of 18
S
1
Supplementary Materials
for
Helix
-
dependent Spin Filtering through the DNA
Duplex
Theodore J. Zwang,
1
Sylvia H
ü
rlimann,
1
Michael G. Hill,
2
Jacqueline K.
Barton
1,*
correspondence
to:
jkbarton@c
altech.edu
This PDF file
includes:
Materials and Methods
Supplementary
Text
Fig
s.
S1
to
S5
Table
S1
S
2
Materials and Methods
DNA Synthesis
All materials for DNA synthesis were purchased from Glen Research. Oligonucleotides
were synthesized on an Applied Biosystems 3400 DNA synthesizer using
phosphoramidite chemistry on a controlled
-
pore glass support. The two strands of a
duplex
were
synthesized separately, purified, stored frozen, then annealed prior to
electrochemical experiments. The 5’
-
end of one strand
was
modified with a C6 S
-
S
phosphoramidite that is later reduced befo
re use. The 5’
-
end of the comple
mentary
strand is either un
modified or modified with an NHS
-
Carboxy
-
dT phosphoramidite for
later coupling with
N
ile blue. High pressure liquid chromatography (HPLC) was
performed using a reverse
-
phase PLRP
-
S column (Agilent) using a gradient of
acetonitrile and 50 mM ammonium acetat
e.
Unmodified DNA.
DNA was synthesized using standard phosphoramidites and
reagents. After synthesis, the DNA was
lyophilized overnight
. It was then cleaved from
the solid support by incubation at 60
o
C with
concentrated (28
-
30%)
NH
4
OH for 12
hours, filter
ed
using CoStar columns
, then dried. The dried DNA film was resuspended in
phosphate buffer (5 mM phosphate,
pH 7, 50 mM NaCl
) and HPLC
-
purified. The DMT
(4,4’
-
dimethoxytrityl) group protecting the 5’
-
end was then removed by incubation with
80% acetic aci
d for 45 minutes. The reaction mixture was dried and
resuspended in
phosphate buffer. The DNA was isolated
using HPLC. The purified
oligonucleotide
was
desalted using ethanol
precipitation
, dried, and the mass was confirmed with Matrix
-
assisted laser desor
ption/ionization
-
time of flight mass spectrometry (MALDI
-
TOF).
S
3
Unmodified
oligonucleotides were then stored at
-
20
o
C
in phosphate buffer until
annealing with their complementary strand.
Thiolated DNA.
DNA was synthesized using standard phosphoramidites an
d
reagents, with the exception of a C6 S
-
S phosphoramidite that was attached to the 5’
-
end.
After synthesis, the DNA was lyophilized overnight. It was then cleaved from the
solid support by incubation at 60
o
C with concentrated (28
-
30%) NH
4
OH for 12 hours
,
filtered using CoStar columns, then dried. The dried DNA film was resuspended in
phosphate buffer (5 mM phosphate, pH 7, 50 mM NaCl) and HPLC
-
purified. The DMT
(4,4’
-
dimethoxytrityl) group protecting the 5’
-
end was then removed by incubation with
80% ac
etic acid for 45 minutes. The reaction mixture was dried and resuspended in
phosphate buffer. The DNA was isolated using HPLC.
The purified
oligonuceotide
was
desalted using ethanol
precipitation
, dried, and the mass was confirmed with Matrix
-
assisted lase
r desorption/ionization
-
time of flight mass spectrometry (MALDI
-
TOF).
Within one week of annealing
and use
, the dithiolated DNA was reduced by resuspending
in 50 mM Tris
-
HCl, pH 8.4, 50 mM NaCl, 100 mM dithiothreitol (Sigma) for 2 hours.
The reduced thiol
-
modified DNA was then purified by size exclusion chromatography
(Nap5 Sephadex, G
-
25, GE Healthcare) with phosphate buffer as the eluent and
subsequently purified using HPLC.
Covalent Coupling of Nile Blue to DNA.
DNA was synthesized using ultramild
pro
cedures from Glen Research. Phophoramidites compatible with ultramild synthesis
were used as well as ultramild Cap A
(5% phenoxyacetic anhydride in THF/Pyridine)
. An
NHS
-
Deoxy
-
dT phosphoramidite was attached to the 5’
-
end. After synthesis, the DNA
was dri
ed overnight in a lyophilizer. Nile blue perchlorate (Sigma) was then dissolved
S
4
(15
mg/ml) in anhydrous N,N
-
dimethylformamide and activated with 10% v/v DIEA
(N,N
-
Diisopropylethylamine). This
N
ile blue solution was then incubated with the DNA
overnight
with gentle shaking. At least 18 hours later, the DNA
solid support
beads were
rinsed with N,N
-
dimethylformamide then dichloromethane. The DNA was then cleaved
from the solid support
by incubating with
0.05
M potassium carbonate in methanol
for 4
hours at
ambient
temperature. The DNA was then
dried, resuspended in phosphate buffer,
and the oligonucleotide was isolated with
HPLC. The purified
oligonucleotide
was
desalted using ethanol
precipitation
, dried, and the mass was confirmed with
MALDI
-
TOF
. Nile blue
-
modified strands were then stored in the dark at
-
20
o
C
in phosphate
buffer until annealing with their complementary strand.
Synthesis of 3’
-
Dabcyl
-
DNA
.
DNA was synthesized using standard
phosphoramidites and reagents, with the exception of 3’
-
Dabcyl
modified controlled pore
glass (CPG)
beads. After synthesis, the DNA was
lyophilized overnight
. It was then
cleaved from the solid support by incubation at
ambient
temperature with concentrated
(28
-
30%)
NH
4
OH for 2 hours, spin filtered to remove the
solid
support
beads, then dried.
The
oligonucleotide
was resuspended in phosphate buffer and HPLC
-
purified. The DMT
(4,4’
-
dimethoxytrityl) group protecting the 5’
-
end was then removed by incubation with
80% acetic acid for 45 minutes, dried, resuspended in phos
phate buffer, and re
-
purified
using HPLC. The purified strand was desalted using ethanol
precipitation
, dried, and the
mass was confirmed with
MALDI
-
TOF
.
Annealing Duplex DNA.
Duplex DNA for electrochemistry was prepared by first
quantifying the complement
ary strands with UV
-
Visible spectroscopy, then
mixing
equimolar (50
M) complementary strands in
200
l
phosphate buffer. The DNA
S
5
solution was then deoxygenated by bubbling argon for at least 5 minutes per ml. Duplex
DNA was then annealed on a thermocycler
(Beckman Instruments) by initial heating to
90
o
C followed by slow cooling over 90 minutes.
DNA Sequences
:
CG
-
repeat
3’
-
GCG CGC GCG CGC GCG C
-
5’
HS
-
C
6
-
5’
-
CGC GCG CGC GCG CGC G
-
3’
C
m
G
-
repeat
3’
-
GC
m
G C
m
GC
m
GC
m
G C
m
GC
m
GC
m
G C
m
HS
-
C
6
-
5’
-
C
m
GC
m
GC
m
G C
m
GC
m
GC
m
G C
m
GC
m
G
-
3’
16bp DNA
3’
-
TGC AGA GTT GAG TGC A
-
5’
HS
-
C
6
-
5’
-
ACG TCT CAA CTC ACG T
-
3’
30bp DNA (well
-
matched)
3’
-
AGA GTT CTT AGC CGT AAT CGA GTT GAC AGT
-
5’
HS
-
C
6
-
5’
-
TCT CAA GAA TCG GCA TTA GCT CAA CTG TCA
-
3’
30bp DNA (C:A
mismatch)
3’
-
AGA GT
C
CTT AGC CGT AAT CGA GTT GAC AGT
-
5’
HS
-
C
6
-
5’
-
TCT CA
A
GAA TCG GCA TTA GCT CAA CTG TCA
-
3’
17 bp NB (Well matched)
3’
-
CGA GTC ATG CTG CAG CT
-
5’
-
NB
HS
-
C
6
-
5’
-
GCT CAG TAC GAC GTC GA
-
3’
17 bp NB (C:A mismatch)
3’
-
CGA GT
C
ATG CTG CAG CT
-
5’
-
NB
HS
-
C
6
-
5’
-
GCT CA
A
TAC GAC GTC GA
-
3’
29 bp NB
3’
-
CAC CGT CCA GTC AGT ACA TAT GAC GTG AT
-
5’
-
NB
HS
-
C
6
-
5’
-
GTG GCA GGT CAG TCA TGT ATA CTG CAC TA
-
3’
43 bp NB
3’
-
CGT CAT CCA CTT AGC ACC GTC CAG TCA GTA CAT ATG ACT TGA T
-
5’
-
NB
HS
-
C
6
-
5’
-
GCA GTA GGT GAA TCG TGG CAG GTC AGT CAT GTA TAC TGA ACT A
-
3’
60 bp NB
3’
-
CAA GTA CTG TAT GCA TGC GTC ATC CAC TTA GCA CCG TCC AGT CAG TAC ATA TGA CTT GAT
-
5’
-
NB
HS
-
C
6
-
5’
-
GTT CAT GAC ATA CGT ACG CAG TAG GTG AAT CGT GGC AGG TCA GTC
ATG TAT ACT GAA CTA
-
3’
29 bp Dabcyl
DAB
-
3’
-
CAC CGT CCA GTC AGT ACA TAT GAC GTG AT
-
5’
HS
-
C
6
-
5’
-
GTG GCA GGT CAG TCA TGT ATA CTG CAC TA
-
3’
The above sequences use the following abbreviations for modifications:
DAB = Dabcyl; NB = Nile Blue;
HS
-
C
6
= hexanethiol linker; C
m
= 5
-
methyl
-
cytosine
S
6
Electrode Fabrication
Single electrode surface fabrication was carried out at UCLA and received as a gift
following the general protocol of R. Naaman
et al
(20). P
-
type
oxidized s
ilicon wafers
were coate
d with 25 nm titanium as an adhesive layer, following deposition with 200 nm
nickel then 10 nm gold.
The surfaces were then cleaved into 1 cm x 1.5 cm rectangles
and used following the preparation below (for a total 1 cm
2
exposed surface area for
experimen
ts following device assembly).
Multiplexed electrode surfaces were fabricated
following a modified version of a published protocol (2
4
) using the gold
-
capped nickel as
the electrode material.
DNA
-
modified electrode preparation
Gold
-
capped nickel surfaces
or multiplexed chips are gently cleaned by rinsing
with ultrapure water, acetone, isopropanol,
and finally a second rinse with
ultrapure water
before drying with argon. They are then cleaned
with a UVO Cleaner Model 42 (Jelight
Co.)
for 20 minutes. Immediately after cleaning the surface, a plastic clamp and
rubber
(BunaN)
gasket
were
affixed to the surface to create a well for liquid and either 50 μM
duplex DNA, single stranded DNA, or no DNA in
phosphate buffer
(pH 7, 5 mM
phosphate,
50 mM NaCl, 100 mM MgCl
2
) to make densely packed films.
The DNA
was
incubated on the surface for 18
-
24 hours. Once the DNA is on the surface, it cannot be
dried without compromising the structure and therefore the measured properties of the
film. The sol
ution was then exchanged 5x with 1 μM mercaptohexanol in
phosphate
buffer (pH 7, 5 mM phosphate, 50 mM NaCl, 5% glycerol) and incubated for 45 minutes.
S
7
Lastly the surface was rinsed at least 5x with either phosphate buffer (pH 7, 5 mM
phosphate, 50 mM NaCl
) for most experiments or tris buffer (pH 7.6, 10 mM Tris, 100
mM KCl, 2.5 mM MgCl
2
, 1 mM CaCl
2
) for experiments using dabcyl
,
and
electrochemical experiments were performed immediately afterwards.
Electrochemical measurements
The central well around the
electrode surface created by the clamp was filled with
one of three buffers prior to electrochemical measurements: a phosphate buffer with MgCl
2
(pH 7, 5 mM phosphate, 50 mM NaCl, 10 mM MgCl
2
),
a phosphate buffer without MgCl
2
(pH 7, 5 mM phosphate, 50 mM
NaCl),
and a tris buffer
(pH 7.6, 10 mM Tris, 100 mM
KCl, 2.5 mM MgCl
2
, 1 mM CaCl
2
)
. The phosphate buffer was used for all experiments,
except those containing dabcyl. The phosphate buffer
without
MgCl
2
was only used to collect
data with experiments compar
ing methylated and unmethylated dsDNA. The tris buffer is
necessary for both the reduction and oxidation of dabcyl to occur within the potential range
that we can scan. Our experiments did not show a statistically significant difference in
the
magnitude of
spin selectivity when comparing DNA sequences of the same length in the
different buffers
, except in the cases where the increased ionic strength helped prevent
electrostatic binding of MB to the phosphate backbone
.
An Ag
Cl/Ag
reference electrode
(Cypress
) was coated with a solidified mixture of 1% agarose and 3M NaCl in water
inside a long, thin pipette tip. The tip was cut so that the salt bridge could connect the
electrode to the buffer from the top of the well. A platinum wire used as an auxiliary
elec
trode was also submerged in the buffer from the top of the well. The working
electrode contacted a dry part of unmodified gold surface. A grounding wire was
S
8
connected to the metallic base of a ring stand. A CH1620D Electrochemical Analyzer
(CH Instruments)
was used to control the electrochemical experiments.
Magnetic field experiments were conducted using a 6619 Guass surface strength
magnetic field neodymium magnet (K&J Magnetics). Alligator clips were replaced with
nonmagnetic stainless steel to minimize
magnetic interference. All other parts of the
assembly were created using plastic to prevent extraneous objects the magnetic field
could influence. Additionally, the strong magnet was waved near the potentiostat during
operation, with no obvious change in
signal that was apparent, to ensure that the magnet
was not interfering with the operation of the potentiostat. Each experiment that was
condu
cted for magnetic field pointed
up vs down were done by comparing the same
surface in the same solution under these varying field directions in order to minimize
variability caused by other factors.
Noncovalent Methylene Blue.
Solutions of 1μM or 10μM methylene blue were
made in phospha
te buffer
(pH 7, 5 mM phosphate, 50 mM NaCl, 10 mM MgCl
2
)
. Cyclic
voltammograms were collected using a scan rate of
0.05,
0.1, 0.5, 1, 5, 10, and 20 V/s,
which scanned from 0V to
-
0.55V
(vs AgCl/Ag)
then back to 0V at least 6 times
consecutively for each e
xperiment. The magnetic field direction was then switched,
scanned, and swit
ched again multiple times. The first
reductive an
d oxidative sweeps
we
re compared because subsequent scans see a diminishing effect. This attenuation is
restored upon waiting for
approximately 30 seconds between scans.
Ruthenium Hexammine.
A c
oncentration of 10 μM Ru(NH
3
)
6
3+
was added to the
phosphate buffer. Cyclic voltammograms were collected at a scan rate of 0.1 V/s, which
scanned from 0V to
-
0.4V
(vs AgCl/Ag)
then back to 0V
at least 6 times consecutively
S
9
for each experiment. The magnetic field direction was then switched, scanned, and
switched again multiple times. The
first
reductive and oxidative sweeps
were
compared.
B
-
to
-
Z DNA Experiments.
Experiments were conducted with
solutions of 1 μM
methylene blue in phosphate buffer
with and without 10 mM MgCl
2
. Cyclic
voltammograms were collected at a scan rate of 0.1 V/s, which scanned from 0V to
-
0.4V
(vs AgCl/Ag)
then back to 0V at least 6 times consecutively for each experiment
. The
magnetic field direction was then switched, scanned, and switched again at least 4 times.
Following these scans, phosphate buffer with 10 mM MgCl
2
was placed in the well by
exchanging the solution 5x. The surfaces were scanned in a similar manner as
before, and
then the solution was again exchanged 5x to replace it with a phosphate buffer without
magnesium. It was
then
scanned
similarly
.
Covalent
Nile Blue.
Experiments were conducted in phosphate buffer with 4
diffe
rent lengths of DNA, each with N
ile
blue covalently tethered at the 5’
-
end away
from the surface. Cyclic voltammograms were collected using a scan rate of
0.05,
0.1,
0.2,
0.5, 1,
2, 3, 4,
5,
7,
10,
11, 12, 13, 14, 15, 17
and 20 V/s, which scanned from 0V to
-
0.55V then back to 0V 20 times consecutively for each experiment. The magnetic field
direction was then switched, scanned, and switched again multiple times
. The first
reductive an
d oxidative sweeps we
re compared.
3’
-
Dabcyl.
Experiments were conducted in tris buffer
(pH 7.6, 10 mM Tris, 100
mM KCl, 2.5 mM MgCl
2
, 1 mM CaCl
2
)
with 1μM methylene blue. Due to the relatively
slow rate of redox chemistry with azobenzene, cyclic voltammograms were coll
ected
using a scan rate
of 10 mV
/s so the peak splitting allowed for both the reductive and
oxidative peaks to lie within the measurable potential range range. The experiments
S
10
scanned from 0V to
-
0.6V
(vs AgCl/Ag)
then to 0.5V and repeated scanning between
0.5V and
-
0.6V 20 time
s consecutively for each experiment. The magnetic field direction
was then switched, scanned, and switched again multiple times. The
first
reductive and
oxidative sweeps we
re compared.
Calculating the surface concentration of DNA
Surface concentration
of DNA was quantified based on the total area of the reductive
signal generated from a cyclic voltammogram of electrostatic binding of Ru(NH
3
)
6
3+
to
the phosphate backbone of dsDNA. This was done under saturation conditions
(10 μM
Ru(NH
3
)
6
3+
)
. The follow
ing equation was used to relate the reductive signal (Q
Ru
) to the
surface concentration of DNA (Γ
DNA
):
In this equation,
A
is the electrode area in cm
2
,
F
is the Faraday constant,
n
is the number
of electrons per reduction event,
z
is the charge on the Ru(NH
3
)
6
3+
, and
m
is the number
of base pairs in the duplex DNA. The surface concentration of dilute DNA films was 8±1
pm/cm
2
and dense DNA films was 40±3 pm/cm
2
.
Circular Dichroism Spectroscopy
An Aviv 62A DS spectropolarimeter wa
s used to collect circular dichroism (CD)
spectra
.
Data were obtained from samples containing 3 μM
d
(
m
CG)
8
or 3 μM
d
(CG)
8
dsDNA in
phosphate buffer (5 mM phosphate, 50 mM NaCl, pH7) using a 1.0 mm path length cell
(Fig
ure
S3
).
Scans were conducted with sam
ples that were incubated
in the presence or
absence of
10 mM MgCl
2
. Data presented in figures represent the average of three scans.
S
11
The B
-
Z transformation is very clearly seen in the CD spectra of the methylated DNA.
Upon addition of magnesium
ion,
there i
s a large decrease in magnitude of ellipticity at
254 nm and 293 nm with isochromism at 277 nm, which is characteristic of the B
-
Z
transformation. The unmethylated DNA does not show any change in CD spectr
um
for
conditions with and without magnesium
ion
, w
hich confirms that it remains in the B
-
form.
Spin Polarization Calculations
The spin polarization (
S
) is defined as:
(1)
in which
I
+
and
I
-
are the intensities of the signals corresponding to the spin oriented
parallel and antiparallel to the electrons
velocity.
(
12)
The spin polarization for electrons
travelling
through
dsDNA is calculated assuming that the injected spin polarization is
23%
which, using equation 1, results in the total amount of each spin injected assumed to
be 0.615 and 0.385, with the majority spin depending on the magnetization direction.
(
20)
The amount of charge transferred to the probe (
Q
)
, which is determined by integr
ating the
current under the reductive or oxidative peak in the cyclic voltammograms,
can be related
to the injected spin polarization by the following equation:
(2)
w
here
η
+
and
η
-
are the yield for the amount of injected spin ori
ented parallel
and
antiparallel, respectively, to the velocity of the electrons that reduce
the probe compared
S
12
to the total amount injected. Therefore the amount of probe reduced can be related as
Q = 0.615
(Yield
Up
) + 0.385 (Yield
Down
) for one magnetization and Q
= 0.615 (Yield
Down
)
+ 0.385 (Yield
Up
) for the other.
Modifying equation 1 to solve for the spin polarization through dsDNA gives equation 3.
(3)
Solving for
η
+
and
η
-
and placing them into equation 3 gives the spin polarization of
electrons mov
ing through dsDNA (
S
DNA
)
.
Our data lead
to the calculation that electrons traveling through a densely packed
monolayer of 16 bp dsDNA to MB must be at least 22.5±1.2% spin polarized at
ambient
temperature. Charge transport through 60 bp dsDNA to a covalently tethered NB probe
must be at least 55±10% spin polarized. More calculations are presented in Supplemental
Table 1
.
It should be evident that decreasing the initial spin polarization will i
ncrease the
calculated spin polarization through dsDNA
;
therefore these calculated values can be
considered
lower estimates
. If the nickel injected 13% spin polarized electrons, then the
electrons passing through 60bp dsDNA would be 100% spin polarized, wh
ich could be
treated as a theoretical maximum.
S
13
Fig
ure
S1
.
Plot of the magnetic field effect observed fo
r DNA with covalently tethered
N
ile blue at different
oligonucleotide
lengths scanned at 20 V/s in phosphate buffer. Each
error bar represents the
standard error from at least 3 separate surfaces. ANOVA shows a
statistically significant difference for the effect being length dependent with p = 0.017.
S
14
S
c
a
n
R
a
t
e
D
e
p
e
n
d
e
n
c
e
W
i
t
h
C
o
v
a
l
e
n
t
N
i
l
e
B
l
u
e
0
0
.
0
2
0
.
0
4
0
.
0
6
0
.
0
8
0
.
1
0
.
1
2
0
.
1
4
0
.
1
6
0
.
1
8
-
4
-
3
-
2
-
1
0
1
2
3
4
C
h
a
n
g
e
I
n
P
e
a
k
S
p
l
i
n
g
(
V
)
L
n
(
S
c
a
n
R
a
t
e
)
M
a
g
n
e
t
U
p
M
a
g
n
e
t
D
o
w
n
M
a
g
n
e
t
U
p
k
=
2
3
.
3
s
-
1
a
=
.
2
M
a
g
n
e
t
D
o
w
n
K
=
2
6
.
6
s
-
1
a
=
.
2
1
Fig
ure
S2
.
Scan Rate dependence of covalently tethered
N
ile blue reduction through
duplex DNA
.
S
15
Figure S3.
Cyclic voltammetry of 29bp dsDNA with covalently tethered dabcyl and
noncovalently intercalated MB. (Left) Cartoon representing the DNA tethered to the
surface and the paths the electrons take from the surface to the
ir respective redox probes
.
(Center Left
) Illust
ration of dabcyl molecule.
(
Center Right
) Representative cyclic
voltammogram of DNA with 1 μM MB with the magnetic field pointing towards the
surface (red, up) or away from the surface (blue, down). The insets show the dabcyl
oxidatio
n (green border) and methylene blue reduction (blue border).
The signal centered
around
-
0.5V
corresponds to
the reduction of dabcyl
, while its oxidation appears at 0.3 V
.
Note that scans were carried out at 10 mV/s given the slow proton
-
coupled redox
reaction.
(Right) The difference in cyclic voltammetry data for the reduction of MB
(blue), and the oxidation of dabcyl (green) between the two magnetic field directions.
S
16
Figure S
4
.
Summary of cyclic voltammetry data for 17bp duplex DNA with a C:A
mismatch 6 nucleotides from the surface.
A,
A mismatch (red) has decreased total yield
of NB reduction when compared to a well
-
matched duplex (blue) .
B
, The spin selectivity
for
probe reduction is the same for DNA with (blue) and without (red) a C:A mismatch.
S
17
Figure S5
.
Circular dichroism spectra of [A] unmethylated d(CG)
8
and [B] methylated
d(
m
CG)
8
in phosphate buffer with [blue] and without [red] MgCl
2
.