(D-1 1995 Oxford University Press
Nucleic Acids Research, 1995, Vol. 23, No. 2
217-222
Efficient pH-independent sequence-specific DNA
binding by pseudoisocytosine-containi ng bis-PNA
Michael Egholm+, Leif Christensen, Kim L. Dueholm, Ole Buchardtt, James Coull1 and
Peter E. Nielsen2,*
Center for Biomolecular Recognition, Department of Organic Chemistry, The H.C. Orsted Institute,
Universitetsparken 5, DK 2100 0 Copenhagen, Denmark, 1Biosearch, Core R & D, 75 A Wiggins Institute,
Bedford, MA 01730, USA and 2Center for Biomolecular Recognition, Department of Medical Biochemistry and
Genetics, Biochemistry Laboratory B, The Panum Institute, Blegdamsvej 3c, DK-2200 N Copenhagen, Denmark
Received November 1, 1994; Revised and Accepted December 16,1994
ABSTRACT
The synthesis and DNA binding properties of bis-PNA
(peptide nucleic acid) are reported. Two PNA segments
each of seven nucleobases in length were connected
in a continuous synthesis via a flexible linker composed of three 8-amino-3,6-dioxaoctanoic acid units.
The sequence of the first strand was TCTCTTT (C- to
N-terminal), while the second strand was TTTCTCT or
TTTJTJT, where J is pseudoisocytosine. These bisPNAs form triple-stranded complexes of somewhat
higher thermal stability than monomeric PNA with
complementary oligonucleotides and the thermal melting transition shows very little hysteresis. When the J
base is placed in the strand parallel to the DNA
complement ('Hoogsteen strand'), the DNA binding
was pH independent. The bis-PNAs were also superior
to monomeric PNAs for targeting double-stranded
DNA by strand invasion.
INTRODUCTION
Homopyrimidine peptide nucleic acids (PNA) (Fig. 1) bind to
complementary DNA or RNA oligonucleotides forming
(PNA)2/DNA (RNA) triplexes of very high thermal stability
(1-3). Due to the high stability of such (PNA)2/DNA triplexes,
binding of PNA to homopurine targets in double-stranded DNA
takes place via strand displacement (1,4,5), instead of conventional triple helix formation as observed with natural oligonucleotides (6,7). While PNA/DNA duplexes are most stable in the
antiparallel configuration (8, PNAs are oriented from the N- to the
C-terminal, thus if the N-end of the PNA faces the 5'-end of the
oligonucleotide the complex is termed parallel), the parallel
orientation seems to be preferred for (PNA)2/DNA triplexes (3)
if only one species of PNA is employed. As would be predicted
from these results, the most stable (PNA)2/DNA triplexes are
formed when the Watson-Crick base pairing PNA strand is in the
anti-parallel orientation relative to the DNA strand (8) and the
Hoogsteen strand is in the parallel orientation relative to the DNA
strand (i.e. the two PNA strands are anti-parallel relative to each
other) (Nielsen et al, in preparation).
As a logical consequence of this, dimeric PNA molecules
(bis-PNAs) with one strand designed for Watson-Crick recognition of DNA (RNA) and the other strand designed for Hoogsteen
recognition of a PNA-DNA duplex should be optimal for
(PNA)2/DNA triplex stability and thus enhance strand displacement binding to double-stranded DNA. We describe here such
bis-PNAs which furthermore contain pseudoisocytosine instead
of cytosine in the 'Hoogsteen strand' to avoid pH sensitivity ofthe
binding.
MATERIALS AND METHODS
Synthesis of N-(2-boc-aminoethyl)-N-(N2(benzyloxycarbonyl)isocytosin-5-ylacetyl)glycine, the
pseudoisocytosine PNA monomer
Methyl a-fornylsuccinate. This procedure is a modification of the
one published by Fissekis and Sweet (9). Sodium methoxide
(40.5 g, 0.75 mol) was suspended in dry ether (500 ml) and stirred
under nitrogen at 0°C. A mixture of dimethyl succinate (65.4 ml,
0.50 mol) and methylformate (123 ml, 2.00 mol) was added
dropwise over 30 min. The reaction mixture was stirred at 0°C for
2 h and then at room temperature overnight. Subsequently, the
reaction mixture was evaporated to a viscous brown residue
which was washed once with petroleum ether and then dissolved
in 3 M hydrochloric acid (160 ml). This solution was made
weakly acidic with concentrated hydrochloric acid and then
extracted with dichloromethane (4x250 ml). The organic phase
was dried (MgSO4), filtered and evaporated under reduced
*To whom correspondence should be addressed
+Present address: Biosearch, Core R & D, 75 A Wiggins Institute, Bedford, MA 01730, USA
tDeceased
218
Nucleic Acids Research, 1995, Vol. 23, No. 2
B
B
N(H
NH
oNJI
PNA (dimer)
H
H
_N2NsH~~~
HN
(,-N
I
HC
..
N
N #,N
N
N
H
G
H
H
I/N
N
H
pseudo-C
H
p'
N
N
C
N
.00N
H
G
Figure 1. Chemical structures of PNA (B is the nucleobase) and of C+-G-C and
J-G-C triplets.
The resulting residue was distilled in a kugelrohr
apparatus at 60°C and 0.6 mbar yielding 52.3 g of a mixture of
the title compound and dimethyl succinate in the molar ratio
80:20 (determined by NMR) as a colorless oil. This mixture might
pressure.
be used directly in the following preparation.
The product might be isolated free of dimethyl succinate by
exchanging the extraction with dichloromethane for a continuous
extraction with diethyl ether. However, in our hands this reduced
the yield to 34%, far from the 62% reported by Fissekis and Sweet
(9). 1H-NMR (DMSO-d&/TMS): 6 = 3.20 (s, 2H, CH2); 3.59 (s,
3H, OMe); 3.61 (s, 3H, OMe); 7.73 (s, 1H, CHOH); 10.86 (br s,
1H, CHQH). 13C-NMR (DMSO-dc/MS): 6 = 28.9 (CH2); 51.0
(OMe); 51.6 (OMe); 102.1 (C=CHOH); 156.6 (CHOH); 168.3
(COO); 171.7 (COO).
Isocytosin-5-ylacetic acid (7). This procedure is a modification of
the one published by Beran et al. (10). Sodium methoxide (41.9
g, 0.78 mol) was dissolved in dry methanol (200 ml) and
guanidine hydrochloride (49.4 g, 0.52 mol) was added. The
mixture was stirred for 10 min under nitrogen at room temperature. A solution of methyl a-formylsuccinate (30.0 g, 0.17 mol)
in dry methanol (100 ml) was added to the mixture. The reaction
mixture was refluxed under nitrogen for 3 h and then stirred at
room temperature overnight. Subsequently, the reaction mixture
was filtered, and the filter cake was washed once with methanol.
The collected filtrate and washing were evaporated under reduced
pressure. The resulting residue was dissolved in water (80 ml) and
the solution was acidified with concentrated hydrochloric acid to
pH 4.2. After having been stirred at 0°C the mixture was filtered,
the precipitate washed once with water and then freeze-dried
leaving 28.29 g (97%) of the title compound as a white solid.
Calculated for C6H7N303 1/2H20: C, 40.45; H, 4.53; N, 23.59.
Found: C, 40.13; H, 4.22; N, 23.26.
Due to the poor solubility properties of the product it was further
characterized as its sodium salt. 7 (0.42 g, 2.5 mmol) and sodium
bicarbonate were dissolved in boiling water (35 ml). The solution
was cooled and evaporated. The residue was dissolved in water (6
ml) and ethanol (4 ml) and isopropanol (8 ml) were added. The
sodium salt of 7 was collected by filtration, washed with absolute
ethanol and petroleum ether and dried to yield 0.31 g (65%) as white
crystals. IH-NMR (D2O/TMS): 6 = 3.10 (s, 2H, CH2COO); 7.40 (s,
1H, H-6). 13C-NMR (DMSO-d4/MS): 6 = 34.8 (CH2COO); 112.0
(C-5); 145.6-146.5 (m, C-2); 155.1 (C-6); 169.4 (C4); 179.3
(CO/OH). MS (FAB+) m/z (%): 192 (100, M+H).
Methyl isocytosin-5-ylacetate (8). Thionylchloride (3.6 ml, 50
mmol) was added to stirred methanol (210 ml) at -40'C under
nitrogen. Isocytosin-5-ylacetic acid (7, 7.0 g, 41 mmol) was added
and the reaction mixture was stirred at room temperature for 1 h,
at 60°C for 3 h and overnight at room temperature. The reaction
mixture was evaporated to dryness and the residue was dissolved
in saturated aqueous sodium bicarbonate (80 ml) giving a foamy
precipitate. 4 M hydrochloric acid was added to pH 6.5 and the
suspension was stirred for 1 h. The precipitate was collected by
filtration, washed with water, recrystallized from water and
freeze-dried yielding 4.66 g (62%) of methyl isocytosin-5-ylacetate as white crystals.
IH-NMR (DMSO-dW/MS): 8 = 3.28 (s, 2H, CH2COO); 3.64
(s, 3H, COOMe); 6.87 (br s, 2H, NH2); 7.54 (s, 1H, H-6).
13C-NMR (DMSO-dW/TMS): 6 = 32.0 (CH2COO); 51.5
(COOMe); 108.4 (C-5); 153.3 (C-2); 156.4 (C-6); 164.0 (C4);
171.8 (CH2COO). MS(FAB+) m/z (%): 184 (100, M+H).
Calculated for C7H9N303 3/2H20: C, 40.00; H, 5.75; N, 19.99.
Found: C, 40.18; H, 5.46; N, 20.30.
Methyl N2-(benzyloxycarbonyl)isocytosin-5-ylacetate (9). Methyl
isocytosin-5-ylacetate (8, 9.5 g, 52 mmol) was dissolved in dry
DMF (95 ml) and the solution was stirred at 0°C under nitrogen.
N-Benzyloxycarbonyl-N'-methylimidazolium triflate (37.99 g,
104 mmol) was added slowly and the reaction mixture was stirred
for 30 min at 0°C and then overnight at room temperature.
Dichloromethane (800 ml) was added and the resultant mixture
was washed with half-saturated aqueous sodium bicarbonate
(2x400 ml), half-saturated aqueous potassium hydrogen sulfate
(2x400 ml) and with brine (lx400 ml). The organic phase was
dried (MgSO4), filtered and evaporated under reduced pressure.
The residue was recrystallized from methanol affording 13.32 g
(81 %) of the title compound as white crystals.
IH-NMR (DMSO-d6/TMS): 6 = 3.43 (s, 2H, CH2COO); 3.67
(s, 3H, COOMe); 5.30 (s, 2H, PhCH2); 7.43-7.52 (m, 5H,
PhCH2); 7.77 (s, 1H, H-6). 13C-NMR (DMSO-d6/TMS): 6= 31.9
(CH2COO); 51.6 (COOMe); 67.0 (PhCH2); 128.1-128.5 (m,
PhCH2); 135.7 (PhCH2); 150.7 (Z-CO); 170.8 (COO). MS
(FAB+) m/z (%): 318 (3.5, M+H) Calculated for C15HI5N305: C,
56.78; H, 4.76; N, 13.24. Found: C, 56.68; H, 4.79; N, 13.28.
Nucleic Acids Research, 1995, Vol. 23, No. 2
N2-(Benzyloxycarbonyl)isocytosin-5-ylacetic acid (10). Methyl
N2-(benzyloxycarbonyl)isocytosin-5-ylacetate (9, 5.2 g, 16
mmol) was suspended in THF (52 ml) and cooled to 0°C. 1 M
lithium hydroxide (49 ml, 49 mmol) was added and the reaction
mixture was stirred at 0°C for 25 min. Additional 1 M lithium
hydroxide (20 ml, 20 mmol) was added and the mixture was
stirred at 0°C for 90 min. The product was precipitated by
acidifying to pH 2 with 1 M hydrochloric acid, collected by
filtration, washed once with water and dried. The yield was 4.12
g (83%) of white crystals.
'H-NMR (DMSO-d6/TMS): 6 = 3.33 (s, 2H, CH2COO); 5.29
(s, 2H, PhCH2); 7.43-7.52 (m, SH, PhCH2); 7.74 (s, 1H, H-6);
11.82 (br s, 3H, exchangable protons). MS (FAB+) m/z (%): 304
(12, M+H). Calculated for C14HI3N3O5: C, 55.45; H, 4.32; N,
13.86. Found: C, 55.55; H, 4.46; N, 13.84.
Ethyl N-(2-Boc-aminoethyl)-N-(N2.(Benzyloxycarbonyl) iso cytosin-5-ylacetyl)glycinate (11). N2-(Benzyloxycarbonyl)isocytosin-5-ylacetic acid (10, 2.0 g, 6.6 mmol) was transferred to a flask
equipped with a stirring bar and a septum through which a flow of
nitrogen was applied. Dry DMF (20 ml) and N-methylmorpholine
(2.2 ml, 19.8 mmol) were added. The mixture was cooled to 0°C
and ethyl N-(2-Boc-aminoethyl)glycinate (1.8 g, 7.3 mmol) and
O-benzotriazol-1-yl-N,N,N',N'-tetramethyluronium hexafluorophosphate (HBTU, 3.0 g, 7.9 mmol) were added. The reaction
mixture was stirred under nitrogen for 4 h followed by addition of
dichloromethane (100 ml). The organic phase was washed with
half-saturated aqueous sodium bicarbonate (2x75 ml), half-saturated aqueous potassium hydrogen sulfate (2x75 ml) and with
brine (1x75 ml), dried (MgSO4), filtered and evaporated under
reduced pressure. The residue was dissolved in ethyl acetate (50
ml), stirred at 0°C for 10 min and filtered through celite which was
washed with ethyl acetate. The collected filtrate and washing were
concentrated to a volume of 10 ml. Diethyl ether (100 ml) was
added and the resultant solution was stirred overnight at room
temperature. The product was collected by filtration, washed once
with diethyl ether and dried to yield 2.6 g (74%) of the title
compound as white crystals.
Due to hindered rotation around the amide bond several of the
NMR signals are comprised of a major (ma) and minor (mi)
component.
1H-NMR (DMSO-d6/TMS): 6 = 1.20-1.30 (m, 3H, CH2CH3);
1.45 (s, 9H, Boc); 3.05-3.52 (m, 6H, NCH2, CH2N, CH2CON);
4.08 and 4.40 (s, ma and s, mi, respectively, 2H, CH2COO); 4.15
and 4.25 (q, ma, J=7 Hz and q, mi, respectively, 2H, QF2CH3);
5.29 (s, 2H, PhCH2); 7.40-7.52 (m, SH, PhCH2); 7.64 and 7.67
(s, mi and s, ma, respectively, 1H, H-6). 13C-NMR (DMSOd64TMS): 6 = 14.1 (CH2CH3); 28.2 (Boc); 30.2 and 30.5 (ma and
mi, respectively, CH2CON); 37.9 and 38.3 (mi and ma, respectively, NCH2); 47.7 and 48.0 (ma and mi, respectively, CH2N);
50.2 (CH2COO); 60.4 and 61.0 (ma and mi, respectively,
CH2CH3); 67.0 (PhCH2); 127.9-128.5 (m, PhCH2); 135.8
(PhCH2); 155.7 (C-6), 169.4 (CON); 170.0 (COO). MS (FAB+)
m/z (%): 532 (3.5, M+H); 432 (3.5, M-Boc+H). Calculated for
C25H33N508: C, 56.49; H, 6.26; N, 13.17. Found: C, 56.46; H,
6.14; N, 12.86.
N-(2-Boc-aminoethyl)-N-(N2-(Benzyloxycarbonyl)isocytosin-5-
ylacetyl)glycine (12). Ethyl N-(2-Boc-aminoethyl)-N-(N2-(benzyloxycarbonyl)isocytosin-5-ylacetyl)glycinate (11, 1.6 g, 3.0 mmol)
219
was dissolved in methanol (16 ml) by gentle heating. The solution
was cooled to 0°C and 2 M sodium hydroxide (23 ml) was added.
The reaction mixture was stirred at room temperature for 75 min
and then cooled to 0°C again. The pH was adjusted to 1.7 and the
product was collected by filtration, washed once with water and
dried to give 1.24 g (82%) of 12 as white crystals. Due to hindered
rotation around the amide bond several of the NMR signals are
comprised of a major (ma) and minor (mi) component. 1H-NMR
(DMSO-d6/TMS): 6 = 1.45 (s, 9H, Boc); 3.05-3.52 (m, 6H,
NCH2, CH2N, CH2CON); 4.01 and 4.29 (s, ma and s, mi,
respectively, CH2COO); 5.29 (s, 2H, PhCH2); 7.40-7.51 (m, SH,
PhCH2); 7.63 and 7.68 (s, mi and s, ma, respectively, 1H, H-6).
13C-NMR (DMSO-d6/TMS): 6 = 28.2 (Boc); 30.2 and 30.5 (ma
and mi, respectively, CH2CON); 37.9 and 38.3 (mi and ma,
respectively, NCH2); 47.5 and 47.9 (ma and mi, respectively,
CH2N); 50.2 (CH2COO); 67.0 (PhCH2); 128.0-128.5 (m,
PhCH2); 135.8 (1hCH2); 150.5 (Z-CO); 155.7 (C-6); 169.9 and
170.3 (ma and mi, respectively, CON); 170.8 and 171.2 (ma and
mi, respectively, COO). MS (FAB+) n/z (%): 504 (16, M+H);
448 (3.5, M-t-Bu+H); 404 (23, M-Boc+H).
PNA oligomerization was performed as previously described
(2,3), and the PNA and egl linker monomers were obtained from
BioSearch (Massachusetts). The PNA oligomers were purified by
reversed phase HPLC and characterized by FAB mass spectroscopy. All PNAs produced an M+1 peak within 1-2 mass units
of the calculated value.
Strand displacement binding of bis-PNAs to a
32P-labeled double-stranded DNA fragment probed by
KMnO4 oxidation
A 32P-end labeled EcoRI-PvuII fragment of the plasmid pTHa
(11) was incubated with PNA in 100 [tl 10 mM Na phosphate, 1
mM EDTA, pH 7 for 60 min at 20°C, and subsequently treated
with KMnO4 (20 mM for 15 s). Following precipitation and
treatment with piperidine (1,7) the samples were analyzed by
electrophoresis in polyacrylamide sequencing gels and the
radioactive DNA bands visualized by autoradiography.
RESULTS AND DISCUSSION
By preparing the bis-PNAs in a continuous synthesis the two
PNA moieties will form an antiparallel clamp. A flexible,
hydrophilic linker based on 8-amino-3,6-dioxaoctanoic acid (egl)
was used to connect the two PNA segments, and computer model
building (using the Biosym software InsightIl and Discover)
indicated that three linker units would be suitable. The bis-PNAs
were synthesized by standard solid phase t-Boc-peptide chemistry as previously reported (2,3), purified by reversed phase
HPLC, and characterized by FAB mass spectrometry.
The hybridization results presented in Table I show that a small
but significant increase in Tm is gained by linking the two PNAs
together (1 versus 4 + 5). As expected, only a small difference is
observed when comparing DNA targets of opposite polarity. The
slight difference is ascribed to the effect of the linker which is
positioned at the 3'- and 5'-end of the oligonucleotide, respectively. The Tm also exhibits a strong pH dependence due to the
requirement for protonation of the cytosines in the Hoogsteen
strand.
220
Nucleic Acids Research, 1995, Vol. 23, No. 2
Table 1. Thermal stabilities (Tm) of bis-PNA-oligonucleotide complexes
PNA
H-TCTCT3-(egl)3-T3CTCT-LysNH2 (1)
H-TJTJT3-(egl)3-T3CTCT-LysNH2c (2)
pH
7
Oligonucleotide Ib
69.0
49.0
9
38.5
5
67.0
64.0
60.5
5
7
9
H-TCTCT3-(egl)3-T3JTJT-LysNH2 (3)
5
66.0
47.0
37.5
50.0
7
40.0
9
-
5
62.0
45.0
34.5
5
7
9
H-TCTCT3-LysNH2 (4)
4 + H-T3CTCT-LysNH2 (5)
7
9
4 + H-T3JTJT-LysNH2 (6)
Tma (oC)
5
7
9
46.0
36.0
37.0
Oligonucleotide Ilb
68.5
52.0
41.0
65.0
48.5
39.0
61.5
60.0
59.0
55.5
39.0
-23.5
65.5
50.0
39.5
48.5
44.0
42.5
aBuffer: 100 mM NaCI, 10 mM Na-phosphate, 0.1 mM EDTA. Heating rate: 0.50C/min at 5-900C
boligonucleotide sequences: I, 5'-d(CGCAGAGA3CGC); II, 5'-d(CGCA3GAGACGC)
cThe hydrogen bond donor/acceptor pattern of pseudoisocytosine is such that dependent on the tautomer it may function
both as a cytosine analog for Watson-Crick base pairing with guanine and as a protonated cytosine analog for Hoogsteen
base pairing with guanine. Due to this dual recognition property we refer to the pseudoisocytosine base as J for the Greek
god Janus with two faces.
It has previously been shown that changing cytosine to
pseudoisocytosine in the Hoogsteen strand (Fig. 1) abolishes the
pH sensitivity of the DNA triplex (12,13) without seriously
affecting the stability of the triplex. Likewise, 8-oxoadenine has
been employed to recognize guanine in the Hoogsteen mode
(14-17). However, synthesis of the PNA pseudoisocytosine
monomer seemed more straightforward (Scheme 1). We therefore synthesized bis-PNAs with thymines and cytosines in the one
strand and thymines and pseudoisocytosines in the other PNA
strand (2 and i). The complexes ofthese PNAs with complementary oligonucleotides showed thermal stabilities at acidic pH (pH
5) comparable to that of the bis-PNA 1 complex. However, in the
complexes where the cytosine PNA strand is anti-parallel to the
DNA target (and thus the pseudoisocytosine strand is parallel),
almost no pH dependence of the Tm is observed (complexes 2/I
and _/II), indicating that the orientation directs the complex
formation in a way that orients the Watson-Crick strand in the
anti-parallel configuration and thus the Hoogsteen strand in the
parallel configuration. Concordently, the complexes 2/II and 3/I
show pH dependence of the Tm, indicating that the cytosine PNA
strand in these cases is involved in Hoogsteen hydrogen bonding.
Finally, we note that the pronounced hysteresis of the melting
behavior usually observed with (PNA)2/DNA triplexes (-30°C
for PNA 4), indicating a very slow rate of triplex formation, is
dramatically reduced in the case of bis-PNAs (2-3°C for PNAs
,-). This is ascribed to the high local concentration of the now
covalently linked second PNA strand. Thus significantly higher
DNA affinity, due to a faster on-rate, would be predicted for
triplex-forming bis-PNAs as compared to monomeric PNAs.
As shown in Table 2, the sequence discrimination of the
bis-PNA 3, as judged from thermal stability measurements, is
better than that found for monomeric PNAs and the very high cost
in stability (-30°C) for a base mismatch reflects the 2-fold
recognition process involving both PNA strands.
Table 2. Effect of base pair mismatches on PNA/DNA thermal stabilities (Tm)a
Oligonucleotide
5'-d(CGCA3GAGACGC)-3'
5'-d(CGCA3CAGACGC)-3'
5'-d(CGCA3AGAGACGC)-3'
5'-d(CGCA3!AGACGC)-3'
5'-d(CGCA3CACACGC)-3'
Tm (°C)
PNA3
60.0
27.0
36.5
23.0
<11
aMeasured at pH 7.0 as described in Table 1.
PNA4+6
44.0
32.5
34.0
33.0
<11
Nucleic Acids Research, 1995, Vol. 23, No. 2
A
NH2
HN
N
ii
I
B
C
221
D
C 1 2 3 4 1 2 3 4 1 2 3 4 1 2 3 4
0
COOH
....;...
8
7
NH-Z
NH-Z
HN
'N
...
0
HN N
0 ~~~~~~~~~~1
* 0e
COOH
COOMe
10
9
NH-Z
NH-Z
HN
HN
N
v
0
^
Ns,, COOEt
11
N
, a
0
0
0
Boc-HN
a
a
Boc-HN
~-
wI
Scheme 1. (i) SOCl2 in MeOH, overnight, 62%. (ii) N-BenzyloxycarbonylN'-methylimiazolium triflate in DMF, overnight, 81 %. (iii) LiOH in
H20/THF, 2h, 83%. (iv) Ethyl N-(2-Boc-aminoethyl)glycinate, HBTU,
N-methylmorpholine in DMF, 4h, 74%. (v) NaOH in H20/MeOH, 75 min,
ACKNOWLEDGEMENTS
This work was supported by ISIS Pharmaceuticals, The Danish
National Research Foundation and the Alfred Benzon Foundation (a fellowship to K.L.D.).
-T#
\C'
82%.
Binding to double-stranded DNA was analyzed using a 32P-end
labeled DNA fragment containing the PNA target and probing for
strand displacement using KMnO4, since thymines in the
displaced loop will react with KMnO4 (1,5). The results (Fig. 2)
show that all three dimeric PNAs bind efficiently to doublestranded DNA, whereas the monomeric PNA 4 does not (only a
very weak thymine oxidation is seen even at the highest PNA
concentration; lane A4). The observed KMnO4 sensitivity of the
thymines outside the PNA target is ascribed to increased DNA
breathing proximal to the bound PNA. The most efficient binding
is, not surprisingly, observed with PNA 3 having the cytosines in
the anti-parallel (Watson-Crick) PNA strand and the pseudoisocytosine in the parallel (Hoogsteen) PNA strand. Furthermore,
the binding did not exhibit the strong pH dependence observed
with PNAs containing cytosines in the Hoogsteen strand (5).
In conclusion, we have shown that bis-PNAs with superior
DNA binding properties can be constructed by incorporation of
nucleobases chosen independently for optimal Watson-Crick and
Hoogsteen recognition.
a
2#
N _COOH
12
*
\
1~~~~
Figure 2. Strand displacement binding of bis-PNAs to a 32P-labeled
double-stranded DNA fragment probed by KMnO4 oxidation. The following
concentrations of PNA were used: 1, 3, 10 or 30 jiM (lanes 1-4, respectively),
and the PNAs were: (A) PNA 4, (B) PNA 1, (C) PNA 2, (D) PNA 3. Lane C
is a control without PNA. An A/G sequence reaction was used to assign the
bands.
REFERENCES
1 Nielsen, P.E., Egholm, M., Berg, R.H. and Buchardt, 0. (1991) Science
254, 1497-1500.
2 Egholm, M., Buchardt, O., Nielsen, P.E. and Berg, R.H. (1992) J. Am.
Chem. Soc. 114, 1895-1897.
3 Egholm, M., Buchardt, O., Nielsen, P.E. and Berg, R.H. (1992) J. Am.
Chem. Soc. 114, 9677-9678.
4 Cherny, D.Y., Belotserkovskii, B.P., Frank-Kamenetskii, M.D., Egholm,
M., Buchardt, O., Berg, R.H. and Nielsen, P. E. (1993) Proc. Natl. Acad.
Sci. USA 90, 1667-1670.
5 Nielsen, PE., Egholm, M. and Buchardt,O (1994) J. Mol. Recognition 7,
164-170.
6 Arnott, S.A., Bond, PJ., Selsing, E. and Smith, P.J.C. (1976) Nucleic Acids
Res. 3, 2459-2470.
7 Moser, H. and Dervan, P.B. (1987) Science 238, 645-650.
222
Nucleic Acids Research, 1995, Vol. 23, No. 2
8 Egholm, M., Buchardt, O., Christensen, L., Behrens, C., Freier, S.M.,
Driver, D.A., Berg, R.H., Kim, S.K., Norden, B. and Nielsen, P.E.(1993)
Nature 365, 566-568.
9 Fissekis, J.D. and Sweet, F. (1970) Biochemistry 9, 3136-3142.
10 Beran et al. (1983) Collect. Czech. Chem. Commun. 48, 292-298.
11 Nielsen, P.E. (1992) Nucleic Acids Res. 20, 2735-2739.
12 Ono, A., Ts'o, POP. and Kan, L.J. (1991) J. Am. Chem. Soc. 113,
4032-4033.
13 Ono, A., Ts'o, P.O.P and Kan, L. (1992) J. Org. Chem. 57, 3225-3230.
14 Miller, P.S., Bhan, P., Cushman, C.D. and Trapane,T.L. (1992)
Biochemistry 31, 6788-6793.
15 Jetter, M.C. and Hobbs, F.W. (1993) Biochemistry 32, 3249-3254.
16 Krawczyk, S.H., Milligan, J.F., Wadwani, S., Moulds, C., Froehier, B.C.
and Matteucci, M.D. (1992) Proc. Natl. Acad. Sci. USA 89, 3761-3764.
17 Davison, E.C. and Johnsson, K. (1993) Nucleosides Nucleotides 12,
237-243.