Functionalization of mono- and oligonucleotides with phosphane ligands by amide bond formation

Article abstract of DOI:10.1002/ejoc.201000125

Seven phosphane-functionalized deoxyuridines have been prepared by amide bond formation between aminodeoxyuridines and phosphanylcarboxylic acids. X-ray crystal structures for two of these new modified nucleosides have been obtained. The same coupling method has been extended to oligonucleotides. The phosphane containing strands have been purified and characterized by MALDI-TOF and, for the first time, ³¹P NMR spectrometry. Coordination of a phosphane-modified 15-mer to a [PdCl(η³-allyl)] moiety has been confirmed by ³¹P NMR spectroscopy.

Full text of DOI:10.1002/ejoc.201000125

FULL PAPER
DOI: 10.1002/ejoc.201000125
Functionalization of Mono- and Oligonucleotides with Phosphane Ligands by
Amide Bond Formation
Marzia Nuzzolo,[a] Arnald Grabulosa,[a][‡] Alexandra M. Z. Slawin,[a]
Nico J. Meeuwenoord,[b] Gijsbert A. van der Marel,[b] and Paul C. J. Kamer*[a]
Keywords: Nucleosides / Oligonucleotides / DNA / Phosphanes / Palladium
Seven phosphane-functionalized deoxyuridines have been
prepared by amide bond formation between aminodeoxy-
uridines and phosphanylcarboxylic acids. X-ray crystal struc-
tures for two of these new modified nucleosides have been
obtained. The same coupling method has been extended to
oligonucleotides. The phosphane containing strands have
been purified and characterized by MALDI-TOF and, for the
first time, 31P NMR spectrometry. Coordination of a phos-
phane-modified 15-mer to a [PdCl(η3-allyl)] moiety has been
confirmed by 31P NMR spectroscopy.
mentary strands can give a complete new catalyst that can
place the substrate of interest in a very different environ-
ment.
Our aim is to functionalize DNA strands with phos-
phane units at specific locations,[18–20] with the objective of
applying them as ligands in homogeneous catalysis. A
method to modify DNA strands with phosphane units via
amide bond formation[21] in mono- and oligonucleotides is
described in this paper. This method is compatible with
water as solvent and results in high conversions.
Introduction
During the past decades there has been an enormous
interest in both academia and industry in the synthesis of
optically pure compounds. One of the most efficient meth-
ods to obtain enantiopure compounds is through asymmet-
ric homogeneous catalysis.[1] A countless number of ligands
have been prepared and screened in the search of catalysts
achieving the selectivity and reactivity needed in several re-
actions.[2] Despite all the effort and knowledge gained over
the years, many reactions in industry still lack a proper
catalytic system, forcing the use of stoichiometric amounts
of reagents. Catalysts are also present in Nature in the form
of enzymes. Enzymes have high selectivity and specificity in
the many biochemical reactions they catalyze. Hybrid cata-
Results and Discussion
For our strategy we required nucleosides modified with
lysts aim to merge the recognizing and chiral properties of aliphatic amino groups to react with activated phosphanyl-
biomolecules with a transition metal active site that gives carboxylic acids. Our synthetic route was based on nucleo-
the desired reactivity.[3–5] Several research groups have com- side 1, prepared by Sonogashira coupling between 5-iodo-
bined the molecular recognition of proteins with transition 2Ј-deoxyuridine (IdU) and N-propynyltrifluoroacetamide
metal catalysis[6–8] and in more recent years the focus has (Scheme 1). The amino group has to be protected to avoid
also spread to the use of DNA combined with transition interference with the catalytic process.
metals. Two methods have been used to attach transition
There are a number of reports on the preparation of
metal fragments to DNA: they are the covalent[9–11] and compound 1 under different conditions.[22–24] Although the
non covalent[12–16] approaches. Both Jäschke[9] and Roelfes reaction was successful under the conditions of Scheme 1,
and Feringa[17] groups have found that the catalytic results we found that during column chromatography purification
were sequence dependent. This discovery is a great triethylammonium iodide coeluted with the desired prod-
advantage of DNA as ligand because simple changes in the uct. This problem has been solved by changing to a poly-
structure of the DNA such as the sequence of the comple- mer-supported base (Amberlite Ira-67) and use of Pd/C as
heterogeneous catalyst[24] but in our hands low conversions
were found even after prolonged reaction times. Therefore,
we decided to use triethylamine and optimize the purifica-
tion method. Treatment of the crude reaction mixture with
solid potassium carbonate in THF neutralized the ammo-
nium salt back to triethylamine, which could be removed
under vacuum. Addition of methanol allowed us to sepa-
rate most of the palladium species as insoluble solids. The
[a] School of Chemistry, University of St Andrews,
KY16 9ST, St Andrews, Fife, United Kingdom
E-mail: pcjk@st-andrews.ac.uk
[b] Leiden Institute of Chemistry, Gorlaeus Laboratory,
University of Leiden,
Einsteinweg 55, 2333 CC, Leiden, The Netherlands
[‡] Present address: Departament de Química Inorgànica,
Universitat de Barcelona,
Martí i Franquès, 1–11, 08028 Barcelona, Spain
Eur. J. Org. Chem. 2010, 3229–3236
© 2010 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
3229

P. C. J. Kamer et al.
FULL PAPER
Scheme 1. Preparation of amino-modified nucleosides 2 and 4.
final residue was recrystallised from cold acetonitrile[23] to this method, we isolated compound 3 in high purity after
render pure 1 in moderate yield with no need of column simply removing the Pd/C by filtration.
chromatography.
The deprotection of the amine group was carried out in
In order to have a more flexible and reactive amino concentrated aqueous ammonia, rendering the compounds
1
group, we considered the hydrogenation of 1. This reaction 2 and 4 as confirmed with H NMR spectroscopy.
has been described before in literature. Grasby and co-
With the modified nucleosides in hand, we turned our
workers[25] prepared this compound by two stepwise re- attention to the coupling with phosphanylcarboxylic acids.
ductions with NiCl2/NaBH4 isolating the (Z)-alkene inter- As our ultimate goal is to perform these couplings with
mediate. We used an excess of reducing reagent and longer DNA strands in an aqueous environment, we sought a suit-
reaction times in order to carry on a one-pot reduction to able method under these conditions and we applied the
the alkane, but only starting material and alkene were iso- method of Jäschke and co-workers.[21] This method is de-
lated. The procedure of Williams and co-workers,[23] was picted in Scheme 2 and consists of the activation of the car-
also tested with 20% PtO2 (Adam’s catalyst) and 10 bar of boxylic acid group with EDC in the presence of NHS to
H2 in methanol at room temperature. After 12 h mainly form a succinimidyl ester,[26] which is stable under aqueous
compound 3 was obtained, but contaminated with small conditions but reactive towards primary amines.
amounts of unidentified impurities. The best results were
The phosphanylcarboxylic acids were activated with
obtained by hydrogenation with 5% Pd/C for 24 h. With EDC/NHS in DMF and reacted in situ with degassed solu-
Scheme 2. Preparation of phosphane-modified nucleobases 6–8.
3230
www.eurjoc.org
© 2010 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Eur. J. Org. Chem. 2010, 3229–3236

Functionalization of Nucleotides with Phosphane Ligands
tions of 2 or 4 in aqueous NaHCO3. After extractive work- tramolecular hydrogen bond between the OH at the 5’ posi-
up and column chromatography we obtained the pure com- tion of the ribose ring and the carbonyl oxygen atom of the
pounds 6–8. The couplings with 2 resulted in much lower amide group.
yields compared to 4, possibly due to the reduced reactivity
Following standard procedures,[27,28] the 5Ј-hydroxy
of the propargylamino group. All these compounds were group of nucleosides 1 and 3 was protected with DMTCl
fully characterized, including the crystal structures for 7o (4,4Ј-dimethoxytrityl chloride) to afford the compounds 9
and 7m (Figure 1).
and 10 respectively (Scheme 3). Interestingly, we could also
obtain 10 by hydrogenation of 9 in excellent yield without
affecting the DMT group. Standard phosphitylation with 2-
cyanoethoxy-(N,N-diisopropylamino)chlorophosphane of 9
and 10 finally afforded 11 and 12 respectively, ready to be
used in the oligonucleotide synthesis.[27,28]
As a model compound, the trimer TXG (X = 12) was
synthesized in solution according to a recently published
procedure.[29] After FPLC purification, MALDI-TOF
analysis confirmed the synthesis of the desired compound.
The phosphane introduction was carried out with 4 equiv.
of ortho-diphenylphosphanylbenzoic acid under the same
conditions used for the mononucleosides. After FPLC puri-
fication with degassed solvents, MALDI-TOF and electro-
spray were inconclusive because no peaks of the expected
product, oxidized product or starting material could be
found. 31P{1H} NMR showed two peaks at 0 and +36 ppm
in approximately 2:1 ratio, likely to correspond to the oxid-
ized product. Apparently, the degassing of eluents for
Figure 1. Ortep representation of 7o (left) and 7m (right). The ther-
mal ellipsoids are drawn at 50% probability. Only one of the inde-
pendent molecules of 7m is represented.
The crystal structures proved the identity of the com- FPLC did not prevent product oxidation. At this point we
pounds and confirmed the expected absolute configurations turned our attention to longer strands, which can be
of the stereogenic carbons on the ribose ring. Bonding dis- purified by precipitation under more oxygen-free condi-
tances and angles are similar to those found in a related tions.
compound reported by us.[10] In spite of that, 7m presents
DNA strands, 15 and 16 bases long, containing the bases
a more open structure than 7o, with the diphenylphosphane 11 and 12 were synthesized on 10 µmol scale by automatic
unit displaced far away from the ribose ring and the elec- solid-supported synthesis, cleaved from the solid support
tron pair at the phosphorus atom pointing outwards the and purified by HPLC. The prepared sequences and their
molecule. This conformation is probably enforced by an in- MALDI-TOF analysis results are listed in Table 1.
Scheme 3. Synthesis of phosphoramidites 11 and 12.
Eur. J. Org. Chem. 2010, 3229–3236
© 2010 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.eurjoc.org
3231

P. C. J. Kamer et al.
FULL PAPER
Table 1. DNA strands with a modified base.
Strand Sequence[a]
best conditions (800 nmol, 16 h of acquisition time, 49,000
scans) we recorded the 31P{1H} NMR spectrum of strand
2H coupled to p-dppba (Figure 2).
Molecular
weight
Found
mass
1
2
2H
3H
4
5Ј-TAAGCCAXCATCCGC-3Ј
5Ј-ATTCGGTXGTAGGCG-3Ј
5Ј-ATTCGGTYGTAGGCG-3Ј
5Ј-CGCCTACYACCGAAT-3Ј
5Ј-GCGGATGXTGGCTTA-3Ј
5Ј-XATTCGGTAGTAGGCG-3Ј
5Ј-YTTCGGTAGTAGGCG-3Ј
5Ј-GCTGATGACTATCGAZ-3Ј
4536
4678
4682
4540
4678
4991
4682
4729
4534
4676
4682
4540
4677
4989
4680
4730
5
6H
7[b]
[a] X = 11, Y = 12, Z = –(CH2)3NH2 at the terminal phosphate
group of dA. [b] This strand (1 µmol) was purchased from Ocimum
Biosolutions but cleaved from the solid support and purified by
HPLC by us.
Cleavage from solid support was performed in aqueous
ammonia at 55 °C overnight. This procedure also released
the free amine from the trifluoroacetyl protecting group.
After purification by preparative HPLC, MALDI-TOF
analysis showed an extra peak in most of the strands with
a mass 53 units higher than expected. This peak corre-
sponds to strands whose aliphatic amino group has been
attacked by the acrylonitrile released from the deprotection
of the phosphates. Treatment of monobase 1 with ammonia
in the presence of acrylonitrile confirmed the high affinity
of the amino group towards acrylonitrile. We were able to
minimise this problem by deprotecting the strands with
aqueous solutions of the more nucleophilic methylamine.
Several attempts were made to couple strands 1, 4 and 5
to o-, p-(diphenylphosphanyl)benzoic acid (dppba) and 5.
The purified strands were coupled with 250–500 equiv. of
activated phosphanylcarboxylic acids under the same con-
ditions discussed above for the monomers and the trimer.
Unfortunately, the coupling reactions with the propargylic
amines exhibited bad reproducibility. Therefore, we discon-
tinued our efforts with propargylic amine-modified strands
1, 2, 4 and 5 and decided to use strand 7, whose amino
group should be more reactive. Gratifyingly, p-dppba was
successfully coupled. The same activated phosphane was
then successfully coupled to strands 2H and 3H with good
reproducibility. The method is also applicable to the impor-
tant class of diphosphane ligands as illustrated by the suc-
cessful coupling of 5 to strand 6H.
Figure 2. 31P{1H} spectrum of strand 2H coupled to p-dppba.
We assigned the peak at δ = –6.8 ppm to the coupled
phosphane and the peaks around –1.3 ppm to the phos-
phates of the backbone. The difference in relaxation times
between the two kinds of phosphorus atoms and the low
concentration of the sample account for the 1.4:14 ratio.
The MALDI-TOF analysis of the same strand is repre-
sented in Figure 3 and shows the masses corresponding to
the oxidized strand (expected at m/z = 4986 units) with the
additional peaks due to multiple Na cations.
Several methods for purification of the modified oligonu-
cleotides were tested and ethanol precipitation proved to be
the most efficient one. After work-up and purification the
strands were analyzed by MALDI-TOF, which gave the
molecular weight corresponding to the phosphane oxides
of the coupled strands. Phosphane oxidation usually occurs
under the conditions of MALDI-TOF analysis.
Figure 3. MALDI-TOF analysis of phosphane-modified strand
2H.
The coupled strand 4 with o-dppba was analyzed also
In situ complexation of a crude sample of a 15-mer with
with electrospray mass spectroscopy, giving the mass of the 0.5 equiv. of the dimer [PdCl(η3-allyl)]2 resulted in a broad
unoxidized oligonucleotide confirming that oxidation of the peak at δ = +28 ppm in the 31P{1H} NMR spectrum indi-
strands takes place during the MALDI-TOF analysis. An- cating that Pd complexation is feasible.
other way to prove that phosphane containing strands were
A preliminary catalytic study was performed with ligands
not oxidized after purification is 31P NMR spectroscopy 6 and 7. Allylic alkylation at 1,3-diphenylallyl acetate with
although it proved challenging due to the extremely small dimethyl malonate and allylic amination with benzylamine
amounts of oligonucleotide (less than 1 µmol). Under our as nucleophile resulted in 100% conversion for ligands
3232
www.eurjoc.org
© 2010 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Eur. J. Org. Chem. 2010, 3229–3236

Functionalization of Nucleotides with Phosphane Ligands
apparatus equipped with a UV/Vis photodiode array detector. The
measurements were performed at 254 nm, with an injection volume
of 1 µL. Compounds 9,[27] 11[27] and 12[28] were prepared according
to literature procedures.
6p, -m and 7p, -m while ligands 6o and 7o gave less than
10% conversions. Unfortunately, no enantioselectivity was
obtained.
We are currently exploring the applicability of this syn-
thetic method to other DNA strands and their use in cataly-
sis.
CCDC-763240 (for 7m) and -763241 (for 7o) contain the supple-
mentary crystallographic data for this paper. These data can be
obtained free of charge from The Cambridge Crystallographic
Data Centre via www.ccdc.cam.ac.uk/data_request.cif.
Conclusions
5-[3-N-(Trifluoroacetyl)aminopropynyl]-2Ј-deoxyuridine (1): This
compound was prepared by a slightly modified procedure than
those previously published.[23,24,30,31] 5-Iodo-2Ј-deoxyuridine
(2.00 g, 5.65 mmol) and CuI (108.6 mg, 0.57 mmol, 0.1 equiv.) were
suspended in 30 mL of anhydrous DMF. The mixture was stirred
for 30 min, protected from light. To this mixture triethylamine
(1.60 mL, 11.50 mmol), N-propynyltrifluoroacetamide (2.60 g,
17.20 mmol)[32] and tetrakis(triphenylphosphane)palladium(0)
(330 mg, 0.28 mmol, 0.05 equiv.) were added in this order. The yel-
low clear solution was stirred for 18 h, protected from light, giving
an orange solution. The solvents were then removed in vacuo and
the residue was dissolved in 70 mL of THF. Solid potassium car-
bonate (approximately 4 g) was added and the suspension stirred
in an open flask for 2 h. The salts were filtered and the liquid was
concentrated to dryness. The gummy residue was treated with
30 mL of methanol causing the precipitation of a yellow solid. This
solid was filtered and discarded. The remaining liquid was concen-
trated to dryness and the methanol treatment was repeated until
no more solid precipitated (usually two or three times). The filtrate
was concentrated under vacuum again and the residue was recrys-
tallised from 30 mL of acetonitrile. After several hours in the
freezer, the desired product was obtained as a yellow powder; yield
1.2 g (58%). Characterization of the product fully agreed with the
literature.[23,24,30,31]
In summary, we have shown that amide bond formation
is an efficient tool to functionalize nucleosides with phos-
phane moieties. The strategy is also applicable to oligonu-
cleotides, but utmost care has to be taken to avoid oxidation
of the phosphane groups. We also coordinated a [PdCl(η3-
allyl)] to one of the phosphane-modified 15-mers.
Experimental Section
General: Chemicals were purchased from Aldrich, Acros, Fluka,
Glen Research and Link Technologies. Triethylamine was distilled
from sodium, acetonitrile over CaH2, THF over sodium/benzophe-
none and ethanol and methanol over Mg/I2; all under nitrogen.
DMF was purchased “Extra dry” and kept over molecular sieves
(4 Å) under argon. Aqueous solutions were degassed by three
freeze/thaw cycles under vacuum and kept under argon. Air and
water sensitive reactions were carried out in dry (from oven) glass-
ware using standard Schlenk techniques under atmosphere of puri-
fied argon. Thin layer chromatography (TLC) was performed on
silica plates (Polygram 0.2 mm with fluorescent indicator UV254).
Silica gel 60, particle size 0.063–0.2 mm from Fluka was used for
flash chromatography purification. Oligonucleotides were synthe-
sized at the University of Leiden with a AKTA-oligopilot DNA
synthesizer. DNA concentration was determined spectrophotomet-
rically when needed. Purification by ion exchange and size ex-
clusion chromatography and desalting were carried out with a
FPLC AktaBasic 100 (P-901) monitored by a UV-900 detector. Pu-
rification of modified oligonucleotides was carried out on a Hi-
Load 16/10 Q Sepharose HP with the following buffers: buffer A
(1  NaCl/10 m NaOH) and buffer B (10 m NaOH). The col-
umn for the desalting was a HiPrep 26/10 Desalting with a linear
gradient of 0.15  NH4HCO3 buffer. HPLC purification of oligo-
nucleotides was carried out in a Waters apparatus equipped with a
Waters 2700 sample manager, Waters 600 controller and Waters
2487 dual λ absorption detector. The reverse phase column used
was a Phenomenex, Clarity® C18 5 µ-oligo-RP, 250ϫ21.20 mm.
NMR spectra were recorded on Bruker Avance spectrometers (300,
400 and 500 MHz). Positive chemical shifts (δ values, ppm) are
given for higher frequency shifts relative to tetramethylsilane (for
1H and 13C), 85% aqueous H3PO4 (31P) and CCl3F (19F). 13C, 31P
and 19F spectra have been recorded 1H-decoupled. The following
abbreviations have been used in the description of the spectra: s,
singlet; d, doublet; t, triplet; m, multiplet. IR spectra were recorded
in a Perkin–Elmer Spectrum GX spectrometer in KBr pellets, with
a window between 4000 and 400 cm–1. MALDI-TOF mass spectra
were recorded in a 4800 Plus MALDI TOF/TOF Analyzer using
a 9:1 3-hydroxypicolinic acid (HPA): ammonium citrate matrix.
This matrix was made with 50 mg/mL of HPA in 50:50 acetonitrile/
water and 50 mg/mL of ammonium citrate in water. The samples
for MALDI-TOF analysis were desalted with a GE Healthcare Il-
lustra micron G-25 column and Millipore C18 tips. Chiral HPLC
analyses were carried out on an Agilent Technologies 1200 Series
5-(3-Aminopropynyl)-2Ј-deoxyuridine (2): The protected base 1
(377 mg, 1.0 mmol) was dissolved in 20 mL of concentrated com-
mercial ammonia and stirred for 12 h. Evaporation of the liquid
furnished the desired deprotected base 2 in quantitative yield as a
yellowish oil. This product was used without any purification for
the couplings with the phosphanylcarboxylic acids. However,
19F{1H} NMR spectroscopy showed the presence of the trifluoro-
acetate anion at δ = –75 ppm. The compound could be used further
as such or alternatively the impurity could be removed by column
chromatography (SiO2, CH2Cl2/MeOH, 70:30), to yield the pure
product as a yellow solid. The yield was quantitative. Characteriza-
tion of the product fully agreed with data from literature.[28]
5-[3-N-(Trifluoroacetyl)aminopropyl]-2Ј-deoxyuridine (3): The mono-
cleoside 1 (2.0 g, 5.3 mmol) and 10% Pd/C (564 mg, 0.53 mmol Pd)
were suspended in 30 mL of dry methanol and hydrogenated at 10
Bar of H2 for 26 h at room temperature. The resulting suspension
was filtered through a pad of celite twice and concentrated in
vacuo. If needed, the crude was purified via a short column (SiO2,
EtOAc/MeOH, 90:10) to yield, after evaporation of the solvent, the
title product as a pale yellow solid; yield 1.88 g (90%). Characteri-
zation of the product fully agreed with data from the literature.[23,25]
5-(3-Aminopropyl)-2Ј-deoxyuridine (4): The protected base
3
(115 mg, 0.3 mmol) was dissolved in 6 mL of concentrated ammo-
nia and stirred for 12 h. Evaporation of the liquid furnished the
desired deprotected base 4 in quantitative yield as a yellowish oil.
This product was used without any purification for the couplings
with the phosphanylcarboxylic acids. The yield was quantitative.
Characterization of the product fully agreed with data from the
literature.[28]
Eur. J. Org. Chem. 2010, 3229–3236
© 2010 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.eurjoc.org
3233

P. C. J. Kamer et al.
FULL PAPER
6o: o-(Diphenylphosphanyl)benzoic acid (306 mg, 1.2 mmol),
EDC·HCl (230 mg, 1.2 mmol) and NHS (173 mg, 1.5 mmol) were
dissolved in 20 mL of dry degassed DMF and stirred for 2 h. The
nucleoside 2 (281 mg, 1.0 mmol) was dissolved in 10 mL of pre-
viously degassed 0.1  NaHCO3 aqueous solution and the acti-
vated phosphane was added via a syringe. A white precipitate
formed but redissolved after a few minutes, giving a clear orange
mixture, which was stirred for 12 h. Water (50 mL), CHCl3 (50 mL)
and brine (ca. 10 mL) were added to the reaction mixture giving a
biphasic system. The organic phase was separated and the aqueous
layer extracted twice with CHCl3 (50 mL). The combined organic
layers were washed with water (100 mL) and dried with Na2SO4.
After filtration, the solvent was removed under vacuum (heating to
60 °C was necessary to remove the last traces of DMF) giving a
gummy white residue. This crude product was purified column
chromatography (SiO2, CHCl3/MeOH, 90:10) to give the title prod-
uct as a white foam after solvent removal; yield 171 mg (30%). IR
Ar), 99.9 (C), 90.0 (C), 89.2 (CH), 87.0 (CH), 75.2 (C), 72.0 (CH),
62.6 (CH2), 41.7 (CH2), 31.0 (CH2) ppm. 31P NMR (121 MHz,
CD3OD): δ = –5.5 ppm. MS (ES+): m/z = 592.02 [M + Na]+.
HRMS: calcd. for C31H28N3NaO6P, 592.1613; found 592.1617.
7o: The same method as for 6o was employed but starting from
0.3 mmol (85.8 mg) of mononucleoside 4; yield 80 mg (46%). IR
(KBr): ν
= 3403, 3053, 2930, 1671 ν(C=O), 1475, 1434, 1275,
˜
max
1
1055, 746, 698 cm–1. H NMR (300 MHz, CD3OD): δ = 7.85 (s, 1
H, =CH), 7.52 (ddd, J = 7.6, 3.6, 1.3 Hz, 1 H, Ar), 7.42 (dt, J =
7.8, 1.3 Hz, 1 H, Ar), 7.36–7.23 (m, 11 H, Ar), 6.96 (ddd, J = 4.1,
3.6, 0.9 Hz, 1 H, Ar), 6.27 (t, J = 6.8 Hz, 1 H, CHOCH), 4.41–
4.36 (m, 1 H, CHOH), 3.90 (dd, J = 6.9, 3.5 Hz, 1 H, CHOCH),
3.77 (dd, J = 12.0, 3.1 Hz, 1 H, CH2OH), 3.68 (dd, J = 12.0, 3.8 Hz,
1 H, CH2OH), 3.18 (t, J = 6.7 Hz, 2 H, NHCH2), 2.28–2.21 (m, 4
H, CHCH2CH, CH2CH2CH), 1.63–1.55 (m, 2 H, CH2CH2CH2)
ppm. 13C NMR (75 MHz, CD3OD): δ = 172.2 (C), 166.0 (C), 152.3
(C), 138.9 (CH), 138.6–128.4 (C, CH), 114.8 (C), 88.9 (CH), 86.4
(CH), 72.2 (CH), 62.9 (CH2), 41.2 (CH2), 39.7 (CH2), 28.9 (CH2),
25.0 (CH2) ppm. 31P NMR (121 MHz, CD3OD): δ = –10.1 ppm.
MS (ES+): m/z = 572.19 [M – H]; 573.20 [M]. HRMS: calcd. for
C31H31N3O6P, 573.1950; found 572.1942. Crystals suitable for X-
ray diffraction were grown from an acetonitrile solution at room
temperature.
(KBr): νmax = 3411, 3054, 1700 ν(C=O), 1653 ν(C=O), 1584, 1465,
˜
1
1434, 1282, 1093, 747, 697 cm–1. H NMR (500 MHz, CD3OD): δ
= 8.34 (s, 1 H, =CH), 7.60 (ddd, J = 7.6, 3.6, 1.0 Hz, 1 H, Ar),
7.43 (td, J = 7.5, 1.1 Hz, 1 H, Ar), 7.27–7.37 (m, 11 H, Ar), 7.01
(ddd, J = 7.6, 4.1, 0.7 Hz, 1 H, Ar), 6.27 (t, J = 6.6 Hz, 1 H,
CHOCH), 4.43–4.41 (m, 1 H, CHOH), 4.13 (s, 2 H, NHCH2), 3.97
(dd, J = 6.5, 3.2 Hz, 1 H, CHOCH), 3.83 (dd, J = 12.0, 3.1 Hz, 1
H, CH2OH), 3.75 (dd, J = 12.0, 3.5 Hz, 1 H, CH2OH), 2.36–2.31
(m, 1 H, CHCH2), 2.26–2.21 (m, 1 H, CHCH2) ppm. 13C NMR
(126 MHz, CD3OD): δ = 171.4 (C), 164.8 (C), 151.2 (C), 145.5
(CH), 142.0 (d, JCP = 24.5 Hz, 1 C), 138.9 (d, JCP = 8.6 Hz, 1 C),
138.8 (d, JCP = 13.6 Hz, 1 C), 135.2–128.6 (CH Ar), 100.0 (C), 89.8
(C), 89.2 (CH), 87.1 (CH), 75.6 (C), 72.1 (CH), 62.7 (CH2), 41.8
(CH2), 31.0 (CH2) ppm. 31P NMR (162 MHz, CD3OD): δ =
–9.5 ppm. MS (ES+): m/z = 592.24 [M + Na]+. HRMS: calcd. for
C31H28N3NaO6P, 592.1613; found 592.1615.
7m: The same method as for 6m was employed but starting from
0.3 mmol (85.8 mg) of mononucleoside 4; yield 142 mg (82%). IR
(KBr): ν
= 3421, 3054, 2926, 1675 ν(C=O), 1539, 1465, 1434,
˜
max
1275, 1093, 744, 696 cm–1. 1H NMR (300 MHz, CD3OD): δ = 7.87
(s, 1 H, =CH), 7.85–7.77 (m, 2 H, Ar), 7. 44 (td, J = 7.6, 0.7 Hz,
1 H, Ar), 7.37–7.26 (m, 11 H, Ar), 6.27 (t, J = 6.7 Hz, 1 H,
CHOCH), 4.41–4.37 (m, 1 H, CHOH), 3.90 (dd, J = 6.7, 3.4 Hz,
1 H, CHOCH), 3.78 (dd, J = 12.1, 3.1 Hz, 1 H, CH2OH), 3.69 (dd,
J = 12.1, 3.7 Hz, 1 H, CH2OH), 3.35 (t, J = 6.8 Hz, 2 H, NHCH2),
2.37 (t, J = 7.2 Hz, 2 H, CH2CH2CH), 2.25–2.21 (m, 2 H,
CHCH2CH), 1.83–1.74 (m, 2 H, CH2CH2CH2) ppm. 13C NMR
(75 MHz, CD3OD): δ = 169.9 (C), 166.1 (C), 152.2 (C), 138.8 (CH),
139.9–128.6 (C, CH), 114.8 (C), 88.9 (CH), 86.4 (CH), 72.2 (CH),
62.8 (CH2), 41.3 (CH2), 49.1 (CH2), 29.2 (CH2), 25.1 (CH2) ppm.
31P NMR (121 MHz, CD3OD): δ = –5.4 ppm. MS (ES+): m/z =
596.11 [M + Na]+; 612.11 [M + O + Na+]. HRMS: calcd. for
C31H32N3NaO6P, 596.1926; found 596.1908. Crystals suitable for
X-ray diffraction were grown from an acetonitrile solution at room
temperature.
6m: The same method as for 6o was employed using m-(diphenyl-
phosphanyl)benzoic acid, starting from 0.3 mmol (84.4 mg) of
mononucleoside 2; yield 107 mg (63%). IR (KBr): ν
= 3423,
˜
max
1686 ν(C=O), 1542, 1459, 1434, 1283, 1093, 746, 696 cm–1. 1H
NMR (400 MHz, CD3OD): δ = 8.31 (s, 1 H, =CH), 7.90 (s, 1 H,
Ar), 7.86 (dt, J = 8.1, 1.5 Hz, 1 H, Ar), 7.82 (dt, J = 7.7, 1.3 Hz,
1 H, Ar), 7.45 (tdd, J = 7.7, 1.3, 0.6 Hz, 1 H, Ar), 7.39–7.26 (m,
10 H, Ar), 6.23 (t, J = 6.6 Hz, 1 H, CHOCH), 4.39–4.37 (m, 1 H,
CHOH), 4.31 (s, 2 H, NHCH2), 3.93 (dd, J = 6.6, 3.3 Hz, 1 H,
CHOCH), 3.80 (dd, J = 12.0, 3.1 Hz, 1 H, CH2OH), 3.72 (dd, J =
12.0, 3.5 Hz, 1 H, CH2OH), 2.33–2.27 (m, 1 H, CHCH2), 2.24–
2.219 (m, 1 H, CHCH2) ppm. 13C NMR (101 MHz, CD3OD): δ =
169.3 (C), 164.7 (C), 151.1 (C), 145.5 (CH), 140.0 (d, JCP = 13.5 Hz,
1 C), 138.0 (d, JCP = 10.9 Hz, 1 C), 137.7 (d, JCP = 15.5 Hz, CH),
135.4 (d, JCP = 7.5 Hz, 1 C), 134.9–128.7 (CH Ar), 99.9 (C), 90.0
(C), 89.1 (CH), 87.0 (CH), 75.2 (C), 72.0 (CH), 62.6 (CH2), 41.7
(CH2), 31.0 (CH2) ppm. 31P NMR (162 MHz, CD3OD): δ =
–5.5 ppm. MS (ES+): m/z = 592.03 [M + Na]+. HRMS: calcd. for
C31H28N3NaO6P, 592.1613; found 592.1605.
7p: The same method as for 6p was employed but starting from
0.3 mmol (85.8 mg) of mononucleoside 4; yield 103 mg (60%). IR
(KBr): ν
= 3423, 3054, 2927, 1686 ν(C=O), 1475, 1434, 1276,
˜
max
1093, 746, 697 cm–1. H NMR (300 MHz, CD3OD): δ = 7.89 (s, 1
H, =CH), 7.79 (d, J = 1.4 Hz, 1 H, Ar), 7.76 (d, J = 1.4 Hz, 1 H,
Ar), 7.39–7.26 (m, 12 H, Ar), 6.27 (t, J = 6.8 Hz, 1 H, CHOCH),
4.42–4.38 (m, 1 H, CHOH), 3.90 (dd, J = 6.7, 3.4 Hz, 1 H,
CHOCH), 3.80 (dd, J = 12.1, 3.1 Hz, 1 H, CH2OH), 3.71 (dd, J =
12.1, 3.7 Hz, 1 H, CH2OH), 3.40 (t, J = 6.7 Hz, 2 H, NHCH2),
2.39 (t, J = 7.2 Hz, 2 H, CH2CH2CH), 2.25–2.22 (m, 2 H,
1
6p: The same method as for 6o was employed but using p-(di-
phenylphosphanyl)benzoic acid; yield 171 mg (30%). IR (KBr):
CHCH2CH), 1.82 (quint, J = 6.9 Hz, 1 H, CH2CH2CH2) ppm. 13
C
NMR (75 MHz, CD3OD): δ = 170.0 (C), 166.1 (C), 152.2 (C),
138.8 (CH), 135.1–128.2 (C, CH), 114.8 (C), 88.9 (CH), 86.3 (CH),
72.1 (CH), 62.8 (CH2), 41.3 (CH2), 40.0 (CH2), 29.2 (CH2), 25.2
(CH2) ppm. 31P NMR (121 MHz, CD3OD): δ = –5.6 ppm. MS
ν
= 3394, 3054, 2963, 1691 ν(C=O), 1535, 1465, 1434, 1263,
˜
max
1094, 802, 746, 696 cm–1. H NMR (300 MHz, CD3OD): δ = 8.32
(s, 1 H, =CH), 7.81–7.78 (m, 2 H, Ar), 7.39–7.26 (m, 12 H, Ar),
6.23 (t, J = 6.6 Hz, 1 H, CHOCH), 4.40–4.35 (m, 3 H, CHOH,
NHCH2), 3.93 (dd, J = 6.6, 3.3 Hz, 1 H, CHOCH), 3.80 (dd, J =
12.0, 3.1 Hz, 1 H, CH2OH), 3.72 (dd, J = 12.0, 3.5 Hz, 1 H,
CH2OH), 2.34–2.20 (m, 2 H, CHCH2) ppm. 13C NMR (75 MHz,
CD3OD): δ = 169.3 (C), 164.7 (C), 151.1 (C), 145.5 (CH), 144.0 (d,
1
(ES+): m/z
=
596.02 [M
+
Na]+. HRMS: calcd. for
C31H32N3NaO6P, 596.1926; found 596.1915.
8: The same method as for 7o was employed but using 4-[bis(2-
diphenylphosphanyl)ethyl]amino-4-oxobutanoic acid (5). Yields in
JCP = 8.9 Hz, 1 C), 137.8 (d, JCP = 6.0 Hz, 1 C), 135.2–128.3 (CH the 10–20% range were obtained. IR (KBr): ν
= 3423, 3054,
˜
max
3234
www.eurjoc.org
© 2010 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Eur. J. Org. Chem. 2010, 3229–3236

Functionalization of Nucleotides with Phosphane Ligands
2923, 1671 ν(C=O), 1475, 1434, 1276, 1096, 742, 698 cm–1. 1H tion in this order. The reaction mixture was stirred for 24 h. Each
NMR (500 MHz, CD3OD): δ = 7.85 (s, 1 H, =CH), 7.39–7.29 (m,
20 H, Ar), 6.25 (t, J = 6.8 Hz, 1 H, CHOCH), 4.39–4.37 (m, 1 H,
solution was filtered through SiO2 and left to evaporate overnight.
A CHIRACEL OD-H chiral column (Daicel Chemical Industry
CHOH), 3.89 (dd, J = 6.3, 3.2 Hz, 1 H, CHOCH), 3.79 (dd, J = Ltd.) was employed using a 99.5:0.5 (hexane/iPrOH) isocratic mix-
12.1, 3.0 Hz, 1 H, CH2OH), 3.69 (dd, J = 12.1, 3.5 Hz, 1 H,
CH2OH), 3.39–3.31 [m, 4 H, (CH2)2N], 3.20–3.12 (m, 2 H,
NHCH2), 2.37 (s, broad, 4 H, CCH2CH2C), 2.30 (m, 4 H,
CH2CH2CH), 2.25 (t, J = 7.8 Hz, 2 H), 2.19–2.16 (m, 2 H,
CHCH2CH), 1.71–1.564 (m, 2 H, CH2CH2CH2) ppm. 13C NMR
(125 MHz, CD3OD): δ = 174.8 (C), 173.6 (C), 165.9 (C), 152.2 (C),
139.4 (d, JCP = 12.6 Hz, 1 C), 139.0 (d, JCP = 12.2 Hz, 1 C), 138.8
(CH), 133.9–133.6 (CH), 130.3–129.6 (CH), 114.8 (C), 88.9 (CH),
86.4 (CH), 72.2 (CH), 62.9 (CH2), 46.6 (d, JCP = 24.7 Hz, CH2),
45.0 (d, JCP = 24.1 Hz, CH2), 41.4 (CH2), 39.2 (CH2), 31.8 (CH2),
29.5 (CH2), 29.2 (CH2), 28.5 (d, JCP = 14.6 Hz, CH2), 27.3 (d, JCP
= 13.8 Hz, CH2), 24.9 (CH2) ppm. 31P NMR (121 MHz, CD3OD):
δ = –21.0, –21.7 ppm. MS (ES+): m/z = 831.02 [M + Na]+; 847.02
[M + O + Na]+. HRMS: calcd. for C44H50N4NaO7P2, 831.3052;
found 831.3040.
ture with a flow of 0.5 mL/min; tR(R) 28.4 min, tR(S) 30.3 min.
Acknowledgments
The authors would like to acknowledge Dr Thomas Lebl and Ms
Melanja Smith for assistance in NMR and Mr Alex Houston, Dr
Catherine Botting and Dr Sally Shirran for MALDI-TOF measure-
ments. Dr Qingzhi Zhang was very helpful with the preparative
HPLC. Financial contributions from the European Union (Marie
Curie Excellence Grants, artizyme catalysis, contract number
MEXT-2004-014320) and from the Wellcome Trust are gratefully
acknowledged.
[1] P. W. N. M. van Leeuwen, Homogeneous Catalysis, Understand-
ing the Art, Kluwer Academic Publishers, Dordrecht, 2004.
[2] A. Börner, Phosphorus Ligands in Asymmetric Catalysis: Syn-
thesis and Applications, John Wiley & Sons, New York, 2008.
[3] C. M. Thomas, T. R. Ward, Chem. Soc. Rev. 2005, 34, 337.
[4] M. T. Reetz, M. Rentzsch, A. Pletsch, M. Maywald, P. Mai-
wald, J. J. Peyralans, A. Maichele, Y. Fu, N. Jiao, F. Hollmann,
R. Mondiere, A. Taglieber, Tetrahedron 2007, 63, 6404.
[5] J. Steinreiber, T. R. Ward, Coord. Chem. Rev. 2008, 252, 751.
[6] M. E. Wilson, R. G. Nuzzo, G. M. Whitesides, J. Am. Chem.
Soc. 1978, 100, 2269.
5-[3-N-(Trifluoroacetyl)aminopropyl]-5Ј-O-(4,4Ј-dimethyltrityl)-2Ј-de-
oxyuridine (10): The monocleoside 9 (2.0 g, 2.94 mmol) and 10%
Pd/C (313 mg, 0.294 mmol of Pd) were suspended in 30 mL of dry
methanol and hydrogenated under 10 bar of H2 for 24 h at room
temperature. The resulting suspension was filtered through a pad
of celite twice and concentrated in vacuo. The crude product was
purified via a short column (SiO2, CH2Cl2/MeOH/NEt3, 89:10:1)
to yield, after removal of the solvent, the title product as a pale
yellow solid; yield 1.91 g (95%). Characterization of the product
fully agreed with data in the literature.[28]
[7] M. T. Reetz, Angew. Chem. Int. Ed. 2001, 40, 284.
[8] T. R. Ward, Chem. Eur. J. 2005, 11, 3798.
Example of Phosphane Modification of a 15-mer DNA Strand: The
DNA strand (15 nmol) was dissolved in a 300 µL of a 0.1 
NaHCO3 aqueous solution. One of the diphenylphosphanylbenzoic
acids (2.3 mg, 7.5 µmol, 500 equiv.), EDC·HCl (1 mg, 9 µmol,
600 equiv.), and NHS (1.4 mg, 7.5 µmol, 500 equiv.) were dissolved
in 600 µL of DMF and stirred for 3 h. The activated phosphane
solution was added to the DNA solution and stirred for 12 h.
600 µL of water were added to the reaction mixture, which was
washed with dichloromethane (3ϫ2 mL). The aqueous layer was
concentrated in vacuo and redissolved in 1 mL of water. 200 µL of
this solution were placed in an Eppendorf vial. 20 µL of a 3 
NaOAc aqueous solution and 700 µL of cold EtOH were added to
the DNA solution which was cooled to 0 °C for 1 h. It was centri-
fuged for 30 min at 4 °C at 13000 rpm. The supernatant was re-
moved and discarded, a cold 70% aqueous EtOH solution was
added and the sample was centrifuged for another 30 min. Removal
of the supernatant left a pellet of purified DNA.
[9] P. Fournier, R. Fiammengo, A. Jäschke, Angew. Chem. Int. Ed.
2009, 48, 4426.
[10] L. Ropartz, N. J. Meeuwenoord, G. A. van der Marel,
P. W. N. M. van Leeuwen, A. M. Z. Slawin, P. C. J. Kamer,
Chem. Commun. 2007, 1556.
[11] N. S. Oltra, G. Roelfes, Chem. Commun. 2008, 6039.
[12] G. Roelfes, A. J. Boersma, B. L. Feringa, Chem. Commun.
2006, 635.
[13] G. Roelfes, Mol. BioSyst. 2007, 3, 126.
[14] A. J. Boersma, B. L. Feringa, G. Roelfes, Angew. Chem. Int.
Ed. 2009, 48, 3346.
[15] F. Rosati, A. J. Boersma, J. E. Klijn, A. Meetsma, B. L. Fer-
inga, G. Roelfes, Chem. Eur. J. 2009, 15, 9596.
[16] R. P. Megens, G. Roelfes, Org. Biomol. Chem. 2010, in press.
[17] A. J. Boersma, J. E. Klijn, B. L. Feringa, G. Roelfes, J. Am.
Chem. Soc. 2008, 130, 11783.
[18] K. Sakurai, T. M. Snyder, D. R. Liu, J. Am. Chem. Soc. 2005,
127, 1660.
[19] R. M. Franzini, E. T. Kool, ChemBioChem 2008, 9, 2981.
[20] K. Furukawa, H. Abe, J. Wang, M. Uda, H. Koshino, S. Tsu-
neda, Y. Ito, Org. Biomol. Chem. 2009, 7, 671.
[21] M. Caprioara, R. Fiammengo, M. Engeser, A. Jäschke, Chem.
Eur. J. 2007, 13, 2089.
[22] F. W. Hobbs, J. Org. Chem. 1989, 54, 3420.
[23] S. E. Lee, A. Sidorov, T. Gourlain, N. Mignet, S. J. Thorpe,
J. A. Brazier, M. J. Dickman, D. P. Hornby, J. A. Grasby, D. M.
Williams, Nucleic Acid Res. 2001, 29, 1565.
[24] N. K. Garg, C. C. Woodroofe, C. J. Lacenere, S. R. Quake,
B. M. Stoltz, Chem. Commun. 2005, 4551.
General Procedure for Allylic Alkylation of 1,3-Diphenylallyl Acet-
ate with Dimethyl Malonate: A mixture of [Pd(η3-C3H5)Cl]2
(0.5 µmol) and ligand (2 µmol) were dissolved in DCM, and stirred
for 30 min. A trace of KOAc, 1,3-diphenylallyl acetate (100 µmol),
diphenyl ether (100 µmol), dimethyl malonate (300 µmol) and BSA
(300 µmol) were added to the metal complex solution in this order.
The reaction mixture was stirred for 24 h. Each solution was fil-
tered through SiO2 and left to evaporate overnight. A CHIRACEL
OD-H chiral column (Daicel Chemical Industry Ltd.) was em-
ployed using a 99:1 (hexane/iPrOH) isocratic mixture with a flow
of 1 mL/min; tR(S) 23.6 min, tR(R) 24.3 min.
[25] S. E. Lee, J. S. Vyle, D. M. Williams, J. A. Grasby, Tetrahedron
Lett. 2000, 41, 267.
[26] C. A. G. N. Montalbetti, V. Falque, Tetrahedron 2005, 61,
10827.
[27] P. R. Hardwidge, D. Lee, T. P. Prakash, B. Iglesias, R. B. Den,
C. Switzer, L. J. Maher III, Chem. Biol. 2001, 8, 967.
[28] J. A. Brazier, T. Shibata, J. Townsley, B. F. Taylor, E. Frary,
N. H. Williams, Nucleic Acid Res. 2005, 33, 1362.
General Procedure for Allylic Amination of 1,3-Diphenylallyl Acet-
ate with Benzylamine: A mixture of [Pd(η3-C3H5)Cl]2 (0.5 µmol)
and ligand (2 µmol) were dissolved in DCM, and stirred for 30 min.
1,3-Diphenylallyl acetate (100 µmol), diphenyl ether (100 µmol)
and benzylamine (300 µmol) were added to the metal complex solu-
Eur. J. Org. Chem. 2010, 3229–3236
© 2010 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.eurjoc.org
3235

P. C. J. Kamer et al.
FULL PAPER
[29]
[31] T. R. Battersby, D. N. Ang, P. Burgstaller, S. C. Jurczyk, M. T.
Bowser, D. D. Buchanan, R. T. Kennedy, S. A. Benner, J. Am.
Chem. Soc. 1999, 121, 9781.
[32] E. J. Trybulski, J. Zhang, R. H. Kramss, R. M. Mangano, J.
Med. Chem. 1993, 36, 3533.
M. C. de Koning, A. B. T. Ghisaidoobe, H. I. Duynstee,
P. B. W. T. Kortenaar, D. V. Filippov, G. A. van der Marel, Org.
Process Res. Dev. 2006, 10, 1238.
[30] N. Lin, J. Yan, Z. Huang, C. Altier, M. Li, N. Carrasco, M.
Suyemoto, L. Johnston, S. Wang, Q. Wang, H. Fang, J. Caton-
Williams, B. Wang, Nucleic Acid Res. 2007, 35, 1222.
Received: January 29, 2010
Published Online: March 23, 2010
3236
www.eurjoc.org
© 2010 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Eur. J. Org. Chem. 2010, 3229–3236