pH-independent recognition of the dG·dC base pair in triplex DNA: 9-Deazaguanine N7-(2′-deoxyribonucleoside) and halogenated derivatives replacing protonated dC

Article abstract of DOI:10.1002/hlca.200690063

Triplex-forming oligonucleotides (TFOs) containing 9-deazaguanine N ⁷-(2′-deoxyribonucleoside) 1a and halogenated derivatives 1b,c were synthesized employing solid-phase oligonucleotide synthesis. For that purpose, the phosphoramidite building

Full text of DOI:10.1002/hlca.200690063

598
Helvetica Chimica Acta – Vol. 89 (2006)
pH-Independent Recognition of the dG·dC Base Pair in Triplex DNA:
9-Deazaguanine N7-(2’-Deoxyribonucleoside) and Halogenated Derivatives
Replacing Protonated dC
by Frank Seela*, Khalil I. Shaikh, and Thomas Wiglenda
Laboratorium für Organische und Bioorganische Chemie, Institut für Chemie, Universität Osnabrück,
Barbarastrasse 7, D-49069 Osnabrück, and Laboratory of Bioorganic and Biophysical Chemistry, Center
for Nanotechnology (CeNTech), Gievenbecker Weg 11, D-48149 Münster
(phone: +495419692791; fax: +495419692370; e-mail: Frank.Seela@uni-osnabrueck.de)
Triplex-forming oligonucleotides (TFOs) containing 9-deazaguanine N7-(2’-deoxyribonucleoside) 1a
and halogenated derivatives 1b,c were synthesized employing solid-phase oligonucleotide synthesis. For
that purpose, the phosphoramidite building blocks 5a–c and 8a–c were synthesized. Multiple incorpora-
tions of 1a–c in place of dC were performed within TFOs, which involved the sequence of five consec-
utive 1a–c·dG·dC triplets as well as of three alternating 1a–c·dG·dC and dT·dA·dT triplets. These
TFOs were designed to bind in a parallel orientation to the target duplex. Triplex forming properties
of these oligonucleotides containing 1a–c in the presence of Na+ and Mg2+ were studied by UV/melt-
ing-curve analysis and confirmed by circular-dichroism (CD) spectroscopy. The oligonucleotides contain-
ing 1a in the place of dC formed stable triplexes at physiological pH in the case of sequence of five con-
secutive 1a·dG·dC triplets as well as three alternating 1a–c·dG·dC and dT·dA·dT triplets. The
replacement of 1a by 9-halogenated derivatives 1b,c further enhanced the stability of DNA triplexes.
Nucleosides 1a–c also stabilized duplex DNA.
Introduction. – Since the discovery of triple-helical DNA [1], triplex-forming oligo-
nucleotides (TFOs) have attracted attention due to their application in the sequence-
specific recognition of duplex DNA thereby blocking gene expression (ꢀantigene strat-
egyꢁ) [2–4]. Biologically active TFOs were shown to induce targeted mutations [5][6].
Conjugates of oligonucleotides with psoralen and alkylating agents have been used to
demonstrate the accessibility of a particular DNA target sequence in cell nuclei [7][8].
TFOs have found application in oligonucleotide diagnostics [9][10].
Essentially two families of DNA triple helices have been characterized i) those con-
taining the pyrimidine·purine·pyrimidine motif, and ii) others forming a purine·puri-
ne·pyrimidine scheme. These two families differ in their third-strand composition and
orientation. In the more commonly described pyrimidine·purine·pyrimidine motif, the
third pyrimidine strand has been shown to bind within the major groove of duplex
DNA in a parallel orientation with respect to the purine strand to form the triple
helix. In this motif, the third-strand recognition occurs by the formation of dT·dA·dT
and dCH+ ·dG·dC base triades through Hoogsteen pairing [2][3]. The necessary proto-
nation of cytosine limits the use of this method. Much effort has been devoted to
increase the stability of triplexes under neutral conditions. The replacement of cytosine
by 5-methylcytosine increases the stability of triplexes but does not alleviate the pH
dependence [11–14]. Oligonucleotides containing N7-glycosylated purine nucleosides,
ꢂ 2006 Verlag Helvetica Chimica Acta AG, Zürich

Helvetica Chimica Acta – Vol. 89 (2006)
599
acyclic derivatives, or peptide nucleic acid (PNA) analogs have been studied for the
same purpose [15–19]. Also base-modified oligonucleotides incorporating pyra-
zolo[4,3-d]pyrimidines [20][21], 5-halogenated pyrimidines [11], or 6-oxocytidines
[22] were investigated. In particular, the guanine N7-(2’-deoxyribonucleoside) 2, a non-
natural purine analogue of a 2’-deoxycytidine which displays the H-bonding pattern of
a protonated cytidine, shows a high third-strand-binding affinity being specific to the
dG·dC base pair, and the corresponding triplexes are stable under physiological pH
[15–18].
Earlier experiments with halogenated derivatives of 7-deazapurine nucleosides
[23–26] or 8-aza-7-deazapurine nucleosides have shown that 7-halogeno substituents
stabilize the DNA duplex structure significantly [27][28]. The incorporation of 7-chlo-
ro-7-deaza-2’-deoxyguanosine into TFOs in place of 2’-deoxyguanosine leads to a tri-
plex stabilization [29]. This phenomenon was studied on nucleosides related to the pu-
rine constituents of the third strand of triplexes but not on those mimicking the proto-
nated dC residues. We are now investigating the effect of bulky substituents introduced
in the third strand of triplex-forming oligonucleotides. For this, we chose the 9-deaza-
purine (=pyrrolo[3,2-d]pyrimidine) system as a purine surrogate which can be glycosy-
lated regioselectively at the 7-position of the base (purine numbering is used through-
out the discussion section) and can be substituted at the 9-position with any substituent
of choice. This report describes on the synthesis of the phosphoramidite building blocks
containing N7-glycosylated 9-deazaguanine 2’-deoxyribonucleosides 1a–c, which are
remarkably stable structural analogs of the purine nucleoside 2. As the base triad
2·dG·dC is structurally related to that of dCH+ ·dG·dC, oligonucleotides were synthe-
sized for the investigation of their third-strand binding. Preliminary results of a part of
this work have been reported in a short communication [30][31].
Results and Discussion. – 1. Synthesis of Monomers. The synthesis of the 9-deaza-
guanine N7-(2’-deoxyribonucleoside) 1a was already reported [32]. Recently, an
improved synthesis protocol for 1a and also for the 9-halogenated derivatives 1b,c
was published [33]. The partially deprotected intermediates 3a–c and 6a–c obtained
during nucleoside synthesis [33] were used as precursors to prepare phosphoramidite
building blocks for solid-phase oligonucleotide synthesis. The two series of starting
materials differ in their protecting design. In the series 3a–c, the lactam moiety of

600
Helvetica Chimica Acta – Vol. 89 (2006)
the 9-deazaguanine residue is unprotected, while this moiety is protected by a (pivaloyl-
oxy)methyl (=(2,2-dimethyl-1-oxopropoxy)methyl) residue in the series 6a–c. At first,
compounds 3a–c or 6a–c were converted to the 4,4’-dimethoxytrityl derivatives 4a–c
and 7a–c, respectively, with (MeO)2TrCl in pyridine (Scheme). Subsequently, the treat-
ACHTREUNG
AHCTREUNG
ment of 4a–c and 7a–c with 2-cyanoethyl diisopropylphosphoramidochloridite yielded
the phosphoramidites 5a–c and 8a–c, respectively, which were further used in solid-
phase oligonucleotide synthesis.
Scheme
a) (MeO)2
A
G
N
To assure the stability and applicability of the amino-protecting groups during oli-
gonucleotide syntheses, the deprotection of the formamidine protecting group of com-
pounds 3a–c was studied UV-spectrophotometrically in 25% aqueous ammonia solu-
tion (408). The half-life value for the deprotection of the 9-unsubstituted compound
3a measured spectrophotometrically at 260 nm (25% aq. ammonia, 408) was 12.3
min, while the 9-substituted derivatives 3b,c show higher half-life values (3b 22.5
min, 3c 40 min). This demonstrates that halogen substituents increase the stability of
the amino-protecting groups in the pyrrolo[3,2-d]pyrimidine system.

Helvetica Chimica Acta – Vol. 89 (2006)
601
All synthetic intermediates were characterized by 1H-, 13C-, and 31P-NMR spectros-
ACHTREUNG
Table 1. 13C-NMR Chemical Shifts of Nucleosides and Derivativesa)
4a
4b
4c
7a
7b
7c
C(2)b), C(2)c)
C(4)b), C(6)c)
C(4a)b), C(5)c)
C(6)b), C(8)c)
C(7)b), C(9)c)
C(7a)b), C(4)c)
Me(Piv)
MeN
154.5d)
155.1d)
113.9
127.8
102.6
144.9
–
154.7d)
155.1d)
113.9
129.6
90.1
144.8
–
34.4
154.7d)
154.8d)
114.1
130.3
59.2
144.8
–
34.5
153.7d)
154.3d)
112.8
127.9
102.9
145.0
26.9
153.9d)
154.2d)
112.7
129.8
90.3
144.9
26.8
34.6
153.7d)
154.0d)
112.9
131.6
59.3
144.8
26.7
34.6
34.4
34.6
N=CH
C(1’)
C(2’)
157.0
85.4d)
157.0
85.5d)
156.9
85.5d)
157.0
85.2d)
39.9
157.0
85.6d)
39.9
156.9
85.7d)
39.9
e
e
e
)
)
)
C(3’)
C(4’)
C(5’)
OCH2
70.5
85.1d)
64.1
–
70.3
85.4d)
63.9
–
70.4
85.0d)
64.0
–
70.5
70.4
70.5
85.4d)
64.1
85.6d)
64.0
85.6d)
64.0
65.5
65.5
65.5
C=O
–
–
–
176.5
176.8
176.8
a) Measured in (D6)DMSO. b)Systematic numbering. c) Purine numbering. d) Tentative. e) Superimposed
by the DMSO signal.
2. Synthesis and Characterization of Oligonucleotides. Oligonucleotides 9–28 (see
Tables 2 and 3 for Formulae) were synthesized in an automated ABI 392-08 DNA syn-
thesizer with the phosphoramidites 5a–c and 8a–c. The oligonucleotides were purified
by reversed-phase HPLC (see Exp. Part). The composition of the oligonucleotides was
determined by HPLC (RP-18) after tandem enzymatic hydrolysis with snake-venom
phosphodiesterase followed by alkaline phosphatase in 0.1M Tris·HCl buffer (pH
8.3) at 378. The HPLC profiles of the reaction products obtained after enzymatic diges-
tion clearly demonstrate that the 9-bromo- and 9-iodo-substituted compounds 1b,c are
much more hydrophobic (Fig. 1,b and c) compared to the unsubstituted nucleoside 1a
(Fig. 1,a). MALDI-TOF mass spectra were measured for the modified oligonucleoti-
des, correct masses were found in all cases (see Exper. Part).
3. Duplexes Containing Base Pairs of the 9-Deazaguanine N7-(2’-Deoxyribonucleo-
side) 1a and the 9-Halogenated Derivatives 1b,c with m5iCd. Before investigating the sta-
bility of triplexes, the effect of compounds 1a–c on the duplex stability was studied.
Previously, our laboratory had reported on the incorporation of N7-glycosylated 2’-
deoxyadenosine and N7-glycosylated 2’-deoxyguanosine 2 in oligonucleotide duplexes
[34][35]. Compound 2 forms stable base pairs with 2’-deoxy-5-methylisocytidine
(=m5iCd) (duplex 11·12, Table 2) [36]. Now, we incorporated the 9-deazaguanine

602
Helvetica Chimica Acta – Vol. 89 (2006)
Fig. 1. Reversed-phase HPLC profiles after the enzymatic hydrolysis of the oligonucleotides a) 25 con-
taining 1a, b) 26 containing 1b, and c) 27 containing 1c, with snake-venom phosphodiesterase and alka-
line phosphatase in 0.1M Tris·HCl buffer (pH 8.3). Gradient: 20 min A, 20–60 min 50% B in A, flow
rate 0.7 ml/min (A=0.1M (Et3NH)OAc (pH 7.0)/MeCN 95 :5, B=MeCN).
A
N7-(2’-deoxyribonucleosides) 1a–c in duplex DNA and studied their base pairing with
m5iCd (Table 2). According to Table 2, compounds 1a–c form base pairs with m5iCd
(which was deduced from sigmoidal melting profiles). Incorporation of nucleoside 1a
(duplex 13·14) results in a Tm value of 408 (Table 2) which is identical to that of the
duplex 11·12 containing the N7-glycosylated guanine 2 but lower than the duplex
9·10 containing the two canonical base pairs at the corresponding positions (Fig. 2,
motif III). The replacement of 1a by the bulkier 9-functionalized nucleosides 1b,c
increases the Tm value by 18 per modification (see duplex 15·14 and 16·14). This indi-
cates that the bulky 9-substituents of N7-glycosylated 9-deazapurines are well accom-
modated in duplex DNA. These results are in line with observations reported for
other halogenated 7-deazapurine and 8-aza-7-deazapurine 2’-deoxyribonucleosides
[23–28]. However, the stabilization by the halogeno substituents in a base pair of 7-
functionalized 7-deazapurine nucleosides with regular glycosylation sites is slightly
higher than those of the uncommonly glycosylated analogs [26]. Regarding the base-
pair motifs formed between 1a–c and m5iCd, we suggest a similar motif as it was pro-
posed for the purine analog 2 (motif II) [35]. In both cases, tridentate base pairs are
formed similarly to the case of the canonical dG·dC base pair (motif I). The lower sta-
bility of the base pairs shown in the motifs II and III might result from the differences in
overlap of the surface areas of the heterocycles (stacking) as well as from weaker H-
bonding (Fig. 2).
4. Triplexes Containing the 9-Deazaguanine N7-(2’-Deoxyribonucleoside) 1a and 9-
Halogenated Derivatives 1b,c. Protonated cytosine is required to form two H-bonds in
the triplex motif IV (Fig. 3). The pKa value of the monomeric 2’-deoxycytidine is 4.3.

Helvetica Chimica Acta – Vol. 89 (2006)
603
Table 2. Tm Values of Oligonucleotide Duplexes Containing 1a–c and 2a)
DGꢀ310
Duplexes
Tm [8]
DH8
DS8
[kcal/mol]
À10.9
À8.4
[kcal/mol]
[kcal/mol]
5’-d(TAG GTC AAT ACT)-3’ (9)
3’-d(ATC CAG TTA TGA)-5’ (10)
5’-d(TAiC iCTC AAT ACT)-3’ (11)
3’-d(AT2 2AG TTA TGA)-5’ (12)
5’-d(TA1a 1aTC AAT ACT)-3’ (13)
3’-d(ATiC iCAG TTA TGA)-5’ (14)
5’-d(TA1b 1bTC AAT ACT)-3’ (15)
3’-d(ATiC iCAG TTA TGA)-5’ (14)
5’-d(TA1c 1cTC AAT ACT)-3’ (16)
3’-d(ATiC iCAG TTA TGA)-5’ (14)
47
40
40
42
42
À89
À253
À57
À156
À74.9
À64.9
À75.8
À212.6
À181.1
À212.8
À8.9
À9.7
À9.8
a) Measured at 260 nm in 0.1M NaCl, 10 mM MgCl2, and 10 mM Na-cacodylate (pH 7.0) with 5 mM of sin-
gle-strand concentration. iC=2’-deoxy-5-methylisocytidine.
Fig. 2. Base pairs formed by 1a–c and 2 with m5iCd
When incorporated into oligonucleotides, the pKa will be higher but will be still below
pH 7.0. This limits the use of triplex motif IV in vivo as an optimal triplex stability
occurs only under acidic conditions; the intracellular pH is around 7.3. The first solu-
tion to provide more stable dCH+ ·dG·dC triplexes was the use of 2’-deoxy-5-methyl-
cytidine instead of dC [11–14]. This stabilizes the triplex significantly although its pKa
value is only 0.2 units higher than that of dC. Thus stabilization cannot be traced back to
pKa changes. Apparently the hydrophobic Me group has a favorable influence on
entropic changes. Various base modifications were performed to overcome this prob-

604
Helvetica Chimica Acta – Vol. 89 (2006)
Fig. 3. Triplex Motifs of Protonated 2’-Deoxycytidine with the dG·dC (motif IV) and of 1a–c with the
dG·dC Base Pair (motif V)
lem[11][15][19–22]. Earlier, the guanine N7-(2’-deoxyribonucleoside) 2 [15][35] has
been incorporated into the TFOs. It was shown that the nucleoside 2 resembles the rec-
ognition site of a protonated 2’-deoxycytidine. A high third-strand-binding affinity spe-
cific to the dG·dC base pair was observed at neutral pH [15–18]. Also related pyra-
zolo[4,3-d]pyrimidine nucleosides (P1 and P2) were investigated [20][21]. Pyrrolo[3,2-d]-
A
within a triplex structure and can be functionalized with various residues including
any kind of reporter groups at the position 9. Therefore, we incorporated the N7-glyco-
sylated 9-deazapurine 2’-deoxyribonucleoside 1a and its 9-halogen derivatives 1b,c into
the TFOs. Its recognition site mimics that of protonated dC. We then examined the for-
mation of triplexes with oligonucleotides containing 15 base pairs with multiple incor-
porations of the modified nucleosides, both involving either a ꢀconsecutiveꢁ run of non-
canonical residues or those following an ꢀalternatingꢁ scheme. The target duplexes
17·18 and 19·20 shown in Table 3 follow the same sequence which has been already
used by McLaughlin and co-workers [37]. The third strands containing the
dCH+ ·dG·dC motif is expected to bind in parallel orientation to the target duplex
by Hoogsteen base pairing. The triplexes differ from each other by the number of
dCH+ ·dG·dC base triads. In one series, the pyrimidine-motif-forming triplex is con-
structed from consecutive dCH+ residues, while in the other, the dCH+ residues are
alternating with dT thereby diluting the number of dCH+ residues within the particular
DNA fragment from five to three.
The triplex 17·18·24 containing alternating dCH+ ·dG·dC and dT·dA·dT base tri-
ads in the third strand (Hoogsteen strand) is composed of unmodified nucleoside resi-
dues (5’-d(TTT CTT TTC TCT CTT)-3’ (24)). It forms a stable triplex under acidic con-
ditions (pH value 6.5) in the presence of spermine and shows a Tm value of 458 for the
Hoogsteen-strand melting (triplex melting) (Table 3). A decrease in the third-strand
melting from 458 to 178 is observed at pH 8.0. This results from the deprotonation of
the 2’-deoxycytosine residue being necessary for triplex formation. No changes in the
higher Tm value (duplex melting) at 728 are observed at pH 6.5 and 8.0. In the case

Helvetica Chimica Acta – Vol. 89 (2006)
605
Table 3. Tm Values of Oligonucleotide Triplexes Containing 9-Deazaguanine N7-(2’-Deoxyribonucleoside)
1a and Its 9-Halogenated Derivatives 1b,c in Alternating and Consecutive Sequencesa)b)
Triplex (alternating)
Tm [8]
pH 6.5 pH 8.0
21
5’-d(TTT 1a TT TT 1a
T
A
T
T
A
T
T
A
T
T
A
T
1a
G
C
1b
G
C
1c
G
C
T
A
T
T
A
T
T
A
T
T
A
T
1a TT)-3’
17 5’-d(GCGCGAAA
18
22
17 5’-d(GCGCGAAA
18
23
17 5’-d(GCGCGAAA
18
24
17 5’-d(GCGCGAAA
18 3’-d(CGCGCTTT
G
C
AA AA
TT TT
G
C
G
C
AACCCGG)-3’ 45/72
TTGGGCC)-5’
45/72
3’-d(CGCGCTTT
5’-d(TTT 1b TT TT 1b
1b TT)-3’
G
C
AA AA
TT TT
G
C
G
C
AACCCGG)-3’ 48/72
TTGGGCC)-5’
47/72c)
3’-d(CGCGCTTT
5’-d(TTT 1c TT TT 1c
1c TT)-3’
G
C
C
G
C
AA AA
TT TT
TT TT
AA AA
TT TT
G
C
C
G
C
G
C
C
G
C
AACCCGG)-3’ 48/72c) 48/72c)
TTGGGCC)-5’
TT)-3’
AACCCGG)-3’ 45/72
TTGGGCC)-5’
3’-d(CGCGCTTT
5’-d(TTT
C
G
C
17/72
Triplex (consecutive)
25 5’-d(TTT 1a TT TT 1a 1a 1a 1a 1a TT)-3’
19 5’-d(GCGCGAAA
G
C
AA AA
TT TT
G
C
G
C
G
C
G
C
G
C
AACCCGG)-3’ 53/74
TTGGGCC)-5’
52/74
20
26
3’-d(CGCGCTTT
5’-d(TTT 1b TT TT 1b 1b 1b 1b 1b TT)-3’
19 5’-d(GCGCGAAA
G
C
AA AA
TT TT
G
C
G
C
G
C
G
C
G
C
AACCCGG)-3’ 57/74
TTGGGCC)-5’
56/74c)
57/74c)
n.t.d)/74
20
27
3’-d(CGCGCTTT
5’-d(TTT 1c TT TT 1c 1c 1c 1c 1c TT)-3’
19 5’-d(GCGCGAAA
G
C
C
G
C
AA AA
TT TT
TT TT
AA AA
TT TT
G
C
C
G
C
G
C
C
G
C
G
C
C
G
C
G
C
C
G
C
G
C
C
G
C
AACCCGG)-3’ 57/74
TTGGGCC)-5’
TT)-3’
AACCCGG)-3’ 24/74
TTGGGCC)-5’
20
28
3’-d(CGCGCTTT
5’-d(TTT
19 5’-d(GCGCGAAA
20 3’-d(CGCGCTTT
a) Measured at 260 nm in 10.5 mM HEPES (=4-(2-hydroxyethyl)-1-piperazineethansulfonic acid) with 50
mM NaCl, 10 mM MgCl2 and 0.5 mM spermine. b) The concentration of single strand was 1 mM. c) Observed
with poor hypochromicity. d) No triplex formed.
of the triplex 19·20·28, in which a third strand contains five adjacent dC residues (5’-
d(TTT CTT TTC CCC CTT)-3’ (28)), the stability of the triplex at pH 6.5 decreases to
248, and no triplex is formed at pH 8.0. Thus the formation of triplexes is restricted to
pH 6.5, while duplex melting stays almost unchanged (Tm 748) (Table 3). Apparently,
the presence of adjacent charged dCH+ residues leads to a charge repulsion resulting
in a destabilization of the triplex structure [16].
The 9-deazaguanine N7-(2’-deoxyribonucleoside) 1a can form a bidentate Hoogs-
teen pair with dG·dC without protonation thereby resembling the dCH+ ·dG·dC
base-pairing motif (see Fig. 3, motif IV and V). For this purpose, oligonucleotides
21–23 and 25–27 were studied next. Typical first derivatives of the melting profiles
are shown in Fig. 4.
The replacement of dC by nucleoside 1a, as shown in Table 3 results in a stable tri-
plex structure 17·18·21 showing Tm values for duplex and triplex melting similar to

606
Helvetica Chimica Acta – Vol. 89 (2006)
those incorporating dCH+ residues, i.e., 17·18·24. The most significant difference
between the triplex 17·18·24 containing dCH+ the triplex and 17·18·21 containing
1a lies in the finding that a stable triplex is formed at pH 8.0 with the latter one
(458) (Fig. 4,a). At pH 8.0, triplex 17·18·24 showed a rather low stability (178). This
demonstrates that triplex formation takes place pH-independently when dC is replaced
by compound 1a. In particular, replacements of dC residues by 1a with five consecutive
incorporations in the triplex 19·20·25 (Table 3) forms a much more stable triplex (Tm
538 at pH 6.5 and 528 at pH 8.0). The stabilizing effect of 1a is much higher in the con-
secutive sequence motif as compared to the alternating arrangement of bases. We pro-
pose that this stabilization is due to an improved base stacking between adjacent N7-
glycosylated guanine residues, the absence of charges, and an increased surface area
compared to the protonated dC residues. Thus the destabilization of triplexes caused
by adjacent charged dCH+ residues can be circumvented when dC is replaced by
nucleoside 1a, and the stable triplexes can be formed under neutral conditions.
As discussed above, we became also interested in studying the influence of bulky 9-
halogeno substituents incorporated in the same triplexes. For that purpose compound
1a was replaced by its 9-bromo and 9-iodo derivatives 1b,c (Table 3). According to
Table 3, triplexes containing 1b,c exhibit significantly higher Tm values than those con-
taining the 9-unsubstituted nucleoside 1a. These results are in line with our expectation
derived from earlier experiments [11]. Table 3 also demonstrates that the effect of sta-
bilization of triplexes by 9-bromo and 9-iodo nucleosides 1b and 1c, respectively, is sim-
ilar. The presence of 1b and 1c in the ꢀalternatingꢁ triplexes 17·18·22 and 17·18·23
leads to Tm values around 488 at both pH 6.5 and 8.0 (Fig. 4,b). This is also much higher
when compared with that of nucleoside 1a. The effect of stabilization by 1b,c is also sim-
ilar in the case of consecutive incorporations; the triplexes 19·20·26 and 19·20·27)
containing five adjacent incorporations of 1b,c show a Tm value of 578 at pH 6.5 and
8.0 (Fig. 4,c). This clearly reveals that the 9-halogeno substituents of 9-deazapurine
nucleoside 1a stabilize DNA triplexes, and the 9-substituents are well accommodated
in triplex DNA. This stabilizing effect can be attributed to the enhanced stacking inter-
actions of the 9-halogenated 9-deazapurine residues within the triplex structure. Fur-
thermore, the halogeno substituents make the triplex structure more hydrophobic lead-
ing to favorable entropic changes. These results clearly demonstrate that the destabili-
zation of triplexes in case of consecutive incorporations of dCH+ caused by charge–
charge repulsion of consecutive incorporations can be alleviated when dC is replaced
by 1a–c.
Next, the CD spectra of the triplexes formed between the duplex target site and
modified oligonucleotides in the case of alternating as well as of consecutive sequences
were measured (Fig. 5). The appearance of an intense negative short wavelength band
(210–220 nm) in the CD spectra strongly supports the formation of triple-stranded
complexes [38–43]. The absorption at 210–220 nm was observed for both the ꢀalternat-
ingꢁ modified triplexes 17·18·21, 17·18·22, and 17·18·23 (Fig. 5,a) as well as for the
triplexes 19·20·25, 19·20·26, and 19·20·27 with consecutive incorporations (Fig.
5,b) but not for the duplexes 17·18 and 19·20.
Conclusions. – Triplex DNA containing the dCH+ ·dG·dC motif is susceptible to
pH changes as only the protonated dC or its 5-methyl derivative are able to form

Helvetica Chimica Acta – Vol. 89 (2006)
607
Fig. 4. First derivatives of the UV/melting curves of the triplexes a) 17·18·21 (containing 1a), b)
17·18·22 (containing 1b) at pH 6.5, and c) 19·20·27 (containing 1c) at pH 8.0. Measured at 260 nm
in 10.5 mM HEPES, 50 mM NaCl, 10 mM MgCl2, and 0.5 mM spermine.

608
Helvetica Chimica Acta – Vol. 89 (2006)
Fig. 5. CD Spectra of DNA triplexes a) containing alternating 1a–c·dG·dC and dT·dA·dT base trip-
lets (triplexes 17·18·21, 17·18·22, and 17·18·23) and b) containing consecutive 1a–c·dG·dC base
triplets (triplexes 19·20·25, 19·20·26, and 19·20·27). Measured in 10.5 mM HEPES, 50 mM NaCl, 10
mM MgCl2, and 0.5 mM spermine at pH 6.5.
Hoogsteen base pairs. This behavior is a drawback for the triplex application in DNA
diagnostics or antigene therapy. The replacement of dC by the uncommonly linked
9-deazapurine N7-deoxyribofuranoside 1a within TFOs results in stable triplex forma-
tion under neutral conditions. Thus the 9-deazaguanine base mimics the recognition
site of a protonated cytosine. The stability of these triplexes is further enhanced
when the nucleoside 1a is substituted at the 9-position by bromo or iodo substituents
(see 1b,c). The effect of triplex stabilization is particularly pronounced in the case of
consecutive incorporations of 1a–c as compared to alternating 1a–c·dT sequence.

Helvetica Chimica Acta – Vol. 89 (2006)
609
This reveals charge repulsion of protonated dC residues which does not occur in tri-
plexes containing 1a–c. All 9-deazapurine nucleosides are almost completely stable
against N–glycosyl bond hydrolysis, and the bulky 9-halogeno substituents are well
accommodated in the DNA triplexes.
We thank Dr. H. Rosemeyer and Dr. Y. He for the NMR spectra. We also thank M. Dubiel, E. Micha-
lek, and S. Budow for their kind help. Financial support by Roche Diagnostics GmbH (Penzberg, Ger-
many) is gratefully acknowledged.
Experimental Part
General. All chemicals were purchased from Aldrich, Sigma, or Fluka (Sigma-Aldrich Chemie
GmbH, Deisenhofen, Germany). Solvents were of laboratory grade. Thin layer chromatography
(TLC): aluminium sheets, silica gel 60 F254, 0.2 mm layer (VWR, Germany). Column flash chromatogra-
phy (FC): silica gel 60 (VWR, Germany) at 0.4 bar; sample collection with an UltroRac II fractions col-
lector (LKB Instruments, Sweden). UV Spectra: U-3200 spectrometer (Hitachi, Tokyo, Japan); lmax (e) in
nm. CD Spectra: Jasco 600 (Jasco, Japan) spectropolarimeter with thermostatically (Lauda-RCS-6 bath)
controlled 1-cm cuvettes. NMR Spectra: Avance-250 or AMX-500 spectrometers (Bruker, Karlsruhe,
1
Germany), at 250.13 MHz for H and 13C; d in ppm rel. to Me4
A
H3PO4 for 31P; J values in Hz. Elemental analyses were performed by Mikroanalytisches Laboratorium
Beller (Gçttingen, Germany).
UV/Melting Curves. The melting temp. were measured with a Cary-1/3 UV/VIS spectrophotometer
(Varian, Australia) equipped with a Cary thermoelectrical controller. The temperature was measured
continuously in the reference cell with a Pt-100 resistor, and the thermodynamic data of duplex formation
were calculated by the Meltwin 3.0 program [44].
Oligonucleotides. The oligonucleotide syntheses were carried out in an ABI 392-08 DNA synthesizer
(Applied Biosystems, Weiterstadt, Germany) at a 1-mmol scale with the phosphoramidites 5a–c and 8a–c
following the synthesis protocol for 3’-(2-cyanoethyl phosphoramidites) (user manual for the 392 DNA
synthesizer, Applied Biosystems, Weiterstadt, Germany). The coupling efficiency was always higher than
97%. After cleavage from the solid support, the oligonucleotides were deprotected in 25% aq. NH3 soln.
for 14–16 h at 608 [45].
Purification of the 5’-dimethoxytrityl oligomers was performed by reversed-phase HPLC (RP-18;
gradient system (A=0.1M (Et3NH)OAc (pH 7.0)/MeCN 95 :5, B=MeCN): 3 min 20% B in A, 12 min
G
20–50% B in A, and 25 min 20% B in A; flow rate 1.0 ml/min). The soln. was dried and treated with
2.5% CHCl2COOH in CH2Cl2 for 5 min at 08 to remove the 4,4’-dimethoxytrityl residues. The detrity-
lated oligomers were purified by reversed-phase HPLC (gradient: 0–20 min 0–20% B in A; flow rate
1.0 ml/min). The oligomers were desalted (RP-18, silica gel) and lyophilized on a Speed-Vac evaporator
to yield colorless solids which were frozen at À248. For UV/melting analysis of the triplexes, the oligo-
nucleotides (single-strand concentration 1 mmol at 260 nm) were mixed in the buffer soln. containing 10.5
mM HEPES (=2-[4-(2-hydroxyethyl)piperazine]-1-ethansulfonic acid), 50 mM NaCl, 10 mM MgCl2, and
0.5 mM spermine, and heated to 858 for 5 min, allowed to cool to r.t. followed by overnight storage at
48. The temp. was measured continuously by heating with the rate of 0.58/min from 208 to 908.
The enzymatic hydrolysis of the oligonucleotides was performed as described by Seela and Becher
[46] with snake-venom phosphodiesterase (EC 3.1.15.1, Crotallus adamenteus) and alkaline phosphatase
(EC 3.1.3.1, Escherichia coli from Roche Diagnostics GmbH, Germany) in 0.1M Tris·HCl buffer (pH 8.3)
at 378. The reaction products were subjected to reversed-phase HPLC (RP-18, 200”10 mm column; gra-
dient: 20 min A, 20–60 min 50% B in A). Quantification of the constituents were made on the basis of the
peak areas, which were divided by the extinction coefficients e260 of the nucleosides: dT 8800, dC 7300,
dA 15400, 1a 5500, 1b 4000, and 1c 4100.
The molecular masses of the oligonucleotides were determined by MALDI-TOF-MS with a Biflex-
III instrument (Bruker Saxonia, Leipzig, Germany) and 3-hydroxypicolinic acid (3-HPA) as a matrix
(Table 4).

610
Helvetica Chimica Acta – Vol. 89 (2006)
Table 4. Molecular Masses ([M+H]+) of the Oligonucleotides Measured by MALDI-TOF Mass Spec-
trometry
[M+H]+ (calc.)
[M+H]+ (found)
5’-d(TA1a 1aTC AAT ACT)-3’ (13)
5’-d(TA1b 1bTC AAT ACT)-3’ (15)
5’-d(TA1c 1cTC AAT ACT)-3’ (16)
5’-d(TTT 1aTT TT1a T1aT 1aTT)-3’ (21)
5’-d(TTT 1aTT TT1a 1a1a1a 1aTT)-3’ (25)
5’-d(TTT 1bTT TT1b T1bT 1bTT)-3’ (22)
5’-d(TTT 1bTT TT1b 1b1b1b 1bTT)-3’ (26)
5’-d(TTT 1cTT TT1c T1cT 1cTT)-3’ (23)
5’-d(TTT 1cTT TT1c 1c1c1c 1cTT)-3’ (27)
3641.6
3799.2
3895.2
4598.0
4647.1
4913.6
5119.5
5101.9
5401.5
3641.1
3801.1
3895.7
4598.3
4647.5
4915.1
5120.5
5101.2
5401.6
5-[2-Deoxy-5-O-(4,4’-dimethoxytrityl)-b-D-erythro-pentofuranosyl]-2-{[(dimethylamino)methylide-
ne]amino}-3,5-dihydro-4H-pyrrolo[3,2-d]pyrimidin-4-one (4a). Compound 3a [33] (375 mg, 1.16 mmol)
was dried by repeated co-evaporation with anh. pyridine (2”3 ml) and dissolved in anh. pyridine (5
ml). Then, 4,4’-dimethoxytrityl chloride (550 mg, 1.62 mmol) was added at r.t. while stirring, and stirring
was continued for another 5 h. The mixture was poured into a 5% aq. NaHCO3 soln. (30 ml) and
extracted with CH2Cl2 (2”30 ml). The combined org. layer was dried (Na2SO4) and evaporated and
the resulting residue subjected to FC (silica gel, column 3”12 cm, CH2Cl2/Me2CO 9 :1 (300 ml), then
G
CH2Cl2/MeOH 95 :5): 4a (650 mg, 89%). Colorless foam. TLC (silica gel, CH2Cl2/MeOH 9 :1): Rf 0.3.
UV (MeOH): 232 (34700), 252 (29600), 293 (21300). 1H-NMR ((D6)DMSO): 11.18 (s, NH); 8.54 (s,
N=CH); 7.40 (d, J=2.85, HÀC(6)); 7.40–7.23 (m, arom. H); 6.94–6.84 (m, arom. H, HÀC(1’)); 6.13
(br. s, HÀC(7)); 5.31 (m, OHÀC(3’)); 4.29 (m, HÀC(3’)); 3.90 (m, HÀC(4’)); 3.72 (s, 2 MeO); 3.10
(m, MeN, CH2(5’)); 2.98 (s, MeN); 2.36–2.27 (m, CH2(2’)). Anal. calc. for C35H37N5O6 (623.70):
C 67.40, H 5.98, N 11.23; found: C 66.49, H 5.88, N 10.92.
7-Bromo-5-[2-deoxy-5-O-(4,4’-dimethoxytrityl)-b-D-erythro-pentofuranosyl]-2-{[(dimethylamino)-
methylidene]amino}-3,5-dihydro-4H-pyrrolo[3,2-d]pyrimidin-4-one (4b). As described for 4a, with 3b
[33] (400 mg, 1.00 mmol), 4,4’-dimethoxytrityl chloride (440 mg, 1.30 mmol), and pyridine (5 ml): 4b
(625 mg, 89%). Colorless foam. TLC (silica gel, CH2Cl2/MeOH 9 :1): Rf 0.3. UV (MeOH): 231
1
(27300), 260 (21800), 297 (16600). H-NMR ((D6)DMSO): 11.39 (s, NH); 8.50 (s, N=CH); 7.62 (s, HÀ
C(6)); 7.39–7.23 (m, arom. H); 6.87–6.84 (m, arom. H, HÀC(1’)); 5.30 (d, J=4.2, OHÀC(3’)); 4.30
(m, HÀC(3’)); 3.91 (m, HÀC(4’)); 3.72 (s, 2 MeO); 3.13–3.08 (m, MeN, CH2(5’)); 3.00 (s, MeN); 2.39
(m, HaÀC(2’)); 2.28 (m, HbÀC(2’)). Anal. calc. for C35H36BrN5O6 (702.59): C 59.83, H 5.16, N 9.97;
found: C 59.82, H 5.17, N 9.83.
5-[2-Deoxy-5-O-(4,4’-dimethoxytrityl)-b-D-erythro-pentofuranosyl]-2-{[(dimethylamino)methylide-
ne]amino}-3,5-dihydro-7-iodo-4H-pyrrolo[3,2-d]pyrimidin-4-one (4c). As described for 4a, with 3c [33]
(250 mg, 0.58 mmol), 4,4’-dimethoxytrityl chloride (271 mg, 0.80 mmol), and pyridine (4 ml): 4c (359
mg, 86%). Colorless foam. TLC (silica gel, CH2Cl2/MeOH 9 :1): Rf 0.27. UV (MeOH): 231 (25600),
262 (20200), 298 (15600). 1H-NMR ((D6)DMSO): 11.33 (s, NH); 8.49 (s, N=CH); 7.62 (s, HÀC(6));
7.39–7.20 (m, arom. H); 6.90–6.86 (m, arom. H, HÀC(1’)); 5.29 (d, J=4.1, OHÀC(3’)); 4.30 (m, HÀ
C(3’)); 3.91 (m, HÀC(4’)); 3.73 (s, 2 MeO); 3.19–3.09 (m, MeN, CH2(5’)); 3.00 (s, MeN); 2.37 (m,
HaÀC(2’)); 2.27 (m, HbÀC(2’)). Anal. calc. for C35H36IN5O6 (749.59): C 56.08, H 4.84, N 9.34; found:
C 56.13, H 4.77, N 9.40.
5-[2-Deoxy-5-O-(4,4’-dimethoxytrityl)-b-D-erythro-pentofuranosyl]-2-{[(dimethylamino)methylide-
ne]amino}-3-[(2,2-dimethyl-1-oxopropoxy)methyl]-3,5-dihydro-4H-pyrrolo[3,2-d]pyrimidin-4-one (7a).
As described for 4a, with 6a [33] (375 mg, 0.86 mmol), 4,4’-dimethoxytrityl chloride (339 mg, 1.0
mmol), and pyridine (5 ml): 7a (550 mg, 86%). Colorless foam. TLC (silica gel, CH2Cl2/MeOH 95 :5):
1
Rf 0.36. UV (MeOH): 236 (23600), 257 (25200), 298 (18600). H-NMR ((D6)DMSO): 8.56 (s, N=CH);
7.72 (s, HÀC(6)); 7.38–7.20 (m, arom. H); 6.87–6.84 (m, arom. H, HÀC(1’)); 6.16 (s, CH2O); 5.37 (d,

Helvetica Chimica Acta – Vol. 89 (2006)
611
J=4.3, OHÀC(3’)); 4.33 (m, HÀC(3’)); 3.93 (m, HÀC(4’)); 3.72 (s, 2 MeO); 3.15–3.10 (m, MeN,
CH2(5’)); 2.98 (s, MeN); 2.43 (m, HbÀC(2’)); 2.38 (m, HaÀC(2’)); 1.08 (s, 3 Me). Anal. calc. for
C41H47N5O8 (737.84): C 66.74, H 6.42, N 9.49; found: C 66.53, H 6.31, N 9.54.
7-Bromo-5-[2-deoxy-5-O-(4,4’-dimethoxytrityl)-b-D-erythro-pentofuranosyl]-2-{[(dimethylamino)-
methylidene]amino}-3-[(2,2-dimethyl-1-oxopropoxy)methyl]-3,5-dihydro-4H-pyrrolo[3,2-d]pyrimidin-4-
one (7b). As described for 4a, with 6b [33] (262 mg, 0.509 mmol), 4,4’-dimethoxytrityl chloride (300 mg,
0.88 mmol), and pyridine (5 ml): 7b (355 mg, 85%). Colorless foam. TLC (silica gel, CH2Cl2/MeOH
95 :5): Rf 0.3. UV (MeOH): 237 (24000), 265 (25900), 312 (19300). 1H-NMR ((D6)DMSO): 8.52 (s, N=
CH); 7.71 (s, HÀC(6)); 7.38–7.20 (m, arom. H); 6.87–6.84 (m, arom. H, HÀC(1’)); 6.16 (s, CH2O);
5.37 (d, J=4.3, OHÀC(3’)); 4.33 (m, HÀC(3’)); 3.93 (m, HÀC(4’)); 3.72 (s, 2 MeO); 3.15–3.10 (m,
MeN, CH2(5’)); 2.98 (s, MeN); 2.40 (m, HbÀC(2’)); 2.32 (m, HaÀC(2’)); 1.08 (s, 3 Me). Anal. calc. for
C41H46BrN5O8 (816.74): C 60.29, H 5.68, N 8.57; found: C 60.22, H 5.78, N 8.45.
5-[2-Deoxy-5-O-(4,4’-dimethoxytrityl)-b-D-erythro-pentofuranosyl]-2-{[(dimethylamino)methylide-
ne]amino}-3-[(2,2-dimethyl-1-oxopropoxy)methyl]-3,5-dihydro-7-iodo–4H-pyrrolo[3,2-d]pyrimidin-4-
one (7c). As described for 4a, with 6c [33] (150 mg, 0.26 mmol), 4,4’-dimethoxytrityl chloride (170 mg,
0.50 mmol), and pyridine (4 ml): 7c (176 mg, 82%). Colorless foam. TLC (silica gel, CH2Cl2/MeOH
95 :5): Rf 0.3. UV (MeOH): 237 (21500), 267 (25000), 313 (18800). 1H-NMR ((D6)DMSO): 8.51 (s, N=
CH); 7.70 (s, HÀC(6)); 7.39–7.22 (m, arom. H, H); 6.88–6.86 (m, HÀC(1’)); 6.17 (s, CH2O); 5.36 (d,
J=4.3, Hz, OHÀC(3’)); 4.33 (m, HÀC(3’)); 3.93 (m, HÀC(4’)); 3.73 (s, 2 MeO); 3.19–3.11 (m, MeN,
CH2(5’)); 2.99 (s, MeN); 2.41 (m, HbÀC(2’)); 2.31 (m, HaÀC(2’)); 1.09 (s, 3 Me). Anal. calc. for
C41H46IN5O8 (863.74): C 57.01, H 5.37, N 8.11; found: C 57.10, H 5.48, N 7.98.
5-[2-Deoxy-5-O-(4,4’-dimethoxytrityl)-b-D-erythro-pentofuranosyl]-2-{[(dimethylamino)methylidene]-
amino}-3,5-dihydro-4H-pyrrolo[3,2-d]pyrimidin-4-one 3’-(2-Cyanoethyl Diisopropylphosphoramidite)
i
(5a). To a soln. of 4a (300 mg, 0.50 mmol) in anh. CH2Cl2 (8 ml) were added Pr2NEt (210 ml, 1.17
A
mmol) and 2-cyanoethyl diisopropylphosphoramidochloridite (166 ml, 0.75 mmol) while stirring under
Ar at r.t. Stirring was continued for another 30 min, and then CH2Cl2 (30 ml) was added. The soln.
was then washed with 5% aq. NaHCO3 soln. (30 ml), the aq. layer extracted with CH2Cl2 (2”30 ml),
the combined org. layer dried (Na2SO4) and evaporated, the resulting residue subjected to FC (silica
gel, column 6”1.5 cm, CH2Cl2/Me2
U
G
7-Bromo-5-[2-deoxy-5-O-(4,4’-dimethoxytrityl)-b-D-erythro-pentofuranosyl]-2-{[(dimethylamino)-
methylidene]amino}-3,5-dihydro-4H-pyrrolo[3,2-d]pyrimidin-4-one 3’-(2-Cyanoethyl Diisopropylphos-
phoramidite) (5b). As described for 5a, with 4b (200 mg, 0.28 mmol), iPr2
ACHTREUNG
AHCTREUNG
ACHTREUNG
AHCTREUNG
AHCTREUNG
5-[2-Deoxy-5-O-(4,4’-dimethoxytrityl)-b-D-erythro-pentofuranosyl]-2-{[(dimethylamino)methylide-
ne]amino}-3-[(2,2-dimethyl-1-oxopropoxy)methyl]-3,5-dihydro-4H-pyrrolo[3,2-d]pyrimidin-4-one 3’-(2-
Cyanoethyl Diisopropylphosphoramidite) (8a). As described for 5a, with 7a (200 mg, 0.28 mmol), iPr2
ACHTREUNG
AHCTREUNG
methylidene]amino}-3-[(2,2-dimethyl-1-oxopropoxy)methyl]-3,5-dihydro-4H-pyrrolo[3,2-d]pyrimidin-4-
one 3’-(2-Cyanoethyl Diisopropylphosphoramidite) (8b). As described for 5a, with 7b (200 mg, 0.24
mmol), iPr2
NEt (67 ml, 0.39 mmol), 2-cyanoethyl diisopropylphosphoramidochloridite (66 ml, 0.30
G
mmol), and CH2Cl2 (4 ml): 8b (150 mg, 60%). Colorless foam. TLC (CH2Cl2/Me2
CO 9 :1): Rf 0.70. 31P-
G
NMR (CDCl3): 150.02; 149.60.

612
Helvetica Chimica Acta – Vol. 89 (2006)
5-[2-Deoxy-5-O-(4,4’-dimethoxytrityl)-b-D-erythro-pentofuranosyl]-2-{[(dimethylamino)methylide-
ne]amino}-3-[(2,2-dimethyl-1-oxopropoxy)methyl]-3,5-dihydro-7-iodo-4H-pyrrolo[3,2-d]pyrimidin-4-
one 3’-(2-Cyanoethyl Diisopropylphosphoramidite) (8c). As described for 5a, with 7c (100 mg, 0.11
mmol), iPr2
NEt (33 ml, 0.19 mmol), 2-cyanoethyl diisopropylphosphoramidochloridite (33 ml, 0.15
G
mmol), and CH2Cl2 (2 ml): 8c. (90 mg, 73%). Colorless foam. TLC (CH2Cl2/Me2
CO 9 :1): Rf 0.68. 31P-
G
NMR (CDCl3): 149.98; 149.57.
REFERENCES
[1] G. Felsenfeld, D. R. Davies, A. Rich, J. Am. Chem. Soc. 1957, 79, 2023.
[2] H. E. Moser, P. B. Dervan, Science (Washington, D.C.) 1987, 238, 645.
[3] T. Le Doan, L. Perrouault, D. Praseuth, N. Habhoub, J.-L. Decout, N. T. Thuong, J. Lhomme, C. HØl-
›ne, Nucleic Acids Res. 1987, 15, 7749.
[4] C. HØl›ne, Anticancer Drug Des. 1991, 6, 569.
[5] A. F. Faruqi, S. H. Krawczyk, M. D. Matteucci, P. M. Glazer, Nucleic Acids Res. 1997, 25, 633.
[6] A. Majumdar, A. Khorlin, N. Dyatkina, F.-L. M. Lin, J. Powell, J. Liu, Z. Fei, Y. Khripine, K. A.
Watanabe, J. George, P. M. Glazer, M. M. Seidman, Nature Genet. 1998, 20, 212.
[7] C. Giovannangeli, S. Diviacco, V. Labrousse, S. Gryaznov, P. Charneau, C. HØl›ne, Proc. Natl. Acad.
Sci. U.S.A. 1997, 94, 79.
[8] E. S. Belousov, I. A. Afonina, I. V. Kutyavin, A. A. Gall, M. W. Reed, H. B. Gamper, R. M. Wydro,
R. B. Meyer, Nucleic Acids Res. 1998, 26, 1324.
[9] C. HØl›ne, in ꢀTriplex Forming Oligonucleotidesꢀ, Eds. C. Malvy, A. Harel-Bellan, L. L. Pritchard,
and M. A. Norwell, Kluwer Academic, 1999.
[10] V. N. Soyfer, V. N. Potaman, ꢀTriple-Helical Nucleic Acidsꢁ, Springer, New York, 1995.
[11] T. J. Povsic, P. B. Dervan, J. Am. Chem. Soc. 1989, 111, 3059.
[12] L. E. Xodo, G. Manzini, F. Quadrifoglio, G. A. van der Marel, J. H. van Boom, Nucleic Acids Res.
1991, 19, 5625.
[13] S. F. Singleton, P. B. Dervan, Biochemistry 1992, 31, 10995.
[14] G. E. Plum, Y.-W. Park, S. F. Singleton, P. B. Dervan, K. J. Breslauer, Proc. Natl. Acad. Sci. U.S.A.
1990, 87, 9436.
[15] J. Hunziker, E. S. Priestley, H. Brunar, P. B. Dervan, J. Am. Chem. Soc. 1995, 117, 2661.
[16] H. Brunar, P. B. Dervan, Nucleic Acids Res. 1996, 24, 1987.
[17] A. St. Clair, G. Xiang, L. W. McLaughlin, Nucleosides Nucleotides 1998, 17, 925.
[18] M. DꢁCosta, V. A. Kumar, K. N. Ganesh, J. Org. Chem. 2003, 68, 4439.
[19] J. Marfurt, C. Leumann, Angew. Chem. 1998, 110, 184.
[20] J. S. Koh, P. B. Dervan, J. Am. Chem. Soc. 1992, 114, 1470.
[21] I. Radhakrishnan, D. J. Patel, E. S. Priestley, H. M. Nash, P. B. Dervan, Biochemistry 1993, 32,
11228.
[22] U. Parsch, J. W. Engels, Chem.–Eur. J. 2000, 6, 2409.
[23] F. Seela, H. Thomas, Helv. Chim. Acta 1995, 78, 94.
[24] F. Seela, M. Zulauf, Chem.–Eur. J. 1998, 4, 1781.
[25] N. Ramzaeva, F. Seela, Helv. Chim. Acta 1996, 79, 1549.
[26] N. Ramzaeva, C. Mittelbach, F. Seela, Helv. Chim. Acta 1997, 80, 1809.
[27] F. Seela, M. Zulauf, J. Chem. Soc., Perkin Trans. 1 1999, 479.
[28] F. Seela, G. Becher, Helv. Chim. Acta 1999, 82, 1640.
[29] Y. Aubert, L. Perrouault, C. HØl›ne, C. Giovannangeli, U. Asseline, Bioorg. Med. Chem. 2001, 9,
1617.
[30] P. Leonard, T. Wiglenda, F. Seela, Nucleosides, Nucleotides, Nucleic Acids 2001, 20, 1279.
[31] T. Wiglenda, Diplomarbeit, Universität Osnabrück, Osnabrück, Germany, 2001.
[32] N. S. Girgis, H. B. Cottam, S. B. Larson, R. K. Robins, J. Heterocycl. Chem. 1987, 24, 821.
[33] F. Seela, K. I. Shaikh, T. Wiglenda, P. Leonard, Helv. Chim. Acta 2004, 87, 2507.
[34] F. Seela, H. Winter, ꢀ10th International Round Tableꢁ, Park City, Utah, USA. 1992.
[35] F. Seela, P. Leonard, Helv. Chim. Acta 1996, 79, 477.

Helvetica Chimica Acta – Vol. 89 (2006)
613
[36] P. Leonard, Ph. D. Thesis, University of Osnabrück, Osnabrück, Germany, 1998.
[37] G. Xiang, R. Bogacki, L. W. McLaughlin, Nucleic Acids Res. 1996, 24, 1963.
[38] D. M. Gray, A. R. Morgan, R. L. Ratliff, Nucleic Acids Res. 1978, 5, 3679.
[39] D. M. Gray, S.-H. Hung, K. H. Johnson, Methods Enzymol. 1995, 246, 19.
[40] K. Liu, H. T. Miles, J. Frazier, V. Sasisekharan, Biochemistry 1993, 32, 11802.
[41] N. Sugimoto, P. Wu, H. Hara, Y. Kawamoto, Biochemistry 2001, 40, 9396.
[42] A. AviÇꢃ, M. Frieden, J. C. Morales, B. G. de la Torre, R. G. Garcꢄa, F. Azorꢄn, J. L. Gelpꢄ, M.
Orozco, C. Gonzꢅlez, R. Eritja, Nucleic Acids Res. 2002, 30, 2609.
[43] S. D. Osborne, V. E. C. Powers, D. A. Rusling, O. Lack, K. R. Fox, T. Brown, Nucleic Acids Res.
2004, 32, 4439.
[44] M. Petersheim, D. H. Turner, Biochemistry 1983, 22, 256.
[45] ꢀUsers Manual of the DNA Synthesizerꢁ, Applied Biosystems, Weiterstadt, Germany.
[46] F. Seela, G. Becher, Nucleic Acids Res. 2001, 29, 2069.
Received November 17, 2005