Intraduplex DNA-mediated electrochemistry of covalently
tethered redox-active reporters
Catrina G. Pheeney
and
Jacqueline K. Barton
*
*
Division of Chemistry and Chemical Engineering, California Institute of Technology, Pasadena,
California 91125, USA
Abstract
Intraduplex DNA-mediated reduction is established as a general mechanism for the reduction of
distally bound stacked redox-active species covalently tethered to DNA through flexible alkane
linkages. Methylene Blue (MB), Nile Blue (NB), and Anthraquinone (AQ) were covalently
tethered to DNA with three different covalent linkages. Using these reporters DNA
electrochemistry was shown to be both DNA-mediated and intra-, rather than inter-, duplex.
Significantly, the charge transport pathway occurring through the DNA
π
-stack is established by
using an intervening AC mismatch to break this path. The fact that the DNA-mediated reduction
of MB occurs primarily via intraduplex intercalation is established through varying the proximity
and integrity of the neighboring duplex DNA.
DNA-modified electrodes are extensively used in both the development of next generation
diagnostic sensors (1–6) and the characterization of ground-state DNA-mediated
electrochemistry (7–15). Many different electrochemical reporters have been developed for
DNA-modified electrodes, including: DNA-binding compounds (1, 5), quantum dots (4),
and metallization (15, 16). Two classes of DNA-based devices have since emerged using the
same covalently tethered intercalative compounds but different mechanisms: DNA
conformation (3, 17–22) and DNA-mediated charge transfer (DNA CT) (6–14).
The DNA-mediated reduction of both freely diffusing and covalently tethered redox-active
reporters has long been established on these DNA-modified electrodes (23–25). The strategy
of covalently tethering redox-active reporters to the DNA, as opposed to the use of freely
diffusing reporters, has been adopted to significantly diminish non-specific signals (11, 17).
Redox-active reporters covalently tethered to DNA have been shown to electronically
couple to the
π
-stack by a variety of mechanisms: end capping (13), intercalation (11, 12),
and direct conjugation (14). However, the recent characterization of covalently tethered
Methylene Blue (MB), an extensively used reporter for DNA-conformation based assays,
has spurred a new debate with regards to the mechanism of its reduction (12, 19–21).
The capacity of these redox reporters to be reduced via an intraduplex DNA-mediated
pathway has been brought into question with alternative mechanisms such as the duplex
tilting to the surface, the charge traveling along the counter ions associated with the sugar
phosphate backbone, or the reporter intercalating in a neighboring duplex (19–21). Here we
Corresponding Author
. jkbarton@caltech.edu.
ASSOCIATED CONTENT
Supporting Information
Experimental procedures and supporting figures. This material is available free of charge via the Internet at
http://pubs.acs.org
.
Notes
The authors declare no competing financial interests.
NIH Public Access
Author Manuscript
J Am Chem Soc
. Author manuscript; available in PMC 2014 October 09.
Published in final edited form as:
J Am Chem Soc
. 2013 October 9; 135(40): 14944–14947. doi:10.1021/ja408135g.
NIH-PA Author Manuscript
NIH-PA Author Manuscript
NIH-PA Author Manuscript
demonstrate the generality of DNA-mediated reduction of covalently tethered reporters by
varying both the redox-active species and the covalent linkage. Critical to these assays is the
use of an intervening AC mismatch to establish that the charge transport pathway is through
the duplex base pair stack. Beyond demonstrating that the reduction is DNA-mediated, we
also provide evidence supporting an intraduplex process by varying the identity and
proximity of neighboring duplex DNA.
The electronic coupling of redox-active reporters that are covalently tethered through a long
(dT-C12-DNA) flexible alkane linkage was explored by linking three different NHS-ester
activated redox reporters (MB, Nile Blue (NB), and Anthraquinone (AQ)) to the same
amine-modified DNA (Figure 1). The reporter-modified DNA was annealed to a thiol-
modified complement prior to the self-assembly on gold electrodes. Consistency between
electrodes was ensured through the use of multiplexed chips containing 16 individually
addressable electrodes (SI) (6). The midpoint potentials of the signals generated from
densely packed DNA monolayers (assembled at 25 μM in the presence of 100 mM MgCl
2
)
labeled with MB, NB, and AQ are −300 mV, −420 mV, and −500 mV, respectively (Figure
1). Interestingly, the oxidative signal sizes are 6.1 ± 0.3 nC (MB), 2.8 ± 0.2 nC (NB), and
2.2 ± 0.4 nC (AQ), which is not consistent across the reporters. The signal sizes, however,
correlate with the binding affinity of reporters based on their ability to thermally stabilize
the duplex DNA (Supporting Table 2). Therefore, MB is well coupled to the
π
-stack
compared to the relatively poorly coupled AQ. The quantity of immobilized DNA was
determined using the electrostatic binding of Ru(NH
3
)
6
3+
at low but saturating conditions
based on titration (1 μM in 5.0 mM phosphate, 50 mM NaCl, pH 7.0) (Figure S1).
Regardless of the redox reporter, the electrodes all were found to have surface coverages
within error of each other, 5.2 ± 0.3 pmol/cm
2
, consistent with that previously reported
based on reporter reduction (12). This ensures that differences in signal sizes are not due to
differences in the quantity of immobilized DNA.
In addition to varying the redox-active species, both the length of the covalent linkage and
the placement of the redox-active reporter were also probed (Figure 2). In addition to the
long C12 linkage appended off the thymine, two new linkages were prepared: a shortened
C8 linkage appended off the modified thymine in the same manner and a C12 linkage
instead appended off the 5
′
hydroxyl (26). Densely packed monolayers were compared for
MB (Figure 2) and NB (Figure S2) covalently tethered to the DNA through these three link-
ages. Taken together, the midpoint potential, peak splitting, and signal size create a distinct
reproducible profile for each linkage (Table S3). Notably, differences are observed in the
signals generated from MB covalently tethered via the same C12 linkages appended to the
thymine and 5
′
hydroxyl as well as between both the short and long linkages appended off
the thymine. These distinct electrochemical profiles indicate that both the placement and
length of the covalent tether play critical roles in dictating the coupling of the reporter with
the base pair stack and therefore the kinetics of the redox signal. For example, the difference
in midpoint potential between the dT-C12 and dT-C8 linkages is attributed to the truncated
linkage, decreasing the binding affinity of the reporter; in fact, the midpoint potential shift
results solely from a shift in the reduction peak, while the oxidative peak remains relatively
constant. This differential behavior of the reductive and oxidative peaks can be understood
based on the lowered binding affinity of the reporter in the reduced form (27). The shift in
the midpoint potential of the 5’ hydroxyl linkage is attributed to binding via an end-capping
mode as opposed to intercalation (13). Additionally, based on poorer coupling, NB
consistently shows a decreased signal size (40% – 80%) compared to MB for all three
linkages under all conditions examined.
The degree of DNA-mediated reduction for this family of redox reporters and linkages was
assessed by introducing a single mismatched base pair intervening between the redox
Pheeney and Barton
Page 2
J Am Chem Soc
. Author manuscript; available in PMC 2014 October 09.
NIH-PA Author Manuscript
NIH-PA Author Manuscript
NIH-PA Author Manuscript
reporter and the gold electrode. This is an
essential
assay of DNA-mediation. DNA CT is
known to be exquisitely sensitive to subtle perturbations to the
π
-stack; therefore the
introduction of even a single intervening mismatched base pair (AC) significantly attenuates
the overall yield of electrons reaching the distally bound reporter when the signals are
generated via a DNA-mediated pathway (7, 24). This diminished yield of reporter reduction
is seen to be associated with incorporation of a single base pair mismatch even in a 100-mer
and over a wide range of temperatures (9, 28). Here direct comparisons between fully well-
matched and AC mismatched 17-mer DNA were performed on the same multiplexed surface
to reduce effects caused by the assembly conditions and surface quality. The percent signal
remaining with a single AC mismatch was determined relative to well- matched DNA for all
linkages and for both MB and NB (Table 1). Although AQ appended by a C12 linkage to the
terminal thymine was also shown to display significant signal attenuation upon introducing a
single AC mismatch (Figure S3), its negative midpoint potential and oxygen- mediated
electrocatalytic activity made signal quantification challenging. Overall, AQ poses the most
challenges to quantification and is poorly coupled to the
π
-stack so it was not further
investigated as an electrochemical reporter for DNA CT.
The degree of signal attenuation upon mismatch incorporation was characterized under both
low and high density conditions. It has previously been established that the DNA-mediated
reduction of MB-dT-C12-DNA is exceptionally sensitive to the assembly conditions of the
electrode, yielding a greater DNA-mediated signal upon decreased surface accessibility of
the reporter (12). All linkages attaching both MB and NB displayed significant signal
attenuation upon incorporation of a single AC mismatch under high density conditions,
validating that the predominant mechanism for reduction is a DNA-mediated pathway
(Figure 2 and Table 1). The quantity of immobilized DNA was again verified by the
addition of Ru(NH
3
)
6
3+
(1 μM). DNA surface coverages were found to be within error of
one another for well-matched or mismatched DNA for both film densities, 5.2 ± 0.3 pmol/
cm
2
and 3.0 ± 0.3 pmol/cm
2
for high and low density respectively (Figure 3 and S4). This
ensured that the observed differences in signal were not a result of altered hybridization
efficiency caused by the incorporation of the AC mismatch. Finally, the scan rate was varied
from 50 mV to 13 V to establish the rate of electron transfer for MB with both dT-C8 and
dT-C12 linkages (Figure S5). Based on the Laviron analysis they displayed similar rates of
electron transfer (2–4 s
−1
), consistent with previously established results for a DNA-
mediated mechanism where tunneling through the alkane-thiol is rate limiting (8, 12).
Under low density conditions, the percent signal remaining upon incorporation of a single
AC mismatch ranged from 60% – 90%, compared to the high density conditions where it
ranged from 10% – 40% across both reporters and all linkages (Table 1). The general nature
of this dependence on film density speaks to the validity of our proposed mechanism, where
surface accessibility dictates whether or not charge transport proceeds via DNA-mediated or
surface-mediated pathways. Additionally, this demonstrates the utility of the truncated dT-
C8 linkage as a means of decreasing the surface accessibility of the reporter as it retains the
most DNA-mediated character under low density conditions.
The proximity of the neighboring duplex DNA within the monolayer was again varied, this
time including the additional assembly condition of 1 mM MgCl
2
, to provide further support
that DNA-mediated reduction is favored under conditions of high surface coverages and not
simply because of the addition of MgCl
2
during assembly. The addition of MgCl
2
during
monolayer formation has been used for over a decade to control the density of DNA
monolayer packing (2, 7). As previously mentioned, 100 mM MgCl
2
has been used to
produce dense DNA monolayers that physically extrude the reporter from directly accessing
the electrode surface therefore enhancing DNA-mediated electrochemistry; however, it has
Pheeney and Barton
Page 3
J Am Chem Soc
. Author manuscript; available in PMC 2014 October 09.
NIH-PA Author Manuscript
NIH-PA Author Manuscript
NIH-PA Author Manuscript
yet to be determined whether this effect results from simply the addition of MgCl
2
during
assembly or from the increased surface coverages obtained at high MgCl
2
concentrations.
Signal attenuation upon introducing a single AC mismatch and the concentration of
immobilized DNA, based on the electrostatic binding of Ru(NH
3
)
6
3+
, were quantified for
MB covalently tethered by both dT-C8 and dT-C12 linkages at three different DNA
monolayer conditions, incubation without MgCl
2
, with 1 mM MgCl
2
, and with 100 mM
MgCl
2
(Figure S7). The observed signal attenuation upon incorporation of an AC mismatch
and the DNA surface coverage in the 1 mM MgCl2 film is not altered compared to signals
for DNA monolayers assembled without MgCl
2
(Figure s7). As the sensitivity to
π
-stack
perturbations is only enhanced at higher DNA surface coverages, this result further supports
the model where the extent of the exposed electrode surface dictates the mechanism of either
DNA-mediated or surface-mediated reduction.
The significance of the neighboring duplex integrity was investigated by introducing
increasing fractions of unlabeled duplex DNA with a single AC mismatch. DNA-modified
electrodes were assembled using a 50/50 mixture of well-matched MB-modified DNA and
unlabeled DNA of varying compositions of well-matched and mismatched DNA. As the
truncated dT-C8 linkage displays DNA-mediated character under both high and low density
conditions it was used to establish whether inter- or intra- duplex intercalation is the
predominant mode of reporter binding to the
π
-stack (Figure 4). Results for the dT-C12
linkage were included for completeness (Figure S6). As the fraction of well-matched MB-
modified DNA remains constant, an attenuation of signal would only be expected in the
presence of higher fractions of mismatched DNA if the reduction occurred through an
interduplex pathway, as opposed to an intraduplex pathway where the signal is expected to
remain unaltered. The predicted extent of signal attenuation for an interduplex pathway was
calculated at each mismatch concentration based on the fraction of mismatched DNA in the
film and the percent signal attenuation previously determined for a 100% mismatched DNA
film for both assembly conditions (Table 1). In all cases, the experimental results of
incorporating mismatched DNA showed significantly less signal attenuation than predicted,
consistent with a predominantly intraduplex pathway (Figure 4). Most notably, for the case
of the low density DNA films, the effect is less than 5% of that predicted, suggesting that
95% of the DNA-mediated signal is generated via an intraduplex pathway. Therefore, there
is roughly 20-fold more intraduplex reduction occurring than interduplex reduction within
these films. Even in the densely packed DNA monolayers where the decreased distance
between duplexes would facilitate interduplex intercalation of the reporter, the observed
attenuation was still only 25% of that predicted, indicating that intraduplex reduction is still
3-fold more favored than an interduplex pathway.
Thus, we have demonstrated that intraduplex DNA CT is the primary mechanism for the
redox-activity of probe molecules that are covalently tethered to a DNA duplex and well
stacked. The extent of DNA-mediated electronic coupling depends not only on how tightly
the redox-active species interacts with the
π
-stack but also the location and structure of the
covalent tether. Finally, the possibility by which charge transport occurs through the counter
ions associated with the sugar-phosphate backbone or the reporter intercalating into the
neighboring duplex are not supported by our results. Instead we observe reporter sensitivity
to intervening
π
-stack perturbations and a tolerance to the integrity of the neighboring
duplex DNA. These results fully support the intraduplex DNA-mediated reduction of these
covalently tethered reporters. Ultimately, delineating the mechanisms of electron transfer in
DNA-modified electrodes is critical for their continued development as useful diagnostic
tools.
Pheeney and Barton
Page 4
J Am Chem Soc
. Author manuscript; available in PMC 2014 October 09.
NIH-PA Author Manuscript
NIH-PA Author Manuscript
NIH-PA Author Manuscript
Supplementary Material
Refer to Web version on PubMed Central for supplementary material.
Acknowledgments
We are grateful for the support of NIH (GM61077). We also thank the Kavli Nanoscience Institute facilities and
staff for help in fabricating the multiplexed chips.
REFERENCES
1. Lam B, Fang Z, Sargent EH, Kelley SO. Anal. Chem. 2012; 84:21. [PubMed: 22142422]
2. Lapierre MA, O’Keefe M, Taft BJ, Kelley SO. Anal Chem. 2003; 75:6327. [PubMed: 14616017]
3. Zhao S, Yang W, Lai RY. Biosens. Bioelectron. 2011; 26:2442. [PubMed: 21081271]
4. Kwon SJ, Bard AJ. J. Am. Chem. Soc. 2012; 134:10777. [PubMed: 22702801]
5. Pheeney CG, Guerra LF, Barton JK. Proc. Natl. Acad. Sci. 2012; 109:11528. [PubMed: 22733728]
6. Slinker JD, Muren NB, Gorodetsky AA, Barton JK. J. Am. Chem. Soc. 2010; 132:2769. [PubMed:
20131780]
7. Kelley SO, Jackson NM, Hill MG, Barton JK. Angew. Chem. Int. Ed. 1999; 38:941.
8. Drummond TG, Hill MG, Barton JK. J. Am. Chem. Soc. 2004; 126:15010. [PubMed: 15547981]
9. Slinker JD, Muren NB, Renfrew SE, Barton JK. Nat. Chem. 2011; 3:228. [PubMed: 21336329]
10. Pheeney CG, Arnold AR, Grodick MA, Barton JK. J. Am. Chem. Soc. 2013; 135:11869. [PubMed:
23899026]
11. Kelley SO, Barton JK, Jackson N, Hill MG. Bioconjugate Chem. 1997; 8:31.
12. Pheeney CG, Barton JK. Langmuir. 2012; 28:7063. [PubMed: 22512327]
13. Boon E, Jackson N, Wightman M, Kelley S, Hill M, Barton JK. J. Phys. Chem. B. 2003;
107:11805.
14. Gorodetsky A, Green O, Yavin E, Barton JK. Bioconjugate Chem. 2007; 18:1434.
15. Weizmann Y, Chenoweth DM, Swager TM. J. Am. Chem. Soc. 2011; 133:3228.
16. Wu J, Chumbimuni-Torres KY, Galik M, Thammakhet C, Haake DA, Wang J. Anal. Chem. 2009;
81:1007.
17. Plaxco KW, Soh HT. Trends Biotech. 2011; 29:1.
18. Bonham AJ, Hsieh K, Ferguson BS, Vallee-Belisle A, Ricci F, Soh HT, Plaxco KW. J. Am. Chem.
Soc. 2012; 134:3346. [PubMed: 22313286]
19. Abi A, Ferapontova EE. J. Am. Chem. Soc. 2012; 134:14499. [PubMed: 22876831]
20. Farjami E, Campos R, Freapontova EE. Langmuir. 2012; 28:16218. [PubMed: 23106377]
21. Yu Z, Lai RY. Anal. Chem. 2013; 85:3340. [PubMed: 23413882]
22. Wu Y, Lai RY. Chem. Comm. 2013; 49:3422. [PubMed: 23503676]
23. Wagenknecht, HA., editor. Charge Transfer in DNA:From Mechanism to Application. Weiheim:
Wiley-VCH; 2005.
24. Genereux JC, Barton JK. Chem. Rev. 2010; 110:1642. [PubMed: 20214403]
25. Murphy CJ, Arkin MR, Jenkins Y, Ghatlia ND, Bossmann SH, Turro NJ, Barton JK. Science.
1993; 262:1025. [PubMed: 7802858]
26. There is precedence for MB-modified DNA with all of the linkages used here: dT-C12 (12, 19),
dT-C8 (5), and 5
′
-C12 (13, 20).
27. Boon EM, Barton JK, Bhaghat V, Nerissian M, Wang W, Hill MG. Langmuir. 2003; 19:9255.
28. Wohlgamuth CH, McWilliams MA, Slinker JD. Anal. Chem. 2013; 85:1462. [PubMed: 23252597]
Pheeney and Barton
Page 5
J Am Chem Soc
. Author manuscript; available in PMC 2014 October 09.
NIH-PA Author Manuscript
NIH-PA Author Manuscript
NIH-PA Author Manuscript
Figure 1.
Variation of redox-active species. Structures (top) and cyclic voltammetry (scan rate = 100
mV/s)(bottom) of MB (blue), NB (purple), and AQ (orange) modified DNA covalently
tethered via a C12 flexible alkane linkage on a modified thymine. DNA-monolayers were
assembled in the presence of 100 mM MgCl
2
, backfilled with 6-mercaptohexanol (1 mM for
45 min), and scanned in spermidine buffer, see SI for full methods.
Pheeney and Barton
Page 6
J Am Chem Soc
. Author manuscript; available in PMC 2014 October 09.
NIH-PA Author Manuscript
NIH-PA Author Manuscript
NIH-PA Author Manuscript
Figure 2.
Variation of covalent linkage. Structures (top) and cyclic voltammetry (scan rate = 100 mV/
s) (bottom) of MB covalently tethered to duplex DNA via three different linkages: 5
′
-C12
(red), dT-C12 (blue), and dT-C8 (green). The signals from well-matched (solid) and AC
mismatched (dashed) 17-mer DNA are both presented, sequences available in SI.
Pheeney and Barton
Page 7
J Am Chem Soc
. Author manuscript; available in PMC 2014 October 09.
NIH-PA Author Manuscript
NIH-PA Author Manuscript
NIH-PA Author Manuscript
Figure 3.
Consistency of Ru(NH
3
)
6
3+
quantification across DNA sequences in high density DNA
monolayers. Representative cyclic voltammograms (scan rate = 100mV/s) of the
quantification of immobilized DNA by the electrostatic binding of Ru(NH
3
)
6
3+
(1 μM) in
phosphate buffer are presented for MB-dT-C8-DNA. Well-matched (blue) and AC
mismatched (red) 17-mer DNA are presented. 4 individual electrodes are presented to
demonstrate the variability observed. Signals from both Ru(NH
3
)
6
3+
and MB are present at
−250 mV and −300 mV respectively. Only the MB signal shows significant signal
attenuation upon mismatch incorporation.
Pheeney and Barton
Page 8
J Am Chem Soc
. Author manuscript; available in PMC 2014 October 09.
NIH-PA Author Manuscript
NIH-PA Author Manuscript
NIH-PA Author Manuscript
Figure 4.
Effect of neighboring duplex integrity for MB-dT-C8-DNA. Electrodes assembled with MB-
modified well-matched DNA and varied fractions of unlabeled well-matched and AC
mismatched DNA. The experimental (blue) reductive signal areas were determined at each
fraction of unlabeled mismatched DNA and normalized to the signal at 100% well-matched
DNA. The predicted values were determined by the total fraction of mismatched DNA times
by the percent signal attenuation for the given linkage and assembly conditions. Electrodes
assembled in the presence (triangle) and absence (square) of 100 mM MgCl
2
are presented.
Pheeney and Barton
Page 9
J Am Chem Soc
. Author manuscript; available in PMC 2014 October 09.
NIH-PA Author Manuscript
NIH-PA Author Manuscript
NIH-PA Author Manuscript
NIH-PA Author Manuscript
NIH-PA Author Manuscript
NIH-PA Author Manuscript
Pheeney and Barton
Page 10
Table 1
Percent signal remaining for MB and NB modified DNA for all covalent linkages examined in high and low
density DNA monolayers.
Covalent Tether
Percent Signal Remaining
(MM/WM*100)
b
,
c
MB-DNA
NB-DNA
High Density
dT-C
8
9
42
dT-C
12
18
N/A
a
5
’-C
12
6
15
Low Density
dT-C
8
65
66
dT-C
12
89
N/A
a
5
’-C
12
61
75
a
Data were not acquired due to low synthetic yields.
b
Errors on all percent signal remaining are ± 1 %.
c
WM is well-matched and MM is AC mismatched.
J Am Chem Soc
. Author manuscript; available in PMC 2014 October 09.