Two successive reactions on a DNA template: A strategy for improving background fluorescence and specificity in nucleic acid detection

Article abstract of DOI:10.1002/chem.201002426

We report a new strategy for template-mediated fluorogenic chemistry that results in enhanced performance for the fluorescence detection of nucleic acids. In this approach, two successive templated reactions are required to induce a fluorescence signal, r

Full text of DOI:10.1002/chem.201002426

DOI: 10.1002/chem.201002426
Two Successive Reactions on a DNA Template: A Strategy for Improving
Background Fluorescence and Specificity in Nucleic Acid Detection
Raphael M. Franzini and Eric T. Kool*[a]
Abstract: We report a new strategy for
template-mediated fluorogenic chemis-
try that results in enhanced perfor-
mance for the fluorescence detection
of nucleic acids. In this approach, two
successive templated reactions are re-
quired to induce a fluorescence signal,
rather than only one. These novel fluo-
rescein-labeled oligonucleotide probes,
termed 2-STAR (STAR=Staudinger-
triggered a-azidoether release) probes,
contain two quencher groups tethered
by separate reductively cleavable link-
ers. When a 2-STAR quenched probe
successively binds adjacent to two
mono-triphenylphosphine-(TPP)-
DNAs or one dual-TPP-DNA, the two
quenchers are released, resulting in a
fluorescence signal. Because of the re-
quirement for two consecutive reac-
tions, 2-STAR probes display an unpre-
cedented level of sequence specificity
for template-mediated probe designs.
At the same time, background emission
generated by off-template reactions or
incomplete quenching is among the
lowest of any fluorogenic reactive
probes for the detection of DNA or
RNA.
Keywords: fluorescence
· nucleic
acids · quenched probes · Stauding-
er reaction · template synthesis
Introduction
probes exhibited substantial background fluorescence, re-
sulting from reactions of the fluorogenic probe with water
or cellular components.
Hybridization-based fluorogenic probes enable the se-
quence-specific detection of nucleic acids in solution phase
and are of increasing interest for applications in biology and
medicine. Dual-probe detection schemes are especially ap-
pealing with this respect because the requirement of two si-
multaneous recognition events for the generation of a read-
out signal engenders the possibility of greater sequence spe-
cificity and reduced background fluorescence relative to
single probe approaches.[1,2] Template-mediated fluorogenic
reactions have been established as a promising class of dual-
probe nucleic acid detection schemes.[3] In this approach, ad-
jacent hybridization of two modified oligonucleotide probes
to the target nucleic acid initiates a chemical reaction,[4] elic-
iting a fluorescence turn-on signal. Templated reactive
probes can provide isothermal signal amplification by a cat-
alytic cycle of target hybridization, reaction, and dissocia-
tion.[5,6] Such probes can further be responsive to single-nu-
cleotide polymorphisms and have been employed in the dis-
crimination of bacteria species by single-nucleotide differen-
ces in rRNA.[7] However, for the detection of targets that
occur in low copy numbers it is essential to minimize back-
ground fluorescence, which can obscure the detection signal
in fluorogenic probes. For example, early templated reactive
To improve the characteristics of template-mediated fluo-
rescence activation schemes, several groups have investigat-
ed varied chemical reactions and fluorescence turn-on strat-
egies for the detection of nucleic acids in vitro[8] and in
cells.[7,9] One recent probe design that has displayed promise
in engendering low background and high sequence specifici-
ty is that of quenched Staudinger-triggered a-azidoether re-
lease probes (Q-STAR probes).[10] These probes consist of a
DNA hybridization sequence labeled with fluorescein,
which is quenched by a dabsyl group attached to the DNA
through an a-azidoether linker.
A triphenylphosphine-
(TPP)-labeled DNA probe reductively cleaves the linker, re-
leasing the quencher and eliciting a fluorescence turn-on
signal; adjacent binding of the two probes on the target
greatly accelerates this reaction. Recent probe designs in
general, and Q-STAR probes in particular, have significantly
improved key aspects of DNA-/RNA-template-mediated
fluorescence activation.[8,10] For example, nonspecific cleav-
age of the a-azidoether linker occurs slowly under cellular
conditions, providing very low background fluorescence.
Furthermore, Q-STAR probes provide an amplified detec-
tion signal, resulting from probe turnover on the target nu-
cleic acid. Because of these favorable properties, Q-STAR
probes are under further investigation for applications in
cellular and medical diagnostics. However, optimizing the
templated chemistry cannot alleviate all sources of back-
ground fluorescence. For example, bimolecular collisions of
the probes engender off-template reactions and incomplete
quenching results in residual fluorescence in the quenched
probes. Furthermore, as for virtually all hybridization
[a] Dr. R. M. Franzini, Prof. Dr. E. T. Kool
Department of Chemistry, Stanford University
Stanford, CA-94305-5080 (USA)
Fax : (+1)650-725-0295
E-mail: kool@stanford.edu
Supporting information for this article is available on the WWW
under http://dx.doi.org/10.1002/chem.201002426.
2168
ꢀ 2011 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Chem. Eur. J. 2011, 17, 2168 – 2175

FULL PAPER
probes, discrimination with single-base sensitivity can be
challenging for weakly destabilizing mismatches such as G–
T or G–A, resulting in selectivity that is lower than ideal.
Here, we describe a new design of reactive probes, aimed
at further lowering background signal and improving se-
quence specificity. This strategy, which may be generalizable
to many kinds of templated chemistries, requires two succes-
sive templated reactions to generate a fluorescence signal.
The reactive (2-STAR) probes are fluorescein-labeled oligo-
nucleotides, containing two 5’-terminal a-azidoether linkers,
each carrying a dabsyl quencher (Scheme 1), in contrast to
previous probe designs that contained a single linker and
quencher. The 2-STAR probes can be activated by two
quencher release reactions upon templated reductive linker
cleavage with two TPP–DNA probes, involving exchange of
the TPP–DNA conjugates (Scheme 1a). Alternatively, a
single activating probe, containing two TPP groups, can
induce 2-STAR activation in
a single binding event
Scheme 1. Illustration of the template-mediated fluorescence activation scheme based on probes, containing two releasable quenchers (2-STAR probes).
a) Template-mediated reaction of a 2-STAR probe with a DNA probe that contains a single triphenylphosphine (TPP) moiety cleaves the a-azidoether
linker, releasing the quencher, and forming triphenylphosphine oxide (TPPO). Exchange of two reacting TPP–DNA probes enables the consecutive re-
lease of both quenchers and fluorescence turn-on. b) Templated reaction of 2-STAR probe with a DNA probe, carrying two TPP groups, which cleaves
both linkers in one binding event. c) Molecular structures of 2-STAR probe (top) and dual-TPP modifications (bottom)
Chem. Eur. J. 2011, 17, 2168 – 2175
ꢀ 2011 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.chemeurj.org
2169

E. T. Kool et al.
(Scheme 1b). Importantly, the release of both quenchers is
required for the occurrence of a fluorescence signal and
mono-quencher intermediates generated by nonspecific re-
actions are non-fluorescent.
Two examples of bifunctional fluorogenic probes for tem-
plated nucleic acid detection have been previously reported.
Furukawa et al. have described probes based on a fluores-
cein profluorophore derivatized with two azidomethyl
groups.[9d] However, in that design a single templated reac-
tion, removing one azidomethyl group, generated a substan-
tial turn-on signal, which obviates some of the possible ben-
efits of doubly reactive probes. In a second report, our labo-
ratory has described fluorescent probes containing two
quenchers, but only one of the quenchers was displaced by a
templated reaction, whereas the second quencher dissociat-
ed by nonspecific hydrolysis.[11] The present probe design re-
quires for the first time two template-mediated reactions to
induce a fluorescence turn-on event. This condition greatly
lowers the probability of background-generating reactions
that can occur off-template and with mismatched targets.
The described 2-STAR probes provide a strong fluores-
cence turn-on signal by a rapid templated reaction with
TPP–DNAs and enable amplified detection of nucleic acid
targets by probe turnover. Moreover, the requirement for
two reactions greatly lowers the fluorescence signal arising
from off-template reactions and allows for an unprecedent-
ed degree of single-nucleotide specificity for templated reac-
tive probes.
Scheme 2. Synthesis of monomethoxytrityl-(MMt)-protected 5’-bis-
amino-modifier phosphoramidite 2. a) 1,3-diamino-2-propanol, dicyclo-
hexylcarbodiimide (DCC), cat. dimethylaminopyridine (DMAP), 86%
yield; b) 2-cyanoethyl-N,N-diisopropyl chlorophosphoramidite, diisopro-
pylethylamine (DIEA), 97% yield.
MALDI-TOF mass spectrometry (see the Supporting Infor-
mation, Table S1).
To evaluate the performance of 2-STAR probes in a
DNA-templated reaction scheme, we selected two 27-nt
DNA targets, differing by a single nucleotide. The sequences
correspond to a ribosomal RNA site that has close homolo-
gy (a single-nucleotide difference) in Escherichia coli and
Salmonella enterica (Table 1, EC-DNA and SE-DNA).[7a]
Table 1. Sequences of probes and templates.
Probe
DNA sequence[a]
2-STAR EC
2-STAR SE
5’-(DabAzL)2-AGTFlCGACA-3’
5’-(DabAzL)2-AGTFlAGACA-3’
mono-quencher SE 5’-DabAzl-AGTFlAGACA-3’
9m TPP–DNA
15m TPP–DNA
9d TPP–DNA
15d TPP–DNA
EC-DNA
5’-CAACCTCCA-TPP-3’
5’-AGGGCACAACCTCCA-TPP-3’
5’-CAACCTCCA-(TPP)2-3’
5’-AGGGCACAACCTCCA-(TPP)2-3’
5’-GATGTCGACTTGGAGGTTGTGCCCTTG-3’
5’-GATGTCTACTTGGAGGTTGTGCCCTTG-3’
Results and Discussion
Preparation of the reactive probes: The 2-STAR probes
were obtained by post-synthetic solid-phase conjugation of
two dabsyl-modified a-azidoether linkers to fluorescein-la-
beled DNAs with two terminal amine functionalities, follow-
ing a previously described protocol.[10] A 5’-bis-amino-modi-
fier was required for the attachment of two quencher release
linkers to the DNA probe. For this purpose, we designed
the monomethoxytrityl-(MMt)-protected 5’-bis-amino-modi-
fier 2, which was introduced as the terminal phosphorami-
dite during DNA solid-phase synthesis and deprotected on
solid support by repeated washes with 2% trichloroacetic
acid in dichloromethane. Compound 2 was conveniently pre-
pared in two steps and 83% overall yield by coupling two
molecules of N-MMt-3-aminopropionic acid[12] to 1,3-diami-
no-2-propanol followed by the formation of the phosphora-
midite (Scheme 2). The same 5’-bis-amino-modifier 2 was
also used for the preparation of the DNA probes with two
TPP groups (dual-TPP–DNAs, Scheme 1c) by post-synthetic
conjugation of 4-(diphenylphosphino)benzoic acid to 3’-
amine-modified DNA synthesized in the reverse (5’!3’) di-
rection. Probes that contained a single releasable quencher
(Q-STAR probes, prepared for comparison to the new
design) or a single TPP moiety (mono-TPP–DNAs) were
synthesized as previously described.[10] Probes were purified
by semi-preparative reverse-phase HPLC and analyzed by
SE-DNA
[a] TFl =fluorescein-labeled deoxythymidine, DabAzL=dabsyl containing
a-azidoether linker, EC=E. coli, SE=S. enterica.
The TPP–DNA probes were designed to bind adjacent to
the 2-STAR probe on the same target site. Two types of
TPP–DNAs were prepared, containing either one or two 3’-
terminal TPP groups. We prepared both short and long ver-
sions of TPP–DNA probes: the shorter probes (DNA 9-
mers) were designed to rapidly bind to and dissociate from
the target under the reaction conditions, whereas the longer
TPP–DNA probes (DNA 15-mers) were expected to bind
tightly to the target and dissociate slowly (or not at all)
under the same conditions. The following nomenclature is
used throughout the text: mono-TPP–DNA short probes
(9m); dual-TPP–DNA short probes (9d); mono-TPP–DNA
long probes (15m); dual-TPP–DNA long probes (15d).
Fluorescence activation of 2-STAR probes with mono-TPP–
DNAs: We initially tested the reactivity of the dual-quench-
er (2-STAR) probes. The 2-STAR probe (sequence EC,
100 nm) was co-incubated with the 9-mer (9m) TPP–DNA
(600 nm) in the presence of the complementary template
2170
www.chemeurj.org
ꢀ 2011 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Chem. Eur. J. 2011, 17, 2168 – 2175

Improving Nucleic Acid Detection
FULL PAPER
lower than that for 9m TPP–DNA. This finding confirms
that slow exchange of TPP–DNA probes impedes two suc-
cessive binding/reaction events and presumably causes sub-
stantial accumulation of non-fluorescent mono-quencher in-
termediates. This outcome provides strong evidence that
probe exchange is required for efficient fluorescence turn-
on of 2-STAR probes by single TPP–DNAs (Scheme 1a).
Having established that short mono-TPP–DNAs can acti-
vate the fluorescence of 2-STAR probes, we examined
whether the double quencher design enhances sequence se-
lectivity and reduces background fluorescence relative to
probes with a single quencher. To assess the sequence spe-
cificity, we monitored the fluorescence activation of 2-STAR
EC as a function of time by using SE-DNA as a single-mis-
match-containing template (Figure 1a and Figure S1 in the
Supporting Information). The effect of the single T–C mis-
match on the rate of fluorescence activation was dramatic.
After 115 min, the 2-STAR emission intensity generated by
9m TPP–DNA and the mismatched template reached only
(0.9ꢀ0.2)% of the value for the complementary EC-DNA
template. By comparison, the corresponding value for the
control single-quencher probes was (11.0ꢀ1.8)% for the
same mismatched target.[10] Therefore, double displacement
probes improve the single-mismatch specificity for this set
of probes by a factor of more than ten. Moreover, the fluo-
rescence generated without the template was as low as (or
slightly lower than) that with the singly mismatched tem-
plate (Figure 1a and Figure S1 in the Supporting Informa-
tion), demonstrating that double quencher release probes
minimize the background fluorescence associated with off-
template reactions of the probes. Furthermore, the fluores-
cence turn-on value of 2-STAR probes (emission of 2-STAR
probes after complete activation divided by emission prior
to addition of TPP–DNA) was remarkably high ((370ꢀ20)-
fold), substantially exceeding the values reported for other
quencher displacement probes including the control single-
quencher probes (61-fold).[10] The high fluorescence turn-on
value can be attributed to the presence of two dabsyl
groups,[13] which quench the fluorophore of 2-STAR probes
with >99.7% efficiency. Interestingly, this quenching effi-
ciency is higher than that reported for molecular beacon
probes containing two dabsyl groups (98.75%),[13] possibly
because of the different arrangement of the quenchers/fluo-
rophore in these probes.
Figure 1. Templated fluorescence activation of 2-STAR probes with reac-
tive probes containing a single TPP moiety. a) Kinetic analysis of tem-
plate-mediated fluorescence activation of 2-STAR EC by mono-TPP–
DNAs. Two TPP–DNA probes of different sequence length were investi-
gated at 378C: 9m TPP–DNA (black trace) and 15m TPP–DNA (grey
trace). Template dependence was assessed by incubating the probes with
the matched template EC-DNA (c), the mismatched template SE-
DNA (a) and without template (g). (For a Figure with increased y
axis see the Supporting Information). b) Comparison of single-nucleotide
discrimination of 2-STAR SE (c) and mono-quencher SE (a) for a
weakly destabilizing A–G mismatch. Sequence specificity was calculated
by dividing the fluorescence intensity for the reaction mediated by the
matched template SE-DNA by the fluorescence intensity for the reaction
mediated by the mismatched template EC-DNA. Error bars represent
the standard deviation of the mismatch sensitivity at the time point of
maximal specificity. Conditions:
cACHTNGUTREN(UNG 2-STAR)=100 nm, cACHTUNGTRNE(NUNG TPP–DNA)=
600 nm, c(template)=100 nm; tris-borate buffer (70 mm, pH 7.55) con-
ACHTUNGTRENNUNG
taining MgCl2 (10 mm), T=378C.
EC-DNA (100 nm). Fluorescence monitoring revealed that
the template efficiently mediated the reaction between 2-
STAR EC and 9m TPP–DNA, generating a strong increase
in fluorescein emission (lex =494 nm; lem =521 nm) (Fig-
ure 1a). The fluorescence activation trace exhibited a brief
delay time at the beginning of the experiment, consistent
with the requirement for two consecutive quencher release
events to occur and buildup of non-fluorescent mono-dabsyl
intermediates. Nevertheless, the overall rate of 2-STAR acti-
vation was rapid and reached 75% fluorescence activation
in less than 40 min.
These results establish that double quencher release
probes enhance the turn-on value of templated fluorescence
activation schemes, whereas greatly suppressing off-template
reaction signals. Moreover, the experiments with the T–C
mismatch-containing SE-DNA template established that
probes, relying on two quencher release events provide
greatly enhanced sequence discrimination by reducing the
signal from the mismatched template. We hypothesize that
the enhancement of sequence specificity for double quench-
er release probes is a stochastic effect; the requirement of
two binding/reaction events to occur for fluorescence activa-
tion magnifies the difference in total signal-generating rates
between the matched and mismatched templates.
To evaluate the role of probe exchange, we investigated
the activation of 2-STAR EC probes by the longer 15-mer
(15m) TPP probe (Figure 1a). The rate of fluorescence acti-
vation of 2-STAR EC by 15m TPP–DNA was considerably
Chem. Eur. J. 2011, 17, 2168 – 2175
ꢀ 2011 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.chemeurj.org
2171

E. T. Kool et al.
To provide an even more stringent test of selectivity, we
further assessed the potential of dual-quencher 2-STAR
probes to enhance the sequence-specificity for a weakly de-
stabilizing A–G mismatch in comparison to single-quencher
Q-STAR probes. We prepared a second 2-STAR probe (2-
STAR SE) complementary to the same target but specific to
the sequence of S. enterica and tested it in combination with
9m TPP–DNA to discriminate between the complementary
target (SE-DNA) and the corresponding sequence of E. coli
(EC-DNA), which contains the mismatch (Table 1). The
minor difference in thermal stability between the matched
(A–T) and the mismatched (A–G) renders the sequence se-
lectivity particularly challenging. Incubation of the control
mono-quencher SE probe (100 nm) with 9m TPP–DNA
(600 nm) and either the matched SE-DNA or the mis-
matched EC-DNA template (100 nm) provided only moder-
ate selectivity for this challenging mismatch; the fluores-
cence signal generated after 115 min in the presence of the
mismatched target EC-DNA reached (23ꢀ7)% of that of
the matched target (see the Supporting Information, Fig-
ure S3). In contrast, the EC-DNA template-mediated reac-
tion of 2-STAR SE was visibly decelerated relative to the
control reaction of Q-STAR SE, providing only (12ꢀ4)%
background signal relative to the matched target (see the
Supporting Information, Figure S2). To analyze this differ-
ence quantitatively over time, we plotted the sequence spe-
cificity of 2-STAR SE and of mono-quencher SE (fluores-
cence emission of the reaction with SE-DNA divided by
that with EC-DNA) as a function of time (Figure 1b). With
this challenging mismatch, the 2-STAR probes reached a
maximal sequence selectivity of (16ꢀ4)-fold after 29.5 min.
In comparison, the sequence selectivity for the mono-
quencher case peaked earlier (17.5 min) and maximized at a
value of only (7.8ꢀ1.3)-fold. Therefore, 2-STAR probes can
enhance single-mismatch selectivity even for challenging tar-
gets by more than a factor of two.
Figure 2. Kinetic analysis of template-mediated fluorescence activation of
2-STAR EC by DNA probes with two TPP modifications. Two dual-
TPP–DNA probes of different length were investigated: 9-mer (9d)
TPP–DNA (black traces) and 15-mer (15d) TPP–DNA (grey traces).
Template dependence was assessed by incubating the probes with the
matched template EC-DNA (c), the mismatch-containing template
SE-DNA (a) and without template (g). Conditions: c(2-STAR
EC)=100 nm,
cACTHNGUTERNNU(G TPP–DNA)=300 nm, cACHTGNUTREN(NGNU template)=100 nm; tris-borate
buffer (70 mm, pH 7.55) containing MgCl2 (10 mm), T=37 8C.
In contrast to the experiments with mono-TPP–DNAs,
which yielded extremely low background signal without
template (Figure 1a), dual-TPP–DNA probes generated a
considerable off-template background signal and lower se-
lectivity against a singly mismatched template (SE-DNA).
The off-template signal for dual-TPP–DNAs was 40-fold
higher than with mono-TPP–DNA probes after 115 min
(Figures 1a and 2); the mismatch specificity at the same
time point was only (2.3ꢀ0.7)-fold for 2-STAR EC and 9d
TPP–DNA. Decreased mismatch specificity and higher off-
template reaction for dual-TPP–DNA probes were antici-
pated because both quenchers of the 2-STAR probe can be
released in a single binding event. However, this back-
ground reaction significantly exceeded the level observed
previously for single-quencher Q-STAR probes with single
TPP–DNA.[10] Therefore, dual-TPP–DNA probes generate
increased background reaction by an additional, unknown
mechanism. One possible explanation for the high back-
ground signal for dual-TPP–DNAs is that the covalent imi-
nophosphorane ylide intermediate for the first Staudinger
reduction persists long enough to initiate the second reduc-
tion prior to dissociation. Alternatively, the two probes may
noncovalently associate because of the increased hydropho-
bicity of the dual-TPP modification.
Fluorescence activation of 2-STAR probes with dual-TPP–
DNAs: In principle, probes that carry two TPP groups
(dual-TPP–DNAs) could trigger the release of both quench-
ers and obviate probe exchange (Scheme 1b). Therefore, we
evaluated the ability of dual-TPP–DNA probes (9d or 15d,
Table 1) to activate 2-STAR probes in a template-dependent
configuration. Under the same conditions as for the templat-
ed reactions with mono-TPP–DNAs, incubation of 2-STAR
EC (100 nm) with 9d TPP–DNA (300 nm) and the EC-DNA
template (100 nm) provided a strong fluorescence turn-on
signal (Figure 2). Fluorescence activation was slightly more
rapid for 9d TPP–DNA than for 9m TPP–DNA, and no lag
phase was observable for 9d/15d TPP–DNAs. The experi-
ment with longer the 15d TPP–DNA probe generated simi-
lar kinetic traces relative to 9d TPP–DNA, which indicates
that fluorescence activation was independent of strand ex-
change. This result is in strong contrast to the mono-TPP ac-
tivator probes, which clearly required multiple binding
events for 2-STAR activation.
2-STAR probes provide amplified detection signal with re-
duced background reaction: In a previous study, mono-
quencher (Q-STAR) probes yielded signal amplification of
>75-fold by an isothermal catalytic cycle of target binding,
reaction, and dissociation.[6,10] In principle, signal amplifica-
tion should also be possible for 2-STAR probes. However,
the requirement for two sequential quencher release reac-
tions may decrease the rate of fluorescence activation and
amplification. Consequently, longer reaction times might be
necessary to engender amplified fluorescence activation for
2-STAR probes.
2172
www.chemeurj.org
ꢀ 2011 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Chem. Eur. J. 2011, 17, 2168 – 2175

Improving Nucleic Acid Detection
FULL PAPER
Figure 4. Comparison of the thermal stability of 2-STAR SE (&) and
Figure 3. Signal amplification for 2-STAR EC and 9m TPP–DNA with
substoichiometric concentrations of the target EC-DNA (c(EC-
DNA=100 (c), 20 (a), 4.0 nm (b), and no template (g), other
conditions are the same as described in Figure 1).
*
mono-quencher SE ( ) probes. The quenched probes were incubated in
buffer for 90 min at different temperatures. Fluorescence emission (y
axis) indicates the fluorescence intensity of the samples as a percentage
of fluorescence intensity of completely activated 2-STAR probes. Condi-
tions:
cACTHNGUTRENN(UG Q-STAR)=100 nm, cACHTUNGERTN(NUNG template)=100 nm; tris-borate buffer
(70 mm, pH 7.55) containing MgCl2 (10 mm).
To test the performance of 2-STAR probes under turn-
over conditions, we incubated 2-STAR EC (100 nm) and 9m
TPP–DNA (600 nm) with the EC-DNA template (100, 20,
and 4 nm) and measured the intensity of fluorescence as a
function of time (Figure 3). Within few hours, the fluores-
cence emission for the substoichiometric concentrations of
EC-DNA approached the level of complete probe activa-
tion, which establishes that 2-STAR probes are able to pro-
vide an amplified signal corresponding to at least fifteen
turnovers. Not surprisingly, the generation of an amplified
signal was slower for 2-STAR probes than for mono-
quencher probes, and this unfavorable tendency intensified
as the concentration of EC-DNA declined. However, this
effect was compensated by an overproportional decrease of
the signal associated with the off-template reaction between
2-STAR EC and 9m TPP–DNA relative to the mono-
quencher EC. After 12 h, the fluorescence of 2-STAR EC in
the absence of template reached only (3.0ꢀ1.0)% of the
value measured for complete conversion, compared to
(17.0ꢀ1.6)% for mono-quencher probes.[10] Thus, the data
show that 2-STAR probes provide an amplified reporter
signal with a higher signal-to-background ratio than corre-
sponding single quencher probes.
temperature, neither probe showed a significant increase in
background fluorescence over 90 min. In contrast, with
higher temperatures (T>508C) the fluorescence level in-
creased substantially. For example, mono-quencher SE incu-
bated for 90 min at 858C reached a fluorescence level corre-
sponding to (13.5ꢀ1.3)% probe activation, a degree that
could be problematic for certain analytical assays. The fluo-
rescence was significantly lower for 2-STAR SE, reaching
only (3.6ꢀ0.4)% conversion under the same conditions.
Thus, the data show that 2-STAR probes have reduced
background fluorescence at elevated temperatures and are
suitable for applications that require prolonged heating.
Conclusion
We have described a novel strategy to increase both the
signal-to-background ratio and the sequence specificity of
fluorescence DNA/RNA detection schemes that employ
templated chemistry. The design of the presented 2-STAR
probes requires that two successive quencher release reac-
tions with TPP–DNA activators occur before fluorescence
emerges, minimizing background signal. For example, an un-
intended reaction with a single TPP group releases one
quencher, but leaves a second quencher to maintain the
dark state of the probe.
Kinetic analysis revealed that activator DNA probes with
a single TPP molecule efficiently induced the fluorescence
of 2-STAR probes in a two-step mechanism in which two
mono-TPP–DNAs consecutively bind a target adjacent to
the 2-STAR probe and successively release both quenchers.
In general, the activation kinetics of 2-STAR was rapid and
enabled the generation of an amplified detection signal. In
contrast to mono-TPP–DNAs, dual-TPP–DNA probes rap-
idly activated the fluorescence of 2-STAR probes without
probe exchange. However, the dual-TPP activator probes
yielded a significant template-independent signal and poor
sequence selectivity.
Thermal stability of 2-STAR probes: Prolonged incubation
in aqueous solution potentially generates background fluo-
rescence for mono-quencher probes because of hydrolysis of
the a-azidoether groups.[14] Although the stability of such
probes is sufficient at physiological conditions, during this
study we observed that these probes release quenchers non-
specifically at elevated temperatures. Thermal instability
could preclude the application of previous mono-quencher
probes in detection assays that require elevated tempera-
tures or thermal cycling. To investigate whether double re-
lease probes can provide enhanced thermal stability, we
compared the fluorescence levels of 2-STAR SE and mono-
quencher SE probes measured after incubation in buffer for
90 min at various temperatures (Figure 4); the fluorescence
intensities were normalized to the fluorescence levels corre-
sponding to complete fluorescence activation. At ambient
Chem. Eur. J. 2011, 17, 2168 – 2175
ꢀ 2011 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.chemeurj.org
2173

E. T. Kool et al.
sopropyl chlorophosphoramidite (183 mg, 0.77 mmol) was added to a so-
lution of N,N’-bis-[3-(4-monomethoxytrityl)aminopropionyl]-1,3-diamino-
2-propanol (300 mg, 0.39 mmol) and diisopropylethylamine (168 mL,
0.97 mmol) in anhydrous MeCN (3 mL). The solution was stirred for
90 min at room temperature under an argon atmosphere and concentrat-
ed under vacuum. The residue was purified by silica column chromatog-
raphy (hexane/EtOAc/MeCN 7:3:1 + 2% TEA) to provide 2 as a white
foam (368 mg, 97%). 1H NMR (400 MHz, CDCl3): d=1.16–1.21 (m,
12H; CH3), 2.35–2.55 (m, 10H; CH2), 2.92–3.00 (m, 1H; CH), 3.08–3.19
(m, 1H; CH), 3.55–3.66 (m, 2H; CH2), 3.70–3.82(m, 10H; CH3 and
CH2), 3.88–3.98 (m, 1H; CH), 6.71–6.81 (m, 6H; Ar-H and NH), 7.13–
7.18 (m, 4H; Ar-H), 7.22–7.28 (m, 8H; Ar-H), 7.32–7.36 (m, 4H; Ar-H),
7.43–7.46 ppm (m, 8H; Ar-H); 13C NMR (400 MHz, CDCl3): d=20.45
(20.38), 24.72 (24.57), 37.30 (37.24), 39.96, 40.56, 43.16 (43.03), 55.12,
58.41 (58.20), 70.37, 113.06, 118.25, 126.17, 127.78, 128.49, 129.78, 138.04
(138.02), 146.19, 157.71, 172.74, 173.14 ppm (the values in parentheses
refer to the signals that are different for the second diastereomer);
HRMS [+ scan]: m/z: calcd for C58H70N6O6P: 977.5089; found: 977.5106.
The 2-STAR probes greatly reduced the fluorescence
background resulting from unwanted side reactions relative
to single quencher probes. This effect was evident for multi-
ple sources of background, including reactions mediated by
mismatched templates, off-template reactions between the
probes, and thermal hydrolysis of the a-azidoether linker.
Furthermore, two dabsyl molecules quench fluorescein more
efficiently than a single one[13] and reduce the inherent back-
ground fluorescence of the quenched probes considerably.
Together, these results clearly illustrate the potential of
double release probes to minimize background signal and
strongly enhance mismatch selectivity for fluorescence-
based genetic identification. The beneficial strategy of carry-
ing out two reactions on a template may be generalizable to
multiple classes of templated fluorogenic reactions.
Unmodified oligonucleotides: Oligonucleotides were synthesized on a
1 mmol scale by an ABI model 392 synthesizer using standard b-cyano-
AHCTUNGERTGeNNUN thylphosphoramidite coupling chemistry. Removal of the protecting
groups and cleavage from the CPG support were carried out by incuba-
tion in concentrated aqueous NH4OH solution at 558C for 14 h. The oli-
gonucleotides were purified by using Poly-Pak II cartridges. The identity
and purity of the probes was confirmed by MALDI-TOF mass spectrom-
etry (see the Supporting Information, Table S1).
Experimental Section
General: Anhydrous solvents were purchased from Fisher Scientific and
used without further purification. Chemical reagents were purchased
from either Sigma–Aldrich or Acros and used without further purifica-
tion. Reagents used for the solid-phase synthesis of oligonucleotides such
as phosphoramidites, solid supports, amino modifiers, and reagent solu-
tions for the synthesizer were acquired from Glen Research (Sterling,
VA, USA). 1H and 13C NMR spectra were recorded on either a Varian
Innova 500 MHz or a Varian Mercury 400 MHz NMR spectrometer.
High-resolution mass spectrometry analyses were performed by the UC
Riverside Mass Spectrometry Facility. Analytical and semi-preparative
high performance liquid chromatography was performed on a LC-CAD
Shimadzu liquid chromatograph, equipped with a SPD-M10A VD diode
array detector and a SCL 10A VP system controller and by using re-
verse-phase C18 columns. Fluorescence measurements were performed
on a Fluorolog 3 Jobin–Yvon fluorophotospectrometer equipped with an
external temperature controller. Oligonucleotide masses were deter-
mined by the Stanford University Protein and Nucleic Acid Facility by
Preparation of 2-STAR probes: Oligonucleotides were synthesized with
the 5’-bis-amino-modifier 2 appended to the 5’ terminus. The MMt pro-
tecting groups were removed on the DNA synthesizer by using alternat-
ing cycles of deprotection reagent (3% trichloroacetic acid in CH2Cl2)
and CH2Cl2 washes. The solid support was added to a solution containing
compound
2 (25 mm), benzotriazol-1-yl-oxytripyrrolidinophosphonium
hexafluorophosphate (PyBOP, 25 mm), and diisopropylethylamine
(50 mm) in DMF (250 mL) and gently shaken at room temperature for 5 h
protected from light. The DMF was decanted, the resin washed twice
with DMF and three times with MeCN, and dispersed in aqueous NH3/
MeNH2 deprotection/cleavage solution (1 mL) and incubated for 1 h at
558C. Beads were removed by filtration and the oligonucleotide probes
were purified by reverse phase HPLC. The purity of the Q-STAR probes
was assessed by analytical HPLC and found to be >95% with a minor
impurity that contains two dabsyl and one fluorescein molecules. Mono-
quencher Q-STAR probes were prepared as described previously.[10]
using
a Perspective Voyager-DE RP Biospectrometry MALDI-TOF
mass-spectrometry instrument with a 3-hydroxypicolinic acid/diammoni-
um hydrogen citrate matrix. Thermal stability experiments were per-
formed on a Flexstation II 384 microplate reader with a 96-well quartz
plate. Oligonucleotide concentrations were determined by UV absorb-
ance at 260 nm for denatured oligonucleotides (T=908C) by using linear
combinations of the nucleobases extinction coefficients.
Dual-TPP–DNA conjugates: Oligonucleotides were prepared by reverse
(5’!3’) synthesis by using the 5’-bis-amino-modifier 2 as the terminal
phosphoramidite for modification of the 3’ terminus. The MMt protecting
groups were removed on the synthesizer by using alternating cycles of
deprotection reagent (3% trichloroacetic acid in CH2Cl2) and CH2Cl2
washes. The solid support was added to a solution containing 4-(diphenyl-
phosphino)benzoic acid (0.1m), 1-ethyl-3-(3-dimethylaminopropyl) carbo-
diimide hydrochloride (0.1m), and diisopropylethylamine (0.2m) in DMF
(500 mL). Air trapped by the solid support was removed by briefly vac-
uumizing the mixture followed by backfilling with argon. The reaction
mixture was gently shaken for 2.5 h at room temperature. DMF was de-
N,N’-bis-[3-(4-Monomethoxytrityl)aminopropionyl]-1,3-diamino-2-propa-
nol (1): N,N’-dicyclohexylcarbodiimide (0.82 g, 4.0 mmol), 1,3-diamino-2-
propanol (0.16 g, 1.8 mmol), and a catalytic amount of 4-(dimethylami-
no)pyridine (50 mg) were added to a solution of N-(4-monomethoxytri-
tyl)-b-alanine[12] (1.45 g, 4.0 mmol) in anhydrous CH2Cl2 (10 mL). The
mixture was stirred for 14 h at room temperature. The formed precipitate
was eliminated by filtration and the product was purified by silica
column chromatography (hexane/EtOAc 2:3 + 0!5% MeOH + 2%
AHCTUNGTREGcNNUN anted, the resin washed twice with DMF and three times with MeCN,
dispersed in aqueous NH4OH/MeNH2 deprotection/cleavage solution
(1 mL) containing the oxygen scavenger tris-(2-carboxyethyl)phosphine
(4 mg) and incubated for 1 h at 558C. The solid support was removed by
filtration, the solution concentrated (60 min) on the Speed-Vac to
remove the volatile amines. Dual-TPP–DNAs were purified by semi-
preparative reverse-phase HPLC concentrated on the Speed-Vac, divided
in aliquots, flushed with argon and stored at ꢁ788C. Samples were used
within one month after preparation to ensure maximal reactivity of the
probes. Mono-TPP–DNAs were prepared as described previously.[10]
triethylamine (TEA)) to provide
1 as a white foam (1.2 g, 86%).
1H NMR (500 MHz, CDCl3): d=2.05 (brs, 2H; NH), 2.36 (t, 3J
(H,H)=
3
6.0 Hz, 4H; CH2), 2.46 (t, J
N
CH2), 3.39–3.45 (m, 2H; CH2), 3.76 (s, 6H; CH3), 4.05–4.18 (m, 1H;
3
3
CH), 6.80 (d, J
(H,H)=9.0 Hz, 4H; Ar-H), 6.96 (t, J
(H,H)=6.0 Hz, 2H;
3
3
NH), 7.17 (t, J
G
ACHTUNGTRENNUNG
Ar-H), 7.33 (d, 3J
G
ACHTUNGTRENNUNG
8.0 Hz, 8H; Ar-H); 13C NMR (500 MHz, CDCl3): d=37.13, 39.93, 42.86,
55.23, 70.52, 113.21, 126.38, 127.92, 128.53, 129.84, 137.85, 146.01, 157.94,
174.12 ppm; HRMS [+ scan]: m/z: calcd for C49H53N4O5: 777.4011;
found: 777.4014.
Kinetic analysis of 2-STAR fluorescence activation: 2-STAR probes
(100 nm) and the corresponding template (100 nm, unless stated different-
ly) were incubated at 378C in tris-borate buffer (70 mm, pH 7.55) con-
taining MgCl2 (10 mm). TPP–DNA (600 nm for mono TPP–DNA and
N,N’-bis-[3-(4-Monomethoxytrityl)aminopropionyl]-1,3-diamino-2-propa-
nol 2-cyanoethyl diisopropylphosphoramidite (2): 2-Cyanoethyl-N,N-dii-
2174
www.chemeurj.org
ꢀ 2011 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Chem. Eur. J. 2011, 17, 2168 – 2175

Improving Nucleic Acid Detection
FULL PAPER
300 nm for dual-TPP–DNA) was added and the fluorescence emission
(lex =494 nm, lem =521 nm) was measured as a function of time.
[5] Z.-Y. J. Zhan, D. G. Lynn, J. Am. Chem. Soc. 1997, 119, 12420–
12421.
[6] T. N. Grossmann, A. Strohbach, O. Seitz, ChemBioChem 2008, 9,
2185–2192.
[7] a) A. P. Silverman, E. T. Kool, Nucleic Acids Res. 2005, 33, 4978–
4986; b) A. P. Silverman, E. J. Baron, E. T. Kool, ChemBioChem
2006, 7, 1890–1894; c) G. P. Miller, A. P. Silverman, E. T. Kool,
Bioorg. Med. Chem. 2008, 16, 56–64.
Analysis of thermal stability of 2-STAR probes: A solution containing
either the mono-dabsyl or 2-STAR probe (100 nm) and the template
strand SE-DNA (100 nm) was incubated for 90 min without TPP–DNA in
tris-borate buffer (70 mm, pH 7.55) containing MgCl2 (10 mm) at the
specified temperature. The solutions were cooled for 10 min on ice. Fluo-
rescence signals were measured by using a 96-well microplate reader. As
a reference for complete conversion, the Q-STAR or 2-STAR probe
(100 nm) was incubated for 5.5 h at 258C with SE-DNA (100 nm) and 9m
TPP–DNA (600 nm).
[8] a) Y. Xu, N. B. Karalkar, E. T. Kool, Nat. Biotechnol. 2001, 19, 148–
152; b) S. Sando, E. T. Kool, J. Am. Chem. Soc. 2002, 124, 2096–
2097; c) J. Cai, X. Li, X. Yue, J. S. Taylor, J. Am. Chem. Soc. 2004,
126, 16324–16325; d) J. Cai, X. Li, J. S. Taylor, Org. Lett. 2005, 7,
751–754; e) T. N. Grossmann, O. Seitz, J. Am. Chem. Soc. 2006, 128,
15596–15597; f) S. Ogasawara, K. Fujimoto, Angew. Chem. 2006,
118, 4624–4627; Angew. Chem. Int. Ed. 2006, 45, 4512–4515;
g) Z. L. Pianowski, N. Winssinger, Chem. Commun. 2007, 3820–
3822; h) Y. Huang, J. M. Coull, J. Am. Chem. Soc. 2008, 130, 3238–
3239; i) R. M. Franzini, E. T. Kool, Org. Lett. 2008, 10, 2935–2938;
j) R. M. Franzini, E. T. Kool, ChemBioChem 2008, 9, 2981–2988;
k) A. Shibata, H. Abe, M. Ito, Y, Kondo, S. Shimizu, K. Aikawa, Y.
Ito, Chem. Commun. 2009, 6586–6588; l) D. Arian, E. Clꢁ, K. V.
Gothelf, A. Mokhir, Chem. Eur. J. 2010, 16, 288–295; m) E.
Jentzsch, A. Mokhir, Inorg. Chem. 2009, 48, 9593–9595; n) D. K.
Prusty, A. Herrmann, J. Am. Chem. Soc., 132, 12197–12199.
[9] a) S. Sando, E. T. Kool, J. Am. Chem. Soc. 2002, 124, 9686–9687;
b) H. Abe, E. T. Kool, Proc. Natl. Acad. Sci. USA 2006, 103, 263–
268; c) H. Abe, J. Wang, K. Furukawa, K. Oki, M. Uda, S. Tsuneda,
Y. Ito, Bioconjugate Chem. 2008, 19, 1219–1226; d) K. Furukawa,
H. Abe, K. Hibino, Y. Sako, S. Tsuneda, Y. Ito, Bioconjugate Chem.
2009, 20, 1026–1036; e) Z. Pianowski, K. Gorska, L. Oswald, C. A.
Merten, N. Winssinger, J. Am. Chem. Soc. 2009, 131, 6492–6497.
[10] R. M. Franzini, E. T. Kool, J. Am. Chem. Soc. 2009, 131, 16021–
16023.
Acknowledgements
We thank the National Institutes of Health for financial support
(GM068122). R.M.F. acknowledges a Gerhard Casper Fellowship from
Stanford University. We also thank the NSF (CHE-0639053) for support
of fluorescence instrumentation in the Department of Chemistry at Stan-
ford University.
[1] D. M. Kolpashchikov, Chem. Rev. 2010, 110, 4709–4723.
[2] Selected examples: a) R. A. Cardullo, S. Agrawal, C. Flores, P. C.
Zamecnik, D. E. Wolf, Proc. Natl. Acad. Sci. USA 1988, 85, 8790–
8794; b) A. Oser, G. Valet, Angew. Chem. 1990, 102, 1197–1200;
Angew. Chem. Int. Ed. Engl. 1990, 29, 1167–1169; c) S. M. Gryaz-
nov, R. Schultz, S. K. Chaturvedi, R. L. Letsinger, Nucleic Acids Res.
1994, 22, 2366–2369; d) S. Tyagi, U. Landegren, M. Tazi, P. M. Lizar-
di, F. R. Kramer, Proc. Natl. Acad. Sci. USA 1996, 93, 5395–5400;
e) P. L. Paris, J. L. Langenhan, E. T. Kool, Nucleic Acids Res. 1998,
26, 3789–3793; f) D. M. Kolpashchikov, J. Am. Chem. Soc. 2005,
127, 12442–12443; g) S. Hasegawa, G. Gowrishankar, J. Rao, Chem-
BioChem 2006, 7, 925–928; h) C. I. Stains, J. L. Furman, D. J. Segal,
I. Ghosh, J. Am. Chem. Soc. 2006, 128, 9761–9765; i) S. Nakayama,
L. Yan, H. O. Sintim, J. Am. Chem. Soc. 2008, 130, 12560–12561;
j) E. Mokany, S. M. Bone, P. E. Young, T. B. Doan, A. V. Todd, J.
Am. Chem. Soc. 2010, 132, 1051–1059.
[11] D. J. Kleinbaum, G. P. Miller, E. T. Kool, Bioconjugate Chem. 2010,
21, 1115–1120.
[12] G. Berube, V. J. Richardson, C. H. J. Ford, Synth. Commun. 1991,
21, 931–944.
[13] C. J. Yang, H. Lin, W. Tan, J. Am. Chem. Soc. 2005, 127, 12772–
12773.
[14] T. L. Amyes, W. P. Jencks, J. Am. Chem. Soc. 1989, 111, 7888–7900.
[3] a) A. P. Silverman, E. T. Kool, Chem. Rev. 2006, 106, 3775–3789;
b) T. Ihara, M. Mukae, Anal. Sci. 2007, 23, 625–629.
[4] X. Li, D. R. Liu, Angew. Chem. 2004, 116, 4956–4979; Angew.
Chem. Int. Ed. 2004, 43, 4848–4870.
Received: August 20, 2010
Published online: January 10, 2011
Chem. Eur. J. 2011, 17, 2168 – 2175
ꢀ 2011 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.chemeurj.org
2175