Synthesis of oligonucleotides with 3′-terminal 5-(3- acylamidopropargyl)-3′-amino-2′,3′-dideoxyuridine residues and their reactivity in single-nucleotide steps of chemical replication

Article abstract of DOI:10.1021/jo052097j

Oligonucleotides with a 3′-terminal 5-alkynyl-3′-amino- 2′,3′-dideoxyuridine residue were prepared, starting from 2′-deoxyuridine. The optimized route employs a 2′,3′-dideoxy- 3′-trifluoroacetamido-5-iodouridine 5′-phosphoramidite as building block for DN

Full text of DOI:10.1021/jo052097j

Synthesis of Oligonucleotides with 3′-Terminal
5-(3-Acylamidopropargyl)-3′-amino-2′,3′-dideoxyuridine Residues
and Their Reactivity in Single-Nucleotide Steps of Chemical
Replication
Patrick Baumhof, Niels Griesang, Michael Ba¨chle, and Clemens Richert*
Institute for Organic Chemistry, UniVersity of Karlsruhe (TH), D-76131 Karlsruhe, Germany
cr@rrg.uka.de
ReceiVed October 6, 2005
Oligonucleotides with a 3′-terminal 5-alkynyl-3′-amino-2′,3′-dideoxyuridine residue were prepared, starting
from 2′-deoxyuridine. The optimized route employs a 2′,3′-dideoxy-3′-trifluoroacetamido-5-iodouridine
5′-phosphoramidite as building block for DNA synthesis and involves on-support Sonogashira coupling
with N-tritylpropargylamine to generate oligonucleotides. The amino group of the propargylamine side
chain was acylated to accelerate primer extension reactions involving the 3′-amino group. Three acyl
groups were identified that decrease the half-life for DNA-templated extension steps with 7-azabenzo-
triazole esters of 2′-deoxynucleotides. The residue of 4-pyrenylbutyric acid was found to accelerate primer
extension reactions and to render them more selective than those of the control primer. With this substituent,
primer extension is also faster than previously measured for three-strand systems involving template,
aminoprimer, and a downstream-binding helper oligonucleotide. Fast-reacting primers might become
useful for genotyping single nucleotides.
Introduction
Primer extension is also the basis of key biomedical applications,
such as PCR and dideoxy DNA sequencing. Both in nature and
in biomedical applications, primer extension requires poly-
merases as catalysts. Studying non-enzymatic alternatives to
primer extension promises insights that might eventually lead
to chemical versions of the polymerase chain reaction (PCR),
the creation of self-replicating systems, and less expensive
methods for determining DNA sequences.
Non-enzymatic replication has been studied for a number of
years, and certain sequences of RNA have been shown to
support the formation of complementary strands from activated
mononucleotides.2-4 Chemical primer extension has also been
Replication is a pivotal reaction of life that precedes cell
division. Bacterial cells achieve high-fidelity replications of
entire genomes in under 20 min by enzymatically reacting
nucleoside triphosphates with the terminal hydroxyl groups of
growing DNA chains at a rate of approximately 1000 nucleotides
per second.1 Replication is based on template-directed oligo-
merization steps in which a primer is extended by one
mononucleotide at a time. Primers are short strands of DNA or
RNA that act as starting points for template-directed syntheses.
* To whom correspondence should be addressed. Tel.: 49 (0) 721 608 2091.
Fax: 49 (0) 721 608 4825.
(1) Voet, D.; Voet, J. G.; Pratt, C. W. Fundamentals of Biochemistry; J.
Wiley & Sons: New York, 1999.
(2) Sulston, J.; Lohrmann, R.; Orgel, L. E.; Miles, H. T. Proc. Natl.
Acad. Sci. U.S.A. 1968, 59, 726-733.
10.1021/jo052097j CCC: $33.50 © 2006 American Chemical Society
Published on Web 01/11/2006
1060
J. Org. Chem. 2006, 71, 1060-1067

Synthesis of Capped Primers
FIGURE 1. Chemical primer extension and primer with cap.
demonstrated for nucleic acids with modified backbones,5-7 or
nucleobases,8 but neither has entered practical applications.
Chemical ligation of short oligomers on a DNA or RNA
template,9-13 has reached a level of efficiency in generating a
complementary strand that makes this reaction attractive for
practical applications and for generating self-replicating sys-
tems.14 To make non-enzymatic primer extension more suitable
for diagnostic applications, however, an acceleration of the
underlying reaction is needed.
To increase the reactivity toward activated mononucleotides,
primers can be equipped with an amino group at their terminus.
Amines are known to react rapidly with activated nucleotides
in template-directed reactions in aqueous solutions.15-17 Further,
leaving groups other than the well-established 2-methylimid-
azolides18 have recently been shown to lead to rate acceleration
by up to 2 orders of magnitude.19 A complementary approach
to accelerate individual steps of chemical replication steps uses
oligonucleotides whose termini are modified with substituents
that assist in the capturing and/or conversion of monomers,
mimicking, in a modest form, the effect of polymerases on
primer extension in the cell. Such an effect has been observed
for “capped” templates.20 For practical applications, however,
it is desirable to achieve this effect without the need to modify
the template, so that DNA from natural sources can be
employed. This calls for modified primers, e.g. in the form of
oligonucleotides featuring both a terminal amino group and a
cap that can assist in reactions with incoming nucleotides (Figure
1). Here, we report on the synthesis of such oligonucleotides
and their increased reactivity in nonenzymatic primer extension
reactions. Rapid single-nucleotide extension reactions allow for
genotyping single-nucleotide polymorphisms with mass spec-
trometry as read-out technique.21-23
(3) Kozlov, I. A.; Orgel, L. E. Mol. Biol. 2000, 34, 781-789.
(4) Hey, M.; Hartel, C.; Go¨bel, M. W. HelV. Chim. Acta 2003, 86, 844-
854.
Results
Oligonucleotides with a 3′-terminal 3′-amino-2′,3′-dideoxy-
uridine residue featuring a substituent at position 5 (1, Figure
1) were the target molecules of our syntheses. Alkynyl groups
at position 5 of pyrimidines are known to stabilize duplexes.24,25
We chose propargylamine as the substituent at the 5 position
of deoxyuridine, as the amino function allows for amide-forming
reactions, so that the large pool of available carboxylic acids
can be accessed in the search for caps that accelerate non-
enzymatic primer extension reactions. Because amide-forming
reactions involving oligonucleotides on controlled pore glass
(cpg) are easiest to perform when the amino group is on the
distal nucleotide of the immobilized chain,26 we opted for DNA
syntheses with 5′- or “reverse” phosphoramidites.27,28
(5) Koppitz, M.; Nielsen, P. E.; Orgel, L. E. J. Am. Chem. Soc. 1998,
120, 4563-4569.
(6) Kozlov, I. A.; Politis, P. K.; Pitsch, S.; Herdewijn, P.; Orgel, L. E.
J. Am. Chem. Soc. 1999, 121, 1108-1109.
(7) Wu, X.; Delgado, G.; Krishnamurthy, R.; Eschenmoser, A. Org. Lett.
2002, 4, 1283-1286.
(8) Chaput, J. C.; Switzer, C. J. Am. Chem. Soc. 2000, 122, 12866-
12867.
(9) Gryaznov, S. M.; Letsinger, R. L. J. Am. Chem. Soc. 1993, 115,
3808-3809.
(10) Gryaznov, S. M.; Schultz, R.; Chaturvedi, S. K.; Letsinger, R. L.
Nucleic Acids Res. 1994, 22, 2366-2369.
(11) Sando, S.; Kool, E. T. J. Am. Chem. Soc. 2002, 124, 9686-9687.
(12) Sando, S.; Abe, H.; Kool, E. T. J. Am. Chem. Soc. 2004, 126, 1081-
1087.
(13) Ficht, S.; Mattes, A.; Seitz, O. J. Am. Chem. Soc. 2004, 126, 9970-
9981.
(14) Zielinski, W. S.; Orgel, L. E. Nature 1987, 327, 346-347. (b)
Sievers, D.; von Kiedrowski, G. Nature 1994, 369, 221-224. (c) Li, T.;
Nicolaou, K. C. Nature 1994, 369, 218-221.
(20) Rojas Stu¨tz, J. A.; Richert, C. J. Am. Chem. Soc. 2001, 123, 12718-
12719.
(15) Lohrmann, R.; Orgel, L. E. Nature 1976, 261, 342-344.
(16) Zielinski, W. S.; Orgel, L. E. Nucleic Acids Res. 1987, 15, 1699-
1715.
(21) Tost, J.; Gut, I. G. Clin. Biochem. 2005, 38, 335-350.
(22) Stoerker, J.; Mayo, J.; Tetzlaff, C. N.; Sarracino, D. A.; Schwope,
I.; Richert, C. Nat. Biotechnol. 2000, 18, 1213-1216.
(23) Tang, K.; Fu, D.-J.; Julien, D.; Braun, A.; Cantor, C. E. Ko¨ster, H.
Proc. Natl. Acad. Sci. U.S.A. 1999, 96, 10016-10020.
(24) Wagner, R. W.; Matteucci, M. D.; Lewis, J. G.; Gutierrez, A. J.;
Moulds, C.; Froehler, B. C. Science 1993, 260, 1510-1513.
(25) Chaput, J. C.; Sinha, S.; Switzer, C. Chem. Commun. 2002, 1568-
1569.
(17) Kanavarioti, A.; Stronach, M. W.; Ketner, R. J.; Hurley, T. B. J.
Org. Chem. 1995, 60, 632-637.
(18) Inoue, T.; Orgel, L. E. J. Am. Chem. Soc. 1981, 103, 7666-7667.
(b) Joyce, G. F.; Inoue, T.; Orgel, L. E. J. Mol. Biol. 1984, 176, 279-306.
(c) Kozlov, I. A.; Orgel, L. E. Mol. Biol. 2000, 34, 781-789.
(19) Hagenbuch, P.; Kervio, E.; Hochgesand, A.; Plutowski, U.; Richert,
C. Angew. Chem. 2005, 117, 6746-6750; Angew. Chem., Int. Ed. Engl.
2005, 44, 6588-6592.
(26) Connors, W. H.; Narayanan, S.; Kryatova, O. P.; Richert, C. Org.
Lett. 2003, 5, 247-250.
J. Org. Chem, Vol. 71, No. 3, 2006 1061

Baumhof et al.
SCHEME 1
Several routes differing in the choice of reagents, protecting
groups, and order of events were tested. The shortest route,
involving a 3′-azide as the latent amine and no protecting groups
at this position up to the oligonucleotide level, proved incompat-
ible with phosphoramidite chemistry. After phosphitylation,
decomposition, probably via intramolecular Staudinger reduc-
tion, set in before coupling to the remainder of the DNA chain
could be achieved. Attempts to introduce the 3′-azido function-
ality via ring opening of 5-iodo-2,3′-anhydrouridine were also
unsuccessful, probably because of the high reactivity of the iodo
functionality. Schemes 1 and 2 show routes that led to full-
length DNA chains on solid support. Either of them starts from
2′-deoxyuridine, which was converted to 2′-deoxy-2,3′-anhydro-
uridine in 85% yield, following literature protocols (Scheme
1).29 Opening of the anhydrouridine with sodium azide in DMF
gave 2,30 and subsequent iodination of position 5 with CAN
and iodine31 gave nucleoside 3 in 56% yield over two steps.
Reduction of the azide proved more troublesome. Hydrogenation
with palladium or platinum catalysts failed to reduce the azide.
Instead, deiodination was observed without reduction of the
azide (or the 5,6 double bond), even after long reaction times.
Staudinger reduction yielded amine 4 in 67% yield, but scale-
up and hydrolysis of the iminophosphorane proved difficult,
and the purification remained cumbersome. Tin dichloride in
water/THF32 reacted smoothly to give amine 4 in 89% yield.
This reaction was also suitable for scale-up to the gram scale.
The route then involved protection of the 3′-amino group with
a trityl group to give 5, followed by Sonogashira coupling with
N-Alloc-protected propargylamine33 to yield 6 (Scheme 1). The
latter reaction gave variable yields, caused by incomplete
conversion, making this route unsatisfactory. Subsequent de-
benzoylation led to 7 in moderate yield, but phosphitylation to
8 proceeded uneventfully. Although the coupling of 8 to the
3′-terminal hydroxyl group of DNA chains was high yielding,
as evidenced by MALDI-TOF mass spectrometry of crude
products obtained by deprotecting analytical samples with
aqueous ammonia, the subsequent removal of the Alloc group
again proceeded unsatisfactorily, with low conversion or
decomposition when attempted under forcing conditions.
Scheme 2 shows the synthesis of a phosphoramidite that
proved more reliable. The Sonogashira coupling was performed
on cpg support,34 and protecting groups were changed to shorten
the route. A trifluoroacetyl group was used to protect 3′-amine
4, giving 9 in reasonable yield. The TFA group later releases
the free amine during the final deprotection step of the DNA.
Debenzoylation of 9 to 10 proceeded without concomitant loss
of the TFA group when 1.5 equiv of 1 M sodium hydroxide
solution in MeOH was used for 1 h at ambient temperature.
Conversion of 10 to phosphoramidite 11 using 2-cyanoethyl-
N,N-diisopropyl-chlorophosphoramidite as the phosphitylating
reagent gave the latter in 80% yield under unoptimized
conditions.
SCHEME 2
A DNA hexamer of the sequence 5′-GCCACG-3′ was
prepared, starting from cpg loaded with a deoxyguanosine
residue (Scheme 3). After incorporation of the 5-iodonucleoside
as the last monomer during automated DNA synthesis, giving
12, on-support Sonogashira reaction34 produced 13. Trityl-
protected propargylamine,35 together with Pd(PPh3)4, CuI, and
diisopropylethylamine (DIEA),36 gave 80-96% conversion after
4 h at room temperature, as detected in MALDI-TOF mass
(32) Lyssenko, K., A.; Lenev, D. A.; Kostyanovsky, R. G. Tetrahedron
2002, 58, 8525-8537.
(33) Brotzel, F. Diploma Thesis, University of Konstanz, Konstanz,
Germany, 2002; see also Supporting Information.
(34) Khan, S. I.; Grinstaff; M. W. J. Am. Chem. Soc. 1999, 121, 4704-
4705.
(35) (a) Koziara, A.; Osowska-Pacewicka, K.; Zawadzki, S.; Zwierzak,
A. Synthesis 1985, 202-204. (b) Vassiliki, T.; Valentine, R.; Alexandros,
S.; Evangelos, Z.; Argyro, E.; Dimitriou, M. Tetrahedron Lett. 2005, 46,
1357-1360.
(36) Kottysch, T.; Ahlborn, C.; Brotzel, F.; Richert, C. Chem. Eur. J.
2004, 10, 4017-4028.
(27) Uhlmann, E.; Engels, J. Tetrahedron Lett. 1986, 27, 1023-1026.
(28) Connors, W. H.; Narayanan, S.; Kryatova, O. P.; Richert, C. Org.
Lett. 2003, 5, 247-250.
(29) Naimi, E.; Zhou A.; Khalili, P.; Wiebe, L. I.; Balzarini, J.; De Clercq,
E.; Knaus E. J. Med. Chem. 2003, 46, 995-1004.
(30) Czernecki, S.; Vale´ry, J.-M. Synthesis 1991, 239-240.
(31) Asakura, J.; Robins, J. R. J. Org. Chem. 1990, 55, 4928-4933.
1062 J. Org. Chem., Vol. 71, No. 3, 2006

Synthesis of Capped Primers
SCHEME 3
spectra of the crude products. The trityl group of 13 was then
removed with the trichloroacetic acid solution in dichlo-
romethane regularly employed for deblocking during DNA chain
assembly. Amine (14) was then coupled to active esters of
carboxylic acids, generating, after ammonia treatment, primers
with an amidic cap (general structure 15, Scheme 3).
â-Ala and â-Ala-Gly-linked constructs, as the ability of this
aromatic moiety to stack on terminal base pairs is well
documented.40 To add diversity, nalidixic acid was coupled to
the â-Ala platform, and anthraquinone carboxylic acid was
coupled to the â-Ala-Gly platform. Again, either of the acyl
residues is known to stabilize duplexes when appended to the
terminus of one strand.41 In addition to cap-bearing primers 16-
21, which were obtained in overall yields between 15% and
41%, unmodified primer 22 was prepared as a control via
standard DNA synthesis to a 3′-amino-3′-deoxyprimer.19,20
With oligonucleotides 16-22 in hand, exploratory primer
extension reactions of the type shown in Figure 1 were
performed. The assays employed azabenzotriazole esters of
deoxynucleotides (25a-t) as activated monomers in aqueous
buffer under conditions recently shown to give rapid reactions
with 3′-amino-terminal primers.19 Two different templates, one
with weakly base-pairing adenine (23) and one with strongly
base-pairing cytosine (24) as the templating base, were employed
(Scheme 4). The A-templated reaction is the one giving the
slowest reactions among the four different template bases (A,
C, G, and T).19 Exploratory assays showed that primers 17-19
do not undergo faster reactions with 25a-t in the presence of
23 than control 22. The remaining three primers (16, 20, and
21) exceeded 22 in reactivity and were selected for kinetic
analysis.
Six different target oligonucleotides with caps were prepared
to test the methodology. Two of these were generated through
direct acylation of 14 with the respective carboxylic acids, two
involved installation of a â-alanine linker, and two required a
linker with a â-alanine and a glycine residue (Figure 2). The
linkers were introduced via peptide coupling with Fmoc-
protected amino acids, using HATU/HOAt, followed by Fmoc
removal by treatment with piperidine/DMF (1:4) for 10 min
and subsequent coupling, following a protocol similar to that
reported for other acylated amino-terminal oligonucleotides.37
For the direct acylation of 14, 4-pyrenylbutyric acid and cholic
acid were used to give 16 and 17, respectively. Either of the
acid residues is known to act as a cap for terminal base pairs,38,39
and either contains an alkyl chain for bridging the distance
between the attachment point to the primer and the stacking
site above the incoming nucleotide (Figure 1). One-step HPLC
purification gave 16 and 17 in yields between 30% and 35%.
Modeling studies showed that an external spacer is necessary
for acid residues lacking an internal alkyl spacer to reach above
the base pair between templating base and incoming nucleotide.
Trimethoxystilbene carboxylic acid was coupled to both the
Figure 3 shows representative spectra and kinetics for the
fastest reaction found, and Table 1 lists half-lives of primer
FIGURE 2. Structures of primers.
J. Org. Chem, Vol. 71, No. 3, 2006 1063

Baumhof et al.
FIGURE 3. Competitive primer extension reactions with primer 16, activated monomers 25a-t, and template 24: (a) MALDI-TOF mass spectra
taken 10 and 120 min after initiation of the reaction. (b) Kinetics of formation of extension products. The legend shows the nucleotides by which
16 was extended; incorrect extension by deoxycytidine residues was below the detection limit.
SCHEME 4
increased the rates of the primer extension up to 4-fold, but
also improved its fidelity compared to control 22, with less
incorporation of mismatched nucleotides that do not fit the
templating base of 23 or 24. Compared to reactions with the
same template performed in the presence of a “helper” oligo-
nucleotide that binds immediately downstream of the templating
nucleotide,19 primer extension with 16 is faster for either
template. Helper oligonucleotides create a binding pocket for
incoming nucleotides by binding immediately downstream of
the templating base. Perhaps most importantly for practical
applications, such as non-enzymatic SNP genotyping,19 the cap
adjusts the rate of the extension reactions on the two different
templates to very similar values, with half-lives of primer
conversion of 0.67 and 0.61 h (Table 1). In the case of the three-
strand system involving a helper oligonucleotide, the values are
0.8 and 1.3 h for the A and C templates, respectively, under
the same conditions.19 Anthraquinone-bearing primer 21 rivals
16 in reactivity and exceeds it in fidelity, when employed on
template 24 but is inferior when reacting in the presence of 23.
TABLE 1. Results of Competitive Template-Directed Primer
Extension Reactions with Activated Monomers 25a-ta
Discussion
product ratio
t1/2 primer
These results establish a synthetic route to oligonucleotides
with 3′-terminal diaminonucleotide residues. They also show
that it is possible to generate primers for non-enzymatic single-
nucleotide extension reactions with higher reactivity and better
selectivity than control strands. This is encouraging, as it might
lead to ways of more efficiently determining nucleotides in DNA
strands without costly enzymes and their triphosphate substrates.
Determining DNA sequences at individual positions known to
harbor variances in the human genome is one focus of current
genomic projects.42,43 The step from caps on templates used in
earlier work20 to caps on primers is critically important for
genotyping applications, as it ensures that unmodified DNA that
comes from biological sources or a PCR reaction starting from
genomic DNA can be probed.
template
primer
(primer + C/T/A/G)b
(h)
23
23
23
23
24
24
24
24
22
16
20
21
22
16
20
21
<1/61/12/26
<1/72/<1/26
7/49/17/27
<1/55/<1/43
<1/9/5/85
<1/6/6/87
6/13/10/71
<1/5/<1/93
2.58
0.67
1.58
1.15
1.45
0.61
1.03
0.62
a Conditions: 36 µM template and primer, 3.6 mM 25a-g, 18 mM 25t,
200 mM HEPES, pH 8.9, 400 mM NaCl, 80 mM MgCl2. b Matched primer
extension products highlighted in boldface.
conversion as well as product distributions for assays with either
template. The “winner” compound, pyrenyl primer 16, not only
(41) Kryatova, O. P.; Connors, W. H.; Bleczinski, C. F.; Mokhir, A. A.;
Richert, C. Org. Lett. 2001, 3, 987-990. (b) Siegmund, K.; Maheshwary,
S.; Narayanan, S.; Connors, W.; Riedrich, M.; Printz, M.; Richert, C. Nucleic
Acids Res. 2005, 33, 4838-4848.
(37) Schwope, I.; Bleczinski, C. F.; Richert, C. J. Org. Chem. 1999, 64,
4749-4761.
(38) Bleczinski, C. F.; Richert, C. J. Am. Chem. Soc. 1999, 121, 10889-
10894. (b) Tuma, J.; Richert, C. Biochemistry 2003, 42, 8957-8965.
(39) Narayanan, S.; Gall, J.; Richert, C. Nucleic Acids Res. 2004, 32,
2901-2911.
(42) Newton-Cheh, C.; Hirschhorn, J. N. Mutat. Res. 2005, 573, 54-
69. (b) Gewin, V. Nature 2005, 435, 1159. (c) Craig, D. W.; Stephan, D.
A. Expert ReV. Mol. Diagn. 2005, 5, 159-170.
(40) Dogan, Z.; Paulini, R.; Rojas Stu¨tz, J. A.; Narayanan, S.; Richert,
C. J. Am. Chem. Soc. 2004, 126, 4762-4763. (b) Tuma, J.; Paulini, R.;
Rojas Stu¨tz, J. A.; Richert, C. Biochemistry 2004, 43, 15680-15687.
(43) Shendure, J.; Porreca, G. J.; Reppas, N. B.; Lin, X.; McCutheon, J.
P.; Rosenbaum, A. M.; Wang, M. D.; Zhang, K.; Mitra, R. D.; Church, G.
M. Science 2005, 309, 1728-1732.
1064 J. Org. Chem., Vol. 71, No. 3, 2006

Synthesis of Capped Primers
The present results are also interesting in the context of primer
extension reactions themselves. They show that these reactions
can occur more rapidly when the incoming nucleotide reacts in
the presence of an aromatic substituent flexibly tethered to the
primer. This is not trivial, as a substituent or cap might also
block primer extension by stacking on the last base pair of the
primer-template duplex, sealing the primer against extension.
Because the caps tested do not feature cationic functional groups
that can activate the monomers toward attack by the 3′-amino
group of the primer, it is likely that the main effect is that of
retaining the monomer at the reaction site. The accelerating
effect of the current caps is similar to that of so-called helper
oligonucleotides that bind immediately downstream of the
templating nucleotide on the template,19 even though the caps
are not anchored in place through base pairing. It is possible
that caps can be evolved that act in the presence of helper
oligonucleotides, strengthening their effect. We have initiated
a project that seeks to identify such caps.44
adjusted mixture,48 suggest that spectrometrically monitored
selection experiments (SMOSE) should be feasible. Last, but
not least, it is worth mentioning that building block 11 might
also become useful for the synthesis of oligonucleotides with
phosphoramidate linkages49 in the interior of the sequence. To
this end, the protecting group scheme shown in Scheme 1 should
be more suitable, as it ensures that the 3′-terminus can be
deblocked selectively during chain assembly. The amino group
at the propargyl side chain can be diversified in combinatorial
searches for substituents that stabilize T-A base pairs via
interactions in the major groove,36,50 thus compensating for
the weakness of U-A base pairs with just two hydrogen
bonds.51
Experimental Section
3′-Azido-5′-O-benzoyl-2′,3′-dideoxyuridine (2). The following
is a modification of the method of Czernecki and Vale´ry.30 To a
solution of 2,3′-anhydro-5′-O-benzoylthymidine30 (0.79 g, 2.51
mmol) in dry DMF (5 mL) was added sodium azide (0.49 g, 7.54
mmol, 3 equiv), and the reaction mixture was heated to 120°C.
After 18 h, the solution was treated with water (200 mL), and the
pH was adjusted with 10% H2SO4 to pH 7. Three extractions with
diethyl ether (200 mL), followed by drying of the combined organic
phases over Na2SO4 and concentration in vacuo, gave a crude
product that was chromatographed on silica (1:2 hexane/ethyl
acetate, RF ) 0.48) to afford 2 as a pale yellow oil (0.51 g, 1.40
mmol, 56%). Spectroscopic data are in agreement with the
literature.30 HR-MS C16H16N5O5 calcd 358.1151, found 358.1135.
3′-Azido-5′-O-benzoyl-2′,3′-dideoxy-5-iodoridine (3). A solu-
tion of 2 (0.56 g, 1.59 mmol), iodine (0.24 g, 0.95 mmol, 0.6 equiv),
and cerium ammonium nitrate (0.44 g, 0.8 mmol, 0.5 equiv) in
acetonitrile (10 mL) was heated to 65°C. After 30 min, the solution
was treated with saturated sodium hydrogen thiosulfate solution
until the brown color of the solution disappeared and was then
extracted with diethyl ether (2 × 150 mL). The combined organic
phases were dried over Na2SO4 and concentrated in vacuo.
Purification of the pale yellow oil on silica (1:1 hexane/ethyl acetate,
RF ) 0.60) gave 3 as a pale yellow oil (0.65 g, 1.33 mmol, 84%).
1H NMR (CDCl3, 500 MHz) δ 2.41 (m, 1H), 2.61 (m, 1H), 4.27
(t, J ) 3.6 Hz, 1H), 4.36 (m, 1H), 4.63 (d, J ) 3.7 Hz, 2H), 6.12
(t, J ) 6.4 Hz, 1H), 7.48 (t, J ) 8.0 Hz, 2H), 7.61 (t, J ) 8.0 Hz,
1H), 7.88 (s, 1H), 8.05 (d, J ) 8.3 Hz, 2H), 9.66 (s, 1H); 13C NMR
(CDCl3, 125 MHz) δ 166.1, 160.1, 149.8, 144.0, 133.7, 129.8,
129.0, 128.9, 86.1, 82.5, 68.9, 63.7, 60.4, 38.3; MS (FAB, 3-NBA)
m/z C16H15IN5O5 calcd 484.0, found 484.1; IR (drift) ν ) 2106.9
Even the rates achieved thus far are sufficient to determine
nucleotides within 1 h, using suitable microarrays45 and mass
spectrometric read-out.19 Rates for primer extension with 16
and both 23 and 24 are the highest on record for OAt esters
25a-t, and up to a factor of 2 greater than those known for the
three-strand system.19 The latter system with its additional helper
oligonucleotide has proven sufficiently reactive for on-chip
genotyping in less than 3 h, starting from 500 fmol of template
DNA.19 The level of selectivity achievable with primers such
as 16 is sufficient for making base calls in SNP genotyping
assays, but probably too low for non-enzymatic replication of
even a very small genome.
Only five different caps were tested in the current exploratory
assays. All of them came from studies focused on sealing blunt
ends of duplexes in the ground state.37-41 Perhaps, optimized
caps for primer extension reactions will look different. They
might exploit interactions with both the substrates and the single-
stranded overhang of the template through a combination of
hemi-intercalation and stacking. To achieve this result, the cap
should be too bulky for full intercalation to avoid sequestering
it in the primer-template duplex upstream of the reaction site.
Further, the linker should be rigid enough to disfavor blocking
of the binding site for the incoming nucleotide. Finally,
functional groups stabilizing the transition state directly or by
attracting Mg2+ ions in a fashion similar to that of polymerases46
could be beneficial.
cm-1
.
5′-O-Benzoyl-2′,3′-dideoxy-3′-amino-5-ioduridine (4). To a
solution of 3 (1 g, 2.07 mmol) in THF (10 mL) was added slowly
a solution of tin dichloride (0.78 g, 4.14 mmol, 2 equiv) in water
(5 mL). After 20 min, the reaction was stopped by adding 100 mL
of saturated NaHCO3 solution, and the pH was adjusted to 8 with
NaOH solution (5 M). After triplicate extraction with 100 mL of
ethyl acetate, the organic phase was dried over Na2SO4 and
concentrated in vacuo. Amine 4 was obtained as a colorless oil
With the route to 3′-amino primers with terminal 5-(acyl-
amidopropargyl)deoxyuridine residues presented here, one can
envision a search for caps via combinatorial libraries whose
components compete in selection assays monitored by MALDI-
TOF MS.47 Peaks for primers with more efficient caps can be
expected to appear first in the mass range of the extension
products. Exploratory experiments leading to small libraries
(chemsets), generated via mixed coupling with a reactivity-
1
(0.94 g, 2.06 mmol, 98%). H NMR (CD3OD, 400 MHz) δ 2.35
(m, 2H), 3.60 (m, 1H), 4.10 (m, 1H), 4.60 (dd, J ) 12.5 Hz, 2.6
Hz, 2H), 6.14 (t, J ) 6.7 Hz, 1H), 7.50 (m, 2H), 7.61 (t, J ) 7.4
Hz, 1H), 8.01 (m, 3H); 13C NMR (CD3OD, 125.8 MHz) δ 166.1,
160.1, 149.8, 144.7, 133.1, 129.2, 128.4, 85.7, 84.7, 67.1, 63.5,
(44) Spies, S.; Griesang, N.; Richert, C. University of Karlsruhe,
Karlsruhe, Germany, 2005. Unpublished results.
(45) Plutowski, U.; Richert, C. Angew. Chem. 2005, 117, 627-631;
Angew. Chem., Int. Ed. 2005, 44, 621-625.
(48) Dombi, K. L.; Steiner, U. E.; Richert, C. J. Comb. Chem. 2003, 5,
45-60.
(46) Doublie, S.; Tabor, S.; Long, A. M.; Richardson, C. C.; Ellenberger,
T. Nature 1998, 391, 251-258.
(47) Altman, R. K.; Schwope, I.; Sarracino, D. A.; Tetzlaff, C. N.;
Bleczinski, C. F.; Richert, C. J. Comb. Chem. 1999, 1, 493-508. (b) Berlin,
K.; Jain, R. K.; Tetzlaff, C.; Steinbeck, C.; Richert, C. Chem. Biol. 1997,
4, 63-77.
(49) Gryaznov, S. M. Biochim. Biophys. Acta 1999, 1489, 131-140.
(50) Barnes, T. W.; Turner, D. H. J. Am. Chem. Soc. 2001, 123, 4107-
4118.
(51) Nguyen, H. K.; Fournier, O,; Asseline, U.; Dupret, D.; Thuong, N.
T. Nucleic Acids Res. 1999, 27, 1492-1498.
J. Org. Chem, Vol. 71, No. 3, 2006 1065

Baumhof et al.
50.8, 40.0; HR-MS (EI) C16H16IN3O5 calcd 457.0135, found
457.0132.
with brine (30 mL). The organic phase was dried over Na2SO4,
the volume was reduced in vacuo to 0.5 mL, and the solution was
added dropwise to pentanes (20 mL). The supernatant was aspired,
and the oily residue was dried in vacuo to give the title compound
(mixture of diastereomers) as a colorless foam (49 mg, 0.06 mmol,
74%). 31P NMR (CD3CN-d3, 300 MHz) δ 150.2, 150.5.
5′-O-Benzoyl-2′,3′-dideoxy-5-iodo-3′-(tritylamino)uridine (5).
To a solution of 4 (0.2 g, 0.43 mmol) in dry CH2Cl2 (5 mL) was
added NEt3 (0.13 mL, 0.56 mmol, 1.3 equiv), followed by the
dropwise addition of trityl chloride (0.16 g, 0.56 mmol, 1.3 equiv)
in CH2Cl2 (2 mL). After 3 h, the reaction was quenched with
saturated aqueous NaHCO3 (100 mL) and extracted three times with
100 mL of CH2Cl2. The combined organic phases were dried over
Na2SO4 and concentrated in vacuo. Purification on silica (20:1
CH2Cl2/MeOH, RF ) 0.56) afforded the title compound, which
5′-O-Benzoyl-2′,3′-dideoxy-5-iodo-3′-(trifluoroacetylamido)-
uridine (9). Amine 4 (0.92 g, 2.01 mmol) was dissolved in 25 mL
of CH2Cl2, and trifluoracetic acid anhydride (640 mg, 3.02 mmol,
1.5 equiv) and pyridine (377 µL, 370 mg, 4.05 mmol, 2 equiv)
were added under stirring. After 30 min, the solution was partioned
between 100 mL of saturated aqueous NaHCO3 and 100 mL of
CH2Cl2. Concentration of the organic phase in vacuo and purifica-
tion on silica [1:2 (v/v) hexanes/ethyl acetate, RF ) 0.75] gave a
1
crystallized upon standing (0.22 g, 0.39 mmol, 73%). H NMR
(CDCl3, 500 MHz) δ 2.41 (m, 1H), 2.61 (m, 1H), 4.27 (t, J ) 3.6
Hz, 1H), 4.36 (m, 1H), 4.63 (d, J ) 3.7 Hz, 2H), 6.12 (t, J ) 6.4
Hz, 1H), 7.37 (m, 10H), 7.45 (m, 2H), 7.61 (m, 7H), 7.88 (s, 1H),
8.05 (d, J ) 8.3 Hz, 3H), 9.2 (s, 1H); 13C NMR (CDCl3, 125 MHz)
δ 166.1, 160.1, 149.8, 144.0, 133.7, 129.8, 129.0, 128.9, 86.1, 82.5,
68.9, 63.7, 60.4, 38.3; FAB-MS (3-NBA) m/z 699.2.
1
colorless solid. (0.76 g, 1.37 mmol, 68%). H NMR (CDCl3, 500
MHz) δ 2.31 (m, 1H), 2.71 (d, J ) 11.1 Hz, 1H), 4.57-4.75 (m,
3H), 6.26 (s, 1H), 7.49 (t, J ) 7.5 Hz, 2H), 7.62 (t, J ) 7.0 Hz,
1H), 7.87 (s, 1H), 8.07 (d, J ) 7.8 Hz, 2H), 8.2 (br s, 1H), 10.3 (br
s, 1H); 13C NMR (CDCl3, 125 MHz) δ 166.3, 144.9, 144.6, 133.4,
133.1, 129.3, 128.4, 128.4, 126.4, 90.6, 86.0, 84.9, 81.8, 63.6, 50.0,
36.5; HR-MS (EI) m/z for C18H15N3F3IO6 calcd 552.9958, found
552.9957.
2′,3′-Dideoxy-5-iodo-3′-(trifluoroacetylamido)uridine (10). To
a solution of 9 (100 mg, 0.18 mmol) in MeOH (4 mL) was added
aqueous NaOH (0.72 mL, 1 M, 4 equiv). After TLC showed that
the reaction was complete (typically 0.5-1 h), water (10 mL) was
added, and the resulting mixture was extracted three times with
100 mL of ethyl acetate. The combined organic phases were dried
over Na2SO4 and concentrated in vacuo. Purification on silica [1:2
(v/v) hexanes/ethyl acetate, RF ) 0.5] gave the title compound as
a colorless solid (0.076 g, 0.17 mmol, 94%). 1H NMR (MeOH-d4,
500 MHz) δ 2.32 (t, J ) 6.6 Hz, 2H), 3.59 (dd, J ) 12.1, 3.3 Hz,
1H), 3.74 (dd, J ) 12.1 Hz, 2.5 Hz, 1H), 3.88 (m, 1H), 4.50 (m,
1H), 6.08 (t, J ) 6.6 Hz), 8.45 (s, 1H); 13C NMR (MeOH-d4, 125
MHz) δ 161.3, 157.6, 157.3, 150.4, 145.6, 114.4, 85.3, 84.7, 60.5,
49.1, 37.1; HR-MS (EI) m/z for C11H11N3F3IO5 calcd 448.9695,
found 448.9694.
2′,3′-Dideoxy-5-iodo-3′-(trifluoroacetylamido)uridin-5′-O-yl-
â-cyanoethyl-N,N-diisopropyl-phosphoramidite (11). To a stirred
solution of 10 (125 mg, 0.27 mmol) and diisopropylethylamine (141
µL, 0.81 mmol, 3 equiv) in anhydrous CH3CN (2.0 mL) was added
2-cyanoethyl-N,N-diisopropyl chlorophosphoramidite (70 µL, 0.81
mmol, 1.1 equiv). The solution was stirred for 30 min. After TLC
indicated complete conversion, CH2Cl2 (20 mL) was added, and
the solution was washed with saturated NaHCO3 solution (2 × 30
mL) and with brine (30 mL). The organic phase was dried over
Na2SO4, and the solvent was evaporated to an approximate volume
of 0.5 mL. This solution was added dropwise to pentanes (20 mL).
The supernatant was aspired, and the oily residue was dried in
vacuo, yielding the title compound (mixture of diastereomers) as a
colorless foam (141 mg, 0.22 mmol, 80%). 31P NMR (CD3CN-d3,
300 MHz) δ 149.9, 150.3.
DNA Synthesis. Oligodeoxynucleotides were prepared on 1
µmol scale on a model 8909 Expedite synthesizer, software 2.01
(Perseptive Biosystems) using the standard protocol, except that
5′-phosphoramidites (reverse phosphoramidites) were used. Cou-
pling of 11 to the unmodified portion of the strand used a 0.4 M
solution in CH3CN that had been dried over 4-Å molecular sieves.
All oligodeoxynucleotides were deprotected and cleaved from the
solid support using concentrated aqueous ammonia for 5 h at 65
°C. After being cooled to room temperature., the supernatant was
collected, and the residue was washed with deionized water. Excess
ammonia was removed from the combined aqueous solutions with
a gentle stream of compressed air, followed by lyophilization to
dryness. The residue was taken up with deionized water, filtered,
and subjected to HPLC purification.
5-[3-N-(Allyloxycarbonyl)aminopropyn-1-yl]-5′-O-benzoyl-
2′,3′-dideoxy-3′-(tritylamino)uridine (6). A solution of 5 (150 mg,
0.21 mmol) in DMF (2 mL) was degassed under reduced pressure
and placed under argon. Then, Pd(PPh3)4 (25.0 mg, 21 µmol, 0.1
equiv), CuI (8.0 mg, 42 µmol, 0.2 equiv), N-alloc-protected
propargylamine (39.0 mg, 0.28 mmol, 1.3 equiv), and NEt3 (0.09
mL, 0.63 mmol, 3 equiv) were added, and the solution was stirred
for 4 h at room temperature. After addition of an aqueous solution
of EDTA (0.14 M), the mixture was extracted with ethyl acetate
(3 × 100 mL), and the combined organic phases were dried over
Na2SO4, concentrated in vacuo, and subjected to chromatography
(silica, 2:1 hexanes/ethyl acetate, RF ) 0.5), yielding a pale yellow
oil (79 mg, 0.11 mmol, 53%). 1H NMR (CDCl3, 500 MHz) δ 1.27
(m, 1H), 1.75 (m, 1H), 1.90 (m, 2H), 3.40 (m, 1H), 3.90 (dq, J )
17.6, 5.2 Hz, 2H), 4.03 (m, 1H), 4.39 (dd, J ) 12.5, 4.0 Hz, 1H),
4.57 (m, 3H), 4.87 (br s, 1H), 5.29 (d, J ) 1.5 Hz, 1H), 5.32 (d, J
) 1.5 Hz, 1H), 5.91 (m, 1H), 5.99 (t, J ) 5.3 Hz, 1H), 7.13-7.29
(m, 10 H), 7.45 (t, J ) 8.2 Hz, 2H), 7.60-7.69 (m, 7H), 7.89 (dd,
J ) 8.0, 0.9 Hz, 2H), 8.84 (s br, 1H); 13C NMR (CDCl3, 125 MHz)
δ 166.2, 161.3, 155.8, 148.8, 145.8, 145.7, 142.1, 133.3, 132.6,
129.6, 129.3, 128.6, 128.4, 128.2, 126.9, 117.8, 99.3, 89.5, 85.7,
84.3, 74.1, 71.2, 65.8, 63.6, 53.6, 40.6, 31.5; FAB-MS (3-NBA)
m/z 711.3.
5-[3-N-(Allyloxycarbonyl)aminopropyn-1-yl]-2′,3′-dideoxy-3′-
(tritylamino) uridine (7). To a solution of 6 (73.0 mg, 0.1 mmol)
in MeOH (3 mL) was added aqueous NaOH (0.3 mL, 1 M, 3 equiv)
at room temperature. After 3 h, water (10 mL) was added, and the
resulting mixture was extracted with ethyl acetate (3 × 100 mL).
The combined organic phases were dried over Na2SO4 and
concentrated in vacuo. Chromatographic purification on silica (1:2
hexanes/ethyl acetate, RF ) 0.6) gave a colorless solid (0.03 g,
1
0.06 mmol, 58%). H NMR (CDCl3, 500 MHz) δ 1.32 (m, 1H),
1.45 (m, 1H), 1.94 (s, 2H), 2.96 (s, 1H), 3.40 (m, 1H), 3.62 (m,
1H), 3.71 (m, 1H), 3.69 (m, 1H), 3.88 (m, 1H), 3.92 (m, 1H), 4.55
(d, J ) 5.3 Hz, 1H), 5.20 (d, J ) 1.3 Hz, 2H), 5.53 (d, J ) 4.9 Hz,
1H), 5.94 (m, 1H), 5.99 (t, J ) 6.3 Hz, 1H), 7.16 (t, J ) 7.1 Hz,
3H), 7.20-7.30 (m, 6H), 7.53 (d, J ) 7.4 Hz, 6H), 7.91 (s, 1H),
9.12 (s, 1H); 13C NMR (CDCl3, 125 MHz) δ 171.2, 161.7, 156.1,
149.1, 146.1, 146.0, 143.8, 132.6, 132.1, 128.6, 128.1, 126.8, 117.9,
98.9, 89.6, 87.3, 85.5, 77.7, 77.2, 74.3, 71.2, 65.9, 61.7, 53.1, 40.4,
31.6; MALDI-TOF MS (linear, positive mode) m/z C35H34N4O6
calcd 606.2 [M + H], 629.1 [M + Na], found 606.8, 629.8.
5-[3-N-(Allyloxycarbonyl)aminopropyn-1-yl]-2′,3′-dideoxy-3′-
(tritylamino) uridin-5′-O-yl-â-cyanoethyl-N,N-diisopropyl phos-
phoramidite (8). To a stirred solution of 7 (50.0 mg, 0.08 mmol)
and diisopropylethylamine (42 µl, 0.24 mmol, 3 equiv) in anhydrous
CH3CN (2.0 mL) was added 2-cyanoethyl-N,N-diisopropyl chloro-
phosphoramidite (30 µL, 0.12 mmol, 1.5 equiv). The solution was
stirred for 30 min at room temperature. After TLC indicated
complete conversion, CH2Cl2 (20 mL) was added, and the solution
was washed with saturated aqueous NaHCO3 (2 × 30 mL) and
Compound 22. The last extension cycle of the DNA synthesis
used a 3′-MMT-protected phosphoramidite of 3′-amino-3′-deoxy-
thymidine (ChemGenes). The MMT group was removed using the
1066 J. Org. Chem., Vol. 71, No. 3, 2006

Synthesis of Capped Primers
deblocking solution for DNA synthesis [3:97 (v/v) trichloroacetic
acid/CH2Cl2] for 10 min prior to cleavage from the solid support.
HPLC 0% B for 5 min, to 30% B in 45 min, tR ) 21.2 min; yield
44%; MALDI-TOF MS calcd for C67H81N27O39P6 ([M - H]-) 2073,
found 2071.
(3.0 mg, 100 µmol) was coupled with HATU, HOAt, and diiso-
propylethylamine following the protocol for 15. Fmoc removal
again followed the protocol for 18/19. Coupling of trimethoxy-
stilbene carboxylic acid or anthraquinone carboxylic acid was
performed as described for 15. Deprotection followed the protocol
given under DNA Synthesis above.
Compound 20. HPLC 0% B for 5 min, to 30% B in 45 min, tR
) 35.9 min; yield 41%; MALDI-TOF MS calcd for C92H106N30O45P6
([M - H]-) 2536, found 2532.
Support 13. A sample of controlled pore glass 12 (5.0 mg, 0.125
µmol loading) in a polypropylene vial was dried at 0.1 Torr for 1
h, and a slurry of Pd(PPh3)4 (8.0 mg, 6.25 µmol, 50 equiv) and
CuI (2.0 mg, 10 µmol, 80 equiv) in dry DMF (200 µL) was added,
followed by diisopropylethylamine (5 µL, 35 µmol, 280 equiv).
After 2 min, N-tritylpropargylamine was added. After the mixture
had been shaken for 5 h, the supernatant was aspired, and the
support was washed once with CH3CN (400 µL), twice with a
solution of ethylenedithiocarbamic acid (sodium salt) in 400 µL of
DMF [0.5% (v/v)], and thrice with CH3CN (400 µL). The support
was then dried at 0.1 Torr. Deprotection of an analytical sample
showed >80% conversion by MALDI-TOF MS. MALDI-MS calcd
for C88H97N28O39P6 ([M - H]-) 2355, found 2359.
Support 14. The trityl group on support 13 was removed as
described for compound 22, above. The deprotection was performed
immediately prior to amide coupling. Deblocking solution (250 µL)
was added to the solid support (10.0 mg, 0.25 µmol), and the
mixture was shaken for 10 min. The supernatant was aspired, and
the support was washed with CH3CN (3 × 300 µL) and then dried
at 0.1 Torr.
Oligonucleotides of General Structure 15. The carboxylic acid
(100 µmol), HATU (15.0 mg, 90 µmol), and HOAt (2 mg, 90 µmol)
were dissolved in DMF (300 µL), diisopropylethylamine (10 µL,
100 µmol, 1 equiv) was then added, and the resulting mixture was
shaken for 5 min. The resulting solution was added to 14, and the
slurry was shaken for 3 h. The supernatant was aspired, and the
support was washed with CH3CN (3 × 300 µL) and then dried at
0.1 Torr. Afterward, the support was either subjected to subsequent
Fmoc removal and coupling (compounds 18-22) or deprotected
with NH4OH, as described for DNA Synthesis, above.
Compound 16. HPLC 0% B for 5 min, to 15% B in 55 min, tR
) 25.4 min; yield 30%, MALDI-MS calcd for C89H96N28O40P6 ([M
- H]-) 2382, found 2388.
Compound 17. HPLC 0% B for 5 min, to 30% B in 45 min, tR
) 40.3 min; yield 35%, MALDI-MS calcd for C93H120N28O43P6
([M - H]-) 2503, found 2509.
Compounds 18 and 19. The coupling of N-Fmoc-â-alanine (3.2
mg, 100 µmol) to support 14 (10.0 mg, 0.25 µmol loading) followed
the protocol given for general structure 15 above. Then, a solution
of piperidine in DMF [1:4 (v/v), 250 µL] was added to the support
to remove the Fmoc groups, and the mixture was shaken for 10
min. The supernatant was aspired, and the support was washed with
CH3CN (3 × 300 µL) and then dried at 0.1 Torr. The support thus
prepared was subjected to an amide-forming coupling reaction with
trimethoxystilbene carboxylic acid52 or nalidixic acid following the
protocol given for 15, above. Deprotection with NH4OH again
followed the general protocol given for DNA synthesis above.
Compound 18. HPLC 0% B for 5 min, to 30% B in 55 min, tR
) 30.2 min; yield 19%; MALDI-TOF MS calcd for C90H103N29O44P6
([M - H]-) 2480, found 2486.
Compound 21. HPLC 0% B for 5 min, to 30% B in 45 min, tR
) 30.0 min; yield 30%; MALDI-TOF MS calcd for C89H96N30O44P6
([M - H]-) 2475, found 2481.
Primer Extension. Template-directed extension reactions of
oligonucleotides were carried out in 2.5 µL of an aqueous buffer
solution at room temperature. The oligonucleotide template (23 or
24) plus one of the primers (16-22) were dissolved separately in
water to give stock solutions (180 µM). An aliquot (0.5 µL) of
each solution was added to 1 µL of a stock solution of the buffer
containing HEPES (0.5 M), NaCl (1 M), and MgCl2 (0.2 M) at pH
8.9 in a polypropylene reaction vessel. An aliquot (0.5 µL) of the
stock solution of azabenzotriazole esters (25a-t) in water, with
equal concentration for 25a-g (18 mM) and 5-fold increased
concentration for 25t (90 mM), was added.19 The final concentra-
tions in the assay mixture were 3.6 mM for 25a-g, 18 mM for
25t, 200 mM for HEPES, 400 mM for NaCl, 80 mM for MgCl2,
36 µM for the respective primer (16-22), and 36 µM for the
template (23 or 24). Samples (0.4 µL) were drawn after 6, 30, 60,
120, and 360 min and 1 day and were diluted with water (15 µL).
The resulting solution was kept over approximately 0.5 mg of
Dowex cation-exchange resin (ammonium form) for 3 min. A
sample (0.5 µL) of the supernatant was used for MALDI-TOF MS.
MALDI-TOF Analysis. Analysis was performed via MALDI-
TOF mass spectrometry as previously described,19 using conditions
that allow for quantitative detection.53 Briefly, aliquots of sample
solution (0.5 µL) were spotted on a stainless steel MALDI target
plate and treated with the matrix/comatrix mixture [0.4 µL, solution
of 2,4,6-trihydroxyacetophenone (0.3 M in EtOH)] and diammo-
nium citrate [0.1 M in H2O, 2:1 (v/v)]. The resulting mixture was
allowed to crystallize at room temperature and subjected to MALDI-
TOF analysis. Three spectra were acquired per time point, using
300 shots per spot. Peak intensities were determined on the basis
of peak heights. Fits to kinetic data were obtained as previously
described.19,20
Acknowledgment. The authors are grateful to T. Lommel,
K. Sto¨ss, Dr. E. Kervio, A. Hochgesand, and F. Brotzel for
sharing experimental results; to S. Vogel for help with kinetic
analyses; and to P. Gru¨nefeld for a literature search. This work
was supported by DFG (Grant 1063/1-3), The University of
Karlsruhe (TH), and Fonds der Chemischen Industrie (Project
164431).
Supporting Information Available: 1H NMR spectra of stable
new nucleosides, protocol for preparation of protected propargyl-
amine, MALDI-TOF mass spectra of modified oligonucleotides,
and additional kinetics for primer extension assays. This material
is available free of charge via the Internet at http://pubs.acs.org.
Compound 19. HPLC 0% B for 5 min, to 30% B in 55 min, tR
) 25.0 min; yield 15%; MALDI-TOF MS calcd for C84H97N31O42P6
([M - H]-) 2398, found 2401.
Compounds 20 and 21. After coupling of Fmoc-â-alanine and
removal of the Fmoc group, as described for 18/19, N-Fmoc-glycine
JO052097J
(52) Amedio, J. C.; Sunay, U. B.; Repic, O. Synth. Comm. 1995, 25,
667-680
(53) Sarracino, D.; Richert, C. Bioorg. Med. Chem. Lett. 1996, 6, 2543-
2548.
J. Org. Chem, Vol. 71, No. 3, 2006 1067