Site-specific prodrug release using visible light

Article abstract of DOI:10.1021/ja800140g

Using a photosensitization-singlet oxygenation-dioxetane cleavage strategy, a photodynamic prodrug system has been developed, whereby drugs bearing carbonyl groups can first be attached to a photosensitizer to give a photosensitizer-drug complex and then released from the complex upon visible light irradiation. Visible light, which has good penetration through tissue, generates singlet oxygen via the photosensitizer, which then releases the prodrug when and where required. With this system, drug mimics and methyl esters of NSAIDs have been successfully incorporated with photosensitizers related to verteporfin and then released by visible light illumination in high to quantitative yields within minutes. Copyright

Full text of DOI:10.1021/ja800140g

Published on Web 03/12/2008
Site-Specific Prodrug Release Using Visible Light
Michael Y. Jiang and David Dolphin*
Department of Chemistry, UniVersity of British Columbia, Room W300 - 6174 UniVersity BouleVard,
VancouVer, BC, V6T 1Z3, Canada
Received January 7, 2008; E-mail: david.dolphin@ubc.ca
Scheme 1. General Diagram of Drug Incorporation and Release
Prodrug approaches are widely used in drug discovery, and the
release of the parent drugs is generally through enzymatic activation
and can be site-specific when unique activating enzymes and
appropriate chemical environment are present at the target site.1
An external and nonenzymatic activation providing more direct
controls over the course of drug release would be attractive. Herein,
we report a novel site-specific prodrug system, in which visible
light is employed to trigger drug release.
UV light triggerable prodrugs have been reported.2 However,
the poor tissue-penetrating nature of UV light, less than 1 mm,
hampers their clinical utility. In fact, only visible light between
650 to 800 nm can penetrate tissue effectively, and some photo-
sensitizers with a strong absorption band within this range have
been developed as photomedicines using photodynamic therapy
(PDT).3 The chemical basis of PDT relates to the ability of the
visible-light-activated photosensitization to convert normal triplet
oxygen into singlet oxygen, which reacts with various biomolecules
to cause cell modification or death.3 Taking advantage of the
reactivity of singlet oxygen, especially its [2 + 2] cycloaddition
with double bonds to ultimately give carbonyl fragments via a
dioxetane intermediate,4 Breslow and co-workers have developed
photocleavable cyclodextrin carriers for photosensitizers.5 Using
the similar photosensitization-singlet oxygenation-dioxetane de-
composition trilogy, we envisioned that a “photodynamic” prodrug
system could be engineered to release drugs bearing carbonyl
functionalities.
In our system, a drug bearing a carbonyl group is incorporated
onto the periphery of a photosensitizer, by a linker, to furnish a
sensitizer-drug complex. The carbonyl group of the drug becomes
part of the double bond linkage. After systemic administration,
selective irradiation at the diseased site by visible light triggers the
photosensitizing moiety of the local complex to generate singlet
oxygen, which then migrates and oxidatively cleaves the olefin
linkage to release the drug (Scheme 1). Due to the fact that the [2
+ 2] cycloaddition generally competes ineffectively with the “ene”
reaction and the [4 + 2] cycloaddition in singlet oxygenation of
alkenes,6 our major challenge was the elaboration of an olefin
linkage to ensure the desired chemoselectivity during photooxy-
genation. Highly electron-rich alkenes, such as those heavily
substituted by hetero groups, have been reported to favor the [2 +
2] mode.7,8 Furthermore, hetero substituents tend to direct the attack
of the singlet oxygen to the side of the double bond, a phenomenon
known as the “cis-directing effect”,9 which can be used to secure
the [2 + 2] selectivity if that side of the olefin lacks an abstractable
allylic hydrogen. Based on these considerations, a 1,2-dihetero-
substituted alkene, preferably in Z-configuration, presents itself as
an ideal olefin linkage for our system. These species can be readily
generated from esters or amides by a recently reported alkoxy-
methylenation methodology using a titanium-carbene complex.10
In addition, Steglich esterification11 provides a facile and mild way
to attach the resulting prodrug module to photosensitizers bearing
a carboxyl group.
Scheme 2. Synthesis of Linker and Photosensitizer-Drug
Complexes
A linker was designed to have functional groups (A and B) at
either end of a spacer arm (Scheme 1). A is a dithioorthoformate
to alkoxymethylenate carbonyl-bearing drugs, while B is a silyl-
protected hydroxyl to esterify carboxyl-bearing photosensitizers.
Simple alkyl chains or complex structures like steroids can be used
as spacers to fine-tune the pharmacokinetic (PK) profile of the
complexes.
As proof-of-principle, linker 1 was readily prepared by a three-
step synthesis from 1,5-pentandiol (Scheme 2). Simple esters
(aliphatic, aromatic and lactone) and amides as drug mimics and
methyl esters of NSAIDs (ibuprofen and naproxen) were connected
with photosensitizers of a tetraphenylporphyrin monoacid derivative
(TPPAD) or benzophenylporphyrin monoacid derivative (BPAD,
verteporfin analogue) to give complexes 2 to 9, comprising both
Z- and E-isomers (Scheme 2).
Photoirradiation of the final complexes was carried out in NMR
tubes at room temperature using filtered visible light. Progress of
the reactions was monitored and quantified by NMR and GC with
internal standardization. Since dioxetanes decompose instantly at
the injection temperature (220 °C), GC offers the advantage of
showing the total amount of releasable drug from the complex. The
step-by-step NMR spectra of the photoirradiation show the rapid
decay of the starting complex, accompanied by the increase of the
released drug and the remaining dioxetane. Signals of the porphyrin
backbone are barely changed as the photooxygenation took place
9
4236
J. AM. CHEM. SOC. 2008, 130, 4236-4237
10.1021/ja800140g CCC: $40.75 © 2008 American Chemical Society

C O M M U N I C A T I O N S
Table 1. Visible-Light-Triggered Drug Release
lectivity of these reactions. Interestingly, the photooxygenation of
a Z/E ) 4:1 diastereomer mixture of 9 showed complete release of
amide from both Z- and E-isomers, although the latter was more
sluggish (entry 13). This could be due to the pronounced activating
effect of the enamino nitrogen. Our results indicated that a
Z-configured complex should be chosen for future clinical applica-
tions.
yield
by NMRb
(%)
yield
by GCc
(%)
concn
(mM)
time
entrya
complex
solvent
(min)
In conclusion, a proof-of-principle photodynamic prodrug system
has been established, which allows for rapid drug incorporation
and highly efficient drug release upon visible light irradiation. The
advantages of this system are multifold: (1) the external, light-
triggered activation would provide superior controls over the
location and the onset of drug release; (2) the system itself is
sufficiently flexible that both the linker and the photosensitizer can
be rationally modified or functionalized; (3) as shown here, well-
established photosensitizers such as verteporfin can be directly
utilized so that their PK and clinical profiles can serve as good
reference points for new photosensitizer-drug complexes; (4) since
esters and amides provide some of the most common prodrug
derivatives, our system provides a solution to deliver these prodrugs
site-specifically; (5) from a PDT perspective, bifunctional drugs
can be developed to have the complementary drug simultaneously
released at the site where PDT is performed, to bring about a
synergistic therapeutic effect.
1
2
3
2Z
2Z
2Z
2Z
2Z
3Z
4Z
5Z
6Z
7Z
8Z
2E
9f
C6D6
CDCl3
CDCl3/CD3OD ) 4:1
CD3COCD3
CD3COCD3
C6D6
C6D6
C6D6
C6D6
C6D6
7
3
4
4
5
4
3
3
8
60
1
3
2.5
6
7
3
93
91
93
>95
>95
>95
94
92
94
88
94
>95
>95
>95
>95
91
4
5d
6
3
7
8
9
10
11
12
13
10
15
4
8
7
90
>95
>95
>95
86
33
88
C6D6
CDCl3
C6D6
>95
35e
>95
2
8
1
5
a All the experiments were carried out at room temperature, and yields
are based on conversions g 95%. bYield of the total [2 + 2] cycloaddition
products. Yield of total releasable esters or amides. dDABCO (6 equiv)
c
e
f
was added. “ene” products were generated in 56% yield. A Z/E ) 4 :1
mixture was used as starting materials.
exclusively on the side chain, leaving the chromophore intact. The
dioxetane intermediate was confirmed by ESI-MS.
Acknowledgment. We thank the Natural Sciences and Engi-
neering Research Council of Canada (NSERC) and QLT Inc. for
their financial support.
The high yields of the visible-light-triggered drug release were
solvent-independent, as evident by the photooxygenation of 2Z in
solvents of varied polarity (Table 1, entries 1 to 4). The involvement
of singlet oxygen in drug release was confirmed as the progression
of the photooxygenation of 2Z in deuterated acetone was severely
hampered when 1,4-diazabicyclo[2.2.2]octane (DABCO), a singlet
oxygen quencher, was added (entry 5). Quantitative and rapid
releases of ethyl benzoate, δ-valerolactone, methyl esters of
ibuprofen, and naproxen were observed in the photooxygenation
of 3Z to 8Z (entries 6 to 11). It is remarkable that the competing
“ene” reaction and the [4 + 2] cycloaddition were completely
suppressed in these reactions. The equal efficiency of drug release
from either TPPAD or BPAD based complexes indicated the choice
of photosensitizer can be flexible. It is noteworthy that even though
some dioxetane intermediates persisted immediately after the end
of the photooxygenation, they completely decomposed within hours
at room temperature. In fact, the simultaneous and complete
dioxetane cleavage for drug release in ViVo would be expected,
due to the ubiquitous presence of amines and trace metals in human
tissues, which are reported to catalyze the dioxetane decomposi-
tion.12 Unfortunately, the “ene” reaction predominated the photo-
oxygenation of 2E, giving the [2 + 2] cycloaddition products in
only 35% yield (entry 12), which is in direct contrast to the complete
[2 + 2] selectivity of the precedent singlet oxygenation of some
E-configurated enediol ethers.7 The dramatic reactivity difference
between 2Z and 2E is attributed to the different results of the cis-
directing effect of the alkoxy substituents: for the Z-isomer, two
same-side alkoxy groups “lock” the attack of singlet oxygen to their
side of the double bond as expected, while, for the E-isomer, each
side of the double bond has an alkoxy group and the extra alkyl
group at the disubstituted side provides an additional directing effect
for the incoming singlet oxygen, leading to the predominant “ene”
reaction. Clearly, the HOMO-LUMO interaction between singlet
oxygen and the olefin substitutents played a more decisive role
than the electron richness of the olefin in determining the chemose-
Supporting Information Available: Experimental details and
compound characterization data. This material is available free of charge
via the Internet at http://pubs.acs.org.
References
(1) Ja¨rvinen, T.; Rautio, J.; Masson, M.; Loftsson, T. In Drug DiscoVery
Handbook; Gad, S. C., Ed.; Wiley-Interscience: Hoboken, NJ, 2005; pp
733.
(2) Reinhard, R.; Schmidt, B. F. J. Org. Chem. 1998, 63 (8), 2434. McCoy,
C. P.; Rooney, C.; Edwards, C. R.; Jones, D. S.; Gorman, S. P. J. Am.
Chem. Soc. 2007, 129 (31), 9572.
(3) Bonnett, R. Chemical Aspects of Photodynamic Therapy; Gordon &
Breach: Australia, 2000. Sternberg, E. D.; Dolphin, D.; Bruckner, C.
Tetrahedron 1998, 54 (17), 4151.
(4) Schaap, A. P.; Zaklika, K. A. In Singlet Oxygen; Wasserman, H. H.,
Murray, R. W., Eds.; Academic Press: New York, 1979; p 173.
(5) Ruebner, A.; Yang, Z. W.; Leung, D.; Breslow, R. Proc. Natl. Acad. Sci.
1999, 96 (26), 14692. Baugh, S. D. P.; Yang, Z. W.; Leung, D. K.; Wilson,
D. M.; Breslow, R. J. Am. Chem. Soc. 2001, 123 (50), 12488.
(6) Gollnick, K.; Kuhn, H. J. In Singlet Oxygen; Wasserman, H. H., Murray,
R. W., Eds.; Academic Press: New York, 1979; p 287. Bloodworth, A.
J.; Eggelte, H. J. In Singlet O2, II. Reaction Modes and Products; Frimer,
A. A., Ed.; CRC Press: Boca Raton, FL, 1985; p 93.
(7) Bartlett, P. D.; Schaap, A. P. J. Am. Chem. Soc. 1970, 92 (10), 3223.
Bartlett, P. D.; Mendenhall, G. D.; Schaap, A. P. Ann. N. Y. Acad. Sci.
1970, 171 (1), 79. Rio, G.; Berthelo, J. Bull. Soc. Chim. Fr. 1971, (10),
3555. Adam, W.; Kades, E.; Wang, X. H. Tetrahedron Lett. 1990, 31
(16), 2259.
(8) Foote, C. S.; Dzakpasu, A. A.; Lin, J. W. P. Tetrahedron Lett. 1975, (14),
1247; Adam, W.; Liu, J. C. J. Chem. Soc., Chem. Comm. 1972, (2), 73.
Adam, W.; Liu, J. C. J. Am. Chem. Soc. 1972, 94 (4), 1206.
(9) Inagaki, S.; Fujimoto, H.; Fukui, K. Chem. Lett. 1976, (7), 749. Rousseau,
G.; Leperchec, P.; Conia, J. M. Tetrahedron 1978, 34 (23), 3483.
Orfanopoulos, M.; Grdina, S. M. B.; Stephenson, L. M. J. Am. Chem.
Soc. 1979, 101 (1), 275. Matsumoto, M.; Kobayashi, H.; Matsubara, J.;
Watanabe, N.; Yamashita, S.; Oguma, D.; Kitano, Y.; Ikawa, H.
Tetrahedron Lett. 1996, 37 (3), 397. Stratakis, M.; Orfanopoulos, M.
Tetrahedron 2000, 56 (12), 1595; Adam, W.; Bosio, S. G.; Turro, N. J.
J. Am. Chem. Soc. 2002, 124 (47), 14004.
(10) Takeda, T.; Sato, K.; Tsubouchi, A. Synthesis 2004, (9), 1457.
(11) Neises, B.; Steglich, W. Angew. Chem., Int. Ed. Engl. 1978, 17 (7), 522.
(12) Bartlett, P. D.; Landis, M. E. In Singlet Oxygen; Wasserman, H. H.,
Murray, R. W., Eds.; Academic Press: New York, 1979; p 243.
JA800140G
9
J. AM. CHEM. SOC. VOL. 130, NO. 13, 2008 4237