©
1996 Oxford University Press
859-867 Footnote
Bis-intercalation of a homodimeric thiazole orange dye in DNA in symmetrical
pyrimidine-pyrimidine-purine-purine oligonucleotides
Bis-intercalation of a homodimeric thiazole orange dye in DNA in symmetrical pyrimidine-pyrimidine-purine-purine oligonucleotides
Lene F.
Hansen
,
Lisbeth K.
Jensen
and
Jens Peter
Jacobsen*
Department of Chemistry, Odense University,
Odense
M, DK-5230
Denmark
Received november 27, 1995;
Revised and Accepted January 10, 1996
ABSTRACT
One- and two-dimensional
1
H NMR spectroscopy were used to characterize the binding of a homodimeric
thiazole orange dye, 1,1
'
-(4,4,8,8-tetramethyl-4,8-diaza-undecamethylene)-bis-4-(3-methyl-2,3-dihydro-(benzo-1,3-thiazole)-2-methylidene)-quinolinium tetraiodide (TOTO), to various double-stranded DNA oligonucleotides containing symmetric (5
'
-pyr-pyr-pu-pu-3
'
)
2
or (5
'
-pu-pu-pyr-pyr-3
'
)
2
sequences. It was found that TOTO binds preferentially to oligonucleotides
containing a (5
'
-CTAG-3
'
)
2
or a (5
'
-CCGG-3
'
)
2
sequence. Binding to the (5
'
-CCGG-3
'
)
2
sequence is less favored than to the (5
'
-CTAG-3
'
)
2
sequence. The complexes of TOTO with d(CGCTAGCGCTAGCG)
2
(
10
) and d(CGCTAGCCGGCG):d(CGCCGGCTAGCG) (
11
) oligonucleotides, each containing two preferential binding sites, was also
examined. In both cases TOTO forms mixtures of 1:1 and 1:2 dsDNA-TOTO complexes in ratios dependent on the relative amount of TOTO and the
oligonucleotides in the sample. Binding of TOTO to the two oligonucleotides is
sequence selective at the (5
'
-CTAG-3
'
)
2
and (5
'
-CCGG-3
'
)
2
sites. The
1
H NMR spectra of both the 1:2 complexes and the three different 1:1 complexes
have been assigned. A slight negative cooperativity is observed in formation of
the 1:2 complexes. The ratio between the two different 1:1 complexes formed
with oligonucleotide
11
is 2.4 in favor of binding to the (5
'
-CTAG-3
'
)
2
site. This is very similar to results obtained when the two sites are in
different oligonucleotides. Thus the distribution of TOTO among the (5
'
-CTAG-3
'
)
2
and (5
'
-CCGG-3
'
)
2
sites is independent of whether the two sites are in the same or two different
oligonucleotides.
INTRODUCTION
Interest in non-radioactive detection of nucleic acids in gels has led to the synthesis
and evaluation of new dyes that form highly fluorescent complexes with double-stranded DNA (dsDNA). Glazer and co-workers (
1
-
3
) have synthesized and characterized a whole new family of high affinity fluorescent dyes forming dsDNA complexes.
Among these, a homodimeric thiazole orange dye, 1,1'-(4,4,8,8-tetramethyl-4,8-diazaundecamethylene)-bis-4 -(3- methyl-2,3-dihydro-(benzo-1,3-thiazole)-2-methylidene)-quinolinium tetraiodide (TOTO, Scheme
1
), binds extremely strongly, but non-covalently, to dsDNA. It forms complexes that are stable under a variety
of conditions, in particular gel electrophoresis. Enhancement of the
fluorescence quantum yield upon binding of TOTO to dsDNA is >3000-fold. Thus this compound allows high sensitivity detection of dsDNA by
fluorescence scanners (
4
,
5
).
The extremely high binding affinity of TOTO for dsDNA has prompted us to carry
out NMR studies to characterize the binding mode. Previously we reported (
6
) a
1
H NMR investigation of dye binding to a number of dsDNA oligomers (from 4 to 12
bp in length). Line broadening arising from chemical exchange between various
binding sites was observed for many of the sequences used in the investigation.
However, sharp lines were obtained for oligomers containing the (5'-CTAG-3')
2
sequence and we were able to show that this sequence constitutes the most
favorable binding site for the dye among the particular sequences that we more
or less randomly chose.
In order to clarify the structural basis for this sequence-selective binding and the origin of the fluorescence enhancement we
reported an NMR structural study of TOTO bound to the oligo- nucleotide d(CGCTAGCG)
2
(
7
). The structure of this dsDNA-TOTO complex revealed that a characteristic feature of the TOTO
chromophore is its ability to adapt to the base pair propeller twist of the
dsDNA, with the benzothiazole ring sandwiched between the pyrimidines and the
quinolinium ring between the purines.
Based on our knowledge of dsDNA-TOTO structure it became obvious that there is a structural preference
for each of the two TOTO chromophores for binding sites containing (5'-pyr-pyr-3'):(5'-pu-pu-3') nucleobases.
Since TOTO is a symmetrical bis-intercalator we have undertaken a systematic investigation of the binding
of TOTO to all symmetrical dsDNA sequences of the (5'-pyr-pyr-pu-pu-3')
2
or (5'-pu-pu-pyr-pyr-3')
2
types in order to clarify to what extent TOTO prefers the (5'-CTAG-3')
2
sequence compared with other sequences of these types.
The very strong affinity of TOTO for dsDNA has prompted us also to investigate
the possible existence of cooperativity in selective binding. We have studied
the binding of TOTO to dsDNA oligonucleotides containing two selective binding
sites, one of them a 14 bp oligonucleotide containing two (5'-CTAG-3')
2
binding sites (Scheme
2
), the other a 12 bp oligonucleotide containing both the (5'-CTAG-3')
2
and (5'-CCGG-3')
2
binding sites (Scheme
2
). The results are discussed in terms of both cooperativity and selectivity.
Scheme 1
Scheme 2
The oligonucleotieds used in this work and some of their TOTO complexes with
the numbering scheme used.
MATERIALS AND METHODS
Materials
Purified DNA oligonucleotides were purchased from DNA Technology (Aarhus, Denmark) and used without further purification. The non-self-complementary single-stranded oligomers were added to an equivalent amount of the
complementary strand and duplexes formed by annealing at 80o C. The dsDNA oligonucleotides used in this work are shown in Scheme
2
.
TOTO is almost insoluble in water and complexes with dsDNA can therefore not be
made by simple titration of the dsDNA with TOTO. Instead, complexes were formed
by the procedure described earlier (
6
,
7
).
The NMR samples were prepared by dissolving the complexes in 0.5 ml 10 mM sodium
phosphate buffer, pH 7.0, 0.05 mM EDTA. For experiments carried out in D
2
O the solid complex, lyophilized twice from D
2
O, was redissolved in 99.96% D
2
O (Cambridge Isotope Laboratories). A mixture of 90% H
2
O and 10% D
2
O (0.5 ml) was used for experiments examining exchangeable protons. The sample
was kept in an NMR tube under nitrogen. The final concentrations of the
complexes were between 1 and 3 mM.
The complexes of the 8 or 9 bp oligonucleotides were made as 1:1 complexes of
dsDNA and TOTO, but often the complex was made in steps, yielding first the 2:1
dsDNA-TOTO product. Complexes with oligonucleotide
10
10
11
11
10
10
2
stream.
Methods
One- and two-dimensional NOESY spectra in H
2
O and D
2
O, as well as two-dimensional TOCSY spectra in D
2
O, were obtained for all the free oligonucleotides and their TOTO complexes
mentioned above. All NMR experiments were performed at 500 MHz on a Varian
Unity 500 spectrometer. The NOESY spectra were acquired with various mixing
times of 50-200 ms in D
2
O using 1024 complex points in t
2
and a spectral width of 5000 Hz. t
1
experiments (512) were recorded using the states phase cycling scheme. Normally
64 scans were acquired for each t
1
value. The TOCSY spectra were obtained with mixing times of 30 and 90 ms in the
TPPI mode using 1024 complex points in t
2
, 512 t
1
experiments and by acquiring 64 scans for each t
1
value. The NOESY spectra in H
2
O were acquired with a spectral width of 10 000 Hz in 2048 complex points using a pulse sequence where the last 90o pulse was replaced by a pulse containing a notch to suppress the solvent
signal (
8
). Spectra were obtained at 25 or 10o C. The acquired data were processed using FELIX (version 2.3; Biosym
Technologies, San Diego, CA). The TOCSY and NOESY spectra were assigned by
conventional methods (
9
-
14
) as described earlier (
6
,
7
). The integration of the signals in the imino region was performed by the line
fitting procedure in FELIX.
RESULTS
The dsDNA oligonucleotides containing a symmetrical sequence of the (5'-pyr-pyr-pu-pu-3')
2
or (5'-pu-pu-pyr-pyr-3')
2
type and used in this work are listed in Scheme
2
. The choice between 8 or 9 bp was originally made to minimize any possible
assignment problems in the spectra of the complexes, but it turned out to have
no practical influence on the results. The aromatic parts of the
1
H NMR spectra of the complexes of TOTO with the various oligonucleotides in D
2
O are shown in Figure
1
. All samples were made using exactly the same procedure. Obviously, the NMR
spectra show that very different complexes between TOTO and the
oligonucleotides have been formed. In all cases comparison with the spectra of
the free dsDNA prove that TOTO actually binds to each of the oligonucleotides,
but yields a quite different number of complexes.
Figure 1
.
The aromatic and H(1') region of the one-dimensional spectrum obtained in D
2
O of the 1:1 dsDNA-TOTO complexes formed by the following oligonucleotides: (
A
)
1
B
)
2
C
)
3
D
)
4
E
)
5
F
)
6
G
)
7
H
)
8
Figure 2
.
The imino proton region of the one-dimensional spectrum obtained in H
2
O of the mixture of
4
9
9
4
The TOTO complexes with oligonucleotide
1
5
6
7
More than one complex, probably two, each with fairly sharp lines, is formed
when TOTO binds to either
2
3
The one-dimensional spectra of the TOTO complexes with
8
9
2
site. The spectral assignments of both the
8
9
6
,
7
).
The one-dimensional spectra of the TOTO complex with
4
8
9
4
8
9
2
site. The cross-peak pattern in the NOESY spectra of the
4
8
9
11
11
vide infra
).
Mixture of oligonucleotides
d(CGCTAGCG)
2
(
9
) and d(CGCCGGCG)
2
(
4
)
In order to establish the relative preference of TOTO for the (5'-CCGG-3')
2
and the (5'-CTAG-3')
2
sequences we performed a competition study between oligonucleotides
4
9
9
4
2
O is shown in Figure
2
. Resonance lines from
9
4
4
9
2
O and in D
2
O, partly by adding yet another portion of
9
The equilibrium between the two complexes
4
9
9
4
has an equilibrium constant defined by
K
= {[
9
4
4
9
The value of
K
was measured from the integral of the imino proton NMR signals shown in Figure
2
. An averaged value of
K
= 6.2 +- 1.2 at 10o C was obtained from spectra of the two mixture concentrations. The
ratio between the
9
4
4
9
Oligonucleotide d(CGCTAGCGCTAGCG)
2
(
10
)
The NMR spectra of the free oligonucleotide
10
2
) in ratios depending on the amount of TOTO added. A characteristic feature of the
1
H NMR spectra of
10
2
binding site (
6
). However, it is not easy from the one-dimensional spectra to distinguish between 1:1 and 1:2
10
10
The 1:2
10
-TOTO complex.
Part of the NOESY spectrum of the 1:2
10
3
. The NOESY spectrum exhibits the characteristic features of dsDNA sequential
connectivities from aromatic H6/H8 protons to both intra- and inter-residue H1'. The sequential connectivities are interrupted at all the 5'-CpT-3' and 5'-ApG-3' base
pair steps. This is a clear evidence of bis-intercalation in both (5'-CTAG-3')
2
sites. The interruptions of the sequential NOE connectivities at the
intercalation sites are also observed in the cross-peak patterns of the two methyl groups of the thymidines, for which the
peaks to C3 H6 and C9 H6, respectively, are missing. The assignments of the
aromatic (H6, H5, H8 and H2) and deoxyribose (H1', H2', H2'', H3' and H4') proton resonances of the dsDNA in
the complex are given in Table
1
and compared with the free dsDNA oligonucleotide. The NOESY spectrum of the
complex in H
2
O shows the normal Watson-Crick NOE connectivity pattern, again interrupted at the two
intercalation sites. Chemical shift values of labile protons are included in
Table
1
.
Figure 3
.
The H(1') to aromatic part of the 200 ms NOESY spectrum of the 1:2
10
The internal NOE connectivities in the chromophore of TOTO are a distinct
feature of the NOESY spectra of the complex (
6
,
7
). Two distinct connectivity patterns are observed, indicating the existence of
two different TOTO ring systems corresponding to intercalation both in the (5'-C3pT4-3'):(5'-A25pG26-3') and the (5'-A5pG6-3'):(5'-C23pT24-3') base pairs. The other intercalation sites yield identical cross-peak patterns. Resonances in the linker of TOTO were assigned by combined use of TOCSY and NOESY cross-peaks. Chemical shift values of TOTO in the 1:2
10
1
.
TOTO
TOTO
TOTO
TOTO
H1
7.55
H9
6.48/6.40
H19'
4.13/4.07
CH
3
22
3.26/3.24
H2
7.44
H10
7.97/7.89
H19''
4.70/4.65
H23'
3.4
H3
7.55
H13
7.11/7.06
H20'
2.5
H23''
3.47
H4
7.44
H14
7.01/6.95
H20''
2.5
H24'
2.5
CH
3
6
3.60/3.55
H15
6.87/6.61
H21'
3.62
H24''
2.5
H8
6.12/6.06
H16
7.35/7.29
H21''
3.67
The values are given at 25o C relative to 2,2-dimethyl-2-silapentane-5-sulfonic acid (DSS). Hydrogen bonded amide
protons are underlined. Values from both parts of the chromophores are
separated by /. H(1-4) and H(20'), H(20''), H(24') and H(24'') could not be assigned
unequivocally.
The observed cross-peaks in the NOESY spectra between protons in TOTO and protons in the
oligonucleotides in the 1:2
10
10
2
sequences. They also show that the structure of the two intercalation sites
resembles the structure of the 1:1 complex with the (5'-CGCTAGCG-3')
2
oligonucleotide (
7
).
The 1:1
10
-TOTO complex.
Mixtures of TOTO and
10
10
10
10
10
10
10
10
10
10
Table 1
(i).
Chemical shift values (in p.p.m.) of the oligonucleotide protons in the 1:2
10
Table 1 (ii).
Ratio of the 1:1 and 1:2
10
-TOTO complexes.
Most parts of the spectra of the various samples of the
10
4
shows the imino region of the one-dimensional
1
H spectrum obtained at various ratios of TOTO and
10
3
.
Figure 4
.
The imino proton region of the one-dimensional spectrum of 1:2
10
o C in H
2
O. The
10
10
10
Table 2
.
Chemical shift values (in p.p.m.) of the 1:2
11
The values are given at 10o C relative to DSS. Hydrogen bonded amide proteins are underlined. Due to
the lack of sequential connectivities in the NOESY spectra at the intercalation
sites it is not possible to distinguish between the resonances from T4d/R20t,
A5e/A21u, C8h/C16p, G9i/G17q and G10j/G22v and between C11k/C23x in the
complex.
Salt dependence.
To the 1:2
10
10
10
Oligonucleotide d(CGCTAGCCGGCG):d(CGCCGGCTAGCG) (
11
)
Mixtures of
11
10
2
.
The 1:2
11
-TOTO complex.
The two-dimensional NOESY and TOCSY spectra of the complex formed with two
equivalents of TOTO show that the 1:2
11
2
and the (5'-CCGG-3')
2
sites. There are four different H9-H10 cross-peaks from TOTO and there are six cytosine H6 shifted upfield,
corresponding to the effect for cytosines in an intercalation site. The
sequential connectivities between H(6)/H(8) and H(1') and between H(6)/H(8) and H(2')/H(2'') are interrupted in eight cases. This is observed
between C3c and T4d, A5e and G6f, C7g and C8h and G9i and G10j in the first
strand. In the second strand it occurrs between C15o and C16p, G17q and G18r,
C19s and T20t and A21u and G22v. The lack of sequential connectivities makes it
impossible to assign some of the nucleotides unambiguously, since interchange
in the connectivity pattern becomes possible. The chemical shift values of the
dsDNA protons in the 1:2
11
2
.
The distinct internal NOE connectivities from the chromophores of TOTO in the
spectra of the 1:2
11
10
The NOESY spectrum of the complex in H
2
O shows the normal Watson-Crick NOE connectivity pattern, again interrupted at the two
intercalation sites. Chemical shift values of labile protons are included in
Table
2
, but overlapping lines prevent a complete unambiguous assignment.
The 1:1
11
-TOTO complex
es
.
Two different 1:1
11
2
site. The minor form has TOTO bis-intercalated in the (5'-CCGG-3')
2
site. The numbering schemes used in the two complexes are shown in Scheme
2
.
The one- and two-dimensional spectra of the mixture with equivalent amounts of TOTO
and
11
11
11
9
-
14
) as described earlier (
6
,
7
). The values obtained (not given here) are in accord with the values for the
1:2
11
Ratio of the 1:1 and 1:2
11
-TOTO complexes.
Similarly to the
10
11
2
O. However, the imino proton region in the spectrum has separate lines that
enable determination. The assignment of the lines in this part of the spectrum
is shown in Figure
5
. The assignment was performed by conventional methods (
9
-
14
). Integration of the line by the line fitting procedure in FELIX makes it
possible to determine the ratios between the different complexes. The results
are given in Table
3
.
Figure 5
.
The imino proton region of the one-dimensional spectrum of 1:2 ctagcccgg-TOTO complexes obtained at 10o C in H
2
O. The
11
11
11
DISCUSSION
Sequence selectivity
The various complexes formed between TOTO and the oligonucleotides as described
in this paper and earlier work (
6
,
7
) prove that the molecule predominantly bis-intercalates in dsDNA. TOTO binds to all the sequences used. Usually a
dynamic mixture of different binding sites exists. The structure of the
9
2
or (5'-pu-pu-pyr-pyr-3')
2
type. Eight different types of such binding sites exist. Each of these are present in one of
the oligonucleotides in Scheme
2
.
TOTO binds preferentially to oligonucleotides containing a (5'-CTAG-3')
2
or a (5'-CCGG-3') sequence. For oligonucleotide
8
2
site than to any other site present in that oligonucleotide. The spectra of the
complex formed with
4
2
dominates over binding to any other site in this oligonucleotide by at least 50-fold, but TOTO binding to the (5'-CTAG-3')
2
site is favored over binding to the (5'-CCGG-3')
2
site.
The degree of sequence selectivity of TOTO is emphasized by complex formation of
TOTO with
10
11
2
sequences are present in the same oligonucleotide or if both a (5'-CTAG-3')
2
and (5'-CCGG-3')
2
sequence are present in the same oligonucleotide. These two oligonucleotides
are fairly large dsDNA fragments and quite a number of possible binding sites
are actually present. Nevertheless, the only binding sites observed are the (5'-CTAG-3')
2
and (5'-CCGG-3')
2
sequences.
Table 3
.
The ratios of the various complexes formed by TOTO with oligonucleotides
10
11
dsDNA
Relative ratios of complexes
[1:2]/[1:1 CTAG]
[1:2]/[1:1 CCGG]
10
0.27 +- 0.06
11
0.41 +- 0.08
1.0 +- 0.2
The structural basis for the sequence selectivity of TOTO for (5'-CTAG-3')
2
has been clarified by determination of the three-dimensional structure of the
9
7
). The ability of each of the two TOTO chromophores to adapt to the base pair
propeller twist of the dsDNA in a (5'-pyr-pyr-3'):(5'-pu-pu-3') site and the
hydrophobic interaction between the thymine methyl groups and the benzothiazole
ring provide the major contributions to the specific (5'-CTAG-3')
2
sequence selectivity of TOTO. Binding of TOTO to a (5'-CCGG-3')
2
site does not have a contribution from a hydrophobic interaction between a
methyl group on the nucleobase and the benzothiazole ring. Consequently,
preference for the (5'-CCGG-3')
2
site is less pronounced than for the (5'-CTAG-3')
2
site.
Oligonucleotide
7
4
8
9
2
site, does not bind TOTO preferentially. This is probably due to the fact that
methyl groups on thymines cause steric hindrance in the case of
7
2
site does adaptation of TOTO to the propeller twist imply that the thymine
methyl is positioned in the favorable position on top of the benzothiazole
ring.
None of the oligonucleotides containing a (5'-pu-pu-pyr-pyr-3')
2
site binds TOTO preferentially. A major reason for this is probably that
adaptation of both chromophores of TOTO to the propeller twist of the dsDNA
leaves the linker chain in an unfavorable position in the minor groove. Unlike
in the (5'-CTAG-3')
2
site, it is not possible in these sites for the linker to cross from one side
of the groove to the other so as to optimize the electrostatic binding energy and the van der Waals contacts between the linker
chain
N
-methyl groups and the walls of the groove. Steric conflict from thymine
methyl groups may also add to the less preferential binding of TOTO to (5'-pu-pu-pyr-pyr-3')
2
sites.
Cooperativity in complex formation
In the case of both
10
11
3
and are simply given by the square of the ratios between the 1:2 and the 1:1
complexes in these samples.
For oligonucleotide
10
10
10
There are two possible paths for formation of the 1:2 complex with
11
2
site yields a value of 0.18 +- 0.07 for the ratio between the equilibrium constants for formation of
the 1:2 and the 1:1 complexes. The equivalent value for the ratio for the
second path is 1.0 +- 0.4, where the 1:1 complex is at the (5'-CCGG-3')
2
site. This implies that there is slight negative cooperativity for the first
path, whereas there is independent complex formation along the second path.
These results are related to the difference in the binding constants at the two
sites.
The ratio between the two different 1:1 complexes formed with
11
2
site. This is very similar to results obtained when the two sites are in
different oligonucleotides, since the ratio between the complex of TOTO with
9
4
2
and the (5'-CCGG-3')
2
sites is independent of whether the two sites are in the same or two different
oligonucleotides.
Comparison with other bis-intercalators
Most other intercalating compounds show little sequence selectivity in binding to various dsDNA oligonucleotides and when they do so the effect is
most convincingly explained as due to interactions of bulky side groups or
linkers (
15
). TOTO belongs to a new structural class of `adaptable intercalators', in
contrast to the common intercalators. The characteristic feature of the TOTO
chromophore is its ability to adapt to the base pair propeller twist of dsDNA. This is reflected in the slow on-off kinetics, the high binding constant and the sequence selectivity.
Most other bis-intercalators have rigid aromatic chromophores that cannot adapt to the
base pair propeller twist. This makes TOTO distinct among bis-intercalators.
There are a number of dsDNA complexes with other bis-intercalators that have been structurally characterized (
16
-
30
). Feigon and co-workers (
21
-
23
) have studied bis-intercalation of the well-known antitumor antibiotic echinomycin in various dsDNA
oligonucleotides. They found that echinomycin bis-intercalates selectively on each side of a 5'-CpG-3' site. In both the octamer d(ACGTACGT)
2
and the decamer d(ACGTATACGT)
2
they found that echinomycin binds to both sites in a fully cooperative manner,
whereas binding to the two sites in a d(TCGATCGA)
2
octamer is independent. Thus the structural changes caused by echinomycin
binding to oligonucleotides may be so large that cooperativity in binding is
possible. This is probably caused by the large cyclic peptide groups linking
the two intercalation chromophores.
In the case of TOTO we have shown (
7
) that the two TOTO chromophores are able to adapt to the base pair propeller
twist of dsDNA and therefore cause less structural change. This is probably why
TOTO binds in an almost non-cooperative manner.
Application of TOTO
The TOTO class of fluorescent bis-intercalators has proven very effective in a variety of DNA detection
applications (
31
,
32
). The finding of significant sequence specificity through interactions of the
intercalating chromophore and base pairs in this system was unexpected.
However, this binding specificity will probably not interfere with its use as a
general DNA stain, although there are situations where the binding preference
could pose problems.
The binding of more than one TOTO molecule in an almost independent manner is
important for the use of TOTO in non-radioactive detection of nucleic acids on gels. Even at relatively high
TOTO concentrations an inhomogeneous staining of nucleic acids on the gel due
to cooperative effects is not to be expected.
There are many applications for sequence-specific dsDNA binding molecules, both as drugs and as probes of dsDNA
structure and function. Intercalation provides a large contact surface between
the ligand and dsDNA which can provide a substantial free energy of binding.
TOTO might serve as a template for the design of new successful sequence-selective DNA binding molecules. The combination of other binding motifs
with an intercalative anchor has been exploited in both natural and synthetic
systems (
33
-
36
). An option to achieve both higher affinity and specificity is to combine the
thiazole orange chromophore with other minor groove-specific agents, such as netropsin or a depsipeptide linker. Such hybrid
molecules could offer extended sequence recognition. Of course, attention must
be paid to allow each component to adopt its optimum binding geometry without
losing the advantage of linking them together.
ACKNOWLEDGEMENTS
The authors are grateful to Professor A.Glazer (University of California,
Berkeley, CA) for providing TOTO and to Dr Paul C.Stein (Odense University) for
many helpful discussions.
REFERENCES
1 Rye ,H.S. , Yue,S., Wemmer,D.E., Quesada,M.A., Haugland,R.P., Mathies,R.A. and Glazer,A.N. (1992 ) Nucleic Acids Res 20 2803 -2812.MEDLINE Abstract
2 Benson ,S.C. , Mathies,R.A. and Glazer,A.N. (1993 ) Nucleic Acids Res 21 5720 -5726.MEDLINE Abstract
3 Benson ,S.C. , Singh,P. and Glazer,A.N. (1993 ) Nucleic Acids Res 21 5727 -5735.MEDLINE Abstract
4 Rye ,H.S. , Yue,S., Quesada,M.A., Haugland,R.P., Mathies,R.A. and Glazer,A.N. (1993 ) Methods Enzymol 217 414 -431.MEDLINE Abstract
5 Rye ,H.S. , Dabora,J.M., Quesada,M.A., Mathies,R.A. and Glazer,A.N. (1993 ) Anal. Biochem 208 144 -150.
6 Jacobsen ,J.P. , Pedersen,J.B., Hansen,L.F. and Wemmer,D.E. (1995 ) Nucleic Acids Res 23 753 -760.MEDLINE Abstract
7 Spielmann ,H.P. , Wemmer,D.E. and Jacobsen,J.P. (1995 ) Biochemistry 34 8542 -8553.MEDLINE Abstract
8 Stein ,P. , Jacobsen,J.P. and Spielmann,H.P. (1995 ) J. Magn. Resonance Ser. B 109 93 -96.
9 Wüthrich ,K. (1986 ) NMR of Proteins and Nucleic Acids
10 Feigon ,J. , Leupin,W., Denny,W.A. and Kearns,D.R. (1983 ) Biochemistry 22 5943 -5951.MEDLINE Abstract
11 Hare ,D.R. , Wemmer,D.E., Chou,S.-H., Drobny,G. and Reid,B.R. (1983 ) J. Mol. Biol 171 319 -336.MEDLINE Abstract
12 Scheek ,R.M. , Boelens,R., Russo,N., Van Boom,J.H. and Kaptein,R. (1984 ) Biochemistry 23 1371 -1376.MEDLINE Abstract
13 Scheek ,R.M. , Russo,N., Boelens,R. and Kaptein,R. (1983 ) J. Am. Chem. Soc 105 2914 -2916.
14 Boelens ,R. , Scheek,R.M., Dijkstra,K. and Kaptein,R. (1985 ) J. Magn. Resonance 62 378 -386.
15 Wilson ,W.D. and Jones,R.L. (1982 ) Intercalation Chem 14 446 -499.
16 Searle ,M.S. (1993 ) Progr. NMR Spectrosc 25 403 -480.
17 Addess ,K.J. and Feigon,J. (1994 ) Biochemistry 33 12386 -12396.MEDLINE Abstract
18 Addess ,K.J. and Feigon,J. (1994 ) Biochemistry 33 12397 -12404.MEDLINE Abstract
19 Addess ,K.J. , Gilbert,D.E., Olsen,R.K. and Feigon,J. (1992 ) Biochemistry 31 339 -350.MEDLINE Abstract
20 Gao ,X. and Patel,D.J. (1989 ) Q. Rev. Biophys 22 93 -138.
21 Gilbert ,D.E. and Feigon,J. (1992 ) Nucleic Acids Res 20 2411 -2420.MEDLINE Abstract
22 Gilbert ,D.E. and Feigon,J. (1991 ) Biochemistry 30 2483 -2494.MEDLINE Abstract
23 Gilbert ,D.E. , Van der Marel,G.A., Van Boom,J.H. and Feigon,J. (1989 ) Proc. Natl. Acad. Sci. USA 86 3006 -3010.MEDLINE Abstract
24 Searle ,S.M. , Hall,J.G., Denny,W.A. and Wakelin,L.P.G. (1989 ) Biochem. J 259 433 -441.MEDLINE Abstract
25 Delepierre ,M. , Milhe,C., Namane,A., Dinh,T.H. and Roques,B.P. (1991 ) Biopolymers 31 331 -353.MEDLINE Abstract
26 Delepierre ,M. , Maroun,R., Garbay-Jaureguiberry,C., Igolen,J. and Roques,B.P. (1989 ) J. Mol. Biol 210 211 -228.MEDLINE Abstract
27 Delepierre ,M. and Dinh,T.H. (1989 ) Biopolymers 28 2115 -2142.MEDLINE Abstract
28 Maroun ,R. , Delepierre,M. and Roques,B.P. (1989 ) J. Biol. Struct. Dyn 7 607 -621.
29 Pothier ,J. , Delepierre,M., Barsi,M., Garbay-Jaureguiberry,C., Igolen,J., Le Bret,M. and Roques,B.P. (1991 ) Biopolymers 31 1309 -1323.MEDLINE Abstract
30 Williams ,L.D. and Gao,Q. (1992 ) Biochemistry 31 4315 -4324.MEDLINE Abstract
31 Perkins ,T.T. , Smith,D.E. and Chu,S. (1995 ) Science 264 819 -822.MEDLINE Abstract
32 Perkins ,T.T. , Quake,S.R., Smith,D.E. and Chu,S. (1995 ) Science 264 822 -826.MEDLINE Abstract
33 Bailly ,C. and Henichart,J.P. (1991 ) Bioconjugate Chem 2 379 -393.
34 Mouscadet ,J.F. , Ketterle,C., Goulaouic,H., Carteau,S., Subra,F., LeBret,M. and Auclair,C. (1994 ) Biochemistry 33 4187 -4196.MEDLINE Abstract
35 Orson ,F.M. , Kinsey,B.M. and McShan,W.M. (1995 ) Nucleic Acids Res 22 479 -484.MEDLINE Abstract
36 Subra ,F. , Carteau,S., Pager,J., Paoletti,J., Paoletti,C., Auclair,C., Gosselin,G. and Imbach,J.L. (1991 ) Biochemistry 30 1642 -1650.MEDLINE Abstract
Return
*
To whom correspondence should be addressed
