Turn On of a Ruthenium(II) Photocatalyst by DNA-Templated Ligation

Article abstract of DOI:10.1002/chem.201804283

Here, the synthesis of a Ru^II photocatalyst by light-directed oligonucleotide-templated ligation reaction is described. The photocatalyst was found to have tremendous potential for signal amplification with >15000 turnovers measured in 9 hours. A templated reaction was used to turn on the activity of this ruthenium(II) photocatalyst in response to a specific DNA sequence. The photocatalysis of the ruthenium(II) complex was harnessed to uncage a new precipitating dye that is highly fluorescent and photostable in the solid state. This reaction was used to discriminate between different DNA analytes based on localization of the precipitate as well as for in cellulo miRNA detection. Finally, a bipyridine ligand functionalized with two different peptide nucleic acid (PNA) sequences was shown to enable template-mediated ligation (turn on of the ruthenium(II) photocatalysis) and recruitment of substrate for templated photocatalysis.

Full text of DOI:10.1002/chem.201804283

DOI: 10.1002/chem.201804283
Full Paper
&
Photocatalyst Activation |Hot Paper|
Turn On of a Ruthenium(II) Photocatalyst by DNA-Templated
Ligation
Marcello Anzola and Nicolas Winssinger*[a]
Abstract: Here, the synthesis of a RuII photocatalyst by light-
directed oligonucleotide-templated ligation reaction is de-
scribed. The photocatalyst was found to have tremendous
potential for signal amplification with >15000 turnovers
measured in 9 hours. A templated reaction was used to turn
on the activity of this ruthenium(II) photocatalyst in re-
sponse to a specific DNA sequence. The photocatalysis of
the ruthenium(II) complex was harnessed to uncage a new
precipitating dye that is highly fluorescent and photostable
in the solid state. This reaction was used to discriminate be-
tween different DNA analytes based on localization of the
precipitate as well as for in cellulo miRNA detection. Finally,
a bipyridine ligand functionalized with two different peptide
nucleic acid (PNA) sequences was shown to enable tem-
plate-mediated ligation (turn on of the ruthenium(II) photo-
catalysis) and recruitment of substrate for templated photo-
catalysis.
Introduction
Templated reactions are accelerated by the high effective con-
centration of reagents achieved in a supramolecular assembly.
Oligonucleotides offer a particularly attractive platform for
templated reaction based on the fact that affinities of the
supramolecular interactions are readily tunable, they can be
multiplexed through unique oligonucleotide sequences and
have important applications.[1–4] This has stimulated significant
efforts in the area, with a growing number of reactions extend-
ing the breadth and scope of templated transformations.
Prominent examples include nucleophilic reactions,[5–7] olefina-
tions,[8–9] Staudinger reactions,[2,10–12] conjugate additions,[13]
native chemical ligations or acyl transfers,[14–17] and cycloaddi-
tions.[18–20] Important applications of these reactions include
the synthesis of small molecule libraries, nucleic acid sensing
or drug uncaging. The majority of the reactions are designed
to bring two reagents, each conjugated to unique oligonucleo-
tides, in proximity upon hybridization to a template (Figure 1).
Template turnover requires the exchange of both reagents.
Most examples of templated reactions do not surpass 100
turnovers relative to the template. We have shown that the
photocatalytic activity of [Ru(bpy)3]2+ type of complexes can
be harnessed to reduce different immolative linkers, enabling
substrate turnover with only one reagent exchanging on the
template. However, these studies clearly point to the fact that
Figure 1. Nucleic acid-templated transformation with: stoichiometric
amounts of exchanging reagents (top); active photocatalyst with exchange
of a single reagent (middle); formation of the active photocatalyst without
exchange of reagents (bottom).
substrate turnover is rate limiting.[21–23] Alternatively, a templat-
ed reaction could be used to generate an active catalyst. Tran-
sition-metal catalysts are capable of high turnover numbers.
Thus, a nucleic acid-templated formation of an active transi-
tion-metal catalyst from catalytically inactive precursors could
be used to convert substrates with high signal amplification
overcoming the rate-limiting step of templated reactions (sub-
[a] M. Anzola, Prof. Dr. N. Winssinger
Department of Organic Chemistry, NCCR Chemical Biology
Faculty of Science, University of Geneva
30 Quai Ernest-Ansermet, 1205 Geneva (Switzerland)
E-mail: Nicolas.winssinger@unige.ch
Supporting information and the ORCID identification number(s) for the
author(s) of this article can be found under:
https://doi.org/10.1002/chem.201804283.
Chem. Eur. J. 2018, 24, 1 – 10
1
ꢀ 2018 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
&
&
These are not the final page numbers! ÞÞ

Full Paper
strate turnover/product inhibition). Surprisingly, this templated cyanate [Ru(bpy)2(phen-NCS)]2+ is commercially available and
reaction format has not received much attention despite some its conjugation to oligonucleotides is straightforward.[21,23,28,29]
previous achievements in the area. One of the first examples While there are more examples of photocatalysis in the litera-
of templated reactions extending beyond questions related to ture with [Ru(bpy)3]2+ than [Ru(bpy)2(phen)]2+ and its related
prebiotic chemistry was reported by Sheppard in 2001.[24] In analogs, both complexes have very close redox potentials and
this example, a metallosalen complex was prepared in a DNA- photophysical properties; hence, the well-established photoca-
templated fashion and shown to be effective in DNA-cleavage talysis of [Ru(bpy)3]2+ can be extrapolated to [Ru-
reactions. Subsequently, Herrmann demonstrated that
a
(bpy)2(phen)]2+. These complexes are generally prepared from
Pd(PPh3)2 catalyst could be generated from two oligonucleo- cis-dichlorobis(2,2’-bipyridine)ruthenium(II) by substitution with
tides coupled to triphenylphosphine that can coordinate to the third bidentate ligand (bipyridine or phenanthroline). Al-
palladium.[25] However, due to competing coordination from though such displacement reactions could be envisioned to be
nucleobases, the chemistry had to be performed under acidic performed in a templated fashion, the high temperatures re-
conditions (pH 5). Tris(bipyridyl)ruthenium(II) complexes are re- quired (typically performed at >1008C) seemed excessive in
markably stable and have enjoyed a renewed interest in the the context of nucleic acid-templated processes. On the other
scope of transformations they can promote.[26] Such complexes hand, it is known that monodentate ligands on bis(bipyridi-
are currently used in humans for photodynamic therapy[27] and ne)ruthenium complexes can exchange under mild conditions
we have shown that ruthenium-photocatalyzed reactions can following photoexcitation at 450 nm.[30–31] This ligand exchange
be performed in cells as well as in living organisms.[21,28] Here, has indeed been used to uncage bioactive compounds in
we demonstrate that a surrogate of tris(bipyridine)rutheniu- animal models.[32] Thus, we reasoned that, although the tem-
m(II) complex can be prepared in a templated fashion, result- plated ligation to form the catalyst would be challenging
ing in a turn-on of the ruthenium complex photocatalytic under thermal conditions, it should proceed under 450 nm irra-
properties (Figure 1).
diation at physiological temperatures. We set out to prepare
ruthenium(II) complexes 1 and 2 with an alkyne on the phe-
nanthroline for conjugation to peptide nucleic acids sequences
(PNA, Scheme 1). The synthesis started with ruthenium trichlor-
Results and Discussion
We had previously used a [Ru(bpy)2(phen)]2+ complex as a ide that was reacted with 1,4-cyclohexadiene to obtain
substitute for [Ru(bpy)3]2+, based on the fact that the isothio- [RuCl2(h-C6H6)]2. This complex was engaged in two successive
Scheme 1. A) Synthetic route for ruthenium(II) complexes. i) EtOH, microwave irradiation, 1308C, 4 min; ii) 1 equiv bpy, MeCN, microwave irradiation, 1208C,
10 min; iii) 1 equiv 5-ethnyl-phenanthroline, DMF, 908C, 8 h. iv) excess pyridine, MeOH/H2O 4:1, 908C, 9 h; v) MeOH/H2O 4:1, 908C, 9 h; vi) BnN3, DMF, CuBr,
TBTA, RT, 1 h. B) General structures of the PNA sequences used in the experiments. Black spheres indicate the optional [2-(2-aminoethoxy)ethoxy]acetic acid
spacer. R=H, CH2OH.
&
&
Chem. Eur. J. 2018, 24, 1 – 10
www.chemeurj.org
2
ꢀ 2018 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
ÝÝ These are not the final page numbers!

Full Paper
substitution reactions with bipyridine and 5-ethynyl-1,10-phe-
nanthroline (commercially available or readily prepared from
phenanthroline in three steps)[33] to afford complex 1. The two
remaining chlorides were further substituted with pyridines to
afford complex 2, which was considered a better starting ma-
terial for templated reactions based on its stability and the use
of related complexes in biological settings.[32] In parallel, we
prepared the templated-ligation product 3 by using 5’-acetam-
ido-(2,2’-bipyridine)-5-carboxamide [prepared from 5’-amino-
(2,2’-bipyridine)-5-carboxylic acid]. Complex 2 was found to
conjugate smoothly to PNAs functionalized with an azide by
copper(I)-catalyzed alkyne-azide cycloaddition (CuAAC). The bi-
pyridine-PNA adduct was prepared by standard coupling con-
ditions.
It is known that [Ru(bpy)3]2+ and its analogues can be used
as photocatalysts in the reduction of N-alkyl-pyridinium pro-
tecting groups with subsequent liberation of a caged function-
al group (ester, carbonate, carbamate, phenolic ether).[34,35]
Using a pyridinium-caged coumarin,[23] the photocatalytic activ-
ity of the precatalyst (complex 1 or 2) was tested, and com-
pared to the catalyst resulting from ligation (complex 3). As
anticipated, the two precatalysts were found to be catalytically
inactive towards the reduction of a caged fluorophore, where-
as complex 3 led to coumarin liberation by photocatalytic pyri-
dinium uncaging (Figure 2A, B). We then tested the per-
formance of catalyst 3 at high substrate concentration and low
catalyst loading to maximize the turnover number.
The reaction was carried out in deaerated phosphate buf-
fered saline (PBS), to prevent oxygen quenching and take full
advantage of the efficiency of the catalyst. Impressively, com-
pound 3 was found to be capable of more than 15000 turn-
overs in 9 hours (Figure 2C). Having demonstrated that only
the tris-bidentate chelated RuII complex 3 is suitable as a cata-
lyst, we next investigated the light-promoted and templated
formation of the catalyst. PNA sequences complementary to
the DNA template were prepared according to the previously
developed Mtt/Boc-based protection strategy with g-modified
PNA monomers[36] (see Table S1 in the Supporting Information
for the full list of sequences and PNA characterization for ex-
plicit structures). PNA 10-mers were used because this length
is sufficient to generate duplexes with DNA that are suitably
long-lived for templated reactions to proceed. As shown in
Scheme 1, complex 2 was introduced at the N-terminus of a
PNA sequence, whereas the bipyridine ligand was coupled to
the C-terminus of another PNA sequence. To discover the most
favorable preorganization of the reagents for a ligation, follow-
ing hybridization, PNAs were prepared with and without a
short polyethlenglycol (PEG) spacer (9 atoms, À9.46 ꢁ). In addi-
tion, different bpy-PNA sequences were prepared to leave a
gap of 1– 4 unhybridized nucleobases between the hybridiza-
tion sites of PNAs on the DNA template. The ligation reactions
were performed at 408C, by irradiation of the mixture with
blue light (455 nm, LED lamp, 1 W). Based on the difference in
luminescence spectra between complexes 2 and 3, we rea-
soned that we could follow the course of the reaction spectro-
scopically. Indeed, a clear bathochromic shift was observed
after 1 h of reaction, with emission maxima shifting from 600
Figure 2. A) Schematic representation of the photocatalytic cleavage of
caged coumarin 4 by different RuII complexes 1–3 at 10% catalyst loading.
B) Kinetic plots of the reaction with complexes 1–3. C) Photocatalytic cleav-
age of caged coumarin 4 by catalyst 3 at different concentrations after
9 hours (the calculated turnover number is indicated at the top of each bar;
350 mm of 4; 10 mm sodium ascorbate; LED 455 nm, 1 W; Buffer: PBS 1x
pH 7.4, 0.05% Tween 20 and 0.1% formamide. lex: 360 nm, lem: 460 nm).
to 630 nm and exaltation of emission. Hence, the reactions
were conveniently monitored by observing the increase of
emission at 630 nm as a function of time (Figure 3A,B). Using
this analysis, we compared the rates for the different ligation
permutations (see Figure S1 in the Supporting Information for
detailed kinetics). When comparing the size of the gap of un-
hybridized nucleobases for the series of reactions without PEG
spacers, the reactions had comparable rates for a gap of 1–3
nucleobases, with a sharp decrease for a gap of 4 nucleobases.
This suggests a cutoff point at which the size of the gap sur-
passed the distance between the linker and reagents required
for an efficient reaction.
The reactions with the PNAs bearing a PEG spacer were
found to proceed slower than the corresponding reaction
without PEGs, which is consistent with a reduced level of pre-
organization in the reactants. Based on these results, further
reactions were carried out with probes lacking PEGs designed
to leave a gap of three nucleobases between hybridization
Chem. Eur. J. 2018, 24, 1 – 10
www.chemeurj.org
3
ꢀ 2018 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
&
&
These are not the final page numbers! ÞÞ

Full Paper
agents and products on the gel. A marginal band correspond-
ing to the product was also observed in lane c (Figure 3C) in
the absence of template. The small amount of product ob-
served under these conditions is attributed to unspecific hy-
bridization of the PNA-Ru with bpy-PNA leading to product
formation.
We next turned our attention to the use of the ligated prod-
uct in catalytic reactions for nucleic acid sensing. To leverage
the benefit of the high turnover potential of ruthenium photo-
catalysis, the reactions should be performed at high substrate
concentrations. Thus, it is very important to have a reporter
system with a very high turn-on ratio such that the product
formed is not overshadowed by the large concentration of
substrate. We had previously made use of a quinazolinone pre-
cipitating dye (QPD) with an azide-triggered immolation.[37]
Contrary to other organic dyes, QPD is strongly fluorescent in
the solid state and further benefits from a large Stokes shift
and high photostability. The fluorescence in the solid state is
given by an intramolecular proton transfer in the excited state
(ESIPT). Considering that the phenolic hydrogen is essential for
the fluorescence of QPD, the replacement with a caging group
precludes any fluorescence. The turn-on performance of this
system is thus far superior to regular caged fluorophore such
as coumarin 4. Based on the more recent discovery that pyridi-
nium-based immolative linker provides a much faster uncaging
than the azide-based immolative linker,[22] we prepared the
new caged QPD 5 (Scheme 2). The pyridinium caging group
was found to render the derivative highly soluble in water and
non-fluorescent. These properties enable the use of high con-
centrations of substrate to maximize the rate of ruthenium(II)-
photocatalyzed uncaging without compromising the signal
with the background of a profluorophore. This new caged
QPD complements the prior caged-QPDs that have been de-
veloped as probes for phosphatases,[38,39] glycosidases,[40] pepti-
dases.[41]
Figure 3. A) Schematic representation of the DNA-templated ligation reac-
tion. B) Luminescence spectra recorded at different ligation times. lex
450 nm for n=3. C) Silver stain SDS-PAGE of ligation reaction after 60 mi-
nutes and HRMS of lane a. 5 mm PNA-Ru, 5 mm bpy-PNA3, 5 mm DNA 1 or
Ctrl DNA, 408C, LED 455 nm 1 W. Buffer: PBS 1x pH 7.4, 0.05% Tween 20
and 0.1% formamide.
:
sites. It is interesting to note the decrease in luminescence
(Figure 3B) at the very beginning of the reaction, which sug-
gests an exchange between one of the two pyridine mono-
dentate ligands of the ruthenium(II) complex with water as the
first step prior to ligation.[32] Based on the titration curve of
complex 3, the yield of ligation was estimated to be >85%.
Performing the reaction in the absence of either template,
light, or the PNA bearing the bipyridine ligand, failed to yield a
bathochromic shift (Figure S2, Supporting Information), sup-
porting that the observed ligation is a light-mediated and tem-
plated reaction. To unambiguously characterize the ligation
product, the system was further analyzed by sodium dodecyl
sulfate polyacrylamide gel electrophoresis (SDS-PAGE) of the
crude reactions and HRMS (Figure 3C). A bright stain for a
higher molecular weight product was obtained only for the
condition including the template and 450 nm light irradiation.
HRMS analysis indeed showed the mass of the ligated product.
Notably, the reaction had to be performed at relatively high
concentration (5 mm; templated reaction with 10-mer PNAs are
designed to operate at nm concentrations) to visualize the re-
Scheme 2. Synthesis of caged QPD 5. i) K2CO3, DMF, 608C, 6 h; ii) nPrOTf,
DCM, À808C to RT, 1 h; iii) 1 equiv 2-amino-5-chlorobenzamide, TsOH, EtOH,
reflux, 6 h; iv) DDQ, 08C to RT, 1 h.
&
&
Chem. Eur. J. 2018, 24, 1 – 10
www.chemeurj.org
4
ꢀ 2018 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
ÝÝ These are not the final page numbers!

Full Paper
The attractive properties of this dye (photostability, spatial
contrast) have already inspired its use in nucleic acid imaging
using an alkaline phosphatase conjugates in fluorescent in situ
hybridization (FISH).[42–43] We reasoned that for soluble assays,
confining the product of the templated reaction on a bead
would localize the precipitate and enhance its detection. In
our first design, streptavidin–agarose beads were loaded with
a biotinylated version of bpy-PNA3 (same sequence as shown
in Figure 3) with a loading of 20 pmolmLÀ1 of beads. The PNA-
containing ruthenium(II) precatalyst 2 (PNA-Ru, same sequence
as shown in Figure 3) was added to the solution, and then
matching DNA 1 was added at different concentrations. After a
two-hour incubation, ligation was performed at 408C using
455 nm LED irradiation. Beads were then placed under a micro-
scope for visualization and the samples were irradiated with
the laser of the microscope (argon laser, 488 nm) in the pres-
ence of profluorophore 5 and sodium ascorbate as the sacrifi-
cial reducing agent (Figure 4A).
Fluorescence of the precipitated QPD was assessed by two-
photon (2P) excitation microscopy (730 nm 2P excitation, 500–
550 nm emission). Although laser irradiation extended far out-
side the beads, we observed bright fluorescent precipitates
only on the beads, and the precipitates were still detectable
down to 0.01 equivalent of template (Figure 4B; see Figure S4
in the Supporting Information for the complete set of images).
The quantification showed a direct correlation between the
measured fluorescence values and the amounts of template
used. The fluorescence of the beads without template or with
1 equivalent of non-complementary DNA was found to be sig-
nificantly different from the lowest concentration tested (Fig-
ure 4B). It should be noted that because the detection can be
confined to
a single bead of 70 mm, the detection of
0.01 equivalent of DNA analyte by the PNA capture probe rep-
resents a calculated detection threshold of 36 amol (amol=
10À18 mol) of DNA per bead (estimate based on a bead loading
of 20 pmolmLÀ1 and bead volume of 0.18 nL) and this detec-
tion threshold is independent of the volume used because it is
confined to the bead. Considering that the formation of the
precipitate is confined to the bead bearing the ligation prod-
uct, we reasoned that beads of distinct size and/or shape
could be derivatized with different PNA sequences to discrimi-
nate between multiple analytes, although the same detection
reaction is used (QPD uncaging). To investigate the feasibility
of this one-color multi-sequence detection, that is, to discrimi-
nate template sequences based on beads size, streptavidin
agarose beads were filtered to separate beads with diameter
bigger than 90 and smaller than 70 mm. The bigger beads
were loaded with the biotinylated version of bpy-PNA3 previ-
ously used and small beads with a biotinylated version of a
bpy-PNA conjugate that hybridizes to miRNA-21 (bpy-PNA5,
see Table S1 in the Supporting Information for sequence and
explicit structure). Equal quantities of the beads were mixed
and incubated alternatively with two different synthetic DNA
templates in the presence of both the PNA-Ru targeting the
different DNA sequences. Subsequent ligation and laser irradia-
tion clearly showed the segregation of the precipitate depend-
ing upon the DNA sequence that was used. The non-comple-
Figure 4. A) Scheme and representative images of QPD precipitation on
beads. 100 nm PNA-Ru, 50 mm 5, 10 mm sodium ascorbate, irradiation for
5 min with 488 nm laser, 100% power. Buffer: PBS 1x pH 7.4, 0.05%
Tween 20 and 0.1% formamide. lex: 730 nm (2P excitation) 20%, lem: 500–
550 nm. Scale bar: 50 mm. Fluorescence is shown as a Z-maximum projec-
tion. B) Quantification of fluorescent precipitate on the beads at different
template loading. N=3, two-tailed unpaired t-tests. * p<0.05, ns: non-signif-
icant.
mentary PNA-Ru probes also served as an internal negative
control (Figure 5).
We next investigated the possibility to apply the same
chemistry to image cellular nucleic acids. Prior experiments
showed that the use of a pyridinium-caged fluorophore is
compatible with fixed cells conditions.[44] We chose to use
MCF-7, a breast adenocarcinoma cell line that expresses mod-
erately high level of miRNA-21, as a model.[45] Specifically, PFA-
fixed cells were incubated with the corresponding PNA probes
Chem. Eur. J. 2018, 24, 1 – 10
www.chemeurj.org
5
ꢀ 2018 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
&
&
These are not the final page numbers! ÞÞ

Full Paper
Figure 5. Representative images of DNA segregation on beads. 100 nm PNA-
Ru and PNA6-Ru, 100 nm DNA 1 (top) or DNA 21 (bottom), 50 mm 5, 10 mm
sodium ascorbate, irradiation for 1 min with laser line 488 nm, 100% power.
Buffer: PBS 1x pH 7.4, 0.05% Tween 20 and 0.1% formamide. lex: 730 nm (2P
excitation) 20%, lem: 500–550 nm. Scale bar: 50 mm.
at 100 nm, 378C for 2 hours. Ligation was performed at the
same temperature, for 1 hour using 455 nm LED irradiation,
then sodium ascorbate and caged QPD 5 were added and the
cells further incubated for 1 hour. Following incubation, the
cells were irradiated with 455 nm LED light for 30 minutes and
the formation of fluorescent precipitates was visualized by mi-
croscopy. As shown in Figure 6, cells incubated with the PNAs
designed to hybridize with miRNA-21 exhibited fluorescent
precipitates localized in the cytosol, whereas no fluorescent
precipitates were observed for the same experiment per-
formed with non-complementary PNA sequences.
The afore-mentioned results make use of a catalyst forma-
tion that is conditional on the presence of a template (turn-on
of catalysis). The reaction that yields the fluorescent readout is
not templated and proceeds best at high substrate concentra-
tions (>10 mm). We reasoned that our system could be extend-
ed to proceed at low substrate concentration, if an overhang
sequence was adjacent to the catalyst for substrate hybridiza-
tion. We had previously shown that a pentamer PNA was suffi-
cient for PNA–PNA templated reactions.[22] This was achieved
using a bifunctional bipyridine bearing a carboxylic acid group
at the 5-position and a Fmoc-protected glycine conjugated to
the 5’-amino group (Figure 7). The use of this bifunctional bi-
pyridine allowed bpy-PNA synthesis to be continued to obtain
a PNA-bpy-PNA adduct, using orthogonal Mtt/Fmoc chemis-
try.[36] To avoid crosstalk between the different stretches of this
bifunctional PNA-bpy-PNA conjugate, the PNA hybridizing to
DNA was prepared using chiral g-l-PNA[22,46], whereas the
stretch hybridizing the substrate was prepared using chiral g-
d-PNA. For the detection, we made use of the caged coumarin
because the reaction was performed in solution at low concen-
trations.
Figure 6. Representative images of templated ligation on fixed MCF-7 cells.
Conditions: 100 nm PNA6-Ru, 100 nm bpy-PNA5 or bpy-PNA3, 20 mm 5,
20 mm sodium ascorbate, irradiation for 30 min with LED 455 nm 1 W.
Buffer: DPBS (À). lex: 730 nm (2P excitation) 20%, lem: 500–550 nm. Scale
bar: 20 mm for A and E, 10 mm for B and F. A)–C) Reactions performed with
the perfect match probes (PNA6-Ru and bpy-PNA5). E–F) Reaction performed
with the scrambled probes (PNA6-Ru and bpy-PNA3). A,E) Overlay of differ-
ential interference contrast (DIC) with fluorescence images. B,F) Expanded
views of the red quadrant in A and E, respectively. C,G) Fluorescence channel
corresponding to B and F respectively, yellow lines denote the area that is
quantified in quadrants E and H, respectively.
The ligation was achieved as before, and the substrate was
then added at a stoichiometric concentration (100 nm). Gratify-
ingly, the templated uncaging of coumarin proceeded well
and reached completion after 60 min, whereas reaction per-
&
&
Chem. Eur. J. 2018, 24, 1 – 10
www.chemeurj.org
6
ꢀ 2018 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
ÝÝ These are not the final page numbers!

Full Paper
Figure 7. Templated release of a caged coumarin. 100 nm PNA probes and DNA, 10 mm sodium ascorbate, 408C, LED 455 nm 1 W. Buffer: PBS 1x pH 7.4,
0.05% Tween 20 and 0.1% formamide. lex: 360 nm, lem: 460 nm.
formed under the same condition but without the template or
with a non-complementary DNA template afforded negligible
conversion. When the reaction was performed under the same
conditions but with bpy-PNA instead of PNA-bpy-PNA (i.e., a
catalyst that cannot recruit the substrate), negligible reaction
was observed, clearly highlighting the importance of substrate
recruitment for reactions at low concentrations. We performed
the same reaction under conditions that would not require
template catalyst formation but would benefit from substrate
recruitment. To this end, we used PNA-Ru(phen)(bpy)2 instead
of PNA-Ru (i.e., the PNA is conjugated to an active catalyst)
and PNA-(Gly)4-PNA instead of the PNA-bpy-PNA. The PNA-
(Gly)4-PNA cannot undergo ligation but can still recruit the
substrate. Indeed, this reaction showed a complete conversion
of the substrate. Notably, the kinetics of this latter reaction are
very similar to those of the reaction performed with the liga-
tion, suggesting that the ligation indeed delivered a fully func-
tional catalyst.
tris-heteroleptic RuII complex from catalytically inactive precur-
sors. The ligation proceeded at physiological conditions,
making use of light-promoted ligand exchange. The products
of the ligations were used to uncage a QPD. Performing the re-
action with probes that capture the ligation product on beads,
we demonstrated a detection level of 36 amol of the DNA ana-
lyte per bead, suggesting that miniaturization of the assay to a
single bead should afford attomole detection of analytes (DNA
or RNA). This detection was multiplexed for multiple analytes
with beads of distinct sizes. This technology was also applied
to miRNA visualization in fixed cells. Finally, we demonstrated
that this ligation also proceeds with a bifunctional bipyridine
having one strand serving in the hybridization to the template
(ligation chemistry, turn-on of photocatalysis) and a second
strand for substrate recruitment (enabling the technology to
operate at low substrate concentration).
Experimental Section
Preparation of final ruthenium(II) or coumarin derivative: Com-
pound 2 (or S18 in the Supporting Information) was dissolved in
50 mL of N-methyl-2-pyrrolidone (NMP) to obtain a 0.1m solution
(5 equiv to 5 mg of resin loaded at 0.2 mmolgÀ1). TBTA (2 mg,
Conclusion
In summary, we have described a new oligonucleotide-tem-
plated ligation for the synthesis of a photocatalytically active
Chem. Eur. J. 2018, 24, 1 – 10
www.chemeurj.org
7
ꢀ 2018 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
&
&
These are not the final page numbers! ÞÞ

Full Paper
4 mmol, 4 equiv) was added as a solid. 15 mL of a 0.4m solution of
CuSO4 in water (6 equiv) was added, followed by 50 mL of 2m
aqueous solution of sodium ascorbate (100 equiv). The yellow solu-
tion was transferred onto the resin and reacted overnight. Final
compounds were purified by reversed-phase HPLC.
radiated with blue light for 30 minutes at room temperature. Cells
were washed three times with cold DPBS and then imaged using a
Zeiss LSM710 with settings lex: 730 nm (2P) 20%, lem: band-pass
filter 500–550 nm.
Cleavage from the resin: The N-terminus of the PNAs was capped
if needed. The resin was suspended in TFA (200 mL for 5 mg resin)
for 1 h. The PNA was then precipitated from diethylether and cen-
trifuged to recover the product as a pellet.
Acknowledgements
The authors thank Marion Pupier for the NMR spectra of the
ruthenium(II) complexes. UniGE bioimaging center provided
the instruments for imaging. We thank the Swiss National Sci-
ence Foundation for generous support (grant 200020 157106).
Synthetic templates: DNA sequences were purchased from Euro-
gentech as solids.
Representative procedure for ligation: Ligation was carried out
by mixing DNA and PNAs in PBS 1x buffer with 0.05% Tween 20
and 0.1% formamide (pH 7.4) to a final concentration of 5 mm and
a final volume of 100 mL in a 96-well plate (Nunc). The plate was
placed at 408C and irradiated with a l=455 nm LED (1 W, 2 cm
distance), unless otherwise stated. At different time intervals, lumi-
nescence was recorded using a SpectraMax M5 Microplate Reader
Conflict of interest
The authors declare no conflict of interest.
from Molecular Devices for a 96-well plate at lex: 450 nm, lem
:
630 nm.
Keywords: DNA · photocatalysis · ruthenium
Beads loading: Streptavidin–agarose resin, high capacity, was pur-
chased from Thermo Scientific. Binding capacity was calculated to
be 550 pmol of biotin 4-nitrophenylestermLÀ1 of settled resin,
based on the supplier specifications. 10 mL of settled resin was
treated with 20 mL of a 9:1 solution of 100 mm biotin (1800 pmol)
and 100 mm biotinylated PNA-bpy3 (200 pmol) by incubating for
1 hour. Beads were washed and resuspended in 200 mL of water.
[1] X. Li, D. R. Liu, Angew. Chem. Int. Ed. 2004, 43, 4848–4870; Angew.
Chem. 2004, 116, 4956–4979.
[2] K. Gorska, N. Winssinger, Angew. Chem. Int. Ed. 2013, 52, 6820–6843;
Angew. Chem. 2013, 125, 6956–6980.
[3] Y. Choi, G. Metcalf, M. H. Sleiman, D. Vair-Turnbull, S. Ladame, Bioorg.
Med. Chem. 2014, 22, 4395–4398.
[4] M. Di Pisa, O. Seitz, ChemMedChem 2017, 12, 872–882.
[5] Y. Xu, N. B. Karalkar, E. T. Kool, Nat. Biotechnol. 2001, 19, 148–152.
[6] S. Sando, E. T. Kool, J. Am. Chem. Soc. 2002, 124, 2096–2097.
[7] H. Abe, E. T. Kool, Proc. Natl. Acad. Sci. USA 2006, 103, 263–268.
[8] Y. Huang, J. M. Coull, J. Am. Chem. Soc. 2008, 130, 3238.
[9] X. H. Chen, A. Roloff, O. Seitz, Angew. Chem. Int. Ed. Engl. 2012, 51,
4479–4483; Angew. Chem. 2012, 124, 4556–4561.
[10] Z. L. Pianowski, N. Winssinger, Chem. Commun. 2007, 3820–3822.
[11] R. M. Franzini, E. T. Kool, ChemBioChem 2008, 9, 2981–2988.
[12] R. M. Franzini, E. T. Kool, J. Am. Chem. Soc. 2009, 131, 16021–16023.
[13] D. Al Sulaiman, J. Y. H. Chang, S. Ladame, Angew. Chem. Int. Ed. 2017,
56, 5247–5251; Angew. Chem. 2017, 129, 5331–5335.
[14] S. Ficht, C. Dose, O. Seitz, ChemBioChem 2005, 6, 2098–2103.
[15] O. Vꢂzquez, O. Seitz, J. Pept. Sci. 2014, 20, 78–86.
[16] O. Zavoiura, U. Resch-Genger, O. Seitz, Bioconjugate Chem. 2018, 29,
1690–1702.
[17] J. Sayers, R. J. Payne, N. Winssinger, Chem. Sci. 2018, 9, 896–903.
[18] K. Fujimoto, S. Matsuda, N. Takahashi, I. Saito, J. Am. Chem. Soc. 2000,
122, 5646–5647.
[19] H. Wu, B. T. Cisneros, C. M. Cole, N. K. Devaraj, J. Am. Chem. Soc. 2014,
136, 17942–17945.
[20] H. Wu, S. C. Alexander, S. Jin, N. K. Devaraj, J. Am. Chem. Soc. 2016, 138,
11429–11432.
[21] K. K. Sadhu, N. Winssinger, Chem. Eur. J. 2013, 19, 8182–8189.
[22] D. Chang, E. Lindberg, N. Winssinger, J. Am. Chem. Soc. 2017, 139,
1444–1447.
QPD precipitation on beads: 50 mL of resuspended beads (2.5 mL
of settled resin) was added to
a 1 mL of 200 nm PNA-Ru
(200 pmol) and 1 mL of DNA 1 at different concentrations in a
35 mm petri dish. The mixture was incubated and shaken for
2 hours in the dark at room temperature. The ligation reaction was
performed at 408C, shining blue light for 1 hour. After the ligation,
sodium ascorbate and 5 were added to a final concentration of
10 mm and 50 mm. Precipitation was elicited by irradiation of the
beads with a 488 nm laser, and fluorescence was recorded using a
Zeiss LSM710 confocal microscope. lex: 730 nm (2P) 20%, lem
:
band-pass filter 500–550 nm.
Fluorescence quantification on beads: Confocal images were ana-
lyzed and quantified using FIJI software. Ch1 (fluorescence chan-
nel) of each image was Z-stacked to its maximum projection. The
image was duplicated for thresholding. The beads were highlight-
ed and added to the regions of interest (ROI) manager, and the
threshold was set at 20–255. ROIs were combined and treated as a
single ROI. Particles of size (mm2) 1-Infinity were analyzed, report-
ing the analysis to the original image. The reported values of mean
grey and area% were multiplied to give the values used for the
statistics, based on the assumption that a higher concentration of
template should give a brighter and denser precipitate on the
beads.
miRNA21 templated ligation: 104 MCF-7 cells were seeded in a
35 mm dish and incubated at 378C for 24 hours. Cells were
washed three times with Dulbecco’s phosphate-buffered saline
(DPBS, without magnesium and calcium), then fixed with a 4% PFA
solution for 20 minutes at room temperature. Following fixation,
cells were washed three times with DPBS and PNA6-Ru and bpy-
PNA5 or PNA6-Ru and bpy-PNA3 (100 nm in DPBS) were added to
the dish. Cells were incubated at 378C for 2 hours, and then
washed once with cold DPBS. Ligation was performed through
455 nm irradiation of the dish at 378C for 1 hour. The cells were
washed again once with cold DPBS, sodium ascorbate (20 mm)
and 5 (20 mm) added. After 1 hour incubation at 378C, cells were ir-
[23] D. L. Chang, K. T. Kim, E. Lindberg, N. Winssinger, Bioconjugate Chem.
2018, 29, 158–163.
[24] J. L. Czlapinski, T. L. Sheppard, J. Am. Chem. Soc. 2001, 123, 8618–8619.
[25] D. K. Prusty, M. Kwak, J. Wildeman, A. Herrmann, Angew. Chem. Int. Ed.
2012, 51, 11894–11898; Angew. Chem. 2012, 124, 12064–12068.
[26] C. K. Prier, D. A. Rankic, D. W. MacMillan, Chem. Rev. 2013, 113, 5322–
5363.
[27] F. Heinemann, J. Karges, G. Gasser, Acc. Chem. Res. 2017, 50, 2727–
2736.
[28] L. Holtzer, I. Oleinich, M. Anzola, E. Lindberg, K. K. Sadhu, M. Gonzalez-
Gaitan, N. Winssinger, ACS Cent. Sci. 2016, 2, 394–400.
[29] M. Rçthlingshçfer, K. Gorska, N. Winssinger, Org. Lett. 2012, 14, 482–
485.
[30] L. Zayat, M. Salierno, R. Etchenique, Inorg. Chem. 2006, 45, 1728–1731.
&
&
Chem. Eur. J. 2018, 24, 1 – 10
www.chemeurj.org
8
ꢀ 2018 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
ÝÝ These are not the final page numbers!

Full Paper
[31] L. Salassa, C. Garino, G. Salassa, R. Gobetto, C. Nervi, J. Am. Chem. Soc.
2008, 130, 9590–9597.
[41] M. Prost, L. Canaple, J. Samarut, J. Hasserodt, ChemBioChem 2014, 15,
1413–1417.
[32] J. C. Griepenburg, T. L. Rapp, P. J. Carroll, J. Eberwine, I. J. Dmochowski,
Chem. Sci. 2015, 6, 2342–2346.
[42] V. B. Paragas, Y. Z. Zhang, R. P. Haugland, V. L. Singer, J. Histochem. Cyto-
chem. 1997, 45, 345–357.
[33] J. Suffert, R. Ziessel, Tetrahedron Lett. 1991, 32, 757–760.
[34] J. B. Borak, D. E. Falvey, J. Org. Chem. 2009, 74, 3894–3899.
[35] J. B. Edson, L. P. Spencer, J. M. Boncella, Org. Lett. 2011, 13, 6156–6159.
[36] D. Chouikhi, M. Ciobanu, C. Zambaldo, V. Duplan, S. Barluenga, N. Wins-
singer, Chem. Eur. J. 2012, 18, 12698–12704.
[43] J. Lu, A. Tsourkas, Nucleic Acids Res. 2009, 37, e100.
[44] J. Saarbach, E. Lindberg, S. Folliet, S. Georgeon, O. Hantschel, N. Wins-
singer, Chem. Sci. 2017, 8, 5119–5125.
[45] K. Ikeda, K. Horie-Inoue, T. Ueno, T. Suzuki, W. Sato, T. Shigekawa, A.
Osaki, T. Saeki, E. Berezikov, H. Mano, S. Inoue, Sci. Rep. 2015, 5, 13170.
[46] I. Sacui, W. C. Hsieh, A. Manna, B. Sahu, D. H. Ly, J. Am. Chem. Soc. 2015,
137, 8603–8610.
[37] K. K. Sadhu, E. Lindberg, N. Winssinger, Chem. Commun. 2015, 51,
16664–16666.
[38] Z. Huang, E. Terpetschnig, W. You, R. P. Haugland, Anal. Biochem. 1992,
207, 32–39.
[39] T. Machida, S. Dutt, N. Winssinger, Angew. Chem. Int. Ed. 2016, 55,
8595–8598; Angew. Chem. 2016, 128, 8737–8740.
Manuscript received: August 22, 2018
[40] Z. Diwu, Y. Lu, R. H. Upson, M. Zhou, D. H. Klaubert, R. P. Haugland, Tet-
rahedron 1997, 53, 7159–7164.
Version of record online: && &&, 0000
Chem. Eur. J. 2018, 24, 1 – 10
www.chemeurj.org
9
ꢀ 2018 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
&
&
These are not the final page numbers! ÞÞ

Full Paper
FULL PAPER
&
Photocatalyst Activation
M. Anzola, N. Winssinger*
&& – &&
Turn On of a Ruthenium(II)
Photocatalyst by DNA-Templated
Ligation
Photocatalysis on! A light-directed oli-
gonucleotide-templated ligation reac-
tion was used to generate an active RuII
photocatalyst. This complex was found
to have tremendous potential for signal
amplification with >15000 turnovers. A
templated reaction was used to turn on
the activity of this ruthenium(II) photo-
catalyst in response to a specific DNA
sequence.
&
&
Chem. Eur. J. 2018, 24, 1 – 10
www.chemeurj.org
10
ꢀ 2018 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
ÝÝ These are not the final page numbers!