Synthesis and application of poly(phenylene ethynylene)s for bioconjugation: A conjugated polymer-based fluorogenic probe for proteases

Article abstract of DOI:10.1021/ja043134b

A set of carboxylate-functionalized poly(phenylene ethynylene)s (PPEs) has been synthesized in which the carboxylic acid groups are separated from the polymer backbone by oligo(ethylene glycol) spacer units. These polymers are soluble in water and organic

Full text of DOI:10.1021/ja043134b

Published on Web 02/18/2005
Synthesis and Application of Poly(phenylene Ethynylene)s for
Bioconjugation: A Conjugated Polymer-Based Fluorogenic
Probe for Proteases
Jordan H. Wosnick,† Charlene M. Mello,‡ and Timothy M. Swager*,†,§
Contribution from the Department of Chemistry and Institute for Soldier Nanotechnologies,
Massachusetts Institute of Technology, Cambridge, Massachusetts 02139, and
U.S. Army Research, DeVelopment and Engineering Command, Natick Soldier Center,
Natick, Massachusetts 01760
Received November 15, 2004; E-mail: tswager@mit.edu
Abstract: A set of carboxylate-functionalized poly(phenylene ethynylene)s (PPEs) has been synthesized
in which the carboxylic acid groups are separated from the polymer backbone by oligo(ethylene glycol)
spacer units. These polymers are soluble in water and organic solvents and have photophysical properties
that are sensitive to solvent conditions, with high salt content and the absence of surfactant promoting the
formation of aggregates of relatively low quantum yield and long fluorescence lifetime. Quenching of these
materials by the dinitrophenyl (DNP) chromophore (KSV ∼ 104) is also highly solvent-dependent. The
presence of carboxylate groups far from the polymer backbone appended to each repeating unit allows for
the postpolymerization modification of these PPEs with peptides by methods analogous to those described
for carboxylate-functionalized small-molecule dyes. Covalent attachment of the fluorescence-quenching
14-mer Lys(DNP)-GPLGMRGLGGGGK to the PPE results in a nonemissive substrate whose fluorescence
is restored upon treatment with trypsin. The rate of fluorescence turn-on in this case is increased 3-fold by
the presence of surfactant, though the actual rate of peptide hydrolysis remains the same. A small-molecule
mimic of the polymer-peptide system shows a smaller fluorescence enhancement upon treatment with
trypsin, illustrating the value of polymer-based amplification in this sensory scheme.
Introduction
solubilizing groups, usually in the form of ionized side chains.
The strong interactions between the resulting conjugated poly-
Electronically conjugated semiconductive polymers such as
poly(phenylene ethynylene)s (PPEs) have been used extensively
by our group and others as signal-transducing elements for
chemosensors.1 In particular, electron-deficient analytical targets
such as nitroaromatics and methyl viologen (N,N′-dimethyl-4,4′-
bipyridinium) efficiently quench the fluorescence of conjugated
polymers by providing a mechanism for the nonradiative
recombination of electron-hole pairs (excitons).2 This phe-
nomenon has been used to create highly sensitive sensors for
viologens and vapors of the explosive trinitrotoluene (TNT)
based on a PPE fluorescence turn-off mechanism. The extraor-
dinary signal amplification provided by exciton migration along
the conjugated polymer backbone has inspired considerable
effort by several groups to apply this concept toward the
detection of analytes of biological interest. For such applications
to be feasible, the inherently hydrophobic nature of conjugated
polymer backbones must be overcome by the addition of water-
electrolytes3 and charged molecules can be exploited for
analytical purposes. For example, methyl viologen quenches the
fluorescence of a sulfonated poly(phenylene vinylene) (PPV)
with a Stern-Volmer quenching constant KSV of ∼107, and
biotinylated viologen derivative was used to create an electro-
statically quenched system in which fluorescence is restored
upon addition of avidin.4 Similarly, extremely efficient fluo-
rescence quenching of conjugated polyelectrolyte fluorescence
has been observed with oppositely charged Fe3+-containing
proteins,5 and polymeric carbohydrate sensors utilizing violo-
gen-boronic acid quenchers have been prepared wherein the
fluorescence quenching ability of the viologen core is reduced
upon complexation to the vicinal diol units of monosaccharides.6
Very recently, PPEs containing pendant mannose units were
shown to be effective in the detection of bacteria containing
complementary receptors capable of polyvalent binding.7 Other
recent applications have included turn-on sensors for proteases
that make use of the release of a nitroanilide group from a
† Department of Chemistry, Massachusetts Institute of Technology.
‡ U.S. Army Research, Development and Engineering Command, Natick
Soldier Center.
(3) Pinto, M. R.; Schanze, K. S. Synthesis 2002, 9, 1293.
(4) For an example of reversible quenching aided by strong electrostatic
quencher-polymer interactions, see Chen, L.; McBranch, D. W.; Wang,
H.-L.; Helgelson, R.; Wudl, F.; Whitten, D. G. Proc. Natl. Acad. Sci. U.S.A.
1999, 96, 12287.
(5) Fan, C.; Plaxco, K. W.; Heeger, A. J. J. Am. Chem. Soc. 2002, 124, 5642.
(6) DiCesare, N.; Pinto, M. R.; Schanze, K. S.; Lakowicz, J. R. Langmuir 2002,
18, 7785.
(7) Disney, M. D.; Zheng, J.; Swager, T. M.; Seeberger, P. H. J. Am. Chem.
Soc. 2004, 126, 13343.
§ Institute for Soldier Nanotechnologies, Massachusetts Institute of
Technology.
(1) (a) Swager, T. M. Acc. Chem. Res. 1998, 31, 201. (b) McQuade, D. T.;
Pullen, A. E.; Swager, T. M. Chem. ReV. 2000, 100, 2537. (c) Swager, T.
M.; Wosnick, J. H. MRS Bull. 2002, 446.
(2) (a) Zhou, Q.; Swager, T. M. J. Am. Chem. Soc. 1995, 117, 12593. (b)
Yang, J. S.; Swager, T. M. J. Am. Chem. Soc. 1998, 120, 5321. (c) Yang,
J. S.; Swager, T. M. J. Am. Chem. Soc. 1998, 120, 11864.
9
3400
J. AM. CHEM. SOC. 2005, 127, 3400-3405
10.1021/ja043134b CCC: $30.25 © 2005 American Chemical Society

Poly(phenylene Ethynylene)s for Bioconjugation
A R T I C L E S
cationic peptide electrostatically bound to an anionic PPE8 or
of strongly quenching chromophores from a biotinylated pep-
tide.9 Biotin-avidin interactions have been used to tether nucleic
acid probes to PPEs for the detection of specific DNA sequences
through fluorescence quenching effects.10 Conjugated polymer-
based biosensors based on fluorescence effects other than
nonradiative quenching have also been reported. The groups of
Leclerc11 and Nilsson12 have made use of changes in the
photophysical properties of cationic polythiophenes on com-
plexation to anionic DNA to create highly sensitive DNA
sensors, while Bazan and co-workers13 have used the interaction
of cationic polyfluorene derivatives with nucleic acids to control
energy transfer between the conjugated polymers and fluoro-
phore-labeled target DNA.
enhancement of 1 order of magnitude upon treatment with the
proteolytic enzyme trypsin.
Results and Discussion
Polymer Synthesis and Properties. Among the standard
techniques used for the derivatization of biomolecules with
small-molecule labels, one of the most common is the formation
of amides by use of biomolecular amine groups and activated
esters of carboxylate-functionalized small molecules. The broad
utility of this approach to biomolecule labeling inspired us to
develop parallel methods for the derivatization of bioreactive
molecules with activated esters of conjugated polymers. Toward
this end, we designed and synthesized a diiodophenylene
monomer suitable for use in PPE synthesis in which carboxylic
acid moieties are separated from the aromatic core by hydro-
philic hexaethylene glycol linkers (Scheme 1). In a variation
of a literature procedure,17 a tetrahydrofuran (THF) solution of
pentaethylene glycol was treated with a substoichiometric
quantity of sodium metal followed by tert-butyl acrylate. Excess
unreacted pentaethylene glycol was easily separated from the
resultant Michael adduct on the basis of the differential water
solubilities of the starting material and product. Conversion of
the remaining free alcohol to its p-toluenesulfonate ester
followed by Williamson etherification with 1,4-diiodohydro-
quinone provided the diester 6 as a clear oil in good yield.
Removal of the tert-butyl groups with neat trifluoroacetic acid
(TFA) provided the acid monomer 8 as a white powder. A
similar monomer 7 based on a tri(ethylene glycol) linker was
prepared by an analogous method. The presence of the hydro-
philic linker groups in 8 and 7 encourages water solubility and
is also expected to reduce nonspecific interactions with bio-
molecules. Sonogashira-Hagihara copolymerization of 8 with
a dialkyne comonomer bearing di(ethylene glycol) side chains
provided polymer 9a, which dissolved in THF, N,N-dimethyl-
formamide (DMF), and water to give brightly fluorescent
solutions. Polymer 9b, which contains tri(ethylene glycol)
comonomer units, was prepared in an analogous manner, and
its extremely high water solubility made it useful in photo-
physical studies where concentrated polymer solutions are
required. Successful gel-permeation chromatography (GPC)
measurements could not be made for these polymers, perhaps
because of nonspecific adsorption to the column packing
material, but we estimate their molecular weights to be in excess
of 25 000 based on dialysis and gel-filtration behavior.
A unifying feature of most of the biosensory methods
described above is the use of simple, unmodified conjugated
polyelectrolytes as sensory signal amplification elements.
Because the signal transduction mechanism in these systems is
reliant on electrostatic effectsswhich mediate a wide variety
of biomolecular interactions involving charge-bearing proteins
and nucleic acidssthese sensor designs necessarily suffer from
the possibility of nonspecific polymer-analyte interactions that
may reduce the sensitivity or fidelity of the system. Previous
research by our group has sought to overcome these limitations
by use of nonionic water-soluble polymers with dendritic side
chains.14 These polymers can be end-capped with hydrogel-
forming, thermally responsive poly(N-isopropylacrylamide)
blocks, allowing for temperature control of biomolecule-
polymer interactions.15 In this paper, we report an alternative
approach to the construction of conjugated polymer biosensors
based on water-soluble PPEs containing carboxylic acid-
terminated oligo(ethylene glycol) side chains. The presence of
multiple carboxylate units tethered to the polymer backbone
allows for postpolymerization activation and bioconjugation of
the polymer side chains to amine-containing molecules in a
manner analogous to that used for small-molecule fluoro-
phores.16 The photophysical properties and quenching behavior
of these polymers toward small-molecule nitroaromatics have
been found to be highly dependent on solution pH and surfactant
concentration. We have exploited the chemical and physical
properties of these materials to generate a polymer-amplified
turn-on fluorogenic probe for proteases based on the covalent
attachment of a fluorescence-quenching 14-mer peptide to the
polymer side chains. The resulting system shows a fluorescence
The spectral properties of 9a and 9b in DMF solution (λabs
430 nm, λem 470 nm) are similar to those of conventional PPEs
in organic solvents. However, we noted that the shapes of the
absorbance and fluorescence spectra in aqueous solution are
highly pH-dependent, displaying features characteristic of
polymer-polymer aggregation in unbuffered solutions that
disappear when the solution is rendered alkaline (pH 11) with
dilute NaOH (data not shown). These observations would seem
to suggest that this polymer undergoes a pH-sensitive aggrega-
tion process in which alkaline environments increase the
negative charge on the carboxylate-functionalized PPE, aug-
menting the water solubility of individual polymer chains and
discouraging interchain interactions. A similar mechanism has
been proposed to explain the pH- and solvent-dependent
spectroscopic features of phosphonate- and sulfonate-substituted
(8) Pinto, M. R.; Schanze, K. S. Proc. Natl. Acad. Sci. U.S.A. 2004, 101, 7505.
(9) Kumaraswamy, S.; Bergstedt, T.; Shi, X.; Rininsland, F.; Kushon, S.; Xia,
W.; Ley, K.; Achyuthan, K.; McBranch, D.; Whitten, D. Proc. Natl. Acad.
Sci. U.S.A. 2004, 101, 7511.
(10) (a) Kushon, S. A.; Ley, K. D.; Bradford, K.; Jones, R. M.; McBranch, D.;
Whitten, D. Langmuir 2002, 18, 7245. (b) Kushon, S. A.; Bradford, K.;
Marin, V.; Suhrada, C.; Armitage, B. A.; McBranch, D.; Whitten, D.
Langmuir 2003, 19, 6456.
(11) (a) Ho, H. A.; Boissinot, M.; Bergeron, M. G.; Corbeil, G.; Dore, K.;
Boudreau, D.; Leclerc, M. Angew. Chem., Int. Ed. Engl. 2002, 41, 1548.
(b) Dore, K.; Dubus, S.; Ho, H. A.; Levesque, I.; Brunette, M.; Corbeil,
G.; Boissinot, M.; Boivin, G.; Bergeron, M. G.; Boudreau, D.; Leclerc, M.
J. Am. Chem. Soc. 2004, 126, 4240.
(12) Nilsson, K. P. R.; Inganaes, O. Nat. Mater. 2003, 2, 419.
(13) (a) Gaylord, B. S.; Heeger, A. J.; Bazan, G. C. Proc. Natl. Acad. Sci. U.S.A.
2002, 99, 10954. (b) Gaylord, B. S.; Heeger, A. J.; Bazan, G. C. J. Am.
Chem. Soc. 2003, 125, 896. (c) Wang, S.; Bazan, G. C. AdV. Mater. 2003,
15, 1425.
(14) Kuroda, K.; Swager, T. M. Chem. Commun. 2003, 26.
(15) Kuroda, K.; Swager, T. M. Macromolecules 2004, 37, 716.
(16) (a) Aslam, M.; Dent, A. Bioconjugation; MacMillan Reference: London,
1998; Chapt. 6. (b) Hermanson, G. Bioconjugate techniques; Academic
Press: San Diego, CA, 1996.
(17) Seitz, O.; Kunz, H. J. Org. Chem. 1997, 62, 813.
9
J. AM. CHEM. SOC. VOL. 127, NO. 10, 2005 3401

A R T I C L E S
Wosnick et al.
Scheme 1. Synthesis of Polymers 9a and 9ba
a Synthetic conditions: (a) (i) Na, THF, (ii) tert-butyl acrylate, 16 h; (b) p-toluenesulfonyl chloride, Et3N, CH2Cl2, 16 h; (c) 2,5-diiodohydroquinone,
K2CO3, KI, 2-butanone, reflux; (d) CF3COOH, neat; (e) Pd(PPh3)4 (cat.), CuI (cat.), morpholine, 60 °C.
PPEs.18 Although the minimization of polymer aggregation
effects would seem to be necessary to achieve the greatest
possible sensitivity in assay schemes that use PPE labels, the
need for stable, near-neutral buffered solutions precludes the
use of a highly alkaline environment to control aggregation in
these applications. Instead, we screened a number of com-
mercially available nonionic surfactants for their ability to
increase the fluorescence quantum yield (QY) of these PPEs in
neutral and buffered solutions. Triton X-100, an oligo(ethylene
glycol) with a single hydrophobic headgroup, was found to be
remarkably effective in this regard: the addition of about 0.5
wt % of this surfactant to a solution of 9b in unbuffered water
increased the QY of the polymer from 0.11 to 0.20. The same
effect was observed in a solution of 9b in a pH 7.5 Tris-buffered
saline (TBS) solution, although the magnitude of the enhance-
ment was not as large (Figure 1). This difference can be
attributed to the larger inherent hydrophobic forces in solutions
of high ionic strength.
The data in Figure 1 clearly show that 30-50% increases in
Figure 1. Fluorescence enhancement of polymer 9b (0.7 µM in repeat
QY are possible with the addition of small concentrations of
units) in Tris-buffered saline (TBS; 50 mM Tris, pH 7.5, 150 mM NaCl,
and 5 mM CaCl2) upon addition of Triton X-100, with approximate quantum
yields. The concentration of polymer is identical in each of the scans.
Triton X-100. Concomitant with the fluorescence intensity
increases is a gradual change in the shape of the fluorescence
spectra, in which the 480 nm red-shifted emission maximum
suggestive of an aggregate19 gives way to a single fluorescence
feature at 460 nm. In this case, the addition of surfactant results
in a fluorescence profile for 9b that resembles the spectral
features of PPEs typically observed in good solvents. The results
of the pH and surfactant experiments suggest that in aqueous
solution the polymer exists in equilibrium between an ag-
gregated, π-stacked form of relatively low QY and a highly
solubilized form of higher QY in which interchain interactions
are minimized. To determine if these two forms can be identified
photophysically, the fluorescence lifetime of 9b was measured
in DMF and three different buffers, both with and without added
surfactant (Table 1).
(18) (a) Pinto, M. R.; Kristal, B. M.; Schanze, K. S. Langmuir 2003, 19, 6523.
(b) Tan, C.; Pinto, M. R.; Schanze, K. S. Chem. Commun. 2002, 446.
(19) McQuade, D. T.; Kim, J.; Swager, T. M. J. Am. Chem. Soc. 2000, 122,
5885.
9
3402 J. AM. CHEM. SOC. VOL. 127, NO. 10, 2005

Poly(phenylene Ethynylene)s for Bioconjugation
A R T I C L E S
interest.21 Internal energy transfer maintains the system in a
nonfluorescent state until the peptide is cleaved. Modifications
of this general design have given rise to substrates in which
the peptide and fluorophore are tethered to a dendrimer22 or a
soluble, nonconjugated polymer support.23 As an initial dem-
onstration of the utility of the carboxylate-containing polymers
described above, we chose to construct a fluorogenic protease
substrate in which a water-soluble PPE acts as both soluble
support and fluorophore. The electron-accepting 2,4-dinitro-
phenylamino group (DNP) stood out as an excellent potential
fluorescence quencher for this system because of the known
sensitivity of PPE fluorescence to the presence of nitroaromatics2
and the commercial availability of Fmoc-protected DNP-
substituted lysine (Fmoc-LysDNP) to facilitate incorporation of
the quenching unit during peptide synthesis.
Table 1. Fluorescence Lifetimes of 9b in Aqueous Buffer
Solutions and in DMFa
surfactant/buffer
pH 5.0 (acetate)
pH 7.3 (PBS)
pH 8.5 (borate)
DMF
no surfactant
0.64 (0.82) 0.64 (0.82) 0.67 (0.93) 0.77
3.81 (0.18) 4.05 (0.18) 3.97 (0.07)
0.1 wt % Triton X-100 0.77 (0.97) 0.70 (0.87) 0.75 (0.98)
3.90 (0.03) 3.24 (0.13) 3.71 (0.02)
0.2 wt % Triton X-100
0.73 (0.92)
4.24 (0.08)
a Fluorescence lifetimes are given in nanoseconds. In cases where the
fluorescence follows a biexponential decay pattern, the two component
lifetimes are given along with their relative contributions in parentheses
(to a total of 1.0). Polymer solutions were excited at 405 nm and
fluorescence was recorded at 470 nm by use of RF phase modulation (see
Supporting Information).
The fluorescence lifetime of 9b in DMF solution was 0.77
ns, a value typical for PPEs in good solvents. In the case of
aqueous solutions, we found that the fluorescence decay was
biexponential, indicating the presence of two distinct fluorescent
species in solution or two radiative decay pathways.20 The
components of the fluorescence decay correspond to a dominant
short-lived species of τ ) 0.6-0.7 nsssimilar to the lifetime
of the polymer in DMFstogether with minor contributions (2-
18% of the observed fluorescence decay) from a longer-lived
3-4 ns component. As Table 1 indicates, the relative popula-
tions of the two species are strongly affected by the addition of
surfactant, which was found to greatly increase the contribution
of the short-lifetime component at the expense of the longer-
lived species. Alkaline buffer environments also generally
favored increased contributions from the short-lifetime com-
ponent. In general, the data indicate that the same conditions
that promote large quantum yields and sharp emission spectra
(organic solvents, alkaline pH values, or surfactant in aqueous
solutions) also result in an increase in the relative contribution
of the short-lived species. These observations are consistent with
a model in which the polymer chains are aggregated in neutral
to acidic solutions and well-solvated in organic solvents, when
the solution is rendered alkaline, or when surfactant is added.
In this case the short-lived component of the fluorescence
lifetime corresponds to the nonaggregated polymer and makes
a larger contribution to the overall fluorescence lifetime when
the relative proportion of polymer chains in this state is
increased. Similarly, the long-lived fluorescence decay can be
assigned to polymer aggregates that are broken up at conditions
of high pH and surfactant concentration and are completely
absent in DMF solutions. These aggregates are expected to form
through strong hydrophobic and π-stacking interactions between
polymer chains, and their long fluorescence lifetime is the
signature of a weakly allowed transition from the lowest excited
state of the aggregate to the ground state.18 In phosphate-
buffered saline (PBS) solutions of 9b, the high ionic strength
of the medium encourages greater aggregation through hydro-
phobic interactions, requiring higher concentrations of surfactant
to break up aggregates relative to buffer solutions lacking added
salts.
To verify the quenching ability of the DNP group toward
carboxylate-functionalized water-soluble PPEs, Stern-Volmer
quenching constants KSV were measured for LysDNP under a
variety of conditions. In unbuffered water solutions, polymer
9b was quenched by free LysDNP with a KSV of 5.8 × 104
M-1. The presence of Triton X-100 at low concentration reduced
this value by a factor of 4 to 1.4 × 104 M-1, similar to the KSV
determined in Tris-buffered saline (1.5 × 104 M-1). Addition
of surfactant to the TBS solution reduced KSV still further to
1.0 × 104 M-1. Comparable behavior was observed for polymer
9a. Given the nanosecond-scale fluorescence lifetimes described
above, the large magnitudes of KSV in all these cases strongly
implies that static binding of LysDNP is responsible for the
observed quenching, and in this context the differences in KSV
observed under various solvent and surfactant conditions can
be understood in terms of variations in the effective hydropho-
bically driven association constant of LysDNP and the PPE.
Here the addition of surfactant and the use of a controlled near-
neutral pH reduces hydrophobic interactions between LysDNP
and the PPE, suggesting that the pH, ionic strength, and
surfactant concentration of the assay medium will be important
in determining residual background fluorescence and maximum
fluorescence enhancement in our proposed protease substrate.
To link the fluorescence-quenching DNP group to the PPE,
we chose to use a peptide sequence (GPLGMRGL) that has
previously been identified as a potent substrate for a variety of
matrix metalloproteases, most notably the cancer-associated
human collagenase III (MMP-13).24 We also noted that the
single arginine residue in the middle of the peptide sequence
renders it susceptible to cleavage by trypsin, an inexpensive
and easily handled digestive enzyme that hydrolyzes peptides
on the carboxyl side of unmodified arginine and lysine residues.
The peptide sequence AcHN-Lys(DNP)-GPLGMRGLGGGG-
Lys-OH was designed to incorporate both the fluorescence-
quenching DNP group as part of the amino acid sequence and
(21) Recent examples of fluorogenic substrates for proteases: (a) Yan, Z.-H.;
Ren, K.-J.; Wang, Y.; Chen, S.; Brock, T. A.; Rege, A. A. Anal. Biochem.
2003, 312, 141. (b) Re´hault, S.; Brillard-Bourdet, M.; Bourgeois, L.;
Frenette, G.; Juliano, L.; Gauthier, F.; Moreau, T. Biochim. Biophys. Acta
2002, 1596, 55. (c) Mu¨ller, J. C. D.; Ottl, J.; Moroder, L. Biochemistry
2000, 39, 5111. (d) Yang, C. F.; Porter, E. S.; Boths, J.; Kanyi, D.; Hsieh,
M.; Cooperman, B. S. J. Pept. Res. 1999, 54, 444.
Polymer-Based Protease Substrate Design and Quenching
Efficiency. In conventional fluorogenic protease substrates, a
fluorescent donor and nonfluorescent acceptor dye are separated
by a peptide sequence known to be cleaved by the enzyme of
(22) McIntyre, J. O.; Fingleton, B.; Wells, K. S.; Piston, D. W.; Lynch, C. C.;
Gautam, S.; Matrisian, L. M. Biochem. J. 2004, 377, 617.
(23) Law, B.; Curino, A.; Bugge, T. H.; Weissleder, R.; Tung, C.-H. Chem.
Biol. 2004, 11, 99. For a review of recent work in this area, see Funovics,
M.; Weissleder, R.; Tung, C.-H. Anal. Bioanal. Chem. 2003, 377, 956.
(24) Deng, S. J.; Bickett, D. M.; Mitchell, J. L.; Lambert, M. H.; Blackburn, R.
K.; Carter, H. L.; Neugebauer, J.; Pahel, G.; Weiner, M. P.; Moss, M. L.
J. Biol. Chem. 2000, 275, 31422.
(20) Lakowicz, J. R. Principles of fluorescence spectroscopy; Kluwer Aca-
demic: New York, 1999.
9
J. AM. CHEM. SOC. VOL. 127, NO. 10, 2005 3403

A R T I C L E S
Wosnick et al.
Scheme 2. Synthesis of Quenched Peptidic (10) and Nonpeptidic (11) Fluorogenic Substratesa
a Substrates are shown fully substituted for simplicity (see text).
a C-terminal lysine residue through which coupling to the PPE
can take place via an amide linkage. The point of attachment is
separated from the cleavage sequence by four glycine spacer
units. The peptide was synthesized by solid-phase techniques
and end-capped with an acetyl group on the resin. The peptide
was then coupled to 9a after activation of the polymer with
excess EDAC and N-hydroxysuccinimide, providing 10 after
dialysis.16 Polymer 11, an analogue of 10 containing a nonpep-
tidic oligo(ethylene glycol) spacer, was prepared by similar
means (Scheme 2). The peptide loading in 10 was estimated
by quantifying the contributions of the DNP and PPE chro-
mophores to the absorption spectrum of dialyzed 10, giving
values between 0.25 and 1.7 peptide chains/phenylene eth-
ynylene repeat unit, depending on the stoichiometry and
activation time of the conjugation reaction. Solutions of 10 were
visibly nonfluorescent, and related control experiments revealed
that prior activation of 9a with carbodiimide was necessary for
the generation of internally quenched substrates.
Protease Detection by Use of a Polymer-Based Fluorogenic
Substrate. To test the efficacy of 10 as a protease sensor, a
sample of 10 (1.1 µM in repeat units) with high peptide loading
(1.7 peptides/repeat unit) was subjected to hydrolysis by trypsin
in dilute solution (3 µg/mL). As shown in Figure 2, the system
exhibits an increase in fluorescence of approximately 1 order
of magnitude after hydrolysis, while the absorbance spectrum
remains largely unchanged. Control experiments (Figure 2b)
showed that turn-on of fluorescence in the presence of trypsin
is rapid but is slowed considerably by the trypsin inhibitor
benzamidine hydrochloride. No fluorescence increase occurs
Figure 2. (a) Absorbance (left) and fluorescence (right) spectra of substrate
10 (1.1 µM in repeat units) before and after treatment with trypsin (3 µg/
mL). (b) Generation of fluorescence as a function of time on treatment of
10 and control substrate 11 with trypsin in the absence and presence of an
inhibitor (benzamidine hydrochloride).
with the nonpeptidic substrate 11 or in the presence of the
inhibitor alone.
Ionic Strength and Surfactant Effects in Fluorescence
Turn-on. To further study the mechanism of fluorescence turn-
on in this system, we examined the overall response and relative
kinetic behavior of 10 lightly loaded with fluorescence-
quenching peptide (0.25 peptide chain/polymer repeat unit) with
trypsin in solutions containing varying concentrations of buffer
(Tris, pH 7.5), salts (NaCl, CaCl2), and surfactant (Triton
X-100). The data in Figure 3 show that larger trypsin-induced
fluorescence enhancements are generally favored by conditions
of lower ionic strength and the absence of surfactant. In the
latter case, the presence of surfactant reduces the overall degree
of fluorescence enhancement by augmenting the fluorescence
of both the initial (internally quenched) and, to a smaller extent,
9
3404 J. AM. CHEM. SOC. VOL. 127, NO. 10, 2005

Poly(phenylene Ethynylene)s for Bioconjugation
A R T I C L E S
Chart 1. Structure of 12, a Small-Molecule Mimic of 10
peptide cleavage, it did increase the rate of fluorescence
generation by a factor of 3. The kinetic and overall fluorescence
enhancement data suggest that a strong hydrophobic effect
mediates the interaction between the PPE backbone and the
fluorescence-quenching peptide fragment both before and after
hydrolysis of the peptide linker. By solubilizing hydrophobic
portions of the oligopeptide, the surfactant reduces quencher-
polymer interactions both before and after peptide cleavage and
leads to a faster rate of fluorescence increase.
To quantify the sensory advantage provided by use of a
conjugated polymer as the fluorophore in 10, we synthesized
the substrate 12, a small-molecule mimic of 10 (Chart 1).
Relative to 10, substrate 12 has a shorter side-chain linker
between the fluorophore and the peptide chain in addition to a
larger effective peptide loading (2 peptide units/molecule).
Although both of these factors would ordinarily be expected to
improve the efficiency of 12 as a fluorogenic probe, its
fluorescence intensity increases only about 2.9-fold upon
hydrolysis by trypsin, an overall increase several times smaller
than that seen for 10. In this case, the favorable contributions
of hydrophobic effects and exciton migration in 10 negate the
effects of the larger fluorophore-quencher distance and lower
quencher loading relative to 12.
Figure 3. Overall fluorescence enhancement in substrate 10 (0.25 peptide/
repeat unit) before and after hydrolysis by trypsin. Numbers in parentheses
indicate the overall fluorescence enhancement. TBS: 50 mM Tris, pH 7.5,
150 mM NaCl, and 5 mM CaCl2. TBS/Triton: as for TBS, with 0.08%
Triton X-100. Tris: 50 mM Tris, pH 7.5.
Table 2. Relative Initial Reaction Rates and Initial Rates of
Fluorescence Increase for the Hydrolysis of 10 under Conditions
of Varying Ionic Strength and Surfactant Concentration
relative rates
conditionsa
peptide cleavage
fluorescence generation
In conclusion, we have prepared water-soluble poly(phenylene
ethynylene)s containing carboxylate-functionalized side chains.
These materials show solvent- and surfactant-dependent pho-
tophysical properties and are efficiently quenched by the
dinitrophenyl (DNP) chromophore. Derivatization of these
polymers with a fluorescence-quenching peptide susceptible to
proteolytic cleavage by trypsin provides a fluorogenic substrate
that shows large fluorescence increases upon hydrolysis by this
enzyme, even in cases of extremely low peptide loading.
TBS
TBS/Triton
Tris
1
1
1.1 ( 0.2
4.1 ( 0.6
2.9 ( 0.2
5.3 ( 0.2
a See caption to Figure 3.
the final (cleaved) forms of 10.25 The fact that this lightly loaded
polymeric substrate shows an overall fluorescence enhancement
of a magnitude similar to that seen in our initial experiment
(1.7 peptide chains per polymer repeat unit) attests to the ability
of relatively few DNP groups to efficiently quench the
fluorescence of an entire PPE chain.
Acknowledgment. This work was supported by the NASA-
NCI Fundamental Technologies for Biomolecular Sensors
program. We thank Jennifer Burzycki and Ghislaine Bailey for
experimental assistance, Samuel Thomas III and Juan Zheng
for helpful discussions, and the MIT Institute for Soldier
Nanotechnologies for the use of HPLC equipment. This paper
reports research undertaken at the U.S. Army Research,
Development and Engineering Command, Natick Soldier Center,
Natick, MA, and has been assigned No. NATICK/TP-05/015
in a series of reports approved for publication.
The kinetic behavior of the lightly loaded protease substrate
described above was investigated by measuring the relative
initial rates of peptide hydrolysis (measured by HPLC) and
fluorescence turn-on under the same set of ionic strength and
surfactant conditions. As shown in Table 2, the initial reaction
rate is greatly increased under conditions of lower ionic strength.
This follows general trends previously reported for the trypsin
hydrolysis of N-R-benzoylarginine ethyl ester (BAEE).26 Inter-
estingly, we found that while the addition of Triton X-100 to
the reaction medium did not significantly change the rate of
Supporting Information Available: Synthetic and experi-
mental procedures (PDF). This information is available free of
charge via the Internet at http://pubs.acs.org.
JA043134B
(25) We have found that the addition of Triton X-100 increases the fluorescence
quantum yield of neutral, PEGylated PPEs in water approximately 14-fold.
Surfactants have previously been found to significantly increase the quantum
efficiency of poly(phenylene vinylene)s: Chen, L. H.; Xu, S.; McBranch,
D.; Whitten, D. J. Am. Chem. Soc. 2000, 122, 9302.
(26) (a) Be´chet, J. J.; Chevallier, J.; Labouesse, J.; Arnould, B.; Aldin-Cicarelli,
S.; Yon, J. Biochim. Biophys. Acta 1968, 151, 165. (b) Castan˜eda-Agullo´,
M.; Del Castillo, L. M.; Whitaker, J. R.; Tappel, A. L. J. Gen. Physiol.
1961, 44, 1103.
9
J. AM. CHEM. SOC. VOL. 127, NO. 10, 2005 3405