Mapping of the functional phosphate groups in the
catalytic core of deoxyribozyme 10–23
Barbara Nawrot, Kinga Widera, Marzena Wojcik*, Beata Rebowska, Genowefa Nowak and
Wojciech J. Stec
Department of Bioorganic Chemistry, Centre of Molecular and Macromolecular Studies of the Polish Academy of Sciences, Lodz, Poland
The RNA-cleaving DNA enzymes, like most ribozymes,
require a divalent metal cation for their cleavage activity
[1]. Among metal ion-dependent DNA enzymes,
deoxyribozyme 10–23, first selected and characterized
by Santoro & Joyce [1,2], has been examined most
extensively both in vitro and in vivo [3–5]. This enzyme
consists of a 15-nucleotide conserved catalytic core and
variable substrate recognition arms (Fig. 1A). Cleavage
of an RNA substrate is highly sequence-specific, and
occurs between the bulged 5¢-purine and paired 3¢-pyr-
imidine nucleosides, resulting in the formation of the
two products, a 5¢-terminal product with a 2¢,3¢-cyclic
phosphate, and a 3¢-terminal product containing an OH
group at its 5¢-end. The enzyme preferentially uses Mg
2+
for its activity, although other divalent metal ions are
accepted as cofactors [1,2,6]. To date, the structure of
the substrate–deoxyribozyme 10–23 active complex
remains unknown [7,8], and the mechanistic details of
the catalytic reaction are not fully understood. There-
fore, much effort has been devoted to determine the role
of individual nucleotides in the 10–23 catalytic core, as
well as their relative importance [9–13]. Despite numer-
ous studies performed on a mutant deoxyribozyme 10–
23 containing chemical modifications inserted into the
catalytic core, the role of particular phosphates within

or 3 mm Mn
2+
. A metal-specificity switch approach permitted
the identification of nonbridging phosphate oxygens (proR
P
or proS
P
)
located at seven positions of the core (P2, P4 and P9–13) involved in direct
coordination with a divalent metal ion(s). By contrast, phosphorothioates
at positions P3, P6, P7 and P14–16 displayed no functional relevance in the
deoxyribozyme-mediated catalysis. Interestingly, phosphorothioate modifi-
cations at positions P1 or P8 enhanced the catalytic efficiency of the
enzyme. Among the tested deoxyribozymes, thio-substitution at position P5
had the largest deleterious effect on the catalytic rate in the presence of
Mg
2+
, and this was reversed in the presence of Mn
2+
. Further experiments
with thio-deoxyribozymes of stereodefined P-chirality suggested direct
involvement of both oxygens of the P5 phosphate and the proR
P
oxygen at
P9 in the metal ion coordination. In addition, it was found that the oxygen
atom at C6 of G
6
contributes to metal ion binding and that this interaction
is essential for 10–23 deoxyribozyme catalytic activity.
Abbreviations

of thiophilic cations such as Mn
2+
,Zn
2+
,orCd
2+
,in
increasing order (the rescue effect). Analysis of these
types of interaction led to a better understanding of the
mechanistic aspects of the action of naturally occurring
catalytic ribozymes: group I and II introns [22–25], the
RNA subunit of RNase P [26,27], and the hammerhead
ribozymes [18,28–30]. The successful application of P-
chiral phosphorothioates in those mechanistic studies
prompted us to establish the role of phosphate groups in
the catalytic core of deoxyribozyme 10–23. First, we
introduced a P-stereorandom single PS linkage in prede-
termined positions of the catalytic core in 16 thio-deoxy-
ribozymes 10–23 (P1–P16; Table 1, entries 2–17), and
conducted metal-specificity switch experiments with
Mg
2+
and thiophilic Mn
2+
. These experiments showed
that catalytically important phosphate groups were
positioned within the catalytic domain of the enzyme.
The role of the particular oxygen atoms of the selected
phosphate groups is also discussed. Moreover, we ana-
lyzed the function of the oxygen moiety at C6 of nucleo-

was replaced by sulfurization [31]. Each oligomer was
an R
P
and S
P
(c. 1 : 1) diastereomeric mixture (Fig. 1).
The activity of thio-substituted deoxyribozymes was
tested against a short target substrate homosequential
with mRNA of aspartyl protease Asp2 (BACE1, acces-
sion number AF190725, between nucleotides 1801 and
1817) (Fig. 1). It has already been demonstrated that
deoxyribozyme 10–23 accepts not only short RNA
substrates but also modified substrates containing a
DNA backbone with RNA nucleotides (5¢-purine and
3¢-pyrimidine ribonucleotides) positioned at the scissile
bond of the target oligonucleotide [32–34]. We pre-
pared a 17-nucleotide chimeric DNAÆRNA substrate
with the sequence 5¢-d(ACAGATGA)GUd(CAACC-
CT)-3¢, which was easier to synthesize and chemically
more stable than an RNA oligonucleotide.
All kinetic experiments were performed at a satur-
ating concentration of the unmodified deoxyribozyme
1 or thio-deoxyribozymes 2–17 (10 lm) with
32
P-labe-
led substrate (0.1 lm) in the presence of 3 mm MgCl
2
.
The cleavage product (9-mer) and the substrate were
quantified by autoradiography following electrophor-

chiral PS internucleotide bonds in PS DNA of S
P
-sense and R
P
-
sense of chirality, respectively.
B. Nawrot et al. Metal-binding site in deoxyribozyme 10–23
FEBS Journal 274 (2007) 1062–1072 ª 2007 Centre of Molecular and Macromolecular Studies of the Polish Academy of Sciences
Journal compilation ª 2007 FEBS
1063
esis in 20% polyacrylamide gels. The observed rate
constants (k
obs
) were calculated according to the equa-
tion given in Experimental procedures, and compared
with the rate constant of the unmodified deoxyribo-
zyme (k
rel
).
The data presented in Table 1 and Fig. 2A indicate
that thio-substitutions at phosphates P2, P4 and P9–13
lowered the k
rel
values by c. 50%. The replacement of
3mm Mg
2+
with 3 mm Mn
2+
resulted in a restoration
of the activity to the level of the k

catalysis steps such as the acceleration of the ribose
2¢-hydroxyl group deprotonation, stabilization of a
negative charge that may develop on the nonbridging
oxygen in a transition state, and ⁄ or stabilization of the
negative charge on the oxygen atom of the 5¢-leaving
group.
Among the tested modified enzymes, the biggest thio
effect (a 16-fold reduction in the cleavage activity;
Table 1, Fig. 2A) was found for the PS enzyme modi-
fied at position P5. The reduction was much bigger
than the two-fold reduction expected if only one of the
diastereomers coordinated the metal ion, suggesting
that the sulfur atoms in both the proR
P
and proS
P
positions hindered direct contact with metal ions.
Interestingly, this PS enzyme regained its activity in
the presence of Mn
2+
, with the k
Mn
obs
value being 176-
fold higher than the k
Mg
obs
value. This value, however,
was still c. 3-fold lower than that measured for the
unmodified reference at the same conditions (Table 1).

)
d
k
Mn
rel
e
k
Mn
obs
⁄ k
Mg
obs
(rescue effect)
1 d(AGGCTAGCTACAACGAT) 0.27 ± 0.028 1 1.0 8.00 ± 0.42 1 30
2 P1 d(A
PS
GGCTAGCTACAACGAT) 0.85 ± 0.042 3.10 0.3 0.36 ± 0.031
f
3.0
f
NA
3 P2 d(AG
PS
GCTAGCTACAACGAT) 0.15 ± 0.018 0.56 1.8 8.10 ± 0.57 1.01 54
4 P3 d(AGG
PS
CTAGCTACAACGAT) 0.24 ± 0.0078 0.89 1.1 8.00 ± 1.10 1.00 33
5 P4 d(AGGC
PS
TAGCTACAACGAT) 0.16 ± 0.0071 0.59 1.7 8.60 ± 1.40 1.08 54

CGAT) 0.14 ± 0.024 0.52 1.9 9.20 ± 0.71 1.20 66
15 P14 d(AGGCTAGCTACAAC
PS
GAT) 0.28 ± 0.014 0.96 1.0 9.30 ± 0.64 1.20 33
16 P15 d(AGGCTAGCTACAACG
PS
AT) 0.38 ± 0.035 1.40 0.7 9.40 ± 0.71 1.20 25
17 P16 d(AGGCTAGCTACAACGA
PS
T) 0.24 ± 0.030 0.89 1.1 7.40 ± 0.78 0.93 31
18 P1 ⁄ P8 d(A
PS
GGCTAGC
PS
TACAACGAT) 0.76 ± 0.080 2.80 0.78 ± 0.048
f
6.5
f
NA
a
The sequences of PS deoxyribozymes 10–23 containing a single PS linkage of stereorandom P-configuration (equal amounts of R
P
and
S
P
diastereomers) in the selected positions of the catalytic core marked from P1 (phosphate bond between A
0
and G
1
) to P16 (phosphate

rel
¼ the ratio
of the k
obs
values of modified deoxyribozyme to unmodified deoxyribozyme, in the presence of Mn
2+
.
f
Reactions were performed in
20 m
M Tris ⁄ HCl (pH 7.5), containing 100 mM NaCl and 0.06 mM Mn
2+
under single-turnover conditions with 0.1 lM 5¢-end
32
P-labeled sub-
strate and 10 l
M deoxyribozyme, at 37 °C. Values of k
obs
for unsubstituted and thio-substituted deoxyribozyme reactions represent mean
values of four independent experiments, and errors indicate deviations between individual experiments. The obtained data were normal-
ized to a k
obs
of 0.12 ± 0.014 min
)1
for reaction of the unmodified deoxyribozyme in 0.06 mM Mn
2+
.
Metal-binding site in deoxyribozyme 10–23 B. Nawrot et al.
1064
FEBS Journal 274 (2007) 1062–1072 ª 2007 Centre of Molecular and Macromolecular Studies of the Polish Academy of Sciences

The catalytic activity of double PS-substituted
deoxyribozyme 10–23
PS modification at positions P1 or P8, surprisingly,
accelerated the cleavage rates (3-fold and 1.6-fold,
respectively) in the presence of Mg
2+
as well as Mn
2+
(Table 1, Fig. 3). The k
obs
and k
rel
values for these
enzymes were calculated from the reactions performed
in 3 mm Mg
2+
or 0.06 mm Mn
2+
. The concentration of
Mn
2+
was reduced 50-fold, because reactions per-
formed in the presence of 3 mm Mn
2+
reached comple-
tion in less than 5 s, making kinetic analysis
impossible. Whereas the P8 substitution had only a
minor effect both in the presence of Mg
2+
and in the

k
Mn
obs
value for this mutant enzyme was three-fold
higher than the k
Mn
obs
value for the unmodified reference.
The obtained data demonstrate that the P1 ⁄ P8
A
unmodified
P2
P3
P4
P5
P6
P7
P9
P10
P11
P12
P13
P14
P15
P16
unmodified
P2
P3
P4
P5

B
Fig. 2. Comparison of the relative rates of cleavage (k
rel
) of thio-
substituted deoxyribozymes 10–23 in the presence of 3 m
M MgCl
2
(A) and 3 mM MnCl
2
(B).
0
1
2
3
4
5
6
7
8
unmodified Mg
2+
unmodified Mn
2+
P1 Mg
2+
P1 Mn
2+
P8 Mg
2+
P8 Mn

could
reach a value of c. 40 min
)1
. This is a four times
higher value than the highest one so far reported in
the literature for catalytic nucleic acids [35]. Moreover,
this double PS congener is c. 150-fold more active in
the presence of Mn
2+
than the unmodified reference in
the presence of Mg
2+
. A possible explanation for these
results is that both pairs of oxygen atoms at the P1
and P8 phosphates do not directly interact with metal
ions, and such a double PS modification, together with
the presence of Mn
2+
, facilitates a catalytically favora-
ble conformation of the 10–23 core. Moreover, one
cannot exclude the possibility that the 10–23 enzyme
operates with two metal ions interacting with different
sets of residues.
Our finding that the introduction of a PS bond at
the P1 site of deoxyribozyme 10–23 causes about
three-fold stimulation of the cleavage rate, irrespective
of the metal ion used, demonstrates that chemical
modifications of the deoxyribozyme backbone can be
used to improve both its stability and its catalytic effi-
ciency in cellular experiments.

Mg
2+
interactions with nonbridging phosphate oxy-
gens. In 3 mm Mn
2+
buffer, the R
P
-PS and S
P
-PS
deoxyribozyme P5-mediated cleavage activity was
significantly enhanced (73-fold and 108-fold increase
of k
obs
values, respectively; Table 2, Fig. 5B). The
observed thio effect and rescue effect values for partic-
ular P-chiral diastereomers slightly differed from those
determined for the diastereomeric mixture of this PS
enzyme, and these differences may result from experi-
mental errors. The remarkable increase of the catalytic
rate for the reactions carried out in the presence of
Mg
2+
and each of the P-chiral diastereomeric deoxyri-
bozymes suggests that Mn
2+
can stimulate 10–23
enzyme activity in a way that depends on the simulta-
neous metal ion interactions with both nonbridging
oxygens at position P5. Thus, earlier suggestions are

D
Fig. 4. Comparison of the Mg
2+
-dependent activity of the unmodi-
fied deoxyribozyme 10–23 with that of thio-substituted deoxyribo-
zyme R
P
-P5 in the presence of 3 mM MgCl
2
. Time course of
cleavage reaction of a chimeric DNAÆRNA oligonucleotide by the
unmodified (A, B) and R
P
-P5 (C, D) deoxyribozymes.
Metal-binding site in deoxyribozyme 10–23 B. Nawrot et al.
1066
FEBS Journal 274 (2007) 1062–1072 ª 2007 Centre of Molecular and Macromolecular Studies of the Polish Academy of Sciences
Journal compilation ª 2007 FEBS
eomeric deoxyribozymes were prepared from diastereo-
meric R
P
dimers and S
P
dimers, G
PS
C [37]. We chose
this sequence following Taira and coworkers’ sugges-
tion that the proR
P
phosphate at position P7 (between

phosphate oxygens at positions P3 and P7 reach sim-
ilar values, indicating a lack of direct coordination of
a metal cation to the proR
P
and proS
P
oxygen atoms
at these positions. These findings further confirm our
previous data obtained for the mixtures of diastereo-
mers of PS deoxyribozymes.
We compared our data with those published for
hammerhead ribozymes containing site-specific PS
modifications at either the proR
P
or proS
P
positions
[17]. Single-turnover relative rates of RNA cleavage,
determined at 10 mm Mg
2+
, were reduced three-fold
for R
P
-PS isomers at positions A
13
and A
14
, and S
P
-PS

observed for hammerhead constructs, except for the
k
rel
value determined in the presence of Mg
2+
for the
R
P
-PS isomer at position U
1.1
of the hammerhead
ribozyme.
Mutational analysis of nucleoside in position 6
of the catalytic core
We were interested in whether there are any other lig-
ands in the 10–23 catalytic core that might be directly
involved in stabilization of the catalytically active
architecture of the deoxyribozyme. As has already
been proven, the hammerhead ribozyme metal-binding
site utilizes both nonbridging oxygen atoms of the A
9
phosphate as well as nitrogen N7 of the subsequent
guanosine unit G
10.1
[38]. We were interested in deter-
mining whether the nucleotide residue following the A
5
unit in deoxyribozyme 10–23 plays any role in cata-
lysis. Although the exact metal-binding site of deoxyri-
bozyme 10–23 is not yet known, it has already been

Mg
obs
(min
)1
)
b
k
Mg
rel
d
Thio effect k
Mn
obs
(min
)1
)
c
k
Mn
rel
e
k
Mn
obs
⁄ k
Mg
obs
f
(rescue effect)
1 Unmodified 0.27 ± 0.028 1 1 8.0 ± 0.42 1 30

b, c
All RNA cleavage reactions were performed in
20 m
M Tris ⁄ HCl (pH 7.5), containing 100 mM NaCl and
b
3mM Mg
2+
or
c
3mM Mn
2+
under single-turnover conditions with 0.1 lM 5¢-end
32
P-
labeled substrate and 10 l
M deoxyribozyme, at 37 °C. Values of k
obs
for nonsubstituted and thio-substituted deoxyribozyme reactions repre-
sent mean values of four independent experiments, and errors indicate deviations between individual experiments.
d
k
Mg
rel
ratio of the k
obs
values for the modified and unmodified deoxyribozymes in the presence of Mg
2+
.
e
k

inosine substitution [11], exchange of the G
6
base with
AP nucleoside resulted in complete loss of catalytic
activity, independent of the metal ion (no substrate
cleavage over 8 h; Table 3). These findings clearly indi-
cate that the oxygen at C6 is essential for the catalytic
activity of deoxyribozyme 10–23, whereas the exo
amino group of G
6
is not of functional importance.
In addition, we extended our mutational analysis to
the nucleoside at position 6 by the replacement of G
6
with a 7-deaza-dG unit. This substitution resulted in a
104-fold loss of activity of the DN
7
-zyme in the pres-
ence of Mg
2+
, suggesting that the N7 nitrogen partici-
pates in the formation of a functionally important
intramolecular hydrogen bond within the deoxyribo-
zyme 10–23 catalytic core. The k
obs
for this enzyme
increased by almost three orders of magnitude upon
addition of Mn
2+
, and was about 30-fold greater than

P
and proS
P
oxygens of P5, and an interaction with
the oxygen ligand at C6 of the subsequent guanosine
nucleotide (Fig. 7). One can argue that in this model
the distances between the oxygen ligands of P5 and the
oxygen of G
6
are too large to be spanned by a single
metal ion. However, it is possible that the architecture
of the active conformation of the catalytic core allows
for such interactions, or that more than one metal ion
is involved in catalysis. Contributions of other ligands
cannot be excluded, and the first candidate is the
proR
P
oxygen of phosphate P9, between the T
8
and A
9
nucleosides (Table 2). It is also possible that other
functional groups of the catalytic core serve as metal
ion ligands, because, as we have already suggested,
there are at least seven more nonbridging phosphate
oxygens, at positions P2, P4, P9, P10, P11, P12 and
P13, which exhibit remarkable thio and rescue effects.
Besides the oxygen ligands of the internucleotide
bonds, some other functional groups, as indicated in
other studies [11], may form intraloop hydrogen bonds

P
S
P
3
P
R
P
3
P
S
P
5
P
R
P
5
P
S
P
7
P
R
P
7
P
S
P
9
P
R

2
(A) and 3 mM MnCl
2
(B).
Metal-binding site in deoxyribozyme 10–23 B. Nawrot et al.
1068
FEBS Journal 274 (2007) 1062–1072 ª 2007 Centre of Molecular and Macromolecular Studies of the Polish Academy of Sciences
Journal compilation ª 2007 FEBS
this model, the plausible ligands for metal coordina-
tion are the proR
P
and proS
P
oxygen atoms of the P5
phosphate, and the proR
P
oxygen at position P9, as
well as the carbonyl oxygen of the guanosine unit at
position 6 of the 10–23 catalytic core. In addition, sev-
eral other phosphate oxygens and nucleobase func-
tional groups can serve as metal-binding ligands
and ⁄ or hydrogen bond acceptors within the catalytic
core, but no detailed information is yet available.
Therefore, further experiments are required to identify
possible metal-binding ligands and to study the struc-
ture of deoxyribozyme 10–23 at the atomic level, either
by molecular modeling or by solution of the crystal
structure.
In addition, our observations that nonbridged oxy-
gens at phosphates at positions P3, P6, P7, P14 and

Mn
rel
d
k
Mn
obs
⁄ k
Mg
obs
e
Unmodified (WT) None 0.27 ± 0.028 1 8.0 ± 0.42 1 30
G
6
fi adenosine
f
ND – – – –
G
6
fi inosine
f
As WT – – – –
s
6
G-zyme G
6
fi 6-thio-dG 0.013 ± 0.0028 0.048 0.37 ± 0.035 0.046 28
AP-zyme G
6
fi 2-aminopurine nucleoside ND – ND – –
DN

.
d
k
Mn
rel
¼ ratio of the k
obs
values of modified deoxyribozyme to unmodified deoxyribozyme, in the presence of Mn
2+
.
e
The values of the res-
cue effect were calculated from k
Mn
obs
⁄ k
Mg
obs
f
.
B. Nawrot et al. Metal-binding site in deoxyribozyme 10–23
FEBS Journal 274 (2007) 1062–1072 ª 2007 Centre of Molecular and Macromolecular Studies of the Polish Academy of Sciences
Journal compilation ª 2007 FEBS
1069
Experimental procedures
Deoxyribozymes and substrate
The unmodified deoxyribozyme and its substrate oligonu-
cleotide (Fig. 1) were synthesized using an ABI 394 DNA
synthesizer (Applied Biosystems, Inc., Foster City, CA) and
commercially available phosphoramidite monomers (Glen

P]ATP and
T4 polynucleotide kinase (Amersham, Little Chalfont,
UK). A mixture containing 10 mm Tris ⁄ HCl (pH 8.5),
10 mm MgCl
2
,7mm 2-mercaptoethanol, 30 lm (0.1 A
260
unit) oligonucleotide, 1 lL (10 lCi) of [c-
32
P]ATP and T4
polynucleotide kinase (6 units) was incubated for 30 min at
37 °C, and then heat denatured and stored at ) 20 °C.
Enzymatic assay
The substrate cleavage reactions were performed under
single-turnover conditions with the DNA enzyme in 100-
fold excess over the substrate. The 5¢-labeled substrate
(0.1 lm) was incubated with deoxyribozyme (10 lm)in
20 mm Tris ⁄ HCl (pH 7.5) containing 100 mm NaCl, and
3mm MgCl
2
or 3 mm MnCl
2
,at37°C. After various
time intervals, 10 lL aliquots were withdrawn, and the
cleavage reaction was stopped by addition of 50 mm
EDTA and by cooling on ice. Before electrophoresis, 8 lL
of formamide containing 0.03% bromophenol blue and
0.03% xylene cyanol was added to each sample, and the
cleavage products were separated from noncleaved sub-
strate by electrophoresis in 20% polyacrylamide gel under

rel
were calculated as a ratio of
(k
obs
M+SD
M
) ⁄ (k
obs
U ) SD
U
), where k
obs
M and
SD
M
, and k
obs
U and SD
U
, are the mean reaction rates
and SD errors for the modified and unmodified enzymes,
respectively. Similarly, lower limits for k
rel
were calculated
from the equation (k
obs
M ) SD
M
) ⁄ (k
obs

O
O
O
O
P
O
O
A
O
O
O
T
O
Mg
2+
O
O
T
O
O
P
O
O O
A
O
?
?
pro-R
P
P5

of modified enzyme.
Acknowledgements
The authors thank Professor J. Connolly of Glasgow
University for critical reading of the manuscript and
valuable suggestions. This work was supported by
the Ministry of Science and Higher Education
(Poland) through the Centre of Molecular and
Macromolecular Studies, Polish Academy of Sciences,
under Decision 70⁄ E-63 ⁄ SN-014 ⁄ 2006 and ICGEB
project CRP ⁄ POL04-01.
References
1 Santoro SW & Joyce GF (1998) Mechanism and utility
of an RNA-cleaving DNA enzyme. Biochemistry 37 ,
13330–13342.
2 Santoro SW & Joyce GF (1997) A general purpose
RNA-cleaving DNA enzyme. Proc Natl Acad Sci USA
94, 4262–4266.
3 Dass CR (2004) Deoxyribozymes: cleaving a path to
clinical trials. Trends Pharmacol Sci 25, 395–397.
4 Silverman SK (2005) In vitro selection, characterization,
and application of deoxyribozymes that cleave RNA.
Nucleic Acids Res 33, 6151–6163.
5 Dass CR (2006) A DNA enzyme that cleaves RNA.
Preclinical anticancer activity of DNA-based cleavage
molecules. Drug Dev Ind Pharm 32, 1–5.
6 Breaker RR & Joyce GF (1994) A DNA enzyme that
cleaves RNA. Chem Biol 1, 223–229.
7 Nowakowski J, Shim PJ, Prasad GS, Stout CD & Joyce
GF (1999) Crystal structure of an 82-nucleotide RNA–
DNA complex formed by the 10–23 DNA enzyme.

phosphorothioate-containing oligoribonucleotides in the
study of the mechanism of cleavage of hammerhead
ribozymes. Nucleic Acids Res 19, 1183–1188.
17 Knoll R, Bald R & Furste JP (1997) Complete identifi-
cation of nonbridging phosphate oxygens involved in
hammerhead cleavage. RNA 3, 132–140.
18 Scott EC & Uhlenbeck OC (1999) A re-investigation of
the thio effect at the hammerhead cleavage site. Nucleic
Acids Res 27 , 479–484.
19 Yoshinari K & Taira K (2000) A further investigation
and reappraisal of the thio effect in the cleavage reac-
tion catalyzed by a hammerhead ribozyme. Nucleic
Acids Res 28
, 1730–1742.
20 Shan SO & Herschlag D (2000) An unconventional ori-
gin of metal-ion rescue and inhibition in the Tetrahy-
mena group I ribozyme reaction. RNA 6, 795–813.
21 Pearson RG (1968) Hard and soft acids and bases,
HSAB, part I: fundamental principles. J Chem Educ 45,
581–587.
22 Piccirilli JA, Vyle JS, Caruthers MH & Cech TR (1993)
Metal ion catalysis in the Tetrahymena ribozyme reac-
tion. Nature 361, 85–88.
23 Padgett RA, Podar M, Boulanger SC & Perlman PS
(1994) The stereochemical course of group II intron
self-splicing. Science 266, 1685–1688.
24 Podar M, Perlman PS & Padgett RA (1995) Stereochemi-
cal selectivity of group II intron splicing, reverse splicing,
and hydrolysis reactions. Mol Cell Biol 15, 4466–4478.
25 Weinstein LB, Jones BC, Cosstick R & Cech TR (1997)

154–160.
30 Suzumura K, Yoshinari K, Tanaka Y, Takagi Y, Kasai
Y, Warashina M, Kuwabara T, Orita M & Taira K
(2002) A reappraisal, based on (31)P NMR, of the
direct coordination of a metal ion with the phosphoryl
oxygen at the cleavage site of a hammerhead ribozyme.
J Am Chem Soc 124, 8230–8236.
31 Zon G & Stec WJ (1991). Oligonucleotides and Analo-
gues: a Practical Approach (Eckstein F, ed.), pp. 87–
108. IRL Press, Oxford.
32 Faulhammer D & Famulok M (1997) Characterization
and divalent metal-ion dependence of in vitro selected
deoxyribozymes which cleave DNA ⁄ RNA chimeric
oligonucleotides. J Mol Biol 269, 188–202.
33 Ota N, Warashina M, Hirano K, Hatanaka K & Taira
K (1998) Effects of helical structures formed by the
binding arms of DNAzymes and their substrates on cat-
alytic activity. Nucleic Acids Res 26, 3385–3391.
34 Ferrari D & Peracchi A (2002) A continuous kinetic
assay for RNA-cleaving deoxyribozymes, exploiting
ethidium bromide as an extrinsic fluorescent probe.
Nucleic Acids Res 30, e112–e120.
35 Breaker RR, Emilsson GM, Lazarev D, Nakamura S,
Puskarz LJ, Roth A & Sudarsan N (2003) A common
speed limit for RNA-cleaving ribozymes and deoxyribo-
zymes. RNA 9, 949–957.
36 Brautigam CA & Steitz TA (1998) Structural principles
for the inhibition of the 3¢-5¢ exonuclease activity of
Escherichia coli DNA polymerase I by phosphorothio-
ates. J Mol Biol 277, 363–377.