Toward a mechanism-based fluorescent assay for site-specific recombinases and topoisomerases: Assay design and syntheses of fluorescent substrates

Article abstract of DOI:10.1021/ja9612920

Site-specific recombinases and topoisomerases break and rejoin the phosphodiester bonds of DNA. Both classes of enzymes do so through the formation of a covalent intermediate involving a phosphodiester bond with a hydroxylated amino acid (usually tyrosine

Full text of DOI:10.1021/ja9612920

12004
J. Am. Chem. Soc. 1996, 118, 12004-12011
Toward a Mechanism-Based Fluorescent Assay for Site-Specific
Recombinases and Topoisomerases: Assay Design and
Syntheses of Fluorescent Substrates
Gagan Panigrahi,† Bao-ping Zhao,† Jiri J. Krepinsky, and Paul D. Sadowski*
Contribution from the Department of Molecular and Medical Genetics, UniVersity of Toronto,
Toronto, Ontario M5S 1A8, Canada
ReceiVed April 19, 1996X
Abstract: Site-specific recombinases and topoisomerases break and rejoin the phosphodiester bonds of DNA. Both
classes of enzymes do so through the formation of a covalent intermediate involving a phosphodiester bond with a
hydroxylated amino acid (usually tyrosine). We have previously shown that oligonucleotides that bear a
3′-phosphoryltyrosine residue linked to the phosphoryl group via a phenolic hydroxyl group are effective substrates
for the assay of ligation by the FLP recombinase and mammalian topoisomerase I. In this article we describe the
synthesis of oligonucleotides bearing several novel 3′-phosphoryl substituents. It is shown that oligonucleotides
bearing a 3′-phosphoryltyrosine residue N-substituted on tyrosine with the bulky fluorescent groups dansyl and pyrene
are ligated effectively by the FLP recombinase, and the dansyltyrosine derivative is used as effectively as the tyrosine
adduct by mammalian topoisomerase I. The dansyl derivatives were completely stable during the syntheses; this
underlines the potential usefulness of the dansylated class of compounds for the development of simplified assays
and for mechanistic studies of breaking-joining enzymes.
Introduction
plasmid.5,6 In Vitro studies have yielded considerable insight
into the catalytic mechanism of the breakage-joining reaction
of this protein.7
Topoisomerases and conservative site-specific recombinases
catalyze the breaking and joining of DNA.1,2 Topoisomerases
are essential enzymes that relieve the superhelical strain
introduced into DNA by the processes of DNA replication and
transcription.1 Site-specific recombinases catalyze a variety of
genome rearrangements of episomal DNAs such as bacterio-
phages and plasmids. Both classes of enzymes share the
property of conserving the energy of the phosphodiester
backbone by forming a covalent phosphodiester bond between
a nucleophilic amino acid on the protein (tyrosine or serine)
and a phosphate terminus on the broken DNA (see Figure 1).
In the first step of these reactions, breakage of the phosphodi-
ester bond is accompanied by formation of a phosphate ester
bond between one of the two DNA fragments and the hydroxyl-
ated amino acid in the enzyme active site (usually tyrosine). In
the second step of the reaction sequence, the 5′-hydroxyl group
of the 2′-deoxyribose on the other DNA fragment displaces the
hydroxylated amino acid and reforms the phosphodiester bond.
In the case of topoisomerases, the breaking-joining reaction
changes the linking number of the DNA. The conservative site-
specific recombinases promote the joining of DNA molecules
in novel recombinations. In addition, they possess either non-
specific or site-specific topoisomerase activity.3,4 The FLP
recombinase is a member of the integrase family of conserva-
tive site-specific recombinases. The enzyme is encoded by the
autonomously replicating 2 µm plasmid of the yeast Saccha-
romyces cereVisiae and promotes the amplification of the
The active site of FLP8,9 contains four residues that are
absolutely conserved among all the integrase family members:
arginine 191, histidine 305, arginine 308, and tyrosine 343. (The
numbers refer to the amino acid residues of the FLP protein.)
The catalytic triad of R191, H305, and R308 has been implicated
in the activation of the scissile phosphodiester bonds of the FLP
recognition target site (FRT site) for cleavage, and R191 and
H305 have been implicated in the rejoining step of the reaction
(ligation). The Y343 residue is the nucleophile that facilitates
cleavage of the phosphodiester bonds and covalently attaches
the protein to the 3′-phosphoryl group at the site of the break.
The ligation stage of the reaction results from a nucleophilic
attack of the 5′-hydroxyl group upon the phosphotyrosine bond
with the resultant reclosure of the phosphodiester backbone and
liberation of the protein.7
We have developed a series of substrates to assay the ligation
step of the reaction. A nicked FRT-site substrate that contained
a 5′-hydroxyl group and an adjacent 3′-phosphotyrosine group
underwent efficient FLP-mediated ligation. This reaction did
not require the nucleophilic tyrosine 343.10 Analogous sub-
strates were synthesized chemically, and the ligation reaction
was extended to other site-specific recombinases and mam-
malian topoisomerase I.11,12
Since topoisomerases are the targets of agents useful in
antimicrobial and antineoplastic chemotherapy,13 we sought to
(5) Hartley, J. L.; Donelson, J. E. Nature 1980, 286, 860.
(6) Futcher, A. B. J. Theor. Biol. 1986, 119, 197.
* Corresponding author (416) 978-6061 (tel), (416) 978-6885 (fax),
P.sadowski@utoronto.ca (e-mail).
(7) Sadowski, P. D. Prog. Nucleic Acids Res. Mol. Biol. 1995, 51, 53.
(8) Argos, P.; Landy, A.; Abremski, K.; Egan, J. B.; Haggard-Ljundquist,
E.; Hoess, R. H.; Kahn, M. L.; Kalionis, B.; Narayana, S. V. L.; Pierson,
L. S.; Sternberg, N.; Leong, J. M. EMBO J. 1986, 5, 433.
(9) Abremski, K.; Hoess, R. H. Protein Eng. 1992, 53, 87.
(10) Pan, G.; Sadowski, P. D. J. Biol. Chem. 1992, 267, 12397.
(11) Pan, G.; Luetke, K.; Juby, C. D.; Brousseau, R.; Sadowski, P. D. J.
Biol. Chem. 1993, 268, 3683.
† The first two authors contributed equally to the experiments in this
paper; order of listing is alphabetical.
X Abstract published in AdVance ACS Abstracts, October 15, 1996.
(1) Champoux, J. J. In DNA Topology and Its Biological Effects;
Cozzarelli, N. R., Wang, J. C., Eds.; Cold Spring Harbor Laboratory: Cold
Spring Harbor, NY, 1990; pp 217-262.
(2) Sadowski, P. D. FASEB J. 1993, 7, 760.
(3) Kikuchi, Y.; Nash, H. Proc. Natl. Acad. Sci. U.S.A. 1979, 76, 3760.
(4) Krasnow, M. A.; Cozzarelli, N. R. Cell 1983, 32, 1313.
(12) Sadowski, P. D.; Pan, G.; Brousseau, R. Methods Mol. Genet. 1995,
6, 186.
S0002-7863(96)01292-9 CCC: $12.00 © 1996 American Chemical Society

Fluorescent Assay for Recombinase and Topoisomerase
J. Am. Chem. Soc., Vol. 118, No. 48, 1996 12005
Figure 1. Chemistry of FLP- or topoisomerase I-mediated cleavage and ligation (top left). A tyrosine hydroxyl acts as a nucleophile to break a
specific phosphoester bond to yield free a 5′-OH and a 3′-enzyme phosphotyrosyl intermediate (top right). A nucleophilic attack of a 5′-OH group
upon the phosphotyrosyl linkage reforms the phosphodiester bond and liberates the enzyme (bottom).
394 DNA/RNA synthesizer. Step 5: Cleavage of the amino acid-
oligonucleotide conjugates from the support and deprotection by
ammonolysis. Step 6: Purification of the conjugates by a passage
through Nensorb cartridges (DuPont) or by HPLC on a C18 column
(Waters, 10 µm, 3.9 × 300 mm) using a gradient of acetonitrile and
triethylamine acetate (0.1 M, pH 7). Electrospray mass spectrometry
(ES-MS) was used to confirm the structure of the purified oligonucleo-
tide conjugates.
Synthesis of Individual Substances. 3′-(O-Phosphotyrosylamide)-
Terminated Oligonucleotides 1. These substances were prepared as
described in ref 12 with the modifications described in the General
Synthetic Procedures section above. Although these are actually
tyrosine amides (see step 5 of General Synthetic Procedures and
Results), for brevity we refer to the substituent as “tyrosine”.
3′-(O-Phospho-N-dansyltyrosylamide)-Terminated Oligonucleo-
tides 2. See Figure 2, left column.
develop fluorescent substrates that might be useful in high-
throughput assays for screening novel therapeutic candidates
available through recent advances in synthetic combinatorial
chemistry. The development of such assays has at least three
requirements: (i) the synthesis of the substrates must be
compatible with commonly used automated methods of oligo-
nucleotide synthesis; (ii) the bulky fluorescent groups must not
interfere with the ligation activity of the enzymes; (iii) it must
be possible to detect both products of the ligation reaction,
namely, the elongated oligonucleotide as well as the liberated
fluorescent compound. In this article we describe the synthesis
of fluorescent tyrosine-oligonucleotide substrates for the FLP
recombinase and mammalian topoisomerase I and the use of
the substrates for the assay of these enzymes.
O-tert-Butyl-N-dansyltyrosine (8). A solution of dansyl chloride
(270 mg, 0.74 mmol) in acetonitrile (30 mL) was added dropwise to a
solution of O-tert-butyltyrosine (160 mg, 0.67 mmol) in lithium
carbonate buffer (pH 9.5, 0.1 M, 60 mL) at 0 °C, and the mixture was
stirred for 2 h. Acetonitrile was removed in Vacuo, and the residual
solution was acidified to pH 7 with 1 N HCl and extracted with ethyl
acetate. The organic phase was dried over Na2SO4 and evaporated to
dryness to yield a yellow solid residue, which was subsequently
subjected to chromatography on a silica gel column. The title
compound 8 (162 mg, 50%) was eluted with hexane:ethyl acetate (1:
2). 1H NMR (CDCl3): 8.49 (d, J ) 8.5 Hz, 1H), 8.20 (d, J ) 8.8 Hz,
1H), 8.18 (dd, J ) 7.3, 1.1 Hz, 1H), 7.52 (dd, J ) 8.8, 7.5 Hz, 1H),
7.46 (dd, J ) 8.5, 7.3 Hz, 1H), 7.19 (d, J ) 7.5 Hz, 1H), 6.82 (d, J )
8.5 Hz, 2H), 6.70 (d, J ) 8.5 Hz, 2H), 5.30 (d, J ) 8.3 Hz, 1H), 4.18-
4.14 (m, 1H), 2.90-2.88 (b, 8H), 1.28 (s, 9H). ES-MS: 469 [(M -
H)-, 100].
The O-tert-butyl-N-dansyltyrosyl-support (9), N-dansyltyrosyl-
support (10), 3′-(O-phospho-N-dansyltyrosylamide)-terminated oligo-
nucleotide, and 3′-(O-phospho-N-dansyltyrosylamide)-terminated oli-
gonucleotide 2 were prepared according to General Synthetic Procedures
(Figure 2, left column).
3′-(N-Phospho-O-tert-butyltyrosylamide)-Terminated Oligonu-
cleotides 3. See Figure 3, right column. The oligonucleotide synthesis
was initiated on the free amino group of the tyrosine.
Experimental Section
Chemical Synthesis of 3′-Substituted Oligonucleotides. General
Methods. Thin layer chromatography (TLC) was performed on silica
gel 60F254 (Merck) or polyamide plates (Schleicher & Schuell) and
visualized either by spraying with 50% aqueous sulfuric acid and
heating at 200 °C or by inspection under UV light (Mineralight). Silica
gel (230-400 mesh; Toronto Research Chemicals) was used for flash
chromatography. All starting materials were dried overnight in Vacuo
(10-3 mmHg) over KOH or P2O5 prior to use, and the solvents were
distilled from appropriate drying agents. Solutions were concentrated
at 1 mmHg pressure in a rotary evaporator.
1H NMR spectra (δ, ppm) were recorded at 500 MHz with a Varian
500 Unity Plus spectrometer at the NMR Spectrometry Laboratory
(Director: Dr. A. A. Grey) of the Carbohydrate Research Centre,
University of Toronto. Spectra were obtained in either CDCl3 or
CD3OD containing a trace of TMS (0.00 ppm). All mass spectra were
recorded with a VG Analytical ZAB-SE or Sciex API III spectrometer
at the Mass Spectrometry Laboratory (Acting Director: M. Cheung)
of the Carbohydrate Research Centre, University of Toronto. Relative
intensities (%) of ion peaks are quoted in parentheses. Fluorescence
spectra were recorded with spectrofluorometer MD 5020, Photon
Technology International. The spectra were obtained in aqueous
solutions.
Oligonucleotides were synthesized on an ABI 394 DNA/RNA
synthesizer (Perkin-Elmer) using standard chemistry unless stated
otherwise.
O-tert-Butyl-N-Fmoc-tyrosyl-support (11). This was prepared
according to step 2 of the general procedure (Figure 3).
O-tert-Butyltyrosyl-support (12). The Fmoc N-protective group
General Synthetic Procedures. We reported previously a scheme
for the synthesis of 3′-phosphoryltyrosine-terminated oligonucleo-
tides.11,12 We followed this synthetic design outlined below with
modifications noted (see steps 1-5 of Figures 2 and 3). Step 1:
Removal of the dimethoxytrityl protecting group from the functionalized
solid support (Teflon-based, Glen Research, or TentaGel-based, Rapp
Polymere, Tu¨bingen, Germany) to form the “activated support” (Figure
2, central column). Step 2: Coupling of protected tyrosine or its
derivatives by an ester bond to the 5′-hydroxyl group obtained in step
1 and capping the unreacted 5′-hydroxyl group. Step 3: Removal of
the tert-butyl group from the phenolic hydroxyl of tyrosine and its
derivatives by a 1-h treatment with 1 mL of 50% trifluoroacetic acid
in methylene chloride at room temperature. Step 4: Standard phos-
phoramidite synthesis of the oligonucleotide on a Perkin-Elmer ABI
t
was removed from the Bu-Tyr-Fmoc Teflon-based solid support (11;
30 mg) by a 30-min treatment with piperidine in dimethylformamide
(20%, 1 mL) at room temperature. The support was washed several
times with DMF and anhydrous acetonitrile and air-dried. The
oligonucleotide synthesis subsequently began at the free amino group
of tyrosine as described for steps 4-6.12
3′-[O-Phospho-N-[4-(1-pyrenyl)butyryl]tyrosylamide]-Termi-
nated Oligonucleotides 4 and 5. See Figure 3, left column. The
O-tert-butyltyrosyl-support (12; 30 mg) was treated with succinimidyl
1-pyrenylbutyrate (10 mg; Molecular Probes, Inc.) in dimethylform-
amide (1 mL) for 12 h at room temperature to yield the 4-(1-pyrenyl)-
butyramide of tyrosine (13). After several washes with DMF and
anhydrous acetonitrile, this substance was air-dried and subjected to
hydrolysis with trifluoroacetic acid, as described in step 3, to yield the
N-[4-(1-pyrenyl)butyryl]tyrosyl-support. After the unreacted amino
(13) Liu, L. F. Annu. ReV. Biochem. 1989, 58, 351.

12006 J. Am. Chem. Soc., Vol. 118, No. 48, 1996
Panigrahi et al.
Figure 2. Schematic description of the step by step synthesis of the oligonucleotide conjugates 2 and 7. The solid support may be either functionalized
Teflon or TentaGel (cf. Experimental Section). The numbers in squares indicate the steps 1-5 (Experimental Section).
groups were capped with acetic anhydride, the protected oligonucleo-
tides terminated with the 3′-O-phosphoryl-N-[4-(1-pyrenyl)butyryl]-
tyrosyl-support (14) were prepared as described in step 4 and depro-
tected to yield a mixture of compounds 4 and 5 according to step 5.
These two oligonucleotide conjugates were separated by reverse-phase
HPLC (step 6). Compound 4 retained its fluorescence, but compound
5 did not fluoresce (see Discussion for more details).
(CDCl3): 7.75 (d, J ) 7.5 Hz, 2H), 7.57 (d, J ) 7.3 Hz, 2H), 7.40 (t,
J ) 7.5 Hz, 2H), 7.31 (t, J ) 7.5 Hz, 2H), 7.04 (d, J ) 8.1 Hz, 2H),
6.77 (d, J ) 8.1 Hz, 2H), 4.85 (s, 1H), 4.76 (b, 1H), 4.41 (d, J ) 6.9
Hz, 2H), 4.21 (t, J ) 6.9 Hz, 1H), 3.41 (q, J ) 6.7 Hz, 2H), 2.74 (t,
J ) 6.7 Hz, 2H). EM-MS: 360 [(M + H)+, 57], 182 [(M - 179 +
2H)+, 53], 179 [(M - 180)+, 100], 121 [(M - 238)+, 56].
N-Fmoc-tyramine O-(2-Cyanoethyl)-N,N-diisopropylphosphor-
amidite (16). To a suspension of N-Fmoc-tyramine (150 mg, 0.42
mmol) in CH2Cl2 (5 mL) containing diisopropylethylamine (0.1 mL)
was added dropwise a solution of 2-cyanoethyl N,N-diisopropyl-
chlorophosphoramidite (191 mg, 0.8 mmol) in CH2Cl2 (2 mL) at 0 °C.
After the addition was completed, the reaction mixture was stirred for
another 30 min, and subsequently a 5% aqueous solution of NaHCO3
(10 mL) was added. The organic layer was separated, dried over
Na2SO4, and evaporated to dryness to give amorphous compound 16,
which was subjected to chromatography on silica gel (60 g). The
elution with hexane:ethyl acetate:triethylamine (2:1 + 1%) gave the
pure compound 16 (156 mg, 60%). 1H NMR (CDCl3): 7.76 (d, J )
7.5 Hz, 2H), 7.57 (d, J ) 7.3 Hz, 2H), 7.39 (t, J ) 7.5 Hz, 2H), 7.30
(t, J ) 7.3 Hz, 2H), 7.07 (d, J ) 8.0 Hz, 2H), 6.98 (d, J ) 8.0 Hz,
3′-O-Phosphoryltyramine-Terminated Oligonucleotides 6. See
Figure 4).
N-Fmoc-tyramine (15). A solution of 9-fluorenylmethyl chloro-
formate (1.10 g, 4.24 mmol) in diethyl ether (10 mL) was added
dropwise to a stirred suspension of tyramine (0.581 g, 4.24 mmol) in
tetrahydrofuran (70 mL) at 0 °C. After the mixture was stirred for
another 2 h, the organic solvents were removed in Vacuo, and ethyl
acetate (10 mL) was added. The solution was washed successively
with 1 N HCl, water, and brine and dried over Na2SO4. The solids
were removed by filtration, ethyl acetate was evaporated to dryness,
and crude N-Fmoc-tyramine was obtained as a white solid. Chroma-
tography on a silica gel column using hexane:ethyl acetate (3:1) as an
eluent gave pure N-Fmoc-tyramine (15; 1.19 g, 78%). 1H NMR

Fluorescent Assay for Recombinase and Topoisomerase
J. Am. Chem. Soc., Vol. 118, No. 48, 1996 12007
Figure 3. Schematic description of the step by step synthesis of the oligonucleotide conjugates 3 and 4. The solid support may be both functionalized
Teflon and TentaGel (cf. Experimental Section). The conjugate 5 accompanied the conjugate 4 (cf. Experimental Section and Discussion). The
numbers in squares indicate the steps 1-5 (Experimental Section).
2H), 4.81 (b, 1H), 4.39 (d, J ) 6.8 Hz, 2H), 4.20 (t, J ) 6.8 Hz, 1H),
3.96-3.85 (m, 2H), 3.77-3.69 (m, 2H), 3.41 (q, J ) 6.6 Hz, 2H),
2.75 (t, J ) 6.6 Hz, 2H), 2.64 (td, J ) 6.4, 1.5 Hz, 2H). 31P NMR
(CDCl3): 147.3.
The oligonucleotides were synthesized in the 5′ f 3′ direction using
5′-CPG support and 3′-dimethoxytritylated 5′-phosphoramidites. The
phosphoramidite of N-Fmoc-protected 4-(2-aminoethyl)phenol (16) was
coupled in the sequence in the last cycle (see Figure 4), and this was
followed by deprotection and purification as described in steps 5 and
6 of the General Synthetic Procedure.
Procedure was followed (steps 1-6) using (4-O-tert-butylphenyl)-
propionic acid.
Synthesis of N-Dansyltyrosine Amide (17). The coupling of O-tert-
butyl-N-dansyltyrosine (210 mg) with TentaGel (S 30 000; 500 mg) in
the presence of hydroxybenzotriazole (20 mg) and dicyclohexylcarbo-
diimide (400 mg) in dimethylformamide (5 mL) at room temperature
was allowed to proceed for 48 h. The TentaGel was then collected by
filtration and washed with dimethylformamide and dry acetonitrile.
To remove the O-tert-butyl group, TentaGel, to which tert-butyl-
N-dansyltyrosine was linked (500 mg), was treated with trifluoroacetic
acid in methylene chloride (50%, 5 mL) for 1 h at room temperature.
The liquids were then removed by filtration, and the TentaGel was
3′-[[4-(Phosphoryloxy)phenyl]propionamide]-Terminated Oligo-
nucleotides 7. See Figure 2, right column. The General Synthetic

12008 J. Am. Chem. Soc., Vol. 118, No. 48, 1996
Panigrahi et al.
using polynucleotide kinase (Figure 5, bottom panel). Each labeled
oligonucleotide was then annealed with two other unlabeled oligo-
nucleotides to form a duplex DNA substrate with a nick at the enzyme
cleavage site and potential ligation site. Three oligonucleotides (50
pmol of radioactive oligonucleotide and 100 pmol of nonradioactive
oligonucletide) were mixed in a 20 µL reaction containing 0.1 mM
NaCl and 5 mM MgCl2. The mixture was heated at 95 °C for 10 min
and then slowly cooled to 25 °C. The substrates were passed through
a G50 Sephadex column equilibrated with 10 mM Tris-HCl and 1 mM
EDTA.
Preparation of FLP, Topoisomerase I, and Other Enzymes.
Wild-type and mutant FLP proteins were prepared as described
previously.14 The purity of the wild-type FLP protein was 95%. The
purity of the mutant proteins was 15-30%. Mammalian topoisomerase
I was purchased from Promega. Polynucleotide kinase and terminal
deoxynucleotidyl transferase were purchased from New England
Biolabs.
Ligation Assays. The FLP and topoisomerase I reactions were done
as described previously.11,12 Where needed, 5 µg of proteinase K and
0.005% SDS were added to the reaction, and the incubation was
continued at room temperature for 10 min. The products were analyzed
by 8% denaturing gel electrophoresis.
Thin Layer Chromatography. The release of the fluorescent
dansyltyrosine (17) from the 3′-phosphoryl end was monitored by thin
layer chromatography. Typically, 100 pmol of substrate and 51 pmol
of FLP were used in a reaction volume of 100 µL. The incubation
was performed at room temperature for 1 h. The reaction mixture was
extracted three times with ethyl acetate, the combined extracts were
evaporated to dryness, and the residue was dissolved with 1 µL of water
and spotted on a polyamide TLC plate (5 × 5 cm). The polyamide
chromatogram was developed with 1.5% formic acid, and 17 exhibited
Rf ) 0.61. Compound 17 showed on the silica gel chromatogram
developed with chloroform:methanol:glacial acetic acid, (15:4:1), Rf
) 0.69.
Figure 4. Schematic description of the synthesis of the oligonucleotide
conjugate 6.
Results
Ligation of 3′-Phosphoryl Derivatives by FLP. We have
previously shown that enzymatically derived oligonucleotides
bearing a 3′-phosphoryltyrosine residue were effective substrates
to measure the ligation activity of the FLP recombinase.10
Synthetic oligonucleotides bearing a 3′-phosphoryltyrosylamide
(hereafter called “phosphoryltyrosine”) were also effectively
used by FLP, other site-specific recombinases, and mammalian
topoisomerase I.11,12
We wished to broaden the spectrum of 3′-phosphoryl sub-
stituents for two reasons. First, the use of fluorescently tagged
leaving groups might facilitate the development of simplified
assays for recombinases and topoisomerases. Second, such
substituents might provide further insight into the flexibility of
the active site and mechanisms of action of such enzymes.
In addition to the 3′-phosphoryltyrosylamide 111,12,15 as a
control, we also made oligonucleotides bearing a 3′-phosphoryl
group linked to N-dansyltyrosylamide 2 (“dansyltyrosine”),
N-(4-butyrylpyrenyl)tyrosylamide 4 (“pyrenotyrosine”), phos-
phoryltyramine 6, and (hydroxyphenyl)propionic acid 7. All
were linked to the 3′-phosphoryl end of the oligonucleotide via
the phenyl hydroxyl group. We also synthesized an oligonucleo-
tide in which the tyrosine was linked to the 3′-phosphoryl group
via its amino group instead of the hydroxyl group to ensure
that no unexpected enzymic reactions took place (see Figure 6
for a depiction of these structures). The synthesis of the pyrenyl
derivative 4 resulted in the formation of a byproduct (5) in about
equal amounts; 5 had lost its fluorophore, but its molecular size
apparently did not differ significantly from 4 (see Discussion).
Figure 5. Sequences of the substrates for FLP and topoisomerase I.
FLP: The horizontal arrows indicate the sequences of inverted
symmetry elements to which FLP binds. The open box represents the
core of the FRT site. The top strand consists of one oligonucleotide,
40 nucleotides in length, and the bottom strand consists of two
oligonucleotides of 22 and 18 nucleotides in length. The caret indicates
the phosphodiester bond of the top strand that is cleaved by FLP.
Topoisomerase I: The minimum sequence needed for topoisomerase
I-mediated cleavage is shown by the box. The top strand is 77
nucleotides in length, and the bottom strand consists of two oligo-
nucleotides, 40 and 35 nucleotides in length.11 The X represents the
presence of 3′-phosphoryl modification.
washed sequentially with methylene chloride, water, and dry acetonitrile
and treated with 30% aqueous ammonia for 2 h. N-Dansyltyrosine
amide (17) was obtained by lyophilization of the mother liquors after
the removal of TentaGel by filtration. 1H NMR (CD3COCD3): 8.39
(d, J ) 8.4 Hz, 1H), 8.16 (d, J ) 8.8 Hz, 1H), 7.94 (dd, J ) 7.3, 1.2
Hz, 1H), 7.42 (dd, J ) 8.8, 7.5 Hz, 1H), 7.39 (dd, J ) 8.4, 7.3 Hz,
1H), 7.12 (d, J ) 7.5 Hz, 1H), 6.64 (d, J ) 8.3 Hz, 2H), 6.32 (d, J )
8.3 Hz, 2H), 3.85 (m, 1H), 2.75 (s, 6H), 2.70 (dd, J ) 13.89, 5.77 Hz,
1H), 2.55 (dd, J ) 13.89, 7.69 Hz, 1H). HR-FAB-MS: calcd for
C21H24N3O4S [(M + H)+], 414.1488; found, 414.1485.
Preparations of Synthetic Activated Substrates. Oligonucleotides
were synthesized on an Applied Biosystems DNA synthesizer (Model
394) and purified either by the cartridge method or by HPLC (see
above). For the FLP reaction a 22-nucleotide oligonucleotide was 3′
labeled with 32P using terminal deoxynucleotidyl transferase and
[32P]ddATP (Figure 5, top panel). For the topoisomerase I reaction
(14) Pan, H.; Clary, D.; Sadowski, P. D. J. Biol. Chem. 1991, 266, 11347.
(15) Zhao, B. P.; Panigrahi, G.; Sadowski, P. D.; Krepinsky, J. J.
Tetrahedron Lett. 1996, 37, 3093.
the oligonucleotides bearing a 3′ modification were 5′ labeled with 32
P

Fluorescent Assay for Recombinase and Topoisomerase
J. Am. Chem. Soc., Vol. 118, No. 48, 1996 12009
Figure 6. Structures of the 3′-phosphoryl substituents used in this study. The oligonucleotides were synthesized as described in the Experimental
Section. The substituents are 1, O-phospho-1-tyrosine amide; 2, dansyltyrosine amide; 3, N-phosphotyrosine; 4, O-phospho-N-pyrenotyrosine
amide; 6, O-phosphotyramine; and 7, O-(phosphohydroxylphenyl)propionate.
nucleotide failed to show any ligated product (Figure 7, lane
6). This shows the importance of the phosphodiester linkage
to the aromatic leaving group. Note that all substrates showed
detectable amounts of a hairpin ligation product10 that resulted
from cleavage of the top strand and ligation to the 3′-32P-labeled
bottom strand. This serves as a useful internal control for the
enzyme activity.
Use of Synthetic Substrates To Assay Various Mutant
FLP Proteins. Much information has been gained about the
mechanism of action of the FLP protein by the study of
mutations of several key residues of the protein. In order to
learn whether the various substrates behaved differently from
wild-type FLP protein, we assayed the ligation activity of several
altered FLP proteins. The results are shown in Figures 8 and
9 and can be summarized as follows. The FLP Y343F16 protein
cannot cleave the FRT site because the nucleophilic tyrosine
residue has been changed to phenylalanine.17,18 FLP Y343F
can effectively ligate the 3′-tyrosyl-oligonucleotide, although
it cannot cleave and covalently attach to the substrate (Figure
8, lane 1). This mutant protein shows the same spectrum of
activities as the wild-type protein (Figure 8, lanes 1-6). FLP
Y343S (lanes 7-12) showed the same properties, although the
levels of activity were uniformly reduced in this experiment.
The arginine residue at position 308 has been implicated in
activation of the scissile phosphodiester bonds of the FRT site
for cleavage by tyrosine 343.19 FLP R308K can perform
efficient ligation of a substrate to which it can covalently
attach.20 It can cleave the tyrosine-terminated oligonucleotide
about 10% as efficiently as FLP+ protein and ligates this
oligonucleotide to a similar extent (Figure 8, lane 13). Interest-
ingly, FLP R308K shows some ligation of the dansyltyrosine
(lane 14) and tyramine (lane 18) substrates but no activity with
Figure 7. Ligation of 3′-phosphoryl-substituted oligonucleotides by
FLP. The substrate is diagrammed at the top. The reaction mixtures
contained 2.4 pmol of substrate and 17 pmol of FLP+ and were
incubated at room temperature for 20 min. The products were analyzed
on an 8% denaturing polyacrylamide gel. LP ) ligation product; S )
substrate; HP ) hairpin product which results from cleavage of the
top strand by FLP and nucleophilic attack of the free 5′-OH group of
the 3′-32P-labeled bottom strand upon the FLP-oligonucleotide com-
plex. The substrates bore the following 3′-phosphoryl substituents: lanes
1 and 2, -O-tyrosine 1; lanes 3 and 4, -O-tyrosine-N-dansyl 2; lanes 5
and 6, -N-tyrosine 3; lanes 7 and 8, -O-tyrosine-N-pyrene 4; lanes 9
and 10, -O-tyrosine-N-pyrene (pyrene modified) 5; lanes 11 and 12,
-O-tyramine 6. Note that in lane 6 the hairpin product is formed even
though the N-linked dansyltyrosine group is inactive for ligation. Some
substrates (lanes 5-12) contained an unknown band that migrated
behind the substrate band(s). This substance did not participate in, or
interfere with, ligation.
These oligonucleotides were annealed to the two other
oligonucleotides in order to assay ligation by the FLP recom-
binase (Figure 7). As shown previously, the tyrosylamide-
terminated oligonucleotide 1 was ligated very efficiently (Figure
7, lane 2). The dansyltyrosine, both pyrenyltyrosine derivatives,
and the tyramine-terminated substrates were also very effective
substrates (Figure 7, lanes 4, 8, 10, 12). An oligonucleotide
bearing a 3′-phosphoryl (hydroxyphenyl)propionate substituent
(7) was also an active substrate for ligation (data not shown).
Since these assays were done with enzymes in stoichiometric
excess, the reactions reflect the final yields of products rather
than rates. Only the N-linked, 3′-phosphoryltyrosine oligo-
(16) FLP Y343F means that the tyrosine residue at position 343 has been
changed to phenylalanine. This nomenclature was applied to the other
mutations as well.
(17) Prasad, P. V.; Young, L.-J.; Jayaram, M. Proc. Natl. Acad. Sci.
U.S.A. 1987, 84, 2189.
(18) Evans, B. R.; Chen, J.-W.; Parsons, R. L.; Bauer, T. K.; Teplow,
D. B.; Jayaram, M. J. Biol. Chem. 1990, 265, 18504.
(19) Parsons, R. L.; Prasad, P. V.; Harshey, R. M.; Jayaram, M. Mol.
Cell. Biol. 1988, 8, 3303.
(20) Zhu, X.-D.; Sadowski, P. D. J. Biol. Chem. 1995, 270, 23044.

12010 J. Am. Chem. Soc., Vol. 118, No. 48, 1996
Panigrahi et al.
Figure 8. Ligation of various 3′-substituted substrates by mutant FLP
proteins. The reaction conditions were the same as in Figure 7. The
substrates were substituted at their 3′-phosphoryl termini with 1 (O-
tyrosine, lanes 1, 7, 13), 2 (O-tyrosine-N-dansyl, lanes 2, 8, 14), 3 (N-
tyrosine, lanes 3, 9, 15), 4 (O-tyrosine-N-pyrene, lanes 4, 10, 16), 5
(O-tyrosine-N-pyrene (modified), lanes 5, 11, 17), and 6 (O-tyramine,
lanes 6, 12, 18). The proteins were FLP Y343F (lanes 1-6), 3.23 pmol;
and FLP Y343S (lanes 7-12), 1.3 pmol; FLP R308K (lanes 13-18),
1.19 pmol.
Figure 10. Ligation by mammalian topoisomerase I. The reaction
mixtures contained either 3′-phosphoryl-O-tyrosine-terminated (lanes
1-4) or 3-phosphoryl-O-tyrosine-N-dansyl-terminated (lanes 5-8)
oligonucleotides that were 5′ labeled with 32P (*). The substrates are
diagrammed at the top. The reaction mixtures contained 10 units (lanes
2, 6) or 20 units (lanes 3, 4, 7, 8) of topoisomerase I. Reaction mixtures
in lanes 1-3 and 5-7 were concentrated and applied directly to the
gel, whereas those in lanes 4 and 8 were treated with SDS and
proteinase K prior to electrophoresis. The 8% sequencing gel was run
for 20 min at 1500 V. CC, covalent complex; LP, ligation product; ID,
incompletely degraded peptides.
Figure 9. Reaction conditions and substrates were used as in Figure
8. Proteins used were FLP R191K (lanes 1-6), 5.95 pmol; FLP TA232
(lanes 7-12), 2.55 pmol; and FLP G328E (lanes 13-18), 3.4 pmol.
that were incompletely removed from the 3′-phosphoryl terminus
by proteinase K.
Fluorescence Spectroscopy. The excitation wavelengths of
dansylated tyrosylamide 17 and the dansylated conjugate 2 were
identical (340 nm), and the emission maximum was the same
for both dansyl derivatives as well (545 nm). Although the ꢀ
value of dansyl is low (3400) in comparison with other
fluorophores such as fluorescein, it was sufficiently high to be
identified in 17 by thin layer chromatography on polyamide
plates using an ordinary UV lamp for TLC (see below). The
spectral properties of pyrene (excitation wavelength at 365 nm
with maximum emission at 450 nm) are not more favorable (in
fact, the Stokes shift is smaller than for the dansyl derivatives),
and the synthesis of “pyrenylated” derivatives presents problems
(see Discussion below).
Identification of the Fluorescent Product of Ligation. One
product of FLP-mediated ligation is a lengthened 32P-labeled
strand that was detected by denaturing acrylamide gel electro-
phoresis. The other should be the free hydroxylated aromatic
leaving group, i.e., the dansylated tyrosylamide 17. In order
to detect this product, a ligation reaction involving the N-
dansylated tyrosine was scaled up; the products were extracted
with ethyl acetate and analyzed by thin-layer chromatography.
As expected, the fluorescent product that extracted into ethyl
acetate showed the same Rf ) 0.61 in 1.5% formic acid
(polyamide plate) as the N-dansyltyrosine amide (17) standard.
Thin-layer chromatography on a silica gel plate in chloroform:
methanol:acetic acid (15:4:1) was also suitable for detection of
N-dansyltyrosine amide (17) (Rf ) 0.69). Similar experiments
were done with the topoisomerase I ligation reaction and also
showed that N-dansyltyrosine amide (17) was produced.
either of the pyrenoyl derivatives (lanes 16, 17). This implies
that FLP R308K cannot cleave and attach to the latter substrates.
The other proteins (Figure 9) behaved as expected based on
their known activities on the tyrosine-terminated substrates.11
FLP R191K was ligation-defective with all substrates (lanes
1-6). FLP TA232, a DNA bending-defective protein, ligated
all the substrates as well as FLP+ (lanes 7-12), whereas FLP
G328E, another bending-defective FLP, ligated none of the
substrates (lanes 13-18).
Assay of Topoisomerase I Using Dansyltyrosine Deriva-
tives. One objective of these studies was to develop fluorescent
substrates that might be useful for screening of compounds that
inhibit topoisomerase I. It was therefore important that the
oligonucleotide bearing the bulky fluorescent adduct be as active
in the ligation assay as that bearing only the tyrosine. We
therefore compared the activity of the two oligonucleotides in
a ligation assay of mammalian topoisomerase I.
As shown in Figure 10, the dansylated 3′-phosphorylated
oligonucleotide was as effective a substrate as that bearing only
the 3′-phosphoryltyrosine (cf. lanes 2-4 vs lanes 6-8). In these
assays we also observed some radioactive material that remained
in the wells of the denaturing polyacrylamide gel (lanes 3, 6,
7). We believe that this was a covalent complex between the
enzyme and the DNA since the material was sensitive to
proteinase K (Figure 10, lanes 4, 8). In these lanes we also
observed the appearance of radioactive material that migrated
slightly more slowly than the substrate (Figure 10, lanes 4, 8,
ID). This material was due to the presence of oligopeptide(s)

Fluorescent Assay for Recombinase and Topoisomerase
J. Am. Chem. Soc., Vol. 118, No. 48, 1996 12011
Inhibition by Camptothecin of Topoisomerase I-Mediated
Ligation. To confirm the validity of the assumption that
inhibitors of recombinases indeed would be detectable using
substrates carrying a fluorescent label, camptothecin, a known
inhibitor of topoisomerase I, has been included in the reaction
mixture. As expected, the ligation reaction was suppressed (data
not shown).
interrupted. Although earlier observations25 indicate sites of
reactivity in pyrene and the exact chemical nature of this
modification is presently under investigation, the activity of 5
in ligation assays underlines the ability of FLP to accommodate
large variations in N-substituents of tyrosine (see below).
N-Linked oligonucleotide conjugates (phosphamides such as
3) were synthesized as control compounds to confirm that they
cannot be intermediates in the reactions catalyzed by recom-
binases. To our knowledge, such oligonucleotides have not been
synthesized before.
Discussion
Chemical Synthesis. The choices of the fluorescent reporter
group were limited by the compatibility of such a group with
the requirements of the synthesis and of the enzymic active site.
Dansyl was the preferred choice since the sulfonamide func-
tionality linking the fluorophore with tyrosine should be stable
under both acid and basic conditions and because many common
fluorophores with higher ꢀ values (e.g., fluorescein) are much
more difficult to use for preparations of analogous conjugates
(they are incompatible with the automated synthesis of oligo-
nucleotides, e.g., lactone in fluorescein).21 The pyrene deriva-
tive, together with other structural variations, served to probe
the requirements of the catalytic site.
3′-“Tyrosinylated” Oligonucleotides. The oligonucleotides
were synthesized using 5′-dimethoxytritylated 3′-phosphor-
amidites on a derivatized tyrosine (or its analogues) linked to a
solid Teflon support through a tyrosine carboxyl in the ester
form using the protocol developed for the CPG (controled pore
glass) support. Alternatively, we used TentaGel N which is a
poly(ethylene glycol) grafted onto poly(styrene). TentaGel
terminated with a hydroxyl group was used for the synthesis of
both the tyrosinylated oligonucleotide 1 and the dansylated
analogue 2 according to the protocol for automated oligonucleo-
tide synthesis on TentaGel.22,23
A reporter group was usually attached to the amino group of
the properly protected tyrosine by solution chemistry (Figure
2). Although in the case of tyrosylated oligonucleotides
ammonolysis is completed in 1 h at room temperature, 2.5 h
was needed in the presence of a bulky reporter group. In all
cases the amino acid carboxyl becomes an amide (cf. Figures 2
and 3).24
The tert-butyl group with which the tyrosine hydroxyl was
protected was removed by treatment with trifluoroacetic acid.
This step was not compatible with the use of CPG solid support
which is labile in the presence of strong acids; therefore, it had
to be performed on Teflon or TentaGel.
3′-“Tyraminylated” Oligonucleotides. Since tyramine does
not contain a carboxyl group to link tyrosine derivatives to a
solid support, these oligonucleotides were synthesized on the
CPG support in the 5′ f 3′ direction using 3′-dimethoxytri-
tylated 5′-phosphoramidites (Figure 4). It is worth mentioning
that although the reactivity of the amino group of tyramine with
9-fluorenylmethyl chloroformate was appreciably higher than
that of the OH group of tyramine, the reaction had to be carefully
controlled to avoid the additional derivatization of the phenolic
hydroxyl group.
Ligation Assays. This study shows the feasibility of attach-
ing bulky substituents to the 3′-phosphoryl terminus of oligo-
nucleotides and using them to assay the ligation activity of the
FLP recombinase and topoisomerase I.
The “ligation pocket” of the FLP protein is surprisingly
flexible since it can accommodate a 3′-phosphoryl terminus that
bears not only a tyrosyl group but also those that contain tyrosine
that has been substituted with dansyl or pyrene groups. While
the wild-type FLP protein is able to remove the tyrosine group
and attach covalently to the 3′-phosphoryl terminus before it
executes ligation (cleavage-dependent ligation),20 a FLP protein
that lacks the nucleophilic tyrosine 343 (FLP Y343F) is
nevertheless able to carry out efficient ligation on all the 3′-
substituted substrates tested. The fact that all the substrates
behaved similarly to the tyrosine-substituted oligonucleotide
with all the FLP proteins tested argues that the mechanism of
ligation of all the modified substrates occurs by the same
chemical mechanism. The importance of the phosphoryl linkage
to the aromatic hydroxyl group was shown by the fact that the
N-linked tyrosine derivative was inactive as a leaving group in
the ligation reaction.
We extended the utility of the FLP ligation reactions by
showing that mammalian topoisomerase I can also utilize the
N-dansylated tyrosine derivative as a ligation substrate. We
observed that these substrates also became covalently attached
to the topoisomerase. It may be that the substrate first undergoes
ligation, and then the protein recleaves and covalently attaches
to the substrate. Alternatively the enzyme may cleave and
covalently attach to the substrate without ligating it. Distinction
between these two possibilities could make the substrates useful
for studying the cleavage and ligation steps of the reaction
separately.
In the case of pyrene-tagged conjugates, the reporter group
was attached to the amino group of tyrosine that had already
been linked to the solid support (cf. Figure 3). We have noted
reduced stability of the pyrene ring system under the conditions
of oligonucleotide synthesis. The expected fluorescing pyren-
ylated derivative 4 was accompanied by the nonfluorescing
oligonucleotide derivative 5. Both substances exhibited sub-
stantially prolonged elution times (35 and 33 min) on HPLC
using a C18 reverse-phase column in trimethylamine acetate (pH
7), while the same nucleotide without the pyrene reporter group
was eluted in 11 min. Clearly, the compound 5 is nonfluores-
cent because the conjugation within the pyrene ring system was
Acknowledgment. The authors gratefully acknowledge
funding of this work by the Natural Sciences and Engineering
Research Council of Canada in the form of a strategic grant (to
P.D.S.).
(21) Himel, C. M.; Chan, L.-M. In Biochemical Fluorescence: Concepts;
Chen, R. F.; Edelhoch, H., Eds.; Marcel Dekker: New York, 1975; pp 607-
632.
(22) Bayer, E.; Bleicher, K.; Maier, M. Z. Naturforsch. 1995, 50B, 1096.
(23) Bayer, E.; Maier, M.; Bleicher, K.; Gaus, H.-J. Z. Naturforsch. 1995,
50B, 671.
(24) All nucleotide conjugates with tyrosine or other carboxylic acids,
e.g., 3′-tyrosyl-oligonucleotide, contain carboxamide rather than the carboxyl
group. The conversion of the carboxyl group to its amide occurs during
the ammonolysis step used to remove the oligonucleotide conjugate from
the solid support.
JA9612920
(25) (a) Graebe, C. Liebigs Ann. Chem. 1871, 158, 294. (b) Goldschmiedt,
G.; Wegscheider, R. Monatshefte Chem. 1883, 4, 238. (c) Bannwarth, W.
HelV. Chim. Acta 1988, 71, 1517. (d) Chaikovskii, V. V.; Novikov, A. N.
Zh. Org. Khim. 1984, 20, 1482. (e) Radner, F. Acta Chem. Scand. 1989,
43, 481.