Conformational analysis of antisense
oligonucleotide/RNA duplexes as substrates for RNase H
Computer simulations give us an opportunity to look inside the mechanism
of interaction of antisense oligonucleotides with RNA strands and the recruitment
of RNase H (a short overview of the literature is
given below). Having in mind these goals we are working on the following:
(1) To draw a correlation between the structure of
the DNA(AON)/RNA hybrid and its properties as a substrate for the RNase
H, as well as to point the crucial structural requirements for the
modified AONs to preserve their RNase H potency. We are investigating the
structural changes in the AON/RNA hybrid duplexes brought about by different
modifications of the sugar in the fused oxetane system, backbone (phosphorothioate
and methylphosphonate) or the base in the antisense strand, and monitor
the effect of these changes on the RNase H recognition and cleavage of
the modified substrates. The investigation is performed by two complementary
methods: (a) high-field NMR studies of the T and C-modified
AON/RNA duplex, which is already in progress, and (b)
by the ab initio (for the individual modified AONs) and MM/MD (for
AON/DNA duplexes) simulations in the aqueous environment. This allows
us to examine the details of local and global structural changes of the
modified hybrid duplexes in comparison with the native counterpart.
Thus, we are investigating conformational changes in
a set of hybrid AON/RNA duplexes (10 and 15 mers), consisting of a series
of analogous single bicyclic psiconucleoside modified AONs with fixed North
conformation (Figure 1
) and the complementary
target RNA. These duplexes have been investigated by us recently [Refs.
deploying a combination of CD and partial RNase H digests on these sets
of hybrid AON/RNA duplexes. It has been shown that the local structure
perturbation upon introduction of a single N-type constrained nucleoside
spreads up to 5 nucleotides toward the 3'-end of the modification site
although no global helical conformational change is detectable by CD spectroscopy.
We expect that the on-going ab initio
and MM/MD study would allow
us to understand a theoretical basis of this phenomenon of conformational
transmission, and show how do the structure and dynamics of the AON/RNA
hybrids dictate their functional properties in the presence of the RNase
H and Mg2+
(2) To model the dynamics of the RNase H interaction
with substrate AON/RNA hybrids, which is expected to shade light
on the recognition, interaction, intermolecular hydrolysis,
and in general the mechanism of the RNase H promoted cleavage reaction
the AON/RNA hybrid duplex. We are going to model docking of the RNase H
to the substrate (i.e. hybrid duplex) following the dynamics of their interaction
(a model structure of the enzyme docked to AON/RNA duplex is shown in Figure
2). Since it is not possible to obtain experimental crystal structure
of this reactive complex, the theoretical modeling seems to be a sound
way to investigate the processes involved in this interaction and the conformational
readjustments that result from the ternary complex formation before the
onset of the hydrolysis. The role of Mg2+ ion in this
cleavage reaction is important to address at the catalytic center of this
Figure 1. Chemical structures of the
oxetane modified T thymine and C cytosine ([1-(1',3'-O-anhydro-b-D-psicofuranosyl)
The calculations described above are to be followed
by the modeling of RNase H docking to the AON/RNA hybrid. We are
planning to use molecular dynamics to find a structure of the reactive
complex and to describe the catalytic action by RNase H. Additionally to
the MD simulation of the reaction itself we are planning to model the cleavage
center of RNase H by means of ab initio simulation. The ultimate
goal of this investigation is to give definitive answers to the questions
about the RNase H promoted cleavage mechanism.
The recruitment by RNase H, an endogenous enzyme that
specifically degrades target RNA in the antisense oligonucleotide (AON)/RNA
hybrid duplex is an important pathway for the antisense action beside the
. RNase H hydrolyses the RNA strand in
an RNA/DNA hybrid in a catalytic manner2,3
. It produces short
oligonucleotides with 5'-phosphate and 3'-hydroxy groups as final products4
Bivalent cations as Mg2+
are found to be
necessary cofactors for enzymatic activity4,5,6
. The enzyme
is widely present in various organisms4
, including retroviruses,
as a domain of the reverse transcriptase7
. The RNase H1 from
Escherichia coli. is the most characterized enzyme in this family8,9,10
Even though the physiological functions of E. coli
RNase H1 have
not been understood clearly, it has been suggested to be involved in DNA
replication and repair11
. This enzyme was found to be required
for the initiation of Col E1 DNA replication in vitro12,13
It was suggested that after RNA has served as a primer, RNase H eliminates
it from the product. This enzyme is also involved in the chromosomal DNA
RNase H promoted cleavage of the viral mRNA via formation of
the duplexes with complementary oligo-DNAs (antisense strand) is one of
the strategies to treat viral infections1,18. Recent isolation
of the human RNase H1 and RNase H2 highlights the importance of the development
of the antisense drugs utilizing this mechanism of action19,20,21,22
It has been suggested that for eliciting the RNase H in AON/RNA hybrid,
the AON part should retain the B-type DNA conformation with 2'-endo sugar
(South-type, S), while the RNA moiety should retain its A-type helix character
with 3'-endo sugar (North-type, N)23. To fulfill these requirements
various modifications of sugar, base as well as of the phosphate backbone
have been attempted and numerous reports are available about these modified
AONs and their antisense action24. Among these, AONs having
one or more conformationally fixed (either in N-25 or S-26
form of sugar pucker) nucleoside residues have been found to be promising
candidates because when they are locked in the N-form, they exhibit high
affinity to the target RNA27. Recently, the locked nucleic acid
(LNA), in which the sugar moiety is fixed in the North conformation, has
shown unprecedented affinity towards RNA25e,h. LNA and other
modifications which have the fixed N-sugar moiety drive the AON helix to
the A-type resulting in RNA/RNA type duplex which accounts for their higher
binding affinity, but this leads to the loss of RNase H action28.
The introduction of conformationally constrained N-methanocarba-thymidine
residue in the N-form25a increased the thermodynamic stability
of AON/RNA duplex, whereas in the S-form26a, a destabilizing
effect was observed. It was later found that multiple introduction of (N)-methanocarba-thymidines,
although increased the thermodynamic stability of the AON/RNA duplex, but
failed to recruit any RNase H activity25c.
It is now quite clear27,28
that all modifications that lead
to preferential North-type sugar, including its constrained form, in an
RNA-type AON result in the loss of RNase H activity, because they resemble
RNA/RNA duplex, except when they appear at the termini or in the middle
in the gapmer-AON29
It has been so far assumed that probably three or four N-type conformational
repeats are necessary to enhance the thermal stability of RNA-type AON/RNA
. Nobody however specifically knows how many North-constrained
nucleosides are required to alter the conformational tolerance of the RNase
H recognition, thereby its substrate specificity, owing to the local structural
perturbations in an RNA-type AON/RNA hybrid.
Recently, we have demonstrated29
that introduction of a single North-constrained nucleoside, [1-(1',3'-O-anhydro-b
) (Figure 1
), in to an AON does
not alter the global helical structure of the corresponding AON/RNA hybrid
as compared to the native counterpart30
Despite the fact that a series of single T
hybrid duplexes showed a drop of ~6ºC/modification in Tm
they were however cleaved by RNase H with comparable efficiency as compared
to the native counterpart. We also found that the target RNA strand in
the hybrid AON/RNA duplex was resistant up to 5 nucleotides towards 3'-end29,30
from the site opposite to T
introduction in the AON strand.
We have also reported31
that the incorporation of up to three T
residues in the modified
AONs, irrespective of their positions in the AON chain, shows the 5 nucleotides
resistance region in the RNA strand from the site opposite to T
in the corresponding AON/RNA hybrid to RNase H. Despite the fact that these
modified AON/RNA duplexes were destabilized (Tm
dropped by 6º
to 20 ºC) compared to the native counterpart, they were found to be
as good substrates for RNase H as the native hybrid duplex. The exact
number of T
modifications in the AON was found to play an
important role in exhibiting resistance towards endonucleotic degradation,
although they did not have any effect against the 3'-exonuclease activity.
We also have shown31
that the minimal introduction of T
residues in the AON strand
can be effectively used as tool to produce desired RNA fragments by engineering
the RNase H cleavage site in the AON/RNA duplex, which we believe, should
have considerable applications in the field of RNA engineering.
The following experimental observations29,30,31
give an insight in to the behavior of various T modified
AON/RNA hybrids towards RNase H cleavage as well as their stability toward
endo and exonucleases:
1) The extent of RNA cleavage in hybrid duplexes
by E. coli RNase H1 in the native hybrid [DNA/RNA] was found to
be 68 ± 3%. The target RNA with all single T, double
and triple T modified AONs, were hydrolyzed under the same
conditions with extend of 51-68 ± 3%.
2) In the AON/ RNA hybrid duplexes with a single mismatch, the RNA
was cleaved at a comparable rate as the native counterpart although the
hybrid shows a loss of 10 - 11ºC in Tm. owing to the mismatch.
They also showed additional cleavage sites. These two observations therefore
show that the recognition of the oxetane-based T vis-a-vis
a mismatch in the AON strand by the target RNA is indeed different,
most probably owing to the fact that T was perhaps partially
3) The five nucleotide resistance rule to the RNase
H cleavage of the RNA in the AON/RNA hybrids in all single T,
double T and triple T modified AONs allowed us to engineer
a single cleavage site in the target RNA by RNase H. The single RNA cleavage
site has been earlier shown to occur in case of 2'-O-methyl modified
chimeric AON/RNA duplex32 in which all the central 2'-deoxynucleotides
except the middle nucleotide have been shown to adopt an RNA-type conformation
by NMR spectroscopy33. Since the CD spectra showed that
all our T modified AON/RNA hybrid duplexes have global structure
that corresponds to DNA/RNA type duplex (indicating that our AONs retain
the B-DNA type helical conformation in the hybrid), we conclude that the
5-nucleotides resistance rule observed with our T modified
AONs is owing to more subtle local microscopic conformational (and/or hydration)
change, which is only detectable by the enzyme, not by the CD.
4) The three T modified AONs gave the endonuclease stability
(with DNase 1) almost 4 fold better (87% of AON remained after 1 h of incubation)
compared to the natural counterpart (19% left), but their 3'-exonuclease
stability was identical to that of the native AON. The 3'-exonuclease stability
was however improved by using three T modifications along
with the 3'-tethering of dipyridophenazine (DPPZ) moiety25b,
in that 85% of AON was intact while the native AON was completely hydrolyzed
after 2h of incubation with SVPDE (note that the endonuclease resistance
remained however unchanged). The RNase H promoted cleavage of this AON/RNA
duplex (59 ± 4%) remained very comparable to that of the counterpart
with the native AON (68 ± 3%) and with three T
modified AON (61 ± 6%), although a gain of 7ºC of Tm
was achieved by this additional 3'-DPPZ modification. This again shows
that the rise of Tm do not necessarily
dictate the RNase H cleavage as was earlier found for some methylphosphonate
and boranophosphates33. It should be however
noted that the presence of the 3'-DPPZ moiety produces an additional cleavage
site. This is most probably owing to the stabilization of the terminal
G-C hydrogen bonding by the 3'-DPPZ group (observed by NMR34)
as well as the recognition of the DPPZ by the enzyme25b
both of which appears to be important for RNase H recognition, binding
and cleavage. Interestingly, amongst all the T modified AONs
studied so far, this is the only example where the 5-nucleotide resistance
rule in the RNA strand is not obeyed.
The enthalpy of stabilization (D
as a measure of hydrogen bonding and stacking interactions35,36
can be used to judge the basepairing ablity of AONs. Experimental
have shown that AON/RNA duplex with three
has a D
Hº value about 120 kJ mol-1
lower than the corresponding AON/RNA hybrid with three A
at the same positions. This large enthalpy difference between mismatched
and modified AON/RNA hybrids clearly rules out the possibility of the complete
loss of hydrogen bonding in the T
modified duplexes. This
observation also supported by the RNase H digestion pattern and the extent
of hydrolysis of the mismatched vis-à-vis
). Since the magnitude of enthalpy for the duplex with
modifications lies between the native and mismatched
duplexes, probably there is a partial loss of basepairing occurs when the
AON containing T
modification forms duplexes with target
RNA. Probably, that might be the reason why there was no duplex formation
detected when ten T
s were introduced in to the AON. It should
also be mentioned that the presence of the oxetane moiety of T
in the minor groove of the duplex might have altered the spine of hydration,
which could also contribute to the destabilization of the duplex. Anyway,
thorough structural investigations are necessary; some of them are underway,
before any conclusion can be drawn. We also hope that results of
the calculations planned will give valuable information about energy and
conformational changes involved in the introduction of mismatches and modified
AONs in the DNA/RNA duplexes.
The RNase H recruiting power of the locked T
modified AONs/RNA hybrids rises questions on the relation between thermodynamic
stability/flexibility of hybrid duplexes and the structure/dynamic vis-à-vis
recognition and structural tolerance by the enzyme37
It is likely that AONs/RNA hybrids perhaps should possess certain degree
of structural flexibility to undergo certain structural readjustments upon
complexation with RNase H / Mg2+
in the minor groove for the
hydrolysis to take place, which perhaps can take place in our loosely hydrogen-bonded
modified AON/RNA hybrid duplexes as well as in some methylphosphonate chimeras27
, and hence their RNase H recruitment capability
remains unaltered (or even better) compared to the native counterpart.
The suggested calculations are expected to give quantitative and qualitative
explanations to the experimentally observed RNase H recognition pattern.
To perform the calcuations the following program packages
have been used: GAMESS US, Gaussian 98, and Amber 6. Due to the size
of the systems under investigation we are in need to use powerful parallel
supercomputers and these programs have demonstrated very good scalability
with almost linear increase of performance versus the number of nodes.
The possibility to use resources of National Supercompter Center (NSC)
in Linköping, Sweden, are greatly acknowledged.
1. Crooke S.T. Biochim. Biophys. Acta. 1999,
2. Stein H., Hausen P. Science. 1969, 166,
3. Hausen P., Stein H. Eur. J. Biochem. 1970,
4. Crouch R.J., Dirksen M.L. In: Linn SM, Roberts R
J, editors. Nuclease. New York, Cold Spring Harbor: Cold Spring Harbor
Laboratory. 1982, p. 211-241
5. Berkower I., Leis J., Hurwitz J. J. Biol.
Chem. 1973, 248, 5914-21
6. Stavrianopoulos J.G., Gambino-Giuffrida A., Chargaff
E. Proc. Natl. Acad. Sci. USA. 1976, 73, 1087-1091
7. Wintersberger U. Pharmac. Ther. 1990, 48,
8. Katayanagi K., Miyagawa M., Matsushima M., Ishikawa
M., Kanaya S., Ikehara M., Matsuzaki T., Morikawa K.. Nature; 1990,
9. Yang W., Hendrickson W.A., Crouch R.J., Satow Y.
1990, 249, 1398-1405
10. Katayanagi K., Miyagawa M., Matsushima M., Ishikawa
M., Kanaya S., Nakamura H., Ikehara M., Matsuzaki T., Morikawa K J.
Mol. Biol. 1992, 223(4), 1029-52
11. Casregola S., Khidhir M., Holland I.B. Mol. Gen.
Genet. 1987, 209, 494-98
12. Itoh T., Tomizawa J. Proc. Natl. Acad. Sci..
1980, 77, 2450-54
13. Dasgupta S., Masukata H., Tomizawa J. Cell.
1987, 51, 1113-22
14. Ogawa T., Pickett G.G., Kogoma T., Kornberg A. Proc.
Natl. Acad. Sci. USA. 1984, 81, 1040-44
15. Kogoma T., Subia N.L., Meyenberg K. Mol. Gen.
Genet. 1985, 200, 103-109
16. Hong X., Kogoma T.J. J. Bacteriol.
1993, 175, 6731-34
17. Horiuchi T., Maki H., Sekiguchi M. Mol.
Gen. Genet. 1985, 195, 17-22
18. Minshull J., Hunt T. Nucleic Acids Res. 1986,
19. Frank P., Braunshofer-Reiter C., Wintersberger U.,
Grimm R., Busen W. Proc. Natl. Acad. Sci. USA. 1988, 95,
20. Cerritelli S.M., Crouch R.J. Genomics. 1988,
21. Wu H.J., Lima W.F., Crooke S.T. Antisense Nucleic
Acid Drug Dev. 1998, 8, 53-61
22. Wu H.J., Lima W.F., Crooke S.T. J. Biol. Chem.
1999, 272, 6146-50
23. (a) Altman, K.-H., Kesselring, R.; Francotte, E.;
Rihs, G. Tetrahedron Lett. 1994, 35, 2331.
(b) Siddiqui, M.A.; Ford, H.; George,
C.; Marquez, V. E. Nucleosides and Nucleotides 1996, 15,
(c) Marquez, V. E.; Siddiqui, M.A.;
Ezzitouni, A.; Russ, P.; Wang, J.; Wanger, W.R.; Matteucci, D. M. J.
Med. Chem. 1996, 39,3739.
(d) Obika S.; Nanbu D., Hari Y., Morio,
K.; In, Y.; Ishida, T.; Imanishi, T.Tetrahedron Lett. 1997, 38,
(e) Koshkin, A. A.; Singh, S. K.; Nielson,
P., Rajwanshi, V.K.; Kumar R.; Meldgaard, M.; Wengel, J. Tetrahedron
1998, 54, 3607.
(f) Wang, G.; Giradet, J, –L.;
Gunic, E. Tetrahedron 1999; 55, 7707.
(h)Wengel, J. Acc. Chem. Res.
1999, 32, 301.
(i) Wang, G.; Gunic, E.; Giradet,
J, –L. Bioorg. Med. Chem. Lett 1999, 9, 1147.
(j) Obika S.; Nanbu D., Hari Y.,
Morio, K.; Andoh, J.; Morio, K.; Doi, T.; Imanishi, T. Tetrahedron Lett.
1998, 39, 5401.
(k) Sekine, M.; Kurasawa, O.;
Shohda, K.; Seio, K.; Wada.; T. J. Org. Chem. 2000, 65, 3571.
(l) Altman, K.-H., Imwinkelried,
R.; Kesselring, R.; Rihs, G. Tetrahedron Lett. 1994, 35,
(m) Obika, S.; Morio,K.; Hari, Y.; Imanishi,
T. Chem.Commun. 1999, 2423.
(n) Obika, S.; Morio,K.; Hari, Y.; Imanishi,
T. Bioorg. Med. Chem. Lett. 1999, 9, 515.
(o) Steffens, R.; Leumann, C. Helv.
Chim. Acta 1997, 80, 2426.
(p) Buff. R, Hunziker J. Bioorg.
Med. Chem. 1998, 521
24. (a) Herdewijn, P. Biochem. Biophys. Acta
1999, 1489, 167.
(b) Herdewijn, P. Leibigs Ann. Chem.
25. (a) Baker, B. F.; Monia, B. P. Biochem. Biophys.
E.; Ossipov, D.; Pradeepkumar, P.I.; Amirkhanov, N.; Chattopadhyaya. J.
2001, 57, 593.
26. (a) Crooke, S.T. Methods Enzymol. 2000, 313,
(b) Crooke, S. T. Biochem. Biophys.
Acta 1999, 1489, 31.
27. Giles, R. V.; Tidd, D. M. Anti-Cancer Drug
Des. 1992, 7, 37
28. Lima, W. F.; Crooke, S. T. Biochemistry
32. Inoue, H.; Hayase, Y.; Iwai S.; Ohtsuka, E.
Lett. 1987, 215, 327.
33. Nishizaki, T.; Iwai, S.; Nakumara, H. Biochemistry
1997, 36, 2577.
34. Rait, V. K.; Shaw, B. R. Nucleic Acid Drug Dev.
1999, 9, 53.
36. Burkard, M. E.; Turner, D. H.; Tinoco, I.; Jr. In
RNA World, 2nd ed.; Gesteland, R. F., Cech, T. R., Atkins, J. F., Eds.
Cold Spring Harbor Laboratory Press: Cold Spring Harbor, NY, 1999; pp 233-264.