|
|
|
|||||||
|
|||||||||
(Received for publication, May 10, 1995; and in revised form, June 20, 1995) From the
The thrombin aptamer is a single-stranded DNA of 15 nucleotides
that was identified by the selection of thrombin-binding molecules from
a large combinatorial library of oligonucleotides. This prototype
aptamer of thrombin has a unique double G-tetrad structure capable of
inhibiting thrombin at nanomolar concentrations through binding to a
specific region within thrombin exosite I. Substitution of arginine 70
in thrombin exosite I with glutamic acid effectively eliminated binding
of the prototype thrombin aptamer. In contrast, aptamers selected
against R70E thrombin were able to bind and inhibit both wild-type and
R70E thrombins, and displayed potassium-independent inhibition.
Aptamers selected against R70E thrombin bound to sites identical or
overlapping with that of the prototype thrombin aptamer. These aptamers
retained the potential to form double G-tetrad structures; however,
these structures would be destabilized by a T
Aptamers are oligonucleotide ligands with high binding affinity
toward specific molecular targets, identified by systematic selection
and amplification of a random sequence library of nucleic acids
(Ellington and Szostak, 1990; Kenan et al., 1994). Using this
selection methodology, a single-stranded DNA sequence of 15
nucleotides, d(GGTTGGTGTGGTTGG), with inhibitory activity toward
thrombin at nanomolar concentrations was discovered (Bock et
al., 1992; Griffin et al., 1993a). NMR studies
demonstrated that this prototype aptamer for thrombin, also referred to
as the prototype, adopted a compact tertiary structure consisting of
two tetrads of guanosine base pairs and three loops with a T-T base
pairing between the two minor loops (Fig. 2D) (Wang et al., 1993a, 1993b; Macaya et al., 1993).
Figure 2:
Differences in binding and structure
between the prototype thrombin aptamer and an aptamer accommodating the
R70E mutation: a model. A, WTssP5 containing the prototype or
prototype-like structures bind to wild-type thrombin (WTIIa) at the prototype thrombin aptamer binding site,
which includes Arg-70 as a key determinant (half-circle).
WTssP5 does not bind R70E thrombin (R70E IIa) because of the
negative effects of the substitution of Arg70 with Glu (triangle). B, R70EssP5 binds R70E thrombin because
structural features that accommodate Glu70 have been recruited during
selection against R70E thrombin. These new structural features do not
prevent R70EssP5 binding to wild-type thrombin, although the affinity
is reduced. C, an aptamer accommodating the R70E mutation. D, the prototype thrombin aptamer. Arrow between
bases, 5`
The
dissociation constant, K The dominant structural features of thrombin include a deep active
site cleft and two positively charged surfaces referred to as exosites
I and II (Bode et al., 1992). Exosite I is the binding site of
multiple macromolecular substrates and ligands of thrombin including
fibrinogen, thrombomodulin, hirudin, and heparin cofactor II (Fenton et al., 1988; Tsiang et al., 1990; Rydel et
al., 1990; Sheehan et al., 1993), and exosite II is
responsible for the interaction with heparin (Church et al.,
1989; Gan et al., 1994; Sheehan and Sadler, 1994). Chemical
modification protection studies (Paborsky et al., 1993) and
site-directed mutagenesis of thrombin (Wu et al., 1992; Tsiang et al., 1995) defined the prototype thrombin aptamer binding
site as a discrete region within thrombin exosite I and identified
arginine 70 as a key residue required for interaction with the
prototype thrombin aptamer. The solution of the crystal structure of
the prototype thrombin aptamer complex with thrombin indicated that the
prototype may interact with both exosites I and II of thrombin
(Padmanabhan et al., 1993). However, thrombin exosite II
mutants were susceptible to inhibition by the prototype (Tsiang et
al., 1995), and analysis of the prototype thrombin aptamer binding
to exosite I mutant, R70A, by surface plasmon resonance
spectroscopy The prototype thrombin aptamer is an inhibitor
of both the procoagulant and the anticoagulant functions of thrombin
(Wu et al., 1992; Griffin et al., 1993a, 1993b; Li et al., 1994). Thrombin exerts its main procoagulant function
by cleaving soluble fibrinogen, which then forms a fibrin clot. When
bound to thrombomodulin, thrombin changes its substrate specificity to
activate protein C, the activated form of which is a major
physiological anticoagulant. The demonstration that thrombin residues
involved in fibrinogen clotting and thrombomodulin binding can be
dissociated (Wu et al., 1991) raised interest as to whether
aptamers could be selected to target a different region on thrombin
than the prototype binding site or to even have inhibitory activities
that could discriminate the procoagulant from the anticoagulant
functions of thrombin. In an effort to probe these possibilities and to
test whether non-G-tetrad structures could be selected (Griffin and
Vermaas, 1995), we conducted aptamer selections using as a target the
R70E thrombin, which was highly refractory to inhibition by the
prototype. Aptamers selected against R70E thrombin were not targeted to
a new binding site but instead accommodated the R70E mutation on
thrombin.
In an attempt to further reduce the diversity of the
pool by selecting for species with higher affinity and slower off rate,
two more rounds of selection were performed under competitive
conditions where the thrombin concentration was not in excess of the
concentration of ssDNA (see ``Experimental Procedures''). The
inhibitory activity of the resultant pools, WTssP6 and WTssP7, at 573
nM toward wild-type thrombin in fibrinogen clotting increased
to 80-90% but remained very low toward R70E thrombin, indicating
that the discriminating property of the pool was essentially unaltered (Table 1). The same tendency was also evident in the protein C
assay. The inhibitory activity of R70EssP6 and R70EssP7 toward both
wild-type and R70E thrombin increased to close to 40% in fibrinogen
clotting and to 26% and 43% toward wild-type and R70E thrombins,
respectively, in the protein C assay. These results showed that the
decrease in protein C activation inhibitory activity for WTssP5 and
R70EssP5 after the fifth round was only transitory and that better
inhibitory sequences eventually emerged after selection under more
stringent conditions.
Since the inhibition assays
suggested further accommodation of the R70E mutation by R70EssP7 after
two additional rounds of selection under more stringent conditions, we
also sequenced ssP7 pools (Fig. 1). Out of 19 clones from
R70EssP7, 13 clones had the same consensus double G-tetrad core
sequence as the clones from R70EssP5 (consensus 70ssP7 core 1) (Fig. 1B). However, five clones (1, 2, 6, 10, and 15)
had a different consensus sequence (consensus 70ssP7 core 2). This
second consensus core sequence had a longer major
loop,(N)
Figure 1:
Sequence of pools after seven rounds of
selection. A, sequence of WTssP7. B, sequence of
R70EssP7. Numbering of the positions is shown under the sequence of the
prototype. The last two rounds of selection were under limiting target
conditions. The G residues that can potentially form G-tetrad base
pairings are in boldface type. Bases in either the 5`- or
3`-flanking regions that can potentially base pair to form a stem
structure are underlined. Bases that cannot base pair are
represented as N. When the flanking sequences cannot form
recognizable duplex regions, they are not shown. The consensus core
sequences for each selection are shown, where N represents
variable bases.
Variant 4 is
identical to the double G-tetrad core sequence of clone 7 from
R70EssP5. The fact that it inhibited neither wild-type nor R70E
thrombins significantly suggested that sequences outside of the double
G-tetrad core sequence might also contribute to stabilize the binding
structure. To test this hypothesis, we assayed variants found in clones
of R70EssP5, with sequences capable of duplex formation flanking the
double G-tetrad core. Variant 5 (clone 5) had an inhibitory activity
very similar to that of the entire pool R70EssP5. When the flanking
sequences of variant 5 were removed in variant 9, inhibitory activity
was essentially abolished, suggesting that the duplex region may have a
stabilizing effect on the binding structure.
When the prototype thrombin aptamer with a double G-tetrad
structure was first discovered, it was not clear whether the emergence
of this particular sequence and structure reflected a single
aptamorigenic determinant on thrombin and a single aptamer consensus
sequence or whether other sequences and aptamorigenic sites existed.
After the initial discovery of the prototype thrombin aptamer,
selection against wild-type thrombin was repeated with several other
types of single-stranded DNA libraries (data not shown). All these
libraries consistently yielded the same consensus sequence, of which
the prototype thrombin aptamer was the highest affinity representative.
This result implied that the difference between the consensus sequences
of WTssP5 and R70EssP5 was not incidental. To probe for the
possibility of aptamer selection by an alternative site on thrombin, we
substantially reduced the affinity of the binding site for the
prototype thrombin aptamer by a non-conservative substitution of
arginine 70 with glutamic acid. Aptamer selection against exosite I
mutant R70E thrombin generated (after five rounds) a sublibrary,
R70EssP5, that had drastically different inhibitory and binding
properties than WTssP5, the corresponding sublibrary generated against
wild-type thrombin. Consistent with the inhibitory and binding
properties of the prototype thrombin aptamer, WTssP5 was only able to
bind and inhibit wild-type thrombin but not R70E thrombin because of
the R70E mutation in its binding site (Fig. 2A). In
contrast, R70EssP5, which was the result of selection against the
mutant thrombin, was able to bind to essentially the same original
binding site and inhibit the mutant thrombin through accommodation of
the R70E mutation (Fig. 2B). This accommodation of
Glu-70 in the mutant thrombin did not affect the ability of R70EssP5 to
bind or inhibit wild-type thrombin. In fact, R70EssP5 was a better
binder and inhibitor of wild-type thrombin than R70E thrombin (Fig. 2B). This could reflect a difference between
arginine and glutamic acid in their interactions with nucleic acids,
with the more acidic residue being less conducive to aptamer selection
from random nucleic acid libraries. With two more rounds of selection
under more stringent conditions, further accommodation of the R70E
mutation occurred as demonstrated by an increase in the inhibitory
activity of R70EssP7 toward R70E thrombin and to a lesser extent toward
wild-type thrombin. Sequence analysis of clones selected against
R70E thrombin revealed again a double G-tetrad consensus sequence,
however, with a key difference consisting of an T This
accommodation of R70E thrombin by the aptamer, restoring binding
affinity during selection, can be likened to extragenic suppressor
mutations that are widespread in nature. In most cases, an extragenic
suppressor mutation arises in a second gene whose product interacts
physically with the product of the originally mutated gene (Jarvik and
Botstein, 1975). Examples of this kind of extragenic suppressor
mutation can be found in the fowl plague virus (Mucke and Scholtissek,
1987), E. coli (Osborne and Silhavy, 1993), and Saccharomyces cerevisiae (Yano et al., 1992). By
taking this concept one step further, the amino acid or nucleotide
substituted in a suppressor mutation may in some cases be in physical
contact with, or in close proximity to, the site of the original
mutation. In such cases, the selection of suppressor mutations can be
applied to identify specific intermolecular interactions as in the
example of the repressor-operon interaction in the Salmonella phage P22 (Youderian et al., 1983). Recently, an
example of intragenic suppression, termed covariation, was found to
preserve a non-Watson-Crick base pairing and Rev responsiveness, in the
human immunodeficiency virus type 1 Rev-responsive element during an
RNA aptamer selection (Bartel et al., 1991). However, the
accommodation we observed in this study is the first example of an
extragenic suppression involving nucleic acid-protein interaction by in vitro genetics. Our aptamer selection using wild-type or
R70E thrombins as targets suggests that unlike antigenic epitopes,
aptamorigenic domains are not widespread on the thrombin molecular
surface, which instead contains only one discrete region of higher
aptamorigenicity. The primary aptamorigenic site on thrombin with
respect to single-stranded DNA is the prototype thrombin aptamer
binding site in thrombin exosite I, which precluded efficient selection
of aptamers binding to other sites. Charge reversal of a key residue
within this site only mildly decreased its aptamorigenicity and led
instead to the selection of aptamers that accommodated the mutation.
Recently, RNA aptamers of thrombin with nanomolar binding affinity have
also been generated after 12 rounds of selection (Kubik et
al., 1994). The RNA thrombin aptamers had a unique hairpin
structure and, in contrast to DNA thrombin aptamers, bound to exosite
II of thrombin. This observation further confirms the idea that the
existence and location of a primary aptamorigenic site on a target
protein is a function of the properties of both the target and the
nucleic acid library.
Volume 270,
Number 33,
Issue of August 18, pp. 19370-19376, 1995
©1995 by The American Society for Biochemistry and Molecular Biology, Inc.
ABSTRACT
INTRODUCTION
EXPERIMENTAL PROCEDURES
RESULTS
DISCUSSION
FOOTNOTES
ACKNOWLEDGEMENTS
REFERENCES
A substitution,
disrupting the T
-T base pairing found in the
prototype. This destabilization appeared to be partially compensated by
newly recruited structural elements. Thus, selection against R70E
thrombin did not lead to aptamers that bound to alternative sites, but
instead to ssDNA structures with a suppressor mutation that
accommodated the mutation in thrombin within a double G-tetrad context.
These results provide insight into the aptamer-thrombin interaction and
suggest that the binding site for the prototype is the dominant
aptamorigenic site on thrombin.
3` phosphodiester linkage; straightbrokenline, hydrogen bonding interaction (when crossed, it indicates an altered or weakened hydrogen bonding
interaction with respect to the prototype); curvedbrokenline, variable length region; curved arrow,
sequences flanking the double G-tetrad core. The T
-T base pairing in the prototype structure is shown. This hydrogen
bonding interaction is absent in the accommodating aptamer structure
because of a substitution of either T
or T by
adenosine. The disruption of this T-T base pairing alters or weakens
the double G-tetrad structure with loss of potassium ion dependence. A
longer major loop and the presence of duplex stem may facilitate the
adoption of an altered G-tetrad structure as well as provide structural
stabilization and additional contact points with thrombin. The overall
effect of these changes is the ability of the new structure to
accommodate the R70E mutation.
, for the
prototype thrombin aptamer interaction with thrombin has been
determined to range from 1.4 to 6.2 nM by various methods (Wu et al., 1992; Davis, 1994; Griffin and Leung, 1995). (
)In addition, this prototype thrombin aptamer has
demonstrated potent anticoagulant properties in both monkey and sheep
(Griffin et al., 1993a; Griffin et al., 1993b). indicated that no significant binding outside
of exosite I existed.
Materials
Plasma thrombin, protein C, and rabbit
thrombomodulin were from Haematologic Technologies (Essex Junction,
VT). R70E prothrombin was purified from the conditioned media of a
stably transfected BHK-21 line expressing the mutant prothrombin as
described previously (Wu et al., 1991). R70E prothrombin was
processed to the thrombin form (abbreviated to IIa in all the tables)
by Oxyuranusscutellatus venom and purified by
Amberlite CG-50 ion-exchange chromatography (Fenton et al.,
1977).Aptamer Selection
The starting pool was a library
of 97-mer oligodeoxyribonucleotides containing a 60-nucleotide random
region flanked by constant regions for primer binding. The diversity of
the starting pool was 4 10
. The selection cycle
involving binding to a thrombin affinity column and polymerase chain
reaction amplification of eluted sequences has been described (Bock et al., 1992). For the first five rounds,
750 pmol of
single-stranded DNA were applied to
4.5 nmol of thrombin bound to
1 ml of concanavalin A-agarose. For rounds 6 and 7,
750 pmol of
single-stranded DNA were applied to
0.1 nmol of thrombin bound to
0.15 ml of concanavalin A-agarose.
Binding Assay
The dissociation constants of the
selected ssDNA (
)pools were determined in a binding assay
using microtiter wells coated with thrombin. Maleic anhydride-activated
microtiter plates (Pierce) were first coated with 15 µM Phe-Pro-Arg-chloromethyl ketone (PPACK) (Sigma) in PBS. After
blocking with 3% BSA and 0.05% Tween in PBS, thrombin in selection
buffer (20 mM Tris acetate, pH 7.5, 140 mM NaCl, 5
mM KCl, 1 mM MgCl, 1 mM CaCl) was added at 750 nM. Immobilization of
thrombin with PPACK as a linker between thrombin and the plate,
resulted in a uniform coating geometry. Unreacted thrombin was removed
by washing the wells with 0.1% BSA, 0.05% Tween in PBS. Single-stranded
DNA uniformly labeled with [
-P]dNTPs was
added to the wells at concentrations of 0-573 nM in
selection buffer and incubated for 2 h at room temperature with
constant rocking. After 2 h, the unbound DNA was removed and the wells
were washed with selection buffer, broken apart and individually
counted. For competition binding, the wells were coated with 1500
nM thrombin, labeled ssDNA was used at 4 nM and
unlabeled ssDNA was used at concentrations of 0-1850 nM.
Inhibition of Fibrinogen Clotting
Plasma thrombin
or R70E thrombin at 18.75 nM were incubated with ssDNA in 200
µl of selection buffer at 37 °C for 1 min. Clotting was
initiated by addition of 50 µl of 2 mg/ml human fibrinogen freshly
diluted in selection buffer from a stock of 10 mg/ml made in
calcium-free PBS. Time in seconds from addition of fibrinogen to clot
formation was measured with a fibrometer. For potassium-independent
inhibition, KCl was omitted from selection buffer and
KHPO was substituted with
NaHPO in the PBS used to make the fibrinogen
stock solution.Inhibition of Protein C Activation
The reaction
contained 3.7 nM plasma thrombin or R70E thrombin, a variable
amount of ssDNA, 2 nM rabbit thrombomodulin, and 887 nM protein C in a total volume of 50 µl. Thrombin and ssDNA were
mixed in 12.5 µl of selection buffer and incubated at room
temperature for 6 min. To start the reaction, 13 µl assay buffer
(50 mM Tris-HCl, pH 8.0, 2 mM CaCl, 100
mM NaCl, 0.1% BSA), 20 µl of rabbit thrombomodulin in
assay buffer, and 4.5 µl of protein C in assay buffer were added
sequentially. The reaction was incubated at 37 °C for 1 h and
stopped by addition of antithrombin III and heparin. The activated
protein C generated was assayed by hydrolysis of the chromogenic
substrate S-2366 (PyrGlu-Pro-Arg p-nitroanilide).Cloning and Sequencing
Selected single-stranded
pools were polymerase chain reaction-amplified in the presence of 5`
and 3` cloning primers containing, respectively, a BamHI site
and an EcoRI site. The resulting double-stranded DNA pool was
cloned into the phagemid vector pUC218 using these two restriction
sites. Ampicillin-resistant, XL1-blue Escherichia coli colonies were picked, and single-stranded DNAs for sequencing were
grown in the presence of the helper phage M13K07. Sequencing was by
dideoxy chain termination, using Sequenase (U. S. Biochemical Corp.).
Selection against Wild-type and R70E
Thrombins
In an attempt to identify aptamers that bound to
alternative sites on thrombin and to further understand how selected
ssDNAs interact with thrombin, we conducted parallel selections against
wild-type thrombin and R70E thrombin, which does not bind the
prototype. In the first round of selection, very few input sequences
bound to either wild-type (0.015%) or R70E (0.016%) thrombins. By the
second round, wild-type and R70E thrombins could be distinguished, with
R70E thrombin binding 30-fold less (0.063%) of the input sequences
than the wild-type (1.906%). However, the few sequences bound to R70E
thrombin were quickly enriched to a level (11.221%) similar to that of
wild-type thrombin (14.479%) by the third round. This suggested that
either sequences capable of binding R70E thrombin were present at a
lower frequency in the starting pool than sequences capable of binding
wild-type thrombin or that the sequences present bound R70E thrombin
with a lower affinity. In either case, R70E thrombin was less capable
of serving as a substrate for aptamer selection using single-stranded
DNA libraries. We describe this property of the target molecule with
respect to a given nucleic acid library as aptamorigenicity. Since
nucleic acid ligands selected from combinatorial libraries have been
compared to monoclonal antibodies (Edgington, 1993), the
aptamorigenicity of a target molecule is analogous to the antigenicity
of an antigen molecule, i.e. its ability to elicit recognition
by antibodies.
Binding Properties of Aptamer Pools Selected against
Wild-type and R70E Thrombins
We tested the ability of pools of
single-stranded DNA after five rounds of selection (ssP5) to bind
either wild-type or R70E thrombins. The pool selected against wild-type
thrombin (WTssP5) bound wild-type thrombin with a K of 40 ± 6 nM but had very little affinity
for R70E thrombin, with a K
> 1.2
mM. In contrast, the pool selected against R70E thrombin
(R70EssP5) bound both R70E thrombin and wild-type thrombin with K
values of 123 ± 37 nM and 45 ± 10 nM, respectively. This result clearly
shows that R70EssP5 is different from WTssP5. Next, WTssP5 and R70EssP5
were compared in a competition binding to wild-type thrombin. WTssP5
and R70EssP5, each competed with itself with K
values of 51 ± 5 nM and 87 ± 15
nM, respectively, consistent with those determined in the
direct binding assay. Moreover, WTssP5 and R70EssP5 competed with each
other for binding wild-type thrombin with K
values of 26 ± 3 nM and 62 ± 5
nM, similar to their respective K
values. This demonstrated that WTssP5 and R70EssP5 bind to
the same or an overlapping site on wild-type thrombin. This result was
confirmed by competition with the prototype thrombin aptamer for
binding wild-type thrombin. Both WTssP5 and R70EssP5 competed with the
prototype with an IC
of
150 ± 7 nM,
showing in addition that the binding site of R70EssP5 is identical or
overlapping with that of the prototype. Therefore, instead of selecting
aptamers that bound to another site on thrombin, structural
accommodation of the R70E mutation must have occurred in R70EssP5 to
offset the negative effect of the thrombin mutation.
Inhibitory Activities of Aptamer Pools Selected against
Wild-type and R70E Thrombins
The inhibitory activities of the
pools selected against wild-type and R70E thrombins were assessed in
both fibrinogen clotting and protein C activation assays (Table 1). The prototype thrombin aptamer was used as a positive
control. WTssP4 and WTssP5 at 573 nM inhibited wild-type
thrombin by 55% in the fibrinogen clotting assay but, like the
prototype, had essentially no inhibitory activity toward R70E. In the
protein C assay, WTssP4 and WTssP5 displayed the same pattern of
inhibition, although the inhibition of wild-type thrombin was less
pronounced than in the clotting assay. Therefore, WTssP4 and WTssP5
behaved just like the prototype. In contrast, R70EssP4 and R70EssP5 at
573 nM inhibited both wild-type and R70E thrombins by
20-28% in the clotting assay. This result is consistent with the
ability of R70EssP5 to bind both wild-type and R70E thrombins. In the
protein C assay, R70EssP4 and R70EssP5 inhibited R70E thrombin to a
greater extent than wild-type thrombin, possibly because the R70E
mutation also decreased the affinity for thrombomodulin (Wu et
al., 1991; Tsiang et al., 1995) and, as a result, allowed
for a more effective competition of the aptamer pool against
thrombomodulin.
Sequence of Pools Selected against Wild-type or R70E
Thrombins
WTssP5 and R70EssP5 were cloned (data not shown). All
12 clones from WTssP5 contained a consensus core sequence with a double
G-tetrad motif similar to that of the prototype: GG (N)T GG(N)
GG (N)
T GG. In addition, sequences flanking
the double G-tetrad core of all the clones can potentially base pair to
form a more or less rigid duplex stem. These double G-tetrad motifs are
consistent with the sequences found in previous pools selected against
wild-type thrombin (Bock et al., 1992). The 11 clones from
R70EssP5 also displayed a consensus core sequence with a double
G-tetrad motif: GG TA GG (N)
GG(N)
T GG (data not shown).
Strikingly, all the positions corresponding to T
of the
prototype, located in the first minor loop, were occupied by adenosine
instead of thymidine. Again, sequences flanking the double G-tetrad
core of most clones could potentially form a duplex stem. These
observations suggested that an adenosine at a position corresponding to
T in the prototype might play a role in the accommodation
of the R70E mutation by the aptamer. and, interestingly, did not have an
adenosine at the position corresponding to T
of the
prototype but instead had an adenosine at the position corresponding to
T of the prototype. In contrast, these two positions were
all occupied by thymidines in the consensus core sequences of both
WTssP5 (see text above) and WTssP7 (Fig. 1A). Since
these two positions were formerly occupied by two thymidines that base
pair in the prototype to stabilize its double G-tetrad structure (Wang et al., 1993b), these results suggested that disruption of
this base pairing by an adenosine substitution at either of these
positions was critical in accommodating the R70E mutation in thrombin.
Another trend after limiting target selection concerned the major loop.
While its length decreased to an optimum of 3 nucleotides in WTssP7,
its average length increased in R70EssP7, suggesting that a longer
major loop might also contribute to accommodation.
Inhibitory Activities of Variants of the Prototype
Thrombin Aptamer
In order to understand which elements in the
sequences of R70EssP5 contributed to accommodation of the R70E
mutation, the ability of various synthetic oligonucleotides containing
the double G-tetrad motif to inhibit fibrinogen clotting was determined (Table 2). Substitution of the guanosine in the major loop of the
prototype with thymidine in variant 1b had only a slight effect (8%
drop in activity), as expected from the consensus WTssP7 core sequence (Fig. 1A). When thymidine at position 4 of the
prototype was substituted with adenosine in variant 1 and variant 3,
inhibitory activity toward wild-type thrombin was essentially abolished
without any gain in inhibitory activity toward R70E thrombin. In
contrast, adenosine substitution at position 12 in variant 2 had little
effect as expected from the consensus WTssP7 core sequence. These
observations suggested that an adenosine substitution at position 4 of
the prototype had a destabilizing effect on the aptamer structure
required for binding wild-type thrombin and that it alone was
insufficient to accommodate the R70E mutation.
Potassium Dependence of WTssP5 and R70EssP5 Inhibitory
Activity
Potassium ions have been reported previously to
stabilize G-tetrads in DNA tetraplex structures (Guschlbauer et
al., 1990) and in the prototype thrombin aptamer structure, which
is necessary for its inhibitory activity (Wang et al., 1993b).
To assess the role of potassium ions on the inhibitory activity of
WTssP5 and R70EssP5, we prepared the single-stranded DNA pools in the
presence or absence of potassium ions and assayed them in wild-type
thrombin catalyzed clotting reactions with and without potassium ions,
respectively (Table 3). When potassium ions were absent, sodium
ions were present as substitutes. WTssP5, like the prototype, was
55-67% less inhibitory in the absence of potassium ions. This is
consistent with the presence of prototype-like core structures in
WTssP5. In contrast, R70EssP5 was equally inhibitory whether in the
presence or absence of potassium ions, suggesting that the inhibitory
structures in R70EssP5 did not require potassium ions for
stabilization. In addition, the oligonucleotide variant 5, which was
derived from clone 5 of R70EssP5 also displayed potassium-independent
inhibition.
A substitution
at positions corresponding to the only two conserved T bases, either
T
or T, in the prototype thrombin aptamer
(Bock et al., 1992). The NMR structure of the prototype
thrombin aptamer revealed that T
and T are
involved in hydrogen bonding interactions with each other and base
stacking interactions with the G-tetrads (Macaya et al., 1993;
Wang et al., 1993b). This suggested that an adenosine
substitution could destabilize the prototype structure as evidenced in
the loss of inhibitory activity (Table 2) and play a key role in
the accommodation of the R70E mutation in thrombin by altering the
prototype structure through loss of the T
-T base pairing (Fig. 2C). A simpler and
non-mutually exclusive interpretation of this observation might be that
T
or T contact arginine 70 in the
aptamer-thrombin complex. However, the crystal structure of the
aptamer-thrombin complex (Padmanabhan et al., 1993) seemed to
suggest otherwise. This destabilizing component of the aptamer
structure appeared to be compensated by other potentially stabilizing
features compatible with the R70E mutation including a duplex region
and a larger major loop (Fig. 1), all of which may increase the
binding free energy to R70E thrombin (Fig. 2C). The
potassium independence of R70EssP5 inhibition may indicate that the
compensating structures had a greater contribution to overall
structural stability than the double G-tetrad structure. This
interpretation does not exclude the formation of an altered double
G-tetrad structure with a different metal ion dependence (Hardin et
al., 1992). Overall, accommodation of the R70E mutation was
accompanied by an increase in the average complexity of the sequences
selected. Because these accommodating sequences were of lower affinity
and less abundant than the prototype sequence, they would not be
observed in a selection against wild-type thrombin.
))
We thank Terry Terhorst, Kim Sweetnam, and Cathy
Sueoka for the synthesis of the oligonucleotides used in this study and
Lisa Paborsky and Jay Toole for critical comments on this manuscript.
©1995 by The American Society for Biochemistry and Molecular Biology, Inc.
CiteULike 
Complore 
Connotea 
Del.icio.us 
Digg 
Reddit 
Technorati What's this?
This article has been cited by other articles:
|
|
 |
|
  K. Ikebukuro, Y. Okumura, K. Sumikura, and I. Karube A novel method of screening thrombin-inhibiting DNA aptamers using an evolution-mimicking algorithm Nucleic Acids Res., July 7, 2005; 33(12): e108 - e108. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 M. Gullberg, S. M. Gustafsdottir, E. Schallmeiner, J. Jarvius, M. Bjarnegard, C. Betsholtz, U. Landegren, and S. Fredriksson Cytokine detection by antibody-based proximity ligation PNAS, June 1, 2004; 101(22): 8420 - 8424. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 A. A. K. Hasan, M. Warnock, M. Nieman, S. Srikanth, F. Mahdi, R. Krishnan, A. Tulinsky, and A. H. Schmaier Mechanisms of Arg-Pro-Pro-Gly-Phe inhibition of thrombin Am J Physiol Heart Circ Physiol, June 5, 2003; 285(1): H183 - H193. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 T. Myles, T. H. Yun, and L. L. K. Leung Structural requirements for the activation of human factor VIII by thrombin Blood, September 26, 2002; 100(8): 2820 - 2826. [Abstract] [Full Text] [PDF]
|
|
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
| All ASBMB Journals | Molecular and Cellular Proteomics |
| Journal of Lipid Research | ASBMB Today |