New strategy for the synthesis of 3′,5′-bifunctionalized oligonucleotide conjugates through sequential formation of chemoselective oxime bonds

Article abstract of DOI:10.1002/chem.200400390

A convenient strategy for the synthesis of bifunctionalized oligonucleotide conjugates bearing two different reporters at the 3′ and 5′ ends of the oligonucleotide is presented. The method involves the preparation of oligonucleotides bearing an aldehyde a

Full text of DOI:10.1002/chem.200400390

FULL PAPER
New Strategy for the Synthesis of 3’,5’-Bifunctionalized Oligonucleotide
Conjugates through Sequential Formation of Chemoselective Oxime Bonds
Om Prakash Edupuganti, Yashveer Singh, Eric Defrancq,* and Pascal Dumy[a]
Abstract: Aconvenient strategy for the synthesis of bifunctionalized oligonucleo-
tide conjugates bearing two different reporters at the 3’ and 5’ ends of the oligonu-
cleotide is presented. The method involves the preparation of oligonucleotides
bearing an aldehyde and/or aminooxy functionality at each end, followed by reac-
tion to form oxime bonds with appropriately functionalized reporters. The conju-
gation reactions are carried out under mild aqueous conditions with good reaction
yield.
Keywords: chemoselectivity · con-
jugation · DNA · oligonucleotides ·
oxime · peptides
Introduction
proaches. The first involves the total synthesis of peptide
and oligonucleotide fragments on the same solid support.
The synthesis of peptide is followed by that of oligonucleo-
tide,[14] or vice versa,[15] and in many cases a branched linker
may be attached to the support to permit the independent
growth of the peptide and oligonucleotide fragments.[15] The
method suffers from the poor compatibility of the peptide
and oligonucleotide chemistries and has not found much
favour in spite of some promising advances in this area.[16]
The second approach (the fragment approach) involves the
separate solid phase assembly of peptide and oligonucleo-
tide, deprotection and purification (when necessary), and
the coupling of the two fragments in solution phase. This is
done by introducing mutually reactive groups into each frag-
ment during or after the solid phase synthesis and involves
the formation of linkages like disulfide,[17] thioether,[18]
amide,[19] maleimide,[20] oxime,[21] thiazolidine,[21a,b] and hy-
drazone.[21b,22] This fragment approach is more frequently
used on account of excellent coupling efficiencies and ease
of purification. The method also offers an opportunity to
ligate a number of other reporter molecules (carbohydrates,
fluorophores, etc.) onto the oligonucleotide.
The use of synthetic oligonucleotides for the specific inhibi-
tion of gene expression represents an attractive therapeutic
approach for the treatment of cancer and various other viral
diseases.[1–3] These oligonucleotides may cause selective in-
hibition of gene expression either by targeting the mRNA
[4,5]
by an antisense or siRNAmechanism,
double stranded
[6]
DNAby triplex formation
or proteins by aptamer selec-
tion.[7] These strategies suffer, however, from the poor cell
penetration and cellular targeting of these agents and sensi-
tivity towards nuclease activity.[8] Avery promising approach
to overcome the problem of poor uptake is to attach these
oligonucleotides to a variety of available cell penetrating
and localizing peptides.[9] It has been reported that such pep-
tide–oligonucleotide conjugates show enhanced cell specific
targeting, cellular uptake efficiency and stability to degrada-
tion in comparison to unmodified oligonucleotides.[9–11] Be-
sides, in many cases, improved binding strength with target
sequence has been observed.[12] Similarly, conjugation with
carbohydrates has also been explored to enhance cell target-
ing via lectin recognition.[13]
In this context, the chemical synthesis of peptide–oligonu-
cleotide conjugates (POCs) has generated considerable in-
terest. Different methods for the preparation of POCs have
been described and can be categorized into two broad ap-
Complementary to this monoconjugation, it would be of
great interest to anchor two different reporter groups to the
oligonucleotides. For instance, attachment of two peptides
with different properties may improve the biological activity
of the oligonucleotide (for example, enhancement of cell
penetration and targeting, increasing of the concentration
into the nucleus or stabilization of the duplex) by using a
specific peptide according to the property desired. Similarly,
bisconjugates could combine a fluorescent moiety with a
peptide residue, the fluorophore serving as a probe to detect
the transport of the peptide–oligonucleotide conjugate. In
this context, the preparation of phosphorothioate oligonu-
[a] Dr. O. P. Edupuganti, Dr. Y. Singh, Dr. E. Defrancq, Prof. P. Dumy
LEDSS, UMR CNRS 5616, ICMG FR2607
UniversitØ Joseph Fourier
BP 53, 38041 Grenoble Cedex 9 (France)
Fax : (+33)476514946
E-mail: Eric.Defrancq@ujf-grenoble.fr
Supporting information for this article is available on the WWW
under http://www.chemeurj.org/ or from the author.
5988
ꢀ 2004 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
DOI: 10.1002/chem.200400390
Chem. Eur. J. 2004, 10, 5988 – 5995

5988 – 5995
cleotides substituted with a 5’-protected thiol function and a
3’-amino group has been described. The 3’-amino functional-
ity was used to ligate a peptide (NLS) and the 5’-thiol to
anchor a fluorescent reporter group.[23a] 3’,5’-Bisconjugates
have also been prepared by on-column derivatization using
photolabile solid phase synthe-
oligonucleotide conjugates bearing two different reporter
groups, both with two different peptides and with a peptide
and a fluorescent probe. The method is based on the se-
quential formation of chemoselective oxime bonds using
two different strategies (Figure 1). The first involves the se-
sis supports.[23d] To our knowl-
edge, little has been reported
on the synthesis of oligonu-
cleotides bearing two different
peptides at the 3’ and 5’ ends.
Therefore, the effort to devel-
op suitable synthetic methods
for the preparation of such bi-
functionalized peptide–oligo-
nucleotide conjugates should
be of great interest.
Earlier work from our labo-
ratory has focussed on the de-
velopment of oxime and thia-
zolidine linkages for the effi-
cient preparation of peptide–
oligonucleotide conjugates. It
Figure 1. General strategy for the preparation of 3’,5’-bisfunctionalized oligonucleotides.
has been shown that chemose-
lective oxime and thiazolidine
quential generation of an aldehyde function on both ends of
the oligonucleotide and subsequent conjugation reaction
with the reporter group containing an aminooxy moiety
(Figure 1a). The second requires the sequential generation
of an aldehyde and an aminooxy function on the oligonu-
cleotide followed by conjugation with appropriately derivat-
ized reporters (Figure 1b).
The relevance of the work has been emphasized by using
two different peptides of biological significance (Figure 2).
Firstly, a cyclopentapeptide consisting of an arginine-gly-
cine-aspartic acid (RGD) motif known for selectivity to-
wards the avb3 integrin receptor.[27] The RGD peptide has
also been studied for tumour targeting[27b] and DNAdeliv-
ery.[27c,d] Secondly, the NLS peptide, a nuclear localizing
signal sequence with the basic peptide APKKKRKVED de-
rived from the simian virus 40 antigen. The oligonucleotide
conjugate with this sequence has been reported and affinity
for the target sequence has been studied.[12b] Acytomegalo-
virus luciferase gene bearing this sequence has also been
studied for transfection[28a] and the ability of this sequence
for non-viral gene delivery has been thoroughly investigated
recently.[28b] The fluorescein derivative 10 has been used as
fluorescent reporter group.
linkages can be successfully employed to prepare peptide–
oligonucleotide conjugates bearing peptides at either the 3’
or 5’ terminus of the oligonucleotide.[21,24] The methodology
has been further explored by our group for the labelling of
oligonucleotides and RNA[25] and for anchoring the oligonu-
cleotide on glass support.[26] Earlier results from our group
and others have shown that oxime and thiazolidine bonds
do have certain advantages over other types of linkages. For
instance, the oxime bonds give high efficiency of coupling,
require the use of no activation or stabilization steps and do
not suffer from non-regiospecific ligation as is the case with
thio or amine based ligation.[21a,b] Since the oxime ligation is
carried out at slightly acidic pH at which the free amino
groups in peptides are protonated, it also helps to solubilize
the peptide in water either alone or with a cosolvent.[21b]
In a recent work, we have described the preparation of
3’,5’-bisfunctionalized oligonucleotide conjugates bearing
same groups at both the termini.[21d] The method involves si-
multaneous conjugation at the two extremities of the oligo-
nucleotide functionalized with aldehyde groups. However,
the procedure has only limited applicability as it does not
permit the anchoring of different groups at the two extremi-
ties of the oligonucleotide. The
preparation of such conjugates
would require the use of differ-
ent
orthogonal
protecting
groups so that the generation
of the aldehyde function and
subsequent conjugation at the
two termini of the oligonucleo-
tide could be carried out in a
sequential fashion. Herein, we
present a new method for the
preparation of 3’,5’-derivatized
Figure 2. Structure of the different peptides P1–P4 and fluorescein derivative 10.
www.chemeurj.org ꢀ 2004 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Chem. Eur. J. 2004, 10, 5988 – 5995
5989

E. Defrancq et al.
FULL PAPER
Results and Discussion
linker 3 carrying the protected diol. It should be noted that
this linker would also aid in oligonucleotide purification be-
cause of its hydrophobic nature.
Bisconjugation via 3’,5’-aldehyde functionality (Scheme 1):
This strategy consists of a first conjugation reaction with the
aldehyde at the 3’ end followed by the liberation of the
second aldehyde moiety at the 5’ end and a subsequent
second conjugation. The generation of aldehyde function at
the 3’ end requires the modification of the solid support.
Since depurination is observed during the acidic deprotec-
tion a post oxidation strategy is preferable for the introduc-
tion of the aldehyde moiety at the 3’ end. In the present
work we have used the commercially available solid support
1 (3-[(4,4’-dimethoxytrityl) glyceryl-1-succinyl] long chain al-
kylamino controlled pore glass, CPG) to introduce the alde-
hyde function. The preparation of oligonucleotides bearing
an aldehyde function at the 5’ end has been described earli-
er by using a modified phosphoramidite linker carrying a
masked 1,2-diol group.[21a] The aldehyde was generated by
mild oxidation with sodium periodate. However, the meth-
oxybenzylidene acetal protection of the 5’-diol linker was
found unsuitable for our strategy. In fact, cleavage of the 5’-
diol protecting group occurred both during the purification
step and the conjugation reaction. The benzylidene acetal
protection was chosen, therefore, as it is reported to be
more stable. The phosphoramidite linker 3 carrying a pro-
tected 1,2-diol was prepared in two steps starting from com-
mercially available 1,2,6-hexanetriol. The 1,2-diol was first
protected as benzylidene acetal 2 using benzaldehyde di-
methyl acetal in the presence of a catalytic amount of PPTS.
Phosphitylation with 2-cyanoethyl diisopropylchlorophos-
phoramidite afforded the desired modified phosphoramidite
The 3’,5’-bifunctionalized undecamer d(5’XCGCACA-
CACGCY3’), in which X represents the 5’-diol linker and Y
the 3’-diol linker, was prepared by automated DNAsynthe-
sis, according to standard b-cyanoethylphosphoramidite
chemistry by using the aforementioned support 1. The phos-
phoramidite 3 was incorporated during the last step of the
automated DNAsynthesis. After cleavage from the solid
support and deprotection of the nucleobases by usual am-
monia treatment (28% ammonia, 16 h at 558C), the oligo-
nucleotide 4 carrying the free 3’-diol and the protected 5’-
diol linker was purified by reverse phase HPLC. Oligonu-
cleotide 4 was obtained as a single peak by reverse phase
HPLC (Figure 3a). No hydrolysis of the 5’-benzylidene pro-
tected group was observed (which would give an oligonu-
cleotide bearing a diol moiety at both ends, unsuitable for
subsequent monoconjugation). Further oxidation with
sodium periodate in water gave the desired 3’-aldehyde oli-
gonucleotide 5 in 70% isolated yield. The 3’-conjugation re-
action was then performed with the RGD peptide P1 bear-
ing an aminooxy function (Figure 2). Reaction was carried
out in ammonium acetate buffer at slightly acidic pH (4.6)
and was monitored by HPLC. The reaction proceeded to
completion within 4 h to yield the 3’-RGD conjugate 6 as
the major product (Figure 3b). No cleavage of the 5’-pro-
tecting group was observed in the slightly acidic conditions
used for the coupling reaction. The 3’-RGD,5’-protected diol
conjugate 6 was obtained after HPLC purification in 58%
isolated yield. The conjugate 6 was then treated with 80%
Scheme 1. Preparation of conjugates 9 and 11. a) benzaldehyde dimethylacetal, PPTS; b) 2-cyanoethyldiisopropylchloro phosphoramidite, DIEA; c) auto-
mated oligonucleotide synthesis; d) NaIO4; e) RGD peptide P1; f) 80% aqueous AcOH; g) NLS peptide P2; h) fluorescein derivative 10. PPTS=pyridi-
nium paratoluenesulfonate; DIEA= diisopropylethylamine.
5990
ꢀ 2004 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.chemeurj.org
Chem. Eur. J. 2004, 10, 5988 – 5995

Peptide–Oligonucleotide Conjugates
5988 – 5995
it should be noted that the
peptide used for the 5’-conju-
gation contains a glyoxylic al-
dehyde group proceeding from
the oxidation of a 1,2-aminoal-
cohol residue of a serine. The
formed oxime is thus a glyoxyl-
ic oxime which has been re-
ported to be more stable than
an aliphatic one.[29]
Introduction of the 3’ linker
was achieved using the same
strategy as above with the glyc-
eryl support 1. Anchoring the
aminooxy moiety at the 5’ end
using the phosphoramidite 12,
in which the aminooxy moiety
is protected as a trityl, has
been reported already.[21a] The
trityl protection has the ad-
vantage that it can be removed
Figure 3. HPLC profiles (detection at 260 nm): a) purified undecamer 3’-diol,5’-protected diol 4; b) crude reac-
tion mixture of 3’-aldehyde containing undecamer 5 with RGD peptide P1 (1st conjugation); c) crude reaction
mixture of deprotection of 6; d) crude reaction mixture of 5’-aldehyde containing undecamer 8 with NLS pep-
tide P2 (2nd conjugation). For the HPLC conditions, see Experimental Section.
aqueous acetic acid solution for 1 h to remove the 5’-pro-
tecting group and give the 3’-RGD,5’-diol conjugate 7. It is
important to note that in these slightly acidic conditions,
which are commonly employed for 5’-detritylation, no hy-
drolysis of the oxime bond at the 3’-end was observed (Fig-
ure 3c shows the crude deprotection mixture which reveals
the presence of a single major peak). Another oxidation
step using sodium periodate in water gave the 3’-RGD,5’-
CHO conjugate 8. No degradation of the oxime bond was
noted during this oxidative treatment. The second conjuga-
tion reaction was carried out with the NLS peptide P2 carry-
ing an aminooxy function (Figure 2) under similar condi-
tions as for RGD incorporation at the 3’ end (i.e., ammoni-
um acetate buffer at pH 4.6). The course of the reaction was
followed by reverse phase HPLC and exclusive formation of
the desired 3’-RGD,5’-NLS conjugate 9 was observed (Fig-
ure 3d). The same protocol was then applied with the ami-
nooxy containing fluorescein derivative 10 leading to the 3’-
RGD,5’-fluorescein conjugate 11. The conjugates 9 and 11
were obtained after purification by HPLC in 55–60% isolat-
ed yields. All the oligonucleotide derivatives prepared
herein were characterized by ES-MS analysis (see Support-
ing Information), which showed an excellent agreement be-
tween the experimentally determined molecular weights and
the calculated values (Table 1).
Table 1. ES-MS analysis.[a]
Oligonucleotide
Mcalcd
Mfound
3’-diol,5’-protected diol 4
3’-CHO,5’-protected diol 5
3’-RGD,5’-protected diol 6
3’-RGD,5’-diol 7
3’-RGD,5’-CHO 8
3’-RGD,5’-NLS 9
3’-RGD,5’-fluorescein 11
3’-diol,5’-protected aminoxy 13
3’-CHO,5’-protected aminoxy 14
3’-NLS,5’-protected aminoxy 15
3’-NLS,5’-RGD 16
3707.5
3676.4
4351.1
4263.0
4230.9
5483.4
4749.5
3862.2
3830.2
5082.6
5482.3
4854.2
4261.2
4822.2
4229.2
5341.3
3708.8
3676.3
4351.1
4262.7
4231.0
5483.6
4750.2
3861.9
3829.1
5081.8
5481.0
4853.2
4260.0
4822.3
4229.1
5340.9
3’-diol,5’-NLS 17a
3’-diol,5’-RGD 17b
3’-CHO,5’-NLS 18a
3’-CHO,5’-RGD 18b
3’-fluorescein,5’-NLS 19
[a] The analysis was done in negative mode. The oligonucleotides and
conjugates were dissolved in CH3CN/H2O/NEt3 50:50:2 (v/v/v). Eluent:
CH3CN/H2O 50:50 (v/v).
under the same conditions as the classical DMT group used
in DNAsynthesis. Furthermore, the purifications by reverse
phase HPLC of crude oligonucleotides were facilitated by
the hydrophobic properties of the trityl group. The 3’,5’-bi-
functionalized undecamer d(5’XCGCACACACGCY ), in
3’
which X represents the 5’-trityl protected aminooxy linker
and Y the 3’-diol linker, was prepared by automated DNA
synthesis according to standard b-cyanoethylphosphorami-
dite chemistry. After cleavage from the support and depro-
tection of bases using the standard protocol, the bifunction-
alized 3’-diol,5’-trityl protected aminooxy oligonucleotide 13
was purified by reverse phase HPLC (Figure 4a).
Bisconjugation via 3’,5’-hetero-bifunctionalized oligonucleo-
tide (Scheme 2): This bifunctionalization strategy offered
two ways to prepare 3’,5’-bisconjugates. One consists of
starting the conjugation with the aldehyde at the 3’ end fol-
lowed by the deprotection of the aminooxy group at the 5’
end and subsequent coupling with an aldehyde-containing
reporter. The second involves the “reverse way” where the
deprotection and coupling at the 5’ end is carried out first,
followed by the oxidation of the 3’-diol and subsequent cou-
pling with an aminooxy-containing reporter. In this strategy
Strategy starting with 3’ conjugation: The oxidative cleav-
age of the diol at the 3’ end of the oligonucleotide was car-
ried out under the same conditions as above, by using
Chem. Eur. J. 2004, 10, 5988 – 5995
www.chemeurj.org
ꢀ 2004 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
5991

E. Defrancq et al.
FULL PAPER
Scheme 2. Preparation of conjugates 16 and 19. a) automated oligonucleotide synthesis; b) NaIO4; c) NLS peptide P2; d) 80% aqueous AcOH, RGD
peptide P3; e) 80% aqueous AcOH, NLS peptide P4; f) fluorescein derivative 10.
NaIO4, leading to 3’-aldehyde-containing oligonucleotide 14.
The 3’-conjugation reaction was then performed with the
NLS peptide P2 in ammonium acetate buffer (pH 4.6).
Under these conditions the 5’-trityl protection was found to
be stable as confirmed by HPLC analysis (Figure 4b depicts
the crude conjugation mixture which reveals a single major
peak). The 3’-NLS,5’-trityl protected aminooxy conjugate 15
was obtained after purification by reverse phase HPLC in
50% yield. The 5’ conjugation was then achieved using a
one-pot trityl cleavage and coupling reaction. This proce-
dure has the advantage that it eliminates problems associat-
ed with the highly reactive aminooxy moiety. Reaction was
carried out in 80% aqueous acetic acid in the presence of
the RGD peptide P3 containing the aldehyde function
(Figure 2). The liberation of the aminooxy moiety at the 5’
end resulted in the immediate conjugation with the aldehy-
dic peptide P3 through the formation of the oxime bond.
The unprotected oligonucleotide intermediate could never
be observed. The course of the reaction was controlled by
careful monitoring on analytical HPLC and revealed some
cleavage of the oxime bond at the 3’ end (Figure 4c depicts
the HPLC profile of the crude reaction mixture which
shows the presence of less than 10% of the product result-
ing from cleavage of the 3’-oxime bond). Purification by re-
verse phase HPLC afforded the 3’-NLS,5’-RGD conjugate
16 in 50% isolated yield.
Figure 4. HPLC profiles (detection at 260 nm): a) purified undecamer 3’-
diol,5’-protected aminooxy 13; b) crude reaction mixture of 3’-aldehyde
Strategy starting with 5’ conjugation: In this case, the first
containing undecamer 14 with NLS peptide P2 (1st conjugation);
conjugation was carried out at the 5’ end of the oligonucleo-
c) crude reaction mixture of conjugate 15 with RGD peptide P3 (2nd
conjugation). For the HPLC conditions, see Experimental Section.
tide leading to the formation of a glyoxylic oxime. As this
5992
ꢀ 2004 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.chemeurj.org
Chem. Eur. J. 2004, 10, 5988 – 5995

Peptide–Oligonucleotide Conjugates
5988 – 5995
Table 2. Melting temperatures of the duplexes formed by hybridization
of the indicated oligonucleotides with the complementary strand
d(GCGTGTGTGCG).[a]
type of oxime has been described as more stable under
acidic conditions we presumed that cleavage during the
second conjugation should be less than that observed in the
case of the strategy starting with 3’ conjugation. The first
conjugation was performed using the same procedure as
above (i.e., one-pot detritylation/coupling). Reaction of the
3’-diol,5’-trityl protected aminooxy oligonucleotide 13 was
carried out in 80% aqueous acetic acid solution in the pres-
ence of the NLS peptide P4 containing the aldehyde moiety
at the N-terminus (Figure 2). Analysis by HPLC showed the
selective formation of the 3’-diol,5’-NLS conjugate 17a
which was purified by reverse phase HPLC (Figure 5a). The
ON
Tm [8C]
unmodified undecamer
3’-RGD,5’-NLS 9
3’-NLS,5’-RGD 16
3’-RGD,5’-fluorescein 11
3’-fluorescein,5’-NLS 19
59.0Æ1
62.0Æ1
62.0Æ1
61.0Æ1
63.0Æ1
[a] Measurements were performed in sodium phosphate buffer (10 mm)
containing EDTA(1 m m) and NaCl (100 mm) at pH 7.
tary strand d(GCGTGTGTGCG). The conjugates 11 and 19
containing the fluorescent label at the 5’ and 3’ ends, respec-
tively, were also studied. The melting temperatures of the
resulting duplexes were determined and the “natural” oligo-
nucleotide d(CGCACACACGC) without any reporter was
also studied for comparison. The melting temperature data
are collected in Table 2. It is evident that duplexes contain-
ing the modified oligonucleotides are slightly more stable
than the duplex containing the natural oligonucleotide. This
could be attributed to the positive charges of the lysine and
arginine side chains present within the peptides. The results
reported herein are consistent with the earlier reported ob-
servations and further emphasize the fact that bisconjuga-
tion does not induce any instability in the duplex.
Conclusion
Figure 5. HPLC profiles (detection at 260 nm): a) crude reaction mixture
of undecamer 3’-diol,5’-protected aminooxy 13 with the NLS peptide P4;
b) crude reaction mixture of conjugate 18a with fluorescein derivative
10.
In conclusion, we present a very convenient and facile strat-
egy to prepare bifunctionalized peptide–oligonucleotide
conjugates containing two different peptides at the 3’ and 5’
ends. The strategy is also applicable to the preparation of
peptide–oligonucleotide conjugates carrying a fluorescent
reporter group. The procedure involves sequential formation
of oxime bonds in a one step ligation that is compatible with
the use of unprotected peptide and oligonucleotide. The
method thus allows the rapid and efficient preparation of
POCs where the cell-penetrating and cell-targeting peptides
can be anchored on the same oligonucleotide. Moreover,
the bifunctionalization of oligonucleotides does not destabi-
lize the duplexes. The results open up new prospects for the
preparation and availability of a large variety of POCs for
molecular biology studies.
3’-diol was then oxidized under the mild NaIO4 conditions
described above leading to the 3’-CHO,5’-NLS conjugate
18a. Asecond conjugation was then performed with the flu-
orescein derivative 10 in ammonium acetate buffer at
pH 4.6. The course of reaction was monitored by HPLC and
reaction completion was achieved after 4 h (Figure 5b). No
traces of compound resulting from the cleavage of the 5’-
oxime could be detected. Purification by HPLC afforded
the desired 3’-fluorescein,5’-NLS conjugate 19. The 3’-
NLS,5’-RGD conjugate 16 was also prepared using the same
strategy starting from the 3’-diol,5’-trityl protected aminooxy
oligonucleotide 13. All the oligonucleotides and conjugates
13–19 were characterized by ES-MS analysis which showed
experimental molecular weights in excellent agreement with
the calculated values (Table 1).
Experimental Section
Materials and methods: All solvents and reagents used were of the high-
est purity available. 1,2,6-Hexanetriol, pyridinium paratoluenesulfonate
and sodium periodate were obtained from Aldrich and used without fur-
ther purification. The solid support 1 (3-[(4,4’-dimethoxytrityl) glyceryl-1-
succinyl] long chain alkylamino CPG) was purchased from Eurogentec.
The peptides P1–P4 were prepared by solid phase peptide synthesis ac-
cording to a reported method.[21a] The aminooxy fluorescein derivative
10[25b] and the phosphoramidite 12[21a] were prepared as previously descri-
bed. The oligonucleotides and conjugates were purified on a m-Bondapak
C-18 column (Macherey-Nagel Nucleosil: 10”250 mm, 7mm) using the
Hybridization properties of the POCs (Table 2): Hybridiza-
tion properties of the conjugates were studied by melting
temperature (Tm) measurements to evaluate the influence of
the anchoring of two reporter groups on the stability of the
duplex oligonucleotide. The conjugates 9 and 16 containing
the RGD and the NLS peptide at the 3’ and 5’ ends respec-
tively, or vice versa, were hybridized with their complemen-
Chem. Eur. J. 2004, 10, 5988 – 5995
www.chemeurj.org
ꢀ 2004 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
5993

E. Defrancq et al.
FULL PAPER
following solvent system: solvent A, 20mm ammonium acetate/CH3CN
95:5 (v/v); solvent B (CH3CN); flow rate, 4 mLminÀ1; a linear gradient
from 0 to 30% B in 20 min was applied. The purity of the product was
assessed on analytical column using the same gradient at a flow rate of
1 mLminÀ1. Mass Spectra were measured on a Delsi-Nermag R10-10 for
EI and DCI and on an Esquire 3000 (Bruker) for ESI. The analysis was
performed in the negative mode for the oligonucleotides and conjugates
using 50% aqueous acetonitrile as eluent. The oligonucleotides and con-
jugates were dissolved in 50% aqueous acetonitrile and 2% NEt3 was
added. 1H and 13C NMR spectra were recorded on Bruker AC 200 spec-
trometer.
3’-RGD,5’-NLS conjugate (9): An aqueous solution of NLS-peptide P2
(4 equiv) was added to
a solution of oligonucleotide 8 (0.11 mg,
0.026 mmol) in 0.1m ammonium acetate buffer (0.1 mL; pH 4.6). The re-
action mixture was stirred for 3 h at room temperature leading to the for-
mation of conjugate 9 which was purified by HPLC (0.08 mg, 55%).
3’-RGD,5’-fluorescein conjugate (11): Conjugation with fluorescein deriv-
ative 10 was achieved using the same protocol as for 9 and led to the con-
jugate 11 in 61% yield.
Bisconjugation via hetero-bifunctionalized oligonucleotide starting with
3’-conjugation
3’-Aldehyde,5’-trityl aminooxy containing oligonucleotide (14): The oxi-
dation of the 3’-diol was achieved using the same protocol as above for 5.
Starting from the 3’-diol oligonucleotide 13 (0.84 mg, 0.217 mmol), the
conjugate 14 was obtained (0.74 mg, 90%).
4-[2-Phenyl-(1,3)-dioxolan-4-yl]-butanol-1-ol
(2):
1,2,6-Hexanetriol
(3.68 g, 27.52 mmol) and a catalytic amount of PPTS (0.3 g) were added
to a solution of benzaldehyde dimethyl acetal (2.095 g, 13.76 mmol) in
dry DMF (30 mL). The reaction mixture was stirred for 24 h at 558C
under anhydrous conditions. The DMF was then removed under vacuum
and the oily residue obtained was dissolved in EtOAc. The organic layer
was washed successively with saturated aqueous NaHSO3, 10% aqueous
NaHCO3 and brine, and dried over anhydrous Na2SO4. The pure product
2 was obtained by purification (2.1 g, 72%) on silica column using
CH2Cl2. 1H NMR (CDCl3): d = 7.52–7.44 (m, 2H; Ar-H), 7.42–7.34 (m,
3H; Ar-H), 5.92 and 5.80 (2 s, 1H; O-CH-O), 4.28–4.08 (m, 2H; CH2O),
3.95–3.71 (m, 3H; CH-O and CH2O), 1.84–1.42 (m, 6H; 3CH2); MS
(DCI/NH3): m/z: 222.8 [M+H]+.
3’-NLS,5’-trityl aminooxy containing conjugate (15): Asolution of the
NLS peptide P2 in water (2 equiv, 0.5 mg) was added, to a solution of the
oligonucleotide 14 (0.74 mg, 0.193 mmol) in 0.1m ammonium acetate
buffer (0.7 mL, pH 4.6). The reaction mixture was stirred for 4 h at room
temperature affording the 3’-NLS conjugate 15 in 50% yield (0.47 mg)
after purification by HPLC.
3’-NLS,5’-RGD conjugate (16): The 3’-NLS conjugate 15 (0.45 mg,
0.088 mmol) was dissolved in an 80% aqueous AcOH solution (0.4 mL)
and the RGD peptide P3 in water (3 equiv, 0.175 mg) was added. The re-
action mixture was stirred overnight at room temperature and the acetic
acid was then lyophilized. The crude mixture was then purified by HPLC
to give the 3’-NLS,5’-RGD conjugate 16 (0.24 mg, 50%).
Phosphoramidite (3): DIEA(302 mg, 2.33 mmol) and 2-cyanoethyl diiso-
propylchlorophosphoramidite (359 mg, 1.5 mmol) were added under
argon to a solution of compound 2 (260 mg, 1.16 mmol) in anhydrous
CH2Cl2 (3 mL). The reaction mixture was stirred at room temperature
for 4 h and CH2Cl2 (20 mL) was then added. The organic layer was
washed successively with 10% aqueous NaHCO3, then with brine and
dried over anhydrous Na2SO4. The crude mixture was purified by silica
gel column chromatography (AcOEt/cyclohexane/Et3N 90:10:2) to give
compound 3 as a white oil (0.320 g, 65%). 1H NMR (CDCl3): d = 7.48–
7.44 (m, 2H; Ar-H), 7.37–7.35 (m, 3H; Ar-H), 5.91 and 5.79 (2 s, 1H; O-
CH-O), 4.27–4.07 (m, 2H; CH2O), 3.87–3.75 (m, 2H; CH2O), 3.70–3.44
(m, 5H; CH2O, 3CH), 2.62 (t, 2H; CH2CN), 1.70–1.45 (m, 6H; 3CH2),
1.18–1.15 (m, 12H; 4CH3); 31P NMR (CDCl3): d = 145.5; MS (DCI/
NH3): m/z: 422.8 [M+H]+.
Bisconjugation via hetero-bifunctionalized oligonucleotide starting with
5’-conjugation
3’-Diol,5’-NLS conjugate (17a): Oligonucleotide 13 (0.95 mg, 0.246 mmol)
was dissolved in an 80% aqueous AcOH solution (0.8 mL) and the NLS
peptide P4 (3 equiv, 0.93 mg) in aqueous solution was added. The reac-
tion mixture was stirred for 8 h at room temperature leading to the for-
mation of the 5’-NLS conjugate 17a, which was purified by HPLC (yield:
60%, 0.7 mg). The 3’-diol,5’-RGD conjugate 17b was obtained in the
same manner using RGD peptide P3 (yield: 57%).
3’-Aldehyde,5’-NLS conjugate (18a): Oxidation of the 3’-diol moiety was
achieved as described for 5 by using NaIO4. The conjugate 18a was ob-
tained in 81% yield after purification by HPLC. The 3’-aldehyde,5’-RGD
conjugate 18b was prepared using the same protocol with the undecamer
17b (yield: 74%).
Oligonucleotide synthesis: Automated DNA synthesis was carried out on
an Expedite DNAsynthesizer (Perkin–Elmer) using standard b-cya-
noethyl nucleoside phosphoramidite chemistry on a 1 mm scale with final
DMT on, using the modified solid support 1. The modified phosphorami-
dites 3 and 12 (0.1 gmLÀ1 in dry acetonitrile) were coupled to the support
bound oligonucleotide during the last step of the automated DNAsyn-
thesis using an extended coupling time (15 min). The oligonucleotides
were cleaved from the solid support by treatment with a 28% ammonia
solution for 2 h and finally deprotected by keeping the ammonia solution
at 558C for 16 h. The 5’-protected oligonucleotides carrying free 3’-diol 4
and 13 were purified by reverse phase HPLC and characterized by
ESMS analysis (Table 1).
3’-Fluorescein,5’-NLS conjugate (19): Asolution of the fluorescein deriv-
ative 10 (2 equiv, 0.06 mg) in DMF was added to a solution of the 3’-alde-
hyde,5’-NLS conjugate 18a (0.25 mg, 0.052 mmol) in 0.1m ammonium ace-
tate buffer (0.22 mL, pH 4.6). The mixture was stirred for 4 h at room
temperature and the crude product was purified by HPLC to afford the
conjugate 19 (0.11 mg, 40%).
3’-NLS,5’-RGD conjugate (16): An aqueous solution of NLS peptide P2
(2 equiv, 0.12 mg) was added to
a solution of oligonucleotide 18b
(0.2 mg, 0.047 mmol) in 0.1m ammonium acetate buffer (0.3 mL, pH 4.6).
The reaction mixture was stirred for 4 h at room temperature and the
conjugate 16 was then purified by HPLC (0.180 mg, 70%).
Bis-conjugation via 3’,5’-aldehyde functionality
3’-Aldehyde containing oligonucleotide (5): NaIO4 (20 equiv, 0.923 mg)
was added to a solution of oligonucleotide 4 (0.8 mg, 0.215 mmol) in
water (1 mL), and the solution was stirred at room temperature for 1 h.
Melting studies: The melting curves (absorbance versus temperature)
were measured at 260 nm on a Lambda 5 UV/Vis spectrophotometer
equipped with a Perkin–Elmer C570–070 temperature controller using a
rate of 18C minÀ1 (from 2 to 808C). Melting experiments were carried
out by mixing equimolar amounts of the two undecamer strands dis-
solved in 10 mm sodium phosphate buffer (pH 7) containing 1 mm EDTA
and 100 mm NaCl. All measurements were done at a concentration of
12mm. Before each melting experiment, samples were heated at 808C for
5 min then cooled slowly. Experiments were carried out in duplicate.
The resulting oligonucleotide
(0.554 mg, 70%).
5 was immediately purified by HPLC
3’-RGD,5’-protected diol conjugate (6): Asolution of RGD peptide P1
(4 equiv) in water was added to a solution of oligonucleotide 5 (0.5 mg,
0.136 mmol) in 0.1m ammonium acetate buffer (0.5 mL, Ph 4.6). The reac-
tion mixture was stirred at room temperature and monitored by HPLC.
Completion of the reaction was achieved in 4 h. Purification by HPLC af-
forded the 3’-conjugate 6 (0.35 mg, 60%).
3’-RGD,5’-diol conjugate (7): Oligonucleotide 6 (0.32 mg, 0.07 mmol) was
treated with an 80% AcOH aqueous solution (0.5 mL) for 1 h. The
acetic acid was then lyophilized affording the oligonucleotide 7 (0.19 mg,
60%), which was used in the next step without further purification.
Acknowledgement
3’-RGD,5’-aldehyde conjugate (8): Oxidative cleavage of the 5’-diol was
performed as above for 5. Starting from oligonucleotide 7 (0.18 mg,
0.042 mmol), the oligonucleotide 8 was obtained after purification by
HPLC (0.12 mg, 70%).
This work was supported by the Centre National de la Recherche Scienti-
fique (CNRS). The “Institut Universitaire de France” is gratefully ac-
knowledged for financial support.
5994
ꢀ 2004 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.chemeurj.org
Chem. Eur. J. 2004, 10, 5988 – 5995

Peptide–Oligonucleotide Conjugates
5988 – 5995
[19] D. L. McMinn, M. M. Greenberg, Bioorg. Med. Chem. Lett. 1999, 9,
547–550.
[20] a) W. Mier, R. Erijta, A. Mohammed, U. Haberkorn, M. Eisenhut,
Bioconjugate Chem. 2000, 11, 855–860; b) N. J. Ede, G. W. Tregear,
J. Haralambidis, Bioconjugate Chem. 1994, 5, 373–378.
[1] S. T. Crooke, Annu. Rev. Med. 2004, 55, 61–95.
[2] Oligonucleotides as Therapeutic Agents (Eds.: D. J. Chadwick, G.
Cardew), Wiley, Chichester, 1997.
[3] Applied Antisense Oligonucleotide Technology (Eds.: C. A. Stein,
A. M. Krieg), Wiley, New York, 1998.
[4] D. D. Ma, T. Rede, N. A. Naqvi, P. D. Cook, Biotechnol. Annu. Rev.
2000, 5, 155–196.
[5] T. Tuschl, ChemBioChem 2001, 2, 239–345.
[6] D. Praseuth, A. L. Guieysse, C. Helene, Biochim. Biophys. Acta
1999, 1489, 181–206.
[21] a) D. Forget, D. Boturyn, E. Defrancq, J. Lhomme, P. Dumy, Chem.
Eur. J. 2001, 7, 3970–3984; b) T. S. Zatsepin, D. A. Stetsenko, A. A.
Arzumanov, E. A. Romanova, M. J. Gait, T S. Oretskaya, Bioconju-
gate Chem. 2002, 13, 822–830; c) B. Cebon, J. N. Lambert, D.
Leung, K. Mackie, K. L. McCluskey, H. Nguyen, C. Tassone, Aust. J.
Chem. 2000, 53, 333–340; d) O. M. Edupuganti, O. Renaudet, E.
Defrancq, P. Dumy, Bioorg. Med. Chem. Lett. 2004, 14, 2839–2842.
[22] a) N. Ollivier, C. Olivier, C. Gouyette, T. Huynh-Dinh, H. Gras-
Masse, O. Melnyk, Tetrahedron Lett. 2002, 43, 997–999; b) S. Rad-
datz, J. Mueller-Ibeler, J. Kluge, L. Wäß, G. Burdinski, J. R. Havens,
T. J. Onofrey, D. Wang, M. Schweitzer, Nucleic Acids Res. 2002, 30,
4793–4802.
[23] a) Y. Aubert, S. Bourgerie, L. Meunier, R. Mayer, A.-C. Roche, M.
Monsigny, N. T. Thuong, U. Asseline, Nucleic Acids Res. 2000, 28,
818–825; b) K. Arar, M. Monsigny, R. Mayer, Tetrahedron Lett.
1993, 34, 8087–8090; c) U. Asseline, E. Bonfils, R. Kurfurst, M. Cha-
signol, V. Roig, N. T. Thuong, Tetrahedron 1992, 48, 1233–1254;
d) J. D. Kahl, D. L. McMinn, M. M. Greenberg, J. Org. Chem. 1998,
63, 4870–4871.
[7] T. Herman, D. J. Patel, Science 2000, 287, 820–825.
[8] J. B. Opalinska, A. M. Gewirtz, Nat. Rev. 2002, 1, 503–514.
[9] a) M. J. Gait, Cell. Mol. Life Sci. 2003, 60, 844–853; b) A. Valkna,
U. Soomets, Ü. Langel, in Cell Penetrating Peptides: Process and Ap-
plications (Ed.: Ü. Langel), CRC Press, Boca Raton, 2002, pp. 347–
363; c) P. M. Fischer, E. Karisz, D. P. Lane, Bioconjugate Chem.
2001, 12, 825–841; d) M. A. Lindsay, Curr. Opin. Pharmacol. 2002,
2, 587–594.
[10] a) A. Astriab-Fisher, D. Sergueev, M. Fisher, B. R. Shaw, R. L. Julia-
no, Pharm. Res. 2002, 19, 744–754, and references therein.
[11] C.-H. Tung, S. Stein, Bioconjugate Chem. 2000, 11, 605–618, and ref-
erences therein.
[12] a) C.-H. Tung, K. J. Breslauer, S. Stein, Bioconjugate Chem. 1996, 7,
529–531; b) B. G. de La Torre, F. Albericio, E. Saison-Behmoaras,
A. Bachi, R. Eritja, Bioconjugate Chem. 1999, 10, 1005–1012.
[13] a) M. A. Maier, C. G. Yannopoulos, N. Mohamed, A. Roland, H.
Fritz, V. Mohan, G. Just, M. Manoharan, Bioconjugate Chem. 2003,
14, 18–29; b) T. L. Sheppard, C.-H. Wong, G. F. Joyce, Angew.
Chem. 2000, 112, 3806–3809; Angew. Chem. Int. Ed. 2000, 39, 3660–
3663; c) K. Matsuura, M. Hibino, Y. Yamada, K. Kobayashi, J. Am.
Chem. Soc. 2001, 123, 357–358.
[14] a) M. Beltran, E. Pedroso, A. Grandas, Tetrahedron Lett. 1998, 39,
4115–4118; b) J.-C. Truffert, U. Asseline, A. Brack, N. T. Thuong,
Tetrahedron 1996, 52, 3005–3016.
[15] F. Bergmann, W. Bannwarth, Tetrahedron Lett. 1995, 36, 1839–1842.
[16] a) M. Antopolsky, E. Azhayeva, U. Tengvall, A. Azhayev, Tetrahe-
dron Lett. 2002, 43, 527–530; b) D. A. Stetsenko, A. D. Malakho,
M. J. Gait, Org. Lett. 2002, 4, 3259–3262; c) A. Sakakura, Y. Haya-
kawa, Tetrahedron 2000, 56, 4427–4435; d) M. Antoposky, A. Az-
hayev, Tetrahedron Lett. 2000, 41, 9113–9117; e) M. Antopolsky, A.
Azhayev, Helv. Chim. Acta 1999, 82, 2130–2140.
[24] D. Forget, O. Renaudet, D. Boturyn, E. Defrancq, P. Dumy, Tetrahe-
dron Lett. 2001, 42, 9171–9174.
[25] a) E. Trevisiol, E. Defrancq, J. Lhomme, A. Laayoun, P. Cros, Eur.
J. Org. Chem. 2000, 3, 211–217; b) E. Trevisiol, E. Defrancq, J.
Lhomme, A. Laayoun, P. Cros, Tetrahedron 2000, 56, 6501–6510.
[26] E. Defrancq, A. Hoang, F. Vinet, P. Dumy, Bioorg. Med. Chem.
Lett. 2003, 16, 2683–2686.
[27] a) M. Aumailley, M. Gurrath, G. Mꢁller, J. Calvete, R. Timpl, H.
Kessler, FEBS Lett. 1991, 291, 50–54; b) R. Haubner, D. Finsinger,
H. Kessler, Angew. Chem. 1997, 109, 1440–1456; Angew. Chem. Int.
Ed. Engl. 1997, 36, 1374–1389; c) M. Colin, M. Maurice, G. Tru-
gnan, M. Kornprobst, R. P. Harbottle, A. Knight, R. G. Cooper,
A. D. Miller, J. Capeau, C. Coutelle, M. C. Brahimi-Horn, Gene
Ther. 2000, 7, 139–152; d) R. G. Cooper, R. P. Harbottle, H.
Schneider, C. Coutelle, A. D. Miller, Angew. Chem. 1999, 111, 2128–
2132; Angew. Chem. Int. Ed. 1999, 38, 1949–1952.
[28] a) M. A. Zanta, P. Belguise-Valladier, J.-P. Behr, Proc. Natl. Acad.
Sci. USA 1999, 96, 91–96; b) K. H. Bremner, L. W. Seymour, A.
Logan, M. L. Read, Bioconjugate Chem. 2004, 15, 152–161.
[29] J. Shao, J. P. Tam, J. Am. Chem. Soc. 1995, 117, 3893–3899.
[17] a) E. Vives, B. Lebleu, Tetrahedron Lett. 1997, 38, 1183–1186; b) M.
Antopolsky, E. Azhayeva, U. Tengvall, S. Auriola, I. Jaaskelainen, S.
Ronkko, P. Honkakosi, A. Urtti. H. Lonnberg. A. Azhayev, Biocon-
jugate Chem. 1999, 10, 598–606.
[18] a) J. G. Harrison, S. Balasubramanian, Nucleic Acids Res. 1998, 26,
3136–3145; b) K. Arar, A.-M. Aubertin, A.-C. Roche, M. Monsigny,
R. Mayer, Bioconjugate Chem. 1995, 6, 573–577.
Received: April 22, 2004
Published online: October 20, 2004
Chem. Eur. J. 2004, 10, 5988 – 5995
www.chemeurj.org
ꢀ 2004 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
5995