Silver nanoassemblies constructed from boranephosphonate DNA

Article abstract of DOI:10.1021/ja400898s

Spatially selective deposition of metal onto complex DNA assemblies is a promising approach for the preparation of metallic nanostructures with features that are smaller than what can be produced by top-down lithographic techniques. We have recently reported the ability of 2′-deoxyoligonucleotides containing boranephosphonate linkages (bpDNA) to reduce AuCl₄ ^-, Ag⁺, and PtCl₄^2- ions to the corresponding nanoparticles. Here we demonstrate incorporation of bpDNA oligomers into a two-dimensional DNA array comprised of tiles containing double crossover junctions. We further demonstrate the site-specific deposition of metallic silver onto this DNA structure which generates well-defined and preprogrammed arrays of silver nanoparticles. With this approach the size of the metallic features that can be produced is limited only by the underlying DNA template. These advances were enabled due to a new method for synthesizing bpDNA that uses a silyl protecting group on the DNA nucleobases during the solid-phase 2′-deoxyoligonucleotide synthesis.

Full text of DOI:10.1021/ja400898s

Article
pubs.acs.org/JACS
Silver Nanoassemblies Constructed from Boranephosphonate DNA
Subhadeep Roy, Magdalena Olesiak, Shiying Shang, and Marvin H. Caruthers*
Department of Chemistry and Biochemistry, University of Colorado, Boulder, Colorado 80309, United States
S
* Supporting Information
ABSTRACT: Spatially selective deposition of metal onto
complex DNA assemblies is a promising approach for the
preparation of metallic nanostructures with features that are
smaller than what can be produced by top-down lithographic
techniques. We have recently reported the ability of 2′-
deoxyoligonucleotides containing boranephosphonate linkages
−
2−
(bpDNA) to reduce AuCl4 , Ag+, and PtCl4 ions to the
corresponding nanoparticles. Here we demonstrate incorpo-
ration of bpDNA oligomers into a two-dimensional DNA array
comprised of tiles containing double crossover junctions. We
further demonstrate the site-specific deposition of metallic
silver onto this DNA structure which generates well-defined
and preprogrammed arrays of silver nanoparticles. With this approach the size of the metallic features that can be produced is
limited only by the underlying DNA template. These advances were enabled due to a new method for synthesizing bpDNA that
uses a silyl protecting group on the DNA nucleobases during the solid-phase 2′-deoxyoligonucleotide synthesis.
any closer than 20−30 nm due to interparticle repulsions.52,53
INTRODUCTION
■
This constrains the size, complexity, and density of metallic
features that can be produced even though the available ‘pixel
size’ in underlying DNA structures can be as small as 3 × 7
nm.23
The ability to manipulate and arrange matter in a
preprogrammed manner at the nanoscale level lies at the crux
of the bottom-up approach for creating functional materials. In
particular, preparation of metallic nanostructures of a defined
geometry leads to materials with optical and electronic
properties that are expected to find applications as components
of future electronics and optical devices.1,2 A promising
approach toward creating such well-defined assemblies of
metal nanoparticles (NPs) having features smaller than what is
possible using top-down technologies is based on the use of
DNA as a scaffold.3−5 DNA is well-suited for construction of
nanostructures due to its programmable self-assembling
properties, a well-defined and rigid secondary structure, and
chemical stability to ambient conditions.6−9 As a result of these
favorable properties, increasingly complex two and three-
dimensional (2D and 3D, respectively) DNA nanostructures
have been realized.10−23 In turn these developments have
fostered efforts toward creating metal assemblies that use DNA
nanostructures as templates.
We recently reported that DNA containing boranephosph-
onate internucleotide linkages (bpDNA), as shown in Figure 1,
reduce many metal ions (AuCl4 , PtCl42−, and Ag+) and
−
produce the corresponding NPs.54 At the same time, previous
research has demonstrated that bpDNA forms standard
Watson−Crick base pairs.55−58 Thus the programmable
Watson−Crick binding properties can be used to incorporate
bpDNA at predetermined sites within a DNA structure, while
the reductive group will lead to metal deposition at these sites
upon exposure to metal salts. In the example outlined in Figure
1, only one of the two alternating tiles (shown as orange in this
example) that comprise this particular DNA array is
programmed to contain bpDNA (see also Figure 4). When
this array is exposed to a metallic salt, metal particles, as
illustrated by the golden colored spheres, will be deposited
selectively at the orange tiles to form a preprogrammed, linear
arrangement of NPs. Moreover because the concentration and
spatial distribution of the reductive boranephosphonate link-
ages can be freely varied even within the same bpDNA, a high
degree of control over the size and density of the metallic
features may be achieved.
DNA can be metallized by reduction of metal ions (Ag,24−28
Au,29,30 Pt,31 Pd,32,33 Ni,34 and Co;35 see ref 36 for a more
exhaustive list of references) bound to the DNA backbone
using exogenous agents, such as sodium borohydride or
covalently attached aldehyde groups.37−39 However these
methods do not utilize the underlying addressability available
within complex 2D and 3D DNA assemblies. Alternatively, gold
nanoparticles (AuNPs) can be attached to thiol functionalized
DNA. This allows site-specific immobilization of these particles
at locations within a DNA assembly that contain a
complementary sequence40−52 Unfortunately the AuNPs in
these experiments are negatively charged and cannot be placed
In spite of these promising attributes, the use of bpDNA for
constructing metal-NP assemblies has not been realized. This is
because none of the synthetic approaches reported pre-
Received: January 26, 2013
Published: April 4, 2013
© 2013 American Chemical Society
6234
dx.doi.org/10.1021/ja400898s | J. Am. Chem. Soc. 2013, 135, 6234−6241

Journal of the American Chemical Society
Article
Figure 1. Schematic depiction of selective metal deposition by controlling the location of bpDNA. In this example only the orange tiles contain
bpDNA. Metal deposition (golden yellow spheres) occurs at the orange tiles when the preassembled array is exposed to a metal salt. The
boranephosphonate linkage is oxidized to a phosphate without degradation of the backbone during this process.54
Figure 2. Synthesis of BIBS protected 2′-deoxynucleoside 3′-phosphoramidites. (i) Et3N, DMAP and 1,4-dioxane; (ii) 1.0 M NH3/MeOH; (iii)
DMT-Cl (1.2 equiv), diisopropylethylamine (2 equiv), pyridine; and (iv) P(OMe)(NiPr2)2 (1.2 equiv), tetrazole (1.1 equiv), and CH2Cl2..
viously59−64 can be used to synthesize bpDNA of sufficient
length and purity necessary for nanotechnology applications. In
this report, we present a new synthetic method that overcomes
these limitations. We demonstrate incorporation of bpDNA
into a 2D DNA nanoarray comprised of tiles containing double
crossover (DX) junctions.63 This particular type of tile was
chosen as a large number of DNA assemblies including
addressable 2D arrays23 and origami structures contain these
DX motifs. We last demonstrate site-specific, in situ deposition
onto these arrays of silver nanoparticles (AgNPs) simply by
treating with solutions containing Ag+.
borane reagents used to generate the boranephosphonate
linkages. Thus our challenge lay in developing a new type of
protecting group that was stable to borane as well as other
reagents used during 2′-deoxyoligonucleotide synthesis.
Recently Liang et al66 reported the protection of amines
using the di-tert-butylisobutylsilyl (BIBS) group. In contrast to
the lability of amines protected with nonhindered silanes,67,68
the BIBS group was shown to be quite stable. In addition no
reactivity towards borane was expected a priori. Thus we sought
to develop a synthetic scheme for preparing BIBS protected 2′-
deoxynucleoside 3′-phosphoramidites that could be used as
synthons for DNA and bpDNA synthesis.
Figure 2 depicts the synthetic scheme that was developed for
the preparation of BIBS protected 2′-deoxynucleoside 3′-
phosphoramidites. Starting with 5′,3′-di-O-acetyl 2′-deoxynu-
cleosides (1, 4, or 7), silylation of the exocyclic amines was
carried out using variations of the reported procedure.65 The
reaction of 5′,3′-di-O-acetyl-2′-deoxycytidine (1) with BIBS-
RESULTS
■
Synthesis of N-BIBS Protected 2′-Deoxynucleoside 3′-
Phosphoramidites. The synthesis of bpDNA is problematic
due to reduction of the amide groups that protect the
nucleobase exocyclic amines of adenine, guanine, and cytosine
during DNA synthesis.60 This reduction is caused by the
6235
dx.doi.org/10.1021/ja400898s | J. Am. Chem. Soc. 2013, 135, 6234−6241

Journal of the American Chemical Society
Article
Figure 3. (a) Synthetic cycle for DNA and bpDNA synthesis using BIBS protected phosphoramidites. (b) Lanes 1 and 2: bpdT21 synthesized using
Et3SiH as scavenger. This mixture was found to contain many degradation products. Lane 3: bpdT21 synthesized using TMPB as a trityl scavenger.
5′-O-DMT-2′deoxythymidine 3′-O-cyanoethyl N,N-diisopropylaminophosphoramidite was used for results shown in lane 1, whereas 5′-O-DMT-2′-
deoxythymidine 3′-O-methyl N,N-diisopropylaminophosphoramidite was used for results displayed in lanes 2 and 3. Lanes 4−6: Reactions mixtures
obtained from the synthesis of bpd(T8CT), bpd(T8AT), and bpd(T8GT), respectively, using BIBS protected phosphoramidites and TMPB as a
scavenger. (c) Denatuing PAGE analysis of reaction mixtures obtained from the synthesis of mixed based sequences bpdB1a (lane 1) and 3−
1bpdB1a (lane 2). (d) 31P NMR spectra of purified bpdB1a (top) and 3−1bpdB1a (bottom). Sequences of bpdB1a and 3−1bpdB1a are shown in
Figure 3d. Subscript ‘p’ and ‘b’ refer to a phosphate and boranephosphonate linkages, respectively.
OTf (Tf = triflate) in the presence of base proceeded in 2 h at
50 °C with good yield (67%), whereas 5′,3′-di-O-acetyl-2′-
deoxyadenosine (4) reacted more slowly and required heating
at 60 °C for 3 days in order to obtain the product in 25−30%
yield. These results are in accordance with the lower
nucleophilicity of the adenine amine. Use of stronger bases,
prolonged reactions times, and higher temperatures (80 °C)
did not improve yields. The reaction of 5′,3′-di-O-acetyl 2′-
deoxyguanosine (7) with 1.2 equiv of BIBS-OTf (60 °C) led to
substitution at the O6 position instead of the N2 amine
(Scheme S1 and Figure S4). However when repeated with a
larger excess (5 equiv) of BIBS-OTf (60 °C) for 3 days, the
bis(silyl) compound (8) was obtained in 74% yield.
and 9. Finally, the 3′ hydroxyl was phosphitylated using O-
methyl-bis(N,N-diisopropylamino)phosphoramidite activated
by tetrazole to yield synthons 11−13 in good yields (85−95%).
bpDNA Synthesis using BIBS-Phosphoramidites and a
Novel Trityl Quencher. Solid-phase DNA synthesis using the
phosphoramidite method is comprised of four steps, as shown
in Figure 3a. For bpDNA synthesis, the oxidation (step 3) of
the phosphite triester to phosphate (C1) with peroxide or
iodine is replaced by boronation with BH3 to yield a
trialkylphosphite borane linkage (C2). Since no examples of
N-silyl nucleobase protecting groups for DNA synthesis exist,
we sought to determine the stability of the BIBS groups during
solid-phase 2′-deoxyoligonucleotide synthesis by preparing
DNA containing only phosphate diester internucleotide
linkages. These experiments demonstrated that 12 and 13
could be used without modification of the standard conditions
Removal of the 5′ and 3′ acetyl groups on 2, 5, and 8 with
1.0 M ammonia in methanol was followed by protection of the
5′ hydroxyl as the dimethoxytrityl (DMT) ether to yield 3, 6,
6236
dx.doi.org/10.1021/ja400898s | J. Am. Chem. Soc. 2013, 135, 6234−6241

Journal of the American Chemical Society
Article
Figure 4. (a) Modified design of Tile B for incorporation of bpDNA. (b−d) AFM images of 2D arrays that incorporate the fully boronated B1a
(bpB1a) in the modified design. The x−y scale for images displayed in (b) and (c) are in μm, while the image in (d) is in nm. The stripes at higher
magnification result from the rows of vertical hairpin loops present in modified tile B.
(data not shown). However for 11 the 3% trichloroacetic acid
(TCA) in dichloromethane solution, which is routinely used for
detritylation, led to partial removal of the BIBS group during
each synthesis cycle (Figure S5a and S6). Further studies also
revealed that this effect was specific to TCA as BIBS removal
was not observed when a 0.5% trifluoroacetic acid (TFA)
solution (Figure S5b) was used. It is known that TCA binds
strongly to 2′-deoxyoligonucleotides on a solid support,69
which leads to a high local concentration of acid. Perhaps this
increased acid concentration causes an enhanced rate of acid-
mediated BIBS cleavage. In contrast TFA likely does not
interact similarly with the 2′-deoxyoligonucleotides. Irrespective
of the rational, substitution of TFA in the detritylation step
(Figure 3a, step 4) led to the preparation of several DNA
oligomers in high yield and purity (Figures S7 and S8 and
Table S2).
A second challenge that must be addressed for the successful
synthesis of bpDNA stems from the reaction of trityl cations,
generated in step 4 (Figure 3a), with the trialkyl phosphite
borane linked bpDNA (C2). This reaction degrades the
growing 2′-deoxyoligonucleotide chain.60,70,71 Although trie-
thylsilane has been described as an efficient trityl cation
scavenger,63 the synthesis of a fully boronated 2′-deoxythymi-
dine 21-mer, (bpdT21; Table S3) using an acid deprotection
mixture containing 50% (v/v) Et3SiH, led to significant
amounts of shorter length products (Figure 3b, lanes 1 and
2). We therefore reasoned that an analog of the trialkylphos-
phite-borane internucleotide linkage as found in C2 (Figure 3a)
would be a more effective quencher. Accordingly bpDNAs
bpdT21, bpd(T8AT), bpd(T8CT), and bpd(T8GT) were
synthesized using 10% (v/v) trimethylphosphite-borane
(TMPB; Figure 3a) as the trityl scavenger in 0.5% TFA/
CHCl3. Denaturing PAGE analysis of these reaction mixtures
showed the formation of single bands (Figure 3b, lanes 3−6)
for each oligomer. MALDI-TOF analysis after elution from the
gel confirmed synthesis of the expected products (Table S3).
In order to more thoroughly test this synthesis method, two
24-mer boranephosphonate DNAs (bpdB1a and 3−1bpdB1a)
that contained all four 2′-deoxynucleosides were synthesized
(see Figure 3d for the sequences of these bpDNAs). bpdB1a
was boronated at every internucleotide linkage, whereas 3−
1bpdB1a contained an alternating pattern of three phosphates
and one boranephosphonate. For 3−1bpdB1a, the phosphite
triester produced after the coupling step (Figure 3a; compound
B) was either oxidized to phosphate (C1, Figure 3a) with 1.0 M
tert-butyl hydroperoxide in CH2Cl2 or to boranephosphonate
(C2, Figure 3a) using a BH3/THF complex. All other
conditions remained the same. Denaturing PAGE analysis of
the reaction mixture obtained from the synthesis of bpdB1a
revealed a single product (Figure 3c, lane 1). However in the
case of 3−1bpdB1a (Figure 3c, lane 2) a more diffuse band,
which indicates the presence of failure products, was observed.
6237
dx.doi.org/10.1021/ja400898s | J. Am. Chem. Soc. 2013, 135, 6234−6241

Journal of the American Chemical Society
Article
These are likely due to incomplete oxidation as t-BuOOH is
known to be less efficient than iodine as an oxidant (iodine
leads to degradation of compound C2 and cannot be used for
synthesis of DNA containing boranephosphonate and
phosphate linkages). Nevertheless both bpDNAs were purified
by preparative gel electrophoresis and confirmed to be the
correct product by MALDI-TOF MS (Table S3) and 31P NMR
(Figure 3d). Thus by using a combination of BIBS protected
phosphoramidites, TFA for detritylation, and TMPB as a trityl
cation scavenger, mixed sequence bpDNAs of lengths far
greater than previously possible could be synthesized.
Incorporation of bpDNA into 2D DNA Arrays. Winfree
et al.10 were the first to describe the use of rigid DNA tiles
containing DX junctions for the assembly of 2D DNA lattices.
This DX motif has continued to be used in a number of DNA
nanostructures, including various origami-based assemblies72 as
well as complex and addressable 2D shapes.23 Because of the
many potential applications of this motif, we chose Winfree et
al’s 2D arrays as a test system for incorporation of bpDNA into
DNA-based assemblies. Also incorporation of bpDNA within
one of the alternating tiles would be expected to produce a
predictable pattern of metal NPs upon in situ reduction of Ag+
ions.
double helices, and deviations from the B-form helical
parameters can lead to failure of structure formation. However
our results show that the introduction of boranephosphonates
that contain a chiral P atom was well-tolerated, and no
discernible effect on formation of arrays was observed. Arrays of
similar quality were also produced with the 3−1bpdB1a
oligomer (Figure S9).
In situ Metallization of bpDNA Containing Arrays. In
order to produce assemblies of AgNPs, we first deposited the
DNA arrays that incorporated the bpdB1a strand on
transmission electron microscopy (TEM) grids. The arrays
were subsequently treated with a 20 mM tris-acetate buffer (pH
7.4) containing 0.5 mM AgNO3 and 10 mM Mg(OAc)2 (20 h
in a humidity chamber). Excess liquid was removed, and the
grids were washed three times with water. Deposition of metal
NPs at the sites containing bpDNA was expected to produce a
striped pattern of AgNPs. In addition previous reports have
shown that deposition of NPs can cause these flat arrays to roll
up into tubular structures.43,47 Consistent with these expect-
ations, transmission electron micrographs of these metalized,
bpDNA containing arrays (Figure 5 and S11) demonstrated
The original design10 contained two alternating tiles (tiles A
and B; Figure 4a). Each tile in turn was comprised of five
individual DNA sequences (A1−A5 and B1−B5). Upon
assembly, each DX tile displayed complementary sticky ends
on the corners which generates intertile contacts. This design
therefore allowed the formation of arrays with an alternating
pattern of tiles. The B1 and B5 sequences of tile B also
contained vertical hairpin loops as topographical markers for
AFM experiments. In order to incorporate bpDNA, the hairpin
loop segment on B1 was deleted, and the resultant strand was
divided into two individual oligomers that were 24 and 23
nucleotides in length (B1a and B1b, respectively). Fully
(bpdB1a) and partially (3−1bpdB1a) boronated versions of
B1a were synthesized (Table S3). Such a split design was
necessary because attempts to synthesize the entire B1
sequence (47 nucleotides without the hairpin loop) as a fully
boronated bpDNA yielded a hydrophobic material that was
insoluble in aqueous solution and could not be characterized.
Boranephosphonates are more lipophilic than phosphates,55
and therefore the hydrophobic nature of this bpDNA oligomer
was not surprising. Although it was possible to synthesize the
47-mer B1 strand as a phosphate-boranephosphonate oligomer,
the expected lower yields for such an oligomer (for instance see
gel pattern for 3−1bpdB1a which is 24 nucleotides in length)
dissuaded us from doing so.
For assembling the 2D arrays, 1 bpDNA and 10 DNA
strands were mixed together in a 20 mM tris-acetate (pH 7.4)
buffer containing 10 mM Mg(OAc)2. The arrays were annealed
by cooling the mixture from 94 °C to room temperature over 2
days in a 2 L water bath. AFM analysis of the annealed structure
(Figure 4b−d) revealed large, flat arrays that were several
micrometers in length and a few hundred nanometers in width.
An average spacing of 31 1 nm between the rows of hairpin
loops was obtained which is in good agreement with an
expected value of 33 nm. As a negative control we also annealed
a solution where all the strands except bpdB1a were included.
When examined by AFM, no arrays were observed which
confirms the requirement for the bpdB1a strand in order to
successfully assemble this 2D array. The rigid DX tiles are
known to be highly sensitive to the structure of the constituent
Figure 5. TEM micrograph from in situ reduction of Ag+ onto arrays
containing bpdB1a. (left) Schematic that shows how repulsions
between metal NPs could lead to folding of flat arrays into tubes. The
folding may occur with (green and blue dashed arrow) or without
offset (black dashed arrow) leading to spiral or stacked arrays of NPs.
The DNA array is omitted in the lower part of the figure for clarity.
(right) Both types of metal arrays are seen in the micrograph. The
inset shows a magnified image of rings of NPs in the stacked
conformation.
formation of tubes with parallel stacked rings of NPs as well as
spiral arrangements with either left- or right-handed helicity.
The different types of tubes are produced as the rolling up can
occur with or without any offset along the boundary that is
being sealed during tube formation. Tubes with different
diameters were also observed which reflects the starting widths
of the flat sheets.
Interestingly the distance between the rows of AgNPs in the
stacked ring conformation was measured to be 23 nm, which is
lower than the expected value of 33 nm (also see Figure S10).
Rothemund et al.73 have shown that formation of tubes from
these types of tiled arrays leads to a gentle bending of the helix
spanning the intertile crossovers. It is possible that in the
present case that the bending effect is exacerbated due to
splitting of the B1 strand into two segments which could
introduce a hinge site. The net effect of such an exaggerated
bending motion would be to bring the rows of tiles (and NPs)
6238
dx.doi.org/10.1021/ja400898s | J. Am. Chem. Soc. 2013, 135, 6234−6241

Journal of the American Chemical Society
Article
closer together. Irrespective of the possible reasons for the
decreased spacing, these results clearly show the site-specific
deposition of AgNPs by in situ reduction of Ag+ by bpDNA.
In order to understand the effect of the number of borane
groups (and therefore the reductive potential) on metal
reduction, arrays containing the 3−1bpdB1a were also treated
with the AgNO3 containing buffer described above. However in
these arrays no metallization was observed. This was also true
when higher concentrations of Ag+ (up to 5 mM), longer
incubation times (up to 3 days), or lower Mg2+ concentrations
(1 mM) were used. These results are consistent with our
previous finding54 that a dithymidine boranephosphonate does
not reduce silver at room temperature, but a fully boronated 21-
mer oligothymidine reacts with Ag+ ions to produce NPs. Thus
a minimum number of borane groups in close proximity
appears to be necessary in order to reduce enough silver ions to
lead to formation of a stable seed particle at room temperature.
When the Ag+ treatment was carried out at a higher
temperature (40 °C), uncontrolled metal deposition that
resulted in ill-defined metal NP arrays was observed (Figure
S12). In addition many unassociated NPs were also seen. These
observations suggest that at higher temperatures the formation
and growth of NPs do not occur solely on the arrays adhered to
the surface. It is likely that unincorporated 3−1bpdB1a
containing tiles that are more mobile at higher temperature
are responsible for this type of growth. In future experiments it
would be instructive to use bpDNAs that contain different
numbers of contiguous boranephosphonate linkages. Current
efforts are directed toward using such oligomers within DNA
nanoassemblies that assemble into discrete structures. These
structures can be purified postassembly, which would remove
artifacts that arise due to unincorporated bpDNA strands.
nate DNA. Subsequently we demonstrated the incorporation of
these oligomers into 2D DNA arrays. In situ reduction of metal
ions by bpDNA located at specific sites within these arrays
demonstrated that specific patterns of AgNPs could be
generated. Additionally bpDNA has been shown to possess
properties useful for biological applications as antisense and
antimir agents.54 Thus the development of these methods for
synthesizing bpDNA should stimulate more thorough inves-
tigations into their biological applications.
EXPERIMENTAL SECTION
■
Automated bpDNA Synthesis. Synthesis was carried out on an
ABI 394 Synthesizer. All syntheses were performed at a 0.2 μmol scale
using a 5′-DMT 2′-deoxythymidine joined to a low volume
polystyrene solid support via a succinate linkage. In addition to 11−
13, 5′-dimethoxytrityl-2′-deoxythymidine-3′-O-methyl-N,N-diisopro-
pylphosphoramidite (Glen Research) was used. For synthesis of 2′-
deoxyoligonucleotides, a standard 0.2 μmole synthesis cycle was used
with an increased coupling time of 120 s. A wash with methanol
following the detritylation step was added. 11 and 12 and
commercially obtained 5′-O-DMT-2′-deoxythymidine 3′-O-methyl
N,N-diisopropylaminophosphoramidite (Glen Research) were dis-
solved in anhydrous CH3CN, whereas 13 was dissolved in CH2Cl2 at a
concentration of 0.1 M. Detritylation was carried out using a 0.5%
solution of TFA in anhydrous CHCl3 that also contained 10% TMPB.
Solutions for boronation (0.05 M BH3-THF complex in THF) and
oxidation (1.0 M t-BuOOH in CH2Cl2) were prepared fresh prior to
use. Reagents for activation (ethylthiotetrazole) and capping were
purchased form Glen Research. A stepwise description of the synthesis
cycle is described in Table S1.
Deprotection was carried out in two steps: The solid support linked
2′-deoxyoligonucleotides were first treated with a 1.0 M solution of
disodium-2-carbamoyl-2-cyanoethylene-1,1-dithiolate in DMF for 1 h
followed by extensive washing with DMF and methanol. The resin was
then dried using a flow of argon. Subsequently these 2′-
deoxyoligonucleotides were desilylated by overnight fluoride treatment
(940 μL DMF + 470 μL Et3N + 630 μL Et3N·(HF)3). The resin was
washed repeatedly with DMF, Millipore water, and methanol and
dried with argon. The resin was then transferred to a glass vial and
suspended in 37% ammonia for 1−2 h, and the ammoniawas removed
by evaporation in a SpeedVac. The cleaved 2′-deoxyoligonucleotides
were dissolved in a 10% acetonitrile-water mixture and used for further
analysis and purification.
DNA Arrays. Preparation of DNA arrays was as described
previously10 except that the buffer was 20 mM tris-acetate (pH 7.6)
containing 10 mM Mg(OAc)2.
Atomic Force Microscopy. AFM images were obtained using the
Cypher (Asylum Research) atomic force microscope. Silicon nitride
levers with a nominal spring constant of 0.35 nm (Bruker) were used.
DNA arrays were deposited on freshly cleaved mica by adding 3 μL of
the array solution, waiting for 1−2 min and then adding 75 μL of
buffer (20 mM tris-acetate (pH 7.6) + 10 mM Mg(OAc)2). Imaging
was carried out under liquid using the same buffer.
Metallization of bpDNA Containing Arrays. 400 mesh carbon-
coated gold grids (Electron Microscopy Sciences) were cleaned by
glow discharge immediately prior to use. The grids were first treated
with a solution of 100 mM Mg(OAc)2 for 5 min, the liquid was
removed, and then the grids were washed with water. Next 2−4 μL of
a solution containing the DNA arrays was added to the grid and
allowed to stand for 10 min at room temperature. Subsequently the
liquid was removed using a Kimwipe, and the grid washed once with
water. At this stage 5 μL of the metallization solution (0.5 mM AgNO3
+ 10 mM Mg(OAc)2 + 20 mM tris-acetate (pH 7.6)) was added on
top of the grid, and the grid was incubated in a humidity chamber at
room temperature for 20 h. For incubation of the grids at 40 °C, a
procedure was developed in order to ensure that the liquid on the
grids did not dry out. In this step the grids were floated onto 50 μL
droplets of the metalizing solution that were put on top of a piece of
DISCUSSIONS
■
Here we have presented a method for in situ site-specific
reduction of metals within a DNA assembly. This process was
made possible by introducing a reductive group that is part of
the DNA backbone. Unlike procedures that use presynthesized
AuNPs, the resolution of the metallic features produced using
bpDNA nanostructures is limited only by the underlying
template. Moreover as the metal deposition step can be carried
out in situ, the need for extra synthetic and purification steps
required for AuNP conjugation to DNA oligomers is
eliminated. Lower yields due to unwanted aggregation of
DNA structures through binding of multiple assemblies to a
single AuNP or through aggregation of the NPs33,34 are also
avoided.
A key development that enabled the use of bpDNA in DNA
nanoassemblies was a new method for synthesizing bpDNA
using 5′-DMT protected N-BIBS derivatives of dA, dG, and dC
phosphoramidites. This represents the first reported use of a
nucleobase protecting group that contains a N−Si bond during
solid-phase 2′-deoxyoligonucleotide synthesis. The unique
properties of these synthons (stability to bases and anhydrous
acids as well a quantitative removal with mild fluoride
treatment) make them ideally suited for synthesis of various
base-sensitive DNA analogs.
CONCLUSIONS
■
We have introduced N-BIBS protected 2′-deoxynucleoside 3′-
phosphosphoramidites as a new synthon that can be used to
significantly extend our ability to synthesize boranephospho-
6239
dx.doi.org/10.1021/ja400898s | J. Am. Chem. Soc. 2013, 135, 6234−6241

Journal of the American Chemical Society
Article
Teflon tape. The grids were then incubated in a humidity chamber that
had been pre-equilibrated in an incubation oven set at 40 °C.
Following incubation, the excess liquid was removed from the grid,
and each grid was washed three times with 5 μL of water. The grids
were allowed to air dry and TEM was carried out using a CM100
transmission electron microscope (FEI, Inc., Hillsboro, OR) operating
at 80 kV.
(20) Andersen, E. S.; Dong, M.; Nielsen, M. M.; Jahn, K.; Subramani,
R.; Mamdouh, W.; Golas, M. M.; Sander, B.; Stark, H.; Oliveira, C. L.
P.; Pedersen, J. S.; Birkedal, V.; Besenbacher, F.; Gothelf, K. V.; Kjems,
J. Nature 2009, 459, 73−76.
(21) Douglas, S. M.; Dietz, H.; Liedl, T.; Hogberg, B.; Graf, F.; Shih,
W. F. Nature 2009, 459, 414−418.
(22) Dietz, H.; Douglas, S. M.; Shih, W. M. Science 2009, 325, 725−
730.
ASSOCIATED CONTENT
(23) Wei, B.; Dai, M.; Yin, P. Nature 2012, 485, 623−627.
(24) Braun, E.; Eichen; Sivan, U.; Ben-Yoseph, G. Nature 1998, 391,
775−778.
■
S
* Supporting Information
Detailed synthetic procedures, characterization data for
compounds reported, HPLC and MALDI-TOF data for
phosphate DNA sequences synthesized using BIBS-phosphor-
amidites as well as additional AFM and TEM images are
provided. This information is available free of charge via the
Internet at http://pubs.acs.org.
(25) Keren, K.; Krueger, M.; Gilad, R.; Benyoseph, G.; Sivan, U.;
Braun, E. Science 2002, 297, 72−75.
(26) Yan, H.; Park, S. H.; Finkelstein, G.; Reif, J. H.; LaBean, T. H.
Science 2003, 301, 1882−1884.
(27) Liu, D.; Park, S. H.; Reif, J. H.; Labean, T. H. Proc. Natl. Acad.
Sci. U.S.A. 2004, 101, 717−722.
(28) Park, S. H.; Barish, R.; Li, H. Y.; Reif, J. H.; Finkelstein, G.; Yan,
H.; LaBean, T. H. Nano Lett. 2005, 5, 693−696.
AUTHOR INFORMATION
■
(29) Patolsky, F.; Weizmann, Y.; Lioubashevski, O.; Willner, I. Angew.
Chem., Int. Ed. 2002, 41, 2323−2327.
Corresponding Author
marvin.caruthers@colorado.edu
(30) Aherne, D.; Satti, A.; Fitzmaurice, D. Nanotechnology 2007, 18,
125205.
Notes
(31) Ford, W. E.; Harnack, O.; Yasuda, A.; Wessels, J. M. Adv. Mater.
The authors declare no competing financial interest.
2001, 13, 1793−1797.
(32) Richter, J.; Mertig, M.; Pompe, W.; Monch, I.; Schackert, H. K.
Appl. Phys. Lett. 2001, 78, 536−538.
ACKNOWLEDGMENTS
■
We would like to thank Prof. Thomas Perkins (CU, Boulder)
and his laboratory members for generous use of their atomic
force microscope, sharing their expertise in imaging techniques,
and many helpful discussions. We also thank Prof Nadrian
Seeman (NYU) for helpful discussions and suggestions
throughout the project. This research was supported by the
University of Colorado.
(33) Geng, Y.; Liu, J.; Pound, E.; Gyawali, S.; Harb, J. N.; Woolley, A.
T. J. Mat. Chem 2011, 21, 12126−12131.
(34) Becerril, H. A.; Ludtke, P.; Willardson, B. M.; Woolley, A. T.
Langmuir 2006, 22, 10140−10144.
(35) Gu, Q.; Haynie, D. T. Mater. Lett. 2008, 62, 3047−3050.
(36) Gu, Q.; Haynie, D. T. DNA-Templated Nanowires: Context,
Fabrication, Properties and Applications In Annual Review of Nano-
Research; Cao, G., Brinker, C. J., Eds.; World Scientific Publishing
Company, Inc.: Hackensack, NJ, 2008; Vol. 2, pp 217−245.
(37) Keren, K.; Berman, R.; Braun, E. Nano Lett. 2004, 4, 323−326.
(38) Burley, G. A.; Gierlich, J.; Mofid, M. R.; Nir, H.; Tal, S.; Eichen,
Y.; Carell, T. J. Am. Chem. Soc. 2006, 128, 1398−1399.
(39) Wirges, C. T.; Timper, J.; Fischler, M.; Sologubenko, A. S.;
Mayer, J.; Simon, U.; Carell, T. Angew. Chem., Int. Ed. 2009, 48, 219−
223.
REFERENCES
■
(1) Maier, S. A.; Atwater, H. A. J. Appl. Phys. 2005, 98, 011101:1−
011101:10.
(2) Willets, K. A.; Van Duyne, R. P. Annu. Rev. Phys. Chem. 2007, 58,
267−297.
(3) Seeman, N. C. Annu. Rev. Biochem. 2010, 79, 65−87.
(4) Samano, E. C.; Pilo-Pais, M.; Goldberg, S.; Vogen, B. N.;
Finkelstein, G.; LaBean, T. H. Soft Matter 2011, 7, 3240−3245.
(5) Tan, S. J.; Campolongo, M. J.; Luo, D.; Cheng, W. Nat.
Nanotechnol. 2011, 6, 268−276.
(6) Seeman, N. C. Nature 2003, 421, 427−431.
(7) Lin, C.; Liu, Y.; Rinker, S.; Yan, H. 2006, 7, 1641−1647.
(8) Aldaye, F. A.; Palmer, A. L; Sleiman, H. F. Science 2008, 321,
1795−1799.
(9) Simmel, F. C. Angew. Chem., Int. Ed. 2008, 47, 5884−5887.
(10) Winfree, E.; Liu, F.; Wenzler, L. A.; Seeman, N. C. Nature 1998,
394, 539−544.
(11) Chen, J. H.; Seeman, N. C. Nature 1991, 350, 631−633.
(12) Yan, H.; Park, S. H.; Finkelstein, G.; Reif, J. H.; LaBean, T. H.
Science 2003, 301, 1882−1884.
(13) Shih, M. W.; Quispe, J. D.; Joyce, G. F. Nature 2004, 427, 618−
621.
(14) Rothemund, P. W. K.; Papadakis, N.; Winfree, E. PLoS Biol.
2004, 2, 2041−2053.
(15) Shih, M. W.; Quispe, J. D.; Joyce, G. F. Nature 2004, 427, 618−
(40) Alivisatos, A. P.; Johnsson, K. P.; Peng, X.; Wilson, T. E.;
Loweth, C. J.; Bruchez, M. P., Jr; Schultz, P. G. Nature 1996, 382,
609−611.
(41) Sharma, J.; Chhabra, R.; Andersen, C. S.; Gothelf, K. V.; Yan,
H.; Liu, Y. J. Am. Chem. Soc. 2008, 130, 7820−7821.
(42) Park, S. Y.; Lytton-Jean, A. K. R.; Lee, B.; Weigand, S.; Schatz,
G. C.; Mirkin, C. Nature 2008, 451, 553−556.
(43) Sharma, J.; Chhabra, R.; Cheng, A.; Brownell, J.; Liu, Y.; Yan, H.
Science 2009, 323, 112−116.
(44) Hung, A. M.; Micheel, C. M.; Bozano, L. D.; Osterbur, L. W.;
Wallraff, G. M.; Cha, J. N. Nat. Nanotechnol. 2010, 5, 121−126.
(45) Ding, B.; Deng, Z.; Yan, H.; Cabrini, S.; Zuckerman, R. N.;
Bokor, J. J. Am. Chem. Soc. 2010, 132, 3248−3249.
(46) Schreiber, R.; Kempter, S.; Holler, S.; Schuller, V.; Schiffels, D.;
Simmel, S. S.; Nickels, P. C.; Liedl, T. Small 2011, 7, 1795−1799.
(47) Shen, X.; Song, C.; Wang, J.; Dangwei, S.; Wang, Z.; Liu, N.;
Ding, B. J. Am. Chem. Soc. 2012, 134, 136−149.
(48) Pearson, A. C.; Liu, J.; Pound, E.; Uprety, B.; Woolley, A. T.;
Davis, R. C.; Harb, J. N. J. Phys. Chem. B 2012, 116, 10551−10560.
(49) Pearson, A. C.; Pound, E.; Woolley, A. T.; Linford, M. R.; Harb,
J. N.; Davis, R. C. Nano Lett. 2011, 11, 1981−1987.
(50) Pal, S.; Deng, Z.; Wang, H.; Zou, S.; Liu, Y.; Yan, H. J. Am.
Chem. Soc. 2011, 133, 17606−17609.
(51) Kuzyk, A.; Schreiber, R.; Fan, Z.; Pardatscher, G.; Roller, E.;
Hogele, A.; Simmel, F. C.; Govorov, A. O.; Liedl, T. Nature 2012, 483,
311−314.
621.
(16) Rothemund, P. W. K. Nature 2006, 440, 297−302.
(17) He, Y.; Ye, T.; Su, M.; Zhang, C.; Ribbe, A. E.; Jiang, W.; Mao,
C. Nature 2008, 452, 198−201.
(18) Yin, P.; Choi, H. M.; Calvert, C. R.; Pierce, N. A. Nature 2008,
451, 318−322.
(19) Yin, P.; Hariadi, R. F.; Sahu, S.; Choi, H. M. T.; Park, S. H.;
LaBean, T. H.; Reif, J. H. Science 2008, 321, 824−826.
6240
dx.doi.org/10.1021/ja400898s | J. Am. Chem. Soc. 2013, 135, 6234−6241

Journal of the American Chemical Society
Article
(52) Pilo-Pais, M.; Goldberg, S.; Samano, E.; LaBean, T. H.;
Finkelstein, G. Nano Lett. 2011, 11, 3489−3492.
(53) Sharma, J.; Chhabra, R.; Liu, Y.; Ke, Y.; Yan, H. Angew. Chem.,
Int. Ed. 2006, 45, 730−735.
(54) Roy, S.; Olesiak, M.; Padar, P.; McCuen, H.; Caruthers, M. H.
Org. Biomol. Chem. 2012, 10, 9131−9134.
(55) Li, P.; Sergueeva, Z. A.; Dobrikov, M.; Shaw, B. R. Chem. Rev.
2007, 107, 4746−4796.
(56) Sood, A.; Shaw, B. R.; Spielvogel, B. F. J. Am. Chem. Soc. 1990,
112, 9000−9001.
(57) Tomasz, J.; Shaw, B. R.; Porter, K.; Spielvogel, B. F.; Sood, A.
Angew. Chem., Int. Ed. 1992, 31, 1373−1375.
(58) Olesiak, M.; Krivenko, A.; Krishna, H.; Caruthers, M. H.
Phosphorus, Sulfur Silicon Relat. Elem. 2011, 186:4, 921−932.
(59) Sergueev, D. S.; Shaw, B. R. J. Am. Chem. Soc. 1998, 120, 9417−
9727.
(60) Higson, A. P.; Sierzchala, A.; Brummel, H.; Zhao, Z.; Caruthers,
M. H. Tet. Lett 1998, 39, 3899−3902.
(61) Zhang, J.; Terhorst, T.; Matteucci, M. D. Tet. Lett. 1997, 38,
4957−4960.
(62) McCuen, H. B.; Noe, M. S.; Sierzchala, A. B.; Higson, A. P.;
Caruthers, M. H. J. Am. Chem. Soc. 2006, 128, 8138−8139.
(63) Shimizu, M.; Saigo, K.; Wada, T. J. Org. Chem. 2006, 71, 4262−
4269.
(64) Fu, T.-J.; Seeman, N. C. Biochemistry 1993, 32, 3211−3220.
(65) Virta, P. ARKIVOC 2009, iii, 54−83.
(66) Liang, H.; Hu, L.; Corey, E. J. Org. Lett. 2011, 13, 4120−4123.
(67) Pratt, J. R.; Massey, W. D.; Pinkerton, F. H.; Thames, S. F. J.
Org. Chem. 1975, 40, 1090−1094.
(68) Overman, L. E.; Okazaki, M. E.; Mishra, P. Tet. Lett. 1986, 27,
4391−4394.
(69) Paul, C. H.; Royappa, A. T. Nucleic Acids Res. 1996, 24, 3048−
3052.
(70) Kawanaka, T.; Shimizu, M.; Shintani, N.; Wada, T. Bioorg. Med.
Chem. Lett. 2008, 18, 3783−3786.
(71) Matsumura, F.; Oka, N.; Wada, T. Org. Lett. 2008, 10, 1557−
1560.
(72) Castro, C. E.; Kilchherr, F.; Kim, D.; Shiao, E. L.; Wauer, P. W.;
Bathe, M.; Dietz, H. Nat. Methods 2011, 8, 221−229.
(73) Rothemund, P. W. K.; Ekani-Nkodo, A.; Papadakis, N.; Kumar,
A.; Fygenson, D. K.; Winfree, E. J. Am. Chem. Soc. 2004, 126, 16344−
16352.
6241
dx.doi.org/10.1021/ja400898s | J. Am. Chem. Soc. 2013, 135, 6234−6241