Photochemically-triggered decarboxylation/deamination of o-nitrodimethoxyphenylglycine

Article abstract of DOI:10.1021/ol9907062

(Matrix presented) o-Nitrodimethoxyphenylglycine (1) decomposes on photolysis at 366 nm to release ammonia and carbon dioxide, via a five-step mechanism (Scheme 1). We have obtained spectroscopic and kinetic data from experiment and calculation which supp

Full text of DOI:10.1021/ol9907062

ORGANIC
LETTERS
1999
Vol. 1, No. 4
619-621
Photochemically-Triggered
Decarboxylation/Deamination of
o-Nitrodimethoxyphenylglycine
Christopher D. Woodrell, Polina D. Kehayova, and Ahamindra Jain*
Department of Chemistry, Swarthmore College, Swarthmore, PennsylVania 19081-1397
ajain1@swarthmore.edu
Received June 2, 1999
ABSTRACT
o-Nitrodimethoxyphenylglycine (1) decomposes on photolysis at 366 nm to release ammonia and carbon dioxide, via a five-step mechanism
(Scheme 1). We have obtained spectroscopic and kinetic data from experiment and calculation which support our mechanism. We have also
confirmed that a carbonic anhydrase inhibitor bearing 1 as a photolabile cage releases a tight-binding inhibitor on photolysis at 366 nm.
Photolabile molecular cages have been proven to be widely
applicable in biochemical systems and in organic synthesis.
Photolabile synthetic amino acids have been used to activate
a protein on photolysis1 and to regulate protein function, thus
aiding in the identification of sites of covalent modification
at cageable side chains2 and critical sites in cellular signal
transduction.3 Incorporation of synthetic amino acids such
that the photolabile bond is in the backbone of the protein
has allowed site-specific proteolysis, toward the goal of
regulation of ion channel proteins.4 Photolabile protecting
groups have also been used in solid-phase oligonucleotide,
organic, and peptide synthesis.5-7
activated by light upon delivery. A hydrophobic inhibitor
of carbonic anhydrase II9 (CA) has been modified with
o-nitrodimethoxyphenylglycine (o-NDMPG, 1), a polar cag-
ing group.10 Nitrophenyl derivatives such as 1 have found
utility in many biochemical systems.1-8 The benzylic car-
boxylate group of our cage increases its water solubility at
neutral pH. Delivering the inhibitor as a prodrug is particu-
larly useful because, in addition to allowing control of
activation, the polar caging group increases the water
solubility of the drug10 without compromising activity of the
inhibitor.9 Methoxy substituents have been shown to increase
the rate of photolysis of nitrophenyl derivatives,5,6 thus their
inclusion in our cage. While many uses of such photolabile
cages have been identified, the intermediates and products
of photodeprotection have not been fully characterized.
Our proposed mechanism for photodeprotection (Scheme
1), based on the work of England et al.,4 Holmes,6 and
McCray and Trentham,8 begins with the generation of aci-
nitro intermediate 2 (Scheme 1) via n f π* excitation of 1
and intramolecular hydrogen atom transfer from the R carbon
of the amino acid. Intermediate 2 cyclizes, affording 3. The
subsequent collapse of 3, in the presence of a proton source,
The requirement of photolysis for activation allows the
introduction of ligands of known concentration and directed
localization.8 Photolabile cages may therefore be of particular
interest in the design of site-specific drugs, which are
(1) Chang, C.; Niblack, B.; Walker, B.; Bayley, H. Chem. Biol. 1995, 2,
391-400.
(2) Miller, J. C.; Silverman, S. K.; England, P. M.; Dougherty, D. A.;
Lester, H. A. Neuron 1998, 20, 619-624.
(3) Pan, P.; Bayley, H. FEBS Lett. 1997, 405, 81-85.
(4) England, P. M.; Lester, H. A.; Davidson, N.; Dougherty, D. A. Proc.
Natl. Acad. Sci. U.S.A. 1997, 94, 11025-11030.
(5) Hasan, A.; Stengele, K. P.; Giegrich, H.; Cornwell, P.; Isham, K.
R.; Sachleben, R. A.; Pfleiderer, W.; Foote, R. S. Tetrahedron 1997, 12,
4247-4264.
(6) Holmes, C. P. J. Org. Chem. 1997, 62, 2370-2380.
(7) Rich, D. H.; Gurwara, S. K. J. Am. Chem. Soc. 1975, 97, 1575-
1579.
(8) McCray, J. A.; Trentham, D. R. Annu. ReV. Biophys. Biophys. Chem.
1989, 18, 239-270.
(9) Doyon, J. B.; Jain, A. Org. Lett. 1999, 1, 183-185.
(10) Kehayova, P. D.; Bokinsky, G. E.; Huber, J. D.; Jain, A. Org. Lett.
1999, 1, 187-188.
10.1021/ol9907062 CCC: $18.00 © 1999 American Chemical Society
Published on Web 07/15/1999

Scheme 1. Proposed Mechanism of Photodecomposition of o-Nitrodimethoxyphenylglycine
results in the elimination of ammonia, cleaving the “photo-
labile” bond and forming R-ketoacid 4. Decarboxylation of
4 gives transient ketene 5, with the o-nitroso substituent
serving as an electron sink. Although 5 should be a high-
energy structure, minimization of the conformation of 4 using
molecular mechanics (MOPAC) in CAChe11 showed that the
carbon-carbon bond to the carboxylic acid is properly
aligned for decarboxylation (Scheme 1). In the last step, the
o-nitroso substituent is restored and a proton is abstracted
from the solvent to form o-nitrosobenzaldehyde derivative
6.12
MOPAC/ZINDO calculations11 afforded calculated elec-
tronic spectra for 1-6. On the basis of the differences evident
in these spectra, we expected to be able to follow the progress
of our mechanism at specific wavelengths. By photolyzing
1 in both protic and aprotic solvents and by varying pH, we
were able to gather further evidence in support of the
structures along the decomposition pathway.
force for the decomposition to 4. Furthermore, 3 may not
exist as a zwitterion in CHCl3. Neutral 3 lacks the necessary
leaving group, disfavoring further decomposition, so for these
two reasons we do not expect to obtain 6 in CHCl3. We
determined that the rate of appearance of 2, based on the
change in absorbance at 400 nm with time, was 0.0073 (
0.0010 s-1.14
Photolysis of 1 at 366 nm in acidic CH3OH gave 4,5-
dimethoxy-2-nitrosobenzaldehyde (6, Scheme 1).12 The
spectra obtained on photolysis as a function of time indicate
the disappearance of absorbances corresponding to 1 at 300
and 345 nm. The N-hydroxyl group of 3 is partially
deprotonated, and the ammonia leaving group is protonated,
facilitating formation of 4. The increase in absorbance at
322 nm reflects formation of 6 at a rate of 0.010 ( 0.001
s-1.
Photolysis of 1 at 366 nm in 20 mM KH2PO4 in the
presence of dilute HCl also resulted in direct product
formation. At both pH 1 and pH 2, absorbances for 1 at 300
and 345 nm decrease in intensity upon irradiation, and a peak
characteristic of product appears near 320 nm (Figure 2).
The rate of product formation at pH 2, as measured by the
change in absorbance at 319.5 nm, was 0.00099 ( 0.00016
s-1. The rate of product formation at pH 1, followed at 322
nm, was 0.0033 ( 0.0007 s-1.
13
Photolysis of 1 in CHCl3 implicated intermediate 2 as
the predominant photoproduct in this aprotic medium
(Scheme 1). Successive irradiation at 366 nm over 5 min
yielded the spectra shown in Figure 1. The disappearance
of 1 and appearance of 2 is evident from the increase in
absorption at 400 nm. We believe that aci-nitro intermediate
2 is in equilibrium with cyclic intermediate 3.
The relative rates of photodecomposition of 1 in protic
solvents support our mechanistic scheme. The faster rate at
The ammonium group of 3 is a good leaving group, but
the neutral N-hydroxyl does not supply adequate driving
Figure 2. Absorbance spectra of 1 in 20mM KH2PO4 (pH 1),
obtained after 0, 30, 60, 90, 150, 210, 270, 360, 420, 540, and 660
s of photolysis.
Figure 1. Absorbance spectra of 1 in CHCl3, obtained after 0, 60,
120, 180, 240, and 300 s of photolysis.
620
Org. Lett., Vol. 1, No. 4, 1999

low pH is consistent with our hypothesis that the amino group
of 3 must be protonated to facilitate formation of 4. In water,
the rate was slower than in CH3OH, which is consistent with
the fact that ionic intermediates 2 and 3 are stabilized by
solvation.
Caged inhibitor 1‚I10 was photolyzed in the manner
described above, to release free I. We confirmed that I binds
to CA via a competitive fluorescence-based assay.15 To a
Figure 3. Deprotection of 1‚I monitored by the decrease in
fluorescence observed on displacement of dansylamide from the
active site of carbonic anhydrase by I.
solution of CA (40 nM) saturated with dansylamide (DNSA,
20 µM) in 20 mM KH2PO4 (pH 8) was added 1‚I (200 nM).
This solution was examined by fluorescence spectropho-
tometry, via measurement of emission from CA‚DNSA on
excitation of CA at 280 nm. As the solution was photolyzed,
I was released from the caged form resulting in a time-
dependent decrease in fluorescence (measured using a SPEX
Fluorolog-3 spectrofluorimeter; Figure 3), as free I displaced
DNSA from the active site of CA. These data were analyzed
to afford a rate of deprotection of 0.0028 ( 0.0007 s-1.16
In conclusion, we have obtained spectroscopic and kinetic
evidence that support the mechanism shown in Scheme 1
for the photolysis of o-NDMPG. The rate and products of
photolysis are solvent and pH-dependent. We have also
confirmed that a CA inhibitor bearing 1 as a photolabile cage
(1‚I) releases the active form of the inhibitor (I) on
photolysis. We are now in the process of developing an
alternate delivery mechanism for our caged CA inhibitor,
involving attachment of 1‚I to a gold surface via a photolabile
tether.
Acknowledgment. Swarthmore College, the donors of
the Petroleum Research Fund, administed by the American
Chemical Society, Merck/AAAS, and the Howard Hughes
Medical Institute are acknowledged for provided funding.
(11) Version 4.0.2, Oxford Molecular Group, 1998. Uncertainties in the
calculated electronic transitions are 10-20%.
(12) 1H NMR (Bruker DRX-400, CDCl3) δ 10.5 (s, 1H), 7.63 (s, 1H),
7.44 (s, 1H), 4.05 (s, 3H), 4.04 (s, 3H).
(13) Typical conditions involved photolysis of 67 µM o-NDMPG in
solvent with a handheld UV lamp at 366 nm, directly in a cuvette. The
reaction was followed with a JASCO V-550 UV/vis spectrophotometer, at
room temperature.
Supporting Information Available: Experimental pro-
cedure for the synthesis of 1 and spectra and kinetic data
not shown in Figures 1-3. This material is available free of
charge via the Internet at http://pubs.acs.org.
(14) The rate of disappearance of o-NDMPG in CHCl3 was also
determined at 302 nm to be 0.0066 ( 0.0013 s-1
.
(15) Chen, R. F.; Kernohan, J. C. J. Biol. Chem. 1967, 242, 5813-5823.
(16) We have also followed deprotection of 1‚I by UV/vis spectropho-
tometry at 395 nm and have obtained a rate of 0.0046 ( 0.0003 s-1 via
that approach.
OL9907062
Org. Lett., Vol. 1, No. 4, 1999
621