Novel method for the immobilization of nucleotides.

Article abstract of DOI:10.1021/ol0273607

[reaction: see text] A novel method for the immobilization of nucleotides has been developed. The strategy employs a highly reactive pyrrolidinium phosphoramidate zwitterion intermediate that undergoes nucleophilic attack by long-chain alkylamine-controlled pore glass (LCAA-CPG) to generate an immobilized nucleotide. Quantification of nucleotide loading was accomplished by acidic hydrolysis of the P-N bond and subsequent HPLC analysis of TMP in the presence of an internal standard. Typical nucleotide loadings of 51-59 micromol/g of support were observed.

Full text of DOI:10.1021/ol0273607

ORGANIC
LETTERS
2
003
Vol. 5, No. 3
41-344
Novel Method for the Immobilization of
Nucleotides
3
†
Caren L. Freel Meyers and Richard F. Borch*
Department of Medicinal Chemistry and Molecular Pharmacology and Cancer Center,
Purdue UniVersity, West Lafayette, Indiana 47907
rickb@pharmacy.purdue.edu
Received November 27, 2002
ABSTRACT
A novel method for the immobilization of nucleotides has been developed. The strategy employs a highly reactive pyrrolidinium phosphoramidate
zwitterion intermediate that undergoes nucleophilic attack by long-chain alkylamine-controlled pore glass (LCAA-CPG) to generate an immobilized
nucleotide. Quantification of nucleotide loading was accomplished by acidic hydrolysis of the P−N bond and subsequent HPLC analysis of
TMP in the presence of an internal standard. Typical nucleotide loadings of 51−59 µmol/g of support were observed.
Numerous methods exist for the attachment of nucleotides
to solid supports; however, the extension of these methods
to the immobilization of modified nucleotides has, in many
cases, demanded alteration of standard solid-phase coupling
ester (1). Cleavage of the nucleotide from the solid support
can be accomplished via acid-catalyzed hydrolysis of the
P-N bond, a potentially desirable alternative to the standard
base-catalyzed cleavage conditions employed in the prepara-
1
and cleavage conditions. Recently, we reported a new
4 3 2
tion of oligonucleotides (concentrated NH OH/CH NH )
4
method for the preparation of nucleoside diphosphates that
employs a highly electrophilic zwitterionic phosphoramidate
using automated DNA synthesis. Furthermore, this method
may provide an attractive means for the preparation of unique
DNA analogues such as DNA-peptide conjugates that might
otherwise require careful consideration for timing of depro-
tection of multiple base-sensitive groups and base-catalyzed
2
intermediate (3) as the phosphorylating reagent (Scheme 1).
The reactivity of the zwitterionic intermediate toward nu-
cleophiles under both aqueous and anhydrous conditions has
been investigated2,3 and has inspired the extension of this
method to the immobilization of nucleotides and other
phosphates. The approach relies upon the assumption that a
primary amine attached to controlled pore glass (CPG) will
undergo nucleophilic attack at phosphorus of zwitterionic
intermediate 3 following activation of the phosphoramidate
5
cleavage from the support. The synthesis of biopolymers
1
c
using photolytic cleavage has provided an additional
method for the production of such molecules under neutral
conditions; however, deprotection via photolysis, although
optimized for the cleavage of oligonucleotides bearing
standard bases and protecting groups, still poses potential
problems that arise from UV-induced damage to modified
biopolymers. Our efforts to develop a novel method for the
attachment of nucleotides to solid supports have resulted in
generation of CPG derivatized with a thymidine phosphor-
amidate. We report herein the coupling of a chlorobutyl
†
Current address: Department of Biological Chemistry and Molecular
Pharmacology, Harvard Medical School, Boston, MA 02115; caren_meyers@
hms.harvard.edu.
(
1) (a) Lyttle, M. H.; Hudson, D.; Cook, R. M. Nucleosides Nucleotides
1
999, 8, 1809-1824. (b) Lyttle, M. H.; Dick, D. J.; Hudson, D.; Cook, R.
M. Nucleic Acids Res. 1996, 24, 2793-2798. (c) Venkatesan, H.; Greenberg,
M. J. Org. Chem. 1996, 61, 525-529. (d) Eritja, R.; Robles, J.; Fernandez-
Forner, D.; Alberico, F.; Giralt, E.; Pedroso, E. Tetrahedron Lett. 1991,
(4) (a) Boal, J. H.; Wilk, A.; Harindranath, N.; Max, E. E.; Kempe, T.;
Beaucage, S. L. Nucl. Acids Res. 1996, 24, 3115-3117. (b) Reddy, M. P.;
Michael, M. A.; Farooqui, F.; Girgis, S. Tetrahedron Lett. 1994, 35, 5771-
4314.
3
2, 1511-1514.
(
2) Freel Meyers, C. L.; Borch, R. F. Org. Lett. 2001, 3, 3765-3768.
3) (a) Freel Meyers, C. L.; Borch, R. F. J. Med. Chem. 2000, 43, 4319-
(
4
327. (b) Tobias, S. C.; Borch, R. F. J. Med. Chem. 2001. 44, 4475-4480.
(5) Basu, S. and Wickstrom, E. Tetrahedron Lett. 1995, 36, 4943-4946.
1
0.1021/ol0273607 CCC: $25.00 © 2003 American Chemical Society
Published on Web 01/15/2003

genolysis. The resulting crude hydrogenolysis mixture was
then added to LCAA-CPG, and the suspension was tumbled
overnight in the presence of excess triethylamine. The
supernatant was removed following centrifugation, and the
solid support was washed several times with water and
acetonitrile to remove residual phosphoramidate and phos-
phate byproducts.
Scheme 1
To assess the loading of immobilized phosphoramidate on
the solid support, thymidine 5′-monophosphate (TMP) was
cleaved using acid-catalyzed hydrolysis and quantified by
HPLC analysis. Briefly, CPG derivatized with thymidine
phosphoramidate (5) was tumbled in the presence of aqueous
acid. Cyclic TMP (internal standard) was added to the
suspension following acidic cleavage, and the supernatant
containing TMP and cyclic TMP was removed following
centrifugation and added to ammonium hydroxide. The
resulting mixture was immediately analyzed by HPLC (98:2
0
2 3 t
.1% TFA in H O/CH CN; TMP R ) 7.9 min, cyclic TMP
R
t
) 10.4 min), and the concentration of TMP was
determined from the ratio cyclic TMP:TMP calculated from
HPLC peak areas. Acidic hydrolysis and HPLC analysis of
the TMP/cyclic TMP mixture was repeated until no ad-
ditional TMP was cleaved from the solid support.
The cleavage of TMP from the solid support was studied
using a variety of aqueous acidic conditions. The cleavage
reaction times and TMP loading are presented in Table 1.
thymidine phosphoramidate to CPG using various phosphor-
amidate ester activation conditions and subsequent nucleotide
cleavage from CPG using optimized aqueous acidic condi-
tions.
CPG coupling reactions were carried out using chlorobutyl
phosphoramidic acid 2 (ca. 120-fold excess, Scheme 2) and
Table 1. Aqueous Acidic Cleavage of TMP from Derivatized
CPG
loading
reaction
conditionsa
reaction
timeb
(µmol/g of
support)
c
entry
1
2
3
25% v/v AcOH/H2O, rt
25% v/v Cl2CHCO2H/H2O, rt
50% v/v formic acid/H2O, rt
5 days
<3 h
8 h
51-53
56-59
53-55
Scheme 2a
a
TMP was stable under all conditions tested. b Reaction times were
determined on the basis of the total time taken to cleave all TMP from the
support. A procedure for the determination of the cleavage rate is described
in Supporting Information. c Minimum of two CPG coupling reactions were
analyzed in each case.
The TMP loading in each case is higher than typical
nucleoside loading on CPG used for the automated synthesis
6
of oligonucleotides (30-40 µmol/g CPG). The shortest
cleavage reaction time for the aqueous acidic conditions
considered is 3 h in 1:3 dichloroacetic acid/water (entry 2).
Treatment of the derivatized CPG with the less toxic 1:1
formic acid/water (entry 3) resulted in complete cleavage of
TMP over a longer time period (8 h), but cleavage was more
rapid under these conditions relative to cleavage carried out
in 1:3 acetic acid/water (5 days, entry 1). For circumstances
in which quantitative recovery of nucleotide from the solid
support is not required, shorter cleavage times can be used.
For example, cleavage is 61, 91, and 98% complete after
a
Reaction conditions: (a) H
2
, Pd/C, THF; (b) TEA, LCAA-CPG,
CH
3
CN/H O, rt, overnight; (c) 50:50 formic acid/water, rt, 8 h.
2
long-chain alkylamino-controlled pore glass (LCAA-CPG).
Chlorobutyl phosphoramidate ester 4 is a stable precursor
to the reactive phosphorylating agent 2 and was used
previously in our laboratory for the preparation of TDP-D-
2
glucose and TDP-L-rhamnose. Activation of benzyl ester 4
(6) Current Protocols in Nucleids Acid Chemistry; Harkins, E. W., Ed.;
to phosphoramidic acid 2 was accomplished via hydro-
Wiley & Sons: New York; 2001, p 3.2.20.
342
Org. Lett., Vol. 5, No. 3, 2003

1.5, 3, and 5 h, respectively, in formic acid/water. It should
be noted that TMP is stable under all cleavage conditions
tested. In addition, the stability of a model oligonucleotide
Scheme 3a
(5′-TGACTGACTGACTGACTGAC-3′) was monitored in
1:3 formic acid/water; <5% of the oligonucleotide had
decomposed after 3 h at room temperature.
Additional investigation of the CPG coupling reaction was
carried out for the purpose of defining a broader set of
reaction conditions that might be employed for the im-
mobilization of oligonucleotides and other biomolecules.
Thus, CPG coupling reactions were carried out under
different aqueous/organic conditions and with varying ratios
of chlorobutyl phosphoramidate 2 and solid support. The
results are summarized in Table 2. Comparable loading of
Table 2. Variation of CPG Coupling Reaction Conditions
loading
ratio of
entry 2:CPG-NH
reaction
conditions
(µmol/g of
a
b
2
support)
1
2
3
4
5
120
120
120
50
TEA (3 equiv), CH
rt, overnight
TEA (3 equiv), 95:5 CH
rt, overnight
TEA (3 equiv), 50:50 CH
rt, overnight
TEA (3 equiv), 95:5 CH
rt, overnight
3
CN,
54-57
54-55
49-53
44-48
40-42
3
CN/H
2
O,
3
CN/H
2
O,
3
CN/H
2
2
O,
5
TEA (3 equiv), 95:5 CH
rt, overnight
3
CN/H
O,
a
CPG-NH2 loading was ∼80 µmol/g of support (5 mg of support used
b
in each case). Minimum of two CPG coupling reactions were analyzed in
each case.
a
Reaction conditions: (a) LCAA-CPG, propylamine, TEA,
CH
h.
3 3
CN, rt, overnight; (b) LCAA-CPG, DBU, CH CN, rt, 12-35
TMP was observed from both partial aqueous and anhydrous
coupling reactions (entries 1-3). Furthermore, TMP loading
was reduced by only ∼20% when the amount of phospho-
ramidate coupling reagent was decreased from a 120-fold
excess to a 5-fold excess (entries 2, 4, and 5). The unexpected
efficiency of nucleotide loading observed for CPG coupling
reactions carried out in aqueous solvent suggests that this
methodology for the immobilization of phosphates can be
extended to the attachment of oligonucleotides and other
biomolecules.
phosphoramidate 4. The activation of 6 (Scheme 3) in a
mixture of acetonitrile and DBU (10 equiv) was carried out
in the presence of LCAA-CPG. The TMP loading observed
in this case (entry 2, Table 3) was somewhat lower than the
TMP loading observed in CPG coupling reactions carried
out following activation via hydrogenolysis but comparable
to nucleotide loading on CPG used for automated DNA
synthesis (30-40 µmol/g). Coupling reactions carried out
in the presence of 1.1 equiv of DBU (entry 1, Table 3)
afforded TMP loading similar to those obtained in reactions
carried out in the presence of 10 equiv DBU. This method
of phosphoramidate activation/immobilization is attractive
because it precludes the need for multiple reaction vessels
and is a viable alternative to activation via hydrogenolysis
for the immobilization of phosphates.
Finally, alternative methods of phosphoramidate activation
have been investigated to identify potential activation condi-
tions for the immobilization of phosphoramidate biomol-
ecules that are sensitive to hydrogenolysis conditions or that
require aqueous activation conditions. The nitrophenylethyl
moiety has been employed previously for the protection of
7
phosphate groups and can be removed via a base-catalyzed
â-elimination reaction with DBU (Scheme 3). Use of the
nitrophenylethyl substituent as an activating group for the
liberation of the reactive phosphoramidate anion 2 circum-
vents the requirement for activation via hydrogenolysis and
provides an alternative method for the immobilization of
sensitive biomolecules. Thus, model phosphoramidate 6 was
synthesized according to the procedure used for benzyl
The quinone activating group of phosphoramidate 7
(Scheme 3) was initially designed to undergo bioreductive
activation for the intracellular delivery of alkylating agents8
and more recently has been used for the intracellular delivery
of nucleotides. During the course of this work, model
phosphoramidate 7 was found to undergo rapid activation
3
via 1,4-addition of propylamine to the quinine moiety
(
7) Ohtsuka, E.; Morioka, S.; Ikehara, M. J. Am. Chem. Soc. 1973, 95,
(8) Flader, C.; Liu, J.; Borch R. F. J. Med. Chem. 2000, 43, 3157-
3167.
8
437.
Org. Lett., Vol. 5, No. 3, 2003
343

Table 3. Dependence of TMP Loading on Method of Phosphoramidate Activation
starting
reaction
loading
(µmol/g of support)a
entry
phosphoramidate
conditions
1
2
3
4
6
6
7
2
DBU (1.1 equiv), CH3CN, rt, overnight
DBU (10 equiv), CH3CN, rt, overnight
H2N(CH2)2CH3 (1.05 equiv), TEA (2 equiv), CH3CN, rt, overnight
TEA (3 equiv), CH3CN, rt, overnight
29-33
25-27
33
53-55
a
A minimum of two CPG coupling reactions was analyzed in each case.
5
followed by elimination of the phosphoramidate anion, a
reaction that can be carried out under both aqueous and
nonaqueous conditions. Thus, a CPG coupling reaction was
carried out using phosphoramidate 7 in the presence of
LCAA-CPG, propylamine and triethylamine (Table 3). In
this case, a loading value comparable to that observed with
the nitrophenethyl phosphoramidate 6 was observed.
In summary, a novel method for the attachment of
nucleotides to solid supports has been developed. The
strategy described here employs a highly reactive phospho-
rylating reagent used previously in this laboratory for the
intracellular delivery of nucleotides and for the synthesis of
sugar nucleoside diphosphates. Aqueous acidic conditions
have been identified such that the intact nucleotide can be
cleaved from the solid support prior to the determination of
nucleotide loading by HPLC. The nucleotide cleavage
conditions employed in this strategy may be advantageous
under circumstances in which the immobilized phosphate is
sensitive to a strong base. TMP loading of 51-59 µmol/g
of support was observed in most cases, offering an increase
over the typical nucleoside loading on CPG used for
oligonucleotide synthesis (30-40 µmol/g of CPG). Alterna-
tive activation methods were explored and afforded some-
what lower loading of nucleotide when compared to coupling
reactions carried out following activation via hydrogenolysis.
Finally, preliminary studies indicate that this strategy for the
attachment of nucleotides to CPG may be extended to the
attachment of oligonucleotides and other phosphate biomol-
ecules. Efforts are underway to explore the scope of this
methodology.
Acknowledgment. Financial support from the National
Cancer Institute (CA34619) is gratefully acknowledged.
Supporting Information Available: Experimental pro-
cedures for phosphoramidate coupling of 2, 6, and 7 to
LCAA-CPG, cleavage of nucleotide from CPG, and deter-
mination of nucleotide loading by HPLC. This material is
available free of charge via the Internet at http://pubs.acs.org.
OL0273607
344
Org. Lett., Vol. 5, No. 3, 2003