Pyrenylmethyldeoxyadenosine: A 3′-cap for universal DNA hybridization probes

Article abstract of DOI:10.1002/chem.200801587

A ligand that stabilizes a three-dimensional structure can be expected to have a positive effect on the specificity with which this structure is formed. Here we report on a ligand covalently linked to an oligonucleotide that increases duplex stability, but decreases base-pairing selectivity at the terminus. The ligand consists of a dangling 2′-deoxyadenosine residue with a pyrenylmethyl substituent at the N6-position, that is, a deoxynucleoside with a covalently linked polycyclic aromatic hydrocarbon (PAH). In the presence of the pyrene-bearing nucleosides the UV melting point (ΔT_m) of duplexes increases by up to 29.1°C. The modified residue lowers the base-pairing fidelity at the terminal and penultimate position of duplexes with a depression in ΔΔT_m observable in 20 out of 24 sequence contexts tested. The effect can be rationalized based on a modeled three-dimensional structure. The results are significant for the understanding of base-pairing fidelity in DNA duplexes as modulated by the presence of a polycyclic aromatic hydrocarbon. The fidelity-decreasing effect may be useful for universal hybridization probes that bind to a broader range of sequences than conventional oligonucleotides.

Full text of DOI:10.1002/chem.200801587

DOI: 10.1002/chem.200801587
Pyrenylmethyldeoxyadenosine:
A 3’-Cap for Universal DNA Hybridization Probes
Michael Printz[a] and Clemens Richert*[a, b]
Abstract: A ligand that stabilizes a
matic hydrocarbon (PAH). In the pres-
ence of the pyrene-bearing nucleosides
the UV melting point (DTm) of duplex-
es increases by up to 29.18C. The
modified residue lowers the base-pair-
ing fidelity at the terminal and penulti-
mate position of duplexes with a de-
pression in DDTm observable in 20 out
of 24 sequence contexts tested. The
three-dimensional structure can be ex-
pected to have a positive effect on the
specificity with which this structure is
formed. Here we report on a ligand co-
valently linked to an oligonucleotide
that increases duplex stability, but de-
creases base-pairing selectivity at the
terminus. The ligand consists of a dan-
gling 2’-deoxyadenosine residue with a
pyrenylmethyl substituent at the N6-
effect can be rationalized based on a
modeled three-dimensional structure.
The results are significant for the un-
derstanding of base-pairing fidelity in
DNA duplexes as modulated by the
presence of a polycyclic aromatic hy-
drocarbon. The fidelity-decreasing
effect may be useful for universal hy-
bridization probes that bind to a broad-
er range of sequences than convention-
al oligonucleotides.
Keywords: base pairing · DNA rec-
ognition · molecular caps · noncova-
lent interactions · oligonucleotides
position, that is,
a deoxynucleoside
with a covalently linked polycyclic aro-
Introduction
duplexes also have a greater percentage of terminal base
pairs, where base-pairing fidelity may be poor.
Base pairing at the termini of duplexes is important for bio-
logical and biotechnological processes. For example, the
step-wise formation of complementary chains during replica-
tion, transcription and PCR[1] occurs at the termini of
double-helical regions. Terminal base pairs also dominate
the molecular recognition in the three-nucleotide duplexes
between codons and anticodons that direct translation in ri-
bosomes.[2] Finally, hybridization between probe strands and
genomic DNA, induced during Southern blotting and ex-
pression analysis on microarrays[3] usually leads to short du-
plexes, where base pairs at the termini are significant, both
energetically and in terms of sequence selectivity. The short-
er the probe, the greater the effect of a single mismatch,
and thus the greater the base-pairing selectivity. But, shorter
Despite the critical contributions to the integrity and rep-
lication of genes and their translation into proteins, less is
known about the base pairing at the termini than about base
pairing in the interior of duplexes. The “wobble hypothe-
sis”[4] has helped to explain the degeneracy of the genetic
code, but does not allow for more than qualitative predic-
tions outside the ribosome, and the fidelity of base pairing,
particularly at the termini, remains difficult-to-predict.[5] Li-
gands that enhance stability and base-pairing fidelity at the
termini when covalently attached to one of the duplex-form-
ing strands have been identified, but the structural diversity
among these ligands, ranging from steroids[6] to aromatic
systems like quinolones,[7,8] and stilbenes,[9,10] has made it dif-
ficult to draw general conclusions.
One class of potential ligands for DNA with obvious
shape complementarity are bi- and polycyclic aromatic com-
pounds. Large aromatic substituents, most prominently
pyrene, are used to assist the detection of nucleic acids in
solution,[11] but also to generate functional molecules with
new properties. This includes fluorescent materials,[12] build-
ing blocks for nanomaterials,[13,14] and triplex-forming
strands.[15] Further, the interactions of large aromatic hydro-
carbons with DNA are important for their toxicity, especial-
ly their carcinogenic potential.[16] Therefore, it is critical to
[a] Dr. M. Printz, Prof. C. Richert
Institut fꢀr Organische Chemie, Universitꢁt Karlsruhe (TH)
76131 Karlsruhe (Germany)
[b] Prof. C. Richert
Institut fꢀr Organische Chemie, Universitꢁt Stuttgart
70569 Stuttgart (Germany)
Fax : (+49)711-685-64321
E-mail: clemens.richert@oc.uni-stuttgart.de
Supporting information for this article is available on the WWW
under http://dx.doi.org/10.1002/chem.200801587.
3390
ꢂ 2009 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Chem. Eur. J. 2009, 15, 3390 – 3402

FULL PAPER
understand how such systems, with their high propensity for
p-stacking, affect the base-pairing fidelity. They may have
the same effect on base-pairing fidelity as the active site of
polymerases, causing an increase in fidelity by stabilizing
canonical base pairs more than mismatched base pairs. They
may also stabilize specific mismatches that they are shape-
complementary to. That PAHs may do neither of the two,
but have a general fidelity-decreasing effect is not obvious.
Here we report on such an effect that we encountered when
testing small molecule ligands covalently attached to the ter-
mini of oligonucleotides (Figure 1),[6,10,17,18] as part of a proj-
ect aimed at generating high fidelity hybridization probes
featuring molecular caps at the 3’-terminus.[19]
was chosen for our screening assay. Both the linker and the
ligand itself were varied. Figure S5 (Supporting Informa-
tion) provides an overview of all structures included in our
screening. The ligand was chosen to be large and rigid
enough for stacking on base pairs. This led to two types of
3’-substituents. One contains carboxylic acid residues at-
tached via a short, prolinol-based linker. Reed and co-work-
ers had previously synthesized acridinylprolinol- and choles-
terylprolinol-substituted oligonucleotides and tested their bi-
ostability, cytotoxity and fluorescence properties.[20] The
other was based on deoxyadenosine residues featuring an ar-
omatic substituent at the N6-position.
The synthesis of prolinol phosphoramidite 2 is shown in
Scheme 1. It starts from commercially available hydroxypro-
line methyl ester 3, which was converted to 4 by a known
procedure.[21] After reduction with LiBH4, hydroxyprolinol 5
was obtained in 92% yield. Phosphitylation gave 2 in 88%
yield.
Figure 1. Duplex formation reinforced by bridging interactions between
3’-cap and terminal base pair.
Results
In our current work, oligodeoxynucleotides with covalently
attached ligands were prepared in search for a general 3’-
cap, that is, a 3’-substituent that can be introduced via phos-
phoramidite chemistry and that provides increased target af-
finity for any sequence. In order to get easily measurable ef-
fects on the duplex melting point and to limit the synthetic
effort for the unmodified portion of the oligonucleotides,
the self-complementary DNA hexamer 5’-TGCGCA-3’ (1)
Scheme 1. Synthesis of prolinol phosphoramidite building block 2.
Scheme 2. Synthesis of oligonucleotides with carboxylic residues attached via a prolinol linker.
Chem. Eur. J. 2009, 15, 3390 – 3402
ꢂ 2009 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.chemeurj.org
3391

C. Richert and M. Printz
Chain assembly involved “reversed” DNA synthesis, using
5’-phosphoramidites, and started from controlled pore glass
(cpg) 6 (Scheme 2). Prolinolphosphoramidite 2 was coupled
to support 7 featuring the unmodified DNA core, yielding 8,
which was deprotected to 9. The Alloc group was removed
under Pd0 catalysis to give 10. Best yields were obtained
under microwave irradiation, using octane thiol as the scav-
enger, as determined by MALDI-TOF mass spectrometry of
analytical samples obtained by treatment with aqueous am-
monia.
After coupling the secondary amino group of 10 to car-
boxylic acids and deprotection with concomitant release
from the support, oligonucleotides of general structure 1P
were obtained. A side product with a 3’-terminal phosphate
group was also detected.[22] Oligonucleotides with the resi-
dues of 2-anthraquinoyl carboxylic acid, 1-pyrenyl carboxyl-
ic acid, cholic acid, and 3-perylenylbutyric acid were thus
prepared. A numbering scheme was employed for the oligo-
nucleotides, in which the number denotes the DNA se-
quence, the letter immediately following this number indi-
cates what linker is being used, and the last two letters are a
short-hand for the stacking moiety (e.g., py for pyrene).
Among the oligonucleotides, the one forming an anthraqui-
none-bearing duplex (1P_aq)2 showed the highest melting
point, with a DTm of +8.38C per residue at 1m NaCl, com-
pared to control duplex (1)2 (Table 1).
Table 1. UV-melting points [8C] and hyperchromicities accompanying
duplex dissociation for self-complementary DNA hexamers of the se-
quence 5’-TGCGCA-3’ with or without a 3’-terminal substituent.[a]
compound
Tm at [NaCl]/8C
150 mm
Hyperchr./%[b]
none
1m
TGCGCA (1)
TGCGCAA
1P_pr
1P_aq
1P_py
1P_ch
1U_py
1N_pr
1R_py
1N_na
1N_bp
1N_py
1N_an
23.2
19.9
26.5
31.7
27.8
28.5
n.d.[c]
42.2
29.7
n.d.[d]
19.2
42.6
n.d.[d]
19.3
18.0
33.0
33.8
33.3
45.1
45.8
42.5
42.1
16.9
56.0
44.4
49.4
34.9
58.7
65.4
32.7
30.7
47.9
34.3
37.3
47.4
50.9
46.6
46.2
26.9
62.3
47.3
52.1
39.0
63.4
63.1
37.0
34.1
51.8
9ꢀ2
11ꢀ1
16ꢀ3
10ꢀ2
14ꢀ1
13ꢀ1
18ꢀ4
6ꢀ1
12ꢀ2
4ꢀ0.5
4ꢀ1
5ꢀ1
2ꢀ0.5
11ꢀ2
12ꢀ2
10ꢀ4
1N_dp
1N_ad
1A*_pr[e]
[a] Conditions: 4 mm strand concentration, 10 mm sodium phosphate
buffer pH 7.0, and NaCl concentration given. The given Tm values are
averages of two heating and two cooling curves. Standard deviation be-
tween ꢀ0.2 and 1.58C for all values reported. See Table S1 (Supporting
Information) for a full listing of experimental errors. [b] Hyperchromicity
between low and high temperature baseline at 260 nm at 1m NaCl con-
centration. [c] Melting point below 158C. [d] No melting point, not suffi-
ciently cooperative transition. [e] Structure shown in Figure 2.
Scheme 3. Synthesis of a phosphoramidite bearing a pyrenyl residue.
Conditions: a) toluene/CH3CN, D, b) THF, TBAF, 23% over 2 steps; c)
DMTrCl, DMAP, DMF, pyridine 68%; d) ClP[N
MeCN, DIEA.
ACHUTNGTREN(NUG iPr)2]OACHTUGNTRNE(NUGN CH2)2CN,
Scheme 4. Synthesis of oligonucleotides with dangling deoxyadenosine residues featuring a perylenyl or pyrenyl residue.
3392
www.chemeurj.org
ꢂ 2009 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Chem. Eur. J. 2009, 15, 3390 – 3402

Universal DNA Hybridization Probes
FULL PAPER
Next, oligonucleotides with the stacking moiety linked to
N6 of a dangling 2’-deoxyadenosine residue were prepared.
Initially, a urea linkage was used. One motivation for syn-
thesizing pyrenylcarbamoyl derivative 11 was a literature
report on a related naphthalene derivative[23] (compound 12,
shown in Figure S1, Supporting Information), which gives a
DTm of +9.78C per residue (1m NaCl) when employed as a
dangling residue on short DNA duplexes. Pyrene isocyanate
13[24] was reacted with protected deoxyadenosine 14 to give
15. Desilylation to 16 was followed by introduction of a
DMTr group, generating 17, which led to phosphoramidite
11 (Scheme 3).
preparation of 22 involved the separation of the desired
isomer of perylene carbaldehyde from the mixture formed
under established reaction conditions.[29,30] The duplex of oli-
gonucleotide 1N_pr, prepared with phosphoramidite 25,
melts 288C higher than (1)2 at 1m NaCl. This massive in-
crease in melting point is identical to that induced by the
current lead structure for non-universal 3’-caps that was
identified in
a
recent combinatorial study[31] (compare
Figure 2). This is a perylene ring system linked to the 2’-po-
sition of a fixed, non-dangling 3’-terminal deoxyadenosine
residue that engages in base pairing with the target strand.
Like pyrene-C-nucleoside 18[25] (Figure S1, Supporting In-
formation), the 3’-phosphoramidites of pyrenylcarbamoyl
adenosine 11 and perylenylmethylamine adenosine 23 were
incorporated at the 3’-terminus of the oligonucleotide 5’-
TGCGCA-3’ via DNA syntheses on commercial solid sup-
port 19 (trade name “Universal Support II”), as shown in
Scheme 4.[26] The duplex formed by oligonucleotide 1U_py,
prepared from 11, melted 7.48C lower than the control
duplex at 1m salt. We reasoned that the urea linker between
adenine and pyrene is too rigid to allow for stacking of the
latter on the terminal base pair.
Accordingly, oligonucleotides with a more flexible meth-
ylene linker were prepared. A synthesis for N6-pyrenyl-
methyl-2’-deoxyadenosine (20, Figure S1, Supporting Infor-
mation) was known.[27] The synthesis of the corresponding
perylene derivative is shown in Scheme 5. It starts from 2’-
deoxyinosine (21).
Figure 2. Structure of oligonucleotide with 2’-perylenyl-substituted deoxy-
adenosine as dangling residue.[31]
Pyrene-C-nucleoside 18 (Figure S1, Supporting informa-
tion), known from earlier work on duplex-stabilizing dan-
gling residues, was tested as a benchmark cap.[25] To prepare
it, pyrene was glycosylated under Friedel–Crafts condi-
tions.[32] When linked to the 3’-terminus of the hexamer, the
C-nucleosides induced no more than 138C melting point in-
crease or 6.58C per residue (1R_py, 1m NaCl, Table 1). This
confirmed that the PAH-bearing residue in 1N_pr provides
an unusually strong duplex-stabilizing effect.
We synthesized additional oligonucleotides with 3’-dan-
gling, N6-substituted deoxyadenosine residues. To facilitate
combinatorial work, a solid-support bound version of the lit-
erature-known inosine building block 26[33] (Scheme 6) was
developed. First, the methodology was tested on 27 in solu-
tion (Scheme 5 and Scheme S1, Supporting Information).[33]
Oxybenzotriazolide 27 was treated with 4-hydroxystilbene,
yielding the expected 6-(aryloxy)purine (28) in 99% yield.
This encouraged us to succinylate 26 (Scheme 6) to 29, and
to couple it to sarcosine-bearing controlled pore glass 30,[34]
to obtain support 31. The sarcosine linker prevents loss of
oligonucleotide chains from the support through intramolec-
ular attack of the deprotonated linker amide on the 3’-ester
functionality under the basic conditions required for the on-
support nucleophilic displacement. An oxalyl linker (Figure
S2, Supporting Information)[35] proved too labile towards
these conditions. Support 31 was treated with 1-(aminome-
thyl)pyrene (Py), 2-(hydroxymethyl)anthraquinone, and 4-
hydroxystilbene in the presence of DIEA, yielding supports
of general structure 32. After chain assembly and deprotec-
Scheme 5. I) Synthesis of phosphoramidite 25 bearing a perylenyl residue.
Reaction conditions: a) DMTrCl, DMAP, pyridine 28% over 2 steps; b)
ClP[N
ACHTUNGTRENNUNG(iPr)2]OACHTUNGTRENNUNG(CH2)2CN, MeCN, DIEA. II) Structure of inosine deriva-
tive 27.[33]
Deoxyinosine 21 was treated with BOP and perylenylme-
thylamine 22,[28] and the resulting nucleoside 23 was protect-
ed to give 24, which was phosphitylated to yield 25. The
Chem. Eur. J. 2009, 15, 3390 – 3402
ꢂ 2009 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.chemeurj.org
3393

C. Richert and M. Printz
tion, the desired N6-substituted product was obtained in
high yield for pyrenyl case (1N_py), but the crude products
of the other two reactions contained oligonucleotide 5’-
TGCGCAA-3’ only, suggesting that the alkoxy substituents
do not survive the final deprotection step with aqueous am-
monia.
As mentioned above, the pyrenylmethyladenine-bearing
duplex (1N_py)2 gave the highest melting point at high salt
concentration (compare Figure 3), 18C higher than that of
the perylenylmethyl counterpart (1N_pr)2. Anthracene de-
rivative (1N_an)2 gave an even higher melting point at
150 mm salt (31.68C higher than that of the control hexa-
mer), but the broad melting transition for this duplex at low
salt concentration, combined with the unusually small hyper-
chromicity value, kept us from pursuing the N_an cap fur-
ther. Since absorbances from the PAH moieties overlap with
the absorbance maximum of the nucleobases at 260 nm,
wavelength scans were acquired for solutions of 1N_pr,
1N_py, 1N_ad from 10 to 908C, followed by UV-melting
curves at 233, 242, 256, or 278 nm to obtain melting transi-
This was confirmed in a model reaction in solution, in-
volving 6-(benzyloxy)purine derivative 33, treated with
aqueous ammonia/acetonitrile. After 3 h, half of the benzyl
ether was aminolyzed to the corresponding adenine 34
(Scheme S3, Supporting Information). Insufficient stability
was also observed for substitution products generated with
aromatic amines (Figure S3, Supporting Information). Ali-
phatic amines gave stable oligonucleotides, though, allowing
for the synthesis of compounds 1N_na–1N_ad (Scheme 6).
The amine starting materials and support-bound intermedi-
ate 32 are shown in Figure S4 (Supporting Information).
Table 1 lists the results of UV-melting curve experiments
with the newly prepared oligonucleotides. This includes data
on a duplex with an unmodified dangling deoxyadenosine,
which induces a melting point increase of no more than
+1.58C per residue at 1m NaCl as a second control.
A dangling deoxyadenosine residue with a 2’-linked pery-
lenylbutyramide residue (1A*_pr, Figure 2) also gives no
more than a weak duplex-stabilizing effect, with a Tm
10.58C lower than that of 1N_pr, confirming that the large
DTm for the latter is not a trivial effect, induced by any con-
struct of this size and lipophilicity.
*
Figure 3. Melting curves of control duplex (1)2
(
: recorded at 260 nm)
*
and (1N_py)2
(
: recorded at 242 nm) at 4 mM strand concentration,
10 mm phosphate buffer pH 7.0, and 1m NaCl.
Scheme 6. Solid support assisted synthesis of 1N_na, 1N_bp, 1N_py, 1N_dp, 1N_ad and 1N_an.
3394
www.chemeurj.org
ꢂ 2009 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Chem. Eur. J. 2009, 15, 3390 – 3402

Universal DNA Hybridization Probes
FULL PAPER
tions with higher hyperchromicity and reliable melting
points. It seems likely that in the case of the self-comple-
mentary hexamers, the helix is short enough to allow for hy-
drophobic interactions between the two aromatic stacking
moieties. These interactions will affect the hyperchromicity
observed at 260 nm, where both the DNA and the aromatic
substituents absorb. Again, this confirmed that the pyrenyl-
methyl substituent at N6 of the dangling deoxyadenosine
residue (1N_py) induces the
for the corresponding unmodified control duplexes. So, in
almost every case, the cap stabilizes a mismatched terminal
base pair more than the canonical base pair. This effect is in
sharp contrast to what has been observed for other duplex-
stabilizing caps, all of which also increase base-pairing fideli-
ty.[10,36,37]
Subsequently, mismatch discrimination was also measured
for the penultimate base pair. This series of experiments in-
Table 2. UV-Melting points [8C] of duplexes of DNA-hybrids 5’-AGGTTGBB-N_py (35, 35c, 35g, 35t, 36c,
36g, 36t) and unmodified oligonucleotides 5’-AGGTTGBB-3’ (40a–40t, 41c–41t) with complementary target
strands and the corresponding mismatched oligonucleotides.[a] The DTm values to the corresponding perfectly
matched duplex are given in parentheses.
highest Tm. Having identified a
lead compound, we then pro-
ceeded to studying the effect
of the N6-pyrenylmethyl-sub-
stituted adenine on base-pair-
ing selectivity. For this, UV-
melting curves were measured
using non-self-complementary
octamers featuring the pyren-
yl-dA residue at the 3’-termi-
nus (35, 35g, 35c, and 35t),
each with a different nucleo-
base at the terminus and target
strands 37a–37t, also shown in
Figure 4. The melting points of
the corresponding duplexes are
listed in Table 2. The melting
point depression caused by the
individual mismatches can also
be obtained from Table S3
Probe strand
Target strand
terminal base pair
5’-BTCAACCT-3’
B=T (37t)
B=G (37g)
B=C (37c)
B=A (37a)
5’-AGGTTGAA-N_py (35)
5’-AGGTTGAC-N_py (35c)
5’-AGGTTGAG-N_py (35g)
5’-AGGTTGAT-N_py (35t)
38.3pm
38.0 (ꢁ0.3)
35.9 (ꢁ2.4)
33.1 (ꢁ10.0)
42.5pm
34.5 (ꢁ6.0)
39.0 (+0.7)
37.5 (ꢁ5.6)
37.6 (ꢁ4.9)
40.5pm
34.8 (ꢁ8.3)
36.4 (ꢁ6.1)
37.2 (ꢁ3.3)
43.1pm
37.5 (ꢁ5.0)
37.8 (ꢁ2.7)
5’-AGGTTGAA-3’ (40a)
5’-AGGTTGAC-3’ (40c)
5’-AGGTTGAG-3’ (40g)
5’-AGGTTGAT-3’ (40t)
29.5pm
25.1 (ꢁ4.4)
25.5 (ꢁ4.0)
23.2 (ꢁ13.9)
35.8pm
24.6 (ꢁ7.4)
27.4 (ꢁ2.1)
25.7 (ꢁ11.4)
27.7 (ꢁ8.1)
32.0pm
23.4 (ꢁ13.7)
26.0 (ꢁ9.8)
25.6 (ꢁ6.4)
37.1pm
26.7 (ꢁ9.1)
26.7 (ꢁ5.3)
Penultimate base pair
5’-TBCAACCT-3’
B=T (38t)
B=G (38g)
B=C (38c)
B=A (38a)
5’-AGGTTGAA-N_py (35)
5’-AGGTTGCA-N_py (36c)
5’-AGGTTGGA-N_py (36g)
5’-AGGTTGTA-N_py (36t)
38.3pm
31.5 (ꢁ6.8)
31.2 (ꢁ7.1)
29.4 (ꢁ20.4)
43.3pm
31.1 (ꢁ13.1)
30.6 (ꢁ7.7)
30.6 (ꢁ19.2)
32.7 (ꢁ10.6)
44.2pm
30.8 (ꢁ19.0)
32.6 (ꢁ10.7)
34.2 (ꢁ10.0)
49.8pm
31.9 (ꢁ11.4)
36.9 (ꢁ7.3)
5’-AGGTTGAA-3’ (40a)
5’-AGGTTGCA-3’ (41c)
5’-AGGTTGGA-3’ (41g)
5’-AGGTTGTA-3’ (41t)
29.4pm
19.3 (ꢁ10.1)
18.0 (ꢁ11.4)
17.6 (ꢁ22.6)
34.3pm
18.4 (ꢁ12.8)
19.9 (ꢁ9.5)
20.6 (ꢁ19.6)
26.0 (ꢁ8.3)
31.2pm
(Supporting
Information),
18.4 (ꢁ21.8)
20.1 (ꢁ14.2)
22.8 (ꢁ8.4)
40.2pm
which compiles the same data
in the form of DTm values.
21.5 (ꢁ12.8)
25.5 (ꢁ5.7)
The oligonucleotides with
the N_py cap exhibited a melt-
ing point-increasing effect for
any of the perfectly matched
Watson–Crick duplexes. But
the drop in melting point for
duplexes with a mismatched
base pair was lower than that
Third base pair
5’-GTBAACCT-3’
B=T (39t)
B=G (39g)
B=C (39c)
B=A (39a)
5’-AGGTTGAC-N_py (35c)
5’-AGGTTGAC-3’ (40c)
26.0 (ꢁ17.1)
15.2 (ꢁ29.1)
20.6 (ꢁ22.5)
15.5 (ꢁ21.6)
43.1pm
37.1pm
21.3 (ꢁ21.8)
10.3 (ꢁ26.8)
[a] Conditions: 3.5 mm concentration of each strand, 10 mm sodium phosphate buffer pH 7.0, 1m NaCl; Tm
values given are averages of two heating and two cooling curves. Standard deviation is between ꢀ0.2 and
0.78C. See Table S2 (Supporting Information) for a full listing of melting points and experimental errors;
pm=perfectly matched duplex.
Figure 4. Oligonucleotides employed for testing mismatch discrimination.
Chem. Eur. J. 2009, 15, 3390 – 3402
ꢂ 2009 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.chemeurj.org
3395

C. Richert and M. Printz
volved cap-bearing oligonucleotides 35 or 36c–36t, unmodi-
fied oligonucleotides 40a, 41c–41t and target strands 38t–
38a, resulting in a total of 2ꢃ16=32 duplexes. Again, a sig-
nificant effect of the cap on duplex stability was measurable.
And again, the drop in melting point for a single mismatch
was lower in the presence of the pyrenyl cap than in the
control case (Table 2 and Figure 5). In some instances, the
and the effect of polycyclic aromatic hydrocarbons on mo-
lecular recognition in particular. One might ask: what are
the factors that make a ligand modulate the affinity and
base-pairing selectivity of an oligonucleotide forming a
duplex with a target strand? Three possible factors leading
to such observations are: a) changes in rigidity (reduced or
increased entropic penalty for formation of a duplex), b)
more extensive molecular contacts, leading to additional
bridging interactions (van der Waals interactions, dipole–
dipole interactions, and hydrogen bonds), and c) structural
features that prevent the formation of alternative structures,
such as the presence of negative charges in the phosphate
backbone of the DNA preventing backfolding.[43]
Based on general principles of molecular recognition, the
fidelity-decreasing effect of the N_py cap is counterintuitive.
Fundamentally, factors that lead to high-affinity binding,
such as shape complementarity and charge complementarity
(including dipole–dipole interactions) should also be the
very factors that lead to high specificity for a target. This ar-
gument is often cited in the context of drug molecules inter-
acting with receptors or active sites of enzymes, and is used
as the basis for setting up screens.[44] Applied to the problem
at hand, a covalently appended ligand should stabilize the
correct or an individual mispaired base pair, to which shape
it is complementary to. The ligand should not have a general
fidelity-reducing effect on neighboring or remote positions.
So, affinity “without” selectivity is quite unexpected for a
nucleotide-based molecular cap, particularly one that is
fairly rigid, consisting of three ring systems (deoxyribose,
purine and pyrene), with few rotatable bonds inbetween. A
rigid, planar aromatic system like pyrene should be an ideal
stacking partner for neighboring base pairs. In fact, pyrenyl-
C-nucleosides have been shown to replace base pairs in a
helix well,[45] and other pyrenyl or perylenyl caps can in-
crease base-pairing fidelity at the terminus.[31,37]
What, then, may be the structural basis of the effect ob-
served? In the present case, the binding surface offered by
the pyrenyl ring may be rigid but too general, accommodat-
ing many different base combinations. The melting curve
data suggest that base pairs can be formed in the pyrenyl-
modified duplexes from the unmodified terminus down to at
least the third position from the terminus. At the third posi-
tion from the terminus, the decrease in base-pairing selectiv-
ity in the presence of the pyrenyl group is less pronounced
than at the terminal and penultimate position. At the latter
two positions, the energetic penalty for violating the
Watson–Crick rules is smaller in the presence of the N_py
cap than in the control case, suggesting that this is where
the pyrene interacts with the DNA. The precise position of
the pyrene residue cannot readily be determined from melt-
ing-point data alone, though. The cap seems to be a more
general, less base specific interaction partner, offering some
interaction mode to any combination of neighboring nucleo-
bases. Based on the depth of the effect, we suspect that the
pyrenyl residue intercalates somewhere between the penulti-
mate and the third base pair. We note that in the case of ad-
ducts between polycyclic aromatic hydrocarbons and duplex
Figure 5. Selected melting curves acquired to determine mismatch dis-
crimination at the terminus, involving duplexes of 5’-AGGTTGAA-N_py
(37; triangles) or 5’-AGGTTGAA-3’ (39a; circles) with perfect match
targets (PM, filled symbols), or singly C-mismatched targets (C-MM,
empty symbols). Conditions: 3.5 mM strand concentration each, 10 mm
phosphate buffer pH 7.0, and 1m NaCl.
melting point depression induced by a mismatch was more
than 48C smaller in the cap-bearing duplex than in the con-
trol duplex.
To test how far the effect of the N_Py cap reaches, a first
set of UV-melting points was measured for duplexes featur-
ing a mismatched base at the third position from the termi-
nus. This experiment involved duplexes between oligonucle-
otides presenting a G at the third position of the probe
strands. Interestingly, the melting point drop for the mis-
match-containing duplexes over the control was between 27
and 228C for the unmodified DNA. These are very high
values, similar to melting point drops observed for com-
plexes of circular DNA with mismatches at the 5th position
(ꢁ22.58C).[38] The melting point drop for the case of the oli-
gonucleotide with the pyrenyl substituent 35c (23 to 178C)
was again lower in two out of three cases.
Discussion
Conventional hybridization probes should bind their target
strands with increased affinity and increased base-pairing fi-
delity.[43] The N_py cap presented here induces increased
target affinity but reduces base-pairing fidelity at the termi-
nus. This may make the N_py substituent interesting for the
construction of “universal” DNA probes. Oligonucleotides
with reduced sequence bias have been proposed for the de-
tection of DNA of any sequence.[39–42] and as universal pri-
mers for degenerate PCR. Unlike high fidelity probes, uni-
versal DNA binders are meant to recognize different nu-
cleotide sequences with similar affinity.
The current results may also be discussed in the context
of the effect of ligands on base pairing in DNA in general,
3396
www.chemeurj.org
ꢂ 2009 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Chem. Eur. J. 2009, 15, 3390 – 3402

Universal DNA Hybridization Probes
FULL PAPER
DNA, intercalation usually occurs at the site of modifica-
tion.[46] For dangling residues, the situation is less clear. In
the present case, the melting point data suggests a more
remote intercalation, probably because the dangling nucleo-
tide acts as a linker and because intercalation is more likely
to occur in the interior, where fraying is reduced.
Our structural proposal was corroborated by a modeled
three-dimensional structure shown in Figure 6. This force
field-minimized structure shows that it is possible for the
pyrenyl moiety to intercalate between the penultimate and
the antepenultimate base pair, while the deoxyadenosine
residue serves as a linker, located largely in the minor
groove. In such a case, increased structural flexibility may
cause reduced base-pairing fidelity, as suggested in Figure 7.
Evidence for increased flexibility in a duplex with an inter-
Figure 7. Proposed “backfolding” of the pyrene-bearing dangling residue
attached at the terminus of one of the oligonucleotides of a duplex.
pyrenylmethyl-2’-deoxyadenosine residue showed the largest
increase in the UV-melting point. This increase is one of the
largest on record, with a DTm of up to 288C for a self-com-
plementary sequence. The melting point increase is most
likely caused by bridging interactions that the pyrene ring
system engages in. These seem to reduce base-pairing fideli-
ty in the molecular environment of the pyrenylmethylade-
1
calated pyrene comes from broad H NMR signals for one
of the neighboring base pairs of a pyrene-C-nucleoside in
the interior of a duplex.[49] Modeling studies on DNA du-
plexes with pyrenyl-C-nucleosides facing abasic sites come
to similar conclusions.[50] Still, the model presented in Fig-
ures 6 and 7 is no more than a working hypothesis, and we
cannot exclude that the pyrenyl moiety is located in one of
the grooves, even though it has no discernable shape com-
plementarity to either of the them. A flexibility-inducing
effect of a polycyclic aromatic hydrocarbon, such as pyrene,
is interesting, given that polymerases accept the triphos-
phates of pyrene-C-nucleosides as substrates.[51–53]
nine residue.
A decrease in DDTm between perfectly
matched and mismatch-containing duplexes that is modest
in its magnitude (a few degrees C in Tm), but highly signifi-
cant in terms of the number of mismatched base pairs toler-
ated, was observed. Intercalation is the most likely mode of
interaction of the pyrene ring. This is a general form of in-
teraction between polycyclic aromatic hydrocarbons and
DNA, begging the question whether the effect on base-pair-
ing fidelity observed is a general one for polycyclic aromatic
hydrocarbons. If this is the case, it might contribute to the
toxicity of these compounds. The affinity-enhancing effect
of the N_py cap may become useful for generating universal
DNA binding probes.
Experimental Section
General: Reagents and building blocks were from Aldrich/Fluka/Sigma
(Deisenhofen, Germany), or Acros (Geel, Belgium). Anhydrous solvents
were purchased over molecular sieves and were used without further pu-
rification. Unmodified oligonucleotides (37–41) were purchased from
Biomers (Ulm, Germany). Reagents for standard DNA synthesis were
purchased from Proligo (Hamburg, Germany). The 5’-phosphoramidites
of unmodified nucleosides used for “reversed” DNA synthesis (dABz
dCBz, dGiBu, dT) were obtained from ChemGenes (Wilmington, MA).
MALDI-TOF mass spectra were acquired on Bruker Daltonics
,
a
Figure 6. Force field-minimized, modeled structure of DNA duplex with
N_py cap intercalating from the minor groove. The pyrene ring stacks be-
tween penultimate and third base pair of the duplex (5’-CGCG-N_py:5’-
CGCG-3’). The structure was generated by using Macromodel[47] (version
maestro-v7.5.106), starting from the conformation of the terminus in the
three-dimensional structure of (5’-ACGCGU-NA)2 [NA=residue of nali-
dixic acid linked to the 2’ position of the 3’-terminal residue].[8] The
graphic was produced with VMD.[48] The chemical structure of the DNA
is shown in Figure S1; the coordinates of the modeled structure are given
in Table S1 (both Supporting Information).
REFLEX IV or BIFLEX IV spectrometer in linear negative mode. The
oligonucleotides were analyzed using a matrix mixture made up from
2,4,6-trihydroxyacetophenone (THAP, 0.3m in ethanol) and diammonium
citrate (0.1m in water) (2:1, v/v). Given calculated masses are average
masses; m/z values are those found for pseudomolecular ions ([MꢁꢁH]
or [M ++H]), detected at the peak maximum. NMR spectra were record-
ed on a Bruker AC250, AM 400 or DRX 500 spectrometer. Analysis
using thin layer chromatography was performed on silica gel 60 precoat-
ed plates from Merck (0.25 mm thickness). Silica gel for flash chromatog-
raphy was 0.063–0.2 mm mesh from Merck (Darmstadt, Germany). All
DNA sequences are given from 5’- to 3’-terminus. Synthesis of b-1’,2’-
dideoxy-1’-(1-pyrenyl)-ribose and b-1’,2’-dideoxy-5’-O-(4,4’-dimethoxy-
trityl)-1’-(1-pyrenyl)-ribose-3’-O-yl-cyanoethyl-N,N-diisopropylphosphor-
amidite (18) were synthesized according to known protocols.[25,32]
Conclusion
In the current study, oligonucleotides with a dangling 3’-ter-
minal residue were tested. Among them, those with an N6-
DNA Synthesis: The unmodified portions of oligodeoxynucleotides were
synthesized with a Perseptive Biosystems 8909 Expedite DNA synthesiz-
Chem. Eur. J. 2009, 15, 3390 – 3402
ꢂ 2009 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.chemeurj.org
3397

C. Richert and M. Printz
er, using the protocol provided by the manufacturer. The employed
DNA synthesis columns were purchased from Prime Synthesis (Aston,
PA).
added. This solution was then added to deprotected “universal support
II” 19 (40 mg, 2 mmol) and treated for 2 h at RT on the shaker. After-
wards the supernatant was removed and the support was washed with
MeCN (3ꢃ800 mL) and oxidizer for DNA synthesis (iodine 1m in pyri-
dine/THF/water 21:77:2 v/v, 800 mL) was added. After 20 min, the support
was washed with DMF, MeCN, and CH2Cl2 until the supernatant was col-
orless and dried in vacuo.
HPLC Purification: HPLC purification of modified oligonucleotides was
performed using a Merck-Hitachi L-6200 system with a photodiode array
detector on a 250 mmꢃ4.6 mm Nucleosil C4 column (Macherey-Nagel,
Dꢀren, Germany). The employed gradient was CH3CN (solvent B) in
0.1m triethylammonium acetate, pH 7, with detection at 260 nm. Deter-
mination of yields of oligonucleotides are based on the intensity of the
product peak in the HPLC trace of the crude or based on the amount of
isolated pure product in comparison to the 1 mmol synthesis scale.
N6-(N’-Pyren-1-ylcarbamoyl)-2’-deoxyadenosine-loaded cpg: Phosphora-
midite 11 (87 mg, 110 mmol, 1 equiv) was dried in a 5 mL vial at 0.1 Torr
for 2 h. An solution of dry MeCN (1 mL), DIEA (330 mmol, 55 mL,
3 equiv) and 2-cyanoethyl-N,N-diisopropyl chlorophosphoramidite
(165 mmol, 40 mg, 1.5 equiv), containing molecular sieves (3) ꢄ), was
added. Carbamate 11 dissolved during the following reaction. After TLC
control (Rf = 0.18, CH2Cl2/MeOH 99:1 +1% NEt3) indicated full con-
version, the solution was diluted with CH2Cl2 (15 mL), washed with sat.
NaHCO3 (10 mL), dried over Na2SO4, and the solvent was evaporated in
vacuo. To the obtained crude product tetrazole solution (220 mmol,
488 mL, 2 equiv, 0.45m in MeCN) and dry MeCN (620 mL) were added.
This solution was then added to deprotected “universal support II” 19
(40 mg, 2 mmol) and treated for 2 h at RT on a shaker. Afterwards the su-
pernatant was removed and the support was washed with MeCN (3ꢃ
800 mL) and oxidizer for DNA synthesis (iodine 1m in pyridine/THF/
water 21:77:2 v/v, 800 mL) was added. After 20 min, the support was
washed with DMF, MeCN, and CH2Cl2 until the supernatant was color-
less and dried in vacuo.
UV-Melting experiments: Acquisition of UV-melting curves was per-
formed on a Perkin–Elmer Lambda 750 spectrophotometer at 260 nm
and 1 cm path length at heating or cooling rates of 18Cminꢁ1. The strand
concentration in case of the measurements with the self-complementary
sequences was 4 mm in 10 mm sodium phosphate buffer pH 7. UV-melting
curves were acquired without NaCl, with 150 mm and 1m NaCl. The melt-
ing curve measurements for mismatch discrimination were performed
using a DNA concentration of 3.5 mm (each strand) in sodium phosphate
buffer (10 mm) and 1m NaCl. UV Winlab 2.0 (Perkin Elmer) was used to
determine melting temperatures. They are averages of the maxima of the
first derivative of the 95-point smoothed curves from heating and cooling
experiments. Calculation of the hyperchromicity was performed using the
difference in absorption between high- and low-temperature baseline di-
vided by the absorption at the low-temperature baseline.
Sarcosine-loaded cpg: To a mixture of standard lcaa-cpg (75 mmolgꢁ1
,
Coupling of hydroxyprolinol linker and its acylation: Immobilized oligo-
nucleotide 7 (5’-TGCGCA-3’, 2 mmol, 60 mg) and hydroxyprolinol phos-
phoramidite 2 (50 mmol, 35 mg, 25 equiv) were dried at 0.1 Torr for
30 min in a 1.5 mL polypropylene vessel. Then dry acetonitrile (250 mL)
and 4,5-dicyanoimidazole (DCI) activator solution for DNA synthesis
(0.25m in acetonitrile, 250 mL) were added. After shaking for 1 h the sup-
port was washed with acetonitrile (3ꢃ800 mL). Then the support was
treated with oxidizer solution for DNA synthesis (iodine 1m in pyridine/
THF/water 21:77:2 v/v, 800 mL) for 15 min on the shaker. The support
was washed with acetonitrile (each 800 mL) until the supernatant was col-
orless and was then dried at 0.1 Torr. The synthesized support 8 was
treated for 15 min with deblock solution for DNA synthesis (1% tri-
chloroacetic acid (TCA) in CH2Cl2, 800 mL) in a 1.5 mL polypropylene
vessel on a shaker. Afterwards the support was washed with deblock so-
lution (each 800 mL) until the supernatant was colorless then it was
washed with CH2Cl2 (4ꢃ800 mL). After drying at 0.1 Torr for 30 min the
support 8 was transferred into a 10 mL microwave vessel with stirring bar
400 mg, 30 mmol) Fmoc-Sar-OH (140 mg, 0.45 mmol, 15 equiv), HATU
(163 mg, 0.43 mmol, 14.25 equiv) and HOAt (61 mg, 0.45 mmol, 15 equiv)
in
a 5 mL vial dry DMF (2.5 mL) and DIEA (148 mL, 0.90 mmol,
30 equiv) were added. The vial was sealed with a glass plug and placed
on a shaker for 2 h. Then the support was washed with DMF, MeOH,
and CH2Cl2. After drying at reduced pressure the same coupling proce-
dure was repeated. Afterwards the loading was determined via the liber-
ated fluorenyl piperidine adduct. In
a 1.5 mL polypropylene vessel
loaded support (3–5 mg) was weighed in and were treated with 20% pi-
peridine in DMF (1 mL) for 1 h on the shaker. The supernatant was di-
luted 1:1 with 20% piperidine in DMF and then the absorption at
301 nm was determined. The used value for the absorption coefficient for
[54]
the liberated fluorenyl piperidine adduct was e301 =7800 Lmolꢁ1 cmꢁ1
.
The determined loading was between 34 and 38 mmolgꢁ1. Afterwards the
support (400 mg) was treated, in a sealed 5 mL vial, with capping re-
agents for standard DNA synthesis [CAP A 1.5 mL (acetic anhydride/2,6-
lutidine/THF 1:1:8), CAP B 1.5 mL (DMAP in THF, 6.5% w/v)] on a
shaker for 20 min. Then the support was washed with DMF, MeOH, and
CH2Cl2 and dried at 0.1 Torr for 30 min. Then the support (400 mg) was
treated with 20% piperidine in DMF (3 mL) for 30 min on a shaker in a
sealed 5 mL vial. Then the support was washed with 20% piperidine in
DMF, DMF, MeOH, and CH2Cl2 and dried in vacuo for 30 min.
and PPh3 (4.5 mg, 17 mmol, 8.5 equiv), [H2NEt2]
16 mmol, 8 equiv) and [Pd(PPh3)4] (22 mg, 19 mmol, 9.5 equiv) were
+A[HCO3]ꢁ (22 mg,
CHTUNGTRENNUNG
AHCTUNGTRENNUNG
added. Then dry DMF (3 mL) and octanethiol (240 mL) were added. The
slurry was stirred and heated in the microwave (40 min, 250 W, 808C,
2.5 bar). Then the support was washed with DMF, acetonitrile, and
CH2Cl2 followed by drying at 0.1 Torr for 30 min. Afterwards the same
procedure was carried out a second time. Then the corresponding carbox-
ylic acid (2-anthraquinoyl carboxylic acid, cholic acid, 3-perylenyl butyric
acid, 1-pyrenyl carboxylic acid) (200 mmol, 1 equiv), HATU (0.95 equiv,
190 mmol, 72 mg), and HOAt (1 equiv, 200 mmol, 27 mg) were dried at
0.1 Torr for 20 min in a 2.0 mL polypropylene vessel. Then dry DMF
(1 mL) and DIEA (2.3 equiv, 460 mmol, 76 mL) were added. After 15 min
at RT, the solution was added to the amino terminal solid support 10
(8 mg, 0.3 mmol) in a 2.0 mL polypropylene vessel. The vessel was placed
for 1 h in a water bath at 508C and was agitated casually. Then the sup-
port was washed several times with DMF, acetonitrile, and CH2Cl2.
O6-(Benzotriazol-1-yl)-2’-deoxyinosine-loaded cpg: Sarcosine modified
support 30 (1.5 g, 105 mmol) was placed in a 25 mL vial under argon.
Then a solution of succinylated inosine derivative 29 (388 mg, 0.5 mmol,
1 equiv), HATU (182 mg, 0.48 mmol, 0.95 equiv), HOAt (67 mg,
0.50 mmol, 1 equiv) and DIEA (164 mL, 1 mmol, 2 equiv) in dry DMF
(8 mL) was added via syringe. The vial was sealed with a glass plug and
was placed on a shaker for 3 h. Afterwards the support was washed with
DMF and CH2Cl2 and was then dried at 0.1 Torr for 1 h. The support was
then stored under an argon atmosphere. The loading was determined as
follows: solid support 31 (3–6 mg) was weighed into a 1.5 mL polypropy-
lene vessel and was treated with deblock solution for DNA synthesis
(1% TCA in CH2Cl2, 800 mL) on a shaker for 20 min. Then the superna-
tant was transferred into a 5 mL vial and was diluted with deblock solu-
tion. The absorption was measured at 505 nm. The value of the absorp-
N6-Perylenylmethyl-2’-deoxyadenosine-loaded cpg: Phosphoramidite 25
(25 mg, 30 mmol, 1 equiv) was dried in a 5 mL vial at 0.1 Torr for 2 h. A
solution of dry MeCN (275 mL), DIEA (15 mL, 90 mmol, 3 equiv) and 2-
cyanoethyl-N,N-diisopropyl chlorophosphoramidite (11 mg, 45 mmol,
1.5 equiv), containing some molecular sieves (3) ꢄ), was added. After
TLC control (Rf = 0.2, CH2Cl2/MeOH 99:1 +1% NEt3) indicated full
conversion, the solution was diluted with CH2Cl2 (15 mL), washed with
sat. NaHCO3 (10 mL), dried over Na2SO4 and the solvent was evaporated
under reduced pressure. To the obtained crude product tetrazole solution
(60 mmol, 133 mL, 2 equiv, 0.45m in MeCN) and dry MeCN (170 mL) were
tion coefficient at 505 nm was taken out of the literature
[55]
78000 Lmolꢁ1 cmꢁ1
.
The determined loading was 32 mmolgꢁ1
.
N6-Arylmethyl-2’-deoxyadenosine-loaded cpg: Inosine-modified support
33 (32 mg, 1 mmol) was placed in a 1.5 mL polypropylene vessel and then
treated with a mixture of the corresponding amine Na–Ad (250 mmol),
dry DMF (500 mL), and DBU (57 mg, 56 mL, 375 mmol). Solid amines
3398
www.chemeurj.org
ꢂ 2009 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Chem. Eur. J. 2009, 15, 3390 – 3402

Universal DNA Hybridization Probes
FULL PAPER
were dried together with the solid support at 0.1 Torr before the reaction
was started. Liquid amines were transferred into the reaction vessel
under a gentle stream of argon. Then the reaction vessel was sealed and
placed in a water bath at 508C and agitated occasionally. After 2 h the
supernatant was removed, the support was washed several times with
DMF, MeCN and CH2Cl2 and was dried in vacuo.
J=7.0 Hz, 1H, H1’), 5.38 (d, J=4.1 Hz, 1H, OH), 5.05 (t, J=5.9 Hz, 1H,
H3’), 4.46 (brs, 1H, OH), 3.93–3.89 (m, 1H, H4’), 3.69–3.50 (m, 2H,
H5’+H5’’), 2.84–2.74 (m, 1H, H2), 2.42–2.33 ppm (m, 1H, H2’); MS
(FAB): m/z: 495.2 [M ++H]; UV/Vis (CHCl3/MeOH 1:1 v/v): lmax (e)=
350 (26300), 280 (44100), 243 nm (50600 molꢁ1 dm3 cmꢁ1).
5’-O-(4,4’-Dimethoxytrityl)-N6-(N’-pyren-1-ylcarbamoyl)-2’-deoxyadeno-
sine (17): Derivative 16 (79 mg, 0.16 mmol, 1 equiv), DMTrCl (65 mg,
0.19 mmol, 1.2 equiv) and DMAP (10 mg, 82 mmol, 0.5 equiv) were dried
at 0.1 Torr for 30 min. The mixture was then dissolved in dry pyridine
(1 mL) and dry DMF (1 mL) and stirred at RT over night. After TLC in-
dicated complete conversion MeOH (0.5 mL) was added and after addi-
tional stirring for 30 min the solution was diluted with CH2Cl2 (20 mL).
Subsequently the reaction mixture was washed with aqueous citric acid
(20 mL, 5% in water) and sat. NaHCO3 (20 mL). Afterwards the organic
layer was dried over Na2SO4 and the solvent was removed under reduced
pressure. The residue was purified by column chromatography (silica gel,
CH2Cl2/MeOH 99:1 + 1% NEt3) yielding compound 17 as light yellow
solid (87 mg, 0.11 mmol, 68%). TLC (CH2Cl2/MeOH 99:1 + 1% NEt3):
Rf = 0.16; 1H NMR (250 MHz, [D6]DMSO): d=12.78 (brs, 1H, NH),
10.64 (brs, 1H, NH), 8.88 (s, 1H, H8), 8.85 (d, J=8.6 Hz, 1H, pyrene),
8.72 (s, 1H, H2), 8.53 (d, J=9.5 Hz, 1H, pyrene), 8.44–8.35 (m, 4H,
pyrene), 8.25–8.12 (m, 3H, pyrene), 7.41–7.24 (m, 9H, DMTr), 6.89–6.81
(m, 4H, DMTr), 6.59 (t, J=6.2 Hz, 1H, H1’), 5.51 (d, J=4.6 Hz, 1H,
OH), 4.62 (m, 1H, H3’), 4.12 (m, 1H, H4’), 3.77 (s, 3H, DMTr), 3.75 (s,
3H, DMTr), 3.27 (m, 2H, H5’+H5’’), 3.06 ppm (m, 1H, H2’); MS
(FAB): m/z: 796.2 [M ++H]; UV/Vis (CHCl3/MeOH 1:1 v/v): lmax (e)=
350 (27300), 280 (54300), 242 nm (69000 molꢁ1 dm3 cmꢁ1).
(R)-N-Allyloxycarbonyl-4-(4,4’-dimethoxytrityloxy)-l-prolinol (5): (R)-N-
Allyloxycarbonyl-4-(4,4’-dimethoxytrityloxy)-l-methyl prolinate (4)[21]
(7.57 g, 14.2 mmol) was dissolved in dry THF (55 mL). At ꢁ58C, solid
LiBH4 (465 mg, 21.4 mmol, 1.5 equiv) was added and the solution was
stirred at RT. After 2 h, the reaction mixture was diluted with CH2Cl2
(300 mL) and washed with brine (400 mL) and sat. NaHCO3 (400 mL).
The organic layer was dried over Na2SO4 and the solvent was removed
under reduced pressure. The crude product was purified by column chro-
matography (silica gel, hexanes/ethyl acetate 3:2 + 2% NEt3) yielding
compound 5 (6.59 g, 13.1 mmol, 92%) as a white foam. TLC (hexanes/
ethyl acetate 3:2 + 2% NEt3): Rf = 0.16; [a]2D0 = +10.5 (c=1 in chloro-
form); 1H NMR (500 MHz, CDCl3): d=7.47 (d, J=7.6 Hz, 2H, DMT),
7.34–7.21 (m, 7H, DMT), 6.83 (d, J=8.8 Hz, 4H, DMT), 5.89 (m, 1H,
CH-Alloc), 5.25–5.19 (m, 2H, CH2-Alloc), 4.59–4.51 (m, 2H, CH2-Alloc),
4.26 (m, 1H, OH), 4.16 (m, 2H, H2’+H4’), 3.79 (s, 6H, DMT), 3.57 (m,
1H, H1’), 3.43 (m, 1H, H1’’), 3.10–3.02 (m, 2H, H5’+H5’’), 1.77 (m, 1H,
H3’), 1.34 ppm (m, 1H, H3’’); 13C NMR (126 MHz, CDCl3): d=158.6,
156.9, 145.3, 136.4, 132.6, 130.0, 128.1, 127.9, 126.9, 117.1, 113.2, 86.6,
71.4, 66.8, 66.0, 59.5, 55.2, 53.2, 35.7 ppm; MS (FAB): m/z: 504.2 [M +
+H].
(R)-N-Allyloxycarbonyl-4-(4,4’-dimethoxytrityloxy)-l-prolinol-1’-O-yl-cy-
anoethyl-N,N-diisopropylphosphoramidite (2): Hydroxyprolinol deriva-
tive 5 (1.00 g, 2.0 mmol) was dissolved in dry MeCN (20 mL) and DIEA
(1313 mL, 7.9 mmol, 4 equiv) and a few grains of molecular sieves (3) ꢄ)
were added. Then the solution was stirred at RT for 30 min. Then 2-cya-
Perylen-3-carbaldehyde: Perylen-3-carbaldehyde was synthesized accord-
ing to the procedure described by Asseline et al.[30] Only one isomer was
isolated by column chromatography and was exclusively used for further
syntheses. Rf = 0.97 (CH2Cl2). Spectroscopic data (1H and 13C NMR) dif-
fered from those given by Asseline et al.[30] and Dale et al.[29] and are pro-
vided below. Yield: 50% pure regioisomer; 1H NMR (400 MHz,
[D6]DMSO): d=10.29 (s, 1H), 9.06 (d, J=9.1 Hz, 1H), 8.51 (d, J=7.8,
2H), 8.47 (d, J=7.6 Hz, 1H), 8.44 (d, J=7.6 Hz, 1H), 8.10 (d, J=7.9 Hz,
1H), 7.94 (d, J=8.0 Hz, 1H), 7.87 (d, J=8.0 Hz, 1H), 7.72 (dd, J=7.7,
8.4 Hz, 1H), 7.60 ppm (q, J=7.9 Hz, 2H); 13C NMR (100 MHz,
[D6]DMSO): d=193.3, 137.1, 136.6, 133.8, 131.4, 130.6, 130.0, 129.7,
129.4, 129.3, 129.0, 128.6, 127.9, 127.2, 127.1, 126.9, 123.8, 123.4, 122.0,
121.3, 119.8 ppm; MS (FAB): m/z: 280.3 [M ++H]; UV/Vis (CHCl3/
noethyl-N,N-diisopropylchlorophosphoramidite
(939 mg,
4.0 mmol,
2 equiv) was added and the reaction mixture was stirred for 30 min at
RT. Then the solution was diluted with CH2Cl2 (200 mL) and washed
with sat. NaHCO3 (200 mL). The organic layer was dried over Na2SO4
and the solvent was evaporated under reduced pressure. The residue was
purified by column chromatography (silica gel, hexanes/ethyl acetate 2:1
+ 2% NEt3) yielding compound 2 as a white foam (1.22 g, 1.7 mmol,
88%) as a mixture of diastereomers. TLC (hexanes/ethyl acetate 2:1 +
2% NEt3): Rf
=
0.42; 31P NMR (101 MHz, CD3CN): d=149.3,
149.0 ppm; MALDI-TOF-MS: m/z: 702.7 [M ++H].
MeOH 1:1 v/v): lmax (e)=470 (19900), 450ACTHUNGETRNNU(G 19600), 266 nm
(107400 molꢁ1 dm3 cmꢁ1).
3’,5’-Bis(O-tert-butyldimethylsilyl)-N6-(N’-pyren-1-ylcarbamoyl)-2’-deox-
yadenosine (15): To
a
solution of pyrene isocyanate 13[56] (988 mg,
Perylen-3-yl-methanol: Perylen-3-carbaldehyde (400 mg, 1.4 mmol) and
NaBH4 (108 mg, 2.8 mmol, 2 equiv) were placed in a 25 mL vial and were
then dissolved in dry THF (10 mL). The solution was stirred at RT. After
2 h, the reaction mixture was diluted with CH2Cl2 (100 mL) and the solu-
tion was washed with brine (100 mL). Then the organic phase was dried
over Na2SO4 and the solvent was removed under reduced pressure. The
residue was purified by column chromatography (silica gel, CH2Cl2)
yielding perylen-3-yl-methanol as a dark yellow solid (191 mg, 0.7 mmol,
47%). TLC (CH2Cl2): Rf = 0.33; 1H NMR (250 MHz, CDCl3): d=8.25–
8.15 (m, 4H), 7.96 (d, J=8.4 Hz, 1H), 7.70 (d, J=7.6 Hz, 2H), 7.59–7.46
(m, 4H), 5.10 ppm (s, 2H); MS (FAB): m/z: 282.3 [M ++H]; UV/Vis
(CHCl3/MeOH 1:1 v/v): lmax (e)=440 (27500), 414 (23200), 394 (11800),
254 nm (35600 molꢁ1 dm3 cmꢁ1).
4.1 mmol, 1 equiv) in dry toluene (22 mL) a solution of 3’,5’-bis(O-tert-
butyldimethylsilyl)adenosine (14) (1.95 g, 4.1 mmol, 1 equiv) in dry
MeCN (30 mL) was added via a syringe under argon at 608C. The reac-
tion mixture was then stirred for 6 h at 608C and at RT over night. After-
wards methanol (30 mL) was added and after 15 min the solution was di-
luted with CH2Cl2 (100 mL). The organic layer was washed twice with a
mixture of brine and sat. NaHCO3 (200 mL, 1:1 v/v). The crude product
was used in the subsequent reaction without further purification. TLC
(hexanes/ethyl acetate 2:1): Rf = 0.34; TLC (CH2Cl2/MeOH 99:1): Rf
0.14; MS (FAB): m/z: 723.3 [M ++H].
=
N6-(N’-Pyren-1-ylcarbamoyl)-2’-deoxyadenosine (16): A solution of de-
rivative 15 (0.50 g, 0.7 mmol, 1 equiv) in THF (5 mL) was treated with
TBAF (1m solution in THF, 1.7 mL, 1.7 mmol, 2.5 equiv). The reaction
mixture was stirred at RT until TLC showed full conversion (1 h). Then
methoxytrimethylsilane (400 mL, 0.30 g, 2.9 mmol, 4.1 equiv) was added
and the solution was stirred for 0.5 h at RT. Afterwards it was diluted
with CH2Cl2 (200 mL) and washed with water (200 mL). The organic
layer was then dried over Na2SO4 and the solvent was evaporated under
reduced pressure. First the residue was purified by column chromatogra-
phy (silica gel, CH2Cl2/MeOH 95:5). Then it was washed with ethanol
yielding (79 mg, 0.16 mmol, 23%) of compound 16 as a light yellow solid.
3-(Azidomethyl)perylene: In a 25 mL vial, perylen-3-ylmethanol (141 mg,
0.5 mmol) and triphenylphosphine (197 mg, 0.8 mmol, 1.5 equiv) were
dissolved in dry DMF (3 mL) and then CBr4 (249 mg, 0.8 mmol,
1.5 equiv) was added at 08C. The solution was stirred for 10 min at 08C.
Afterwards NaN3 (194 mg, 3.0 mmol, 6 equiv) was added and the slurry
was stirred at RT for 24 h. Then the slurry was diluted with CH2Cl2
(100 mL) and the solution was washed with water (2ꢃ80 mL), dried over
Na2SO4 and the solvent removed under reduced pressure. The crude
product was purified by column chromatography (silica gel, hexanes/
1
TLC (CH2Cl2/MeOH 95:5): Rf = 0.14; H NMR (250 MHz, [D6]DMSO):
CH2Cl2 2:1) yielding 3-(azidomethyl)perylene as a dark green solid
1
d=12.70 (brs, 1H, NH), 10.54 (brs, 1H, NH), 8.89 (s, 1H, H8), 8.77 (d,
J=8.2 Hz, 1H, pyrene), 8.72 (s, 1H, H2), 8.44 (d, J=9.1 Hz, 1H,
pyrene), 8.34–8.25 (m, 4H, pyrene), 8.15–8.02 (m, 3H, pyrene), 6.48 (t,
(113 mg, 0.4 mmol, 74%). TLC (hexanes/CH2Cl2 2:1): Rf = 0.4; H NMR
(250 MHz, [D6]DMSO): d=8.44–8.35 (m, 4H), 7.92 (d, J=8.4 Hz, 1H),
7.81 (d, J=8.2 Hz, 2H), 7.63 (t, J=7.7 Hz, 2H), 7.54 (t, J=7.8 Hz, 2H),
Chem. Eur. J. 2009, 15, 3390 – 3402
ꢂ 2009 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.chemeurj.org
3399

C. Richert and M. Printz
4.87 ppm (s, 2H); MS (FAB): m/z: 307.3 [M ++H]; UV/Vis (CHCl3/
MeOH 1:1 v/v): lmax (e)=441 (31400), 415 (24100), 393 (11300), 254 nm
(33600 molꢁ1 dm3 cmꢁ1).
(0.2 mmol, 22 mg, 0.3 equiv) and DIEA (0.4 mmol, 61 mL, 0.5 equiv)
were added. After 4 h the solution was diluted with CH2Cl2 (50 mL) and
the solution was washed with aqueous citric acid (5% in water, 50 mL)
and water (50 mL). The organic phase was then dried over Na2SO4 and
the solvent was evaporated under reduced pressure. The residue was pu-
rified by column chromatography (silica gel, conditioned with CH2Cl2/
MeOH 95:5, v/v +2% DIEA, elution with CH2Cl2/MeOH 95:5, v/v)
yielding compound 29 as a white solid (388 mg, 0.5 mmol, 68%). TLC
(CH2Cl2/MeOH 95:5): Rf = 0.26; 1H NMR (500 MHz, CDCl3): d=8.35
(s, 1H, H8), 8.34 (s, 1H, H2), 8.17 (d, J=8.5 Hz, 1H), 7.56 (dd, J=7.8,
6.9 Hz, 1H), 7.51 (d, J=8.1 Hz, 1H), 7.48 (d, J=6.9 Hz, 1H), 7.40 (d, J=
7.1 Hz, 2H, DMTr), 7.32–7.18 (m, 7H, DMTr), 6.82 (d, J=8.1 Hz, 4H,
DMTr), 6.54 (dd, J=8.1, 6.9 Hz, 1H, H1’), 5.59 (d, J=5.6 Hz, 1H, H3’),
4.36 (m, 1H, H4’), 3.80 (s, 6H, DMTr), 3.52–3.44 (m, 2H, H5’+H5’’),
3.08–3.02 (m, 1H, H2), 2.76–2.68 ppm (m, 5H, H2+succinyl); 13C NMR
(126 MHz, CDCl3): d=176.2, 171.5, 159.1, 158.7, 153.6, 151.5, 144.3,
143.5, 135.4, 135.3, 130.0, 129.2, 128.9, 128.8, 128.1, 127.9, 127.9, 127.8,
127.1, 124.9, 120.6, 119.9, 113.3, 113.2, 108.7, 86.9, 85.1, 84.6, 75.6, 63.6,
55.3, 38.1, 29.0, 28.7 ppm; MALDI-TOF-MS: m/z: 792.6 [M ++Na]; UV/
Vis (CHCl3): lmax (e)=281 (6300), 241 nm (23100 molꢁ1 dm3 cmꢁ1).
Perylen-3-yl-methylamine (22): In a 50 mL vial, 3-(azidomethyl)perylene
(113 mg, 0.4 mmol) and Pd/C (30 mg) were treated with methanol
(30 mL). The resulting slurry was stirred under a hydrogen atmosphere
for 2 h. Afterwards the slurry was filtered over kieselguhr and the filter
cake was washed with hot methanol (200 mL). Then the solvent was re-
moved under reduced pressure and the residue was purified by column
chromatography (silica gel, CH2Cl2/MeOH 99:1 + 1% NEt3) yielding
compound 19 as a dark green solid (53 mg, 0.2 mmol, 51%). TLC
(CH2Cl2/MeOH 99:1
+ 1% NEt3): Rf =
0.10; 1H NMR (250 MHz,
[D6]DMSO): d=8.39 (t, J=8.3 Hz, 4H), 7.98 (d, J=8.2 Hz, 1H), 7.79
(dd, J=7.6, 5.7 Hz, 2H), 7.64–7.51 (m, 4H), 4.20 ppm (s, 2H); MS
(FAB): m/z: 279.3 [M ++H].
5’-O-(4,4’-Dimethoxytrityl)-N6-(perylen-3-yl-methyl)-2’-deoxyadenosine
(24): In a 25 mL vial, 2’-deoxyinosine (21; 40 mg, 0.16 mmol, 1 equiv),
(benzotriazol-1-yloxy)tris(dimethylamino)phosphonium hexafluorophos-
phate (BOP) (83 mg, 0.19 mmol, 1.2 equiv) and perylen-3-yl-methylamine
(22; 53 mg, 0.19 mmol, 1.2 equiv) were dissolved in dry DMF (2 mL) and
DIEA (68 mL, 0.24 mmol, 1.5 equiv) was added. Afterwards the slurry
was stirred at RT for 5 d. The red slurry was then diluted with CH2Cl2
(100 mL) and the solution washed with water (80 mL) and dried over
Na2SO4. The solvent was then removed under reduced pressure. The
crude product (72 mg, 0.14 mmol, 1 equiv) was dissolved in dry pyridine
(1.4 mL), then DMAP (10 mg, 82 mmol, 0.6 equiv) and DMTrCl (47 mg,
0.14 mmol, 1 equiv) were added and the solution was stirred at RT for
1.5 h. Then methanol (1 mL) was added and after stirring for additional
30 min the solution was diluted with CH2Cl2 (100 mL). Then the solution
was washed with aqueous citric acid (5% in water, 50 mL) and sat.
NaHCO3 (50 mL). The organic layer was dried over Na2SO4 and the sol-
vent removed under reduced pressure. The residue was purified by
Analytical data for modified oligonucleotides that were HPLC purified
Compound 1U_py: Yield: 29% (HPLC traces); HPLC: CH3CN gradient
0 ! 10% in 1 min, 10 ! 24% in 39 min, tR = 19 min; MALDI-TOF
MS: m/z: calcd for C85H94N29O40P6 [MꢁꢁH]: 2347.7, found 2345.2.
Compound 1N_pr: Yield: 19% (HPLC traces) HPLC: CH3CN 0 ! 5%
in 2 min, 5 ! 30% in 28 min, tR = 24 min; MALDI-TOF MS: m/z:
calcd for C89H97N28O39P6 [MꢁꢁH]: 2368.7, found 2367.7.
Compound 1R_py: Yield: 25% (HPLC traces); HPLC: CH3CN gradient
0 ! 5% in 2 min, 5 ! 23% in 28 min, tR = 22 min; MALDI-TOF MS:
m/z: calcd for C79H90N23O39P6 [MꢁꢁH]: 2171.5, found 2172.3.
Compound 1A*_pr: Yield: 33% (HPLC traces); HPLC: CH3CN gradi-
ent 0 ! 5% in 2 min, 5 ! 30% in 28 min, tR = 27 min; MALDI-TOF
MS: m/z: calcd for C92H102N29O40P6 [MꢁꢁH]: 2439.8, found 2437.1.
Compound 1P_pr: Yield: 33% (HPLC traces); HPLC: CH3CN gradient
0 ! 3% in 2 min, 3 ! 24% in 28 min, tR = 27 min; MALDI-TOF MS:
m/z: calcd for C87H99N24O39P6 [MꢁꢁH]: 2290.7, found 2291.3.
Compound 1P_aq: Yield: 9% (total yield); HPLC: CH3CN gradient 0
! 2% in 1 min, 2 ! 25% in 29 min, tR = 26 min; MALDI-TOF MS:
m/z: calcd for C78H89N24O41P6 [MꢁꢁH]: 2204.5, found 2205.3.
Compound 1P_py: Yield: 9% (total yield); HPLC: CH3CN gradient gra-
dient 0 ! 20% in 30 min, 20 ! 90% in 7 min, tR = 35 min; MALDI-
TOF MS: m/z: calcd for C80H91N24O39P6 [MꢁꢁH]: 2198.6, found 2198.7.
Compound 1P_ch Yield: 18% (total yield); HPLC: CH3CN gradient gra-
dient 0 ! 5% in 1 min, 5 ! 25% in 29 min, tR = 30 min; MALDI-TOF
MS: m/z: calcd for C87H121N24O42P6 [MꢁꢁH]: 2360.9, found 2360.8.
column chromatography (silica gel, conditioned with CH2Cl2
+ 2%
NEt3, eluted first with CH2Cl2 + 2% NEt3, then with CH2Cl2/MeOH
99:1 + 2% NEt3) yielding compound 24 as a brown solid (over two
steps, 33 mg, 0.04 mmol, 28%). TLC (CH2Cl2/MeOH 99:1 + 1% NEt3):
Rf = 0.26; 1H NMR (250 MHz, CDCl3): d=8.40 (s, 1H, H8), 8.20–8.08
(m, 4H, perylene+H2), 7.89 (d, J=8.4 Hz, 1H, perylene), 7.80 (s, 1H,
perylene), 7.86 (d, J=8.1 Hz, 2H, perylene), 7.52–7.43 (m, 4H, perylene),
7.39–7.35 (m, 2H, DMTr), 7.28–7.17 (m, 7H, DMTr), 6.77 (d, 4H, J=
8.9 Hz, DMTr), 6.37 (t, J=6.4 Hz, 1H, H1’), 6.28 (brs, 1H, NH), 5.21 (m,
2H, perylene-CH2-NH), 4.63 (m, 1H, H3’), 4.08 (m, 1H, H4’), 3.40–3.31
(m, 2H, H5’+H5’’), 2.81–2.70 (m, 1H, H2’), 2.49–2.40 ppm (m, 1H,
H2’’).
3’,5’-Bis-(O-tert-butyldimethylsilyl)-O6-(4-styrylphenyl)-2’-deoxyinosine
(28): To a mixture of inosine derivative 27 (100 mg, 0.17 mmol), Cs2CO3
(109 mg, 0.33 mmol, 2 equiv) and (E)-4-hydroxystilbene (66 mg,
0.33 mmol, 2 equiv) dry toluene (1.7 mL) was added and the slurry was
stirred for 2 h at 1058C. Then the slurry was diluted with CH2Cl2 (50 mL)
and washed with water (2ꢃ50 mL), dried over Na2SO4 and the solvent
was removed under reduced pressure. The residue was purified by
column chromatography (silica gel, CH2Cl2) yielding compound 28 as a
Compound 1N_na: Yield: 20% (total yield); HPLC: CH3CN gradient 0
! 3% in 1 min, 3 ! 14% in 34 min, tR = 34 min; MALDI-TOF MS:
m/z: calcd for C79H93N28O39P6 [MꢁꢁH]: 2244.6, found 2245.8.
Compound 1N_bp Yield: 25% (total yield); HPLC: CH3CN gradient 0
! 3% in 1 min, 3 ! 14% in 34 min, 14 ! 90% in 8 min, tR = 42 min;
MALDI-TOF MS: m/z: calcd for C81H95N28O39P6 [MꢁꢁH]: 2270.7, found
2268.9.
white foam (110 mg, 0.17 mmol, 99%). TLC (CH2Cl2): Rf
= 0.06;
1H NMR (250 MHz, CDCl3): d=8.41 (s, 1H, H8), 8.31 (s, 1H, H2), 7.50–
7.40 (m, 4H), 7.29–7.23 (m, 2H), 7.18 (m, 2H), 7.15 (m, 1H), 7.00 (d, J=
3.6 Hz, 2H), 6.43 (t, J=6.4 Hz, 1H, H1’), 4.53 (m, 1H, H3’), 3.93 (m, 1H,
H4’), 3.84–3.66 (m, 2H, H5’+H5’’), 2.54 (m, 1H, H2’), 2.38 (m, 1H,
H2’’), 0.81 (d, J=2.5 Hz, 18H, TBDMS), 0.00 ppm (s, 12H, TBDMS);
MS (FAB): m/z: 659.3 [M ++H]; UV/Vis (CHCl3/MeOH 1:1 v/v): lmax
(e)=311 (2900), 299 (3100), 237 nm (1400 molꢁ1 dm3 cmꢁ1).
Compound 1N_py: Yield: 29% (HPLC traces); HPLC: CH3CN gradient
0 ! 20% in 30 min, tR = 23 min; MALDI-TOF MS: m/z: calcd for
C85H95N28O39P6 [MꢁꢁH]: 2318.7, found 2316.2.
Compound 1N_an: Yield: 20% (total yield); HPLC: CH3CN gradient 0
! 3% in 1 min, 3 ! 18% in 39 min, tR = 34 min; MALDI-TOF MS:
m/z: calcd for C84H97N28O39P6 [MꢁꢁH]: 2308.7, found 2308.7.
O6-(Benzotriazol-1-yl)-5’-O-(4,4’-dimethoxytrityl)-3’-O-(succin-1-yl)-2’-
deoxyinosine (29): In a 10 mL vial, O6-(benzotriazol-1-yl)-5’-O-(4,4’-di-
methoxytrityl)-2’-deoxyinosine (26) (500 mg, 0.7 mmol, 1 equiv) was dis-
solved in dry CH2Cl2 (3.7 mL) and DIEA (2.2 mmol, 368 mL, 3 equiv)
and some grains of molecular sieves (3) ꢄ) were added and the solution
was stirred at RT for 20 min. After subsequent addition of succinic anhy-
dride (0.9 mmol, 97 mg, 1.3 equiv) the reaction mixture was stirred for
additional 3 h 20 min. Then another portion of succinic anhydride
Compound 1N_dp: Yield: 17% (total yield); HPLC: CH3CN gradient 0
! 7% in 2 min, 7 ! 30% in 28 min, tR = 24 min; MALDI-TOF MS:
m/z: calcd for C82H97N28O39P6 [MꢁꢁH]: 2284.7, found 2283.8.
Compound 1N_ad: Yield: 27% (total yield); HPLC: CH3CN gradient 0
! 23% in 35 min, tR
= 30 min; MALDI-TOF MS: m/z: calcd for
C79H101N28O39P6 [MꢁꢁH]: 2252.7, found 2252.2.
3400
www.chemeurj.org
ꢂ 2009 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Chem. Eur. J. 2009, 15, 3390 – 3402

Universal DNA Hybridization Probes
FULL PAPER
Compound 35: Yield: 25% (total yield); HPLC: CH3CN gradient 0 !
2% in 1 min, 2 ! 17% in 29 min, tR = 29 min; MALDI-TOF MS: m/z:
calcd for C107H120N39O50P8 [MꢁꢁH]: 3000.1, found 2998.8.
Compound 35c: Yield: 25% (total yield); HPLC: CH3CN gradient 0 !
2% in 1 min, 2 ! 18% in 29 min, tR = 23 min; MALDI-TOF MS: m/z:
calcd for C106H120N37O51P8 [MꢁꢁH]: 2976.1, found 2975.2.
Compound 35g: Yield: 40% (total yield); HPLC: CH3CN gradient 0 !
2% in 1 min, 2 ! 18% in 29 min, tR = 26 min; MALDI-TOF MS: m/z:
calcd for C107H120N39O51P8 [MꢁꢁH]: 3016.1, found 3014.3.
Compound 35t: Yield: 32% (total yield); HPLC: CH3CN gradient 0 !
2% in 1 min, 2 ! 18% in 28 min, tR = 21 min; MALDI-TOF MS: m/z:
calcd for C107H121N36O52P8 [MꢁꢁH]: 2991.1, found 2990.3.
Compound 36c: Yield: 27% (total yield); HPLC: CH3CN gradient 0 !
2% in 1 min, 2 ! 18% in 29 min, tR = 26 min; MALDI-TOF MS: m/z:
calcd for C106H120N37O51P8 [MꢁꢁH]: 2978.1, found 2977.9.
Compound 36g: Yield: 30% (total yield); HPLC: CH3CN gradient 0 !
2% in 1 min, 2 ! 18% in 29 min, tR = 30 min; MALDI-TOF MS: m/z:
calcd for C107H120N39O51P8 [MꢁꢁH]: 3016.1, found 3016.5.
Compound 36t: Yield: 27% (total yield); HPLC: CH3CN gradient 0 !
2% in 1 min, 2 ! 21% in 29 min, tR = 27 min; MALDI-TOF MS: m/z:
calcd for C107H121N36O52P8 [MꢁꢁH]: 2991.1, found 2992.5.
Wamberg, E. B. Pedersen, Helv. Chim. Acta 2004, 87, 469–478;
c) A. T. M. Zafrul Azam, M. Hasegawa, T. Moriguchi, K. Shinozuka,
Bioorg. Med. Chem. Lett. 2004, 14, 5747–5750; d) I. Luyten, P. Her-
dewijn, Eur. J. Med. Chem. 1998, 33, 515–576.
[16] a) E. L. Kennaway, I. Hieger, Br. Med. J. 1930, 1, 1044–1046;
b) N. C. Yang, A. J. Castro, M. Lewis, T. W. Wong, Science 1961, 134,
386–387; c) H. Pitot, Fundamentals of Oncology, 3rd ed., Marcel
Dekker, New York, 1986.
[17] L. Zhang, H. Long, G. E. Boldt, K. D. Janda, G. C. Schatz, F. D.
Lewis, Org. Biomol. Chem. 2006, 4, 314–322.
[18] N. P. Puri, E. Zamaratski, C. Sund, J. Chattopadhyaya, Tetrahedron
1997, 53, 10409–10432.
[19] S. Al-Rawi, C. Ahlborn, C. Richert, Org. Lett. 2005, 7, 1569–1572.
[20] a) M. W. Reed, A. D. Adams, J. S. Nelson, R. B. Meyer, Bioconju-
gate Chem. 1991, 2, 217–225; b) M. W. Reed, E. A. Lukhtanov, V. V.
Gorn, D. D. Lucas, J. H. Zhou, S. B. Pai, Y.-C. Cheng, R. B. Meyer,
J. Med. Chem. 1995, 38, 4587–4596.
[21] J. C. Verheijen, A.-M. M. Van Roon, A. C. van der Laan, G. A. van
der Marel, J. H. Van Boom, Nucleosides Nucleotides 1999, 18, 493–
508.
[22] C. Richert, P. Grꢀnefeld, Synlett 2007, 1–18.
[23] S.-I. Nakano, Y. Uotani, S. Nakashima, Y. Anno, M. Fuji, N. Sugi-
moto, J. Am. Chem. Soc. 2003, 125, 8086–8087.
[24] K. Shimada, T. Oe, Anal. Sci. 1990, 6, 461–463.
[25] K. M. Guckian, B. A. Schweitzer, R. X.-F. Ren, C. J. Sheils, P. L.
Paris, D. C. Tahmassebi, E. T. Kool, J. Am. Chem. Soc. 1996, 118,
8182–8183.
[26] a) A. V. Azhayev, Tetrahedron 1999, 55, 787–800; b) A. V. Azhayev,
M. L. Antopolsky, Tetrahedron 2001, 57, 4977–4986.
[27] Z.-K. Wan, E. Binnun, D. P. Wilson, J. Lee, Org. Lett. 2005, 7, 5877–
5880.
Acknowledgements
The authors thank M. Rçthlingshçfer, C. Deck, K. Imhof, K. Mꢀller, and
K. Heimann for a review of the manuscript. This work was supported by
the DFG (grant FOR 434 and RI 1063/4-4 to C.R.). M.P. was partially
supported by a predoctoral fellowship from the State of Baden-Wꢀrttem-
berg.
[28] M. Masuko, S. Ohuchi, K. Sode, H. Ohttani, A. Shimadzu, Nucleic
Acids Res. 2000, 28, e34.
[29] T. J. Dale, J. Rebek, J. Am. Chem. Soc. 2006, 128, 4500–4501.
[30] U. Asseline, E. Cheng, Tetrahedron Lett. 2001, 42, 9005–9010.
[31] M. Printz, C. Richert, J. Comb. Chem. 2007, 9, 306–320.
[32] S. Hainke, S. Arndt, O. Seitz, Org. Biomol. Chem. 2005, 3, 4233–
4238.
[33] S. Bae, M. K. Lakshman, J. Am. Chem. Soc. 2007, 129, 782–789.
[34] T. Brown, C. E. Pritchard, G. Turner, S. A. Salisbury, J. Chem. Soc.
Chem. Commun. 1989, 891–893.
[35] R. H. Alul, C. N. Singman, G. Zhang, R. L. Letsinger, Nucleic Acids
Res. 1991, 19, 1527–1532.
[36] C. F. Bleczinski, C. Richert, J. Am. Chem. Soc. 1999, 121, 10889–
10894.
[1] K. B. Mullis, Angew. Chem. 1994, 106, 1271–1276; Angew. Chem.
Int. Ed. Engl. 1994, 33, 1209–1213.
[2] J. A. Lake, Sci. Am. 1981, 245, 84–97.
[3] a) S. P. A. Fodor, J. L. Read, M. C. Pirrung, L. Stryer, A. T. Lu, D.
Solas, Science 1991, 251, 767–773; b) M. C. Pirrung, Angew. Chem.
2002, 114, 1326–1341; Angew. Chem. Int. Ed. 2002, 41, 1276–1289.
[4] F. H. C. Crick, J. Mol. Biol. 1966, 19, 548–555.
[5] U. Plutowski, C. Richert, Angew. Chem. 2005, 117, 627–631; Angew.
Chem. Int. Ed. 2005, 44, 621–625.
[6] J. Tuma, C. Richert, Biochemistry 2003, 42, 8957–8965.
[7] J. Tuma, W. H. Connors, D. H. Stitelman, C. Richert, J. Am. Chem.
Soc. 2002, 124, 4236–4246.
[37] S. Narayanan, J. Gall, C. Richert, Nucleic Acids Res. 2004, 32, 2901–
2911.
[38] E. T. Kool, Acc. Chem. Res. 1998, 31, 502–510.
[39] M. Berger, Y. Wu, A. K. Ogawa, D. L. McMinn, P. G. Schultz, F. E.
Romesberg, Nucleic Acids Res. 2000, 28, 2911–2914.
[40] B. R. Babu, J. Wengel, Chem. Commun. 2001, 20, 2114–2115.
[41] B. R. Babu, A. K. Prasad, S. Trikha, N. Thorup, V. S. Parmar, J.
Wengel, J. Chem. Soc. Perkin Trans. 1 2002, 22, 2509–2519.
[42] D. Loakes, Nucleic Acids Res. 2001, 29, 2437–2447.
[43] C. Ahlborn, K. Siegmund, C. Richert, J. Am. Chem. Soc. 2007, 129,
15218–15232.
[8] K. Siegmund, S. Maheshwary, S. Narayanan, W. Connors, M. Rie-
drich, M. Printz, C. Richert, Nucleic Acids Res. 2005, 33, 4838–4848.
[9] J. Tuma, R. Paulini, J. A. Rojas Stꢀtz, C. Richert, Biochemistry 2004,
43, 15680–15687.
[10] Z. Dogan, R. Paulini, J. A. Rojas Stꢀtz, S. Narayanan, C. Richert, J.
Am. Chem. Soc. 2004, 126, 4762–4763.
[11] K. Yamana, H. Zako, K. Asazuma, R. Iwase, H. Nakano, A. Mura-
kami, Angew. Chem. 2001, 113, 1138–1140; Angew. Chem. Int. Ed.
2001, 40, 1104–1106.
[44] B. E. Eaton, L. Gold, D. A. Zichi, Chem. Biol. 1995, 2, 633–638.
[45] a) R. X.-F. Ren, N. C. Chaudhuri, P. L. Paris, S. I. V. Rumney, E. T.
Kool, J. Am. Chem. Soc. 1996, 118, 7671–7678; b) S. Smirnov, T. J.
Matray, E. T. Kool, C. de los Santos, Nucleic Acids Res. 2002, 30,
5561–5569.
[46] T. Hendrickson, W. C. Still, J. Comput. Chem. 1990, 11, 440–467.
[47] a) H. J. Yeh, J. M. Sayer, X. Liu, A. S. Altieri, R. A. Byrd, M. K.
Lakshman, H. Yagi, E. J. Schurter, D. G. Gorenstein, D. M. Jerina,
Biochemistry 1995, 34, 13570–13581; b) J. L. Schwartz, J. S. Rice,
B. A. Luxon, J. M. Sayer, G. Xie, H. J. Yeh, X. Liu, D. M. Jerina,
D. G. Gorenstein, Biochemistry 1997, 36, 11069–11076; c) D. E.
Volk, J. S. Rice, B. A. Luxon, H. J. Yeh, C. Liang, G. Xie, J. M.
Sayer, D. M. Jerina, D. G. Gorenstein, Biochemistry 2000, 39,
[12] a) J. Gao, C. Straessler, D. Tahmassebi, E. T. Kool, J. Am. Chem.
Soc. 2002, 124, 11590–11591; b) A. T. Krueger, E. T. Kool, J. Am.
Chem. Soc. 2008, 130, 3989–3999.
[13] a) F. A. Aldaye, H. F. Sleiman, J. Am. Chem. Soc. 2007, 129, 13376–
13377; b) F. A. Aldaye, H. F. Sleiman, Angew. Chem. 2006, 118,
2262–2267; Angew. Chem. Int. Ed. 2006, 45, 2204–2209.
[14] a) J. Zimmermann, M. P. J. Cebulla, S. Monninghoff, G. von Kie-
drowski, Angew. Chem. 2008, 120, 3682–3686; Angew. Chem. Int.
Ed. 2008, 47, 3626–3630; b) L. H. Eckardt, K. Naumann, W. M.
Pankau, M. Rein, M. Schweitzer, N. Windhab, G. von Kiedrowski,
Nature 2002, 420, 286.
[15] a) I. Van Daele, N. Bomholt, V. V. Filichev, S. Van Calenbergh, E. B.
Pedersen, ChemBioChem 2008, 9, 791–801; b) K. Walczak, M.
Chem. Eur. J. 2009, 15, 3390 – 3402
ꢂ 2009 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.chemeurj.org
3401

C. Richert and M. Printz
14040–14053; d) P. Pradhan, S. Tirumala, X. Liu, J. M. Sayer, D. M.
Jerina, H. J. Yeh, Biochemistry 2001, 40, 5870–5881; e) D. E. Volk,
V. Thiviyanathan, J. S. Rice, B. A. Luxon, J. H. Shah, H. Yagi, J. M.
Sayer, H. J. C. Yeh, D. M. Jerina, D. G. Gorenstein, Biochemistry
2003, 42, 1410–1420; f) B. Mao, Z. Gu, A. Gorin, J. Chen, B. E.
Hingerty, S. Amin, S. Broyde, N. E. Geacintov, D. J. Patel, Biochem-
istry 1999, 38, 10831–10842; g) I. S. Zegar, S. J. Kim, T. N. Johansen,
P. J. Horton, C. M. Harris, T. M. Harris, M. P. Stone, Biochemistry
1996, 35, 6212–6224; h) B. Feng, A. Gorin, A. Kolbanovskiy, B. E.
Hingerty, N. E. Geacintov, S. Broyde, D. J. Patel, Biochemistry 1997,
36, 13780–13790; i) N. Zhang, C. Lin, X. Huang, A. Kolbanovskiy,
B. E. Hingerty, S. Amin, S. Broyde, N. E. Geacintov, D. J. Patel, J.
Mol. Biol. 2005, 346, 951–965.
[50] G. Cui, C. Simmerling, J. Am. Chem. Soc. 2002, 124, 12154–12164.
[51] E. T. Kool, Curr. Opin. Chem. Biol. 2000, 4, 602–608.
[52] L. Sun, K. Zhang, L. Zhou, P. Hohler, E. T. Kool, F. Yuan, Z. Wang,
J. S. Taylor, Biochemistry 2003, 42, 9431–9437.
[53] L. Sun, M. Wang, E. T. Kool, J.-S. Taylor, Biochemistry 2000, 39,
14603–14610.
[54] W. S. Newcomb, T. L. Deegan, W. Miller, J. A. Porco, Biotechnol.
Bioeng. 1998, 61, 55–60.
[55] S. Mahajan, A. Garg, M. Goel, P. Kumar, K. C. Gupta, Anal. Bio-
chem. 2006, 351, 273–281.
[56] S. Kazutake, O. Tomoyuki, Anal. Sci. 1990, 6, 461–463.
[48] W. Humphrey, A. Dalke, K. Schulten, J. Mol. Graphics 1996, 14,
33–38.
[49] S. Smirnov, T. J. Matray, E. T. Kool, C. de los Santos, Nucleic Acids
Res. 2002, 30, 5561–5569.
Received: August 1, 2008
Revised: November 15, 2008
Published online: February 13, 2009
3402
www.chemeurj.org
ꢂ 2009 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Chem. Eur. J. 2009, 15, 3390 – 3402