A nucleic acid dependent chemical photocatalysis in live human cells

Article abstract of DOI:10.1002/chem.200902377

Only two nucleic acid directed chemical reactions that are compatible with live cells have been reported to date. Neither of these processes generate toxic species from nontoxic starting materials. Reactions of the latter type could be applied as gene-specific drugs, for example, in the treatment of cancer. We report here the first example of a chemical reaction that generates a cytotoxic drug from a nontoxic prodrug in the presence of a specific endogeneous ribonucleic acid in live mammalian cells. In this case, the pro-drug is triplet oxygen and the drug is singlet oxygen. The key component of this reaction is an inert molecule (InP-2′-OMe-RNA/Q-2′-OMe-RNA; P: photosensitizer; Q: quencher), which becomes an active photosensitizer (InP-2′-OMe-RNA) in the presence of single-stranded nucleic acid targets. Upon irradiation with red light, the photosensitizer produces over 6000 equivalents of toxic singlet oxygen per nucleic acid target. This reaction is highly sequence specific. To detect the generation of singlet oxygen in live cells, we prepared a membrane-permeable and water-soluble fluorescent scavenger, a derivative of 2,5-diphenylisobenzofurane. The scavenger decomposes upon reaction with singlet oxygen and this is manifested in a decrease in the fluorescence intensity. This effect can be conveniently monitored by flow cytometry.

Full text of DOI:10.1002/chem.200902377

DOI: 10.1002/chem.200902377
A Nucleic Acid Dependent Chemical Photocatalysis in Live Human Cells
Dumitru Arian,[a] Emiliano Clꢀ,[b] Kurt V. Gothelf,[b] and Andriy Mokhir*[a]
Abstract: Only two nucleic acid direct-
ed chemical reactions that are compati-
ble with live cells have been reported
to date. Neither of these processes gen-
erate toxic species from nontoxic start-
ing materials. Reactions of the latter
type could be applied as gene-specific
drugs, for example, in the treatment of
cancer. We report here the first exam-
ple of a chemical reaction that gener-
ates a cytotoxic drug from a nontoxic
prodrug in the presence of a specific
endogeneous ribonucleic acid in live
mammalian cells. In this case, the pro-
drug is triplet oxygen and the drug is
singlet oxygen. The key component of
this reaction is an inert molecule (InP–
photosensitizer produces over 6000
equivalents of toxic singlet oxygen per
nucleic acid target. This reaction is
highly sequence specific. To detect the
generation of singlet oxygen in live
cells, we prepared a membrane-perme-
able and water-soluble fluorescent
scavenger, a derivative of 2,5-diphenyl-
2’-OMe-RNA/Q–2’-OMe-RNA;
P:
photosensitizer; Q: quencher), which
becomes an active photosensitizer
(InP–2’-OMe-RNA) in the presence of
single-stranded nucleic acid targets.
Upon irradiation with red light, the
AHCTUNGTRENNiGUN sobenzofurane. The scavenger decom-
poses upon reaction with singlet
oxygen and this is manifested in a de-
crease in the fluorescence intensity.
This effect can be conveniently moni-
tored by flow cytometry.
Keywords: live cells · nucleic acids ·
photocatalysis · photosensitizers ·
singlet oxygen
Introduction
kidney 293 (HEK293) and HL-60 cells.[5] Although these
chemical processes are highly sequence specific, they exhibit
relatively low maximal turnover numbers (TONmax) of 7–92.
This limits their applicability to the detection of abundant
RNAs.
Several types of nucleic acid controlled chemical reactions
have been reported, including templated ligation, cleavage,
group transfer, and a variety of photoinduced processes.[1]
A
majority of them were designed for in vitro applications,[1–3]
and only two such reactions were found to be compatible
with living cells. In particular, Kool and co-workers have
shown that ligation between 3’-nucleophilic phosphoro-
thioate and 5’-electrophilic probes is triggered by endogene-
ous RNAs, both in bacteria and in mammalian cells.[4] Three
research groups have independently reported that the Stau-
dinger reaction is accelerated by RNAs in human embryonic
Nucleic acid dependent reactions that are compatible with
live cells are especially interesting, because they can be ap-
plied not only for monitoring endogeneous RNAs in situ but
also for the RNA-dependent release of cytotoxic drugs
(gene-encoded prodrugs).[1] Such drugs can, for example, se-
lectively kill cancer cells, which are usually genetically dif-
ferent from normal cells. The latter possibility has not been
demonstrated yet. In general, the adaptation of chemical re-
actions to live-cell conditions is a challenging task,[5,6] be-
cause of the poor membrane permeability and toxicity of
the reagents and the presence of a rather complex mixture
of different compounds in the cell, which may facilitate un-
desired side reactions. Several nucleic acid directed reac-
tions that convert prodrugs into drugs have been reported to
work in vitro.[7] For example, Saito and co-workers and
Tanabe, Nishimoto, and co-workers have described the stoi-
chiometric UV-light-induced generation of model drugs. The
addition of the specific nucleic acid increases the yields of
the products in these reactions by 2- to 7-fold.[7a,b] Gothelf
and co-workers have reported a reaction that can generate
[a] D. Arian, Dr. A. Mokhir
Anorganisch-Chemisches Institut
Ruprecht-Karls-Universitꢀt Heidelberg
Im Neuenheimer Feld 270, 69120 Heidelberg (Germany)
Fax : (+49)6221-548439
E-mail: Andriy.Mokhir@urz.uni-heidelberg.de
[b] Dr. E. Clꢁ, Prof. Dr. K. V. Gothelf
Centre for DNA Nanotechnology
Department of Chemistry and iNANO
Aarhus University
Langelandsgade 140, 8000 ꢂrhus C (Denmark)
Supporting information for this article is available on the WWW
under http://dx.doi.org/10.1002/chem.200902377.
1
3
toxic O2 (the drug) from nontoxic O2 (the prodrug).[7c] In
288
ꢃ 2010 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Chem. Eur. J. 2010, 16, 288 – 295

FULL PAPER
particular, they have prepared a duplex from the (pyropheo-
phorbide-a)–oligodeoxyribonucleotide conjugate (P–ODN1;
P: pyropheophorbide-a photosensitizer; Figure 1) and the
(Scheme 1). The reaction of this compound with InCl3 in
ꢀ
benzene furnished InP Oallyl in 62% yield. The allyl group
was cleanly removed in the presence of [Pd
A
ꢀ
obtain InP OH. The complex was conjugated to different
ODN and 2’-OMe-RNA sequences by using solid-phase syn-
thesis.
Figure 1. Nucleic acid induced catalytic decomposition of a scavenger
(dye). The dye reacts with singlet oxygen to produce a colorless product.
“P” on a gray background indicates a deactivated photosensitizer, “P” on
a white background (inside the star) indicates the activated photosensitiz-
er, and “Q” in black-colored circles indicates a quencher; all sequences
are arbitrary.
complementary Q–ODN2 (Q: quencher). In this associate,
the P is fully quenched by the Q, whereas in the presence of
an analyte nucleic acid, the P–ODN1 conjugate is set free
and the P is reactivated. In the active form, the P induces
the generation of 1O2 when exposed to light. This effect
could be observed only at rather high concentrations of the
duplex (5 mm) and in the D2O buffer,[7c] conditions that are
not compatible with live cells. This reaction is not catalytic,
1
because the P decomposes quickly in the presence of O2.
Herein, we report on two key modifications of the known
1
system: we use a more sensitive method for O2 monitoring
and substitute the P for a more potent and stable photosen-
sitizer. The resulting nucleic acid controlled photocatalyst is
active in aqueous buffer, exhibits the highest turnover fre-
quency and number among the known analogous chemical
1
catalysts, and is able to generate toxic O2 in the presence of
specific nucleic acids in live cells.
ꢀ
ꢀ
Scheme 1. A) Preparation of photosensitizers P OH and InP OH:
a) Allyl-OH, H2SO4; b) 2,4,6-collidine, 1708C; c) [Pd(PPh3)4], (Et2NH2)-
(HCO3); d) InCl3; e) [Pd(PPh3)4], (Et2NH2)(HCO3). B: Sequences of the
AHCTUNGTRENNUNG
G
N
ACHTUNGTRENNUNG
Results and Discussion
conjugates and nucleic acids used in this study. BHQ-3: black-hole
quencher-3.
In3+ complexes with P and its derivatives are substantially
more active photodynamic agents than the parent ligands.[8]
One of the reasons for this could be that they are better
photosensitizers under physiological conditions. Therefore,
we decided to substitute the pyropheophorbide-a in P–ODN
Next, we developed a fluorescent scavenger (denoted as
1
“dye” in Figure 1) for monitoring the formation of O2 in
1
live human cells. O2 was previously detected by monitoring
for the In3+ complex, InP. It has been reported that InP
its weak phosphorescence at 1270 nm.[7c] Fluorescent probes
offer higher 1O2 sensitivity and easier signal detection.
Known reagents of this type include trans-1-(2-methoxyvi-
nyl)pyrene (MVP), dansyl-2,2,5,5-tetramethyl-2,5-dihydro-
1H-pyrrole (DanePy), fluorescein-based compounds such as
9-[2-(3-carboxy-9,10-dimethyl)anthryl]-6-hydroxy-3H-xanth-
en-3-one (DMAX) or 9-[2-(3-carboxy-9,10-diphenyl)anthr-
yl]-6-hydroxy-3H-xanthen-3-one (DPAX), 1,3-diphenyliso-
benzofuran (DPBF), and single-oxygen sensor green
ꢀ
OH, which is required for the synthesis of the ODN conju-
ꢀ
gates, can be obtained by treatment of P OH with InCl3 in
benzene.[8] In our hands, this reaction yielded only traces of
the complex. We believe the reason for this is that both re-
ꢀ
agents, InCl3 and P OH, are almost insoluble in benzene.
For an alternative metal-complex formation procedure, we
ꢀ
used a soluble allyl ester of pyropheophorbide-a (P Oallyl)
ꢀ
ꢀ
in place of the P OH. P Oallyl was obtained by acid-cata-
lyzed transesterification of phytyl pheophorbide-a, followed
by decarboxylation of the resulting allyl pheophorbide-a
1
(SOSG). They have been used for monitoring O2 in vitro.
DanePy and SOSG have been applied for the detection of
Chem. Eur. J. 2010, 16, 288 – 295
ꢃ 2010 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.chemeurj.org
289

A. Mokhir et al.
1O2 in plant cells and plants.[9] No probe has been demon-
1
strated to be applicable for monitoring O2 in mammalian
cells. The lifetime of 1O2 in aqueous buffer is very short
(ꢁ3.5 ms)[10] and it is further reduced by intracellular
Figure 2. Monitoring uptake of fluorogenic 1O2 scavengers (dyes I–III) in
live HL-60 cells by using flow cytometry. The intensity of emission at
530 nm (excitation at 488 nm) is plotted versus the number of events
(cells). The background trace is a negative control (no dye was added).
Cells (106 cellsmLꢀ1) were incubated in RPMI medium containing 10%
fetal calf serum (FCS). Dyes were added at the final concentration of
5 mm. The measurements were conducted 10 min after addition of the
dyes to the cell suspension. Fluorescence intensity is given in arbitrary
units.
1
Figure 3. Monitoring of the generation of O2 in live HL-60 cells by using
dye III and flow cytometry. Cells (106 cellsmLꢀ1) were first loaded with
Scheme 2. Water-soluble derivatives (dyes I–III) of 1,3-diphenylisobenzo-
furan prepared for the detection of singlet oxygen (1O2) in live mammali-
an cells.
ꢀ
the corresponding amount of InP OH (concentrations shown on the
plot), washed twice with phosphate-buffered saline (PBS), and incubated
with dye III (5 mm) in RPMI with 10% FCS for 10 min. The cells were
then irradiated with red light (635 nm). The mean fluorescence of the
cells that were not exposed to the light was taken as 100%.
perimental Section), which turned out to be soluble in water
at pH 7 up to concentrations of at least 3 mm. Doubly posi-
tively charged dye II is not membrane permeable, whereas
monocharged I and negatively charged III are permeable
(Figure 2). The latter two compounds were not toxic to HL-
60 cells at concentrations of up to 5 mm. By using flow cy-
tometry, we tested whether dyes I and III are suitable for
concentration of the photosensitizer (Figure 3; Supporting
Information, Figure S27). Unfortunately, at high concentra-
tions, dye I causes denaturation of ODN duplexes, for exam-
1
ple, the deactivated catalyst (Figure 1), and thus triggers O2
generation in the absence of a nucleic acid target. We sup-
pose that this effect is due to unspecific interactions of the
positively charged dye with the negatively charged nucleic
acid duplexes. This is confirmed by the fact that negatively
charged dye III does not affect the duplex stability. The
1
monitoring the formation of O2 in live HL-60 cells. First,
the cells were loaded with photosensitizer InP–OH (Sup-
1
porting Information, Figure S26). O2 was generated by ex-
posure of the cells to red light (635 nm). Both dyes were
shown to be stable in the cells in the absence of the photo-
sensitizer. In its presence, they decomposed efficiently and,
thereby indicated the formation of 1O2. As expected, the
rate of the decomposition reaction is proportional to the
1
latter compound was used in all further experiments for O2
detection because of this favorable property.
Under physiological conditions and at low concentration
(300 nm), the known P–ODN1 does not catalyze the photo-
290
www.chemeurj.org
ꢃ 2010 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Chem. Eur. J. 2010, 16, 288 – 295

Controlled Photocatalysis in Live Cells
FULL PAPER
decomposition of dye III in vitro, whereas its In3+ deriva-
tive, InP–ODN1, is an efficient catalyst of this reaction
(Figure 4). The activity of the latter compound can be con-
Figure 5. Fluorescence spectra (excitation at 635 nm) of InP–ODN2–Q
(1 mm) under different conditions. Concentrations of DNA15 and
DNA16: 5 mm. Buffer: 10 mm phosphate buffer, pH 7, with 150 mm NaCl.
Fluorescence intensity is given in arbitrary units: a.u.
accelerates the decomposition of dye III by 6.5-fold with re-
Figure 4. Decomposition of dye III (0.23 mm) upon illumination with red
light (l=635 nm, 2.59 mJsꢀ1) under aerobic conditions in phosphate
buffer (10 mm, pH 7, with 150 mm NaCl) in the presence of P–ODN1
spect to the background reaction (Figure 6).
&
^
(300 nm, ) or InP–ODN1 (300 nm, ) or in the absence of any photosen-
~
sitizer (background, ). Solid lines represent regions of constant rate of
dye III decomposition. The experimental error of the determined concen-
trations of dye III was found to be less than 5%.
trolled by nucleic acids. In particular, InP–ODN1 forms a
stable duplex with Q–ODN7, which is practically not fluo-
rescent (Supporting Information, Figure S28). Complemen-
tary DNA12 (1 equiv) releases InP–ODN1 from the duplex;
this release is manifested by an approximately 30-fold in-
crease in the fluorescence intensity. The InP–ODN1
(300 nm) activated in this way speeds up the photodecompo-
sition of 700-fold excess dye III by a factor of 6.1 relative to
the background reaction (Figure 4). A turnover number
(TON) of 377 is achieved at a constant turnover frequency
(TOF) of (10ꢂ1) minꢀ1 within 40 min. After this time, the
reaction slows down. A maximal TON of 512 is reached
within 100 min. The reaction can be also controlled by RNA
targets. In particular, RNA13, which has the same sequence
as DNA12, produces a TON of 376 under similar conditions.
Next, we tested whether this assay can detect single mis-
matches in nucleic acids. The duplex InP–ODN1/Q–ODN7
was found to be insufficiently sequence specific: in the best
case, the photodecomposition of dye III (0.23 mm) in the
presence of mismatched DNA14 (T8!A8 mutation) is only
1.4 times slower than that in the presence of matched
DNA12. To improve this property, we substituted duplex
InP–ODN1/Q–ODN7 for a hairpin-structured conjugate
InP–ODN2–Q (Scheme 1B). Such probes are known to be
highly sequence specific.[12] As expected, InP–ODN2–Q is
only weakly fluorescent in the folded state. Complementary
DNA15 opens up the hairpin and restores its fluorescence,
whereas mismatched DNA16 (G15!C15 mutation) has no
effect (Figure 5). Correspondingly, the photodecomposition
of dye III (0.23 mm) by InP–ODN2–Q (300 nm) in the pres-
ence of DNA16 is not distinguishable from that in the ab-
sence of any DNA. In contrast, the complementary DNA
Figure 6. Decomposition of dye III (0.23 mm) upon illumination with red
light (l=635 nm, 2.59 mJsꢀ1) under aerobic conditions in phosphate
buffer (10 mm, pH 7, with 150 mm NaCl) in the presence of InP–ODN2–
~
^
Q (300 nm) alone ( ) or with mismatched DNA16 (3 mm, ) or matched
*
DNA15 (3 mm, ). The experimental error of the determined concentra-
tions of dye III was found to be less than 5%.
Furthermore, we explored the TON that this reaction is
able to generate. Concentrations of the deactivated catalyst
and DNA12 were kept constant at 300 nm while the concen-
tration of dye III was increased. At the highest concentra-
tion of substrate tested (3 mm, 104 excess), a TON of 6373
could be achieved in 44.6 h. In 24 h, this reaction produces a
TON of 4533, which represents an average TOF of
3.1 minꢀ1. The best DNA-dependent chemical reaction re-
ported to date[13] generates a TON of 402 during the same
time period, with a TOF of 0.28 minꢀ1 (Table 1). Important-
ly, under these comparable conditions, our reaction has a
signal-to-noise ratio (S/N) that is more than three times
better: S/N=7.2 rather than S/N=2.[13] The dependence of
the TON on the concentration of dye III is linear (Figure 7).
No saturation is observed, even at the highest concentration
of dye used. This indicates that the catalyst, InP, is not
bleached even after the generation of over 6000 equivalents
of 1O2. Therefore, we expect that this photoreaction may
produce even larger TONs at higher concentrations of dye
III.
Chem. Eur. J. 2010, 16, 288 – 295
ꢃ 2010 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.chemeurj.org
291

A. Mokhir et al.
Table 1. Comparison of the reaction reported in this paper with the best
nucleic acid dependent chemical reactions reported to date.
streptolysin O (SLO). The cells were then loaded with dye
III and exposed to the red light. O2 formation correlated
1
A
E
TOF
TON,
time[c]
Reference
with a decrease in the fluorescence intensity of dye III. This
signal was quantified by the flow cytometry. We observed
that the scrambled probe (InP–2’-OMe-RNA5/Q-2’-OMe-
[b]
ratio[a]
[minꢀ1
]
104
103
104
102
104
104
50
2
7
0.28
0.15
0.16
13
0.03
0.06
402, 24 h [13]
215
226
50
[13]
[15]
[16]
1
RNA10) does not generate O2 upon exposure of the cell to
red light for up to 25 min, which indicates that the duplex is
stable in the cell within this time period (Figure 8). Under
the same conditions, the positive control InP–2’-OMe-RNA
40
[17]
92
[18]
1
7
[5a]
produces large amounts of O2. We were pleased to observe
5
0.02
3.4
5
[2b]
this work
that the 28S rRNA probe (InP–2’-OMe-RNA6/Q-2’-OMe-
104
7
4533,
24 h
6373,
44.6 h
RNA11) is activated in the cell, which is reflected in the
1
2.4
this work
time-dependent generation of O2 (Figure 8). The scrambled
1
&
probe ( in Figure 8, bottom) starts producing O2 25 min
[a] Ratio of initial rates of chemical reactions in the presence and ab-
sence of a template nucleic acid. [b] TOF: turnover frequency. [c] TON:
turnover number.
after the beginning of the reaction. When we introduced a
one-hour incubation period before starting the experiment,
Figure 7. Dependence of the maximal turnover numbers (TON) in the
photochemical decomposition of dye III in the presence of [InP–ODN1/
Q–ODN7] (300 nm) and [DNA12] (300 nm) on the concentration of dye
III. Samples were irradiated with red light (l=635 nm, 2.59 mJsꢀ1).
Buffer: 10 mm phosphate, pH 7, with 150 mm NaCl. The solid line corre-
sponds to a linear fit of the experimental data shown as &.
Finally, we studied whether our duplexes could be activat-
ed in live cells by endogeneous RNAs. First, we prepared
two pairs of deactivated catalysts containing 2’-OMe RNAs
in place of ODNs. 2’-OMe RNAs are known to be signifi-
cantly more resistant to intracellular nucleases than natural
ODNs.[15] In each pair, one duplex (the scrambled probe)
was designed to be inert, whereas the second one was sup-
posed to bind 28S rRNA in the cell and produce the active
photosensitizer (the 28S RNA probe). In the first pair, the
duplexes consisted of a 16-mer InP–2’-OMe-RNA and a 22-
mer Q-2’-OMe-RNA, whereas the strands were longer in
the second pair: correspondingly, a 20-mer and a 30-mer.
The shorter duplexes were shown to be unstable in the cell
(Supporting Information, Figure S29). Therefore, all further
experiments were conducted with the longer ones. The deac-
tivated catalysts (InP–2’-OMe-RNA/Q-2’-OMe-RNA) were
brought into HL-60 cells by using a pore-forming peptide,
Figure 8. Monitoring of the formation of 1O2 in live HL-60 cells by using
flow cytometry. Concentration of dye III: 5 mm; concentration of each
conjugate: 1 mm. Top) Intensity of emission at 530 nm (excitation at
488 nm) is plotted versus the number of events (counts=cell numbers):
trace 1: negative control (no conjugates were added); trace 2: positive
control (InP–2’-OMe-RNA5); trace 3: 28S rRNA probe (InP–2’-OMe-
RNA6/Q–2’-OMe-RNA11); trace 4: scrambled probe (InP–2’-OMe-
RNA5/Q–2’-OMe-RNA10). Cells were irradiated with red light (635 nm)
for 25 min before the measurements. Bottom) The mean fluorescence of
cells that were not incubated with any conjugates was taken as 100%
(horizontal line). The trace numbering is the same as in (A).
292
www.chemeurj.org
ꢃ 2010 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Chem. Eur. J. 2010, 16, 288 – 295

Controlled Photocatalysis in Live Cells
FULL PAPER
mixture was slowly (2 h) heated up to 228C and then stirred overnight at
this temperature. The resulting suspension was cooled down to 08C, and
an aqueous solution of hydrochloric acid (15%, 15 mL) was added drop-
wise. The reaction mixture turned deep red. The cooling bath was re-
moved, and the mixture was left to stir under argon for 1 h. The aqueous
solution was washed with diethyl ether several times. It was then neutral-
ized with Et3N, and the product was extracted with diethyl ether. The
combined diethyl ether extracts were dried over Na2SO4, the diethyl
ether was evaporated, and the residue was purified by column chroma-
tography by using a mixture of hexane and Et2O (2:1). The yield of prod-
uct 4 (yellow crystalline compound) was 2.5 g (29%); TLC (EtOAc/
hexane 1:4): Rf =0.34; 1H NMR (600 MHz, [D6]acetone): d=7.83 (m,
4H), 7.79 (m, 2H), 6.92 (m, 2H), 6.88 (m, 4H), 3.01 ppm (s, 12H); ESI-
MS, positive mode: m/z calcd for C24H24N2O [M]2+: 178.09; found:
178.09.
the activities of the scrambled and 28S rRNA probes
became indistinguishable from each other (Supporting Infor-
mation, Figure S30). These facts indicate that the 2-OMe-
RNA duplexes are not fully stable in the cell within the
time period of the experiment, which is in agreement with
reported data.[14] This limits the amplification potential of
our assay. We hypothesize that the background process of
duplex decomposition could be either slowed down or fully
eliminated by the substitution of 2’-OMe-RNAs for more
stable nucleic acid mimics, for example, peptide nucleic
acids (PNAs) and hybrids of locked nucleic acids (LNAs)
and 2’- OMe-RNAs. We are currently exploring these possi-
bilities in our laboratories.
1-(4-Dimethylaminophenyl)-3-(4-trimethylaminoniumphenyl)isobenzofur-
an iodide (dye I): A mixture of compound 4 (178 mg, 0.5 mmol), K2CO3
(500 mg), and excess of MeI (1 mL) in acetone (10 mL) was stirred at
228C for 24 h under argon. The insoluble material was filtered off and
washed with acetone several times. The filtrate was collected, and the
acetone was evaporated to yield dye I (200 mg, 81%) as red crystals:
1H NMR (200 MHz, [D6]dimethylsulfoxide (([D6]DMSO)): d=8.15–8.01
(m, 4H), 7.99–7.87 (m, 4H), 7.20–7.01 (m, 2H), 6.88 (d, 2H), 3.66 (s,
9H), 3.01 ppm (s, 6H); ESI-MS, positive mode: m/z calcd for C25H27N2O
(the cationic part of dye I): 371.13; found: 371.20; UV/Vis (in PBS con-
taining 10 mm phosphate and 150 mm NaCl, pH 7): l=302, 437 nm; fluo-
Conclusion
We have developed a nucleic acid dependent photosensitiz-
er, which is functional under physiological conditions, is able
to generate over 6000 equivalents of toxic singlet oxygen
per equivalent of catalyst, and is highly sequence specific.
We have demonstrated that the activity of this compound in
live human cells is efficiently controlled by specific endoge-
neous ribonucleic acids. This is the first example of a live-
cell-compatible gene-specific photosensitizer. Such agents
may be used to obtain highly selective photodynamic drugs.
rescence spectrum (in PBS buffer as above): lexcitation =437 nm; lemission
=
595 nm.
1,3-Di(4-trimethylaminoniumphenyl)isobenzofuran diiodide (dye II): A
mixture of compound 4 (178 mg, 0.5 mmol), K2CO3 (500 mg), and an
excess of MeI (1 mL) in N,N-dimethylformamide (DMF; 10 mL) was
stirred at 228C for 24 h under argon. The solvent was removed under
0.01 mbar pressure at 508C, the residue was washed with methanol sever-
al times, and the filtrate was collected. The solvent was evaporated, and
the residue was recrystallized from methanol to yield dye II (208 mg,
65%) as orange crystals: 1H NMR (200 MHz, [D6]DMSO): d=8.31–8.26
(m, 4H), 8.14–8.09 (m, 6H), 7.29–7.24 (m, 2H), 3.70 ppm (s, 18H); ESI-
MS, positive mode: m/z calcd for C26H30N2O (the dicationic part of dye
II): 193.11; found: 193.05; UV/Vis (in PBS buffer as above): l=274,
Experimental Section
Commercially available chemicals of the best quality from Aldrich/
Sigma/Fluka (Germany) were obtained and used without purification.
The reagents for DNA synthesis were obtained from Proligo Biochemie
GmbH (Germany), Glen Research (USA), or Link Technologies (UK).
HPLC-purified DNAs and RNAs were purchased from IBA GmbH
(Germany). MALDI-TOF mass spectra were recorded on a Bruker
BIFLEX III spectrometer. The matrix mixture (2:1 v/v) was prepared
from 6-aza-2-thiothymine (ATT, saturated solution in acetonitrile) and
diammonium citrate (0.1m in water). Samples for mass spectrometry
were prepared by the dried-droplet method by using a 1:2 probe/matrix
ratio. Mass accuracy with external calibration was 0.1% of the peak
mass, that is, ꢂ6.0 at m/z 6000. Preparative and analytical HPLC was per-
formed at 608C on a Shimadzu liquid chromatograph equipped with a
UV detector and a Macherey–Nagel Nucleosil C4 250ꢄ4.6 mm column.
UV/Vis spectra were measured on a Varian Cary 100 Bio UV/Vis spec-
trophotometer by using 1 cm optical path black-wall absorption semi-mi-
crocuvettes (Hellma GmbH, Germany) with a sample volume of 0.7 mL.
These cuvettes were also used for photochemical experiments. Light
from a red laser (635 nm, 2.59 mW) was applied from the top of the cuv-
ette by using an optical fiber. Fluorescence spectra were acquired on a
Varian Cary Eclipse fluorescence spectrophotometer by using black-wall
413 nm; fluorescence spectrum (in PBS buffer as above): lexcitation
413 nm; lemission =453, 478 nm.
=
1,3-Di(4-(4,4-dimethyloxazolin-2-yl)phenyl)isobenzofuran (8): Grignard
reagent 7 was prepared from 2-(4-bromophenyl)-4,4-dimethyl-2-oxazo-
line[20] (6; 5.1 g, 20.0 mmol) and Mg turnings (0.51 g, 21 mmol) in THF
(20 mL).[21] After all of the Mg turnings had dissolved, the solution was
added dropwise to a solution of methyl 2-formylbenzoate pseudoester (3)
in THF (25 mL) at 08C. The cooling bath was removed, and the mixture
was left to stir overnight. The aqueous HCl solution (10%, 15 mL) was
then added at 08C, and stirring was continued for an additional hour.
The water layer was washed several times with diethyl ether. (The prod-
uct of the reaction stays in the aqueous phase.) The aqueous phase was
carefully neutralized with sodium hydroxide solution, and the product
was extracted with diethyl ether. The combined diethyl ether extracts
were dried over Na2SO4, and the solvent was evaporated. Recrystalliza-
tion of the residue from diethyl ether yields the analytically pure product
(3.95 g, 85%): 1H NMR (200 MHz, CDCl3): d=8.03 (m, 8H), 7.88 (m,
2H), 7.11 (m, 2H), 4.14 (s, 4H), 1.42 ppm (s, 12H); ESI-MS, positive
mode: m/z calcd for C30H29N2O3 [M+H+]+: 465.21; found: 465.21.
fluorescence semi-microcuvettes (Hellma GmbH, Germany) with
a
sample volume of 0.7 mL. The fluorescence of dyes I–III in live HK-60
cells was quantified by using an Accuri C6 Flow cytometer. The data
were processed by using the CFLow Plus (Accuri) and FCS Express V3
(De Novo Software) software packages.
1,3-Di(4-carboxyphenyl)isobenzofuran (dye III): A solution of compound
8 (1.0 g, 2.2 mmol) in aqueous HCl (3m, 50 mL) was heated to reflux for
1 h. The precipitate formed was filtered, washed with water, and redis-
solved in 20% sodium hydroxide in methanol (prepared from a 50% so-
lution of sodium hydroxide in water). The resulting solution was heated
to reflux for another hour. The sodium hydroxide solution was then acidi-
fied with aqueous HCl (9m) until pH 1 was reached; this resulted in pre-
cipitation of dye III. The precipitate was filtered, washed with water, and
then dried at 0.01 mbar pressure. The yield of dye III was 0.51 g (65%):
TLC (THF/CH2Cl2 1:9, with a drop of acetic acid): Rf =0.57; 1H NMR
1,3-Di(4-dimethylaminophenyl)isobenzofuran (4; Scheme 2): A Grignard
reagent (2) was first prepared from 4-bromo-N,N-dimethylaniline (1,
12 g, 60.0 mmol) and Mg turnings (1.46 g, 60.0 mmol) in tetrahydrofuran
(THF; 40 mL) by using a standard protocol. The resulting solution was
added dropwise to solution of methyl 2-formylbenzoate pseudoester[19]
(3; 4.0 g, 24.4 mmol) in THF (30 mL) at 08C under argon. The reaction
Chem. Eur. J. 2010, 16, 288 – 295
ꢃ 2010 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.chemeurj.org
293

A. Mokhir et al.
(200 MHz, [D6]DMSO): d=13.06 (brs, 2H), 8.13 (m, 4H), 8.06 (m, 6H),
7.20 ppm (m, 2H); ESI-MS, negative mode: m/z calcd for C22H13O5 [dye
IIIꢀH+]ꢀ: 357.08; found: 357.07; UV/Vis (in PBS buffer as above): l=
5.95–5.85 (m, 1H), 5.33–4.96 (m, 2H), 5.32–5.08 (m, 2H), 4.62 (m, 2H),
4.56 (m, 1H), 4.32 (m, 1H), 3.82–3.75 (m, 2H), 3.65 (s, 3H), 3.35 (s, 3H),
3.30 (s, 3H), 2.86–2.68 (m, 1H), 2.61 (m, 1H), 2.50 (m, 1H), 2.24–2.10
(m, 1H), 2.02 (m, 3H), 1.74–1.68 ppm (m, 3H); 13C NMR (CDCl3): d=
195.24, 172.59, 168.94, 160.68, 155.89, 153.72, 150.51, 146.43, 146,22,
145.31, 144.44, 140.40, 136.62, 135.48, 134.80, 131.95, 129.20, 122.67,
118.62, 107.77, 105.70, 100.32, 92.38, 77.20, 65.38, 50.83, 49.59, 48.11,
30.22, 29.69, 23.21, 19.58, 17.29, 12.73, 12.21, 11.03 ppm; high-resolution
ESI-MS, positive mode: m/z calcd for C36H36N4O3In [MꢀClꢀ]+: 687.1821;
found: 687.1813.
282, 430 nm; fluorescence spectrum (in PBS buffer as above): lexcitation
430 nm; lemission =476, 502 nm.
=
Pheophytin a: Spirulina powder (500 g) was mixed with acetone (2 L)
and liquid nitrogen (2 L) was slowly added while the mixture was stirred
with a mechanical stirrer. The mixture was then allowed to melt, and the
resulting slurry was heated to reflux for two hours and filtered through a
Buchner funnel. The precipitate obtained was washed twice with acetone
and discarded. Acetic acid (6 mL) was added to the filtrate and all vola-
tile compounds were removed at 0.01 mbar pressure. The crude product
was purified by column chromatography to give pheophytin a (9; yield:
5 g). Conditions for column chromatography: carotenoids (red, yellow
bands) were eluted with 5% petroleum ether in CH2Cl2; a black band
corresponding to the product was eluted with 20% ethyl acetate in
ꢀ
In(pyropheophorbide a)Cl (In
and PPh3 (1.1 mg, 4.13 mmol) in CH2Cl2 (1 mL) were added to suspension
containing In(pyropheophorbide a allyl ester)Cl (20 mg, 27.67 mmol) and
G
G
(NH2Et2)
was completed in 15 min, according to TLC analysis. The resultant solu-
tion was directly loaded onto a chromatography column. In(P-OH)
ACHTUGNERTN(NUNG HCO3) (20.9 mg, 160 mmol) in CH2Cl2 (5 mL). The reaction
ACHTUNGTRENNUNG
1
CH2Cl2: TLC (CH2Cl2/EtOAc 15:1): Rf =0.64; the H and 13C NMR spec-
(17 mg, 90.2%) was eluted with 5% methanol in CH2Cl2 (with a few
drops of acetic acid): TLC (CH2Cl2/MeOH 50:1): Rf =0.25; 1H NMR:
(400 MHz, CDCl3): d=9.68 (s, 1H), 9.45 (s, 1H), 8.26 (s, 1H), 7.97–7.90
(m, 1H), 6.23–6.07 (m, 2H), 4.76 (m, 1H), 4.60 (m, 1H), 3.94–3.72 (m,
tra of the product were in agreement with literature data;[22] ESI-MS,
positive mode: m/z calcd for C55H75N4O5 [M+H+]+: 871.57; found
871.48.
3H), 3.64 (s, 3H), 3.53 (s, 1H) 3.29 (s, 3H), 3.28 (s, 3H), 1.74–1.71ACHTUNGTRENNUNG(m,
Pheophorbide a allyl ester: Pheophytin a (3 g, 3.44 mmol) and sulphuric
acid (2 mL) were stirred in allyl alcohol (100 mL) at 228C for 48 h. The
solution was diluted with water (200 mL), neutralized with aqueous satu-
rated NaHCO3, and extracted twice with dichloromethane. The combined
organic extracts were dried over Na2SO4, the solvent was removed, and
the residue was dried at 0.01 mbar pressure. The crude product was puri-
fied by column chromatography to yield pheophorbide a allyl ester (10;
1.4 g, 64%, black solid): TLC (CH2Cl2/EtOAc 15:1): Rf =0.44; 1H NMR:
(400 MHz, CDCl3): d=9.47 (s, 1H), 9.32 (s, 1H), 8.56 (s, 1H), 7.97–7.91
(m, 1H), 6.28–6.14 (m, 2H), 6.27 (s, 1H), 5.82–5.71 (m, 1H), 5.20–5.09
(m, 2H), 4.50 (m, 2H), 4.45 (m, 1H), 4.24 (m, 1H), 3.90 (s, 3H), 3.68 (s,
3H), 3.64 (m, 2H), 3.39 (s, 3H), 3.18 (s, 3H), 2.65 (m, 1H), 2.54 (m, 1H),
2.37 (m, 1H), 2.24 (m, 1H), 1.83 (m, 3H), 1.69–1.65 (m, 3H), 0.52 (s, 1H,
NH), ꢀ1,66 ppm (s, 1H, NH); 13C NMR (CDCl3): d=189.60, 172.53,
172.14, 169.59, 161.11, 155.59, 150.93, 149.62, 145.16, 142.01, 137.90,
136.45, 136.22, 131.90, 131.81, 129.03, 128.95, 122.74, 118.38, 105.19,
104.38, 97.49, 93.08, 77.20, 65.21, 64.69, 52.84, 51.08, 50.09, 31.08, 29.75,
23.07, 19.39, 17.38, 12.09, 12.07, 11.17 ppm; high-resolution ESI-MS, posi-
tive mode: m/z calcd for C38H41N4O5 [M+H+]+: 633.3071; found:
633.3096.
3H), 1.50–1.00 (m, 4H), 1.30 ppm (m, 3H); 13C NMR (CDCl3): d=
195.54, 173.02, 168.67, 160.69, 155.89, 153.62, 150.55, 146.58, 146.34,
144.83, 144.50, 139.72, 135.97, 134.29, 132.55, 129.46, 122.03, 107.53,
105.42, 99.78, 92.45, 77.20, 50.59, 47.88, 47.34, 30.56, 28.56, 22.34, 19.61,
17.25, 12.78, 12.28, 11.10 ppm; high-resolution ESI-MS, positive mode:
m/z calcd for C33H32InN4O3 [MꢀClꢀ]+: 647.1508; found: 647.1501. This
complex was prepared before by a different method.[8]
Synthesis of conjugates of photosensitizers with either ODNs or 2’-OMe-
RNAs: Solid-support-bound 5’-amino-modified ODNs (MMT-HN–
ODNs; MMT: mono(4-methoxyphenyl)diphenyl) and their 2-OMe-RNA
analogues (MMT-HN–2’-OMe-RNAs) were prepared on an Expedite
8909 PNA/DNA synthesizer by using standard DNA or 2’-OMe-RNA
phosphoramidites (A and C: benzyl protected; G: dmf protected), stan-
dard solid supports (controlled pore glass (CPG)) with the first nucleo-
tide attached), and 5’-MMT-amino-modifier phosphoramidites from
Links Technologies (UK). 5’-MMT-HN–ODN2–BHQ-3 was prepared
analogously by using 3’-BHQ-3 CPG (Glen Research, USA) in place of
the standard CPG solid support. The MMT group was then cleaved by
treatment of the CPG-bound conjugates (1 mmol scale synthesis) with
1% trifluoroacetic acid (TFA) in CH2Cl2 (1 mL) for 3 min, washing of
the samples with 1% TFA in CH2Cl2 (2ꢄ2 mL) and CH2Cl2 (3ꢄ2 mL),
and drying at 0.01 mbar pressure for 1 h. Finally, the photosensitizers
were coupled under standard conditions. Typical protocols for these syn-
thetic steps are given below.
Pyropheophorbide a allyl ester: Pheophorbide a allyl ester (1.4 g,
2.21 mmol) was dissolved in 2,4,6-collidine (70 mL) and stirred for 5 h at
1708C under an argon atmosphere. After this, the solvent was evaporated
at 0.01 mbar pressure, and the crude product was purified by column
chromatography to yield the product (902 mg, 71%) as a black solid:
TLC (CH2Cl2/EtOAc 15:1): Rf =0.30; 1H NMR: (400 MHz, CDCl3): d=
9.40 (s, 1H), 9.30 (s, 1H), 8.55 (s, 1H), 7.99–7.92 (m, 1H), 6.28–6.13 (m,
2H), 5.88–5.78 (m, 1H), 5.29–5.09 (m, 2H), 5.26–5.15 (m, 2H), 4.54 (m,
2H), 4.50 (m, 1H), 4.31 (m, 1H), 3.63 (s, 3H), 3.60 (m, 2H), 3.40 (s, 3H),
3.17 (s, 3H), 2.72 (m, 1H), 2.59 (m, 1H), 2.33 (m, 2H), 1.83 (m, 3H),
1.68–1.65 (m, 3H), 0.38 (s, 1H, NH), ꢀ1.76 ppm (s, 1H, NH); 13C NMR
(CDCl3): d=196.16, 172.68, 171.29, 160.18, 155.08, 150.65, 148.92, 144.87,
141.46, 137.76, 136.08, 135.96, 135.72, 131.90, 131.47, 129.15, 128.22,
122.45, 118.55, 105.97, 103.96, 97.08, 92.95, 77.20, 65.25, 51.60, 49.92,
48.03, 31.02, 29.79, 23.13, 19.38, 17.40, 12.08, 12.00, 11.16 ppm; high-reso-
lution ESI-MS, positive mode: m/z calcd for C36H39N4O3 [M+H+]+:
575.3017; found: 575.3000.
Coupling of photosensitizers to nucleic acids and their mimics: The car-
ꢀ
ꢀ
boxylic acid (P OH or InP OH, 100 mmol), O-(benzotriazol-1-yl)-
N,N,N’,N’-tetramethyluronium hexafluorophosphate (HBTU; 34 mg,
90 mmol), and 1-hydroxy-1H-benzotriazole (HOBT; 14 mg, 100 mmol)
were dissolved in DMF (1 mL), and N,N-diisopropylethylamine (DIEA;
38 mL, 220 mmol) was added. This solution was vortexed and immediately
added to the CPG-bound nucleic acid strand. The slurry obtained was
left for 1 h with vigorous stirring. The CPG was filtered, washed with
DMF (2ꢄ1 mL) and CH3CN (2ꢄ1 mL), and dried under vacuum
(10ꢀ2 mbar). The conjugates were deprotected and cleaved from the solid
support with 27% aqueous ammonia solution at 228C for 24 h, followed
by removal of excess ammonia with a stream of nitrogen, removal of the
solid support by filtration, lyophilization of the filtrate, dissolution of the
residue in water, and HPLC purification.
In(pyropheophorbide a allyl ester)Cl: Pyropheophorbide a allyl ester
(226 mg, 0,393 mmol), InCl3 (1.356 g, 6.13 mmol), NaOAc (2.26 g,
27.6 mmol), and K2CO3 (2.26 g, 16.4 mmol) in benzene (68 mL) were
heated to reflux for 18 h under an argon atmosphere. The suspension was
then neutralized with acetic acid and extracted with several portions of
water. The combined aqueous phases were extracted with CH2Cl2. The
combined organic phases were dried over Na2SO4, and the solvent was
evaporated in vacuo. The crude product was purified by column chroma-
tography to give the product (175 mg, 62%) as a black solid: TLC
Determination of conjugate concentration by UV spectroscopy: HPLC-
purified conjugates were lyophilized and dissolved in water (250 mL).
Small samples from these solutions were diluted 250-fold with water, and
their absorbances at 260 nm were measured. These values were used for
calculation of the concentrations of the conjugates. Extinction coeffi-
cients of the conjugates were calculated as the sum of the extinction coef-
ficients of the corresponding natural nucleotides and photosensitizers.
1
ꢀ
(CH2Cl2/EtOAc 15:1): Rf =0.36; H NMR: (200 MHz, CDCl3): d=9.69 (s,
The extinction coefficient of the previously uncharacterized InP OH was
1H), 9.49 (s, 1H), 8.53 (s, 1H), 7.98–7.91 (m, 1H), 6.27–6.13 (m, 2H),
determined to be 12000mꢀ1 cmꢀ1 at 260 nm.
294
www.chemeurj.org
ꢃ 2010 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Chem. Eur. J. 2010, 16, 288 – 295

Controlled Photocatalysis in Live Cells
FULL PAPER
Conditions of the HPLC purification of the conjugates: For the majority
of the conjugates, the following gradient was used (solvent B: 1% H2O
in MeOH; solvent A: 0.1m triethylammonium acetate buffer, pH 7):
5 min at 0% B, 0!90% B in 25 min, 10 min at 90% B. For InP–ODN2–
Q, the following gradient was used (solvents as above): 5 min at 0% B,
0!90% B in 20 min, 10 min at 90% B. For P–ODN1, the following gra-
dient was used (solvent B: CH3CN; solvent A: 0.1m triethylammonium
acetate buffer, pH 7): 0!70% B in 35 min, 10 min at 90% B.
Nielsen, A. H. Thomsen, F. Kirpekar, K. V. Gothelf, J. Org. Chem.
2004, 69, 2240–2250; g) J. L. Czlapinski, T. L. Sheppard, J. Am.
Chem. Soc. 2001, 123, 8618–8619; h) J. L. Czlapinsky, T. L. Shep-
pard, ChemBioChem 2004, 5, 127–129; i) C. S. Andersen, H. Yan,
K. V. Gothelf, Angew. Chem. 2008, 120, 5651–5654; Angew. Chem.
Int. Ed. 2008, 47, 5569–5572.
[3] Recent papers on organic nucleic acid directed reactions: a) J. M.
Heemstra, D. R. Liu, J. Am. Chem. Soc. 2009, 131, 11347–11349;
b) T. N. Grossmann, O. Seitz, Chem. Eur. J. 2009, 15, 6723–6730;
c) R. M. Franzini, E. T. Kool, ChemBioChem 2008, 9, 2981–2988.
Recent papers on applications of template reactions in the discovery
of new drugs: d) B. N. Tse, T. M. Snyder, Y. Shen, D. R. Liu, J. Am.
Chem. Soc. 2008, 130, 15611–15626 and references therein; e) M. H.
Hansen, P. Blakskjær, L. K. Petersen, T. H. Hansen, J. W. Højfeldt,
K. V. Gothelf, N. J. V. Hansen, J. Am. Chem. Soc. 2009, 131, 1322–
1327.
P–ODN1: Yield=37%; HPLC: Rt =27.5 min; MALDI-TOF MS: m/z
calcd for C183H229N59O91P15 [MꢀH]ꢀ: 5173.20; found: 5173.20.
InP–ODN1: Yield=36%; HPLC: Rt =25.9 min; MALDI-TOF MS: m/z
calcd for C183H226N59O91P15In [Mꢀ2H+]ꢀ: 5285.00; found: 5285.40.
InP–ODN2–Q: Yield=21%; HPLC: Rt =20.4 min; MALDI-TOF MS:
+
ꢀ
m/z calcd for C332H405N116O165P28InNa [M+Na+ 3H ] : 9660.92; found:
9663.02.
ꢀ
InP–2’-OMe-RNA3: Yield=14%; HPLC: Rt =27.0 min; MALDI-TOF
[4] a) S. Sando, H. Abe, E. T. Kool, J. Am. Chem. Soc. 2006, 128, 263–
268; b) H. Abe, E. T. Kool, Proc. Natl. Acad. Sci. USA 2006, 103,
263–268.
+
ꢀ
ꢀ
MS: m/z calcd for
6102.44.
C207H261N63O117P16In [M 2H ] : 6113.71; found:
[5] a) Z. Pianowski, K. Gorska, L. Oswald, C. A. Merten, N. Winssing-
er, J. Am. Chem. Soc. 2009, 131, 6492–6497; b) K. Furukawa, H.
Abe, K. Hibino, Y. Sako, S. Tsuneda, Y. Ito, Bioconjugate Chem.
2009, 20, 1029–1036; c) R. M. Franzini, E. T. Kool, J. Am. Chem.
Soc. 2009, DOI: 10.1021/ja904138v.
[6] Ruthenium catalyst active in live cells: a) C. Streu, E. Meggers,
Angew. Chem. 2006, 118, 5773–5776; Angew. Chem. Int. Ed. 2006,
45, 5645–5648. Staudinger reaction with one of the substrates at-
tached to the surface of live cells: b) E. Saxon, C. R. Bertozzi, Sci-
ence 2000, 287, 2007–2010; c) M. J. Hangauer, C. R. Bertozzi,
Angew. Chem. 2008, 120, 2428–2431; Angew. Chem. Int. Ed. 2008,
47, 2394–2397, and references therein.
[7] a) A. Okamoto, K. Tanabe, T. Inasaki, I. Saito, Angew. Chem. 2003,
115, 2606–2608; Angew. Chem. Int. Ed. 2003, 42, 2502–2504; b) K.
Tanabe, H. Nakata, S. Mukai, S.-I. Nishimoto, Org. Biomol. Chem.
2005, 3, 3893–3897; c) E. Clꢁ, J. W. Snyder, N. V. Voigt, P. R. Ogilby,
K. V. Gothelf, J. Am. Chem. Soc. 2006, 128, 4200–4201.
[8] Y. Chen, X. Zheng, M. P. Dobhal, A. Gryshuk, J. Morgan, T. J.
Dougherty, A. Oseroff, R. K. Pandey, J. Med. Chem. 2005, 48, 3692–
3695.
[9] X. Ragꢅs, A. Jimꢆnez-Banzo, D. Sꢅnchez-Garcꢇa, X. Batllori, S.
Nonell, Chem. Commun. 2009, 2920, and references therein.
[10] C. Schweitzer, R. Schmidt, Chem. Rev. 2003, 103, 1685–1758.
[11] B. Stevens, S. R. Perez, J. A. Ors, J. Am. Chem. Soc. 1974, 96, 6846–
6850.
InP–2’-OMe-RNA4: Yield=11%; HPLC: Rt =26.8 min; MALDI-TOF
+
ꢀ
ꢀ
MS: m/z calcd for C208H262N66O117P16In [M 2H ] : 6168.81; found:
6151.45.
InP–2’-OMe-RNA5: Yield=14%; HPLC: Rt =26.7 min; MALDI-TOF
+
ꢀ
ꢀ
MS: m/z calcd for C248H316N76O145P20In [M 2H ] : 7415.61; found:
7414.55.
InP–2’-OMe-RNA6: Yield=14%; HPLC: Rt =26.4 min; MALDI-TOF
+
ꢀ
ꢀ
MS: m/z calcd for C250H317N81O145P20In [M 2H ] : 7510.61; found:
7502.38.
Cellular experiments: Streptolysin O (SLO) was used to reversibly per-
meabilize HL-60 cells in the suspension according to a previously report-
ed protocol.[23] In particular, SLO (1000 UmLꢀ1) was activated in PBS
containing 5 mm dithiothreitol and 0.05% bovine serum albumin (BSA)
for 2 h at 378C. HL-60 cells (1ꢄ106 cellsmLꢀ1) were washed twice with
PBS, resuspended in the serum-free medium (100 mL), permeabilized by
the addition of an optimized amount of SLO (10–15 U for 106 cells) and
variable concentrations of conjugates, and then incubated at 378C for
10 min. The cellular membrane was sealed by the addition of RPMI-1640
medium (300 mL) with 10% FCS and further incubation at 378C for 20–
30 min. The processes of permeabilization and sealing were monitored by
flow cytometry. After the sealing, the cells were washed with PBS and re-
suspended in the medium (300 mL) with 10% FCS containing 5 mm of a
singlet-oxygen scavenger (dye I or dye III). The cells were irradiated
with the red light for definite periods of time.
[12] S. Tyagi, D. P. Bratu, F. R. Kramer, Nat. Biotechnol. 1998, 16, 49–53.
[13] T. N. Grossmann, O. Seitz, J. Am. Chem. Soc. 2006, 128, 15596–
15597.
[14] J. F. Milligan, M. D. Matteucci, J. C. Martin, J. Med. Chem. 1993, 36,
1923–1937.
[15] C. Dose, S. Ficht, O. Seitz, Angew. Chem. 2006, 118, 5495–5499;
Angew. Chem. Int. Ed. 2006, 45, 5369–5373.
[16] Z.-Y. J. Zhan, D. G. Lynn, J. Am. Chem. Soc. 1997, 119, 12420–
12421.
Acknowledgements
We thank the German Research Council (Emmy-Noether Program,
grant no. MO1418/1-2 for A.M.), the Danish National Research Founda-
tion, and the Ruprecht-Karls-Universitꢀt Heidelberg for funding this
work.
[17] Y. Xu, N. B. Karalkar, E. T. Kool, Nat. Biotechnol. 2001, 19, 148–
152.
[18] H. Abe, E. T. Kool, J. Am. Chem. Soc. 2004, 126, 13980–13986.
[19] T. M. Cresp, R. G. F. Giles, M. V. Sargent, C. Brown, D. OꢈN. Smith,
J. Chem. Soc. Perkin Trans. 1 1974, 2435–2447.
[1] Recent reviews on nucleic acid directed reactions: a) A. P. Silver-
man, E. T. Kool, Chem. Rev. 2006, 106, 3775–3789; b) M. F. Jacob-
sen, E. Clꢁ, A. Mokhir, K. V. Gothelf, ChemMedChem 2007, 2, 793–
799; c) T. N. Grossmann, A. Strohbach, O. Seitz, ChemBioChem
2008, 9, 2185–2192.
[20] A. I. Meyers, D. L. Temple, D. Haidukewych, E. D. Mihelich, J. Org.
Chem. 1974, 39, 2787–2793.
[2] Papers on metal-complex-catalyzed chemical nucleic acid directed
reactions: a) I. Boll, E. Jentzsch, R. Kraemer, A. Mokhir, Chem.
Commun. 2006, 3447–3449; b) I. Boll, R. Kraemer, J. Brunner, A.
Mokhir, J. Am. Chem. Soc. 2005, 127, 7849–7856; c) J. Brunner, A.
Mokhir, R. Kraemer, J. Am. Chem. Soc. 2003, 125, 12410–12411;
d) F. H. Zelder, J. Brunner, R. Kraemer, Chem. Commun. 2004,
902–903; e) N. Graf, M. Gçritz, R. Krꢀmer, Angew. Chem. 2006,
118, 4117–4119; Angew. Chem. Int. Ed, 2006, 45, 4013–4015; f) M.
[21] J. Einhorn, Synlett 2006, 2035–2038.
[22] K. M. Smith, D. A. Goff, D. J. Simpson, J. Am. Chem. Soc. 1985,
107, 4946–4954.
[23] R. V. Giles, J. Grzybowski, D. G. Spiller, D. M. Tidd, Nucleosides
Nucleotides 1997, 16, 1155–1163.
Received: August 27, 2009
Published online: November 5, 2009
Chem. Eur. J. 2010, 16, 288 – 295
ꢃ 2010 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.chemeurj.org
295