Triggering DNAzymes with light: A photoactive C8 thioether-linked adenosine

Article abstract of DOI:10.1021/ja046964y

Herein we report evidence for a light-inducible DNAzyme. In so doing, we also disclose the synthesis and photochemical properties of a novel nucleoside: 8-(2-(4-imidazolyl)ethyl-1-thio)-2′-deoxyriboadenosine (d1). The light sensitivity of (d1) was evaluated via an examination of the photoinduced reactivation of DNAzyme 8-17E from an inactive form that contained a single nucleotide (d1) modification. Restoration of DNAzyme activity results from a photoinduced reversion of (d1) to unmodified deoxyadenosine. Deuterium studies indicate that water is the source of hydrogen in the C8-H product and not the alkylthio group, suggesting that reversion of (1) to adenosine is not a consequence of simple homolysis of the C8-S bond but of an unprecedented photochemical conversion. This adenosine, which affords significant control of catalytic reactivation of a DNAzyme, may find general use in photodecaging other biological systems. Copyright

Full text of DOI:10.1021/ja046964y

Published on Web 09/16/2004
Triggering DNAzymes with Light: A Photoactive C8 Thioether-Linked
Adenosine
Richard Ting, Leonard Lermer, and David M. Perrin*
Department of Chemistry, The UniVersity of British Columbia, VancouVer, BC, V6T-1Z1, Canada
Received May 23, 2004; E-mail: dperrin@chem.ubc.ca
Light provides an effective means of willfully inducing reactivity
to study kinetically complex biological processes and to localize
drug action for photodynamic therapy. In most cases, photoacti-
vation is a consequence of irreversible photodeprotection to cleanly
unleash the biologically active molecule of interest. Examples in-
clude substrates such as ATP,1a GABA,1b estradiol,1c N-ras-pep-
tides,1d and an mRNA target for ribozymes,1e,f allosteric ligands
such as cAMP1g and inositol triphosphate,1h toxins such as ricin,1i
antibodies to protein-A,1j and enzymes such as thrombin1k,l and
RNaseA.1m
Of the many classes of biologically relevant effectors and
catalysts, catalytic nucleic acids offer unique properties for control-
ling gene expression by catalyzing highly efficient M2+-dependent
sequence-specific RNA cleavage.2 To achieve such control, at least
two schemes might be envisioned. Covalent attachment of a light-
responsive group could (a) constrain the 5′ and 3′ guide sequences
to perturb substrate binding3 or (b) perturb the chemical/confor-
mational nature of the interceding “active site” sequence to impede
catalysis. Veiling the active site of a protein enzyme is achieved
by covalently attaching a photoactive group (e.g., nitrobenzyl,
orthocinnamate) to a nucleophile within the active site, wherein
there is often considerable choice with respect to locus and
reactivity. In contrast to proteins, the “active site” of an unmodified
DNAzyme, absent a 2’OH, affords only the nucleobase amines and
carbonyls for covalent protection.
tion, and (c) noted stabilities of both the C8 purinyl radical and
anion10 that might favor either homo- or heterolysis of the C-S
thioether to yield adenosine (vide infra).
Figure 1. Left: (1 or d1). Right: 8-17E with four singly modified dAs
(1-4). Bold A1 shows inactive substitution. Top: RNA substrate.
The 8-17E DNAzyme,11 which exhibits robust Zn2+-de-
pendent RNase activity, was chosen to test this strategy; each
deoxyriboadenosine was replaced with d1 to give four variants
(8-17E-A1-4) in anticipation that at least one would be completely
devoid of catalytic activity.12 Of these modified variants, A4 and
A3 remained active, A2 exhibited 50% reduced activity, and A1
was utterly inactive (see Supporting Information). This activity
profile confirms a recent report that identified A1 in the stem loop
to be essential for activity.13 Irradiation of 8-17E-A1 (0.3 µM) in
degassed buffer for 8 min with a 10 mW Xe-Hg arc (band-pass
filter cutoff > 310 nm) restored ∼30% activity of 8-17E (Figure
2). Irradiation with a 254 nm UV lamp (∼1 mW) or with a 8 mM
dye laser at 283 nm for 10 minutes gave similar results.
With regard to oligonucleotides, a few reports describe photo-
deprotection of the exocyclic nucleobase amines protected with a
nitroveratrylcarbonyl.4 Irradiation promotes a cascade of elimina-
tions to yield a nitrosobenzaldehyde and an N-carboxynucleoside,
which presumably decarboxylates without detection or report.5
Deprotection yields were variably high upon extended irradiation
times: 25-240 min. More rapid deprotection (5-25 min irradia-
tion) was observed with 2,2′-bis(dinitrophenylethoxy)carbonyls.
Such deprotection would suggest a useful, yet thus far unexplored,
means for unveiling a nucleic acid catalyst.
Our work on modified DNAzymes6 led us to develop an
adenosine analogue, 8-(2-(4-imidazolyl)ethyl-1-thio)-2′-deoxyribo
adenosine (d1, X ) H) and its corresponding phosphoramidite
(Figure 1), with an unprecedented photoactivity (λ-max ) 279 (
1 nm) for unveiling DNAzyme activity. Herein, the development
of a photoactive DNAzyme exploits the weak carbon-sulfur bond
(∼68 kcal/mol), which is energetically on par with carbon-halogen
bonds and which could be expected to display photoactivity
reminiscent of 6-thiopurines7 and arylthioethers.8 In contrast to
substitution with a monovalent halogen, substitution with divalent
sulfur permitted introduction of additional chemical functionality,
in this case an ethylimidazole that could perturb the M2+ dependence
of DNAzymes. The C8 locus was substituted in recognition of (a)
the synthetic ease of displacement on the 8-bromo precursor by an
alkylsulfide, (b) potential for an altered syn-anti preference around
the C1′-N9 purine bond9 that would perturb DNAzyme conforma-
Figure 2. RNA Cleavage activity 8-17E-A1 before and after 8-min
irradiation. RNA cleavage analyzed 5-60 min.
Single turnover kinetics on irradiated 8-17E-A1 returned a kcat
value of 0.113 min-1 that was indistinguishable from that of 0.096
min-1 obtained for unmodified 8-17E, which was irradiated to
control for UV damage (see Supporting Information). MALDI-TOF
analysis of 8-17E-A1 irradiated within the 3-HPA/citrate comatrix
indicated that d1 had converted to dA and not other plausible
products of intermediate steric constraint such as 8-oxo-dA or
8-mercapto-dA (see Supporting Information).
To further characterize this reaction in terms of the final purine
product, a ribonucleoside 1 (X ) OH) was prepared, and its
1
decomposition was characterized by UV-vis, H NMR, and ESI
spectroscopy. A solution of 1 in degassed CH3OH/H2O was
irradiated at 280 nm with a 10 mW dye laser. Within 8 min (t1/2
≈
2 min), the λ-max had totally shifted to ∼260 nm, a characteristic
signature of adenosine (Figure 3). This spectral shift supports the
hypothesis that 1 reverted to adenosine and not to 8-oxo- and
8-mercaptoadenosine, which exhibit λ-maxima at 270 and 310 nm,
respectively.14 Indeed, a small shoulder at 310 nm transiently ap-
9
12720
J. AM. CHEM. SOC. 2004, 126, 12720-12721
10.1021/ja046964y CCC: $27.50 © 2004 American Chemical Society

C O M M U N I C A T I O N S
peared, suggesting the possibility of a discrete intermediate. Fol-
lowing irradiation in CH3OH, ESI analysis of the crude gave one
peak MH+ of A: found/calculated 268.1 (Supporting Information),
which substantiated reversion to adenosine as initially hypoth-
esized. 1H NMR also indicated a reversion to A (Supporting Infor-
mation). This degradation raises mechanistic questions regarding
C-S bond cleavage.
study that should identify the fugitive alkylthio fragment and full
mechanism of fragmentation. The application of this chromophore
is highlighted by the generation of a DNAzyme that is photoacti-
vated over a period of minutes at relatively low wattage compared
to most photodeprotection procedures. The ease at which this anion
is generated may find practical synthetic utility. Furthermore, this
study suggests a “post-selection” synthetic strategy for conferring
photoactivity on other DNAzymes that should be readily applicable
to ribozymes where 1 may be incorporated to distinguish slow fold-
ing from fast cleavage. In addition, d1 may find utility in generating
an anion on DNA for electron transport studies. Phosphates of 1
may find use in the study of adenosine processing enzymes (e.g.,
kinases, polymerases, cAMP-diesterases). Finally, the purine may
be used in “capture and release” linkers for proteomics.18
Acknowledgment. We thank Mr. Qichi Hu for assistance with
laser work. R.T. was supported by a Gladys Estella Laird post-
graduate fellowship. D.M.P. is the recipient of a Michael Smith
Junior Career Scholar Award. Additional support came from
NSERC, UBC startup funds, and CFI.
Figure 3. Time course of UV irradiation at 280 nm of 40 µM 1.
Supporting Information Available: Synthetic and kinetic protocols
and spectral data. This material is available free of charge via the
Internet at http://pubs.acs.org.
Whereas both homolysis and heterolysis have been proposed for
arylthioether fission, homolysis generally predominates.7 With no
precedent for this thiopurine fragmentation and only scant data in
general for thioether photolyses, we considered both homo- and
heterolysis. Homolysis would generate a C8 radical that would ab-
stract an H-atom from a suitable carbon, of which the likely can-
didate would be a departing imidazolylethylthiyl radical. Heterolysis
would normally generate a sulfide and a C8 carbocation that would
quench to give 8-oxoadenosine (unobserved). Heterolysis with in-
verse electronic demand or photoinduced SET into the purine15a,b
followed by C-S homolysis would ultimately yield an unstable
sulfenic acid and a σ-anion at C8 that would protonate in water to
give adenosine. N7 protonation could stabilize this anion as an ylid
or as a carbene (Scheme 1).
References
(1) (a) Kaplan, J. H.; Forbush, B.; Hoffman, J. F. Biochemistry 1978, 17,
1929-1935. (b) Weiboldt, R.; Ramesh, D.; Carpenter, B. K.; Hess, G. P.
Biochemistry 1994, 33, 1526-1533. (c) Cruz, F. G.; Koh, J. T.; Link, K.
H. J. Am. Chem. Soc. 2000, 122, 8777-8778. (d) Volkert, M.; Uwai, K.;
Tebbe, A.; Popkirova, B.; Wagner, M.; Kuhlmann, J.; Waldmann, H. J.
Am. Chem. Soc. 2003, 125, 12749-12758. (e) Chaulk, S. G.; MacMillan,
A. M. Angew. Chem., Int. Ed. 2001, 40, 2149-2152. (f) Chaulk, S. G.;
MacMillan, A. M. Nucleic Acids Res. 1998, 13, 3173-3178. (g) Nargeot,
J.; Nerbonne, J. M.; Engels, J.; Lester, H. A. Proc. Natl. Acad. Sci. U.S.A.
1983, 80, 2395-2399. (h) Walker, J. W.; Feeney, J.; Trentham, D. R.
Biochemistry 1989, 28, 3272-3280. (i) Goldmacher, V. S.; Senter, P.
D.; Lambert, J. M.; Blattler, W. A. Bioconjugate Chem. 1992, 3, 104-
107. (j) Singh, A. K.; Khade, P. K. Bioconjugate Chem. 2002, 13, 1286-
1291. (k) Turner, A. D.; Pizzo, S. V.; Rozakis, G.; Porter N. A. J. Am.
Chem. Soc. 1988, 110, 244-250. (l) Arroyo, J. G.; Jones, P. B.; Porter,
N. A.; Hatchel, D. L. Thromb. Haemostasis 1997, 78, 791-793. (m)
Hiraoka, T.; Hamachi, I. Bioorg. Med. Chem. Lett. 2003, 13, 13-15.
(2) Kurreck, J.; Bieber, B.; Jahnel, R.; Erdmann, V. A. J. Biol. Chem. 2002,
277, 7099-7107.
Scheme 1. Photoinduced SET Followed by Homolysis of the C-S
Bonda
(3) The reader’s attention is directed to: Liu, Y.; Sen, D. J. Mol. Biol.,
published online July 28, 2004, doi: 10.1016/j.jmb.2004.06.060.
(4) Alverez, K.; Vasseur, J.-J.; Beltran, T.; Imbach, J.-L. J. Org. Chem. 1999,
64, 6319-6328.
a Shown is the carbanion form. Alkyl fragment is unknown.
(5) Il’ichev, Y. V.; Schworer, M. A.; Wirz, J. J. Am. Chem. Soc. 2004, 126,
4581-4595.
Since each mechanism stipulates different sources of hydrogen
at C8, deuterium labeling would differentiate the two. A quantity
of 25 nmol 1 (40 µM) was irradiated in either 1:1 D2O/CD3OD or
pure D2O, diluted in CH3OH, lyophilized, and resuspended in 1
mL of CH3OH to protonate the exchange-labile ND2 and 2′,3′,5′-
ODs. In both cases, ESI+ mode gave a single peak (MH+ 269.1118
( 0.1) with negligible values at 270 or higher, indicating stable
(6) Lermer, L.; Roupioz, Y.; Ting, R.; Perrin, D. M. J. Am. Chem. Soc. 2002,
124, 9960-9961.
(7) (a) Nair, V.; Richardson, S. G.; Coffman, R. E. J. Org. Chem. 1982, 47,
4520-4524. (b) Fleming, S. A.; Rawlins, D. B.; Samano, V.; Robins, M.
J. J. Org. Chem. 1992, 57, 5968-5976.
(8) (a) Fleming, S. A.; Jensen, A. W. J. Org. Chem. 1993, 58, 7135-7137.
(b) Cheng C.; Stock L. M. J. Org. Chem. 1991, 56, 2436-2443.
(9) Neidle, S.; Sanderson, M. R.; Subbiah, A.; Chattopadhyaya, J. B.; Kuroda,
R.; Reese, C. B. Biochim. Biophys. Acta 1979, 565, 379-386.
(10) (a) Zady, M. F.; Wong, J. L. J. Org. Chem. 1979, 44, 1450-1454. (b)
Leonard, N. J.; Bryant, J. D. J. Org. Chem. 1979, 44, 4612-4616.
(11) (a) Santoro, S. W.; Joyce, G. F. Proc. Natl. Acad. Sci. U.S.A. 1997, 94,
4262-4266. (b) Li, J.; Zheng, W. C.; Kwon, A. H.; Lu Y. Nucleic Acids
Res. 2000, 28, 481-488.
1
monodeuteration. Following irradiation of 1 in CD3OD, H NMR
showed loss of H8 (Supporting Information). H/D exchange on C8
under ambient conditions and upon photoexcitation is negligible
and cannot explain these data.16 As H8 derives from water (or
solvent) and not carbon, an intermediate C8 anion is favored over
a C8 radical, which would otherwise have to abstract a hydrogen
atom from water giving a high-energy hydroxyl radical.
In conclusion, the salient issues in this communication include
the disclosure of a photoactive alkylthioadenosine analogue that
undergoes rapid and high-yielding cleavage, resulting in reversion
to adenosine by what appears to be a novel mechanism. 8-Thioether-
linked purines have been but minimally explored, and this report
is the first to define the photochemical properties thereof.17 Other
thioether derivatives, including 8-ethylthio-adenosine, were equally
reactive, indicating that the imidazole was not responsible for this
photoactivity (data not shown) and will be described in a future
(12) 8-17E variants here were synthesized on CPG-solid phase, deprotected,
and purified. Activity was assayed in either 5 mM Mg2+ at pH 7.5 or 100
µM Zn2+ at pH 6.8, per ref 11a,b (Supporting Information).
(13) Cruz, R. P. G.; Withers, J. B.; Li, Y. F. Chem. Biol. 2004, 11, 57-67.
(14) (a) Bodepudi, V.; Shibutani, S.; Johnson, F. Chem. Res. Toxicol. 1992, 5,
608-617. (b) Shuman, D. A.; Bloch, A.; Robins, R. K.; Robins, M. J. J.
Med. Chem. 1969, 12, 653-657.
(15) (a) Shima, K.; Sazaki, A.; Nakabayashi, K.; Yasuda, M. Bull. Chem. Soc.
Jpn. 1992, 65, 1472-1474. (b) Marciniak, B.; Hug, G. L.; Bobrowski,
K.; Kozubek, H. J. Phys. Chem. 1995, 99, 13560-13568.
(16) Shelton, K. R.; Clark, J. M. Biochemistry 1967, 6, 2735-2739.
(17) (a) Gendron, F.-P.; Halbfinger, E.; Fischer, B.; Duval, M.; D’Orleans-
Juste, P.; Beaudoin, A. R. J. Med. Chem. 2000, 43, 2239-2247. (b) He,
H. Z.; Llauger, L.; Rosen, N.; Chiosis, G. J. Org. Chem. 2004, 69, 3230-
3232.
(18) Bottari, P.; Aebersold, R.; Turecek, F.; Gelb, M. H. Bioconjugate Chem.
2004, 15, 380-388.
JA046964Y
9
J. AM. CHEM. SOC. VOL. 126, NO. 40, 2004 12721