Synthesis of 3′-BODIPY-labeled active esters of nucleotides and a chemical primer extension assay on beads

Article abstract of DOI:10.1002/ejoc.201000210

A solution-phase synthesis of active esters of 3′-fluoropho:relabeled deoxynucleoside 5′-monophosphates was developed for thymine and cytosine as nucleobases by using two different: BODIPY dyes. Starting from the respective 2′-amino2′,3′-dideoxynucleoside

Full text of DOI:10.1002/ejoc.201000210

Evaluation Only. Created with Aspose.PDF. Copyright 2002-2021 Aspose Pty Ltd.
FULL PAPER
DOI: 10.1002/ejoc.201000210
Synthesis of 3Ј-BODIPY-Labeled Active Esters of Nucleotides and a Chemical
Primer Extension Assay on Beads
Kerstin Gießler,[a] Helmut Griesser,[b] Daniela Göhringer,[b] Thomas Sabirov,[b] and
Clemens Richert*[a,b]
Keywords: DNA / Oligonucleotides / Templated synthesis / Fluorescence
A solution-phase synthesis of active esters of 3Ј-fluorophore-
labeled deoxynucleoside 5Ј-monophosphates was developed
for thymine and cytosine as nucleobases by using two dif-
ferent BODIPY dyes. Starting from the respective 2Ј-amino-
2Ј,3Ј-dideoxynucleoside-5Ј-monophosphate, the fluorescent
oxyazabenzotriazolides can be prepared in one-pot pro-
cedures involving labeling and activation. Screening of a
range of supports led to a chemical primer extension assay
on beads with in situ detection of nucleobases in target DNA
through optical read-out.
monophosphate with a primer, as directed by the templat-
ing base of a DNA or RNA strand. Labeled monomers for
CPE such as 2[8] and 3[9] (Figure 1) have been developed.
Chemical primer extension is sensitive[10] and rapid,[11]
and it allows sequence information on DNA and RNA tem-
plates to be read,[12] but known versions have been limited
to assays with a purification step before detection.[9] For
example, in a published version of CPE with labeled mono-
mers and read-out on DNA microarrays, labeled oligo-
nucleotides had to be separated from excess monomers by
spin columns.[9] The main cause for this undesirable feature
of the assays was the background signal from unincorpo-
rated monomer on the surface of the microarrays, which
was caused by unfavorable physical properties of the labeled
monomers that we were unable to wash off from the sur-
faces of the microarray.
Further, the labeled monomers employed had to be pre-
pared on a solid support[9] by using several costly phos-
phoramidites, limiting the scale on which the fluorescent
building blocks could be prepared. Here we report fluores-
cent oxyazabenzotriazole esters of acylated 3Ј-amino-2Ј,3Ј-
dideoxynucleoside-5Ј-monophosphates[14] with more favor-
able properties, together with primer extension reactions
that can be performed directly on beads with subsequent
washing and optical read-out.
Introduction
Labeled nucleotides are important tools in biomedicine
and molecular biology. Radioactively labeled ATP is per-
haps the most universal labeling reagent for DNA and
RNA strands that are to be detected by gel electrophoresis
and autoradiography.[1] Although radioactively labeled nu-
cleic acids can be detected with exquisite sensitivity and ac-
curacy, the radiation hazard complicates assays and adds
a significant extra cost to experimental procedures. Sanger
dideoxy sequencing with fluorophore-labeled terminator
deoxynucleoside–triphosphates[2] (e.g. 1, Figure 1) has been
the backbone of genome sequencing methods that led to
the elucidation of the human genome, as well as numerous
other genomes from the different kingdoms of life. This
method has been automated to an impressive degree, but
it is probably too costly to compete with new generation
sequencing methods in the long term.[3,4]
One approach to reducing the cost and effort of sequenc-
ing and genotyping is to avoid enzymes. Chemical primer
extension (CPE) is a nascent technology that produces oli-
gonucleotides containing nucleobases complementary to
the sequence of a DNA template without enzymes.[5] It is
an extension of methodologies developed in the context of
studies of primitive self-replicating chemical systems.[6,7] It
relies on the reaction of an active ester of a nucleoside
For our current work, we chose BODIPY dyes as fluoro-
phores for labeling. The fluorescent monomers were pre-
pared by solution-phase syntheses and then employed in
primer extension assays on beads, demonstrating that these
monomers allow a nucleobase in a DNA template on the
very support on which the extension assay is performed to
be read-out, as required for efficient genotyping of single
nucleotide polymorphisms (SNPs) and other genetic varia-
tions.
[a] Institut für Organische Chemie, Universität Karlsruhe (K.I.T.),
76131 Karlsruhe, Germany
Fax: +49-711-685-64321
E-mail: clemens.richert@oc.uni-stuttgart.de
[b] Institut für Organische Chemie, Universität Stuttgart,
70569 Stuttgart, Germany
Supporting information for this article is available on the
WWW under http://dx.doi.org/10.1002/ejoc.201000210.
Eur. J. Org. Chem. 2010, 3611–3620
© 2010 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
3611

Evaluation Only. Created with Aspose.PDF. Copyright 2002-2021 Aspose Pty Ltd.
K. Gießler, H. Griesser, D. Göhringer, T. Sabirov, C. Richert
FULL PAPER
linkage through attack of the amino-terminal primer.[14]
Primer extension to 8c/t can be accelerated by employing a
third oligonucleotide (i.e., 4) that binds immediately down-
stream from the templating base and that offers additional
stacking interactions to the monomer. Such additional oli-
gonucleotides are also dubbed “helper oligonucleotides”.[10]
In our first assays, monomers with cyanine dyes as 3Ј-
labels were employed, including compound 3 (Figure 1).
With these, attempts to establish assay conditions for in situ
detection of single bases in DNA through primer extension,
followed by washing and optical detection on the same sup-
port, failed. Among the surface chemistries tested were
glass slides decorated with an amino-terminal capture
strand that had reacted with aldehyde groups,[15,16] NHS
ester groups (CodeLink activated slides),[17] or isocyanate
modified slides.[18] These monomer/surface combinations
were also unsuccessful when combined with different pas-
sivation/blocking methods.[15] Treating the different surfaces
with a 3.6 m solution of 3 in the absence of any DNA
confirmed that the overly strong background signal was due
to unspecific adsorption of the fluorophore-labeled nucleo-
tide. The problem with the background signal from the cy-
anine (Cy)-labeled nucleotides persisted, even when the un-
activated monophosphate was used, the contact time was
limited to 1 min, and when the washing buffer contained
up to 0.2% SDS (sodium dodecyl sulfate).
This prompted us to search for smaller dyes with more
suitable properties. Among the fluorophores that fulfilled
fundamental criteria, such as not being quenched by DNA,
having a low tendency to intercalate, being stable under the
coupling and assay conditions, and being accessible as a
series of at least four different structures with sufficiently
different emission spectra for each of the different nucleo-
bases, the BODIPY dyes appeared most attractive. This
class of dyes appeared readily accessible synthetically and
has previously been proposed as fluorophores for auto-
mated DNA sequencing.[19] The NHS esters of the
BODIPY chromophores with emission maxima similar to
those of fluorescein and tetramethylrhodamine are com-
mercially available under the trade names BODIPY-FL and
BODIPY-TMR, albeit at very significant costs. We initially
chose BODIPY-labeled nucleotides 21 and 22 as target mo-
lecules.
Figure 1. Fluorophore-labeled monomers that were developed for
dideoxy sequencing (i.e., 1)[2c] or chemical primer extension (i.e., 2,
3).[8,9]
Results and Discussion
First, we synthesized 14 by following literature proto-
Syntheses
cols[20,21] (see Supporting Information). This route pro-
The chemical primer extension system employed for the duced large amounts of side products during an early con-
current study is shown in Scheme 1. It consists of primer 5 densation step and required a costly trichloroethyl ester.
and a DNA 60mer template (7a/g) that is held on a support When ester cleavage was attempted under basic conditions,
through base pairing to a capture oligonucleotide (6). Hy- rather than the enzymatic protocol published, low yields
bridization-based immobilization of templates ensures that were obtained. To avoid these problems, we employed dif-
assays can be performed in parallel fashion, with different ferent syntheses for both 14 and its redshifted analog 13
templates hybridized to capture oligonucleotides on dif- (Scheme 2). Both routes started from pyrrole-2-carbal-
ferent spots of a microarray or different beads of a bead dehyde and methyl (triphenylphosphoranylidene)acetate,
population. Bead-based sequencing of DNA in a massively which were reacted together to give 9 in two steps by a
parallel fashion has been described.[13]
Horner–Wadsworth–Emmons reaction, followed by hydro-
To determine the nucleobase of interest in the template genation on Pd/C.[22] For the route to 14, the required pyr-
strand, a labeled mononucleotide binds to the templating role aldehyde for the subsequent step was commercially
base (i.e., BЈ), followed by formation of a phosphoramidate available. For the route leading to 13, aldehyde 10 was pre-
3612
www.eurjoc.org
© 2010 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Eur. J. Org. Chem. 2010, 3611–3620

Evaluation Only. Created with Aspose.PDF. Copyright 2002-2021 Aspose Pty Ltd.
BODIPY-Labeled Active Esters of Nucleotides and Chemical Primer Extension
Scheme 1. Chemical primer extension reactions.
pared in three steps by pyrrole synthesis,[23] with N-allyl-4-
methoxybenzamide,[24] followed by Vilsmayer Haack for-
mylation in an overall yield of 70%. BODIPY dyes 11 and
12 were then prepared in moderate to good yield (41 and
71%, respectively) by using the method of Malan et al.[21]
Liberation of the carboxylic acids from esters proved chal-
lenging at first. Even after careful optimization studies in-
volving a series of bases (LiOH, KOH, Li2CO3, Na2CO3,
K2CO3, Cs2CO3), solvents, and reaction temperatures, basic
conditions gave no more than 63% yield of 14.
In the case of BODIPY dye 13, all attempts to hydrolyze
the ester moiety by using alkaline hydroxides, carbonates,[25]
or pig liver esterase (PLE)[21] again led to impure product
in low yield. Synthetic procedures that involved a free acid
as the starting material and that avoided the hydrolytic
step[26] were considered, but reported yields are so low that
they were not pursued. Hydrolysis under acidic conditions
was studied instead. Sioni et al. describe the synthesis of
dye 15 through ester hydrolysis with hydrochloric acid in
THF under reflux conditions, resulting in 50% yield.[26] An-
other literature-known acidic hydrolysis,[27] performed with
phosphoric acid, in our hands, gave the BODIPY acid in
equal yield, but in lower purity. We found that hydrolysis
of methyl esters 11 and 12 with more concentrated hydro-
chloric acid (4.5 ) in THF (1:1) at room temperature fur-
nished free carboxylic acids 13 and 14 in analytically pure
form in a yield of 81 and 72%, respectively.
Scheme 2. Synthesis of BODIPY dyes 13 and 14 from pyrrol-2-
carbaldehyde and structure of commercial labeling reagent 15.
Eur. J. Org. Chem. 2010, 3611–3620
© 2010 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.eurjoc.org
3613

Evaluation Only. Created with Aspose.PDF. Copyright 2002-2021 Aspose Pty Ltd.
K. Gießler, H. Griesser, D. Göhringer, T. Sabirov, C. Richert
FULL PAPER
With the two different fluorophores in hand, we then
turned to the labeling reaction. Initially, a three-step pro-
cedure was employed, starting with separate activation of
the BODIPY-FL propionic acid 14 to its OAt ester, fol-
lowed by coupling to 3Ј-amino-3Ј-deoxythymidine-5Ј-mo-
nophosphate (16)[14] at pH 8.5. Amide 18 was then activated
at its phosphate moiety by treatment with 1-ethyl-3-(3-di-
methylaminopropyl)carbodiimide (EDC) and 7-aza-1-hy-
droxybenzotriazole (HOAt) to give labeled active ester 21
in 67% yield. The absorption and fluorescence spectra of
monomers 21 and 22 are shown in Figure 2.
Scheme 3. Synthesis of activated BODIPY-labeled nucleotides 21
and 22.
nonucleoside (16 or 23), and addition of more EDC and
HOAt. The reaction solution was then directly applied to
an RP-18 cartridge, from which 21 or 24 were eluted in a
one-step purification process. The modest solubility of 24
in pure water also allows precipitation as a purification
method, producing 24 in ca. 70% yield, contaminated with
some 13. Even with the cartridge step, the one-pot ap-
proach gave a very convenient protocol, producing labeled
Figure 2. Optical properties of labeled nucleotides: (a) absorption
spectra and (b) fluorescence spectra of 21 and 22.
In our initial work, the second monomer to be employed and activated monomers within a few hours without any
in chemical primer extension assays (compound 22) was HPLC or other costly or time-consuming steps.
synthesized starting from N-benzoyl-protected 3Ј-amino-
Chemical primer extension assays were performed on dif-
2Ј3Ј-dideoxycytidine-5Ј-monophosphate[14] 17 (Scheme 3) ferent surfaces.[16,28] Initially, a capture oligonucleotide with
and commercial BODIPY ester 15. N-Hydroxysuccinimide a lysine residue at its 5Ј-terminus, identical in structure to
ester 15 was coupled to 3Ј-amine 17 in aqueous DMF in the that given in the literature[9] was immobilized on planar
presence of N,N-diisopropylethylamine (DIEA), resulting in glass slides with epoxy groups on their surface through nu-
the protected, labeled cytidine monophosphate 19 that was cleophilic ring opening at pH 11. Templates 7a and 7g were
deprotected with aqueous ammonia to give 20. Reaction then hybridized to different spots of the DNA-bearing sur-
with EDC and HOAt at pH 5 gave monomer 22 for chemi- face, together with primer 5 and helper strand 4. Finally, 22
cal primer extension assays.
was added to the entire slide surface in extension buffer
To make the methodology more attractive for routine containing 0.08  Mg2Cl, 0.4  NaCl, and 0.2  HEPBS
biomedical use, two steps were taken. The first was to de- [N-(2-hydroxyethyl)piperazine-NЈ-(4-butanesulfonic acid)],
velop a one-pot protocol for labeling and active ester for- followed by primer extension and washing [1ϫSSC (saline–
mation. The second was to employ acid 13 (rather than the sodium citrate) buffer containing 0.2% SDS and 1ϫSSC
extremely costly 15) for labeling 3Ј-amino-2Ј,3Ј-dideoxycyt- buffer for 1 min at room temp.] and fluorescence imaging.
idine monophosphate, to access red monomer 24 at afford- The expected stronger fluorescence signal from spots with
able cost. First, aminonucleotide 23 (Scheme 4) was gener- the templating base was not observed. Attempts to block
ated through treatment of 17 with aqueous ammonia. For unspecific adsorption and/or reactions on the surface by
the one-pot conversion, BODIPY dyes 13 or 14 were then using different passivation agents did not solve this problem
treated with EDC at 0 °C, followed by addition of the ami- (Table 1).
3614
www.eurjoc.org
© 2010 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Eur. J. Org. Chem. 2010, 3611–3620