Synthetic access to the chemical diversity of DNA and RNA 5′-aldehyde lesions

Article abstract of DOI:10.1021/jo502170e

Hydrogen atom abstraction from the C5′-position of nucleotides in DNA results in direct strand scission by generating alkali-labile fragments from the oxidized nucleotide. The major damage consists in a terminus containing a 5′-aldehyde as part of an othe

Full text of DOI:10.1021/jo502170e

Featured Article
pubs.acs.org/joc
Synthetic Access to the Chemical Diversity of DNA and RNA
′‑Aldehyde Lesions
5
́
Remy Lartia* and Jean-Franc o̧ is Constant
́ ́ ́
Departement de Chimie Moleculaire, UMR CNRS 5250, Universite Grenoble Alpes, BP 53, 38041 Grenoble Cedex 9, France
*
S Supporting Information
ABSTRACT: Hydrogen atom abstraction from the C5′-position of
nucleotides in DNA results in direct strand scission by generating alkali-
labile fragments from the oxidized nucleotide. The major damage consists
in a terminus containing a 5′-aldehyde as part of an otherwise undamaged
nucleotide. Moreover it is considered as a polymorphic DNA strand break
lesion since it can be borne by any of the four nucleosides encountered in
DNA. Here we propose an expeditious synthesis of oligonucleotides
(
ON) ending with this 5′-aldehyde group (5′-AODN). This straightfor-
ward and cheap strategy relies on Pfitzner−Moffatt oxidation performed
on solid support followed by a transient protection of the resulting
aldehyde function. This method is irrespective of the 5′-terminal
nucleobase and most interestingly can be directly extended to RNA to produce the corresponding 5′-AORN. We also report
preliminary results on recognition of 5′-AODN by base excision repair (BER) enzymes.
INTRODUCTION
neocarzinostatin under aerobic conditions, followed by
■
(
extensive RP-HPLC purification. The yields were generally
DNA is continuously exposed to various damaging conditions
oxidizing agents, ionizing radiations or endogenous species)
that produce strand breaks, cross-links or damaged nucleotides
4
low and highly sequence-dependent.
More recently, Greenberg et al. described their synthesis by
1
automated incorporation of the thymidine phosphoramidite 2
through radical abstraction of hydrogen atoms. DNA strand
9
(
T-Al) at the 5′ end of an ODN (Scheme 1). After complete
breaks terminated by a 5′-aldehyde 1 (hereafter called 5′-
deprotection of the modified ODN, the aldehyde function was
released upon oxidation by sodium periodate. However, this
approach suffered from three drawbacks: (i) synthesis from
AODN, cf. Scheme 1) is one common 5′-end damage produced
2
from either 3′ (internucleotidic) or 5′ hydrogen abstraction.
9
,10
thymidine requires 10 steps with an overall 24% yield,
(ii)
Scheme 1. Structure of the 5′-Aldehyde 1 and of T-Al
9
the synthesis of the other phosphoramidites (A, C and G
derivatives) is not described and would require supplementary
protection steps, (iii) the synthesis of 3′-unmodified RNA
fragments ending with a 5′-aldehyde (5′-AORN) is not
possible, since the final treatment with periodate ring-opens
the ribose moiety. As a consequence, the precise investigation
of the biochemistry of this lesion still requires the design of a
versatile and user-friendly synthetic pathway giving access to its
polymorphism.
Phosphoramidite 2
Generation of a 5′-aldehyde strand break was obtained by
5
In previous studies Greenberg et al. have shown that the
lesion formed from oxidation of thymidine is not a substrate for
nucleotide excision repair pathway but that its excision can be
achieved by strand displacement synthesis by DNA polymerase
β (Pol β) in the presence or absence of flap endonuclease 1
3
,4
many molecular systems, such as enediyne, porphyrinic rings
6
or copper complexes. Analytical methods were also developed
7
in order to quantify this lesion. This strand break, which
retains a nucleotide fragment terminated by a non-natural 5′-
aldehyde extremity is most likely cytotoxic for cells if not
8
8
(FEN1). In the present study we decided to test a collection of
removed by DNA repair process.
enzymes representative of the base excision repair (BER)
pathway. Because of the fact that the 5′-aldehyde lesion is
produced opposite its cognate nucleotide, it is not expected to
be a specific substrate for a DNA glycosylase, the first step in
A precise study of the biochemistry of this lesion requires the
development of a versatile synthetic route giving access to DNA
fragments and taking into account its polymorphism since eight
different nucleotides are expected (A, T, G, C and their RNA
counterparts).
The first method was based on reaction of a self-
complementary duplex ODN with an enediyne drug such as
Received: September 22, 2014
©
XXXX American Chemical Society
A
dx.doi.org/10.1021/jo502170e | J. Org. Chem. XXXX, XXX, XXX−XXX

The Journal of Organic Chemistry
Featured Article
BER. Nevertheless, because of the propensity of the 4′
hydrogen to be abstracted through β-elimination reaction
Scheme 3. General Synthetic Pathway from CPG-Linked 5′-
a
AODN to Deprotected 5′-AODN
(Scheme 2), we considered judicious to test a series of mono-
and bifunctional glycosylases.
Scheme 2. β- and γ-Elimination Pathways from 5′-AODN
a
Structures of CPG-linked 5′-AODN 3a, of CPG-linked DPI-
Both types of enzymes use amino acid residues as
nucleophiles to eliminate abnormal nucleobases leaving a 2′-
deoxyribose residue (abasic or AP site). Furthermore, bifunc-
tional glycosylases perform a β-elimination reaction on 2′
position of deoxyribose following the modified nucleobase
removal. Even if it is not very likely that the 5′-aldehyde lesion
is a specific substrate for these enzymes, we cannot rule out a
favorable interaction leading to a significant β-elimination
reaction producing a 5′-phosphate end with the loss of a
nucleoside. The bifunctional glycosylases considered were
hOGG1 (human 7,8-dihydro-8-oxoguanine-DNA glycosylase),
Nth (Escherichia coli endonuclease III), Fpg (E. coli
formamidopyrimidine-DNA glycosylase) and the monofunc-
tional glycosylases are MUG (E. coli mismatch specific uracil-
DNA glycosylase), TagI (E. coli 3-methyladenine-DNA-
glycosylase I) and ANPG 40 and 60 (truncated human
alkylpurine-DNA N-glycosylase).
protected 5′-AODN 3b, of DPI-protected 5′-AODN 3c and of 5′-
Bz
iBu
Ac
AODN 1. B* = A, T, G or C; B = A, T, G or C; gray ball stands
for the CPG support. (a) N,N′-diphenylethylenediamine, PTSA or
DDQ (see text); (b) 40% aq. methylamine 65 °C, 10 min; (c) RP-
HPLC purification; (d) AcOH 80%, 5 min.
synthesis, 5′-oxidation and protection by DPI) were obtained as
demonstrated by RP-HPLC analysis (see Table 1).
Table 1. Protection of a Model CPG-Linked 5′-AODN by
N,N′-Diphenylethylenediamine Following Method A (PTSA
Activation) or Method B (DDQ Activation)
overall reaction yields (%)
method A
method B
model CPG-linked 5′-AODN ( NTn3′) n = 4 n = 11 n = 4 n = 11
5′
N = A
N = T
N = G
N = C
65
66
67
73
63
57
71
73
74
80
71
77
65
69
50
68
RESULTS AND DISCUSSION
■
Synthesis of 5′AODN. We previously developed a method
to master the oxidation of the 5′- primary alcohol of a
controlled-pore glass (CPG) linked ODN just prior to the final
11
alkaline deprotection step. Under mild on-support Pfitzner−
Moffatt conditions, roughly 70−90% of the primary alcohol are
converted into aldehyde, whatever the nature of the last
nucleobase. Such modified ODNs are known to be
quantitatively converted to truncated 5′-phosphorylated
ODNs through β-elimination during the final alkaline
deprotection step (Scheme 2). Nevertheless, we believed that
a suitable protection of the aldehyde prior to the final
deprotection step could give an easy access to 5′-AODN in
solution.
The strong liphophilicity of the DPI group enabled RP-
HPLC purification: the DPI-protected 5′-AODN was easily
separated from failure sequences, unreacted 5′-AODN, β-
11
elimination product and methylthiomethyl adduct (see Figure
1). DPI-protected 5′-AODN was then deprotected by treat-
ment with 80% aqueous AcOH, which is commonly used in
ODN chemistry. Full deprotection was achieved in less than 5
min at rt. Altogether these elements show that 5′-AODN can
be synthesized and purified as routinely as nonmodified ODNs,
following procedures similar to the well-known “DMT-On” and
“DMT-Off” strategies.
Ideally, CPG-linked 5′-AODN 3a (Scheme 3) have to be
suitably derivatized to resist the strong alkaline deprotection
conditions and then be easily separated from unreacted and
truncated ODNs. Moreover, the 5′-aldehyde group has to be
rapidly recovered by a simple treatment. Keeping these criteria
in mind, we retained the N,N′-diphenylimidazoline (DPI)
group as an ideal candidate for transient protection of the
aldehyde function in the CPG-linked 5′-AODN (Scheme 3) .
DPI group is known to withstand ammonia deprotection and is
We next extended our reaction conditions to various ODNs
sequences. Whatever their length, ODNs were efficiently
converted into 5′-AODNs. Slightly lower yields are observed
for longer sequences, most likely due to overall elongation
chain yield (see Table 2).
Notably, the reaction properly worked whatever the nature of
the 5′-nucleobase (entries 2−5), although for a given length
slightly better yields were obtained with pyrimidines (entries 2
and 4 vs 3 and 5).
We further extended our method to oligoribonucleotides
(ORNs) oxidation. To the best of our knowledge, this is the
first method allowing the chemical modification of the 5′ end of
an ORN without using specifically designed phosphoramidite
synthons.
1
2,13
easily removed by mild acidic treatment.
Model ODNs were treated by N,N′-diphenylethylenedi-
amine in the presence of PTSA (para-toluene sulfonic acid,
method A) or DDQ (2,3-dichloro-5,6-dicyano-para-benzoqui-
1
4
none, method B). The length of the ODNs and the nature of
the final nucleobase were both screened. Whatever the reaction
conditions, good overall reaction yields (comprising automated
B
dx.doi.org/10.1021/jo502170e | J. Org. Chem. XXXX, XXX, XXX−XXX

The Journal of Organic Chemistry
Featured Article
CPG linked 5′-AORN were synthesized according to our
11
previously described method. Unfortunately, when N,N′-
diphenylethylenediamine was used as protecting group, no
DPI-protected 5′-AORN was recovered. We suspected that the
conditions classically used for deprotecting the 2′-OtBDMS
group of tritylated RNA (i.e., TEA·3HF in DMSO/TEA 2:1 at
6
5 °C, 60 min; called method C) were still too harsh for the
16
DPI group. Using milder conditions (NH F in a DMSO/
4
water/CH NH mixture at pH = 8.5; method D), 21% of the
3
2
model DPI-protected 5′-AORN U were recovered (see Table
8
3
).
Table 3. Protection Reaction Yields of CPG-Linked 5′-
AODN and 5′-AORN by N,N′-Diphenylethylenediamine
Derivatives Following Method A
5
′- DPI-protected
5′- DPI-protected
b c
a
AODN
AORN (yield )
c
yield
%)
tR
(min)
method
C
method
D
(
N,N′-diphenyl
65
15.1
traces
21
ethylenediamine
1
1
1
5 (4-Br)
86
42
66
16.5
14.6
17.1
34
28
34
58
29
57
6 (4-CN)
7 (3,5-bis Cl)
a
b
Sequence of the model CPG-linked 5′-AODN was AT . Sequence
4
c
of the model CPG-linked 5′-AORN was U . Overall reaction yields
8
(
ODN chain elongation, 5′-alcohol oxidation, protection by N,N′-
diphenylethylenediamine derivatives and deprotection) were given.
Estimated by RP-HPLC.
In order to improve the reaction yields, other protecting
groups were screened. The DPI acid-sensitivity can be indeed
lowered by introducing electron-withdrawing groups on phenyl
rings and thus N,N′-bis(4-bromophenyl)-ethylenediamine 15,
N,N′-bis(4-cyanophenyl)-ethylenediamine 16 and N,N′-bis-
Figure 1. RP-HPLC profiles of 5′-AODN 13. (a) Protected as DPI
derivative (crude deprotection mixture); (b) protected as DPI
derivative (after RP-HPLC purification); (c) unprotected (after 5
min 80% AcOH treatment). Peak 1: failure sequence and 5′-
unoxidized ODN (natural ODN); peak 2: β-elimination product;
peak 3: methylthiomethyl adduct; peak 4: DPI-protected 5′-AODN
and peak 5:5′-AODN 13.
(
3,5-dichlorophenyl)-ethylenediamine 17 were synthesized.
First of all, we evaluated their efficiency on a CPG-linked 5′-
11
AODN model (sequence AT ) and reagents 15−17 gave good
4
yields (from 42 to 86%, see Table 3, column 2) and compared
favorably with their unsubstituted parent reagent (65% yield).
We then used reagents 15−17 to protect CPG-linked 5′-
AORN (sequence U ), and the reaction yields were enhanced
8
Table 2. ESI-MS and HPLC Data of Protected 5′-AODN (4−13) and Derivatized 5′-AODN
DPI-protected 5′-AODN
5′-AODN
MW
c
MW
a
b
entries
sequence of the 5′-AODN
TTG TT (4)
theor.
exp.
tR (min)
yield (%)
tR (min)
theor.
exp.
1
2
3
4
5
6
7
8
9
1675.5
3787.8
3778.8
3803.8
3763.8
5040.0
5032.0
7533.5
5329.1
6762.3
3522.8
1675.2
3787.2
3779.3
3803.3
3764.3
5039.6
5033.0
7533.7
5328.9
6762.2
3523.3
15.40
16.83
18.18
16.85
17.43
17.10
17.33
17.50
17.00
17.20
17.51
76
74
80
71
77
49
54
49
42
52
50
12.55
13.30
13.37
13.27
13.32
13.14
13.21
13.50
13.30
13.50
13.40
1636.6
3748.9
3739.9
3764.9
3724.9
5001.2
4992.6
7494.6
5290.2
6724.3
3483.9
1636.3
3749.1
3740.3
3765.1
3725.3
5002.3
4994.1
7497.0
5290.2
6725.0
3483.9
ATT TTT TTT TTT (5)
TTT TTT TTT TTT (6)
GTT TTT TTT TTT (7)
CTT TTT TTT TTT (8)
TCA GCT AGA CCA TGC A (9)
CCA CGC ATC GCT GGT A (10)
TTC AGA TAC TTA GCA TGG ACA ACA (11)
TCG ACA TCG ACA TCG CA (12)
TCC ATA TCC AAT TCA CAT ACT C (13)
TAG CTT GAC CG (14)
1
0
1
1
a
b
For clarity purpose, the 5′-oxidized nucleoside is underlined. Global yields of synthesis are given (ODN elongation, on-support oxidation and
protection as DPI derivatives) and were determined by UV monitoring on crude reaction mixture (see SI for consideration on UV absorbance of the
DPI group). 5′-AODNs were derivatized and analyzed by ESI-MS as n-decyloxime ethers.
c
15
C
dx.doi.org/10.1021/jo502170e | J. Org. Chem. XXXX, XXX, XXX−XXX

The Journal of Organic Chemistry
Featured Article
a
even when method C was used for the 2′-OtBDMS groups
removal (from 0 to 34%). However, better yields were obtained
with deprotection method D (see Table 3). Interestingly, the
presence of halogen atoms in 15 and 17 facilitates the
purification step by increasing the retention times of the
resulting 5′-DPI-protected AODN (Table 3, column 3).
With these conditions in hand, we were able to synthesize
and purify 5′-AORN (see Table 4).
Scheme 4. Reduction of 5′-AODN
a
ODN sequence was 5′CGCAGAGACGC . (a) NaBR , R O.
3′
4
2
Table 4. ESI-MS and HPLC Data of DPI-Protected 5′-
AORN (Method A)
Action of BER Enzymes. The β-elimination activity of
various DNA glycosylases was evaluated on ternary complexes
containing a 3′-end labeled 5′-AODN (17 mer, the 3′ end
labeling adds an adenosin residue) and a 5′ phosphate ODN
MW
b
tR
yield
(%)
a
c
entries
sequence
UUU UUU UU (18a)
theor.
exp.
(min)
(
15 mer) hybridized with a 32 mer complementary strand.
d
e
1
2
3
2736.2
2718.2
4144.5
2736.2
2715.8
4145.0
21.00
21.80
17.90
57
58
41
Despite the fact that it is well-known that MUG, Tag1 and
ANPG are not able to cleave an abasic site (2′-deoxyribose or
AP site), their effect on the 5′-aldehyde lesion was investigated
since this lesion was shown to be very sensitive to β-elimination
UUU UUU UU (18b)
UGA CGA UCG UUA
e
(
19b)
4
5
6
7
8
AAU CCA AUC UGA
6643.8
6623.8
3816.4
3756.3
3736.4
6644.2
6624.4
3815.0
3755.1
3735.1
19.56
20.72
20.52
19.42
20.72
47
17
55
56
32
8
d
e
reaction.
UUC CCA AG (20a)
As can be seen in Figure 2, the 5′-aldehyde lesion is resistant
to the AP lyase activity of Fpg, Nth and hOGG1 and is not
either substrate for the monofunctional glycosylases Mug,
Tag1, ANPG 40 and 80.
The absence of cleavage does not exclude the formation of
tight complexes, and the intrinsic reactivity of the lesion may
provoke a covalent cross-link with the interacting protein. In
our case, when the DNA/protein complex is formed, the
aldehyde group may react with close amino groups forming
unstable Schiff base, which can be chemically reduced. Even in
the presence of a large excess (50 mM) of sodium
triacetoxyborohydride, no covalent trapping could be detected
for the seven tested enzymes (data not shown).
AAU CCA AUC UGA
UUC CCA AG (20b)
UAG CUU GAC CG
d
(
21a)
GUU UCU UAU GU
d
(
22a)
GUU UCU UAU GU
e
(
22b)
a
b
For clarity purpose, the 5′-oxidized nucleoside is underlined. Global
yields of synthesis are given (ODN elongation, on-support oxidation
and protection as DPI derivatives) and were determined by UV
monitoring on crude reaction mixture. Theoretical mass of protected
c
5
′-AORN containing the most abundant combination of isotopes was
d e
given. Protected by 15 as 4,4′-dibromo DPI. Protected by 17 as
3
,3′,5,5′-tetrachloro DPI.
CONCLUSION
■
Both reagents 15 and 17 gave consistently good results,
5′-AODN, as well as the unprecedented 5′-AORN, can now be
accessed in two working days irrespective of the nature of the
final nucleobase, using cheap and simple procedure. The
versatility and the efficiency of this method compares favorably
with previously described chemical method, which required
long and tedious synthetic work and was limited to the access
of 5′-AODN ending with a thymidine. In this paper we show
that the 5′-aldehyde lesion is not substrate for several major
BER enzymes. The user-friendly method developed here can be
a milestone, as it gives an easy access to all possible 5′-aldehyde
containing nucleic acids and should help in understanding the
enzymatic reparation mechanisms of this lesion.
although 4,4′-bis-bromoDPI derivatives seem to slightly better
withstand the 2′-OtBDMS deprotection conditions, especially
for longer sequences (compare entries 4 and 6 to 5 and 7).
Both protecting groups were removed by treating with 80%
AcOH during less than 5 min.
It should be noted that method A (PTSA) was used to
achieve the aldehyde protection of the CPG-linked 5′-AORN
by substituted DPI 15−17. Little 5′-AORN was recovered
when method B (DDQ) was used (<10%).
For both 5′-AODN and 5′-AORN (DPI-protected or not), it
should be mentioned that they cannot be stored for long
periods, even frozen at −80 °C, and should be used in the days
following their synthesis (see Experimental Section for more
details).
Works are currently on progress in our lab on the 5′-
aldehyde chemistry and on its biochemistry in RNA and DNA
(notably in a context of radio-induced clusters of lesions).
To further assess the presence of the aldehyde moiety in our
synthesized 5′-AODN, the model 5′-AODN 23 (MW = 3267.6
EXPERIMENTAL SECTION
Oligonucleotides Synthesis and Purification. Oligodeoxyribo-
nucleotides synthesis was performed following the phosphoramidite
strategy on 1 μmol scale, by using classical conditions. Briefly, 0.2 M
■
17
amu) was subjected to NaBH and NaBD reductions
4
4
affording products 24 and 25 with respectively +2.1 and +3.0
amu. (Scheme 4).
4
,5-dicyanoimidazole in MeCN was used as activator, 0.02 M iodine
5
′-AODN’s are stable in aqueous solutions. Half-life of ca.
solution in THF/water/pyridine as oxidizer, acetic anhydride in THF
and 16% N-methylimidazole in THF as cap A and B reagents, 3% TCA
in DCM as deblock reagent and acetonitrile (<30 ppm water) as
washing solvent. Preloaded 500 Å CPG (controlled pore-glass
support) was used. Oligoribonucleotides were synthesized by using
the same protocol except that 0.25 M ETT solution was used as
activator, nucleosides were introduced as 2′-tBDMS derivatives, and
coupling times were extended from 20 to 300 s.
1
00 h was reported by Greenberg for hybridized 5′-AODN in
9
PBS buffer (pH 7.2 at 37 °C). In our hands, nonhybridized 5′-
AODNs were also extremely stable whatever the nature of the
last nucleobase: only 15% of 5′-AODN 5−8 decomposed by β-
elimination after 10 days in PBS buffer (pH 8.0 at rt). On the
other hand, heating at 60 °C dramatically increased the
decomposition rate (half-life 15.1 h, data not shown)
D
dx.doi.org/10.1021/jo502170e | J. Org. Chem. XXXX, XXX, XXX−XXX

The Journal of Organic Chemistry
Featured Article
32
Figure 2. Cleavage assay of 5′ AODN by various BER proteins. 500 nM ternary complex containing a 3′- P-labeled AODN were incubated with 100
nM DNA repair enzyme at 37 °C for 30 min. Lanes: T, no enzyme; NaOH, treated with NaOH; 1, incubated in the presence of hOGG1; 2, Fpg; 3,
Nth; 4, MUG; 5, TagI; 6, ANPG 40; 7, ANPG 80. The products of the reaction were analyzed as described under Experimental Section.
Oxidation of the 5′-Alcohol of CPG-Linked 5′-ODN/ORN to
CPG-Linked 5′-AODN/5′-AORN. CPG-linked ODNs were oxidized
In method C or D, the sample was diluted with water to a final
volume of 1.5 mL and purified by size exclusion column (NAP 25)
before RP-HPLC analysis and purification.
RP-HPLC analyses were achieved by using C18 column (length 250
mm, diam 4.6 mm; particle size 5 μm, porosity 100 Å) at 1 mL/min
flow. Eluent A was a 50 mM TEAA solution pH = 7 in water/MeCN
11
according to a published protocol. Briefly, CPG-linked ODN were
gently stirred in a diisopropylcarbodiimide/dichloroacetic acid/DMSO
(
19:1.7:200; v/v) mixture for 30 min at rt.
Protection from CPG-Linked 5′-AODN/5′-AORN to DPI-
Protected CPG-Linked 5′-AODN/5′-AORN. Supported 5′-
AODN/5′-AORN was protected either by (i) PTSA or (ii) DDQ.
PTSA (Method A). PTSA solution was prepared by dissolving 2 mg
p-toluenesulfonic acid (10 μmol) in 1 mL of DMF. CPG-linked 5′-
AODN (ca. 4 mg, initial loading 30−40 μmol/g) was reacted with 200
μL of a N,N′-diphenylethylendiamine solution (23 mg; 110 μmol in 1
mL of the aforementioned PTSA in DMF solution, or 32 mg (87
μmol) of 15 or 30.4 mg (87 μmol) of 17). The suspension was
vortexed and centrifuged to ensure a complete wetting of the resin.
The mixture was left overnight at rt without stirring. The colorless
supernatant was discarded, and the resin was thoroughly washed by
(
95:5, v/v) and eluent B a MeCN/water (9:1, v/v) solution.
Absorbance was monitored at 260 nm. For purification, a 10 mm
diameter column was used with a 4 mL/min flow. Protected 5′-AODN
and 5′-AORN were analyzed and purified using the following gradient:
0
% B during 2 min, then 0% to 45% B in 23 min.
Purified DPI-protected 5′-AODN/5′-AORN fully decomposes in a
few weeks (deprotection of the 5′-aldehyde) even stored at −20 °C.
Deprotection of DPI-Protected 5′-AODN into 5′-AODN. DPI-
protected 5′-AODN was treated with 400 μL 80% aqueous AcOH for
5
min at rt. 5′-AODN solution was diluted with 400 μL water and was
concentrated to 400 μL final volume in a speedvac evaporator (it is
MeCN (2 × 1 mL) and by Et O (2 × 1 mL). CPG-linked protected
2
advisible to not evaporate sample to dryness since heat released by the
5
′-AODN was then air-dried.
11
apparatus favors the β-elimination reaction ). Water was then added
DDQ (Method B). CPG-linked 5′-AODN (ca. 4 mg) was reacted
(
(
ca. 1 mL), and the 5′-AODN was purified on steric exclusion column
with 200 μL of a N,N′-diphenylethylendiamine solution (23 mg, 110
NAP 25) and lyophilized. 5′-AODN was stored at −20 °C.
μmol in 1 mL 96% EtOH solution). Then, 17 μL of a DDQ solution
Lyophilized 5′-AODN slowly decompose even stored at −80 °C
upon few weeks and should be preferably keep in 0.1% AcOH
solution.
(
26 mg, 114 μmol in 1 mL 96% EtOH solution) were added. The
initially yellow solution instantaneously turned to dark blue and faded
to greenish yellow. The suspension was vortexed and centrifuged to
ensure a complete wetting of the resin. The mixture was left overnight
at rt without stirring. The dark yellow supernatant was discarded and
the resin was thoroughly washed by MeCN until removal of a dark
Derivatization of 5′-AODN to Their n-Decyloxime Ether. 5′-
AODNs were analyzed by ESI-MS as n-decyloxime ether derivative
obtained by reacting with n-decyloxiamine synthesized as previously
1
8
described. 5′-AODN (ca. 50 nmol) was dissolved in 50 μL 0.4 M
powder formed during the reaction and by Et O (2 × 1 mL). CPG-
2
AcONH buffer (pH = 4.6) and 25 μL of a solution of n-decyloxiamine
4
linked protected 5′-AODN was then air-dried.
(10 equiv) in MeCN were added and stirred overnight. Resulting
Cleavage of CPG-Linked DPI-Protected 5′-AODN/5′-AORN
to DPI-Protected 5′-AODN/5′-AORN. DPI group is very acid
sensitive, and it should be noted that evaporation to dryness or
prolonged heating can lead to partial or full deprotection of the DPI-
protected 5′-AODN. In all cases, pH have to be maintained above 8
conjugates were purified by RP-HPLC
Synthesis of Substituted N,N′-Diphenylethylendiamine.
N,N′-Bis(4-bromophenyl)ethylenediamine (15). N,N′-Diphenylethy-
lenediamine (500 mg; 2.36 mmol) was dissolved in 20 mL of CHCl .
3
Tetraethylammonium tribromide (2.27 g; 4.72 mmol; 2 equiv) was
then added portionwise. After ca. 20 min, crude reaction mixture was
evaporated to dryness. The resulting residue was purified by silica gel
column chromatography (CH Cl /pentane mixture, from 1:2 to 1:1,
before NH OH or methylamine evaporation. It is advisible to add a
4
slightly volatile base, such as diisopropylethylamine (DIEA).
Final deprotection was performed as follows. (i) For ODN: dried
resin was treated with 30% aqueous NH OH (55 °C, 16 h) or 40%
2
2
4
1
3
then CH Cl ) to afford a pale yellow powder (585 mg, 67%):
NMR (75 MHz, CDCl ) 146.9; 132.1; 114.7; 109.6; 43.2; H NMR
C
aqueous methylamine (MA, 65 °C, 15 min). Ammoniac or MA was
removed with a speedvac evaporator; DPI-protected 5′-AODN were
then purified by RP-HPLC. (ii) For ORN, following method C: by
treatment with 500 μL 40% methylamine solution at 65 °C for 15 min.
Sample was cooled down and placed in a speedvac unit until complete
evaporation of methylamine. Sample was next frozen and lyophilized.
ODNs were taken up in 115 μL DMSO and treated with 60 μL TEA
and then by 75 μL TEA·3HF. The solution was heated for 90 min at
2
2
1
3
(
300 MHz, CDCl ) 7.16 (d, 2H, J = 8.9 Hz, H ); 6.40 (d, 2H, J = 8.9
3
A
r
Hz, H ); 3.58 (br, 2H, NH); 3.24 (s, 4H, CH ); MS (IE+, CH Cl )
Ar
2
2
2
+
m/z 369.0 (51%), [M + H] ; 371.0 (100%) ; 373.0 (49%); mp 110.0
C ± 0.1 °C (lit. mp 108 °C).
N,N′-Bis(4-cyanophenyl)ethylenediamine (16). Ethylenediamine
167 μL; 2.50 mmol), 4-fluorobenzonitrile (667 mg; 5.51 mmol; 2.2
1
9
°
20
(
6
5 °C. (iii) For ORN, following method D: by treating with 200 μL of
equiv) and potassium carbonate (690 mg; 5.00 mmol; 2 equiv) were
refluxed in 2.5 mL of dimethylacetamide overnight. The crude reaction
mixture was then cooled to room temperature, diluted with 10 mL
a 40% methylamine solution at 65 °C for ca. 10 min. Sample was
cooled down, centrifuged and the supernatant was transferred in a
plastic vial. 600 μL DMSO were added and the resulting solution was
cooled to −10 °C before addition of 100 μL of 70% aqueous glycolic
AcOEt/THF (1:1) and washed by a 5% NaHCO
water. The organic phase was dried over Na SO
solution, brine and
and evaporated to
4
3
2
acid solution. Solution was vortexed, 200 μL of NH F solution (40%
dryness. The resulting oil was purified by silica gel column
chromatography (cyclohexane/AcOEt mixture, from 5:1 to 1:1) to
afford a white powder (420 mg, 64%): 13C NMR (75 MHz, DMSO-
d6) 152.0; 133.4; 120.5; 111.8; 95.8; 41.1; 1H NMR (300 MHz,
DMSO-d ) 7.50 (d, 2H, J = 8.9 Hz, H ); 5.80 (br, 2H, NH); 6.71 (d,
4
aqueous) were added and solution was vortexed again. To avoid
aldehyde deprotection or phosphate migration, the pH value was kept
between 8 and 9 (tested on pH paper) by adding either aqueous
methylamine solution or glycolic acid solution. Sample was heated at
6
Ar
6
5 °C over 1 h.
2H, J = 8.9 Hz, H ); 3.35 (br, 4H, CH ); MS (IE+, CH Cl ) m/z
Ar 2 2 2
E
dx.doi.org/10.1021/jo502170e | J. Org. Chem. XXXX, XXX, XXX−XXX

The Journal of Organic Chemistry
Featured Article
+
19
2
63.1 (100%), [M + H] ; mp 198.5 °C ± 1.3 °C (lit. mp 205−206
ACKNOWLEDGMENTS
This work has been partially funded by the labex ARCANE
ANR-11-LABX-0003-01). The Nanobio-ICMG platforms
PSB, PSM, FR 2603) are acknowledged for providing analysis,
synthesis and purification facilities. Authors thank R. Gueret
́
and L. Fort for running ESI-MS analysis.
■
(
(
°C).
1
3
N,N′-Bis(3,5-dichlorophenyl)ethylenediamine (17). 1,2-Dibro-
moethane (1.05 g; 5.59 mmol) and 3,5-dichloroaniline (5.5 g; 33.9
mmol; 6 equiv) were dissolved in 5 mL of dimethylacetamide and
refluxed for 10 h. The crude reaction mixture was evaporated to
dryness, taken up in 20 mL of AcOEt and washed by a 5% NaHCO3
solution, brine and water. The organic phase was dried over Na SO4
2
REFERENCES
and evaporated to dryness. The resulting oil was purified by silica gel
column chromatography (cyclohexane/AcOEt 9:1) to afford a yellow
■
(1) (a) Pitie,
́
M.; Pratviel, G. Chem. Rev. 2010, 110, 1018−1059.
1
3
powder (948 mg, 48%): C NMR (75 MHz, DMSO-d ) 150.8; 134.4;
(b) Pogozelski, W. K.; Tullius, T. D. Chem. Rev. 1998, 98, 1089−1108.
(2) (a) Lahoud, G. A.; Hitt, A. L.; Bryant-Friedrich, A. C. Chem. Res.
Toxicol. 2006, 19, 1630−1636. (b) Dizdaroglu, M. Cancer Lett. 2012,
327, 26−47.
6
1
1
14.3; 110.0; 41.4; H NMR (300 MHz, DMSO-d ) 6.56 (br, 6H,
6
H ); 6.30 (br, 2H, NH); 3.20 (br, 4H, CH ); MS (IE+, CH Cl ) m/z
Ar
2
2
2
+
3
°
48.0 (78%), [M + H] ; 350.0 (100%); 352.0 (48%); mp 125.8 ± 1.0
13
C (lit. mp 127−129 °C).
(3) (a) Kappen, L. S.; Goldberg, I. H.; Liesch, J. M. Proc. Nat. Acad.
Sci. U. S. A. 1982, 79, 744−748. (b) Christner, D. F.; Frank, B. L.;
Kozarich, J. W.; Stubbe, J.; Golik, J.; Doyle, T. W.; Rosenberg, I. E.;
Krishnan, B. J. Am. Chem. Soc. 1992, 114, 8763−8767.
Preparation of Ternary Complex Containing 5′AODN.
5
3
AODN fragment ( ′CCACGCATCGCTGGTA ′) was 3′-end labeled
by Terminal Deoxynucleotidyl Transferase (Sigma-Aldrich, France) in
3
2
(4) Sugiyama, H.; Fujiwara, T.; Kawabata, H.; Yoda, N.; Hirayama,
N.; Saito, I. J. Am. Chem. Soc. 1992, 114, 5573−5578.
the presence of [α- P]ATP (3000 Ci/mmol, PerkinElmer). The
radiolabeled oligonucleotide was hybridized with 10% molar excess of
the template strand and of the flanking oligonucleotide. Hybridization
was conducted at 90 °C (5 min) followed by slow cooling to room
temperature in the enzyme incubation buffer. For most enzymes, the
buffer was 70 mM Hepes-KOH (pH 7), 100 mM KCl, 1 mM EDTA
except for hOGG1 (commercial buffer: 10 mM Tris-HCl (pH 7.9), 50
(
́
5) (a) Pitie, M.; Pratviel, G.; Bernadou, J.; Meunier, B. Proc. Natl.
Acad. Sci. U. S. A. 1992, 89, 3967−3971. (b) Zaidi, R.; Bryant-
Friedrich, A. C. Radiat. Res. 2012, 177, 565−572.
(
4
(
6) Pitie,
856−4864.
7) Chan, W.; Chen, B.; Wang, L.; Taghizadeh, K.; Demott, M. S.;
Dedon, P. C. J. Am. Chem. Soc. 2010, 132, 6145−6153.
8) Jung, K.-Y.; Kodama, T.; Greenberg, M. M. Biochemistry 2011,
0, 6273−6279.
9) Kodama, T.; Greenberg, M. M. J. Org. Chem. 2005, 70, 9916−
924.
10) (a) Wang, G.; Middleton, P. J. Tetrahedron Lett. 1996, 37,
739−2742. (b) O-Yang, C.; Wu, H. Y.; Fraser-Smith, E. B.; Walker,
K. A. M. Tetrahedron Lett. 1992, 33, 37−40.
11) Sallamand, C.; Miscioscia, A.; Lartia, R.; Defrancq, E. Org. Lett.
012, 14, 2030−2033.
12) Matsuda, A.; Inada, M.; Nara, H.; Ohtsuka, E.; Ono, A. Bioorg.
́
M.; Burrows, C. J.; Meunier, B. Nucleic Acids Res. 2000, 28,
mM NaCl, 10 mM MgCl , 1 mM DTT).
2
Enzymes. hOGG1 (human 8-oxoG-DNA glycosylase) was from
New England Biolabs; TagI (3-methyladenine-DNA glycosylase I),
MUG (mismatch-specific uracil-DNA glycosylase), Fpg (formamido-
pyrimidine-DNA glycosylase), Nth (endonuclease III) and ANPG 40
and 80 proteins (human truncated alkyl-N-purine-DNA glycosylase
missing 63 amino acids at C terminal end and 73 amino acids at N
terminal end respectively) were generous gifts from M. Saparbaev
(
5
(
9
(
2
(
Institut Gustave Roussy, Villejuif, France)
(
2
(
Cleavage Assay. Enzymatic activities were measured in the
32
appropriate incubation buffer. Typically, 10 nM 3′- P-labeled of the
ternary complex oligonucleotides were incubated with 10 nM DNA
repair enzyme at 37 °C for 30 min (10 μL). The reaction was stopped
by adding 10 μL of loading buffer (formamide containing 0.1% TFA)
and analyzed by 20% denaturing PAGE. Strand 1 (complementary
Med. Chem. Lett. 1993, 3, 2751−2754.
(13) Ono, A.; Okamoto, T.; Inada, M.; Nara, H.; Matsuda, A. Chem.
Pharm. Bull. 1994, 42, 2231−2237.
(14) Vanden Eynde, J. J.; Delfosse, F.; Lor, P.; van Haverbeke, Y.
Tetrahedron 1995, 51, 5813−5818.
5
strand): ′TTA CCA GCG ATG CGT GGG AGC GTG AAT TCA
3
5
3
TC ′; strand 2:3′ phosphate ODN: ′GAT GAA TTC ACG CTCp ′;
(15) Angeloff, A.; Dubey, I.; Pratviel, G.; Bernadou, J.; Meunier, B.
Chem. Res. Toxicol. 2001, 14, 1413−1420.
5
3
strand 3:5′ AODN: ′CHO−CCA CGC ATC GCT GGT A ′.
Analysis of radiolabeled oligonucleotides was carried out using a
Typhoon 9410 Phosphorimager and ImageQuant TL software.
Experiments were carried out at least three times.
(16) Zewge, D.; Gosselin, F.; Sidler, R.; DiMichele, L.; Cvetovich, R.
J. J. Org. Chem. 2010, 75, 5305−5307.
(17) Pratviel, G.; Duarte, V.; Bernadou, J.; Meunier, B. J. Am. Chem.
Soc. 1993, 115, 7939−7943.
To attempt to trap DNA−protein complexes, the reaction was
performed in the same conditions, in the presence of triacetoxybor-
ohydride (50 mM final concentration), quenched with loading buffer,
and analyzed as described above.
(
18) Ewan, H. S.; Muli, C. S.; Touba, S.; Bellinghiere, A. T.;
Veitschegger, A. M.; Smith, T. B.; Pistel, W. L., II; Jewell, W. T.; Rowe,
R. K.; Hagen, J. P.; Palandoken, H. Tetrahedron Lett. 2014, 55, 4962−
965.
4
(
19) Berg, S. S. J. Chem. Soc. 1960, 5172−5176.
(20) Vamecq, J.; Maurois, P.; Pages, N.; Bac, P.; Stables, J. P.;
Gressens, P.; Stanicki, D.; Vanden Eynde, J. J. Eur. J. Med. Chem. 2010,
5, 3101−3110.
ASSOCIATED CONTENT
■
*
S
Supporting Information
4
1
H and 13C NMR spectra of compounds 15−17, RP-HPLC
chromatograms of 5′-AODN 4−14 both protected and
derivatized and 5′-AORN 18a−22b and ESI-MS analysis of
5
′-AODN 4−25 are provided. This material is available free of
charge via the Internet at http://pubs.acs.org.
AUTHOR INFORMATION
■
Corresponding Author
*E-mail: remy.lartia@ujf-grenoble.fr.
Notes
The authors declare no competing financial interest.
F
dx.doi.org/10.1021/jo502170e | J. Org. Chem. XXXX, XXX, XXX−XXX