are demonstrated in a model assay for palladium.

Full text of DOI:10.1039/c2cc17566e

View Article Online / Journal Homepage / Table of Contents for this issue
Dynamic Article Links
ChemComm
Cite this: Chem. Commun., 2012, 48, 3018–3020
www.rsc.org/chemcomm
COMMUNICATION
Design of small molecule reagents that enable signal amplification via an
autocatalytic, base-mediated cascade elimination reactionw
Hemakesh Mohapatra, Kyle M. Schmid and Scott T. Phillips*
Received 3rd December 2011, Accepted 26th January 2012
DOI: 10.1039/c2cc17566e
This Communication describes three small molecule reagents
that amplify the signal for a detection event via an autocatalytic
reaction. Two signals are obtained from each reagent: (i) the
dibenzofulvene chromophore and (ii) piperidine, which can be
visualized using a pH indicator dye. The reagents are demon-
strated in a model assay for palladium.
Herein, we demonstrate that the reagents shown in Fig. 1
amplify signal through an autocatalytic reaction (we describe
kinetics that are characteristic of an autocatalytic reaction, as
well as demonstrate that the quantity of piperidine is amplified
during the reaction), and we show that reagent 1 can be used
in combination with a second reagent (a detection reagent,9
e.g., reagent 4, Fig. 2) to provide sensitive detection of a model
analyte.3 The goals of this Communication are to describe the
design of these types of signal amplification reagents, to
demonstrate that they can be paired with a detection reagent
in a model assay, and to establish the physical-organic foundation
that will be needed to modify and optimize the reagents for future
use in specific analytical assays.
The developing world poses many challenges for diagnostic
tests and the corresponding reagents.1 Reagents used in these
environments must be stable for prolonged periods at the
elevated temperatures encountered in the developing world
(refrigeration is often lacking)2 and must be selective and
sensitive enough to enable trace-level detection of biomarkers
of disease. Herein, we describe the design of a new small
molecule reagent that provides a starting point for addressing
these issues. Our reagent is stable, is easily modified, and is
capable of amplifying signal through an autocatalytic reaction
(i.e., one of the products of the signal amplification reaction
catalyzes its own formation).
The design of amplification reagents 1, 2, and 3 involved
several considerations that are worth noting. First, an auto-
catalytic amplification reaction requires that the signal propagating
the reaction (e.g., piperidine) must not be consumed, and therefore,
we chose to employ acid–base chemistry that enables rapid
recycling of the propagating signal. Second, we reasoned that
the most general signal amplification reagent would provide
more than one type of amplified signal to accommodate
multiple strategies for reading the output of the assay. In the
case of reagents 1, 2, and 3, we included two output signals:
the dibenzofulvene chromophore (which is easily quantified
using UV/vis spectroscopy) and base (which can be quantified with
a pH indicator; this readout will be particularly convenient for
assays in resource-limited environments). Third, to enable flexibility
This reagent is similar conceptually to an autoinductive
signal amplification reagent that we reported recently,3 but
also differs in the following important ways: (i) the signal that
initiates and propagates the amplification reaction, (ii) the signal
that is produced upon amplification, and (iii) the potential for
rapid signal amplification. Small molecule reagents related to our
autoinductive reagent also have been reported,4–7 but none
amplifies signal autocatalytically, which is, theoretically, a
more efficient amplification process than autoinductive signal
amplification. Silver reduction is one of few diagnostically
relevant autocatalytic signal amplification processes, but it
uses thermally-unstable silver(I) salts, and therefore, is of
limited utility in remote environments that lack refrigeration.8
The wide use of silver reduction in laboratory diagnostics,
however, highlights the power of an autocatalytic signal
amplification reagent, and points to the need for a similar
reagent for use in resource-limited environments.
The Pennsylvania State University, 104 Chemistry Bldg.,
University Park, PA 16803, USA. E-mail: sphillips@psu.edu;
Fax: 814 865 5235; Tel: 814 867 2502
w Electronic supplementary information (ESI) available: synthesis and
characterization of reagents 1, 2, and 3, supporting figures, and tables
of data. See DOI: 10.1039/c2cc17566e
Fig. 1 Balanced autocatalytic reaction sequence for reagents 1–3
when exposed to a trace quantity of piperidine (green). Reagent 3
releases a second equivalent of piperidine (blue) through additional
reactions that are not shown.
c
3018 Chem. Commun., 2012, 48, 3018–3020
This journal is The Royal Society of Chemistry 2012

View Article Online
consistent with short oligomers (1–4 repeating units; Fig. S2, ESIw),
which likely arise from oligomerization of azaquinone methide
under the reaction conditions. The LC-MS data accounts for
all components of reagent 1, and demonstrates that 1 is
capable of releasing piperidine when the Fmoc group is
cleaved. In contrast, when stored as a solid at 20 1C open to
the air and in the absence of piperidine, reagent 1 showed no
signs of decomposition for at least 14 days (14 days was the
duration of the stability experiment).
The ability of reagent 1 to provide an amplified signal
becomes apparent when 1 is exposed to various substoichio-
metric quantities of piperidine (Fig. 3). The graph in Fig. 3
reveals four key features of the reagent: (i) it is consumed
completely, even when exposed to only 0.001 equivalents of
piperidine; (ii) the background reaction in the absence of applied
piperidine is negligible over the course of the experiment; (iii) it is
capable of 41000Â signal amplification; and (iv) the kinetics
profile of its reaction with piperidine is characteristic of an
autocatalytic reaction (i.e., it displays a sigmoidal response
curve;11 in contrast, if each equivalent of piperidine were
consumed when reacting with 1, then the reaction would
display a logarithmic response curve).
Fig. 2 Signal amplification reagents 1, 2, and 3 are designed to
amplify signal after 4 releases piperidine in response to a specific
activity-based detection event.3,9,10 In the tandem reaction depicted
above, the selectivity for the assay is provided by the activity-based
detection reagent, and the sensitivity is provided by reagent 1, 2, or 3
(these reagents amplify both the dibenzofulvene chromophore and
piperidine). The notation ‘‘n’’ refers to the number of cycles of the
autocatalytic reaction.
in the degree of amplification, we connected the Fmoc group
with piperidine through an aniline linker that accommodates
the release of one (reagent 1), two (reagent 3), or even three
(compound not shown) equivalents of piperidine per reaction
with piperidine. Fourth, the rate of amplification must be made
as fast as possible (without causing background reactions);
therefore, we included a methyl ether at the 3-position of the
aniline linker (reagent 1) to enhance the rate of release of
piperidine (vide infra). Finally, the design of the reagent must
be coupled with a strategic synthetic plan since reagents that
disassemble autocatalytically are especially sensitive to trace
levels of decomposition (i.e., release of piperidine) during the
preparation of the reagent. An efficient synthetic route to
reagent 1 is shown in Scheme 1; the routes to reagents 2 and
3 are shown in Schemes S1 and S2 (ESIw).
To determine whether our reagent design was effective, we
first examined the products that are formed when 1 is exposed
to piperidine. Time-dependent LC-MS analysis (Fig. S1, ESIw)
of the reaction mixture obtained when 1 reacts with 0.02 equiv
of piperidine in 50 : 1 DMSO–H2O at 18 1C revealed clean,
quantitative conversion of 1 to dibenzofulvene within 6 h;
the LC-MS data also revealed an increase in the intensity of
the mass signal for piperidine as the reaction progressed. The
aniline intermediate resulting from deprotection of the Fmoc
group was too short-lived to be observed using LC-MS; we
presume that release of CO2 and piperidine occurs quickly to
form azaquinone methide. Although azaquinone methide was
not present in the LC-MS spectra, we did observe masses
The design of reagent 1 includes a methyl ether at the
3-position on the aniline linker; we anticipated that the
presence of this substituent would accelerate the rate of
azaquinone methide formation and, therefore, increase the
rate of the amplification reaction.12 As predicted, when 1 is
exposed to 0.01 equiv of piperidine, 1 is consumed 2.3Â faster
than 2 (which lacks the methyl ether) under identical reaction
conditions (Fig. S3, ESIw).
A second method for increasing the rate of the amplification
reaction is to release more than one copy of piperidine from
the aniline linker. Reagent 3 releases two equivalents of
piperidine per reaction with piperidine, and, when exposed
to 0.01 equiv of piperidine (Fig. S3, ESIw), 3 is consumed in
16 h. In comparison, 18 h are necessary for complete
Fig. 3 Quantification of dibenzofulvene when 1 is exposed to substoichio-
metric quantities of piperidine. The graph provides the normalized absor-
bance of dibenzofulvene at 305 nm during the course of the signal
amplification reaction, and as a function of the number of equivalents
of piperidine that were exposed to 1. The experiments were conducted
in triplicate, and the error bars reflect the standard deviations from the
average values. The concentration of 1 was 40 mM in 45 : 5 : 1
DMSO–THF–H2O, and the experiments were conducted at 18 1C.
Scheme 1 Synthesis of signal amplification reagent 1.
c
This journal is The Royal Society of Chemistry 2012
Chem. Commun., 2012, 48, 3018–3020 3019

View Article Online
consumption of reagent 2 (which releases only one equivalent
of piperidine) under the same reaction conditions. Further-
more, at the end of the amplification reaction 3 will have
released twice the quantity of base as 2. Ultimately, the ability
to easily tune the rate of the amplification reaction by modifying
the structure of the reagent is a useful characteristic of the general
reagent design shown in Fig. 1.
detection for this proof-of-concept assay is 12 ppm Pd, which
is approximately the government-regulated threshold level of
palladium permitted in drugs;10,14 (ii) the assay is capable of
distinguishing samples with Pd concentrations that differ by
only 2 ppm; and (iii) amplification reagent 1 can be paired
effectively with an activity-based detection reagent (e.g., 4)
that releases piperidine in response to a specific analyte.
In conclusion, we demonstrated that reagents 1, 2, and 3 are
capable of amplifying a signal via a base-catalyzed autocatalytic
reaction. The pairing of reagents 1, 2, and 3 with different
activity-based detection reagents opens the possibility of
conducting future assays that are both selective and sensitive
for a variety of analytes. The amplification reagents do not yet
provide signal amplification with a rate that is ideal for point-
of-care assays, but they do serve as a valuable starting point
for further optimization. Alternatively, this type of auto-
catalytic amplification reagent may be useful more immediately
in other contexts, including stimuli-responsive materials,15,16
in which an amplified response to a specific chemical signal
is needed, but where the time required to generate an amplified
response is less important. Experiments in this direction are
in progress.
Additional evidence that reagent 1 (and, by analogy, reagents
2 and 3) amplifies signal autocatalytically is provided by the
observation that the quantity of piperidine (the free base) is
amplified during the course of the reaction. To measure the
quantity of piperidine free base, we exposed aliquots of the
reaction mixture to a solution of the pH indicator bromocresol
green and measured the change in color of the solution. In the
absence of base (t = 0 h), the solution of bromocresol green has
negligible absorbance at 625 nm, but as the quantity of base
increases due to the autocatalytic reaction, the absorbance of
bromocresol green at 625 nm increases sigmoidally (Fig. S4,
ESIw). In fact, the quantity of piperidine (as reflected by the
normalized absorbance of bromocresol green at 625 nm)
increases with a rate and absorbance–time profile that is
superimposable on the rate and absorbance–time profile for
the formation of dibenzofulvene under the same reaction
conditions (Fig. S4, ESIw), thus demonstrating that both
signals (base and dibenzofulvene) are generated with equal
rates. Control experiments using either 3-methoxyaniline or
4-aminobenzyl alcohol (40 mM) in place of 1 reveal that neither
aniline compound is sufficiently basic to deprotonate bromocresol
green and cause the absorbance at 625 nm (Fig. S5, ESIw).
Exposure of bromocresol green to 40 mM piperidine in 50 : 1
DMSO–H2O, in contrast, induces a large absorbance at 625 nm
(Fig. S5, ESIw).
Having established that reagents 1, 2, and 3 amplify signal
via autocatalytic reactions, we next evaluated whether they
could be used in tandem with an activity-based detection
reagent to provide sensitive detection of an analyte (see Fig. 2
for a depiction of this concept).3,7 We tested this idea using
Pd(II) as a model analyte,13 reagent 1 for signal amplification,
and reagent 4 as the activity-based detection reagent (this
compound incorporates an allyl group as the substrate for
detecting palladium). The model assay was conducted as
follows: a 100 mL solution of reagent 4 (0.2 M) and PhSiH3
(438 mM) in THF was mixed in a 1 : 1 ratio with a solution of
Pd(OAc)2 (the model analyte) and Bu3P (40.5 mM) in THF.
After a 1-h incubation period (in which piperidine was released
from reagent 4 via a catalytic reaction with Pd), the detection
solution was diluted with water (40 mL), and a 50-mL aliquot
was transferred to a solution of the amplification reagent 1
in DMSO (56.8 mM, 360 mL). The piperidine released from
the detection event then initiated the signal amplification
reaction with 1.
We acknowledge financial support from DARPA (N66001-
09-1-2111 and HR0011-11-1-0007), the Arnold and Mabel
Beckman Foundation, the Camille and Henry Dreyfus Founda-
tion, Mr Louis Martarano, and The Pennsylvania State University.
We thank Dr Landy K. Blasdel for assistance in preparing the
manuscript.
Notes and references
1 P. Yager, T. Edwards, E. Fu, K. Helton, K. Nelson, M. R. Tam
and B. H. Weigl, Nature, 2006, 442, 412; D. Mabey, R. W. Peeling,
A. Ustianowski and M. D. Perkins, Nat. Rev. Microbiol., 2004,
2, 231; D. A. Giljohann and C. A. Mirkin, Nature, 2009, 462, 461.
2 M. Urdea, L. A. Penny, S. S. Olmsted, M. Y. Giovanni, P. Kaspar,
A. Shepherd, P. Wilson, C. A. Dahl, S. Buchsbaum, G. Moeller
and D. C. Hay Burgess, Nature, 2006, 444, 73.
3 M. S. Baker and S. T. Phillips, J. Am. Chem. Soc., 2011, 133, 5170.
4 R. Perry-Feigenbaum, E. Sella and D. Shabat, Chem.–Eur. J.,
2011, 17, 12123.
5 E. Sella, R. Weinstain, R. Erez, N. Z. Burns, P. S. Baran and
D. Shabat, Chem. Commun., 2010, 46, 6575.
6 N. C. Gianneschi, S. T. Nguyen and C. A. Mirkin, J. Am. Chem.
Soc., 2005, 127, 1644.
7 E. Sella, A. Lubelski, J. Klafter and D. Shabat, J. Am. Chem. Soc.,
2010, 132, 3945.
8 G. R. Newman and B. Jasani, J. Pathol., 1998, 186, 119.
9 D. G. Cho and J. L. Sessler, Chem. Soc. Rev., 2009, 38, 1647.
10 A. L. Garner, F. L. Song and K. Koide, J. Am. Chem. Soc., 2009,
131, 5163.
11 M. Frenklach and D. Clary, Ind. Eng. Chem. Fundam., 1983,
22, 433.
12 K. M. Schmid, S. T. Phillips, unpublished results.
13 More highly developed assays for Pd can be found in references 9
and 10 and in other references either cited by 9 or 10, or that cite 9
or 10.
14 C. E. Garrett and K. Prasad, Adv. Synth. Catal., 2004, 346, 889.
15 A. P. Esser-Kahn, S. A. Odom, N. R. Sottos, S. R. White and
J. S. Moore, Macromolecules, 2011, 44, 5539.
16 Related base amplification reagents have been developed as photo-
base generators for photolithography. See for example:
K. Arimitsu, M. Miyamoto and K. Ichimura, Angew. Chem., Int.
Ed., 2000, 39, 3425; K. Ichimura, J. Photochem. Photobiol., A,
2003, 158, 205.
To determine the effectiveness of this tandem sequence, we
performed several experiments with various initial quantities
of palladium. In all cases, the signal amplification reaction was
allowed to proceed for 16 h, after which we measured the
absorbance of dibenzofulvene at 305 nm. The dose–response
curve shown in Fig. S6 (ESIw) reveals that (i) the limit of
c
3020 Chem. Commun., 2012, 48, 3018–3020
This journal is The Royal Society of Chemistry 2012