of 18
S1
SUPPORTING INFORMATION
Intraduplex DNAmediated electrochemistry of covale
ntly
tethered redoxactive reporters
Catrina G. Pheeney and Jacqueline K. Barton*
*Division of Chemistry and Chemical Engineering,
California Institute of Technology, Pasadena, Calif
ornia 91125, USA
S2
Materials
Chemicals and reagents used in the preparation of
the activated redox active reporters
were purchased from Sigma and used without further
purification. MB′ was prepared based on
previously established protocols (1). All phosphor
amidites and DNA synthesis reagents were
purchased from Glen Research. Silicon wafers used
for the fabrication of multiplexed chips
were purchased from SiliconQuest.
Synthesis of proprionic acid modified Nile Blue (NB
′).
See Supporting Scheme 1 for the synthetic strategy
for the preparation of
NB′
.
5-(Dimethylamino)-2-nitrosophenol (
1
) preparation
.
1
was prepared according to the
procedure described by Moura (2) by dissolving 35(d
imethylamino)phenol (8.26 g, 50 mmol) in
concentrated hydrochloric acid (25 mL) chilled to 0
ºC followed by the slow addition of sodium
nitrite (4.0g, 58 mmol). The reaction mixture was
maintained at 0 ºC and the addition of sodium
nitrite was slowed if orange fumes were observed.
The reaction was left to proceed for 30 min –
1 hr, until the contents solidified. The precipita
te was isolated by vacuum filtration and washed
with chilled diluted hydrochloric acid (1 N). The
yellow5brown solid (5.6 g, 29 mmol, 57%
yield) was confirmed as the desired product by
1
H NMR (300 MHz, DMSO5d
6
): δ 7.6657.24 (m,
1H), 7.2456.78 (m, 2H), 3.36 (s, 6H) and ESI5MS in
7:3 CHCl
3
:MeOH by the observed [M+H]
+
peak at 167.0 g/mol.
1
was dried and used in following reactions without
further purification.
3-(naphthalen-1-ylamino)propanoic acid (
2
) preparation
.
2
was prepared by dissolving
naphthylamine (2.0 g, 14 mmol) in water (15 mL) fol
lowed by the addition of sodium hydroxide
(6 M, 5.0 mL) and equimolar amounts of dissolved ch
loroproprionic acid (1.50 g, 15 mmol).
The reaction was brought to a reflux and allowed to
proceed for 12524 hours. The excess
S3
solvent was removed under reduced pressure and the
desired product was purified by silica
chromatography (CHCl
3
:MeOH 20:1) with an R
f
of 0.5. The clear fraction was dried to yield a
white solid product that was confirmed as the desir
ed product (
2
) by
1
H NMR (300 MHz,
CD
3
OD): δ 80.557.87 (m, 1H), 7.8557.62 (m, 1H), 7.5357
.23 (m, 3H), 7.18 (t,
J
= 8.4 Hz, 1H),
6.65 (t,
J
= 8.0 Hz, 1H), 3.59 (dt,
J
= 9.4, 7.0 Hz, 2H), 2.75 (t, 2H) and ESI5MS in 7:3
CHCl
3
:MeOH by the observed [M+H]
+
peak at 216.1 g/mol.
Proprionic acid modified Nile Blue (
NB′
) preparation
.
NB′
was prepared based on
previously established procedures by Moura (3). Eq
uimolar portions of
1
(0.17 g, 1.0 mmol) and
2
(0.22 g, 1.0 mmol) were combined and dissolved in
dimethylformamide (DMF) (6 mL). The
reaction mixture was heated to 70 ºC for 24 hours.
The crude reaction mixture was dried under
reduced pressure and purified by dry silica chromat
ography (4). Pure
NB′
was eluted with 6:1
cholorform: methanol as a dark blue compound. The
identity was verified by ESI5MS in
MeOH:H
2
O 5:1 based on the observed [M]
+
peak at 362.3 g/mol.
NHSester activation of carboxylic acid modified re
porters
MB′, NB′, and anthraquinone525carboxylic acid (AQ′
) were all activated to the NHS5
esters immediately prior to coupling to amino5modif
ied DNA to aid the yield of amide bond
formation. The same procedure as previously used f
or the activation of MB′ was applied to the
activation of all three of these redox active repor
ters (1). Briefly, the carboxylic acid modified
reporter (0.022 mol) was stirred at room temperatur
e in DMF with excess N,N′5
dicyclohexylcarbodiimide (9.3 mg, 0.045 mmol) and N
5hydroxysuccinimide (5.2 mg,
0.045 mmol) for 12524 hours. The solvent was subse
quently removed under reduced pressure
and the reaction mixture was re5suspended in DMSO (
20 KL per 1 mg activated).
S4
Preparation of probe and thiol modified oligonucleo
tides
Thiol5modified and amino5modified DNA were synthes
ized and purified based on
previously established protocols (1, 5). The proce
dure for DNA purification was unaltered
across all three different amino modifiers utilized
: dT5C6, dT5C4, and 5′5C6. The purified
amino5modified DNA was covalently modified with NHS
5ester activated MB′, NB′, and AQ′
using roughly 105fold excess reporter in an aqueous
basic solution (0.1 M NaHCO
3
). The same
thiol5modified complement stand was used for all du
plex DNA, to negate variations caused by
the thiol driven self5assembly of the DNA. Both w
ell5matched and AC mismatched sequences
were prepared of a CG5rich 175mer duplex to investi
gate the effect due to π5stack perturbations
(Supporting Table 1).
Supporting Table 1. DNA sequences used for electro
chemical measurements
Name
Sequence
Thiol5ssDNA
5'5 HS 5 C
6
– GAC TGA CCT CGG ACG CA 53'
Well5matched ssDNA
(WM)
a, b
5'5
T
G CGT CCG AGG TCA GTC 53'
AC5mismatched ssDNA
(MM)
a, b
5'5
T
G CGT CC
A
AGG TCA GTC 53'
a
T
is the site of the amino5modified nucleotide used
to produce the dT5C6 and dT5C4 linkages.
b
5′5C6 linkage is produced by an additional 5′ phos
phoramidite.
Electrochemistry of DNAmodified electrodes
S5
Multiplexed chips, with 16 individually addressabl
e electrodes, were prepared and used
based on previously established protocols (6, 10).
These 165electrodes were divided into 4
isolated quadrants, each with 4 electrodes, allowin
g for 4 different typical of DNA or
morphologies to be simultaneous compared without an
y variability introduced due to changing
the underlying gold surface. Electrodes were expos
ed to duplex thiol5modified DNA (20 KL of
25 KM) in phosphate buffer (5.0 mM phosphate, 50 mM
NaCl, pH 7) and allowed to self5
assemble for 12516 hours in a humid environment. M
gCl
2
was added to solutions during DNA
self5assembly as indicated in the text to alter the
immobilization of the DNA. After DNA self5
assembly, electrodes were washed with excess phosph
ate buffer and incubated with 65
mercaptohexanol (1 mM for 45 min) in phosphate buff
er with 5% glycerol. Electrodes were
washed again with excess phosphate buffer.
Electrochemical measurements were performed with a
CHI620D Electrochemical
Analyzer and a 165channel multiplexer from CH Instr
uments. A three5electrode setup was used
with a common Pt auxiliary and a quasi Ag/AgCl refe
rence electrode (Cypress Systems) placed
in the central well of the clamp. Cyclic voltammet
ry data were collected at 100 mV/s over a
window of 0 mV to 50.6 mV versus Ag/AgCl unless oth
erwise indicated. Every scan consisted
of 8 sweeps of the potential window (4 reductive an
d 4 oxidative) and the last two sweeps are
used for quantification and to present all data. O
nly signals that were stable across repetitive
scans and were distinct to the given reporter were
used for this investigation. The electrodes
were exposed to the various running conditions in t
he following sequence: an initial background
scan acquired in phosphate buffer (5.0 mM phosphate
, 50 mM NaCl, pH 7), DNA quantification
by Ru(NH
3
)
6
3+
(1 KM in phosphate buffer), excess washing with ph
osphate buffer until
Ru(NH
3
)
6
3+
signal is no longer present, final scan in spermid
ine buffer (5.0 mM phosphate, 50
S6
mM NaCl, 4 mM MgCl
2
, 4 mM spermidine, 50 PM EDTA, 10% glycerol, and pH
7.0). Unless
otherwise indicated presented and quantified data f
or reporter signals is determined from CV in
spermidine buffer. The signal averages and associa
ted error are for a given data set (comparison
of reporters vs. comparison of linkages) has many o
ther factors that can alter the signal size
including surface and self5assembly quality. To co
ntrol for this, the data was acquired in distinct
data sets such that results being directly compared
are always from multiplexed chips running in
close succession. Therefore, the data presented is
either quadrant averages (4 electrodes) or an
average of
n
quadrants (
n
x 4 electrodes) acquired within no more than a few
days of each other
and is indicated where applicable.
The Laviron analysis was used to determine the ele
ctron5transfer rate constant (
k
) for
MB5modified DNA (7, 8). A plot of the peak shift (
E
pc
– E
o’
, mV), where E
pc
is the peak
potential at a given scan rate and E
o’
is the midpoint potential determined at 50 mV, ver
sus the
ln(scan rate), where the scan rate (
v
, mV/s) is varied from 50 mV/s to 13 V/s was constr
ucted.
The electron transfer rate can then be derived from
the linear portion of this plot (RE vs ln(
v
), see
Figure S5) based on Equation 1. However, the rates
determined using this analysis are useful for
comparisons and can only be assessed with confidenc
e as order of magnitude estimates.
Equation 1.
Formula used to determine the rate of reduction (
k
, s
51
) based on the Laviron
anaylsis (7, 8).


 


ln 



Quantification of immobilized DNA
The quantity of immobilized DNA within a given DNA
monolayer was quantified based
off the area of reductive signal generated from the
electrostatic binding of Ru(NH
3
)
6
3+
. A
S7
concentration of 1 mM Ru(NH
3
)
6
3+
was chosen based on a titration of both DNA surfac
e
coverages and Ru(NH
3
)
6
3+
concentrations. The surface coverage of DNA (Γ) w
as calculated
using equation 1 where
n
is the number of electrons per reduction event,
F
is the Faraday
constant,
A
is the electrode area in cm
2
(2 cm
2
),
z
is the charge on the Ru(NH
3
)
6
3+
,
m
is the
number of base in the duplex DNA (17 mer), and fina
lly
Q
is the reductive signal from
Ru(NH
3
)
6
3+
in phosphate buffer minus the redox reporter signa
l in spermidine buffer. As the
contribution of reporter reduction could not be dec
onvoluted from the Ru(NH
3
)
6
3+
signal it was
subtracted even though it was consistently negligib
le and less than 10% of the total Ru(NH
3
)
6
3+
signal. Therefore the Q
Ru
used for determining the surface coverage is obtai
ned by taking the
area of the Q
Ru + MB
and subtracting out the Q
MB
(obtained after washing off the Ru(NH
3
)
6
3+
).
Equation 2.
Formula used to determine the surface coverage (Γ
, pmol/cm
2
) of immobilized
DNA based on Ru(NH
3
)
6
3+
binding.
Γ









S8
Supporting Scheme 1.
Synthetic strategy for the preparation of proprio
nic acid modified Nile
Blue (NB′).
S9
Supporting Table 2. Thermal Stability of Duplex DN
A
a
Redox Active Species
T
M
(°C)
Stabilization
Methylene Blue
63.7 °C
+ 1.6 °C
Nile Blue
63.1 °C
+ 1 °C
Anthraquinone
62.3 °C
+ 0.2 °C
None
62.1 °C
5
a
Duplex DNA (1 KM) was incubated with the redox act
ive species (5 KM) in phosphate buffer
(5.0 mM phosphate, 50 mM NaCl, pH 7.0) and the abso
rbance at 260 nm was monitored every
0.5 °C from 25 5 90 °C. The maximum of the derivat
ive of the absorbance temperature trace was
determined to extract the melting temperature under
each condition.
S10
Figure S1.
Electrostatic binding of Ru(NH
3
)
6
3+
to immobilized DNA. Representative cyclic
voltammograms (scan rate = 100 mV/s) of MB5dT5C125D
NA5modified electrodes acquired after
the addition of Ru(NH
3
)
6
3+
(1 KM) in phosphate buffer (5.0 mM sodium phosphat
e, 50 mM
NaCl, pH 7.0) are presented. The concentration of
MgCl
2
present during DNA self5assembly
was varied to alter the overall DNA morphology: wit
hout MgCl
2
(blue), with 1 mM MgCl
2
(red),
and with 100 mM MgCl
2
(green). The representative traces for the three d
ifferent DNA
morphologies were obtained on a single multiplexed
chip with a common Ru(NH
3
)
6
3+
solution.
The area of the total reductive signal minus the ar
ea of the reporter reductive signal, acquired
after washing off the Ru(NH
3
)
6
3+
, was used to determine the surface coverage of DNA
on the
electrode as outlined above.
S11
Figure S2.
Variation of covalent linkage for NB5DNA. Cyclic
voltammetry (scan rate = 100
mV/s), in spermidine buffer (5.0 mM phosphate, 50 m
M NaCl, 4 mM MgCl
2
, 4 mM spermidine,
50 PM EDTA, 10% glycerol, and pH 7.0), of NB covale
ntly tethered to duplex DNA via three
different linkages: 5′5C12 (red), dT5C12 (blue), an
d dT5C8 (green). The monolayers of 175mer
well5matched DNA were assembled in the presence of
100 mM MgCl
2
to yield high density
DNA films. See Supporting Table 1 for sequences an
d Figure 2 for structures of linkages.
S12
Supporting Table 3. Electrochemical parameters for
MB signals
Linkage
Midpoint
Potential (mV)
Peak
Splitting (mV)
Signal Size (nC)
a
dT5C12
5300
35
9.1
dT5C8
5330
124
8.4
5′5C12
5280
26
10.2
.
a
Signal sizes were determined from the average of 4
electrodes within a
quadrant on a given multiplexed chip. The data use
d to compare all three
linkages was from the same multiplexed chip and the
same thiol strand was used
to help negate surface and self5assembly affects.
S13
Figure S3.
AQ5dT5C125DNA response to π5stack perturbations.
Cyclic voltammograms (scan
rate = 100 mV/s) of AQ5dT5C125DNA modified electrod
es self5assembled in the presence of 100
mM MgCl
2
and acquired in de5oxygenated spermidine buffer (5
.0 mM phosphate, 50 mM NaCl,
4 mM MgCl
2
, 4 mM spermidine, 50 PM EDTA, 10% glycerol, and pH
7.0). The sensitivity to
the introduction of a single perturbation to the π5
stack is demonstrated by the direct comparison
of 175mer well5matched (blue) and AC mismatched (re
d) DNA (see Supporting Table 1 for
sequences). Despite de5oxygenation of the multiple
xed chips, the extent of the background
signals at the negative potentials of AQ allowed fo
r only qualitative comparisons to be
performed.
S14
Figure S4.
Consistency of Ru(NH
3
)
6
3+
quantification across DNA sequences. Cyclic
voltammograms (CVs) of the quantification of immobi
lized DNA by the electrostatic binding of
Ru(NH
3
)
6
3+
(1 KM) in phosphate buffer are presented for MB5dT
5C85DNA. CVs for low density
DNA monolayers and both well5matched (blue) and AC
mismatched (red) 175mer DNA are
presented. For each given condition 4 individual e
lectrodes are presented to demonstrate the
variability observed. Overall, the DNA sequence is
shown to have no significant impact on the
signal from Ru(NH
3
)
6
3+
binding indicating that there is not significant d
e5hybridization of the
mismatched duplex at the surface of the electrode.
S15
Figure S5.
Scan rate dependence. The scan rate was varied f
rom 50 mV to 13 V and the shift in
the reduction potential was determined relative to
midpoint potential quantified at 50 mV for MB
covalently tether to DNA by both dT5C8 (blue) and d
T5C12 linkages (red). Three different
assembly conditions were also examined for each lin
kage: without MgCl
2
(triangle), with 1 mM
MgCl
2
(square), and with 100 mM MgCl
2
(diamond). Cyclic voltammetry was acquired in
spermidine buffer (5.0 mM phosphate, 50 mM NaCl, 4
mM MgCl
2
, 4 mM spermidine, 50 PM
EDTA, 10% glycerol, and pH 7.0) with well5matched D
NA. The rate of electron transfer was
determined based on the Laviron analysis (7, 8) and
was found to be 2 5 4 s
51
for all conditions.
S16
Figure S6.
Effect of neighboring duplex integrity for MB5dT5
C125DNA. Electrodes assembled
with MB5modified well5matched DNA and varied fracti
ons of unlabeled well5matched and AC
mismatched DNA. The predicted (red) and experiment
al (blue) reductive signal areas were
determined at each fraction of unlabeled mismatched
DNA and normalized to the signal at 100%
well5matched DNA. Electrodes assembled in the pres
ence (triangle) and absence (square) of 100
mM MgCl
2
are presented.
S17
Figure S7.
Electrochemistry as a function of assembly condit
ions. The percent signal remaining
(MM/WM*100) (red) and Ru(NH3)63+ signals (black) fo
r MB modified DNA with both dT5C8
(dark) and dT5C12 (light) linkages were examined.
DNA5modified electrodes were assembled
without MgCl2, with 1 mM MgCl2, and with 100 mM MgC
l2. The error was determined from
across 3 sets each containing 4 electrode replicate
s.