Nucleic acid binding properties of thyminyl and adeninyl pyrrolidine-amide oligonucleotide mimics (POM).

Article abstract of DOI:10.1039/b315768g

Adeninyl POM was prepared using solid-phase peptide chemistry and shown to exhibit higher affinity for complementary DNA and RNA than the corresponding adeninyl PNA.

Full text of DOI:10.1039/b315768g

Nucleic acid binding properties of thyminyl and adeninyl
pyrrolidine-amide oligonucleotide mimics (POM)†
T. H. Samuel Tan, David T. Hickman, Jordi Morral, Ian G. Beadham and Jason Micklefield*
Department of Chemistry, University of Manchester Institute of Science and Technology, PO Box 88,
Manchester, UK M60 1QD. E-mail: jason.micklefield@umist.ac.uk; Fax: 0161 236 7677; Tel: 0161 200
4509
Received (in Cambridge, UK) 3rd December 2003, Accepted 22nd December 2003
First published as an Advance Article on the web 28th January 2004
Adeninyl POM was prepared using solid-phase peptide chem-
istry and shown to exhibit higher affinity for complementary
DNA and RNA than the corresponding adeninyl PNA.
alkylated to provide the cis-alcohol 14. Inversion of the C4-
hydroxyl group to give the trans-alcohol 16 and introduction of N3-
benzoylthymine using Mitsunobu reactions afforded the
(2R,4R)-pyrrolidine 17. Reduction and Fmoc protection gave the
ester 7. Surprisingly the N3-benzoyl protecting group was lost
during transformation 17 ? 7. Following this route the adeninyl
monomer 21 was also prepared. In this case tosylation of the trans-
alcohol 16 ? 18, or the cis-alcohol 14 ? 18, followed by
displacement with N6-benzoyladenine gave the adeninyl derivative
19. Azide reduction, Fmoc protection and cleavage of the tert-butyl
ester gave the required adeninyl Fmoc-amino acid 21.
Earlier we had prepared a thyminyl POM pentamer with an N-
terminal phthalimido group (Phth-T5-POM) by solution phase
synthesis.7 This precluded a direct comparison with PNA, since T5-
PNA requires terminal lysine residues to avoid aggregation and aid
aqueous solubility. The thyminyl POM pentamer, LysPOM(T)5-
LysNH2, with N- and C-terminal lysine residues was therefore
prepared on Rink-amide MBHA resin. Coupling reactions involved
pre-activation of the T-Fmoc-amino acid 8 (2 equiv.), TBTU (1.9
There has been considerable interest in modified nucleic acids and
oligonucleotide mimics due to their utility as therapeutic1–3 or
diagnostics agents and as tools in molecular biology or target
validation.2–4 Of the many mimics that have been developed to date
probably the most widely studied are peptide nucleic acids
(PNA).3,5 Despite success in many applications, PNA suffers from
poor aqueous solubility and a tendency to aggregate. In addition,
PNA is achiral and can bind in both a parallel and antiparallel
orientation with nucleic acids, which can jeopardise sequence
specificity. Of late, attention has focused on mimics which contain
pyrrolidine-amide backbones.6,7 Many of these mimics were
designed as conformationally restricted chiral PNA analogues6 and
a number including mimics 2, 3, and 4 (Fig. 1), exhibit promising
properties.
Previously we introduced pyrrolidine-amide oligonucleotide
mimics (POM).7 Unlike the other pyrrolidine containing mimics
POM was not designed as a PNA analogue, instead it was intended
as a configurational and conformational mimic of natural nucleic
acids. Indeed evidence indicates that the (2R,4R)-pyrrolidine ring in
POM adopts a similar conformation to ribose in RNA.7 In addition,
protonation of the pyrrolidine N-atom ensures POM possesses
excellent water solubility and exhibits electrostatic attraction with
oppositely charged nucleic acid targets. It was therefore of little
surprise to discover that a T5-POM, prepared by solution phase
synthesis,7 exhibits very high affinity for both RNA and DNA.
Remarkably however the T5-POM exhibits a very strong kinetic
selectivity for RNA over DNA. In this paper we describe the
efficient solid-phase synthesis of both thyminyl and adeninyl POM
and compare their DNA and RNA binding properties with PNA.
Initially we chose to utilise Fmoc peptide chemistry for the
synthesis of oligomers. Accordingly the phthalimido group of 57
was replaced by Fmoc (Scheme 1). Subsequent Boc deprotection
followed by N-alkylation with tert-butyl bromoacetate and then
acidolysis gave the thyminyl Fmoc amino acid 8. This route, which
follows on from the earlier synthesis,7 is not ideal for the other
POM monomers (A, C and G) since the base is introduced into the
pyrrolidine ring early in the synthesis. A more convergent route to
monomer 8 was therefore developed. This involved reduction of the
ester 9 followed by mesylation and displacement with azide to give
12. Acidolysis revealed the amino alcohol 13, which was N-
Scheme 1 a) 40% aq. MeNH2, 50 °C, 2 h; b) Fmoc-Cl, 10% (w/v) aq.
Na2CO3, dioxane, 0 °C ? rt, 18 h; c) 20% CF3COOH in CH2Cl2, rt, 4 h; d)
BrCH2CO2tBu, DIPEA, CH2Cl2, 0 °C ? rt, 24 h; e) 4 M HCl–dioxane,
CH2Cl2, rt, 24 h; f) LiBH4, THF, 0 °C 1 h, rt 1 h; g) CH3SO2Cl, DIPEA,
CH2Cl2, 0 °C, 15 min; h) NaN3, DMF, 80 °C, 16 h; i) 4 M HCl–dioxane,
CH2Cl2, 0 °C, 15 min, rt, 2 h; j) HCO2H, PPh3, DIAD, THF, 230 °C ? rt,
18 h; k) conc. aq. NH3–CH3OH, 2 h; l) N3-benzoylthymine, PPh3, DIAD,
THF, 220 °C ? rt, 18 h; m) CH3OTs, PPh3, DIAD, THF, 210 °C ? rt, 18
h; n) TsCl, pyridine, 18 h; o) N6-benzoyladenine, K2CO3, 18-crown-6,
DMF, 80 °C, 18 h; p) sat. H2S in 60% aq. pyridine, 18 h; q) Fmoc-Cl,
DIPEA, CH2Cl2, 0 °C, 18 h. Bz = benzoyl, Boc = tert-butoxycarbonyl,
DIAD = diisopropyl azodicarboxylate, DIPEA = diisopropylethylamine,
Fmoc = fluoren-9-ylmethoxycarbonyl, Phth = phthalimido, Ts = tosyl.
Fig. 1 POM 17 and examples of related PNA analogues.6
†
This paper is dedicated to Professor Heinz G. Floss on the occasion of
his 70th birthday.
Electronic supplementary information (ESI) available: experimental
details. See http://www.rsc.org/suppdata/cc/b3/b315768g/
516
C h e m . C o m m u n . , 2 0 0 4 , 5 1 6 – 5 1 7
T h i s j o u r n a l i s © T h e R o y a l S o c i e t y o f C h e m i s t r y 2 0 0 4

equiv.), HOBT (2 equiv.) and DIPEA (5 equiv.) with an extended
coupling time of 4 h. The coupling reactions were monitored by the
Kaiser test and coupling efficiencies were calculated to be ca. 95%
by UV determination of the dibenzofulvene–piperidine adduct after
Fmoc deprotection. Cleavage from the resin was achieved using
TFA. An adeninyl POM pentamer, LysPOM(A)5NH2, with an N-
terminal lysine, was similarly prepared except prior to cleavage
from the resin benzoyl deprotection was carried out using 1 : 1
conc. aq. ammonia and dioxane at 55 °C for 16 h. Each oligomer
was purified by reverse phase (C8) HPLC to greater than 95%
purity and the identity of the products was confirmed by
electrospray ionisation mass spectrometry (see ESI).
affinity for the poly(dA). Similar results were also obtained with
r(A)20 and d(A)20 demonstrating that the length of the target nucleic
acid strand has little effect on the thermodynamics and kinetics of
hybridisation in this case. In addition LysPOM(T)5LysNH2 does
not hybridise with non-complementary poly(rC), (rU) and (rG)
indicating that base pairing specificity is preserved. Moreover
LysPOM(T)5LysNH2 is able to discriminate against a single base
pair mismatch as evidenced by the drop in the denaturation Tm with
r(A)5 and r(AAGAA) (DTm = 216.3 °C). In contrast, the Tm
increases dramatically for LysPNA(T)5LysNH2 with r(A)5 upon
insertion of the the single T.G mismatch. This is perhaps due to
formation of an atypical complex or aggregate.
The DNA and RNA binding properties of the POM 5mers were
determined using UV thermal denaturation/renaturation experi-
ments and compared with the corresponding PNA 5mers (Table 1)
and DNA 20mers (see ESI). For LysPOM(T)5LysNH2 and
poly(rA) an hyperchromic shift of 32% was observed with a
denaturation melting temperature (Tm) of 44.8 °C (Tm/base 9.0 °C).
Upon renaturation a 28% hypochromic shift was obtained with a Tm
of 43.6 °C. The corresponding PNA exhibited slightly higher
affinity for poly(rA) (Tm/base 10.8 °C). In contrast d(T)20 with
poly(rA) exhibits a considerably lower Tm per base of 2.2 °C. Very
significant hysteresis was observed in the heating/cooling curves,
between LysPOM(T)5LysNH2 and poly(dA), resulting in denatura-
tion and renaturation Tms that differ by 19.3 °C. Little or no
hysteresis was observed in corresponding PNA and d(T)20 melting
experiments. This indicates that the binding of POM with DNA is
much slower than with RNA and is consistent with our previous
observations for Phth-T5-POM.7 Furthermore this confirms that the
unusual kinetic selectivity for RNA over DNA does not depend on
the nature of the N- or C-terminal functionality but is intrinsic to the
thyminyl-POM sequence. Notably when poly(dA) and LysPOM(T-
)5LysNH2 were incubated at room temperature for 48 h prior to UV
thermal denaturation the Tm observed was 65 °C (Tm/base = 13
°C), indicating that despite very slow binding POM has very high
The adeninyl POM pentamer, LysPOM(A)5NH2, exhibits con-
siderably higher affinity for poly(rU) and poly(dT) (Tm/base 10.0
°C and 14.1 °C respectively) than the corresponding PNA (Tm/base
6.2 and 9.7 °C). However noticeable hysteresis between the cooling
and heating curves for LysPOM(A)5NH2 with both poly(rU) (DTm
= 6.8 °C) and poly(dT) (DTm = 14.7 °C) is observed. This
suggests that the kinetic selectivity for RNA over DNA is less
pronounced in this case. In contrast PNA exhibits little or no
hysteresis with either DNA or RNA targets. With r(U)20 and d(T)20
similar results were obtained with LysPOM(A)5NH2 exhibiting
considerably higher Tms than PNA. Noticeably the renaturation
curve between LysPOM(A)5NH2 and d(T)20 shows a major
inflection at 53.0 °C with a minor inflection at 31.8 °C which
suggests that triplexes (T.AT), as well as duplexes are formed in
this case. Notably it was evident that whilst neither LysP-
NA(A)5NH2 nor d(A)20 shows evidence of hybridisation with
r(U)5, LysPOM(A)5NH2 exhibits a clear hyperchromic shift with a
Tm of 18.3 °C.
In conclusion an efficient solid-phase synthesis of both thyminyl
and adeninyl POM from Fmoc amino acids 8 and 21 has been
developed. LysPOM(T)5LysNH2 binds with slightly lower affinity
than LysPNA(T)5LysNH2 to complementary RNA and DNA, but
exhibits a marked kinetic selectivity for RNA over DNA. In
contrast LysPOM(A)5NH2 exhibits considerably higher affinity for
complementary RNA and DNA than PNA, with a slight kinetic
preference for RNA over DNA. These results corroborate our
earlier findings and confirm that POM containing both pyrimidine
and purine bases possess excellent nucleic acid binding properties.
This combined with the chiral, cationic backbone and excellent
aqueous solubility makes POM a potential successor to PNA.
This work is supported by the EPSRC (studentship to DTH) and
the BBSRC (research grant 36/B15998).
Table 1 Tms for POM and PNA 5mers vs. complementary nucleic acids
Tm/°Ca (% hypochromicb/hyperchromic shiftc)
LysPOM(T)5LysNH2
LysPNA(T)5LysNH2
cooling
heating
cooling
53.3 (31)
44.2 (25)
heating
54.1 (32)
48.6 (29)
Poly(rA)
Poly(dA)
43.6a (28)b
25.1 (17)
44.8a (32)c
54.4 (17)
65.0d (28)d
40.2 (31)
51.6 (24)
34.5 (15)
18.2 (14)
r(A)20
d(A)20
r(A)5
37.2 (31)
21.6 (21)
13.6 (14)
47.2 (27)
35.3 (26)
18.0 (26)
52.8 (7)
47.9 (28)
37.4 (27)
19.5 (28)
60.6 (8)
Notes and references
1 J. Micklefield, Curr. Med. Chem., 2001, 8, 1157; J. Kurreck, Eur. J.
Biochem., 2003, 270, 1628.
r(AAGAA) 14.2 (12)
2 M. Petersen and J. Wengel, Trends Biotechnol., 2003, 21, 74.
3 P. E. Nielsen, Curr. Opin. Biotechnol., 2001, 12, 16.
4 F. A. Rogers, K. M. Vasquez, M. Egholm and P. M. Glazer, Proc. Natl.
Acad. Sci. USA, 2002, 99, 16695; H. Parekh-Olmedo, M. Drury and E. B.
Kmeic, Chem. Biol., 2002, 9, 1073.
LysPOM(A)5NH2
43.4 (35)b
LysPNA(A)5NH2
Poly(rU)
Poly(dT)
r(U)20
50.2 (36)c
70.3 (31)
46.0 (43)
53.8 (37)
30.4 (24)
47.9 (36)
24.1 (25)
34.6 (35)
31.0 (25)
55.6 (28)
38.8 (42)
53.0e (22)e
31.8f (16)f
< 10 (12)
48.4 (36)
24.8 (25)
34.8 (34)
d(T)20
5 B. Hyrup and P. E. Nielsen, Bioorg. Med. Chem., 1996, 4, 5; P. E.
Nielsen, Acc. Chem. Res., 1999, 32, 624.
r(U)5
18.3 (11)
n.b.g
n.bg
6 G. Lowe and T. Vilaivan, J. Chem. Soc., Perkin Trans. 1, 1997, 539–560;
G. Lowe, T. Vilaivan and M. S. Westwell, Bioorg. Chem., 1997, 25, 321;
M. D’Costa, V. Kumar and K. N. Ganesh, Org. Lett., 1999, 1, 1513; A.
Püschl, T. Tedeschi and P. E. Nielsen, Org. Lett., 2000, 2, 4161; T.
Vilaivan, C. Khongdeesameor, P. Harnyuttanakorn, M. S. Westwell and
G. Lowe, Bioorg. Med. Chem. Lett., 2000, 10, 2541; M. D’Costa, V.
Kumar and K. N. Ganesh, Org. Lett., 2001, 3, 1281; V. Kumar, P. S.
Pallan, Meena and K. N. Ganesh, Org. Lett., 2001, 3, 1269; T. Vilaivan
and G. Lowe, J. Am. Chem. Soc., 2002, 124, 9326.
7 D. T. Hickman, P. M. King, M. A. Cooper, J. M. Slater and J. Micklefield,
Chem. Commun., 2000, 2251; D. T. Hickman, T. H. S. Tan, J. Morral, P.
M. King, M. A. Cooper, J. M. Slater and J. Micklefield, Org. Biomol.
Chem., 2003, 1, 3277.
a All melting experiments were carried out with 42 mM (conc. in bases) of
each strand in 10 mM K2HPO4, 0.12 M K+, pH 7.0 (total volume 1.0 cm3).
UV absorbance (A260) was recorded with heating at 5 °C min21 from 23 °C
to 93 °C, cooling at 0.2 °C min21 to 15 °C and heating at 0.2 °C min21 to
93 °C. The Tm was determined from the 1st derivative of the slow heating
and cooling curve. b Hypochromic and c hyperchromic shifts are indicated
in parentheses and were calculated as follows: [Abs 93 °C 2 Abs 15 °C] x
100/Abs 93 °C. d LysPOM(T)5LysNH2 and poly(dA) were incubated for 48
h before being subjected to slow thermal denaturation (0.2 °C min21).
e Probable single strand to duplex and f duplex to triplex transitions. g No
binding (hyperchromicity) was observed.
C h e m . C o m m u n . , 2 0 0 4 , 5 1 6 – 5 1 7
517