Cooperative Veratryle and Nitroindoline Cages for Two-Photon Uncaging in the NIR

Article abstract of DOI:10.1002/chem.201601109

Tandem uncaging systems in which a two-photon absorbing module and a cage moiety, linked via a phosphorous clip, that act together by F?rster resonance energy transfer (FRET) have been developed. A library of these compounds, using different linkers and c

Full text of DOI:10.1002/chem.201601109

DOI: 10.1002/chem.201601109
Full Paper
&
Photochemistry |Hot Paper|
Cooperative Veratryle and Nitroindoline Cages for Two-Photon
Uncaging in the NIR
Eduardo Cueto Diaz, SØbastien Picard, Maxime Klausen, Vincent Hugues, Paolo Pagano,
Emilie Genin, and Mireille Blanchard-Desce*[a]
Abstract: Tandem uncaging systems in which a two-photon
absorbing module and a cage moiety, linked via a phospho-
rous clip, that act together by Fçrster resonance energy
transfer (FRET) have been developed. A library of these com-
pounds, using different linkers and cages (7-nitroindolinyl or
nitroveratryl) has been synthesized. The investigation of
their uncaging and two-photon absorption properties dem-
onstrates the scope and versatility of the engineering strat-
egy towards efficient two-photon cages and reveals surpris-
ing cooperative and topological effects. The interactions be-
tween the 2PA module and the caging moiety are found to
promote cooperative effects on the 2PA response while ad-
ditional processes that enhance the uncaging efficiency are
operative in well-oriented nitroindoline-derived dyads. These
synergic effects combine to lead to record two-photon un-
caging cross-section values (i.e., up to 20 GM) for uncaging
of carboxylic acids.
Introduction
trations. Additional key criteria such as uncaging sensitivity,
solubility, dark stability are also to be considered.[7–9] Moreover,
enlarging the concept of uncaging to two-photon excitation in
the near infrared (NIR) range has been a major breakthrough
as it offers intrinsic 3D resolution, reduced out-of-focus
damage and improved penetration in tissues.[18,19] This realiza-
tion initiated the development of specifically designed two-
photon sensitive protecting groups combining large two-
photon absorption (2PA) cross-section (s2) at suitable wave-
lengths (i.e., in the biological spectral window) with appropri-
ate uncaging quantum yield (Qu) and thus displaying high 2P
uncaging sensitivity (quantified by the 2P uncaging cross-sec-
tion, du =s2Qu).[20] One of promising ways towards enhanced
two-photon sensitivity is based on using well-known protect-
ing groups and embedding them within conjugated di-
polar,[21–25] quadrupolar[26–28] and octupolar[29] architectures to
improve their 2PA properties. This strategy led in several cases
to very interesting cages with du values over 1 GM.[21–28] Yet,
a drawback of this strategy is the possible alteration of the un-
caging efficiency as a result of the participation of the caging
subunit in the intramolecular charge redistribution responsible
for large 2PA.[30,31] This may affect the nature of the excited
state and lead to a decrease of the uncaging quantum yield as
compared to that of the isolated uncaging unit. The outcome
of this strategy has appeared to be extremely dependent on
the nature of the cage, as well as on the nature and position
of the structural modifications. Results are hardly rationalized
and this strategy, while being of major interest, remains mostly
empirical. Hence, one of the current challenges is the develop-
ment of versatile and modular routes which could provide
more sensitive 2P cages and easily be transposable to different
families of PPGs, while retaining their crucial characteristics.
Within this context, we focused our efforts on a less-explored
Photolabile protecting groups (PPGs) were historically used
first by organic chemists and became a valuable tool for appli-
cations in multistep organic synthesis, combinatorial chemistry
and solid-phase synthesis.[1–6] Several families of such com-
pounds have been described.[7–9] Their key common feature is
that they offer protection which is orthogonal to most other
protecting strategies as their cleavage requires only light irradi-
ation (classically in the near UV or visible range), without the
addition of chemical reagents. Moreover, by using focused
light pulses acting as external triggers, PPG removal can be
achieved with unique spatial and temporal control. Thanks to
these advantages, photosensitive groups have gained increas-
ing popularity in material sciences[10,11] as well as in biol-
ogy.[2,12,13] Numerous caged proteins and caged biomolecules,
the biological activities of which are temporarily masked by
the presence of a PPG, have been developed to investigate
cellular functions[14–16] or limit toxic effects in drug delivery.[17]
The selection of a cage mostly results from a compromise be-
tween its characteristics and the requirements dictated by the
target application. While systems for photochemical delivery of
bioactive volatiles should display slow kinetics to ensure long-
lasting effects, time-resolved studies of fast biological process-
es require PPGs that ensure very fast release and therefore
generate a sudden local surge of signaling molecule concen-
[a] Dr. E. Cueto Diaz, Dr. S. Picard, M. Klausen, V. Hugues, P. Pagano,
Dr. E. Genin, Dr. M. Blanchard-Desce
University of Bordeaux, Institut des Sciences MolØculaires
(UMR5255 CNRS), 351 cours de la liberation, 33450, Talence (France)
E-mail: mireille.blanchard-desce@u-bordeaux.fr
Supporting information for this article is available on the WWW under
http://dx.doi.org/10.1002/chem.201601109.
Chem. Eur. J. 2016, 22, 10848 – 10859
10848
ꢀ 2016 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim

Full Paper
strategy relying on the decoupling between the 2P excitation
and uncaging processes.
We recently developed tandem systems[32] where a 2P ab-
sorber subunit acts as an intramolecular sensitizer and trans-
fers its excitation energy to the grafted caging unit with the
aim to preserve its key chemical (including dark stability) and
photochemical (including uncaging kinetics) characteristics. By
combining two 7-nitroindolinyl (NI)-caged glutamates with
a hydrophilic quadrupolar tailored 2P chromophore, we pre-
pared triads affording a tenfold greater 2P-induced release of
glutamate (du =0.5 GM at 730 nm) in comparison with the
parent NI-protecting group. Pursuing the same goal, triads op-
erating by intramolecular electron transfer have been very re-
cently reported as 2P sensitive protecting groups of amino
acids with record 2P uncaging sensitivity.[33] Aiming at further
increasing the 2PA ability in the 700–800 nm region and as-
suming that the presence of a second caging unit within the
triad could affect the 2P uncaging efficiency of the overall ar-
chitecture (due to excitation energy trapping by the cage by-
product), we recently[34] elaborated dissymmetric synergic
dyads incorporating a nitroveratryl (NV)-derived protecting
group. These systems show enhanced s2 values (up to 300
GM), sixfold improved 2P uncaging sensitivity in the NIR as
well as the possibility to perform 2P photolysis at longer wave-
lengths (i.e., 800 nm) as compared to the isolated NV cage.
Our present goal is to investigate the scope as well as the ver-
satility of our strategy and develop dyads with 2P uncaging
sensitivity over 10 GM by using PPGs with larger Qu values.
Herein, we report the synthesis of a series of dyads bearing dif-
ferent 2PA subunits (either pure quadrupoles or slightly dis-
symmetrized derivatives) shown in Figure 1.
Scheme 1. Schematic representation of a group of dyads having the same
sensitizer and caging unit (NV) but different length in the linker.
This choice was motivated by the better (e.g., more than ten-
times higher) uncaging quantum yield of NI compared to NV
cages as well as key additional features of interest for applica-
tions in neurosciences. These include very good dark stability
in biological environments and submicrosecond time scale re-
lease of neuroactive amino acids such as glutamate or GABA
upon flash photolysis[35–38]). These two PPGs display moderate
2P uncaging efficiencies at 730 or 740 nm (du values ranging
from 0.03 GM[39] for NV to 0.06 GM[38] for NI) but the NI PPG
moiety was quite attractive in view of its use in neurosciences
due to the above-mentioned features.
We used linkers of different size and structure enabling the
grafting of the caging module to the phosphorus-based clip.
We selected either flexible aliphatic spacers (propyle C3 or
hexyle C6) or semirigid analogues featuring a triazole unit
(methyltriazole C1N3, propyltriazole C3N3, hexyltriazole C6N3;
Scheme 1 and Scheme 2). In addition to the popular NV caging
moiety, NI was selected as a UV absorbing protecting group.
We selected acetic acid as a general model of caged (bio)-
molecules featuring a COOH function (including neuro-amino
acids). We investigated both the photophysical and 2PA prop-
erties of all new dyads as well as their uncaging ability under
one-photon excitation in the UV (i.e., 365 nm) and two-photon
excitation in a NIR spectral window of interest for biological
applications (i.e., 700–1000 nm). This study provides evidence
that modulation of the spacer slightly affect the photochemical
properties and uncaging sensitivities while the 2PA response
of the dyads displays cooperative effect (either positive or neg-
ative) depending on the proximity of the polar uncaging
moiety to the 2PA module. Importantly, the modification of the
PPG (from NV to NI) allows the achievement of a major in-
crease in the uncaging quantum yield. In combination with po-
tentially large 2PA responses, this leads to very large 2P uncag-
ing action cross-section in the NIR region, demonstrating the
strength of the implemented engineering strategy.
Figure 1. Design of dyads for two-photon induced acetic acid release by
combination of a fluorene-cored quadrupolar or dissymmetric two-photon
absorber and a caging moiety as 4,5-dimethoxy-2-nitrobenzyl (NV) or 7-ni-
troindoline (NI).
Chem. Eur. J. 2016, 22, 10848 – 10859
www.chemeurj.org
10849
ꢀ 2016 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim

Full Paper
Scheme 3. Synthesis of NV subunits 6 and 8. Reagents and conditions: a) 1-
bromo-3-chloropropane or 1-bromo-6-chlorohexane, K2CO3, DMF, RT, over-
night (98% of 1a, 95% of 1b); b) nitric acid, 08C to RT, then NaBH4, metha-
nol or THF, RT, overnight (69% of 2a and 65% of 2b); c) NaN3, DMSO, 708C,
4 h (99%); d) Ac2O, Et3N, DMAP, CH2Cl2, RT, overnight (94%); e) 5,
CuSO4·5H2O, l-ascorbic acid sodium salt, TBAF, DMSO, RT, overnight (70%);
f) Ac2O, DMAP, Et3N, CH2Cl2, RT, overnight (95% of 7a, 99% of 7b); g) NaI,
acetone, 608C, 2 days, then hydroquinone, K2CO3, DMF, 408C, 36 h (51% of
8a, 47% of 8b).
by the reduction of the aldehyde group provided alcohols
2a[34] and b efficiently. Nucleophilic substitution of the chlorine
with sodium azide led to compound 3 which was subsequent-
ly treated with acetic anhydride to afford the benzylic ester 4
with an excellent yield. Finally, the introduction of the phenol
terminal grafting moiety was achieved using a one-pot two-
step protocol involving the in situ deprotection of the TMS-
protected alkyne 5[34] followed by a Huisgen 1,3-dipolar cyclo-
addition with azide 4 and the NV-caged acetate module 6 was
isolated in 70% yield.
The graftable NV derivatives 8a and 8b, bearing propyl and
hexyl spacers, respectively, were obtained from the key inter-
mediates 2a and 2b in a two-step sequence. Esterification
with acetic anhydride afforded compounds 7a[34] and b with
a chlorine extremity. Following this second route, the introduc-
tion of the phenol terminal grafting moiety was achieved by
means of a nucleophilic substitution with hydroquinone which
led to 8a[34] and b in moderate yields. The NI-based (14)
caging module featuring a phenol pendant moiety were pre-
pared in four steps, from the known N-acetyl-4-hydroxyindo-
line (9) derivative (Scheme 4). Protection of the phenol group
by using tert-butyldimethylsilyl chloride afforded compound
10. A nitration reaction under mild conditions in the presence
of copper nitrate[40] was subsequently performed to obtain
a mixture of both 5- and 7-nitro regioisomers 11 a and 11 b.
These products were separated by column chromatography on
silica gel, thus yielding 11 a and 11 b as pure regioisomer com-
pounds. The 7-nitro isomer 11 b was converted to the corre-
sponding propargylated derivative 12 by means of a silyl
group deprotection reaction followed by propargylation. Ap-
plying a one-pot procedure, 11 b was therefore treated with
tetra-n-butylammonium fluoride (TBAF) and the resulting un-
stable phenolate was treated in situ with propargyl bromide to
Scheme 2. Comparison of dyads bearing similar sensitizer and linker but dif-
ferent caging units (NV and NI).
Results and Discussion
Synthesis of the graftable subunits
The synthesis of the graftable 6-nitroveratryl acetate subunit 6
bearing a hexyltriazole spacer was achieved by adapting the
synthetic route recently reported for the preparation of the an-
alogue with a shorter propyltriazole linker.[34] The NV derivative
6 was obtained in six steps starting from vanillin in a satisfacto-
ry 40% overall yield (Scheme 3). Alkylation of vanillin with
either 1-bromo-3-chloropropane or 1-bromo-6-chlorohexane in
the presence of anhydrous potassium carbonate in DMF af-
forded compounds 1a[34] and b in almost quantitative yields.
The nitration of benzaldehydes 1a–b with nitric acid followed
Chem. Eur. J. 2016, 22, 10848 – 10859
www.chemeurj.org
10850
ꢀ 2016 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim

Full Paper
cently reported (see the Supporting Information for the syn-
thesis of M4).[34,42,43] All synthetic intermediates as well as graft-
1
able subunits were fully characterized by H and 13C NMR spec-
troscopy and mass spectrometry.
Synthesis of the synergic dyads
The preparation of the small library of synergic structures bear-
ing various absorbing and caging modules is depicted in
Scheme 5. Taking advantage of the efficient nucleophilic sub-
stitutions of the chlorine atoms in the P(S)Cl2 moiety, by phe-
nolic derivatives, the synthetic pathway was based on the se-
quential functionalization of the clipping scaffold 15.[34] Impor-
tantly, the mono-functionalization of only one PÀCl bond
among the two is possible thanks to the significantly lower re-
activity of the P(S)(OAr)Cl moiety compared to P(S)Cl2. To pre-
pare dyads featuring NV and NI moieties, we then applied
a two-step one-pot procedure involving first the coupling of
15 with a stoichiometric amount of caging subunits 6 or 8a–
b or 14, followed by the substitution of the remaining Cl in
the presence of the phenolic anion of chromophoric subunits
Scheme 4. Synthesis of MNI14. Reagents and conditions: a) TBDMSCl, imid-
azole, DMF, RT, overnight (96%); b) Cu(NO3)2·3H2O, CH2Cl2/Ac2O, RT, over-
night (26% of 11 a, 28% of 11 b); c) propargyl bromide, KI, TBAF, THF, RT,
overnight (93%); d) l-ascorbic acid Na salt, CuSO4·5H2O, DMSO, RT, over-
night (92%).
give 12 with excellent yield. Huisgen 1,3-dipolar cycloaddition
with 4-azidophenol[41] 13 provided the graftable NI subunit 14.
The synthesis of the graftable dissymmetric two-photon ab-
sorbing modules M1–4 (Scheme 5) was achieved by stepwise
functionalization of a diidofluorene core involving an in situ
deprotection/double Sonogashira coupling protocol, as we re-
M1–4
.
The synthesis of the dyads was conducted in the dark at
room temperature under inert atmosphere. A THF solution of
one equivalent of the graftable caging module was added
dropwise to a THF solution containing the phosphorus-based
clip 15 and cesium carbonate. The reaction proceeded
smoothly in few hours and the reaction progress was moni-
1
tored by 31P and H NMR spectroscopy. The completion of the
first step was demonstrated by the appearance of a singlet at
approximately 69 ppm in the 31P spectrum attributed to the
P(S)(OAr)Cl group (Figure 2).
The completion of the first step was further confirmed by
the disappearance of the signals corresponding to the phenol
1
moiety of the starting material 6 or 8a–b or 14 in the H NMR
spectrum. A THF solution of the graftable 2PA module M1–4
was then added dropwise to the reaction mixture in order to
substitute the second PÀCl bond. Once again, 31P NMR spec-
troscopy was an invaluable tool for monitoring the reaction
progress. The total disappearance of the signal at approximate-
ly 69 ppm and appearance of a more shielded singlet at
63.8 ppm proved the completion of the second step after 18 h
(Figure 2). Cesium salt residues were removed by centrifuga-
tion of the reaction mixture. The D1 series of dyads with alkyl
or hexyltriazole spacers as well as dyads D2–4’ incorporating an
NI-caged acetate were then obtained with moderate yields as
pure compounds after column chromatography (which leads
to loss of compound explaining the modest yield in purified
1
dyads). Characterization by H and 31P NMR spectroscopy and
MALDI-TOF mass spectrometry unambiguously proved the
structure and the high purity of all the synthesized dyads.
Photophysical and two-photon absorption properties
Scheme 5. Synthesis of dyads. Reagents and conditions: a) 6 or 8a–b, 15,
Cs2CO3, THF, RT, 16 h and then M1 or M4, THF, RT, 18 h (8% for D1(C3)[Ac],
12% for D1(C6)[Ac], 24% for D1(C6N3)[Ac], 59% for D4(C3)[Ac]); b) 14, 15,
Cs2CO3, THF, RT, 16 h and then M2, M3 or M4, THF, RT, 18 h (23% for
D2’(C1N3)[Ac], 28% for D3’(C1N3)[Ac], 25% for D4’(C1N3)[Ac]).
The photophysical and two-photon absorption (2PA) proper-
ties of the dyad compounds were investigated in chloroform
and the corresponding data are gathered in Table 1. For the
Chem. Eur. J. 2016, 22, 10848 – 10859
www.chemeurj.org
10851
ꢀ 2016 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim

Full Paper
Figure 2. 31P NMR monitoring in CDCl3 of the synthesis of the D3’(C1N3)[Ac]: dyad: spectrum of 15 (bottom); spectrum of the reaction mixture 15 and 14
(middle); spectrum of the D3’(C1N3)[Ac] synergic dyad (top).
sake of comparison, the characteristics of the 2PA modules are
also included.
end-group increases. Importantly, the superimposition of the
emission spectra of the 2PA modules M1–M4 with the absorp-
tion spectra of the NV and NI cages highlights overlaps
(Figure 3) which markedly decrease on going from M1 to M2,
M4 and M3, thus suggesting that resonant-energy transfer from
M1 modules to the close NV or NI cages should be more effi-
cient. This trend is confirmed when comparing the fluores-
cence quantum yields of isolated 2PA modules and those of
the corresponding dyads. Indeed, dyads D1, bearing M1 as
a 2PA module, show a marked fluorescence quantum yield re-
duction (varying between 45 and 30% depending on the
nature of the linker between 2PA modules and uncaging
moiety). In contrast dyads D2, D3 and D4 having M2, M3 and M4
two-photon absorbing modules show smaller fluorescence
Photophysical properties
The 2PA modules as well as the dyads display an intense ab-
sorption band in the near-UV region with maximum molar ex-
tinction coefficient values ranging from 6.5 to 9.0”
104 mÀ1 cmÀ1, which are at least tenfold larger than that of the
NV and NI cages. The isolated 2PA modules show intense fluo-
rescence emission in the near-UV–blue visible region with
large fluorescence quantum yields (ꢀ0.70). As anticipated,
a bathochromic shift of both the absorption and emission
bands is observed when the strength of the electron-releasing
Table 1. Photophysical data of dyads and of their corresponding 2PA subunit and uncaging modules in CHCl3.
max
max
ref
max
max
max
Cpd
labs
emax
[mÀ1 cmÀ1
e365nm
[mÀ1 cmÀ1
lem
Ff[a]
Qu/Qu
e
maxQu
2labs
[nm]
l2PA
s2
du
[GM][d]
[b]
[nm]
]
]
[nm]
[mÀ1 cmÀ1
]
[nm]
[GM][c]
NVOAc
MNI
M1
341
302
358
358
358
358
358
368
367
367
375
373
375
385
386
386
5.9”103
4.8”103
7.2”104
7.5”104
9.0”104
8.1”104
8.3”104
7.8”104
7.6”104
6.9”104
6.5”104
6.8”104
6.5”104
7.9”104
7.0”104
7.1”104
4.0”103
1.9”103
6.6”104
6.8”104
7.2”104
7.4”104
7.5”104
7.7”104
7.6”104
6.9”104
6.1”104
6.5”104
6.2”104
5.6”104
5.4”104
5.4”104
–
–
–
–
1
1
–
35
384
–
146
197
174
128
–
48
1510
–
55
490
–
–
30
55
50
0.03[e]
0.06[f]
–
0.11
0.11
0.10
0.05
–
0.10
3.7
–
0.25
2.3
384
383
383
383
383
435
433
434
446
446
446
431
431
431
0.70
0.41
0.45
0.49
0.38
0.69
0.57
0.58
0.65
0.57
0.57
0.75
0.65
0.67
716
716
716
716
716
736
734
734
750
746
750
770
772
772
720
730
730
730
730
750
750
770
800
800
800
710
710
710
D1(C3)[Ac]
D1(C3N3)[Ac][34]
D1(C6)[Ac]
D1(C6N3)[Ac]
M2
D2(C3)[Ac][34]
D2’(C1N3)[Ac]
M3
D3(C3)[Ac][34]
D3’(C1N3)[Ac]
M4
0.32
0.36
0.36
0.26
–
0.11
0.26
–
0.13
0.09
–
0.13
0.19
45
30
195
160
170
240
310
300
1370
1300
1270
D4(C3)[Ac]
D4’(C1N3)[Ac]
53
1130
0.98
20
[a] Fluorescence quantum yield, standard: quinine in 0.5m H2SO4 (F=0.546). [b] One-photon uncaging sensitivity derived from comparative one-photon
max
photolysis experiments at 365 nm, using Qu =0.006 for NVOAc and Qu =0.08 for MNI. [c] Two-photon absorption cross-section at l2PA derived from TPEF
experiments (1 GM=10À50 cm4 sÀ1). [d] Two-photon uncaging sensitivity (du =s2maxQu) [e] From reference [39] at 740 nm, [f] From reference [38] at 730 nm.
Chem. Eur. J. 2016, 22, 10848 – 10859
www.chemeurj.org
10852
ꢀ 2016 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim

Full Paper
Only M1 and M4 are essentially quadrupolar in nature.
Hence, their 2PA response at twice the maximum (one-photon)
absorption wavelength is rather modest (Table 1) in relation
with the fact that the lowest excited is strongly one-photon al-
lowed but almost two-photon forbidden. The higher energy,
strongly two-photon allowed, excited state is clearly observed
for M4 (at 710 nm) while it falls out of the NIR range in the
case of M1 (i.e., at l2PA <700 nm). Conversely, M2 and M3 are
not purely quadrupolar as they bear two different electron-do-
nating end-groups, thus relieving the symmetry interdiction
for 2PA to the lowest excited state. As a result, both show sig-
nificant 2PA responses in the 750–800 nm spectral range corre-
sponding to the transition to the lowest (strongly one-photon
allowed) excited state, with respective maxima located at 750
Figure 3. Normalized absorption spectra of NV and NI cages and emission
spectra of 2PA chromophores M1–4 in chloroform.
max
and 800 nm, and corresponding s2
values of, respectively,
approximately 160 GM and about 300 GM (Table 1). Quite inter-
estingly, we observe that the close proximity between the un-
caging (NV or MNI) moiety and 2PA modules (M1–M4) influen-
ces their 2PA response.
quantum yield decrease (16–17% for dyads D2 and D2’, 11–
13% for dyad D4 and D4’ and 12% for dyads D3 and D3’) which
is consistent with the decreasing overlap of the absorption
spectrum of the acceptor (NV or NI) and emission of the donor
(i.e., M2, M4, M3).
It is interesting to note that the M3 2PA module, though dis-
playing blue-shifted absorption as compared to M4, shows red-
shifted emission, thus leading to slightly smaller overlaps. This
might be related to the intrinsic electronic dissymmetry of the
M3 module compared to M4. This dissymmetry promotes sym-
metry breaking in the excited state leading to more polar ex-
cited states which are stabilized in medium to high polarity en-
vironments, leading to a red-shifted emission.[44]
Comparison of the 2PA spectra of the series of D1 dyads
with that of their 2PA module indeed demonstrates the influ-
ence of the proximity of the uncaging moiety on the dyads
2PA response (Figure 4). The presence of the uncaging moiety
increases the maximum 2PA response in the 700–1000 nm
Comparison of D1 dyads, bearing the M1 2PA module, shows
that the length and nature of the linker between the 2PA and
uncaging subunits slightly influence their photochemical be-
havior. In the case of aliphatic linkers, the shortest linker seems
to lead to the most pronounced decrease of the fluorescence
quantum yield (see comparison of D1(C3)[Ac] and D1(C6)[Ac]
dyads). On the other hand the presence of the triazole moiety
in the linker seems to have a distinct influence. Indeed
D1(C3N3)[Ac] dyad shows a slightly more pronounced fluores-
cence reduction than dyad D1(C6)[Ac] (36 versus 30%), al-
though the length of their spacers are similar. Additionally,
D1(C6)[Ac] dyad which has the longest spacer shows the larg-
est fluorescence quenching which amounts up to 46%. This
unusual observation points to the possible role of the triazole
moiety as a passive chromophore mediating resonant-energy
transfer between the 2PA module and uncaging moieties.[45] Fi-
nally, we observe that dyads D2 and D2’ as well as D3, D3’ and
D4, D4’ show a similar fluorescence decrease as expected from
the overlap between M2–M4 in the emission spectrum and the
NV and MNI absorption spectra (Figure 3).
Two-photon absorption properties
Both 2PA modules and corresponding dyads display a broad
2PA band in the NIR region the intensity of which strongly de-
pends on the nature of the 2PA module (Table 1 and the Sup-
porting Information).
Figure 4. Two-photon absorption spectra of D1 dyads (top) and D3-D3’dyads
(bottom) compared to their 2PA module (M3).
Chem. Eur. J. 2016, 22, 10848 – 10859
www.chemeurj.org
10853
ꢀ 2016 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim

Full Paper
spectral range (i.e., biological window) by up to 85% (Table 1).
We note that this effect decreases with increasing distance be-
tween the uncaging and 2PA module being maximum for the
D1(C3)[Ac] dyad (in which the two modules are the closest)
and vanishing for the D1(C6N3)[Ac] dyad having the longest
spacer. This can possibly be related to the effect of the electro-
static electric field generated by the dipolar uncaging moiety
(thus depending markedly on proximity) on the 2PA module.
Previous reports have shown that the two-photon absorption
response of both dipolar and quadrupolar chromophores can
indeed be influenced by electrostatic interactions.[43,46] As a con-
sequence the D1(C3)[Ac] dyad shows a maximum 2PA cross-
section value at 720 nm which, while modest, is still more than
one order of magnitude larger than that of the isolated NV
cage. Hence, the dyad strategy also provides an indirect way
to enhance the 2PA response by taking advantage of the dipo-
lar nature of uncaging moieties and electrostatic effects.
Figure 5. Kinetics of acetic acid photorelease induced upon irradiation at
365 nm for the series of D1-based dyads and the reference NVOAc cage.
We also observe a definite effect of the spatial proximity be-
tween uncaging moieties and 2PA on the 2PA responses of dis-
symmetric dyads D2-D2’ and D3-D3’. Quite interestingly, a de-
crease of the 2PA response by about 15% is observed for D2-
’
D2 dyads compared to their 2PA module M2, whereas an in-
crease of 27% is observed for D3-D3’ dyads (Figure 4 and
Table 1). This reverse trend can be related to the opposite ori-
entation of the resulting dipole of the dissymmetric 2PA
module M2 and M3 with respect to the caging dipolar moieties.
D4-D4’ dyads display an intermediate behavior with only a very
slight decrease of the peak 2PA response (Table 1 and the Sup-
porting Information).
Figure 6. Comparison of the kinetics of acetic acid photoinduced release
(upon irradiation at 365 nm) between NV- and MNI-derived dyads.
As seen from Table 1, the length and the nature of the
spacer are found to slightly affect the uncaging sensitivity of
the series of D1 dyads (Figure 5). D1(C6N3)[Ac] dyad displays
the lowest relative Qu ratio (0.26), while D1(C3N3)[Ac] and
D1(C6)[Ac] exhibit the highest Qu ratio value (0.36). This sug-
gests an optimum linker length that leads to the highest un-
caging quantum yield. As a result, the one-photon uncaging
sensitivities (i.e., emaxQu) of D1 dyads are thus enhanced by
a factor 4 to 6 in comparison with NVOAc (Table 1). In contrast
D2, D3 and D4 dyads show lower relative Qu ratio values than
D1 dyads, in agreement with the smaller overlaps between the
One- and two-photon uncaging studies
One-photon uncaging properties
The uncaging properties of the dyads were first investigated
by performing one-photon photolysis of samples in CDCl3
upon excitation at 365 nm. For direct comparison, photolysis
of the isolated NV and NI-caged acetate were also carried out
in the same experimental conditions (i.e., at the same absorb-
ance at 365 nm and the same excitation power).
NV absorption spectrum and the different 2PA modules (M2–4
)
The time courses for the photolysis reactions were moni-
emission spectra.
1
tored by H NMR analysis. The production of the photoreleased
As a result D2, D3 and D4 dyads show one-photon uncaging
sensitivities which are only slightly larger than that of NVOAc
(Table 1). On the other hand, we observe that D2’, D3’ and D4’
dyads undergo much faster one-photon photolysis at 365 nm
compared to their NV-bearing counterparts (Figure 6 and the
Supporting Information). This demonstrates the scope of the
implemented strategy in terms of sensitizing various cages.
Indeed by using more efficient (i.e., having larger uncaging
quantum yield) cages such as NI, dyads with larger uncaging
sensitivities are achieved (typically from 500 to 1500mÀ1 cmÀ1,
which is fourfold larger than that of isolated MNI). Quite inter-
estingly, we notice that the Qu ratio values are different for the
analogous NV and MNI dyads. Indeed D2’(C1N3)[Ac] and
D4’(C1N3)[Ac] dyads display significant Qu ratio values (0.26 and
0.19)—larger than that of D2(C3)[Ac] and D4(C3)[Ac] dyads—de-
spite the very small spectral overlap between absorption of NI
acetic acid was evidenced by the onset of a singlet at
2.10 ppm (dCH3), while the signal of the caged acetyl group at
2.16 or 2.23 ppm (dCH3), in the case of NV and MNI-based
dyads, respectively, concomitantly vanished. Quantitative anal-
ysis of the amount of photoreleased acetic acid allowed the
plotting of the conversion rate over time. All derivatives dis-
play first-order kinetics profiles from which the corresponding
slope values (photolysis rate constants) could be derived
(Figure 5 and Figure 6). Comparison of the slope values, which
are proportional to e(365nm)Qu, provided relative Qu values for
each dyad with respect to that of the corresponding isolated
ref
caging moiety (Qu/Qu values reported in Table 1). One-photon
uncaging sensitivities of dyads were then calculated from the
Qu values reported in the literature for NV[39] (0.006), MNI[38]
(0.08) cages.
Chem. Eur. J. 2016, 22, 10848 – 10859
www.chemeurj.org
10854
ꢀ 2016 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim

Full Paper
and emission of M2 and M4. In contrast, D3’(C1N3)[Ac] dyad
shows a low Qu ratio value (0.09), slightly smaller than that of
D3(C3)[Ac]. Hence, the Qu ratio value increases for the D2’ dyad
as compared to the D2 dyad whereas it slightly decreases for
the D3’ dyad as compared to the D3 dyad. This singular behav-
ior suggests that additional processes contribute to the uncag-
ing mechanism in the case of D2’ and D4’ dyads. This could be
related in particular to triplet–triplet sensitization (as reported
earlier in the case of NI-derived dyads[47]) or to electron transfer
mediated uncaging.[33,48,49] The contribution of electron transfer
from the excited 2PA module to the MNI moiety might be sig-
nificant in particular within D4’ (M4 having the strongest elec-
tron-donating end-groups)[33] as well as for the D2’ dyad
(where electron transfer is favored from a topological point of
view, due to the closer proximity of the strongest electron-do-
nating end-group).
caging quantum yields. As previously reported, D2-D3 dyads
show larger 2P uncaging sensitivities (up to 0.25 GM).[34] Nota-
bly, the D4 dyad shows a larger 2P uncaging sensitivity (i.e.,
1 GM), leading to the most efficient uncaging system among
the NV-derived dyads.
More importantly, replacing the NV unit by an NI caging
moiety having a tenfold larger uncaging quantum yield led to
a
major increase in the 2P uncaging cross-sections.
D2’(C1N3)[Ac], D3’(C1N3)[Ac] and D4’(C1N3)[Ac] dyads were
found to be up to more than two orders of magnitude more
sensitive than the NI reference, thus affording large du values
of, respectively, 3.7 GM (at 770 nm), 2.3 GM (at 800 nm) and
20 GM (at 710 nm). In particular, the D4’(C1N3)[Ac] dyad, which
benefits both from the strongly two-photon allowed transition
at 710 nm and significant uncaging quantum yield (in relation
with the contribution of different mechanisms in addition to
FRET) show the highest two-photon uncaging cross-section. To
the best of our knowledge, this record value overcomes those
reported up to now for most functional uncaging systems for
the release of acids, opening a new avenue for uncaging appli-
cations in neurosciences.
As a consequence, the one-photon uncaging sensitivity
(emaxQu) of D2’(C1N3)[Ac] shows a 30-fold enhancement com-
pared to D2(C3)[Ac] and that of D4’(C1N3)[Ac] a 20-fold en-
hancement compared to D4(C3)[Ac] while D3’(C1N3)[Ac shows
only a ninefold enhancement compared to D3(C3)[Ac]. This
markedly different enhancement factor clearly emphasizes the
role of the topology of the dyads in triggering cooperative ef-
fects on uncaging efficiency of the NI cage.
Based on these results, 2P photolysis experiments were con-
ducted with the NI-derived dyads which show large 2P uncag-
ing action cross-sections. Namely, the dyads were dissolved in
CDCl3 and irradiated at wavelengths corresponding to maxi-
mum 2PA responses within the spectral range of interest for
biological applications (690–1000 nm) using a Ti-sapphire laser
delivering 140 fs pulses at 80 MHz repetition rate. After few
hours of irradiation (this typical duration is required due to the
extremely small fraction of the total volume of the cuvette
(mL) addressed with highly confined two-photon excitation),
the production of 2P photoreleased acetic acid was assessed
Two-photon uncaging properties
Maximum 2P uncaging action cross-sections (du =s2maxQu)
were then derived for all dyads from the experimentally deter-
max
mined s2
and Qu values, assuming that one-photon and
two-photon uncaging quantum yields are the same (Table 1).
Except for the twofold less sensitive D1(C6N3)[Ac] derivative,
D1-based dyads have threefold larger 2P uncaging sensitivities
(ꢀ0.10 GM at 730 nm) compared to the NV reference, whatev-
er the spacer featured. This results from the variation of the
2PA responses which compensates for the variation of the un-
1
by H NMR analysis (Figure 7 and the Supporting Information).
Due to the experimental conditions (total volume, high repeti-
tion rate of the fs laser and duration of the experiment), the
observed kinetics of the overall process depends both on the
Figure 7. 1H NMR spectrum (300 MHz) of D4’(C1N3)[Ac] in CDCl3 at t=0 (bottom) and after 4 h two-photon irradiation at 720 nm (top).
Chem. Eur. J. 2016, 22, 10848 – 10859 www.chemeurj.org ꢀ 2016 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
10855

Full Paper
2P-uncaging action cross-section (which influences the photo-
lysis kinetics in the focal spot) and on the diffusion of reagent
(dyad) and side-product (dyads having released acetic acid) in
and out of the focal spot.
Experimental Section
Synthesis
Synthesis of D1-C3[Ac]: Compound 15 (23.3 mg, 78.5 mmol) and
Cs2CO3 (102 mg, 0.31 mmol) under inert atmosphere were sus-
pended in 3 mL of anhydrous THF. Then 8a (27.6 mg, 71 mmol),
dissolved in 2 mL of THF, was added. The reaction mixture was
stirred at room temperature for 16 h. M1 (53 mg, 79 mmol), dis-
solved in 2 mL of anhydrous THF, was added and the mixture was
stirred for another 18 h. The solvent was removed under reduced
pressure. The crude was dissolved (partially) in CH2Cl2. The suspen-
sion was centrifuge-sedimented and the supernatant was evapo-
rated. The crude was purified on silica gel (gradient eluent CH2Cl2/
EtOAc, 1:0 to 9:1) to give D1-C3[Ac] (10 mg, 8%) as a pale yellow
powder. 1H NMR (300 MHz, CDCl3): d=0.59–0.71 (m, 10H), 1.03–
1.15 (m, 4H), 1.95–1.99 (m, 4H), 2.16 (s, 3H), 2.23–2.33 (m, 4H),
Conclusions
The route toward more efficient 2P sensitive PPGs that we
have explored relies on the binding, via a phosphorous clip, of
a tailored 2P absorber with a caging module meant to be sen-
sitized through energy transfer from the two-photon excited
2PA module. A small library of dyads featuring modulations of
the spacer (length and nature), 2PA module and PPG (NV, NI)
was successfully prepared. The comprehensive study demon-
strates the modularity and versatility of this synergic strategy
and reveals interesting cooperative and topological effects. In
particular, the dipolar interactions between the 2PA modules
and the caging moieties are found to promote cooperative
effect (either negative or positive) on the 2PA response of the
dyads compared to that of isolated 2PA modules. This effect is
modulated by the length of the linkers as well as the nature
and topology of the 2PA module. As intended, replacing the
NV uncaging moiety by a more efficient one having a tenfold
larger uncaging quantum yield (e.g., NI) indeed leads to much
larger two-photon uncaging sensitivity. Moreover, additional
contributions to the uncaging process are found to increase
the global uncaging efficiency of specific MNI-derived dyads.
Indeed, electron transfer seems to promote uncaging efficiency
in specific cases where both the electron-donating capacity of
the 2PA module and the topology of the dyad (i.e., relative po-
sitioning of the electron-donating site and NI moiety) are fa-
vorable leading to increased uncaging efficiency by 20- or 30-
fold with respect to their NV-derived counterparts. When com-
pared to the tenfold enhancement between single NV and MNI
cages, this unveils interesting cooperative uncaging effects op-
erating within NI-derived dyads.
3
3.30 (d, JHP =10.6 Hz, 3H), 3.84 (s, 3H), 3.85 (s, 3H), 3.93 (s, 3H),
4.09–4.18 (m, 6H), 4.26 (t, J=6.1 Hz, 2H), 5.49 (s, 2H), 6.81–6.98 (m,
12H), 7.13–7.16 (m, 4H), 7.53–7.48 (m, 8H), 7.61–7.75 ppm (m, 6H);
13C NMR (75 MHz, CDCl3): d=14.0, 21.0, 23.2, 26.0, 29.1, 29.4, 29.9,
3
33.2 (d, Jcp=13.2 Hz, N-Me), 40.4, 55.3, 55.5, 55.5, 56.5, 63.5, 64.4,
64.6, 64.8, 66.2, 89.3, 89.9, 109.8, 111.0, 114.2, 114.4, 114.7, 115.2,
115.6, 115.7, 120.0, 122.3, 122.6, 122.6, 125.9, 127.1, 128.0, 128.6,
130.7, 133.2, 139.4, 139.6, 140.1, 140.6, 144.5, 147.7, 151.2, 154.0,
156.2, 156.3, 159.0, 159.8, 160.9, 170.5 ppm; 31P NMR (121.4 MHz,
CDCl3): d=64.8 ppm (s, P0); ESIHMRS: m/z: calcd for C75H76N3O13PS
[M]+: 1289.4836; found: 1289.4782 [M]+.
Synthesis of D1-C6[Ac]: Compound 15 (45 mg, 0.15 mmol) and
Cs2CO3 (197 mg, 0.6 mmol) under inert atmosphere were suspend-
ed in 3 mL of anhydrous THF. Then 8b (65 mg, 0.15 mmol), dis-
solved in 2 mL of THF, was added. The reaction mixture was stirred
at room temperature for 16 h. M1 (77 mg, 0.11 mmol), dissolved in
2 mL of anhydrous THF, was then added and the reaction mixture
was stirred for another 18 h. The solvent was removed under re-
duced pressure. The crude was dissolved (partially) in CH2Cl2. The
suspension was centrifuge-sedimented and the supernatant was
evaporated. The crude was purified on silica gel (gradient eluent
CH2Cl2/EtOAc, 1:0 to 9:1) to give D1-C6[Ac] (25 mg, 12%) as a pale
yellow powder. 1H NMR (400 MHz, CDCl3): d=0.54–0.70 (m, 10H),
1.0.3–1.16 (m, 4H), 1.53–1.55 (m, 4H), 1.77–1.92 (m, 4H), 1.96–2.01
3
(m, 4H), 2.16 (s, 3H), 2.23–2.27 (m, 2H), 3.30 (d, JHP =10.6 Hz, 3H),
In combination with the large 2PA response of the 2PA
module bearing the stronger electron-donating end-groups,
this effect leads to the measurement of two-photon uncaging
cross-section values for uncaging of organic acids (i.e., up to
20 GM). This value overcomes those reported to date for the
best functional uncaging systems and lies in the range of the
values sought for efficient and confined photorelease of neuro-
transmitters (such as glutamate or GABA) in biological environ-
ments under two-photon excitation in the NIR.[50] In that re-
spect, the excellent dark hydrolytic stability of the NI moiety
and its ability to release l-glutamate on a submicrosecond
time scale upon flash photolysis make such dyads very promis-
ing for further developments in the neurosciences. Building on
the present results, we will thus proceed to replace caged
acetic acid by neurotransmitters such as glutamate or GABA
and convey solubility in physiological conditions by incorporat-
ing suitable hydro-solubilizing groups.[23,51,52] This further
chemical modification will also be of major interest for direct
comparison of the best synergic cages with model cages in
identical conditions.
3.85 (s, 6H), 3.91 (t, J=6.4 Hz, 2H), 3.95 (s, 3H), 4.06–4.19 (m, 6H),
5.49 (s, 2H), 6.78–6.98 (m, 12H), 7.11–7.16 (m, 4H), 7.48–7.53 (m,
8H), 7.61–7.70 ppm (m, 6H); 13C NMR (100 MHz, CDCl3): d=14.0,
3
21.0, 23.2, 25.8, 26.0, 26.0, 28.9, 29.3, 29.4, 33.3 (d, JCP =12.8 Hz),
40.4, 55.3, 55.5, 55.5, 56.5, 63.5, 64.6, 64.8, 68.2, 69.5, 89.3, 89.9,
109.5, 110.9, 114.2, 114.3, 114.7, 115.1, 115.1, 115.6, 115.6, 115.8,
116.2, 120.0, 122.3, 122.5, 122.6, 122.6, 122.6, 125.9, 126.8, 128.0,
3
128.6, 130.7, 133.2, 139.3, 139.5, 140.2, 140.6, 144.3 (d, JCP =7.1 Hz
CH=N), 144.4, 144.5, 147.9, 151.2, 153.9, 156.3, 156.5, 159.0, 159.8,
160.9, 170.5 ppm; 31P NMR (121.4 MHz, CDCl3): d=64.8 ppm (s, P0);
ESIHMRS: m/z: calcd for C78H82N3O13PS [M]+: 1331.5306; found:
1331.5350 [M]+.
Synthesis of D1-C6N3[Ac]: Compound 15 (50 mg, 0.17 mmol) and
Cs2CO3 (219 mg, 0.67 mmol) under inert atmosphere were sus-
pended in 5 mL of anhydrous THF. Then 6 (77 mg, 0.14 mmol), dis-
solved in 2 mL of THF, was added. The mixture was stirred at room
temperature for 18 h. M1 (114 mg, 0.17 mmol), dissolved in 2 mL of
anhydrous THF, was added and the reaction mixture was stirred for
another 18 h. The solvent was removed under reduced pressure.
The crude was dissolved (partially) in CH2Cl2. The suspension was
centrifuge-sedimented and the supernatant was evaporated. The
crude was purified on silica gel (gradient eluent CH2Cl2/EtOAc, 1:0
Chem. Eur. J. 2016, 22, 10848 – 10859
www.chemeurj.org
10856
ꢀ 2016 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim

Full Paper
to 9:1) to give D1-C6N3[Ac] (57 mg, 24%) as a pale yellow powder.
1H NMR (300 MHz, CDCl3): d=0.60–0.71 (m, 10H), 1.04–1.13 (m,
4H), 1.40–1.62 (m, 4H), 1.83–2.01 (m, 6H), 2.16 (s, 3H), 2.25–2.27
3JHH =8.8 Hz, 2H), 6.98 (s, 1H), 7.13–7.15 (m, 4H), 7.40–7.49 (m,
3
8H), 7.61–7.63 (m, 3H),7.69 (d, JHH =8.8 Hz, 2H), 7.78 ppm (s, 1H);
13C NMR (150 MHz, CDCl3): d=12.4, 14.0, 14.1, 20.5, 21.0, 23.2, 26.0,
3
3
(m, 2H), 3.33 (t, JHP =10.7 Hz, 3H), 3.84 (ls, 6H), 3.94 (s, 3H), 4.03–
29.0, 29.5, 33.1 (d, JCP =12.8 Hz), 40.4, 45.8, 49.7, 50.8, 55.2, 55.5,
3
4.18 (m, 6H), 4.39 (t, JHH =7.1 Hz, 2H), 5.49 (s, 2H), 6.82–6.98 (m,
56.4, 63.4, 64.4, 65.8, 66.1, 88.3, 88.6, 90.9, 91.3, 108.8, 109.8, 110.0,
110.9, 111.3, 111.5, 114.3, 115.1, 119.8, 122.6 (m), 122.9, 125.6, 125.7,
127.1, 127.9, 128.6, 130.5, 133.0, 133.1, 139.4, 139.5, 140.1, 140.3,
144.4 (d, 3JCP =7.0 Hz, CH=N), 147.5, 147.7, 148.1, 151.0, 153.9,
156.1, 156.2, 160.8, 170.4 ppm; 31P NMR (121.4 MHz, CDCl3): d=
64.7 ppm (s, P0); MALDI-TOF C83H94N5O11PS calcd for [M]+: 1399.6,
found [M+K]+: 1438.4.
9H), 7.15–7.18 (m, 2H), 7.28–7.31 (m, 2H), 7.48–7.52 (m, 8H), 7.62–
7.69 (m, 6H), 7.76 ppm (d, 3JHH =8.4 Hz, 2H); 13C NMR (75 MHz,
CDCl3): d=13.9, 21.0, 23.2, 25.5, 26.0, 26.3, 28.7, 29.4, 30.3, 33.2 (d,
3JCP =13.1 Hz), 40.4, 50.4, 55.2, 55.4, 56.5, 63.4, 64.6, 64.8, 69.3, 89.3,
89.8, 109.5, 110.9, 114.2, 114.3, 114.7, 115.2, 115.6, 119.4, 120.0,
122.1, 122.2, 122.3, 122.6, 125.9, 126.9, 127.9, 128.1, 128.6, 130.7,
133.2, 139.5, 139.8, 140.1, 140.6, 144.4 (d, 3JCP =7.4 Hz, CH=N),
147.2, 147.8, 150.7, 151.2, 153.9, 156.3, 159.0, 159.8, 160.9,
170.5 ppm; 31P NMR (121.4 MHz, CDCl3): d=63.8 ppm (s, P0);
MALDI-TOF C80H83N6O12PS calcd for [M]+: 1383.6, found: 1383.7.
Synthesis of D4’-(C1N3)[Ac]: Compound 15 (22.5 mg, 76 mmol) and
Cs2CO3 (99 mg, 0.30 mmol) under inert atmosphere were suspend-
ed in 1 mL of anhydrous THF. Compound 14 (30 mg, 76 mmol), dis-
solved in 4 mL of THF, was added. The mixture was stirred at room
temperature for 6 h. Thereafter, the suspension was filtered and
the solvent was evaporated. The crude, M4 (54 mg, 68 mmol) and
Cs2CO3 (99 mg, 0.30 mmol) were re-dissolved in 2 mL of anhydrous
THF, and the mixture was stirred for 18 h. After completion of the
reaction, the suspension was filtered and the solvent was removed
under reduced pressure leading to a yellowish crude that was puri-
fied on silica gel (gradient eluent CH2Cl2/AcOEt, 1:0 to 9:1) to give
Synthesis of dyad D3’-(C1N3)[Ac]: Compound 15 (30 mg, 0.1 mmol)
and Cs2CO3 (131 mg, 0.40 mmol) under inert atmosphere were sus-
pended in 5 mL of anhydrous THF. Then 14 (36 mg, 91 mmol), dis-
solved in 2 mL of THF, was added. The reaction mixture was stirred
at room temperature for 18 h. M3 (63 mg, 91 mmol), dissolved in
2 mL of anhydrous THF, was added and the reaction mixture was
heated to 408C and stirred for another 18 h. Then, the solvent was
removed under reduced pressure. The crude was dissolved (partial-
ly) in CH2Cl2. The suspension was centrifuge-sedimented and the
supernatant was evaporated. The crude was purified on silica gel
(gradient eluent CH2Cl2/EtOAc, 1:0 to 9:1) to give D3’-(C1N3)[Ac]
D4’-(C1N3)[Ac] (25 mg, 23%) as
a
brownish powder. 1H NMR
3
(600 MHz, CDCl3): d=0.53–0.63 (m, 4H), 0.67 (t, JHH =7.3 Hz, 6H),
0.97 (t, 3JHH =7.4 Hz, 6H), 1.07–1.11 (m, 4H), 1.21 (t, 3JHH =7.0 Hz,
3H),1.34–1.39 (m, 4H), 1.54–1.63 (m, 4H), 1.94–1.99 (m, 4H), 2.23
3
1
(s, 3H), 3.09 (t, JHH =7.9 Hz, 2H), 3.26–3.35 (m, 7H), 3.45–3.52 (m,
(25 mg, 38%) as a pale yellow powder. H NMR (600 MHz, CDCl3):
2H), 3.73 (t,3JHH =5.7 Hz, 2H),3.84 (s, 3H), 4.08 (t, JHH =5.9 Hz, 2H),
3
3
d=0.57–0.61 (m, 4H), 0.68 (t, JHH =7.4 Hz, 6H), 1.06–1.11 (m, 4H),
4.21 (t, 3JHH =8.1 Hz, 2H), 5.36 (s, 2H),6,59 (d, 3JHH =8.8 Hz, 2H),
1.21 (t, 3JHH =7.1 Hz, 3H), 1.96–1.99 (m, 4H), 2.24 (s, 3H), 3.09 (t,
3
3
3
3
3JHH =8.1 Hz, 2H), 3.33 (t, JHP =10.7 Hz, 3H), 3.49 (q, JHH =7.0 Hz,
6,67 (d, JHH =8.7 Hz, 2H), 6.81–6.85 (m, 3H), 6.94 (d, JHH =8.6 Hz,
2H), 7.15 (d, 3JHP =8.1 Hz, 2H), 7.39–7.48 (m, 10H), 7.61–7.68 (m,
7H), 7.74 (d, 3JHP =9.0 Hz, 1H), 7.96 ppm (s, 1H); 13C NMR (150 MHz,
CDCl3): d=12.4, 14.0, 14.2, 20.5, 23.2, 23.5, 26.0, 26.5, 29.5, 29.8,
2H), 3.84–3.85 (m, 6H), 4.09 (t, 3JHH =8.1 Hz, 2H), 4.22 (t, JHH
8.1 Hz, 2H), 5.37 (s, 2H), 6.68 (d, 3JHH =8.8 Hz, 2H), 6.81–6.85 (m,
=
3
3
3
3H), 6.90 (d, JHH =8.8 Hz, 2H), 6.94 (d, JHH =8.8 Hz, 2H), 7.14–7.15
(m, 2H), 7.41–7.44 (m, 4H), 7.46–7.52 (m, 6H), 7.63–7.69 (m, 7H),
7.74 (d, 3JHH =9.0 Hz, 1H), 7.96 ppm (s, 1H); 13C NMR (150 MHz,
CDCl3): d=12.4, 14.0, 14.2, 22.5, 23.2, 23.5, 26.0, 26.5, 29.8, 33.2 (d,
3JCP =13.1 Hz), 40.4, 45.8, 49.7, 50.5, 55.2, 55.5, 55.6, 62.4, 65.9, 88.6,
89.3, 89.8, 91.0, 107.6, 110.1, 111.6, 114.2, 114.4, 115.2, 115.6, 119.9,
120.0, 121.4, 122.0, 122.2, 122.6, 122.6, 122.9, 123.2, 123.3, 123.4,
125.6, 125.7, 125.9, 127.6, 128.6, 130.6, 130.7, 133.1, 133.2, 134.0,
3
33.2 (d, JCP =13.1 Hz), 40.5, 45.8, 49.7, 50.5, 50.8, 55.2, 55.5, 62.4,
65.9, 88.3, 88.6, 90.9, 91.4, 107.6, 108.9, 110.1, 111.4, 111.6, 114.4,
115.2, 119.8, 121.4, 122.0, 122.6, 123.0, 123.2 (2 s), 123.4, 125.6,
125.7 (2 s), 127.6, 128.6, 130.5, 133.0, 134.0, 135.8, 137.0, 140.1 (2 s),
140.2, 140.4, 143.8, 144.3 (d, 3JCP =6.9 Hz), 147.5, 148.1, 151.0,
151.2, 156.3, 157.4, 161.1, 168.8 ppm; 31P NMR (121.4 MHz, CDCl3):
d=63.7 ppm (s, P0); MALDI-TOF C83H90N9O8PS calcd for [M]+:
1403.6, found [M]+: 1403.8.
3
135.8, 137.0, 140.1, 140.2, 140.7, 143.8, 144.3 (d, JCP =7.1 Hz, CH=
N), 147.5, 151.1, 151.3, 151.3, 156.3, 157.4, 159.8, 161.1, 168.8 ppm;
31P NMR (242.8 MHz, CDCl3): d=63.7 ppm (s, P0); MALDI-TOF
C76H75N8O9PS calcd for [M]+: 1306.6, found [M+Na]+: 1330.5.
Photophysical studies and photolysis experiments
All photophysical studies were performed with freshly prepared
air-equilibrated solutions at room temperature (298 K). UV/Vis ab-
sorption spectra of 10À5 m solutions were recorded on a Jasco V-
670 spectrophotometer. Steady-state fluorescence measurements
were performed on dilute solutions (ca. 10À6 m, optical density
<0.1) contained in standard 1 cm quartz cuvettes by using a Fluo-
rolog spectro fluorometer. Emission spectra were obtained, for
each compound, under excitation at the wavelength of the absorp-
tion maximum. Fluorescence quantum yields were measured ac-
cording to literature procedures.
Synthesis of D4-(C3)[Ac]: Compound 15 (20.0 mg, 67 mmol) and
Cs2CO3 (88 mg, 0.27 mmol) under inert atmosphere were suspend-
ed in 1 mL of anhydrous THF. Compound 8a (30 mg, 76 mmol), dis-
solved in 3 mL of THF, was added. The mixture was stirred at room
temperature for 20 h. Thereafter, the suspension was filtered and
the solvent evaporated. The crude, M4 (48 mg, 61 mmol) and
Cs2CO3 (88 mg, 0.27 mmol) were re-dissolved in 2 mL of anhydrous
THF, and the mixture was stirred for 18 h. After completion of the
reaction, the suspension was filtered and the solvent was removed
under reduced pressure leading to a yellowish crude that was puri-
fied on silica gel (eluent-DCM 100%) to give D4-(C3)[Ac] (50.2 mg,
2PA cross-sections (s2) were derived from the two-photon excited
fluorescence (TPEF) cross-sections (s2Ff) and the fluorescence
emission quantum yield (Ff). TPEF cross-sections were measured
relative to fluorescein in 0.01m aqueous NaOH,[53] using the well-
established method described by Xu and Webb[54] and the appro-
priate solvent-related refractive index corrections.[55] The quadratic
dependence of the fluorescence intensity on the excitation power
was checked for each sample and all wavelengths. Measurements
were conducted using an excitation source delivering fs pulses. A
1
59%) as a yellow powder. H NMR (600 MHz, CDCl3): d=0.59–0.63
(m, 4H), 0.68 (t,3JHH =7.4 Hz, 6H), 0.97 (t,3JHH =7.4 Hz, 6H), 1.07–
1.11 (m, 4H), 1.21 (t,3JHH =7.1 Hz, 3H), 1.34–1.40 (m, 4H), 1.56–1.61
(m, 4H), 1.96–1.99 (m, 4H), 2.16 (s, 3H), 2.29–2.33 (m, 2H), 3.28–
3.31 (m, 7H), 3.49 (q,3JHH =7.1 Hz, 2H), 3.73 (t,3JHH =6.0 Hz, 2H),
3.84 (s, 3H), 3.93 (s, 3H), 4.07 (t,3JHH =6.1 Hz, 2H),4.12 (t,3JHH
5.9 Hz, 2H), 4.26 (t,3JHH =6.2 Hz, 2H), 5.49 (s, 1H), 6.60 (d,3JHH
=
=
9.1 Hz, 2H),6.67 (d, 3JHH =9.1 Hz, 2H),6.79–6.84 (m, 4H), 6.94 (d,
Chem. Eur. J. 2016, 22, 10848 – 10859
www.chemeurj.org
10857
ꢀ 2016 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim

Full Paper
[12] A. Banerjee, C. Grewer, L. Ramakrishnan, J. Jäger, A. Gameiro, H.-G. A.
Breitinger, K. R. Gee, B. K. Carpenter, G. P. Hess, J. Org. Chem. 2003, 68,
8361–8367.
[13] G. C. R. Ellis-Davies, Nat. Methods 2007, 4, 619–628.
[14] A. C. Rea, L. N. Vandenberg, R. E. Ball, A. A. Snouffer, A. G. Hudson, Y.
Zhu, D. E. McLain, L. L. Johnston, J. D. Lauderdale, M. Levin, Chem. Biol.
2013, 20, 1536–1546.
[15] L. Fournier, C. Gauron, L. Xu, I. Aujard, T. Le Saux, N. Gagey-Eilstein, S.
Maurin, S. Dubruille, J.-B. Baudin, D. Bensimon, ACS Chem. Biol. 2013, 8,
1528–1536.
Chameleon Ultra II (Coherent) was used generating 140 fs pulses
at an 80 MHz repetition rate. The excitation was focused into the
cuvette through a microscope objective (10”, NA 0.25). The fluo-
rescence was detected in epifluorescence mode with a dichroic
mirror (Chroma 675dcxru) and a barrier filter (Chroma e650sp-2p)
by a compact CCD spectrometer module BWTek BTC112E. Total
fluorescence intensities were obtained by integrating the corrected
emission.
The one-photon uncaging sensitivity of dyads was determined by
irradiating solutions of dyads and NVOAc or MNI (used as stand-
ards for D1-D4 and D2’-D4’ dyads, respectively) having the same ab-
sorbance at 365 nm. The solutions (0.6 mL) prepared in CDCl3 were
introduced in a quartz NMR tube which was irradiated with three
UV-lamps (VilberLourmat, Model VL-6.LC) arranged in a triangular
configuration. The conversion rate (X) was monitored over time by
1H NMR spectroscopy by following the decrease of the singlet at
2.16 or 2.23 ppm (corresponding to caged acetyl CH3-COOR), in
the case of NV and MNI-based dyads respectively, and onset of the
singlet at 2.10 ppm (corresponding to released acetic acid). A first
order kinetics was observed for all derivatives. Comparison of the
conversion kinetics for dyads and reference PPG (i.e., NVOAc or
MNI) allowed to calculate the e365Qu(dyad)/e365Qu (NVOAc) ratio
values for D1-D4 dyads and e365Qu(dyad)/e365Qu (MNI) ratio values
for D2’-D4’ dyads. Using e365 values derived from the absorption
spectra, the ratio values (Qu(dyad)/Qu (NV) for D1-D4 dyads and
Qu(dyad)/Qu (MNI) or D2’-D4’ dyads) could be calculated.
[16] D. E. McLain, A. C. Rea, M. B. Widegren, T. M. Dore, Photochem. Photo-
biol. Sci. 2015, 14, 2151–2158.
[17] R. Horbert, B. Pinchuk, P. Davies, D. Alessi, C. Peifer, ACS Chem. Biol.
2015, 10, 2099–2107.
[18] W. Denk, J. Strickler, W. Webb, Science 1990, 248, 73–76.
[19] W. R. Zipfel, R. M. Williams, W. W. Webb, Nat. Biotechnol. 2003, 21,
1369–1377.
[20] G. Bort, T. Gallavardin, D. Ogden, P. I. Dalko, Angew. Chem. Int. Ed. 2013,
52, 4526–4537; Angew. Chem. 2013, 125, 4622–4634.
[21] N. Gagey, P. Neveu, L. Jullien, Angew. Chem. Int. Ed. 2007, 46, 2467–
2469; Angew. Chem. 2007, 119, 2519–2521.
[22] L. Donato, A. Mourot, C. M. Davenport, C. Herbivo, D. Warther, J. LØo-
nard, F. Bolze, J.-F. Nicoud, R. H. Kramer, M. Goeldner, Angew. Chem. Int.
Ed. 2012, 51, 1840–1843; Angew. Chem. 2012, 124, 1876–1879.
[23] G. C. R. Ellis-Davies, J. M. Amatrudo, J. P. Olson, G. Lur, C. Q. Chiu, M. J.
Higley, ACS Chem. Neurosci. 2014, 5, 64–70.
[24] C. Tran, T. Gallavardin, M. Petit, R. Slimi, H. Dhimane, M. Blanchard-
Desce, F. C. Acher, D. Ogden, P. I. Dalko, Org. Lett. 2015, 17, 402–405.
[25] N. Komori, S. Jakkampudi, R. Motoishi, M. Abe, K. Kamada, K. Furukawa,
C. Katan, W. Sawada, N. Takahashi, H. Kasai, Chem. Commun. 2015, 51,
331–334.
[26] S. Gug, F. Bolze, A. Specht, C. Bourgogne, M. Goeldner, J.-F. Nicoud,
Angew. Chem. Int. Ed. 2008, 47, 9525–9529.
Acknowledgements
[27] P. Dunkel, C. Tran, T. Gallavardin, H. Dhimane, D. Ogden, P. I. Dalko, Org.
Biomol. Chem. 2014, 12, 9899–9908.
[28] H. K. Agarwal, R. Janicek, S.-H. Chi, J. W. Perry, E. Niggli, G. C. R. Ellis-
Davies, J. Am. Chem. Soc. 2016, 138, 3687–3693.
[29] S. Picard, E. Genin, G. Clermont, V. Hugues, O. Mongin, M. Blanchard-
Desce, New J. Chem. 2013, 37, 3899–3913.
[30] I. Aujard, C. Benbrahim, M. Gouget, O. Ruel, J.-B. Baudin, P. Neveu, L. Jul-
lien, Chem. Eur. J. 2006, 12, 6865–6879.
[31] S. Gug, S. Charon, A. Specht, K. Alarcon, D. Ogden, B. Zietz, J. LØonard,
S. Haacke, F. Bolze, J.-F. Nicoud, ChemBioChem 2008, 9, 1303–1307.
[32] S. Picard, E. J. Cueto-Diaz, E. Genin, G. Clermont, F. Acher, D. Ogden, M.
Blanchard-Desce, Chem. Commun. 2013, 49, 10805–10807.
[33] K. A. Korzycka, P. M. Bennett, E. J. Cueto-Diaz, G. Wicks, M. Drobizhev, M.
Blanchard-Desce, A. Rebane, H. L. Anderson, Chem. Sci. 2015, 6, 2419–
2426.
[34] E. J. Cueto Díaz, S. Picard, V. Chevasson, J. Daniel, V. Hugues, O. Mongin,
E. Genin, M. Blanchard-Desce, Org. Lett. 2015, 17, 102–105.
[35] M. Canepari, G. Papageorgiou, J. E. T. Corrie, C. Watkins, D. Ogden, J.
Physiol. 2001, 533, 765–772.
This work has been supported by funding from the European
Community’s Seventh Framework Program under TOPBIO proj-
ect-grant agreement no. 264362. M.B.D. thanks the Conseil RØ-
gional d’Aquitaine for funding (Chaire d’Accueil grant and fel-
lowship to V.H.). We also acknowledge financial support from
Agence Nationale pour la Recherche (Grant 2010 ANR-10-
BLAN-1436) and University of Bordeaux for a fellowship to M.K.
This work also received funding from the People Programme
(Marie Curie Actions) of the European Union’s Seventh Frame-
work Programme FP7/2007-2013/ under REA grant agreement
number 607721 (fellowship to P.P.). The authors are grateful to
Jean-Pierre Majoral for generous gift of MMH-PSCl2.
Keywords: caged compounds · Fçrster resonance energy
transfer · photolysis · two-photon absorption · two-photon
uncaging
[36] M. Canepari, L. Nelson, G. Papageorgiou, J. E. Corrie, D. Ogden, J. Neuro-
sci. Methods 2001, 112, 29–42.
[37] G. Lowe, J. Neurophysiol. 2003, 90, 1737–1746.
[38] M. Matsuzaki, G. C. R. Ellis-Davies, T. Nemoto, Y. Miyashita, M. Iino, H.
Kasai, Nat. Neurosci. 2001, 4, 1086–1092.
[39] T. Furuta, S. S.-H. Wang, J. L. Dantzker, T. M. Dore, W. J. Bybee, E. M. Call-
away, W. Denk, R. Y. Tsien, Proc. Natl. Acad. Sci. USA 1999, 96, 1193–
1200.
[1] J. A. Barltrop, P. Schofield, Tetrahedron Lett. 1962, 3, 697–699.
[2] J. Engels, E. J. Schlaeger, J. Med. Chem. 1977, 20, 907–911.
[3] J. H. Kaplan, B. Forbush, J. F. Hoffman, Biochemistry (Mosc.) 1978, 17,
1929–1935.
[4] V. N. Rajasekharan Pillai, Synthesis 1980, 1980, 1–26.
[5] Z. Li, W. E. Lee, W. C. Galley, Biophys. J. 1989, 56, 361–367.
[6] C. G. Bochet, J. Chem. Soc. Perkin Trans. 1 2002, 125–142.
[40] J. Corrie, G. Papageorgiou, 7-Nitroindoline Derivatives and Their Uses
2004, WO2004085394 (A1).
ˇ
[7] P. Klµn, T. Solomek, C. G. Bochet, A. Blanc, R. Givens, M. Rubina, V. Popik,
[41] Q. Cai, W. Zhu, H. Zhang, Y. Zhang, D. Ma, Synthesis 2005, 2005, 496–
499.
[42] O. Mongin, T. R. Krishna, M. H. V. Werts, A.-M. Caminade, J.-P. Majoral, M.
Blanchard-Desce, Chem. Commun. 2006, 915–917.
[43] C. Rouxel, M. Charlot, O. Mongin, T. R. Krishna, A.-M. Caminade, J.-P. Ma-
joral, M. Blanchard-Desce, Chem. Eur. J. 2012, 18, 16450–16462.
[44] C. Katan, M. Charlot, O. Mongin, C. Le Droumaguet, V. Jouikov, F. Teren-
ziani, E. Badaeva, S. Tretiak, M. Blanchard-Desce, J. Phys. Chem. B 2010,
114, 3152–3169.
A. Kostikov, J. Wirz, Chem. Rev. 2013, 113, 119–191.
[8] C. Brieke, F. Rohrbach, A. Gottschalk, G. Mayer, A. Heckel, Angew. Chem.
Int. Ed. 2012, 51, 8446–8476; Angew. Chem. 2012, 124, 8572–8604.
[9] H. Yu, J. Li, D. Wu, Z. Qiu, Y. Zhang, Chem. Soc. Rev. 2010, 39, 464–473.
[10] W. Lu, C. Tian, P. Thogaripally, J. Hu, P. Wang, Chem. Commun. 2013, 49,
9636–9638.
[11] K. A. Mosiewicz, L. Kolb, A. J. van der Vlies, M. P. Lutolf, Biomater. Sci.
2014, 2, 1640–1651.
Chem. Eur. J. 2016, 22, 10848 – 10859
www.chemeurj.org
10858
ꢀ 2016 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim

Full Paper
[45] D. L. Andrews, J. S. Ford, J. Chem. Phys. 2013, 139, 014107.
[46] F. Terenziani, M. Morone, S. Gmouh, M. Blanchard-Desce, ChemPhys-
Chem 2006, 7, 685–696.
[47] G. Papageorgiou, M. Lukeman, P. Wan, J. E. T. Corrie, Photochem. Photo-
biol. Sci. 2004, 3, 366–373.
[48] K. Lee, D. E. Falvey, J. Am. Chem. Soc. 2000, 122, 9361–9366.
[49] D. E. Falvey, C. Sundararajan, Photochem. Photobiol. Sci. 2004, 3, 831–
838.
[52] G. C. R. Ellis-Davies, M. Matsuzaki, M. Paukert, H. Kasai, D. E. Bergles, J.
Neurosci. 2007, 27, 6601–6604.
[53] M. A. Albota, C. Xu, W. W. Webb, Appl. Opt. 1998, 37, 7352.
[54] C. Xu, W. W. Webb, J. Opt. Soc. Am. B 1996, 13, 481.
[55] M. H. V. Werts, N. Nerambourg, D. PØlØgry, Y. L. Grand, M. Blanchard-
Desce, Photochem. Photobiol. Sci. 2005, 4, 531–538.
[50] N. I. Kiskin, R. Chillingworth, J. A. McCray, D. Piston, D. Ogden, Eur. Bio-
phys. J. 2002, 30, 588–604.
Received: March 8, 2016
Revised: May 19, 2016
[51] G. Papageorgiou, D. Ogden, J. E. T. Corrie, Photochem. Photobiol. Sci.
2008, 7, 423–432.
Published online on June 27, 2016
Chem. Eur. J. 2016, 22, 10848 – 10859
www.chemeurj.org
10859
ꢀ 2016 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim