Deactivation behavior and excited-state properties of (coumarin-4- yl)methyl derivatives. 1. Photocleavage of (7-methoxycoumarin-4-yl)methyl- caged acids with fluorescence enhancement

Article abstract of DOI:10.1021/jo9910233

The photochemistry of the (7-methoxycoumarin-4-yl)methyl (MCM) carboxylates 3a-d, the mesylate 4, and the phosphates 5a-e has been examined under near physiological conditions in acetonitrile or methanol/aqueous HEPES buffer solution, respectively. Analysis of photoproducts as well as measurements of photochemical quantum yields, fluorescence quantum yields, and lifetimes for the excited singlet state verified the similar photochemical and photophysical behavior of all the esters studied here. 4- (Hydroxymethyl)-7-methoxycoumarin (2) and the corresponding free acids were obtained as major products upon irradiation. The rates of deactivation of the excited MCM derivatives 3a-5e were found to be dependent on the leaving group ability of the anion concerned as well as on the solvent polarity. The polarity dependence and the exclusive formation of ¹⁸O-labeled 2 during irradiation of 5a in ¹⁸O-labeled water indicate that photocleavage of the excited singlet state of the MCM caged compounds 3a-5e proceeds via a photo S(N)1 mechanism (solvent-assisted photoheterolysis).

Full text of DOI:10.1021/jo9910233

J . Org. Chem. 1999, 64, 9109-9117
9109
Dea ctiva tion Beh a vior a n d Excited -Sta te P r op er ties of
(Cou m a r in -4-yl)m eth yl Der iva tives. 1. P h otoclea va ge of
(7-Meth oxycou m a r in -4-yl)m eth yl-Ca ged Acid s w ith F lu or escen ce
En h a n cem en t
Bjo¨rn Schade,† Volker Hagen,† Reinhard Schmidt,‡ Ralph Herbrich,‡ Eberhard Krause,†
Torsten Eckardt,† and J u¨rgen Bendig*,§
Research Institute of Molecular Pharmacology, Alfred-Kowalke-Strasse 4, D-10315 Berlin, Germany,
Institute of Physical and Theoretical Chemistry, J ohann-Wolfgang-Goethe University Frankfurt/
Main, D-60439 Frankfurt/ Main, Germany, and Institute of Chemistry, Humboldt University Berlin,
Hessische Strasse 1-2, D-10115 Berlin, Germany
Received J une 28, 1999
The photochemistry of the (7-methoxycoumarin-4-yl)methyl (MCM) carboxylates 3a -d , the mesylate
4, and the phosphates 5a -e has been examined under near physiological conditions in acetonitrile
or methanol/aqueous HEPES buffer solution, respectively. Analysis of photoproducts as well as
measurements of photochemical quantum yields, fluorescence quantum yields, and lifetimes for
the excited singlet state verified the similar photochemical and photophysical behavior of all the
esters studied here. 4-(Hydroxymethyl)-7-methoxycoumarin (2) and the corresponding free acids
were obtained as major products upon irradiation. The rates of deactivation of the excited MCM
derivatives 3a -5e were found to be dependent on the leaving group ability of the anion concerned
as well as on the solvent polarity. The polarity dependence and the exclusive formation of 18O-
labeled 2 during irradiation of 5a in 18O-labeled water indicate that photocleavage of the excited
singlet state of the MCM caged compounds 3a -5e proceeds via a photo SN1 mechanism (solvent-
assisted photoheterolysis).
In tr od u ction
Among the numerous caging groups3 described until
now the (7-methoxycoumarin-4-yl)methyl (MCM) caging
group fulfills the requirement to a high degree. It was
shown that the MCM moiety is favorable for the caging
of phosphates, i.e., of adenosine cyclic-3′,5′-monophos-
phate (cAMP)4 and guanosine cyclic-3′,5′-monophosphate
(cGMP), as well as their derivatives. The main advan-
tages of these caged compounds are their high photoef-
ficiencies and hydrolytic stabilities. Several photochemi-
cal studies of such MCM esters of cGMP and cAMP4
indicate that the efficiency of photocleavage, and there-
fore the liberation of the second messenger molecule,
depends on the structure of the caged cyclic nucleotide
as well as on the environment (especially the solvent
polarity).
Recent investigations have focused on the photochem-
istry of 1-naphthylalkyl5 and benzyl6 derivatives. In these
cases a heterolytic bond cleavage or a competition
between homolytic versus heterolytic cleavage were
respectively proposed.4,7 However, there has previously
been no mechanistic study describing the photolytic
pathway of MCM derivatives. The question of whether a
The liberation of bioactive effector molecules from their
photosensitive precursors (caged derivatives) is an at-
tractive method which is used to an increasing extent
for studies of the functionality and action of biomolecules
inside living cells.1,2 Since the photolytic cleavage of such
precursors can be both temporally (short generating
pulses) and spatially controlled, it is possible to trigger
the biological response with almost no delay. The re-
sponse can be analyzed by procedures such as im-
munological, electrophysiological, or imaging techniques.
Especially for the latter method, the use of caging
(protecting) groups leading to highly fluorescent products
during the photolytic liberation of the biomolecule is
particularly favorable. The release of such fluorescent
photoproducts should offer the possibility of monitoring
the phototriggering reaction and additionally, in combi-
nation with fluorescence microscopic procedures, to es-
timate the rate of photolysis and/or the concentration and
local distribution of the liberated biomolecule.
(3) For reviews see: (a) Pillai, V. N. R. Org. Photochem. 1987, 9,
225-323. (b) Pillai, V. N. R. Synthesis 1980, 1-26.
(4) (a) Furuta, T.; Torigai, H.; Sugimoto, M.; Iwamura, M. J . Org.
Chem. 1995, 60, 3953-3956. (b) Bendig, J .; Helm, S.; Hagen, V. J .
Fluoresc. 1997, 7, 357-361. (c) Furuta, T.; Iwamura, M. Caged
Compounds. In Methods of Enzymology; Marriott, G., Ed.; Academic
Press: San Diego, CA, 1998; Vol. 291, pp 50-63.
(5) (a) Itoh, Y.; Gouki, M.; Goshima, T.; Hachimori, A.; Kojima, M.;
Karatsu, T. J . Photochem. Photobiol. A 1998, 117, 91. (b) Pincock, J .
A. Acc. Chem. Res. 1997, 30, 43. (c) Arnold, B.; Donald, L.; J urgens,
A.; Pincock, J . A. Can. J . Chem. 1985, 63, 3140.
(6) Hilborn, J . W.; MacKnight, E.; Pincock, J . A.; Wedge, P. J . J .
Am. Chem. Soc. 1994, 116, 3337.
* Corresponding author. E-mail: juergen)bendig@chemie.hu-ber-
lin.de.
† Research Institute of Molecular Pharmacology.
‡ J ohann-Wolfgang-Goethe University Frankfurt/Main.
§ Humboldt University Berlin.
(1) (a) Corrie, J . E. T.; Trentham, D. R. In Bioorganic Photochem-
istry; Morrison, H., Ed.; Wiley: New York, 1993; Vol. 2, pp 243-305.
(b) Adams, S. R.; Tsien, R. Y. Annu. Rev. Physiol. 1993, 55, 755-784.
(c) McCray, J . A.; Trentham, D. R.; Annu. Rev. Biophys. Biophys. Chem.
1989, 18, 239-270.
(2) (a) Hagen, V.; Dzeja, C.; Bendig, J .; Baeger, I.; Kaupp, U. B. J .
Photochem. Photobiol. B: Biol. 1997, 42, 71-78. (b) Hagen, V.; Dzeja,
C.; Frings, S.; Bendig, J .; Krause, E.; Kaupp, U. B. Biochemistry 1996,
35, 7762-7771.
(7) Givens, R. S.; Matuszewski, B. J . Am. Chem. Soc. 1984, 106,
6860-6861.
10.1021/jo9910233 CCC: $18.00 © 1999 American Chemical Society
Published on Web 11/12/1999

9110 J . Org. Chem., Vol. 64, No. 25, 1999
Schade et al.
Sch em e 1. P a th w a ys of th e P h otolytic Clea va ge of MCM Der iva tives: (a ) Solven t-Assisted
P h otoh eter olytic Clea va ge w ith Ion P a ir F or m a tion ; (b) P h otosolvolysis Rea ction
Sch em e 2. In vestiga ted MCM Ca ged Com p ou n d s
photo SN1 reaction (solvent-assisted photoheterolysis,
Scheme 1a) or of a photosolvolysis (Scheme 1b), charac-
terized by a nucleophilic attack of the solvent at the Cd
O-, PdO-, or SdO center in the primary step, is
involved remained unclear. Additionally, the deactivation
behavior of the excited MCM caged compounds is char-
acterized by the competition of photophysical and pho-
tochemical processes.4 The understanding of the solvent
influence on the efficiency of the photoreaction is ren-
dered more difficult by the fact that the photophysical
deactivation behavior of MCM derivatives, in terms of
fluorescence quantum yields and lifetimes, already shows
a distinctive solvent dependence in general.8 The more
complex solvent influence observed experimentally is
presumed, therefore, to be a superimposition of two
independent relationships.
We are currently investigating the suitability of the
MCM caging group for the deactivation of biologically
active substrates with different chemical functional
groups. In the present paper we report the photochemical
and photophysical deactivation behavior of MCM-pro-
tected model compounds and of the MCM-caged biomol-
ecules cAMP and cGMP (Scheme 2). To investigate the
structural influence on photochemistry, on a possible
fluorescence enhancement, and on the rate of cleavage
during photolysis a series of photolabile (7-methoxycou-
marin-4-yl)methyl (MCM) derivatives was prepared and
examined (Scheme 2, compounds 3a -5e). Starting with
the basic chromophore 1 and the photoproduct 2, a
systematic variation of the ester moiety was effected from
the carboxylates 3a -d via the phosphates 5a -e to the
mesylate 4. A reaction pathway is proposed concerning
the photolytic cleavage of MCM derivatives.
Resu lts a n d Discu ssion
Syn th esis of MCM Der iva tives. 4-(Hydroxymethyl)-
7-methoxycoumarin (2), a reference substance for ana-
lytical HPLC and fluorescence measurements, was pre-
pared according to Sehgal and Seshadri9 via hydrolysis
of MCM acetate formed from 6 by refluxing in acetan-
hydride/sodium acetate (Scheme 3a). The MCM carboxyl-
ic acid esters 3a -d were synthesized using a procedure
described by Wilcox et al.10 starting from 6 and the free
carboxylic acids activated by potassium fluoride. This
very mild and neutral method was chosen in order to
retain the possibility of converting more sensitive car-
(9) Sehgal, J . M.; Seshadri, T. R. J . Sci. Ind. Res. 1953, 12B, 346.
(10) Wilcox, M.; Viola, R. W.; J ohnson, K. W.; Billington, A. P.;
Carpenter, B. K.; McCray, J . A.; Guzikowski, A. P.; Hess, G. P. J . Org.
Chem. 1990, 55, 1585-1589.
(8) Seixas de Melo, J . S.; Becker, R. S.; Macanita, A. L. J . Phys.
Chem. 1994, 98, 6054-6058.

(Coumarin-4-yl)methyl Derivatives
J . Org. Chem., Vol. 64, No. 25, 1999 9111
Sch em e 3. Syn th esis of MCM Der iva tivesa
a
Key (a) i, acetanhydride/NaOAc, reflux, 2 h; ii, HCl/EtOH, reflux, 1 h; iii, diethyl phosphoric acid chloride/pyridine, 0 °C-rt, 4 h; iv,
tetra-n-butylammonium salt of CAMP/CH3CN, reflux, 5 h; v, acid/acetone/KF, rt (reflux), 1 d (3 h); (b) vi, CGMP dihydrate/CH3CN/
DMSO, 50 °C, 1 d; vii, methanesulfonic acid/CHCl3, rt, 2 h.
boxylic acids into their corresponding photolabile MCM
esters. The synthesis procedure led to high or almost
quantitative yields, respectively. Attempts to obtain
MCM mesylate (4) via conversion of 2 with methane-
sulfonic acid chloride or methanesulfonic acid anhydride
led to decomposition of 2. 4 was obtained in moderate
yields starting from the 4-(diazomethyl)-7-methoxycou-
marin (7).11 MCM diethyl phosphoric acid ester (5a )12 was
synthesized by reaction of diethyl phosphoric acid chlo-
ride with 2 in pyridine, adapting a procedure described
by Givens for benzyl phosphates13 (Scheme 3a).
For the synthesis of MCM-caged cAMP (5b,c) we used
the alkylation of the tetra-n-butylammonium salt of
cAMP with 4-(bromomethyl)-7-methoxycoumarin (6). This
method afforded 5b,c in relatively high yields as a
mixture of diastereomers (axial 5b and equatorial 5c).
Our synthesis procedure is superior to the silver(I) oxide
method described by Iwamura et al.14 because of better
yields, a higher diastereomeric excess in favor of 5b, and
the shorter reaction time as well as easier workup.
Obviously the increase of nucleophilic reactivity of the
nucleotide phosphate anion by tetra-n-butylammonium
salt formation is more effective than the Lewis acid
activation by silver(I) oxide. Contrary to the synthesis
of 5b,c, attempts to obtain MCM-caged cGMP 5d ,e by
reaction of 6 with the tetra-n-butylammonium salt of
cGMP or by reaction of 6 with the free acid of cGMP in
the presence of silver(I) oxide failed. Only traces of the
target molecule were obtained. Finally the alkylation of
the free acid of cGMP with 7 was successful. The
diastereomeric mixtures of 5b,c and 5d ,e were separated
into the pure diastereomers by RP-HPLC on a prepara-
tive scale.
The isomeric species were assigned by 31P NMR as
reported for the axial and equatorial isomers of known
cAMP and cGMP esters.15 The 31P NMR signals at higher
field (δ -5.1) correspond to the axial isomers 5b,d and
those at lower field (δ -3.6 for 5c and -4.0 for 5e) to the
equatorial isomers.
Gr ou n d -Sta te a n d UV/Vis Sp ectr oscop ic P r op er -
ties of th e P h otola bile Com p ou n d s. For our investi-
gations acetonitrile/HEPES buffer and methanol/HEPES
buffer solutions were used, which correspond broadly to
physiological conditions. The absorption spectra of the
“basic chromophore” 1 and 3a -5e differ only insignifi-
cantly from each other and are characterized by two
intensive absorption maxima around λ ) 325 nm and λ
) 217 nm (Table 1). The bathochromic shift of the long-
wavelength absorption of the MCM moiety of about 15
nm compared to the unsubstituted (coumarin-4-yl)methyl
system is due to the “push-pull effect” between the
methoxy donor and the carbonyl acceptor. In the case of
the bichromophoric systems 3b-d and 5b-e a third
absorption band, characterized by a maximum in the
region λ ) 210-290 nm, is due to local excitation of the
benzoic acids and purine bases, respectively. A compari-
son with the absorption spectra of the partial chro-
mophores shows that the spectra of 3b-d and 5b-e are
a superimposition of these partial spectra. The long-
wavelength absorption of the MCM chromophore of all
photolabile derivatives with regard to its position as well
(11) Ito, K.; Sawanobori, J . Synth. Commun. 1982, 12 (9), 665-671.
(12) First described by: Givens, R. S.; Matuszewski, B. See ref 7.
(13) Givens, R. S.; Matuszewski, B.; Athey, P. S.; Stoner, M. R. J .
Am. Chem. Soc. 1990, 112, 6016-6021.
(14) Furuta, T.; Torigai, H.; Osawa, T.; Iwamura, M. J . Chem. Soc.,
Perkin Trans 1 1993, 3139-3142.
(15) (a) Engels, J . Bioorg. Chem. 1979, 8, 9-16. (b) Nerbonne, J .
M.; Richard, S.; Nargeot, J .; Lester, H. A. Nature 1984, 310, 74-76.

9112 J . Org. Chem., Vol. 64, No. 25, 1999
Schade et al.
Ta ble 1. UV/Vis Sp ectr oscop ic P r op er ties of th e
In vestiga ted MCM Der iva tives [50 µM Solu tion s in
Aceton itr ile/HEP ES Bu ffer Solu tion (30/70) Excep t for
5b-e: 10 µM in Meth a n ol/HEP ES (20/80)]
bination (k0). As with the photolysis of other phosphoric
acid esters,13 it is not clear whether the ion pair formation
proceeds directly from the S1 state by heterolytic bond
cleavage (k1) or by homolytic cleavage followed by single
electron transfer (kset). However, formation of a singlet
radical pair from S1 is of minor importance, since in that
case one should expect occurrence of hydrogen abstrac-
tion products derived from the escaped radicals, formed
from the triplet state after intersystem crossing. How-
ever, only traces of 1 were found (e0.5% compared to the
formation of 2). Furthermore no phosphorescence which
would be characteristic for the population of any triplet
states could be detected in any of the photolabile MCM
derivatives nor the basic chromophors 1 and 2. Therefore
further kinetic considerations will be made on the basis
of heterolytic bond cleavage from the S1 state. On the
basis of these results, we exclude a significant participa-
tion of the triplet state during photolysis contrary to
Furuta and Iwamura,23 who proposed a triplet mecha-
nism for the photolytic cleavage of (7-methoxycoumarin-
4-yl)methyl diethyl phosphate.
ꢀ/L
ꢀ/L
λ
max/nm
mol-1 cm-1
λ
max/nm
mol-1 cm-1
1
321
218
320
219
324
219
13.600
15.900
13.300
16.000
13.500
15.800
4
325
219
324
219
328
258
207
325
258
208
327
276
253
220
325
276
252
209
13.000
15.700
13.900
15.900
13.200
14.900
34.900
13.300
15.000
34.000
13.300
12.000
15.300
22.500
13.300
11.800
15.200
22.300
2
5a
5b
3a
3b
3c
324
260
215
324
230
222
13.600
19.000
24.500
13.000
19.000
20.900
5c
5d
3d
324
241
13.700
26.600
5e
The SN1 mechanism of the reaction characterized by
solvent-assisted photoheterolysis (marked by the ion-pair
stability) as the rate-determining step is also supported
by the fact that higher values for Φchem were obtained
with an increasing leaving group ability of OX (see Table
2; Φchem (4) > Φchem (5a ) > Φchem (3a -d )). The ability of
good leaving groups to stabilize the corresponding anion
leads to an encouragement of the photoreaction which is
characteristic of the SN1 case. From the cationic view the
stabilization of the coumarinylmethyl cation should
benefit from the increased electron density within the
aromatic system, which is given according to eq 1 by the
methoxy substituent in position 7.
as to its intensity is only insignificantly influenced by
the “leaving groups” OX (see Table 1). There were no
indications in the absorption spectra from which we could
infer conjugative or other interactions (through-space,
hydrogen bonds, charge transfer) within the ground state.
Mech a n istic P a th w a y of th e P h otolytic Clea va ge.
Verification of the reaction pathway including potential
intermediates and transition states is the basis for the
qualitative discussion of the photophysical and -chemical
data obtained. Excitation of solutions of the photolabile
MCM esters 3a -5e in acetonitrile/HEPES buffer solution
(30/70) or methanol/HEPES buffer solution (20/80) within
the long-wavelength absorption maximum with UV ir-
radiation of 333 nm led to photochemical cleavage with
release of the corresponding carboxylic acids (3a -d ),
methansulfonic acid (4), and phosphoric acids (5a -e). 2
was obtained with yields g95% as major product of the
photolysis besides the released acids (Scheme 4a). This
result is in agreement with observations made by Givens
and Matuszewski7 for the photolytic cleavage of MCM
phosphates, by Furuta et al.4 for MCM-cAMP, and by
Hagen et al.16 for MCM-caged 8-Br-cAMP and MCM-
caged 8-Br-cGMP. The photochemical quantum yields
Φchem of these reactions are summarized in Table 2.
To elucidate the reaction pathway of the photolytic
cleavage after electronic excitation, especially in terms
of a mechanistic distinction between a photo SN1 or a
photosolvolysis reaction, a solution of 5a in acetonitrile/
18O labeled water (30/70) was irradiated. ESI mass
spectrometric analysis of the separated reaction products
showed that only photolysis product 2 contained the 18O
isotope; the liberated diethylphosphoric acid was not
labeled (Scheme 4b). From this result we conclude that
the photolytic reaction of the excited-state proceeds
according to Scheme 1a under cleavage of the CH2-OP
bond, while the CH2O-P bond remains intact (Scheme
1b). Consequently an ion pair is postulated as the decisive
intermediate (Scheme 5) which undergoes a nucleophilic
attack of OH- or water, respectively (in the case of
aqueous solvent mixtures), after a successful escape
reaction with the solvent in competition with the recom-
Equation 2 is valid for the photochemical quantum
yield. In general, an effective photochemical cleavage is
observed, if (a) the rate of competing photophysical
deactivation (fluorescence and nonradiative; kfn + knr) is
low and therefore the “physical” lifetime of the excited
state is high, (b) the stability of the coumarinylmethyl
cation intermediate and of the anionic product, respec-
tively, is high, such that the rate of the ion formation
(k1) can compete successfully with the photophysical
deactivation, and (c) the escape process (kesc) is favored
(17) Dippy, J . F. J . J . Chem. Soc. 1938, 9, 1222-1227.
(18) Vandenbelt, J . M.; Henrich, C.; Vanden Berg, S. G. Anal. Chem.
1954, 26, 726.
(19) Willi, A. V.; Stocker, J . F. Helv. Chim. Acta 1955, 38, 1279.
(20) Wideqvist, B. Ark. Kemi 1951, 2, 383-386.
(21) Marziano, N. C.; Sampoli, M.; Gonizzi, M. J . Phys. Chem. 1986,
90 (18), 4347-4353.
(22) Galkin, V. I.; Sayakhov, R. D.; Garifzyanov, A. R.; Cherkasov,
R. A.; Pudovik, A. N. Dokl. Chem. (Engl. Transl.) 1991, 318, 114-
116.
(16) Hagen, V.; Bendig, J .; Frings, S.; Wiesner, B.; Schade, B.;
Helms, S.; Lorenz, D.; Kaupp, U. B. J . Photochem. Photobiol. B: Biol.,
in press.
(23) Furuta, T.; Whang, S. S.-H.; Dantzker, J . L.; Dore, T. M.; Bybee,
W. J .; Callaway, E. M.; Denk, W.; Tsien, R. Y. Proc. Natl. Acad. Sci.
U.S.A. 1999, 96, 1193-1200 and references cited herein.

(Coumarin-4-yl)methyl Derivatives
J . Org. Chem., Vol. 64, No. 25, 1999 9113
Sch em e 4. P h otolysis of MCM Der iva tives
Ta ble 2. P h otop h ysica l a n d P h otoch em ica l Dea ctiva tion Beh a vior Da ta of th e Exa m in ed MCM Der iva tives
kf/109 s-1
kfn/109 s-1
kchem/109 s-1
knr/109 s-1
b
Φchem
Φf
λmaxem/nm
Φnr
τf/ns
pKa
3a
3b
3c
3d
4
5a
0.0043
0.0045
0.0052
0.0064
0.081
0.037
0.13
0.070
0.21
0.092
0.37
0.42
0.43
0.08
0.0072
0.052
0.030
0.040
0.014
0.030
400
400
400
402
394
404
400
400
400
401
0.63
0.58
0.56
0.91
0.91
0.91
0.84
0.89
0.78
0.89
1.41
4.8917
4.4118
3.9919
3.5420
-1.5421
0.7122
.5
0.71
0.26
0.0030
0.444
,0.2
<0.2
,0.2
.5
>5
0.15
0.036
0.26
0.65
0.41
0.19
4.20
4.56
4.55
5ba
5ca
5d a
5ea
a
b
Values in methanol/(aqueous) HEPES buffer solution (20:80 v/v). pKa value of the corresponding acid.
Sch em e 5. Rea ction P a th w a y a n d Mech a n ism of th e P h otolytic Clea va ge
due to the stability of the solvated anions and can also
compete successfully with the ion recombination (k0 based
on pseudo-internal conversion).
3a -5e with regard to the spectral position. However,
there are considerable differences in the intensity of
emission (more than 1 order of magnitude). 2 shows a
fluorescence quantum yield Φf of 0.1-1 in polar protic
solvents. On the other hand Φf is only 0.0014 in cyclo-
hexane.8 Also by using polar protic solvents we observed
a systematic dependence of Φf and τf on the solvent
polarity. It appears that the participation of fluorescence
as a main deactivation process increases with the water
content of the solvent mixture (Table 3). From photoa-
coustic measurements and from the absence of a phos-
phorescence spectrum at 77 K (ethanolic solution) we
P h otop h ysica l Dea ctiva tion Beh a vior of 2. With
regard to a better understanding of the fluorescence
enhancement caused by the appearance of 2 during the
irradiation of 4 and 5a -e, we studied the photophysics
of 2 in more detail. Furthermore 2 was used as a model
compound in order to examine the environmental (sol-
vent) influence on the deactivation behavior of the MCM
system in general. The fluorescence emission spectrum
of 2 is quite similar to those of the photolabile derivatives

9114 J . Org. Chem., Vol. 64, No. 25, 1999
Schade et al.
Ta ble 3. Solven t Dep en d en ce of th e P h otop h ysica l Dea ctiva tion of 2
ratio aqueous HEPES buffer solution/methanol ratio aqueous HEPES buffer solution/acetonitrile
Φf Φnr Φf Φnr
Φf/ns Φfn/ns kf/109 s-1 kfn/109 s-1 knr/109 s-1 τf/ns τfn/ns kf/109 s-1 kfn/109 s-1 knr/109 s-1
0.012 0.988 <0.2 16.67
x
0/100
20/80
40/60
60/40
70/30
80/20
100/0
THF
0.10 0.91
0.31 0.69
0.47 0.53
0.59 0.41
0.43
0.74
1.10
1.63
4.53
2.39
2.34
2.76
2.33
1.35
0.91
0.61
0.22
0.42
0.43
0.36
2.11
0.93
0.48
0.25
>5
2.50
0.060
0.31
0.31
0.30
0.30
4.94
2.19
1.39
0.85
0.45
0.13 0.88
0.18 0.82
0.26 0.74
0.40 0.60
0.40 3.20
0.59 3.28
0.87 3.35
1.33 3.33
1.70
1.15
0.75
0.65 0.35
2.20
3.38
0.46
0.30
0.16
0.55 0.45
1.88 3.42
0.53
0.29
0.24
0.007 0.993 ,0.2
dioxane 0.006 0.994 ,0.2
28.57
33.33
.5
.5
0.035
0.030
4.97
4.97
Sch em e 6. Sch em a tic P r esen ta tion of th e Ter m
Sch em es for th e Low est Lyin g Excited Sta tes of 4
in (a ) Non p ola r a n d (b) P ola r P r otic Solven ts
conclude that the isc process is not significant in the
deactivation of the S1 state (Φisc , 0.05). Analysis of the
rate constants for the radiative (kfn) and the nonradiative
(knr ≈ kic) deactivation as competing processes (see Table
3) shows that the solvent effect is mainly due to knr.
Whereas kfn appears almost independent of the solvent,
knr increases up to 1 order of magnitude within the series
from water to the pure solvents acetonitrile and methanol
respectively (Table 3). This behavior determines the
relative positions of the S1 (nπ*) and S2 (ππ*) states in
the term scheme of 2 (and the MCM chromophore in
general) and derives from the solvent dependence of the
distance between these energy levels. Scheme 6 presents
the position of these levels for 1 which is comparable to
2 due to the isoelectronic behavior with regard to n and
π electrons (for quantum mechanical considerations, see
ref 8). In a nonpolar aprotic environment there is a
dominating nπ* character for S1 despite the coupling and
mixing of the S1 and S2 states. This results in a high
probability for the nonradiative transition and therefore
in a low radiative efficiency of fluorescence for the
transition S1(nπ*) f S0 (Scheme 6a). In the case of polar
protic solvents the S1(nπ*) state is of higher energy and
the energy of the S2(ππ*) state is reduced, which should
lead to an approach or even inversion of the energy levels.
Therefore the deactivation behavior of the S1 state is
characterized by a ππ* state or by a state with a
dominating ππ* character. As a consequence, the disad-
vantage of the internal conversion process due to the
lower vibronic coupling results in an increasing partici-
pation of fluorescence in the deactivation. The data
summarized in Table 3 also illustrate this effect for
aqueous solvent mixtures. Whereas the rate of fluores-
cence kfn remains nearly constant, knr decreases with
increasing molar ratio of water leading to an increase of
Φf and τf.
F igu r e 1. Relationship between photochemical quantum yield
chem, and leaving group ability, represented by the pKa values
of the corresponding acid.
Φ
cien cy of th e Ca ged Com p ou n d s. The investigated
caged derivatives have the same coumarinyl moiety as
the chromophoric system. The absorption properties show
that there is no interaction, at least in the ground state,
which would significantly change the electronic system
with regard to its energy and symmetry (transition
probability). It can thus be concluded that in principle
the rates of photophysical deactivation kfn and kic (esti-
mated for 1 and 2) are comparable for all the compounds
investigated. According to eqs 3 and 4 therefore, the
(3)
(4)
progressive change in Φf and τf which were found by
structural variation (Table 2) must be due to the struc-
tural dependence of k1. A high value for k1 should occur
if both the cation within the ion pair (constant in the
MCM case) and the anion formed (leaving group) are
stable or stabilized, respectively. The latter can be seen
as the reason for the decrease of fluorescence quantum
yield and therefore the increase of the efficiency of
photochemistry with better leaving group ability. The pKa
values given in Table 2 describe the anion stability and
accordingly represent the leaving group ability. Since the
mesylate and diethyl phosphate are superior to carboxyl-
ates in this respect, the derivatives 4 and 5a exhibit a
higher photoreactivity. Figure 1 illustrates this relation-
ship.
Str u ctu r a l a n d Solven t Dep en d en cies of th e P h o-
toch em ica l a n d F lu or escen ce Dea ctiva tion Effi-

(Coumarin-4-yl)methyl Derivatives
J . Org. Chem., Vol. 64, No. 25, 1999 9115
As expected, the progressive change of the electronic
structure of the anion has significant effects as observed
for the substituted MCM benzoic acid esters (3b-d ).
Thus the appearance of the electron-withdrawing cyano
group at the para position of the benzoic acid ester (3d )
leads to an increased photochemical quantum yield,
whereas methoxy substitution (3b) leads to a decreased
rate of the photoreaction. Besides the direct structural
influence due to the stabilization of the leaving group,
an additional structural influence of the nucleoside
moiety can be observed for the phosphoric acid esters 5b-
e. Both compounds show an increased photochemical
quantum yield compared to the diethyl phosphate 5a .
This effect can be explained by additional stabilization
of the MCM cation (by partial charge transfer) and/or the
anion (by hydrogen bonding) within the ion pair inter-
mediate by the purine base, respectively. The question
as to what extent an exciplex interaction between the
purine base and the coumarinyl moiety occurs as an
element of this stabilization within the excited S1 state24
cannot be answered. On the basis of the differences
concerning the donor ability of adenine and guanine, the
stabilization of the intermediate is greater for the cGMP
compared to the corresponding cAMP derivatives, as
reflected by higher values for the photochemical quantum
yield. Thus the one electron standard oxidation potential
E°(N•*/N) in acetonitrile has values of 1.47 V (guanosine)
and 1.94 V (adenosine), respectively.24
F igu r e 2. Competition between photophysical (Φf) and pho-
tochemical (Φchemical) deactivation as a function of the polarity
of the aqueous solvent mixture for 5a .
Furthermore a small but significant difference is
observed by comparison of the deactivation parameters
for the diastereomers (Table 2). Due to steric reasons an
interaction between the purine base and the positively
charged center should be favored within the axial con-
figuration. As a result of the competing processes of
photochemistry and fluorescence, the axial isomers 5b,d
show higher photochemical quantum yields, whereas a
stronger fluorescence is observed for the equatorial
isomers 5c,e.
F igu r e 3. Schematic relationship between the kinetic data
for the deactivation of excited MCM derivatives and increasing
solvent polarity.
As in the case of 2 the photophysical and photochemical
deactivation behavior of the caged compounds is strongly
dependent on the solvent and the composition of the
solvent mixtures. Thus the efficiency of the photoreaction
increases continuously with increasing amount of the
stronger polar protic solvent component whereas the
fluorescence quantum yield increases at the beginning
and decreases after reaching a maximum during the
same solvent variation. Figure 2 represents this effect
for the diethyl phosphate 5a . A very similar behavior was
observed for the photolabile MCM derivatives 3, 4, and
5b-e. The interpretation of the increase of Φchem is given
by the opportunity of better stabilization of the ion pair
by solvation which leads to an acceleration of k1 and,
therefore, according to eq 2 to the consequent continuous
increase of Φchem. Additionally the polarity dependence
of the rate of photolysis is further evidence for the SN1
character of the photolytic cleavage. With regard to the
solvent dependence of the fluorescence quantum yield
there is a superimposition of two opposing effects: first,
the solvent dependent rate of nonradiative deactivation
(kic ≈ knr) which decreases with increasing polarity and,
second, the rate of the photochemical reaction k1 which
increases with increasing polarity. The fluorescence
quantum yield Φf is described by eq 3. Therefore the
global effect is the result of the initially dominating
decrease of kic ≈ knr from the less polar solvent composi-
tion. This effect becomes overcompensated progressively
by the contrary dependence of k1 on the solvent polarity.
That relationship is schematically illustrated in Figure
3.
In conclusion, MCM esters liberate the corresponding
acids during photolysis and are suited for the creation
of caged compounds and therefore for the triggering of
potential biologically active molecules with such func-
tional groups. As we could show, this liberation proceeds
in the nanosecond time region, confirming the usefulness
of the MCM moiety as a caging group. A strong polarity
of the solvent favors the rate of photolysis. Analysis of
the photolytic products and of the kinetic data as well
as the results of isotope labeling studies show that the
reaction proceeds via a photo SN1 mechanism (solvent-
assisted photoheterolysis), with formation of an ion pair
as intermediate. Mechanistically determined, the pho-
tocleavage is most effective in the case of good leaving
groups such as sulfonic and phosphoric acid derivatives.
In these cases the photolytic liberation is combined with
a marked fluorescence enhancement which should be
suited for the visualization of the uncaging process inside
living cells. With regard to the fluorescence enhancement
(24) Seidel, C. A. M.; Schulz, A.; Sauer, M. H. M. J . Phys. Chem.
1996, 100, 5541-5553.

9116 J . Org. Chem., Vol. 64, No. 25, 1999
Schade et al.
extracted with 2 × 40 mL of ethyl acetate. The combined
organic phases were dried with sodium sulfate, and the solvent
was evaporated in a vacuum. Flash chromatography of the
residue through a short column with chloroform yielded 576
mg (1.86 mmol; 98%) of 3c as a pale yellow solid, mp 151 °C:
1H NMR (CDCl3) δ 3.88 (3H, s), 5.50 (2H, s), 6.44 (1H, s), 6.86-
6.91 (2H, m), 7.44-7.62 (4H, m), 8.84-8.12 (2H, m); 13C NMR
(CDCl3) δ 56.5, 62.3, 101.9, 110.6, 111.3, 113.3, 125.1, 129.3
(2C), 129.6, 130.5 (2C), 134.4, 150.0, 156.2, 161.5, 163.6, 166.3;
ESI mass spectrum m/e 311.2 [M + H]+. Anal. Calcd for
and kinetics of photocleavage the MCM system is at the
limits of its efficiency in the case of phosphates and
sulfonates.
Further studies with regard to the influence of the type
and position of substituents on the coumarinyl moiety
on the “cationic stabilization” of the caged derivatives are
in progress.
Exp er im en ta l Section
C
18H14O5: C, 69.67; H, 4.55. Found: C, 69.62; H, 4.52.
Ma ter ia ls a n d Gen er a l Meth od s. 7-Methoxy-4-methyl-
coumarin and 4-(bromomethyl)-7-methoxycoumarin were pur-
chased from Sigma (Germany). 4-(Diazomethyl)-7-methoxy-
coumarin (7) was synthesized according to Ito25 starting with
selective oxidation of 1 to the carbaldehyde, subsequent
formation of the tosylhydrazone, and cleavage with triethy-
lamine to the diazo compound. TLC plates, silica 60-F254, were
from E. Merck (Darmstadt, Germany). Silica gel (30-60 µm)
for flash chromatography was purchased from J . T. Baker (The
Netherlands). Acetonitrile from Riedel-deHae¨n (Germany) was
HPLC grade. All other chemicals and solvents were reagent
grade and were used without further purification. Water was
purified with a Milli Q system (Millipore, Germany). Analytical
HPLC was run on a Hewlett-Packard HP 1100 system with
DAD and fluorescence detection; for preparative HPLC a
Shimadzu LC-8A system with UV detection (SPD-6AV) was
used. A C18 column (Spherisorb ODS 2, 5 µ, 250 × 4 mm,
Polymer Laboratories Ltd., U.K.) was used for HPLC analysis
at a flow rate of 1 mL/min at 20 °C with an injection volume
of 20 µL. Preparative HPLC was run using a Nucleosil 100-7
C18 column (7 µ, 250 × 20 mm, Machery-Nagel, Germany) at
a flow rate of 10 mL/min. Eluent A was water, and eluent B
acetonitrile.1H NMR and 13C NMR were recorded on a Gemini
200 spectrometer (Varian) using TMS as internal standard.
31P NMR spectra were recorded using a Bruker DRX 600
spectrometer with 85% phosphoric acid as external standard.
All melting points are uncorrected.
(7-Meth oxycou m ar in -4-yl)m eth yl-4-cyan oben zoate (3d).
This was made analogously to 3c using 517 mg (1.92 mmol)
of 6, 423 mg (2.88 mmol) of 4-cyanobenzoic acid, and 277 mg
(4.77 mmol) of potassium fluoride in 20 mL of acetone. After
3 h of reflux acetone was evaporated; the residue was
suspended in chloroform and washed with diluted sodium
chloride solution. The organic phase was washed with NaHCO3
in order to remove unconverted 4-cyanobenzoic acid. After
neutralization the solvent was evaporated. Yield: 540 mg (1.61
mmol; 83.8%) of 3d as a yellow solid. The crude product was
further purified by preparative HPLC using a linear gradient
5-80% B in 45 min; tR ) 43 min (λDet ) 320 nm) followed by
lyophilization: mp 243 °C; 1H NMR (CDCl3) δ 3.90 (3H, s),
5.55 (2H, s), 6.41 (1H, s), 6.88-6.94 (2H, m), 7.46-7.51 (1H,
m), 7.78-7.82 (2H, m), 8.19-8.24 (2H, m); 13C NMR (CDCl3)
δ 56.5, 63.0, 102.0, 111.0, 111.2, 113.6, 117.9, 118.4, 125.8,
129.0 (2C), 133.3 (2C), 133.5, 149.2, 156.4, 161.4, 163.8, 164.8;
ESI mass spectrum m/e 336.3 [M + H]+. Anal. Calcd for
C
19H13O5: C, 68.06; H, 3.91; N, 4.18. Found: C, 68.12; H, 3.98,
N, 4.22.
(7-Meth oxycou m ar in -4-yl)m eth ylm eth an esu lfon ate (4).
To a suspension of 474 mg (2.19 mmol) of 3 in 10 mL of
chloroform was added 277 mg (2.88 mmol) of methanesulfonic
acid in 1 mL of chloroform dropwise under vigorous stirring
at room temperature. The mixture was refluxed for 3 h.
Evaporation of the solvent and recrystallization of the residue
from THF yielded 162 mg (0.57 mmol, 26%) of 11 as a pale
yellow solid. The compound was further purified by prepara-
tive HPLC using a linear gradient 5-50% B in 45 min,
isocratic 50% B from 45 to 50 min; tR ) 45 min (λDet ) 320
nm) followed by lyophilization: mp 159 °C; 1H NMR (DMSO-
d6) δ 3.37 (3H, s), 3.87 (3H, s), 5.54 (2H, s), 6.38 (1H, s), 7.03-
7.05 (2H, m), 7.62-7.67 (1H, m); 13C NMR (DMSO-d6) δ 39.0,
56.3, 66.9, 101.3, 110.2, 110.3, 112.7, 126.2, 149.0, 155.3, 160.1,
163.0; ESI mass spectrum m/e 285.1 [M + H]+. Anal. Calcd
for C12H12O6S: C, 50.70; H, 4.25. Found: C, 50.65; H, 4.35.
(7-Met h oxycou m a r in -4-yl)m et h yl(d iet h yl)p h osp h a t e
(5a ). A 1.031 g (5.0 mmol) amount of 2 was suspended in 9
mL of pyridine and cooled to 0 °C. Under vigorous stirring
1.035 g (6.0 mmol) of diethylphosphoric acid chloride was
added dropwise. After being stirred at 0 °C for 2 h, the mixture
was allowed to warm to room temperature under continued
stirring for 2 h. Water was added to dissolve precipitated
pyridinium chloride. The mixture was extracted with ether (4
times), and the organic layer was washed with 2 × 10 mL of
1 N H2SO4, 10 mL of NaHCO3 (5%), 3 × 10 mL of water, and
10 mL of brine. The combined organic phases were dried over
MgSO4 and evaporated. The residue was dissolved in a small
amount of ether, and pentane was added. After refrigeration
overnight, 1.34 g (3.91 mmol; 78.2%) of 5a was obtained as a
white solid, mp 65 °C: ESI mass spectrum m/e 343.1 [M +
H]+. NMR data are in agreement with ref 14.
Syn t h eses. (7-Met h oxycou m a r in -4-yl)m et h yl-n -h ep -
ta n oa te (3a ). To a suspension of 283 mg (1.05 mmol) of 6 in
20 mL of acetone were added 122 mg (2.1 mmol) of potassium
fluoride and 137 mg (1.05 mmol) of n-heptanoic acid. The
mixture was stirred at room temperature for 1 d, and then
the solvent was removed in vacuo. The residue was suspended
in 20 mL water and extracted with 2 × 40 mL of ethyl acetate.
The combined organic phases were dried with sodium sulfate,
and the solvent was evaporated in a vacuum. Flash chroma-
tography of the residue with petroleum ether/ethyl acetate (8:2
v/v) yielded 279 mg (0.876 mmol; 82.6%) of 3a as a pale yellow
1
solid: mp 94 °C; H NMR (CDCl3) δ 0.89 (3H, t, J ) 6 Hz),
1.27-1.36 (8H, m), 2.46 (2H, t, J ) 7.5 Hz), 3.90 (3H, s), 5.31
(2H, s, 6.35 (1H, s), 6.87-6.89 (2H, m), 7.42-7.44 (1H, m); EI
mass spectrum m/e (relative intensity) 319 (M+ + 1, 4), 318
(M+, 16), 189 (6), 175 (2), 129 (2), 85 (33). Anal. Calcd for
C
18H22O5: C, 67.91; H, 6.97. Found: C, 67.8; H, 6.97.
(7-Me t h oxycou m a r in -4-yl)m e t h yl-4-m e t h oxyb e n zo-
a te (3b). This was made analogously to 3a using 465 mg (1.68
mmol) of 6, 215 mg (1.41 mmol) of 4-methoxybenzoic acid, and
204 mg (3.51 mmol) of potassium fluoride in 20 mL acetone.
Yield: 432 mg (1.38 mmol; 90.0%) of 3b as a pale yellow solid,
mp 158 °C. 1H NMR (CDCl3): δ 3.88 (6H, s), 5.47 (2H, s), 6.43
(1H, s), 6.85-6.98 (4H, m), 7.47-7.51 (1H, m), 8.04-8.08 (2H,
m). 13C NMR (CDCl3): δ 56.0, 56.3, 61.9, 101.8, 110.4, 111.2,
113.2, 114.4 (2C), 121.8, 125.0, 132.4 (2C), 150.1, 156.1, 161.5,
163.4, 164.5, 165.9. ESI mass spectrum: m/e 341.2 [M + H]+.
Anal. Calcd for C19H16O6: C, 67.05; H, 4.74. Found: C, 67.16;
H, 4.86.
(7-Meth oxycou m a r in -4-yl)m eth ylben zoa te (3c). Analo-
gously to 3a a mixture of 510 mg (1.90 mmol) of 6, 287 mg
(2.35 mmol) of benzoic acid, and 237 mg (4.08 mmol) of
potassium fluoride was refluxed in 20 mL of acetone for 3 h.
The resulting mixture was suspended in 20 mL of water and
(7-Meth oxycou m a r in -4-yl)m eth yl Ad en osin e Cyclic-3′,
5′-m on op h osp h a te (5b,c). A mixture of 323 mg (1.2 mmol)
of 6 and 243 mg of the dihydrate of the tetra-n-butylammo-
nium salt of cAMP (prepared from cAMP by passing through
an acidic Dowex ion-exchange column equilibrated with a
solution of tetra-n-butylammonium hydroxide followed by
lyophilization) was refluxed in 30 mL of acetonitrile in the dark
for 5 h. After evaporation of the solvent the residue was
washed with 20 mL of water, dried, dissolved in a small volume
of chloroform/methanol (1:1 v/v), and purified by flash chro-
matography. Elution using chloroform removed unconverted
6 from the column. Elution with chloroform/methanol (2.5:97.5
(25) Ito, K.; Maruyama, J . Chem. Pharm. Bull. 1983, 31 (9), 3014-
3023.

(Coumarin-4-yl)methyl Derivatives
J . Org. Chem., Vol. 64, No. 25, 1999 9117
to 5:95 v/v) yielded 133 mg (64.4%) of the pure diastereomeric
mixture of 5b,c in a 80:20 ratio (axial/equatorial) as a colorless
solid after evaporation and lyophilization. The isomers were
separated from each other by preparative HPLC using a linear
gradient 10-45% B in 105 min (λDet ) 320 nm) followed by
lyophilization.
irradiation absorption spectra were recorded and concentra-
tions of the remaining caged compounds were measured by
HPLC. The photochemical quantum yields, Φchem, defined as
the ratio of caged molecules converted to the amount of photons
absorbed, were determined as described2b using potassium
ferrioxalate actinometry26 according to the equation Φchem
)
5b: 1H NMR (DMSO-d6) δ 3.86 (3H, s), 4.29 (1H, dt, J )
10.2 and 4.8 Hz), 4.46 (1H, t, J ) 9.8 Hz), 4.68-4.73 (2H, m),
5.38 (1H, dd, J ) 9.8 and 5.0 Hz), 5.48 (2H, d, J ) 6.3 Hz),
6.06 (1H, s), 6.43 (1H, s), 6.46 (1H, s), 6.96 (1H, dd, J ) 8.8
and 2.5 Hz), 7.08 (1H, d, J ) 2.4 Hz), 7.40 (2H, s br), 7.72
(1H, d, J ) 8.9 Hz), 8.10 (1H, s), 8.35 (1H, s); 31P NMR (DMSO-
d6) δ -5.1; ESI mass spectrum m/e 518.3 [M + H]+.
(dc/dt)Iabs-1V. The initial slope dc/dt of the experimentally
obtained concentration/irradiation time function was deter-
mined directly using the HPLC peak areas of the caged
compounds. The value of Iabs (absorbed light intensity at t ) 0
and R ) 333 nm in “mol of photons” s-1) was calculated using
the absorption factor R (taken from the UV spectra at t ) 0
and R ) 333 nm) multiplied by the light intensity of the HBO
500 exposure tool (determined using the actinometer com-
pound). The volume V of solution irradiated was 3 mL.
1
5c: H NMR (DMSO-d6) δ 3.87 (3H, s), 4.51 (2H, m), 4.71
(1H, d, J ) 4.7 Hz), 4.76 (1H, d, J ) 12.4 Hz), 5.37 (1H, m),
5.41 (2H, m), 6.08 (1H, s), 6.39 (1H, s), 6.39-6.50 (1H, s br),
7.01 (1H, d, J ) 8.8 Hz), 7.07 (1H, s), 7.40 (2H, s br), 7.69 (1H,
d, J ) 8.8 Hz), 8.19 (1H, s), 8.40 (1H, s); 31P NMR (DMSO-d6)
δ -3.6; ESI mass spectrum m/e 518.3 [M + H]+.
F lu or escen ce Mea su r em en ts. Fluorescence spectra were
measured using a MPF-2A fluorescence spectrometer (Hitachi-
Perkin-Elmer) combined with a correction and digitalization
unit. The 10 µM solutions of the compounds under investiga-
tion were placed in quartz cuvettes with a path length of 1
cm and excited in a perpendicular arrangement. In the case
of the highly reactive compounds 4 and 5a -e, the excitation
intensity (controlled by the excitation slit) was very low and
the registration time was very short (10 s) to prevent excessive
photolysis.
The fluorescence quantum yields were determined at 298
K by the relative method27 using quinine sulfate as a standard
(Φf ) 0.545 in 0.1 N H2SO4). At excitation wavelength (328-
335 nm) the absorbance values of the solutions of the standard
and the investigated compound were identical. The different
refractive indices of the solutions were taken into account.
The time-resolved fluorescence decay measurements (pulse
sampling method) were performed using a nitrogen laser (λ )
337 nm) as excitation source and a transient recorder for the
decay registration. The wavelength of the fluorescence detec-
tion was λem ) 405 nm (interference filter). Details of the
equipment and the deconvolution procedure of the experimen-
tal decay curve are described in ref 28.28 The time resolution
achieved was about 250 ps.
Electr osp r a y Ma ss Sp ectr om etr y. ESI mass spectrom-
etry was performed on a triple quadrupole instrument (TSQ
700, Finnigan MAT, Germany) equipped with an electrospray
ion source (API-ESI) operating in the positive mode with a
capillary temperature of 200 °C and a voltage of 4.5 kV. The
sample concentration was 150 pmol/µL in acetonitrile/acetic
acid (99:1, v/v). The samples were introduced into the ion
source at a constant flow rate of 3 µL/min. Nitrogen was used
as sheath gas at a pressure of 3.4 bar. The spectra represent
an average sum of 64 scans.
(7-Meth oxycou m a r in -4-yl)m eth yl Gu a n osin e Cyclic-
3′,5′-m on op h osp h a te (5d ,e). A mixture of 191 mg (0.5 mmol)
of cGMP dihydrate and 216 mg (1.0 mmol) of 7 was stirred in
20 mL of DMSO and 20 mL of acetonitrile at 50 °C in the dark
for 24 h. Acetonitrile was evaporated under reduced pressure,
and DMSO was removed by repeated extraction with ether.
The residue was purified by flash chromatography. Elution
with methanol/chloroform (1:19 to 1:4 v/v) yielded 65 mg
(22.8%) of the pure diastereomeric mixture of 5d ,e in a 50:50
ratio as a colorless solid after evaporation and lyophilization.
Preparative HPLC permitted complete separation into the
diastereomers d and e using a linear gradient 5-45% B in
105 min (λDet ) 320 nm) followed by lyophilization.
5d : 1H NMR (DMSO-d6) δ 3.86 (3H, s), 4.23 (1H, dt, J )
10.3 and 4.9 Hz), 4.55 (1H, t, J ) 9.8 Hz), 4.58 (1H, t, J ) 4.8
Hz), 4.68 (1H, ddd, J ) 22.1, 9.3 and 4.8 Hz), 4.83 (1H, dd, J
) 9.8 and 5.0 Hz), 5.45 (2H, d, J ) 6.1 Hz), 5.84 (1H, s), 6.38
(1H, d, J ) 4.9 Hz), 6.49 (1H, s), 6.53 (2H, s br), 6.97 (1H, dd,
J ) 8.9 and 2.5 Hz), 7.06 (1H, d, J ) 2.4 Hz), 7.72 (1H, d, J )
8.9 Hz), 7.94 (1H, s), 10.72 (1H, s br); 31P NMR (DMSO-d6) δ
-5.1; ESI mass spectrum m/e 534.3 [M + H]+. Anal. Calcd for
C
21H20N5O10P‚2H2O: C, 44.30; H, 4.25; N, 12.30. Found: C,
44.30; H, 4.08; N, 11.42.
1
5e: H NMR (DMSO-d6) δ 3.87 (3H, s), 4.43-4.46 (1H, m),
4.48 (1H, m), 4.58 (1H, t, J ) 4.7 Hz), 4.72-4.76 (1H, m), 5.18
(1H, dd, J ) 9.4 and 5.0 Hz), 5.41 (2H, m), 5.85 (1H, s), 6.31
(1H, d, J ) 4.5 Hz), 6.39 (1H, s), 6.60 (2H, s br), 7.00 (1H, dd,
J ) 8.9 and 2.5 Hz), 7.07 (1H, d, J ) 2.4 Hz), 7.68 (1H, d, J )
8.9 Hz), 10.74 (1H, s); 31P NMR (DMSO-d6) δ -4.0; ESI mass
spectrum m/e 534.3 [M + H]+. Anal. Calcd for C21H20N5O10P‚
H2O: C, 45.74; H, 4.02; N, 12.70. Found: C, 45.88; H, 3.99; N,
11.79.
Ack n ow led gm en t. We thank S. Helm, B. Gentsch,
and J . Lossmann for technical assistance. We are
grateful to Dr. R. Winter for NMR investigations. The
authors thank the Forschungsinstitut J u¨lich GmbH, the
Deutsche Forschungsgemeinschaft (DFG), and the Fonds
der chemischen Industrie for financial support.
Stea d y-Sta te UV/Vis Sp ectr oscop y, P h otolysis Exp er i-
m en ts, a n d Mea su r em en ts of th e P h otoch em ica l Qu a n -
tu m Yield s. UV/vis spectra were recorded with a U-3410
spectrophotometer (Hitachi, J apan). Photolysis was carried out
using a high-pressure mercury lamp (HBO 500, Oriel) with
controlled light intensity and a metal interference filter of 333
nm (Schott, Germany). For quantum yield determinations the
irradiated solutions were analyzed using HPLC. The 20 µM
solutions of the MCM-caged compounds in aqueous HEPES
buffer solution (0.01 M HEPES/NaOH, 0.12 M NaCl, 3 mM
KCl, 1 mM CaCl2, 1 mM MgCl2, pH 7.2)/30% acetonitrile or
20% methanol, respectively, were placed in quartz cuvettes
with a path length of 1 cm and irradiated for periods ranging
from 3 to 90 s in steps of 3-10 s. Before and after each
J O9910233
(26) Kuhn, H. J .; Braslawsky, S. E.; Schmidt, R. Pure Appl. Chem.
1989, 61, 187-210.
(27) Demas, J . N.; Crosby, G. A. J . Phys. Chem. 1971, 75, 991; EPA
Newsl. 1986, 28, 21.
(28) Grewer, C.; Brauer, H.-D. J . Phys. Chem. 1993, 97, 5001-5006.