Polymer-DNA hybrids as electrochemical probes for the detection of DNA

Article abstract of DOI:10.1021/ja046931i

The syntheses of several norbornene block copolymers containing oligonucleotide and ferrocenyl side chains and their use in the electrochemical detection of DNA are described. Two kinds of DNA-containing block copolymers with either ferrocenyl or dibromoferrocenyl groups were prepared via ring-opening metathesis polymerization (ROMP). Based on these two distinct ferrocene derivatives, a triblock copolymer labeling strategy was developed. With this strategy, the identity of DNA target can be determined by the E _1/2s of the ferrocenyl moieties and the ratio of peak currents. These polymers exhibit predictable and tailorable electrochemical properties, high DNA duplex stability, and unusually sharp melting transitions, which are highly desirable characteristics for DNA detection applications. Significantly, single-base mismatches could be easily detected using two distinct block copolymers as dual-channel detection probes in an electrochemical DNA detection format.

Full text of DOI:10.1021/ja046931i

Published on Web 01/07/2005
Polymer-DNA Hybrids as Electrochemical Probes for the
Detection of DNA
Julianne M. Gibbs, So-Jung Park, Donde R. Anderson, Keith J. Watson,
Chad A. Mirkin,* and SonBinh T. Nguyen*
Contribution from the Department of Chemistry and Institute for Nanotechnology,
Northwestern UniVersity, 2145 Sheridan Road, EVanston, Illinois 60208-3113
Received May 25, 2004; E-mail: chadnano@northwestern.edu; stn@northwestern.edu
Abstract: The syntheses of several norbornene block copolymers containing oligonucleotide and ferrocenyl
side chains and their use in the electrochemical detection of DNA are described. Two kinds of
DNA-containing block copolymers with either ferrocenyl or dibromoferrocenyl groups were prepared via
ring-opening metathesis polymerization (ROMP). Based on these two distinct ferrocene derivatives, a triblock
copolymer labeling strategy was developed. With this strategy, the identity of DNA target can be determined
by the E1/2s of the ferrocenyl moieties and the ratio of peak currents. These polymers exhibit predictable
and tailorable electrochemical properties, high DNA duplex stability, and unusually sharp melting transitions,
which are highly desirable characteristics for DNA detection applications. Significantly, single-base
mismatches could be easily detected using two distinct block copolymers as dual-channel detection probes
in an electrochemical DNA detection format.
Introduction
havior in several DNA detection systems have led to significant
improvements in the selectivities of such assays.7-9 In addition,
During the past few years there has been growing interest in
the synthesis and utilization of nanostructured materials in which
multiple DNA strands of known sequence are attached to a
central “core”. These materials include oligonucleotide-modified
gold nanoparticles and oligonucleotide-functionalized den-
drimers.1-4 These particular materials have been utilized in a
number of DNA detection applications,4-9 and their novel
properties have led to major advantages with respect to
sensitivity, selectivity, ease of data readout and analysis, and
even cost. For example, in the case of the oligonucleotide-
modified gold nanoparticles, many (over 100) thiol-modified
DNA strands are attached to the surface of citrate-stabilized gold
nanoparticles via a ligand exchange process.1,10 The resulting
oligonucleotide-modified nanoparticles after complexation with
other nanoparticles that have been modified with complementary
oligonucleotide sequences exhibit extraordinarily sharp melting
profiles. Recognition and application of this cooperative be-
both the dendrimer- and nanoparticle-based strategies have led
to novel signal amplification strategies, which have improved
the sensitivity of assays based upon them by many orders of
magnitude when compared with molecular fluorophore probe
technologies.4,8,9
Recently, we reported a methodology for chemically attaching
multiple DNA strands to well-defined ROMP-based (ROMP
) ring-opening metathesis polymerization)11 organic polymers
and block copolymers of norbornene derivatives (Scheme 1).12
In this approach, the metathesis catalyst Cl2(PPh3)2RudCHPh
(1) was used to polymerize a novel norbornenyl monomer
containing a diphenylacetylene-benzyl alcohol spacer (2). Post-
polymerization modification of the resulting polymer with
2-cyanoethyl tetraisopropylphosphorodiamidite (3) led to the
isolation of a polymer that can be readily coupled to DNA using
standard solid-phase synthesis techniques.13 Significantly, ROMP-
based polymers modified with complementary oligonucleotide
strands reversibly hybridized to form aggregates with very sharp
“melting” characteristics similar to those observed in the
oligonucleotide-modified gold nanoparticle systems.1,10
Due to the living nature of ROMP catalyzed by 1 and its
functional group tolerance, a wide variety of functional groups
can be incorporated into ROMP-based polymer-DNA hybrid
structures in a highly tailorable and controlled manner, allowing
one to encode such structures with identification “tags”. These
hybrids can then be used to build materials with recognition
(1) Mirkin, C. A.; Letsinger, R. L.; Mucic, R. C.; Storhoff, J. J. Nature 1996,
382, 607-609.
(2) Shchepinov, M. S.; Dalova, A. U.; Bridgman, A. J.; Southern, E. M. Nucleic
Acids Res. 1997, 25, 4447-4454.
(3) Shchepinov, M. S.; Mir, K. U.; Elder, J. K.; Frank-Kamenetskii, M. D.;
Southern, E. M. Nucleic Acids Res. 1999, 27, 3035-3041.
(4) Collins, M. L.; Irvine, B.; Tyner, D.; Fine, E.; Zayati, C.; Chang, C.; Horn,
T.; Ahle, D.; Detmer, J.; Shen, L.-P.; Kolberg, J.; Bushnell, S.; Urdea, M.
S.; Ho, D. D. Nucleic Acids Res. 1997, 25, 2979-2984.
(5) Elghanian, R.; Storhoff, J. J.; Mucic, R. C.; Letsinger, R. L.; Mirkin, C. A.
Science 1997, 277, 1078-1081.
(6) Storhoff, J. J.; Elghanian, R.; Mucic, R. C.; Mirkin, C. A.; Letsinger, R. L.
J. Am. Chem. Soc. 1998, 120, 1959-1964.
(7) Taton, T. A.; Mirkin, C. A.; Letsinger, R. L. Science 2000, 289, 1757-
1760.
(8) Taton, T. A.; Lu, G.; Mirkin, C. A. J. Am. Chem. Soc. 2001, 123, 5164-
(11) Trnka, T. M.; Grubbs, R. H. Acc. Chem. Res. 2001, 34, 18-29.
(12) Watson, K. J.; Park, S.-J.; Nguyen, S. T.; Mirkin, C. A. J. Am. Chem. Soc.
2001, 123, 5592-5593.
(13) Brown, T.; Brown, D. J. S. in Oligonucleotides and Analogues; Eckstein,
F., Ed.; Oxford University Press: New York, 1991.
5165.
(9) Park, S.-J.; Taton, T. A.; Mirkin, C. A. Science 2002, 295, 1503-1506.
(10) Demers, L. M.; Mirkin, C. A.; Mucic, R. C.; Reynolds, R. A.; Letsinger,
R. L.; Elghanian, R.; Viswanadham, G. Anal. Chem. 2000, 72, 5535-5541.
9
1170
J. AM. CHEM. SOC. 2005, 127, 1170-1178
10.1021/ja046931i CCC: $30.25 © 2005 American Chemical Society

Polymer−DNA Hybrids as Electrochemical Probes for DNA
A R T I C L E S
Scheme 1. Synthesis of Polymer-DNA Hybrids
capabilities and other physical properties and as probes in novel
detection systems. Herein, we demonstrate the generality and
the utility of this hybrid polymer-DNA approach in a scheme
for the electrochemical detection of DNA. Specifically, we have
synthesized both diblock and triblock copolymers of oligo-
nucleotides and ferrocene derivatives 4 and 5 and characterized
their electrochemical and DNA hybridization properties. The
diblock copolymers were then used successfully in the electro-
chemical detection14-19 of pM-levels of oligonucleotides where
the presence of an electrochemical signal after DNA hybridiza-
tion signifies the presence of a target strand (vide infra). With
two different diblock copolymer-DNA probes, small differ-
ences in DNA sequence mismatches (single-base pair) can be
reliably distinguished. When using a triblock copolymer-DNA
hybrid with two ferrocene blocks possessing widely spaced
redox potentials, the number of polymer-DNA electrochemical
signaling probes can be significantly increased by tailoring the
peak positions and the intensity ratios of the ferrocene blocks,
allowing one to detect multiple targets simultaneously and
accurately.
Although most commercially available detection systems use
fluorescence spectroscopy, numerous electrochemically based
sensing methods have been reported in recent years. The interest
in electrochemical detection stems from the simplicity of the
required voltammetric instrumentation as well as its potential
for miniaturization and use in point-of-care assays.18,19 The
electrochemical component of these detection schemes varies
widely; some systems rely on DNA intercalators, the direct
oxidation of nucleotides, or signal amplification based on
enzymatic processes. Other approaches have employed oligo-
nucleotides covalently modified with redox-active small mol-
ecules20 in strategies similar to fluorescent-based molecular
probe assays.16,17,21-28 When compared with these electrochemi-
cally active molecular probes, the block copolymers reported
herein exhibit several desirable DNA hybridization properties
such as larger binding constants (manifested in higher Tm) and
sharper melting transitions which lead to better target dif-
ferentiation capabilities (vide infra) and higher sensitivities. As
a basis for comparison, we developed a molecular probe that is
structurally similar to a repeating unit of our polymer probe
and acquired a series of melting profiles to further evaluate and
substantiate the advantages of our ROMP polymer-based
systems.
Results and Discussion
Syntheses of Diblock Copolymer-DNA Hybrids. Block
copolymer hybrids (Hybrid I-Hybrid IV) of DNA and redox-
active molecules (4, 5) were synthesized by attaching oligo-
nucleotides to the preformed ROMP polymers (Scheme 1). Each
block copolymer was synthesized by reacting the DNA-
modifiable monomer 2 with Grubbs catalyst 1 until all of the
monomer had been consumed and then polymerizing either
monomer 4 or 5, followed by termination with ethyl vinyl ether,
to yield diblock copolymers poly2-poly4 and poly2-poly5,
respectively. These polymer precursors were then coupled to
the 5′ end of oligonucleotides to yield Hybrid I-Hybrid IV
(20) Immoos, C. E.; Lee, S. J.; Grinstaff, M. W. J. Am. Chem. Soc. 2004, 126,
10814-10815.
(21) Brazill, S. A.; Kim, P. H.; Kuhr, W. G. Anal. Chem. 2001, 73, 4882-
4890.
(22) Brazill, S. A.; Kuhr, W. G. Anal. Chem. 2002, 74, 3421-3428.
(23) Brazill, S.; Hebert, N. E.; Kuhr, W. G. Electrophoresis 2003, 24, 2749-
2757.
(14) Kelley, S. O.; Boon, E. M.; Barton, J. K.; Jackson, N. M.; Hill, M. G.
Nucleic Acids Res. 1999, 27, 4830-4837.
(15) Napier, M. E.; Loomis, C. R.; Sistare, M. F.; Kim, J.; Eckhardt, A. E.;
Thorp, H. H. Bioconjugate Chem. 1997, 8, 906-913.
(16) Ihara, T.; Maruo, Y.; Takenaka, S.; Takagi, M. Nucleic Acids Res. 1996,
24, 4273-4280.
(24) Yu, C. J.; Wang, H.; Wan, Y.; Yowanto, H.; Kim, J. C.; Donilon, L. H.;
Tao, C.; Strong, M.; Chong, Y. J. Org. Chem. 2001, 66, 2937-2942.
(25) Cosnier, S.; Gondran, C.; Dueymes, C.; Simon, P.; Fontecave, M.; Decout,
J.-L. Chem. Commun. 2004, 1624-1625.
(17) Yu, C. J.; Wan, Y.; Yowanto, H.; Li, J.; Tao, C. L.; James, M. D.; Tan, C.
L.; Blackburn, G. F.; Meade, T. J. J. Am. Chem. Soc. 2001, 123, 11155-
11161.
(18) Wang, J. Anal. Chim. Acta 2002, 469, 63-71.
(19) Drummond, T. G.; Hill, M. G.; Barton, J. K. Nat. Biotechnol. 2003, 21,
1192-1199.
(26) Kim, K.; Yang, H.; Park, S. H.; Lee, D.-S.; Kim, S.-J.; Lim, Y. T.; Kim,
Y. T. Chem. Commun. 2004, 1466-1467.
(27) King, G. C.; di Giusto, D. A.; Wlassoff, W. A.; Giesebrecht, S.; Flening,
E.; Tyrelle, G. D. Hum. Mutat. 2004, 23, 420-425.
(28) Mucic, R. C.; Herrlein, M. K.; Mirkin, C. A.; Letsinger, R. L. J. Chem.
Soc., Chem. Commun. 1996, 555-557.
9
J. AM. CHEM. SOC. VOL. 127, NO. 4, 2005 1171

A R T I C L E S
Gibbs et al.
Table 1. DNA Sequences of Hybrids
polymer precursor (ratio)
sequence
I
II
III
IV
V
poly2-poly4 (17:10)
poly2-poly4 (17:10)
poly2-poly5 (17:10)
poly2-poly5 (17:10)
poly2-poly4-poly5 (17:10:5) 5′-T10 ATC CTT ATC AAT ATT-3′
poly2-poly4-poly5 (17:5:5) 5′-T10 AAT ATT GAT AAG GAT-3′
5′-T3 ATC CTT ATC AAT ATT-3′
5′-T3 AAT ATT GAT AAG GAT-3′
5′-T3 ATC CTT ATC AAT ATT-3′
5′-T3 AAT ATT GAT AAG GAT-3′
VI
VII poly2-poly5-poly4 (17:10:5) 5′-T10 TAA CAA TAA TCC CTC-3′
VIII poly2-poly5 (17:10) 5′-T3 ATC CTT AGC AAT ATT-3′
Figure 2. (A) Cyclic voltammograms of Hybrid I (dotted line) and Hybrid
III (solid line) in CH2Cl2 with [nBu4N]PF6 as electrolyte. The E1/2 values
are 30 mV and 350 mV (versus Fc/Fc+) for Hybrid I and Hybrid III,
respectively, as expected from the redox potentials of 4 and 5. (B)
Absorption spectra of a solution containing complementary polymer-DNA
hybrids (Hybrid I:Hybrid II) above and below the DNA melting
temperature, indicating that the aggregation is driven by DNA hybridization.
The mixture of Hybrid III and Hybrid IV behaves the same way and
shows similar temperature-dependent UV-vis absorption spectra.
derivatives with a variety of electron-donating or -withdrawing
substituents. To demonstrate this capability, we chose two
different norbornenyl-modified ferrocene derivatives, 4 and 5,
which exhibit E1/2s separated by 320 mV. Electrochemical
measurements were carried out on the hybrids by casting thin
films of these materials on Au electrodes (see Supporting
Information). Cyclic voltammetry of the resulting films in CH2-
Cl2 with [nBu4N]PF6 as the electrolyte revealed stable and
reversible waves associated with oxidation and reduction of the
ferrocenyl blocks of the hybrid materials. The films are stable
in the organic environment; multiple scans generated consistent
results. As shown in Figure 2A, the E1/2 value for Hybrid I
was found to be 30 mV (versus Fc/Fc+), while the E1/2 of
Hybrid III was 350 mV (versus Fc/Fc+) (the formal potential
value of Fc/Fc+ in CH2Cl2 with [nBu4N]PF6 is 0.46 V versus
SCE).30 These values are almost identical with the redox
potentials observed for the corresponding monomers 4 and 5.
Figure 1. (A) UV-vis absorption spectrum of Hybrid I in water. (B)
Ion-exchange HPLC chromatogram of purified Hybrid I monitored by UV-
vis absorption at 260 nm (λmax for DNA) and 310 nm (λmax for diphenyl-
acetylene). Both chromatograms show a single peak at the same retention
time, indicating that DNA is indeed coupled to the polymer backbone.
(Scheme 1). DNA sequences and the length of each polymer
block are given in Table 1.
Due to the hydrophilicity of the DNA block, the resulting
polymer-DNA hybrids were soluble in aqueous media. The
number of DNA strands attached to each hybrid was estimated
from the UV-vis absorption spectra of the polymer conjugates
in water (Figure 1A).29 Ion-exchange HPLC was performed on
purified Hybrid I (Figure 1B). One major peak at 25 min was
observed at both 260 nm (λmax of DNA) and 310 nm (λmax of
the diphenylacetylene spacer in the HPLC eluent), which further
demonstrates that the oligonucleotides are indeed coupled to
the polymer backbone. Preliminary results from capillary
electrophoresis experiments indicate that our polymer-DNA
hybrids are monodisperse.
DNA Recognition Properties of Diblock Copolymer-DNA
Hybrids. As outlined in Table 1, complementary polymer-
DNA hybrid structures (Hybrid I and Hybrid II from poly2-
poly4 and Hybrid III and Hybrid IV from poly2-poly5) were
prepared for DNA recognition studies. When these comple-
mentary hybrids were mixed together in equal amounts (60 µM,
20 µL) in a PBS buffer (0.3 M NaCl, 10 mM phosphate, pH
7), extended polymer aggregates formed and precipitated from
the solution in a few seconds. This observation demonstrates
that the organic polymer portions of the hybrids do not inhibit
the recognition properties of the DNA side chains. The UV-
vis spectrum of the aggregate solution showed a significantly
The redox characteristics of the electrochemically active
portion of the polymer can be tailored by using ferrocene
(29) Based on the molar absorptivity of DNA at 287 nm and the diphenylacety-
lene spacer at 287 and 307 nm, we determined that there are, on average,
five DNA strands per single polymer chain, which translates to ∼30%
occupation of the 17 available DNA coupling sites on the poly2 block.
This value was calculated by first determining the polymer concentration
based on the absorbance at 307 nm, which is only due to the polymer
component and not the DNA. Knowing the polymer concentration and its
molar absorptivity at 287 nm allowed us to subtract the polymers’
absorbance at 287 nm from the total absorbance. The DNA concentration
was determined from the remaining absorbance at 287 nm and the molar
absorptivity value at 287 for the DNA sequences. The DNA concentration
divided by the polymer concentration gives the average value of DNA per
polymer.
(30) Connelly, N. G.; Geiger, W. E. Chem. ReV. 1996, 96, 877-910.
9
1172 J. AM. CHEM. SOC. VOL. 127, NO. 4, 2005

Polymer−DNA Hybrids as Electrochemical Probes for DNA
A R T I C L E S
Table 2. DNA Sequences Used in Polymer vs Molecular Probe
Comparison
precursor
sequence
Fc-DNA I
Fc-DNA II
polyDNA I
polyDNA II
DNA I
6
6
5′-T3 ATC CTT ATC AAT ATT-3′
5′-T3 AAT ATT GAT AAG GAT-3′
5′-T3 ATC CTT ATC AAT ATT-3′
5′-T3 AAT ATT GAT AAG GAT-3′
5′-T3 ATC CTT ATC AAT ATT-3′
5′-T3 AAT ATT GAT AAG GAT-3′
poly2 (17-mer)
poly2 (17-mer)
unmodified
DNA II
unmodified
Based on theoretical models, the local high density of DNA
strands on our polymer backbone (as compared to free DNA in
solution) appears to be crucial for inducing sharp thermal
denaturation transitions in the resulting polymer-DNA hybrid.31
As the oligonucleotide-modified gold nanoparticle probes offer
high selectivity due to their sharp melting profiles,7-9 the unique
melting properties of the block copolymer-DNA hybrids point
toward their potential use in high selectivity applications such
as single-nucleotide polymorphism (SNP) detection.
Thermal Denaturation Studies of Polymer and Molecular
Probes. We hypothesize that the unique melting profiles of our
polymer probes, as well as that of DNA-modified gold nano-
particles, result from the presence of multiple DNA linkages in
close proximity. Thus, we predicted that molecular probes would
not exhibit such sharp melting transitions.
Figure 3. Thermal denaturation curves of networks formed from (A)
Hybrids I:II and (B) Hybrid III:IV. Insets: the first derivatives of the
thermal denaturation curves (fwhm ) 2 °C). Two different block copolymer
aggregates show similar melting transitions, indicating that the type of
ferrocene would not affect the melting properties of DNA. Note that Hybrids
I and III have the same DNA sequences, as do Hybrids II and IV (Table
1). Compared to the thermal denaturation curve for duplex DNA formed
from unmodified oligonucleotides with the same sequence (C), the polymer
aggregates show a much higher melting temperature and a sharper melting
transition.
To test this prediction, we synthesized 5′-ferrocenyl labeled
oligonucleotides (Fc-DNA I and Fc-DNA II) and homopoly-
mer-DNA hybrids (polyDNA I and polyDNA II) (Table 2).
These polymer hybrids were synthesized by coupling oligo-
nucleotides to the polymer precursor, poly2, following a
previous method.12 The 5′-ferrocenyl labeled oligonucleotides
(Fc-DNA I and Fc-DNA II) consist of a diphenylacetylene
linker in addition to the ferrocenyl moiety (Figure 4A). The
molecular probe was synthesized by converting one of our
ferrocenyl monomer precursors12 into a ferrocenyl alcohol with
a three-carbon spacer. The (3-hydroxypropyl)ferrocene com-
pound was then modified with a diphenylacetylene alcohol
group following a synthetic strategy similar to that used for the
synthesis of the norbornenyl monomer 2 (Scheme 2).12 The
resulting ferrocene derivative 11 was treated with chlorophos-
phoramidite and coupled to DNA in the same manner as the
homopolymers. The diphenylacetylene functionality increased
the retention time of the oligonucleotide sequence significantly
during HPLC purification and absorbed strongly at 310 nm,
allowing for easy purification and characterization of the
molecular probe.
The similarity in the hydrophobic structure of the ferrocenyl
molecular probe and the DNA side chains on the polymer allows
for an accurate comparison of their melting properties. Primarily,
the two types of probes differ in that the latter possesses multiple
DNA strands in close proximity. Because any contribution of
hydrophobicity32 to the melting temperature should be ap-
proximately the same for both systems, comparing their thermal
denaturation profiles should delineate whether the polymeric
nature of the probe has a significant effect on DNA melting
behavior.
reduced signal at 260 nm and increased intensity at higher
wavelengths due to scattering from the polymer aggregates
(Figure 2B). Upon heating the aggregate solution to 70 °C, a
temperature that is above the polymer-bound DNA’s melting
temperature (Tm ) 62 °C), polymer-DNA hybrids can be
redispersed in solution as evidenced by UV-vis spectroscopy
(Figure 2B). The temperature dependence of aggregate formation
indicates that the aggregates are indeed formed by sequence-
specific DNA hybridization rather than nonspecific, hydrophobic
interactions.
The thermal denaturation curves of these aggregates were
obtained by monitoring their absorbance at 260 nm as a function
of temperature (Figure 3). The block copolymer conjugates show
unique DNA hybridization properties that differ from unmodi-
fied DNA. First, the aggregates formed from block copolymer-
DNA (Figure 3A,B) show higher thermal stabilities (13 °C
higher in melting temperatures) than unmodified duplex DNA
(Figure 3C) of the same sequence at comparable concentrations.
Similar high thermal stabilities have been observed in DNA
dendrimers3 and can be attributed to the presence of multiple
DNA linkages which results in cooperative binding. This
attribute suggests that the block copolymer-DNA hybrids might
be useful as probes in DNA detection systems. Second, the
aggregates formed from the hybrids show unusually sharp
melting curves; the full widths at half-maximum (fwhm) for
the first derivatives of the melting curves were ∼2 °C (Figure
3A and B, inset). This is an unusual characteristic, as comparable
“soft” nanostructure probes made of dendrimers apparently do
not exhibit such a narrow melting transition.3 The transitions
observed for the novel probes reported herein are highly
reminiscent of the sharp melting transition observed in the
oligonucleotide-modified gold nanoparticle system, which has
proven to be very useful in DNA diagnostics.1,5,6
Thermal denaturation curves were acquired using UV-vis
spectroscopy as previously described. Two sets of hybridization
(31) Jin, R.; Wu, G.; Li, Z.; Mirkin, C. A.; Schatz, G. C. J. Am. Chem. Soc.
2003, 125, 1643-1654.
(32) Guckian, K. M.; Schweitzer, B. A.; Ren, R. X. F.; Sheils, C. J.; Tahmassebi,
D. C.; Kool, E. T. J. Am. Chem. Soc. 2000, 122, 2213-2222.
9
J. AM. CHEM. SOC. VOL. 127, NO. 4, 2005 1173

A R T I C L E S
Gibbs et al.
fwhma
Table 3. Thermal Stability Values
melting
temp ( C)
<∆HVH>
(kcal/mol)
hybridization pair
°
(°C)
DNA I and II
49.0
50.3
46.5
51.0
61.5
99.7
96.3
101
95.2
452
9.4
DNA I and Fc-DNA II
DNA I and polyDNAb II
Fc-DNA I and Fc-DNA II
polyDNAb I and polyDNAb II
10.4
9.2
8.3
2.6
a fwhm: Full width at half-maximum of the first derivative (d(R)/d(1/
T) vs T). b polyDNA was prepared from poly2 (17-mer), and the average
number of DNA attached to the polymer was determined to be five.
hybridization experiment, the polyDNA I:polyDNA II hybrid-
ization pair shows a greatly increased Tm (Tm ) 61.5 °C, ∆Tm
) 10 °C) and sharpened melting transition (fwhm ) 2.6 °C)
while the molecular probe-molecular probe (Fc-DNA I:Fc-DNA
II) melting (Tm ) 51 °C, fwhm ) 8.3 °C) remains similar to
that of unmodified DNA (Tm ) 49 °C, fwhm ) 9.4 °C) (Figure
4C). These results suggest that hydrophobicity does not
contribute significantly to the melting properties of the attached
oligonucleotides. Further, our experiments indicate that multiple
DNA linkages do have a significant effect on DNA melting,
but only when both strands of the hybridized duplex are bound
to a structure containing multiple DNA strands. This conclusion
could have large ramifications for probe optimization in chip-
based assays, if one assumes that a DNA-modified surface acts
as such a multiple-linked structure.
To quantify the difference in melting behavior, we performed
simple thermodynamic calculations for each melting profile. Our
calculations were a modification of two literature procedures33,34
for calculating the van’t Hoff mean enthalpic change, <∆HVH>,
which assumes a two-state transition model. There are limita-
tions to this model, particularly in that it ignores the contribu-
tions of cooperativity (a multistate transition).33,35 In addition,
it does not account for the polymer-DNA duplexes forming
larger aggregate structures. Nevertheless, the van’t Hoff en-
thalpic calculation is a useful first-order approach for comparing
the enthalpic difference involved in the dehybridization of probe-
hybrid materials.
Figure 4. (A) Structure of Fc-DNA. (B) Thermal denaturation curves of
polyDNA II, Fc-DNA II, and an unmodified oligonucleotide (DNA II)
hybridized with a complementary, unmodified oligonucleotide (DNA I).
(C) Thermal denaturation curves of polyDNA, Fc-DNA, and an unmodified
oligonucleotide hybridized with their complementary analogue.
experiments were performed on each type of probe (polyDNA
and Fc-DNA, Table 2). In the first set of experiments, each
type of probe was hybridized with unmodified complementary
oligonucleotide strands (DNA I). In the second set of experi-
ments, each probe type was hybridized to its respective
complementary probe, i.e., Fc-DNA with its complementary Fc-
DNA and polyDNA with its complementary polyDNA. Inter-
estingly, the melting profiles in the first data set are almost
identical to one another as well as to the melting curve of the
unmodified DNA duplex (Figure 4B). In contrast, in the second
The values for <∆HVH> are listed in Table 3. With the
exception of the polyDNA I:polyDNA II hybridization pair,
all other hybridization pairs have <∆HVH> values of ap-
Scheme 2. Synthesis of the Molecular Probe Precursor
9
1174 J. AM. CHEM. SOC. VOL. 127, NO. 4, 2005

Polymer−DNA Hybrids as Electrochemical Probes for DNA
A R T I C L E S
proximately 100 kcal/mol. Within experimental error it is
difficult to deconvolute the effect of the polymer and molecular
probes on the resulting duplexes where the complementary
partner is either an unmodified DNA strand or a molecular
probe. However, the polyDNA:polyDNA hybridization pair has
a <∆HVH> value that is 4.5 times greater than any of the other
hybridization pairs. This significant increase in thermal stability
may be crucial to the success of assays that use genomic DNA
with complex secondary structures. In addition, the melting
ranges, listed as fwhm in Table 3, suggest that polymer probes
should be able to discriminate SNPs better than molecular probes
under stringent thermal conditions (or salt concentrations).7-9
Thus, a comparison of melting profiles indicates that polymer
probes such as the ones described herein could be better
equipped for use in DNA detection than their molecular probe
counterparts.
Syntheses of Triblock Copolymer-DNA Hybrids. Oligo-
nucleotides have been modified with several redox-active
compounds.20,36-39 For example, Mucic et al. have developed
5′-ferrocenyl modified oligonucleotides as molecular probes for
use in three-strand electrochemical detection assays.28 Deriva-
tizing ferrocene for covalent incorporation into oligonucleotides
has also led to successful detection systems such as those
reported by Yu et al.17,24 and Kuhr et al.21-23 One advantage of
our polymer-DNA hybrid approach is the polyvalent cooper-
ativity of the pendant DNA strands, which provide more specific
discrimination of single-pair mismatches than analogous mo-
lecular probe systems (vide supra). The signaling amplification
capabilities are also better for polymer-DNA hybrids, because
polymers with larger blocks of labeling molecules can be
synthesized, allowing for magnified detection signals. Addition-
ally, the functional group tolerance and synthetic control that
are native to ROMP polymerization enable us to develop a
number of distinctly labeled probes quite easily. For example,
by varying the molar ratio of two redox-active groups on the
same polymer-DNA detection probe, numerous unique signa-
tures based on not only the positions of the redox signals but
also the amplitude ratios of these signals are possible. Such a
system would have many advantages over molecular probe
systems, particularly in multiplexing assays, because only a few
electrochemical tags would be required to generate numerous
barcodelike signatures, detectable by simple data-readout ca-
pabilities, using both identity and quantitative dimensions.
Based on our block copolymer strategy, in theory any DNA
strand in a hybrid can be labeled with a unique electrochemical
polymer tag. The detection of multiple DNA strands in a mixture
can then be carried out by monitoring the electrochemical signal
and referencing it to a particular probe associated with a target
DNA sequence. However, the practical requirement that the
different redox signals arising from the electrochemically active
blocks be reasonably spaced apart to achieve complete signal
discrimination only allows for the incorporation of approxi-
mately four different indicators in a detection system employing
Figure 5. Schematic illustration of the triblock copolymer labeling strategy.
the diblock copolymer-DNA hybrid strategy. To increase the
number of potential indicators, we synthesized triblock copoly-
mers containing two different ferrocenyl derivatives, 4 and 5.
By adjusting the relative lengths of the two redox-active blocks
one can generate several different tags corresponding to distinct
combinations of the two electrochemically active monomers
(Figure 5).40 Triblock copolymer precursors were synthesized
by polymerizing the DNA-modifiable diphenylacetylene mono-
mer 2, then adding ferrocene monomer 4, and finally dibromo-
ferrocene monomer 5 to the growing polymer chain. Each
monomer was added to the polymerization mixture only after
the previous type of monomer had been consumed. For proof-
of-concept experiments, polymer precursors with approximately
2:1, 1:1, and 1:2 ratios of 4 and 5 were synthesized. Gel-
permeation chromatographic (GPC) data for these polymers all
showed a single peak, characteristic of a monodisperse sample,
indicating that all three monomers are incorporated into a single
polymer chain. The poly2-poly4-poly5 precursors were then
coupled to DNA to generate triblock copolymer-DNA Hybrids
V-VII (Table 1). In the design of triblock copolymers, we
employed a T10-spacer instead of the T3-spacer used for diblock
copolymers to improve the water solubility of the final hybrids.
As expected, triblock copolymer-DNA hybrids behave similarly
to diblock copolymer hybrids as indicated by their UV-vis
spectra and melting profiles (Figure 6).
Cyclic voltammograms of the triblock copolymer-DNA
hybrids exhibit two distinct redox peaks at 30 and 350 mV
(versus Fc/Fc+) (Figure 7). Ratios of the peak current were found
to be ∼ 2.7:1, 1:1, and 1:2.7 for Hybrid V, Hybrid VI, and
Hybrid VII, respectively. Even though the peak current ratios
are slightly different from the monomer ratios that we initially
used (2:1, 1:1, and 1:2), these peak ratios and positions are
consistent with those observed for the polymer precursors prior
to DNA coupling. In fact, the cyclic voltammograms of the
(33) Marky, L. A.; Breslauer, K. J. Biopolymers 1987, 26, 1601-1620.
(34) Borer, P. N.; Dengler, B.; Tinoco, I., Jr.; Uhlenbeck, O. C. J. Mol. Biol.
1974, 86, 843-53.
(35) Marky, L. A.; Canuel, L.; Jones, R. A.; Breslauer, K. J. Biophys. Chem.
1981, 13, 141-149.
(36) Beilstein, A. E.; Grinstaff, M. W. J. Organomet. Chem. 2001, 637-639,
398-406.
(40) Consider the theoretical case of triblock copolymers with a maximum
number of 5 repeating units per each electrochemically active block, distinct
ratios which can be observed electrochemically would consist of the
following: (0:5), (1:3), (1:4), (1:5), (2:3), (2:4), (2:5), (3:1), (3:2), (3:4),
(3:5), (4:1), (4:2), (4:3), (4:5), (5:0), (5:1), (5:2), (5:3), (5:4), (5:5).
(37) Hurley, D. J.; Tor, Y. J. Am. Chem. Soc. 1998, 120, 2194-2195.
(38) Weizman, H.; Tor, Y. J. Am. Chem. Soc. 2002, 124, 1568-1569.
(39) Hurley, D. J.; Tor, Y. J. Am. Chem. Soc. 2002, 124, 3749-3762.
9
J. AM. CHEM. SOC. VOL. 127, NO. 4, 2005 1175

A R T I C L E S
Gibbs et al.
Scheme 3. DNA Detection Based on Polymer-DNA Probes
solution of a (0.1 mM) to a part of the gold electrode (0.5 × 1
cm2) for 16 h. Then, the electrode was washed with water and
dipped in a 1 mM aqueous solution of mercaptohexanol for 5
min. This passivation procedure inhibits nonspecific binding of
hybrid molecules to the gold electrodes. Finally, the electrode
was washed with copious amounts of water and used for DNA
detection studies.
To evaluate the performance of our polymer-DNA probes,
synthetic target DNA (a′b′, 10 nM) and Hybrid I (1.2 nM) were
cohybridized to a-modified electrodes in PBS solution for 4 h
at room temperature (Scheme 3). As a control experiment,
another a-modified electrode was treated with Hybrid I without
target DNA under identical conditions. After washing the
electrodes with PBS solution, alternating current (AC) volta-
mmograms were acquired (Figure 8). The electrode treated with
complementary target shows the desired redox signal due to 4
while the control sample generates no significant signal.
To test the sensitivity of the method described in Scheme 3,
we have attempted to detect target at concentrations ranging
from 10 nM to 1 pM. The a-modified electrodes were treated
with a series of target solutions (10 nM-1 pM) and polymer-
DNA probe Hybrid I (2.4 nM) for 6 h. Then, AC voltammo-
grams were obtained for each sample. As presented in Figure
8, we could easily detect target concentrations as low as 100
pM based on this detection method, which is an order of
magnitude more sensitive than that reported by Yu et al. for
their electrochemical molecular probe system.17 When the target
concentration was less than 100 pM, the signal level was
indistinguishable from the background and was comparable to
the signal intensity of the control electrode when only treated
with Hybrid I (Figure 8, inset).
Figure 6. (A) Absorption spectra of triblock copolymer aggregates formed
from Hybrids V and VI above and below the DNA melting temperature.
(B) A thermal denaturation curve for the aggregates (Inset: first derivative).
Figure 7. Cyclic voltammograms in CH2Cl2 with [nBu4N]PF6 as electrolyte
of (A) Hybrid V, a triblock copolymer-DNA with a 10:5 ratio of poly4:
poly5, (B) polymer precursor of Hybrid V, (C) Hybrid VI, a triblock
copolymer-DNA with a 5:5 ratio of poly4:poly5, and (D) Hybrid VII, a
triblock copolymer-DNA with a 5:10 ratio of poly4:poly5. These triblock
copolymer-DNA hybrids exhibit two distinct redox peaks with expected
E1/2 and peak intensity ratios. The detailed shape of the cyclic voltammo-
grams caused by the diffusion-limited current and the charging current is
dependent upon polymer film conditions. The cyclic voltammogram of the
polymer precursor was obtained from a solution sample rather than a film
cast on an electrode surface.
polymer-DNA hybrids and their polymer precursors are almost
superimposable (Figure 7A,B). These results not only demon-
strate the versatility of our synthetic methodology but also
suggest the possibility of using the electrochemical signals as a
type of barcode to tag many DNA targets present in a single
sample solution.
DNA Detection Using Diblock Copolymer-DNA Hybrids.
The block copolymer conjugates were utilized as DNA probes
in a three-component sandwich-type electrochemical detection
strategy (Scheme 3). In a typical experiment, a disulfide-
modified oligonucleotide, a, was immobilized on a freshly
prepared Au electrode (0.5 × 2 cm2) by applying 0.3 M PBS
Figure 8. AC voltammograms in CH2Cl2 with [nBu4N]PF6 as electrolyte
for electrodes treated with different target concentrations. Inset: target
concentration vs peak current. The horizontal dotted line is the peak current
of the sample without the target DNA.
9
1176 J. AM. CHEM. SOC. VOL. 127, NO. 4, 2005

Polymer−DNA Hybrids as Electrochemical Probes for DNA
A R T I C L E S
Figure 9. Signal amplification using polymer-DNA probes. (A) A scheme
illustrating the addition of multiple layers to the electrode surface controlled
by the recognition properties of DNA. (B) AC voltammograms in CH2Cl2
with [nBu4N]PF6 as electrolyte of multilayer block copolymer-DNA hybrids
assembled on a gold substrate.
The number of ferrocenyl groups bound to the gold electrodes
was estimated to be 1.1 × 1013, 3.4 × 1012, and 5.2 × 1011 in
the presence of 9.0 × 1012, 9.0 × 1011, and 9.0 × 1010 target
strands (total sample volume ) 1500 µL, area ) 0.5 cm2),
respectively, using eq 141 where Iavg is the peak current, n is
the number of electrons, T is the temperature, Eac is the peak
amplitude, f is the frequency, F is the Faraday constant, R is
the gas constant, and Ntot is the total number of moles of redox
active species corresponding to the peak.
Figure 10. (A) Sequences of the target and probe DNAs used for dual-
channel detection of a single-base mismatch. Note that Hybrid I and Hybrid
VIII were prepared from 4 and 5, respectively. (B) AC voltammograms in
CH
2Cl2 with [nBu4N]PF6 as electrolyte for electrodes treated with either
Target I or Target II in the presence of both detection probes. Only one
major peak is shown at the expected potential for each case.
Iavg(E0) ) 2nfNtot sin h(nFEac/RT)/cos h(nFEac/RT) (1)
strongly on the number of ferrocenyl groups present in each of
the hybrids used for building the successive layers.
It is noteworthy that the estimated number of ferrocenyl groups
can be higher than the number of target DNA molecules present
in solution. The higher ratio of ferrocenyl groups to target DNA
is possible because each single polymer-DNA probe possesses
about 10 ferrocenyl groups, on average. Thus the hybridization
of one target DNA strand can, in principle, result in the
attachment of 10 ferrocenyl groups to the electrode surface,
thereby providing a form of signal amplification. Hence,
detection methods based on these novel polymer-DNA probes
can be more sensitive than methods utilizing oligonucleotides
modified with only single ferrocenyl moieties, and the sensitivity
of a detection system based on this approach should scale to
some extent with the size of the block of redox groups.
Signal Amplification via Multilayering. Since our block
copolymer-DNA hybrids possess multiple oligonucleotide side
chains, in theory extra DNA strands are still available for further
assembly of block copolymers after target hybridization if this
latter process is less than 100% efficient or if the target
concentration is less than polymer-bound DNA. To address the
possibility of using these extra oligonucleotides for signal
enhancement, a second and third layer of poly4-containing block
copolymer-DNA hybrids (Hybrid II and Hybrid I, respec-
tively) were assembled successively onto the electrode surface
after the initial hybridization of target DNA and Hybrid I
(Figure 9A and B). As the number of layers increased, the peak
currents increased, showing that signal amplification via mul-
tilayering is indeed possible (Figure 9B). In these experiments,
the sensitivity is increased proportionally to the number of
layers. Thus, the magnitude of amplification will depend
Detection of Single-Base Mismatches. In conventional DNA
chip-based detection systems, two different color fluorophores
are often used in each experiment for more accurate detection
and to provide signal ratioing capabilities. In the DNA detection
scheme reported herein, two types of hybrids with different
ferrocenyl moieties can be used in a manner similar to the
electrochemically based dual signaling procedure introduced by
Yu et al.17 The almost identical melting profiles of Hybrid
I:Hybrid II and Hybrid III:Hybrid IV indicate that the nature
of the ferrocenyl group does not significantly affect the melting
properties of DNA (Figure 3A and B). This observation suggests
that two different polymer-DNA probes synthesized from 4
and 5 can be utilized for the dual-channel detection of single-
base mismatches. To evaluate the feasibility of this idea, two
synthetic target DNA strands (Target I and Target II in Figure
10A) were prepared. Target I and II were designed to have
only a single base difference at the position indicated by the
arrow in Figure 10A. Hybrid I was synthesized as a signaling
probe for Target I, and Hybrid VIII was synthesized as a
signaling probe for Target II. In these experiments, we prepared
two sample solutions (PBS, 1500 µL) containing (1) Hybrid I,
Hybrid VIII, and Target I and (2) Hybrid I, Hybrid VIII,
and Target II. Concentrations for polymer probes and target
DNA were 1.2 nM and 10 nM, respectively, for both samples.
Then, gold electrodes modified with capture DNA strands were
immersed in the sample solutions and kept at 43 °C. After 4 h
of incubation, the electrodes were washed with PBS buffer and
dried with flowing N2. Finally, AC voltammograms were
obtained for each of the electrodes (Figure 10B). In each case,
the redox signals attributable to the correspondingly matched
signaling probes were the major peaks observed. The electrode
(41) Sumner, J. J.; Weber, K. S.; Hockett, L. A.; Creager, S. E. J. Phys. Chem.
B 2000, 104, 7449-7454.
9
J. AM. CHEM. SOC. VOL. 127, NO. 4, 2005 1177

A R T I C L E S
Gibbs et al.
treated with the sample containing Target I produced a major
signal at ∼30 mV (versus Fc/Fc+, Figure 10B), corresponding
to the presence of ferrocenyl Hybrid I on the surface, and a
trace signal at 350 mV, corresponding to the mismatched probe.
The electrode treated with Target II produced a major signal
at ∼350 mV (versus Fc/Fc+, Figure 10B), corresponding to the
presence of dibromoferrocenyl Hybrid VIII on the surface, and
a minor signal at 30 mV, corresponding to the mismatched
probe. When the electrodes were incubated at room temperature,
they show two strong redox peaks rather than a single one. Thus,
selectivity in our strategy can easily be achieved using thermal
stringency conditions. An increased hybridization temperature
destabilizes only the mismatched probes, giving rise to one
major redox peak that corresponds to the probe that is perfectly
complementary to the target. These results demonstrate that the
block copolymer probes can be used for the accurate detection
of single base mismatches.
In summary, we have reported a new approach for preparing
block-copolymer probes for DNA detection. Using this ap-
proach, we have prepared polymer-DNA hybrids with tailorable
and well-controlled redox characteristics. These polymers pos-
sess the anticipated electrochemical properties that were de-
signed into them but also unusual and unanticipated DNA
hybridization properties. In particular, they exhibit stronger
binding enthalpies and sharper melting profiles than the oligo-
nucleotides from which they are made. These properties have
allowed us to develop an electrochemical detection system that
exhibits point-mutation selectivity and better target differentia-
tion capabilities than those observed for single-oligonucleotide
probes. Also, due to their strong binding constants, as reflected
in increased Tm’s, they should compete more effectively for
target than conventional single-oligonucleotide probes. Finally,
although only two types of probes have been evaluated for the
proof-of-concept studies reported herein, in principle, a very
large number of probes can be designed by systematically
changing the length and composition of the redox-active blocks
in the hybrid structures.
Acknowledgment. C.A.M. and S.T.N. acknowledge the
National Science Foundation (Grant Nos. EEC-0118025 and
DMR-0094347) and the AFOSR (Grant No. F49620-01-1-0303)
for the financial support of this research. We thank Professors
Irving M. Klotz and George C. Schatz, Drs. Richard Gurney
and Jung Hyurk Lim, and Messrs. Shad Thaxton and Zhi Li
for helpful discussions. J.M.G. and S.J.P. contribute equally to
this work.
Supporting Information Available: Synthetic procedures and
characterization data for polymer-DNA hybrids. Methods and
data for hybridization and detection experiments. This material
is available free of charge via the Internet at http://pubs.acs.org.
JA046931I
9
1178 J. AM. CHEM. SOC. VOL. 127, NO. 4, 2005