Facile synthesis of a fluorescent deoxycytidine analogue suitable for probing the RecA nucleoprotein filament

Article abstract of DOI:10.1021/ol0167863

equation presented We report the synthesis of the fluorescent 2′-deoxycytidine analogue 5-methylpyrimidin-2-one nucleoside, its incorporation at three specified sites in a single 60-nucleotide DNA molecule, and the use of its total and polarized intrinsic

Full text of DOI:10.1021/ol0167863

ORGANIC
LETTERS
2001
Vol. 3, No. 24
3919-3922
Facile Synthesis of a Fluorescent
Deoxycytidine Analogue Suitable for
Probing the RecA Nucleoprotein
Filament
Scott F. Singleton,* Feng Shan, Matthew W. Kanan, Catherine M. McIntosh,
Chad J. Stearman, Jeremiah S. Helm, and Kristofor J. Webb
Department of Chemistry and Department of Biochemistry and Cell Biology,
Rice UniVersity, P.O. Box 1892 MS #60, Houston, Texas 77251-1892
sfs@rice.edu
Received September 20, 2001
ABSTRACT
We report the synthesis of the fluorescent 2′-deoxycytidine analogue 5-methylpyrimidin-2-one nucleoside, its incorporation at three specified
sites in a single 60-nucleotide DNA molecule, and the use of its total and polarized intrinsic fluorescence to characterize RecA−DNA complexes.
The repair of damaged DNA is essential to the maintenance
of heritable genetic information. The Escherichia coli RecA
protein plays indispensable roles in the detection of and
response to genomic DNA damage, and its study has
provided a paradigm for understanding recombinational DNA
repair in all phylogenetic kingdoms.1 The initiating signal
in vivo is single-stranded DNA (ssDNA) which is the
substrate upon which a RecA filament forms (Figure 1). A
combination of in vivo and in vitro studies suggests that this
nucleoprotein filament is essential for directing subsequent
DNA repair. To address fundamental issues concerning the
initiation of DNA repair processes, the key RecA-ssDNA
intermediate must be understood.
filaments to be probed directly. To our knowledge, this is
the first demonstration that such intrinsically fluorescent
We report the synthesis and characterization of oligo-
deoxyribonucleotides (ODNs) containing an intrinsically
fluorescent analogue of 2′-deoxycytidine (5-methylpyrimdin-
2-one nucleoside, m5K) incoroporated at multiple positions.
The total and polarized fluorescence emission signals of these
ODNs allow RecA-single-stranded-DNA (RecA-ssDNA)
Figure 1. Cartoons depicting the collapsed and extended RecA-
ssDNA complexes formed in the absence and presence of ATPγS,
respectively.
(1) Roca, A. I.; Cox, M. M. Prog. Nucleic Acid Res. Mol. Biol. 1997,
56, 129-223.
10.1021/ol0167863 CCC: $20.00 © 2001 American Chemical Society
Published on Web 10/26/2001

synthetic DNAs can be used to characterize definitively the
assembly of the RecA-ssDNA complexes that initiate DNA
repair responses.
The report11 of ODNs comprised exclusively of pyrimidin-
2-one nucleotides led us to consider the properties and
synthesis of m5K.
Nucleoside m5K possesses high intrinsic fluorescence12
and is, formally, a conservatively modified derivative of 2′-
deoxycytidine (Scheme 1). The fluorescence properties of
A wealth of low-resolution structural information indicates
that the helical RecA nucleoprotein filament in the absence
of a cofactor (or with ADP) is “collapsed” (Figure 1), while
an “extended” filament (50% increase in the helical pitch)
is observed with nonhydrolyzable analogues of ATP.2 In
addition, the RecA protein filament comprises at least two
distinct DNA binding sites, each of which can accommodate
3-4 nucleotides (nts) per RecA monomer. The active, repair-
initiating filament contains ssDNA bound at only one site
(site I) with an overall DNA-protein stoichiometry of 3 nts
per monomer. At a minimum, any solution-phase method to
analyze the filaments must recapitulate these observations.
Biochemical assays such as those based on the RecA ATP
hydrolysis activity have been traditionally used to study the
RecA‚ATP‚DNA filament. However, it is not possible to
study the collapsed filament by such a method. Thus, there
are few definitive structural and functional studies which
directly compare the collapsed and extended RecA filaments.
Spectrofluorometry provides a powerful method for moni-
toring RecA-DNA filaments in real time. Unfortunately, the
natural DNA bases are only minimally luminescent3 and the
RecA tryptophans do not serve as reporters for DNA
binding.4 The use of extrinsic fluorophores such as fluores-
cein is not optimal because of the potential for the reporter
moiety to influence the interactions under study,5 leading to
an abnormal DNA-protein stoichiometry.6 Moreover, poly-
nucleotides containing 1,N6-ethenoadenine (ꢀA), one of the
most commonly used fluorsecent base replacements, show
a stoichiometry of 1:6 resulting from two DNA molecules
binding RecA.7 However, the natural bases of DNA can be
replaced with isomorphic fluorescent analogues. In particular,
the adenine analogue 2-aminopurine (2AP) has been used
in an extensive variety of biochemical experiments. Never-
theless, DNA-containing single-site 2AP substitutions was
not suited to the quanitative characterization of RecA-DNA
complexes.8,9 When attempting to overcome this problem by
using 2AP at multiple positions within a single ODN, we
have observed the significant accumulation of non-full-length
oligomers during both synthesis and storage, presumably as
a result of the hydrolytic proclivity of the 2AP heterocycle.10
Scheme 1. Structures of m5K and ODNs
the 5-methylpyrimdin-2-one heterocycle are sensitive to its
microenvironment and can serve as as a reporter for DNA
structural dynamics.12 On the basis of emission changes
previously described for ꢀA-labeled DNA,13 we anticipated
that RecA binding to ssDNA containing m5K would produce
an emission increase. Importantly, this nucleoside analogue
is synthetically accessible from thymidine (dT).
The first syntheses of pyrimindin-2-one nucleosides were
effected in low yield by direct reduction of the corresponding
thymidine or uridine derivative with sodium amalgam.14 The
approach was later improved, and the synthesis of m5K was
achieved via oxidation of 4-hydrazinopyrimidinone nucleo-
sides and rearrangement (with extrusion of dinitrogen) of
the intervening diazene.15 In our hands, the adaptation of
the latter approach to larger scale provided several difficulties
stemming from the need for robust protecting groups during
the oxidation of 4-hydrazinopyrimidinone by Ag2O in
refluxing aqueous dioxane (or EtOH). Thus, we were led to
investigate alternative routes inspired by the mild, homoge-
neous oxidation of monosubstituted hydrazines.16 The op-
portunity for further improvement was afforded by the
reported ability of 4-triazolopyrimidinone nucleosides related
(2) Egelman, E. H. J. Mol. Biol. 2001, 309, 539-542.
(3) Daniels, M.; Hauswirth, W. Science 1971, 171, 675-677.
(4) Dombroski, D. F.; Scraba, D. G.; Bradley, R. D.; Morgan, A. R.
Nucleic Acids Res. 1983, 11, 7487-7504. Morrical, S. W.; Lee, J.; Cox,
M. M. Biochemistry 1986, 25, 1482-1494. Eriksson, S.; Norde´n, B.;
Takahashi, M. J. Biol. Chem. 1993, 268, 1805-1810.
(5) Volodin, A. A.; Smirnova, H. A.; Bocharova, T. N. FEBS Lett. 1994,
349, 65-68.
(10) Fujimoto, J.; Nuesca, Z.; Mazurek, M.; Sowers, L. C. Nucleic Acids
Res. 1996, 24, 754-759.
(11) Zhou, Y.; Ts'o, P. O. P. Nucleic Acids Res. 1996, 24, 2652-2659.
(12) Wu, P.; Nordlund, T. M.; Gildea, B.; McLaughlin, L. W. Biochem-
istry 1990, 29, 6508-6514.
(6) Gourves, A. S.; Defais, M.; Johnson, N. P. J. Biol. Chem. 2001, 276,
9613-9619.
(7) Menetski, J. P.; Kowalczykowski, S. C. J. Mol. Biol. 1985, 181, 281-
295. Zlotnick, A.; Mitchell, R. S.; Steed, R. K.; Brenner, S. L. J. Biol.
Chem. 1993, 268, 22525-22530.
(13) ꢀA-Labeled DNA shows the following trend in emission intensi-
ties: free DNA < RecA-DNA < RecA-ATPγS-DNA (Cazenave, C.;
Toulme´, J. J.; He´le`ne, C. EMBO J. 1983, 2, 2247-2251).
(14) Laland, S. G.; Hanssen, G. S. Biochem. J. 1964, 90, 76-81.
Helgeland, L.; Tipson, R. S. Biochim. Biophys. Acta 1964, 87, 353-355.
(15) (a) Gildea, B.; McLaughlin, L. W. Nucleic Acids Res. 1989, 17,
2261-2281. (b) Connolly, B. A.; Newman, P. C. Nucleic Acids Res. 1989,
17, 4957-4974.
(8) 2AP is characterized by a relatively low emission quantum yield and
excitation and emission spectra that substantially overlap those of RecA.
These limitations are exacerbated by the fact that RecA protein also causes
a significant background problem from light scattering.
(9) A single 2AP substitution in dsDNA can cause local structural and
dynamic perturbations (Nordlund, T. M.; Xu, D.; Andersson, S.; Nilsson,
L.; Rigler, R.; Gra¨slund, A.; McLaughlin, L. W.; Gildea, B. Proc. SPIE
1990, 1204, 344-353).
(16) See: Myers, A. G.; Finney, N. S.; Kuo, E. Y. Tetrahedron Lett.
1989, 30, 5747-5750, and references therein.
3920
Org. Lett., Vol. 3, No. 24, 2001

Synthesis of a monomer suitable for automated, solid-
phase DNA synthesis was initiated by protection of the 5′-
hydroxyl of m5K as the 4,4′-dimethoxytrityl (DMT) ether.
5 was treated with 2-cyanoethyldiisopropylphosphorochloridite
in the presence of Hu¨nig’s base in dry methylene chloride
to give the desired final phosphoramidite 6. This monomer
was incorporated into ODNs 30a and 60a using standard
automated DNA synthesis protocols in quantitative yield.
Following deprotection and purification, analysis of nucleo-
side composition, capillary electropherograms, and MALDI-
TOF mass spectrometric data confirmed that bona fide m5K
was present in the desired position(s) within each ODN.
Scheme 2a
The RecA-DNA complexes described here involve a
60mer ODN containing nucleoside m5K at three positions
(ODN 60a, Scheme 1).19 The emission of the fluorophore
within the ODN is quenched ca. 80% relative to the free
nucleoside,20 presumably due to base stacking interactions.12
An enhancement of the emission is observed upon RecA
addition to ODN 60a in the absence of ATPγS (Figure 2),
a (a) TESCl, pyridine; (b) 1,2,4-triazole, Et3N, POCl3, CH3CN;
(c) NH2NH2, 1,4-dioxane; (d) PTAD, 1,4-dioxane; (e) AcOH/THF/
H2O (4/4/1, v/v/v); (f) DMTCl, pyridine; (g) â-cyanoethyldiiso-
propylphosphorochloridite, DIPEA, CH2Cl2.
to 2 (Scheme 2) to undergo facile nucleophilic addition-
elimination reactions.17 Thus, we envisioned that an improved
synthesis of m5K would proceed via the intermediacy of 2,
followed by the displacement of triazole with hydrazine and
mild, homogeneous oxidation of the resulting 4-hydrazino-
pyrimidinone nucleoside 3.
Figure 2. Normalized emission intensities (Iem/Iem°) and fluores-
cence anisotropies (r ) (I - I )/(I + 2I )) collected for 60a alone
||
||
(ODN) or in the presence of either RecA (+ RecA) or RecA and
ATPγS (+ RecA + ATPγS). Fluorescence data were collected
using 50 nM 60a and excitation and emission wavelengths of 315
and 385 nm, respectively, with spectral bandwidths of 4 nm. See
Supporting Information for details.
The requisite 4-triazolopyrimidinone nucleoside (2), ap-
propriately protected as the di-O-triethylsilyl ether, is avail-
able in 65% yield from dT following etherification of the
dT sugar hydoxyls and subsequent transformation at O4.17
The triazole group of 2 was directly substituted by hydrazine
under mild conditions, and the intermediate (3) was oxidized
using N-phenyltriazolinedione (PTAD) without purification
to give the protected nucleoside 4 in greater than 50% yield
in two steps from 2. The triethylsilyl groups were readily
removed by AcOH in aqueous THF to afford fluorescent
nucleoside m5K in 81% yield. This synthetic approach
provides an efficient method for large-scale production of
m5K. Importantly, the key intermediacy of 2 opens up the
possibility that other milder oxidizing agents might be
employed. Ultimately, ODNs containing 4-triazolopyrimi-
dinone could be converted to 5-methyl-2-pyrimidinone on
the solid support following automated solid-phase synthesis
by using mild displacement-oxidation conditions.18
providing a direct spectroscopic measure of RecA-DNA
complex formation. In contrast, there is no signal change
when RecA binds nonfluorsecent ODN 60b, establishing the
absence of interference from the tryptophan fluorescence
background. Although we anticipated that the DNA fluo-
rescence would change upon association with the RecA‚
ATPγS filament, the emission intensities of the complexes
(18) For examples of this “convertible nucleoside” approach, see ref 17
as well as the following: MacMillan, A. M.; Verdine, G. L. J. Org. Chem.
1990, 55, 5931-5933. Gao, H.; Fathi, R.; Gaffney, B. L.; Goswami, B.;
Kung, P.-P.; Rhee, Y.; Jin, R.; Jones, R. A. J. Org. Chem. 1992, 57, 6954-
6959. Kim, S. J.; Stone, M. P.; Harris, C. M.; Harris, T. M. J. Am. Chem.
Soc. 1992, 114, 5480-5481.
(19) The ODN sequences used here were derived from those capable of
exhibiting RecA-DNA pairing activitiy in the presence of ATPγS
(Podyminogin, M. A.; Meyer, R. B.; Gamper, H. B. Biochemistry 1995,
34, 13098-13108). As a control we performed a spectroscopic ATP
hydrolysis assay to show that these ODNs were competent for activitating
the RecA protein (details are available in the Supporting Information).
(20) Details and graphics are available in the Supporting Information.
(17) Xu, Y.-Z.; Zheng, Q.; Swann, P. F. J. Org. Chem. 1992, 57, 3839-
3845.
Org. Lett., Vol. 3, No. 24, 2001
3921

in the presence of ATPγS are not significantly different from
those of the complexes formed in its absence. Thus, for these
particular DNA sequences, the steady-state fluorescence
signals of m5K display only modest differences between the
collapsed and extended RecA-DNA complexes. The lack
of a diagnostic difference in fluorescence intensity between
the collapsed and extended RecA-DNA complexes led us
to explore other fluorescence observables. The binding of
RecA protein to the ODNs results in an increase in the
polarized emission21 from m5K (Figure 2). Importantly, the
anisotropy data serves as a good discriminator of the different
RecA-DNA filaments, suggesting that anisotropy may prove
useful for future time-dependent studies of RecA-DNA
complex assembly.
monomer to DNA nucleotide ratios of 1:2.5 ( 0.4 in the
absence of cofactor for ODN 60a. In the presence of ATPγS,
the value was 1:2.7 ( 0.4. The observation that the
anisotropy changes saturate at the same points provides
strong corroborative evidence that the signal changes are
monitoring complex formation directly.
Importantly, the titration results of Figure 3 show that m5K
does not disturb RecA-DNA interactions. Specifically, these
RecA-DNA complexes have the canonical 3:1 nts to RecA
monomer stoichiometry determined by other techniques.1
This is the first demonstration of properly assembled RecA-
DNA filaments using intrinsic DNA fluorescence. By
comparison, changes in the DNA structure caused by ꢀA7,13,22
and fluorescein5,6 labeling results in the binding of two
ssDNA molecules and a stoichiometry of 1:6. Hence, the
use of ODNs containing the intrinsic DNA fluorophore m5K
offer the unique opportunity to compare directly DNA
binding at site I in the collapsed and extended RecA
filaments.
Monitoring total DNA emission as a function of increasing
protein concentration provided evidence that the signal
changes are saturatable (Figure 3). Moreover, the saturation
We have presented proof of principle for monitoring the
assembly of RecA nucleoprotein filaments using intrinsic
DNA fluorescence. Emission from m5K uniquely character-
izes collapsed and extended RecA-ssDNA filaments with
appropriate stoichiometries. In the future, the construction
of a kinetic scheme describing RecA-DNA complex forma-
tion will form the basis for elucidating the detection of and
response to DNA damage directed by RecA. Such use of
instrinsic fluorescence from chemically synthesized nucleo-
sides to study the thermodynamics and kinetics of recom-
bination molecular motors promises to play a prominent role
in a new era of biochemistry for RecA and related DNA
motor proteins.
Figure 3. Total (left) and polarized (right) fluorescence titrations
of ODN 60a with RecA in either the absence (squares) or presence
(circles) of ATPγS.
Acknowledgment. We thank Dr. Alberto Roca for
technical assistance. Funding was provided by grants from
the NIH (GM58114) and the Welch Foundation (C-1374).
Supporting Information Available: Protein and DNA
preparation procedures, detailed fluorescence spectral data,
and steady-state ATP hydrolysis kinetics for RecA-ODN
complexes. This material is available free of charge via the
Internet at http://pubs.acs.org.
behavior allowed the stoichiometry of the RecA-DNA
complexes to be determined. An equivalence point analysis22
of the DNA binding isotherms in Figure 3 yielded RecA
(21) Lakowicz, J. R. Principles of Fluorescence Spectroscopy, 2nd ed.;
Plenum Publishing: New York, 1999; Chapter 10.
(22) Silver, M. S.; Fersht, A. R. Biochemistry 1982, 21, 6066-6072.
OL0167863
3922
Org. Lett., Vol. 3, No. 24, 2001