Synthesis of guanosine 5′-conjugates and their use as initiator molecules for transcription priming

Article abstract of DOI:10.1039/b716151d

We have synthesised two guanosine derivatives that are linked to biotinylated adenosine moieties by using two different strategies, one that includes synthetic steps on the solid phase and another one that is performed entirely in solution. The synthesised derivatives were shown to function as initiator molecules in transcription priming experiments. The incorporation efficiency was determined to be approximately 2%. Even though this value is rather low, the use of either molecule in selection experiments seems reasonable. Basically, RNA libraries with sequence complexities of 10 ¹⁵ to 10¹⁶ can be generated. Labelling of such a library with our initiator molecule would still produce 10¹³ to 10 ¹⁴ labelled/functionalised sequences, and thus sufficient sequence space for selection. The Royal Society of Chemistry.

Full text of DOI:10.1039/b716151d

View Article Online / Journal Homepage / Table of Contents for this issue
PAPER
www.rsc.org/obc | Organic & Biomolecular Chemistry
Synthesis of guanosine 5ꢀ-conjugates and their use as initiator molecules for
transcription priming†
Jo¨rn Wolf, Valeska Dombos, Bettina Appel and Sabine Mu¨ller*
Received 19th October 2007, Accepted 19th December 2007
First published as an Advance Article on the web 23rd January 2008
DOI: 10.1039/b716151d
We have synthesised two guanosine derivatives that are linked to biotinylated adenosine moieties by
using two different strategies, one that includes synthetic steps on the solid phase and another one that
is performed entirely in solution. The synthesised derivatives were shown to function as initiator
molecules in transcription priming experiments. The incorporation efficiency was determined to be
approximately 2%. Even though this value is rather low, the use of either molecule in selection
experiments seems reasonable. Basically, RNA libraries with sequence complexities of 1015 to 1016 can
be generated. Labelling of such a library with our initiator molecule would still produce 1013 to 1014
labelled/functionalised sequences, and thus sufficient sequence space for selection.
diol. It can be specifically oxidised with periodate, and the
Introduction
obtained dialdehyde can be reacted with an amine or hydrazine-
The discovery that RNA is capable of catalysing a wide variety of
functionalised derivative of the compound of interest to yield
the reactant–RNA conjugate.5 Alternatively, the reactant can be
conjugated to the 5ꢀ-end of the RNA molecules of the library. A
possible route to this goal is 5ꢀ-thiophosphorylation by chemical
or enzymatic means, and subsequent reaction of the nucleophilic
thiol with a suitably activated derivative of the compound to be
attached.6,7
In addition to post-transcriptional strategies, 5ꢀ-end modifi-
cation of RNA can be also achieved co-transcriptionally by
transcription priming.8 T7 RNA polymerase is the most widely
used enzyme for preparation of RNA by run-off transcription in
vitro. It has been shown that 5ꢀ-modified guanosine derivatives are
accepted instead of GTP as starting nucleoside in the initiation
step, but owing to the missing 5ꢀ-triphosphate, cannot be used
during elongation.9 Thus, conjugates of guanosine or guanosine
monophosphates have been shown, if provided in excess, to be
incorporated at the 5ꢀ-end of the produced RNA molecules by T7
RNA polymerase.4,6,9–13
Within a project that focuses on the selection of ribozymes
that support deamination of nucleosides as an important step
in RNA editing, we have synthesised biotinylated adenosine and
cytidine derivatives and linked those to guanosine. The obtained
conjugates were subsequently used for transcription priming in
order to generate a library of RNA molecules carrying for example
compounds 1 or 2 (Fig. 1) at their 5ꢀ-end. In the strategy we have
devised, selection will not rely on reaction of an attached reactant
with a tagged reactant as described above. On the contrary, we
intend to immobilise the entire library on a solid phase and to select
for species that are released into solution upon reaction. For this
purpose, the exocyclic amino group at C-6 of the adenine residue is
linked to biotin. This decoration allows for a selection procedure
involving immobilisation of the biotinylated RNA library on a
streptavidin-coated surface. Upon incubation of the immobilised
library under suitable reaction conditions, beneficial variants will
be released into solution and can be filtered from non-functional
members of the library. Here we describe the synthetic route to
chemical reactions has significantly supported the hypothesis of an
‘RNA world’, where RNA molecules were carriers of information
as well as functional players in the amplification, adaptation and
realisation of this program.1 To further extend the repertoire of
RNA-catalysed reactions, a number of studies have been carried
out using the powerful method of in vitro selection, whereby a
catalyst for a specific reaction is selected from a random RNA
library. Thus, RNAs that catalyse classical organic reactions such
as, for example, a Diels–Alder reaction,2 a Michael addition3 or
an aldol reaction,4 have been developed.
The typical assay for selection of a catalyst to enhance reaction
between two reactants A and B requires linking of a potential
reactant to the members of the RNA library, preferably via a
flexible tether, while the other reactant is decorated with a suitable
affinity tag, e.g. biotin. RNAs that accelerate the reaction of
the attached reactant with the second reactant carrying the tag
become tagged themselves, and subsequently can be separated
by affinity chromatography. While biotinylation is rather easily
accomplished, conjugation of a potential reactant to the members
of the RNA library is a more difficult task. In vitro selection
involves iterative rounds of selection and amplification requiring
enzymatic procedures for reverse transcription of RNA into DNA
and transcription of amplified DNA back into RNA. Therefore,
modification of the RNA molecules in the library has to be
repeated in each round. The most common strategies involve post-
transcriptional protocols, attaching the specific reactant either
to the 3ꢀ-end or to the 5ꢀ-end of RNAs in the library. 3ꢀ-End
modification uses the unique properties of the 3ꢀ-terminal cis-
Ernst-Moritz-Arndt-Universita¨t Greifswald, Institut fu¨r Biochemie, Bioor-
ganische Chemie, Felix-Hausdorff-Straße 4, 17487, Greifswald, Germany.
E-mail: sabine.mueller@uni-greifswald.de; Fax: +49 (0) 3834 864471;
Tel: +49 (0) 3834 8622843
† Electronic supplementary information (ESI) available: Additional exper-
imental information. See DOI: 10.1039/b716151d
This journal is
The Royal Society of Chemistry 2008
Org. Biomol. Chem., 2008, 6, 899–907 | 899
©

View Article Online
Fig. 1 Adenosine–guanosine conjugates used in transcription priming
reactions. 1 was obtained by solid phase phosphoramidite chemistry; 2
was synthesised in solution.
compounds 1 and 2 and their evaluation as initiator molecules in
transcription priming.
Fig. 2 Solid phase synthesis of 1. Coupling of components 3 and 4
by phosporamidite chemistry and successive deprotection yields com-
pound 1.
Results and discussion
In our effort to synthesise initiator molecules that allow for the
attachment of a functionalised adenosine residue to the 5ꢀ-end
of RNA by transcription priming, we have prepared two different
adenosine–guanosine conjugates, 1 and 2. Synthesis of the initiator
molecules was accomplished by modifying the guanosine and
adenosine moieties separately and eventually coupling them via
a hexaethylene glycol tether.14
Both compounds 1 and 2 carry a biotin residue at the N6-
position of the adenine base. The guanosine moiety is linked
to hexaethylene glycol via its 5ꢀ-phosphate. The glycol linker in
turn is connected to a 2ꢀ-phosphate (in the case of 1) or 5ꢀ-
phosphate (in the case of 2) of the sugar moiety of adenosine.
Two related strategies were applied to the synthesis of the
guanosine 5ꢀ-conjugates 1 and 2, one involving solid phase
coupling of nucleoside derivatives, the other involving full solution
chemistry.
7 was then accomplished by photochemically induced substitution
of the exocyclic amino group by iodine.
To this end, 6 was reacted with isopentyl nitrite in an iodine-
donating solvent (CH2I2) at 60 ◦C followed by irradiation with
visible light according to a procedure introduced by Nair and
Richardson.15 This procedure allows for the synthesis of purinyl
halogenides, which are well suited for the introduction of alkyl
residues at the N6-position of adenosine. There are other ways of
fulfilling this task,16 many of which, however, require the use of
HMPT, known to be a rather strong carcinogen.17 Synthesis of the
iodine derivative, as carried out here, avoids the use of this highly
carcinogenic solvent and thus seemed to be the more convenient
procedure. Furthermore, it has been successfully applied by other
groups and it was reported that successful substitution can be
achieved even in the absence of irradiation.18 Thus, we were able
to obtain the acetylated 9-iodoadenosine 7 in 43% yield.
Next we introduced a linker to the N6-position to be used
for attachment of biotin. The 3,6-dioxa-8-Nꢀ-BOC-aminooctyl
linker was synthesised from unprotected 2,2ꢀ-(ethylenedioxy)bis-
(ethylamine) by reaction with di-tert-butyldicarbonate based on a
method reported by Sigal et al.19 The BOC-protected linker was
coupled to 7 by reaction in ethanol under reflux for 6 hours. The
concomitant deacetylation of the ribose moiety was completed
by treatment with 25% aqueous ammonium hydroxide, and the
product 8 was obtained after chromatographic purification in 64%
yield.
Strategy 1: Solid phase coupling of functionalised adenosine to
guanosine
In our first approach, coupling of a suitably modified guanosine
derivative 4 to the biotinylated adenosine 3 was carried out on the
solid phase using phosphoramidite chemistry and an automated
DNA/RNA-synthesiser, yielding 1 as a 2ꢀ,5ꢀ-tethered dinucleotide
derivative in nanomolar quantity (Fig. 2).
Synthesis of the adenosine moiety started with acetylation of the
2ꢀ-, 3ꢀ- and 5ꢀ-hydroxy groups of adenosine 5 (Fig. 3). Although
it does not obstruct the following reactions, the aminoacetylated
side product was removed by silica gel chromatography to isolate
the pure triacetylated product 6. Synthesis of the 6-iodo derivative
Further synthetic steps involved renewed protection of the
3ꢀ- and 5ꢀ-OH groups, removal of the BOC group from the
aliphatic linker amino function, coupling it with biotin, and
900 | Org. Biomol. Chem., 2008, 6, 899–907
This journal is
The Royal Society of Chemistry 2008
©

View Article Online
Fig. 3 Introduction of the 3,6-dioxa-8-Nꢀ-BOC-aminooctyl linker at
the N6-position of the adenine base. Reagents and conditions: I) 4 eq.
acetic anhydride, 4-DMAP, pyridine, 24 h, RT, 50%. II) 30 eq. methylene
◦
iodide, 19 eq. isoamyl nitrite, 5 h, 60 C, 43%. III) 4.8 eq. N-BOC-2,2ꢀ-
(ethylenedioxy)bis(ethylamine), ethanol, 25% NH4OH, 64%. IV) 1 eq.
1,3-dichloro-1,1ꢀ,3,3ꢀ-tetraisopropyldisiloxane, pyridine, 2 h, RT, 84%.
V) 50% CF3COOH in CHCl3, 20 min, RT; 0.8 eq. 2-succinimi-
do-1,1,3,3-tetramethyluronium tetrafluoroborate (TSTU), 0.8 eq. bi-
otin, overnight, RT, 67%. VI) 1.8 eq. (2-cyanoethyl-N,N-diisopropyl)-
chlorophosphoramidite, 7.5 eq. diisopropylethylamine, CH2Cl2, ethyl
acetate, 76%.
Fig. 4 A) HPLC reverse phase purification of 1; A254 represents the
absorption at 254 nm. B) ESI-spectrum of the pooled HPLC-fractions
from peak II; r.I. represents the relative intensity. The peak at 1312.5
corresponds to molecule 1. For further details refer to main text.
from uncoupled conjugate 4. Smaller peaks surrounding the main
peak may result from incompletely deprotected initiator molecule
1 (peak III). The still present protecting groups lead to a higher
hydrophobicity, resulting in a longer retention time relative to the
fully deprotected product.
Since it can be carried out in part by an automated process,
the method described represents a relatively quick route for
synthesis of the initiator molecule 1. However, it has to be noted
that a large excess of the phosphoramidite 3 (15-fold over solid
phase reactant) is required for coupling in solid phase synthesis.
Furthermore, solid phase coupling with our facilities is limited
to the lmol scale, thus requiring iterative coupling reactions in
order to produce 3 in larger amounts. In order to overcome this
restriction, a second method was developed, which is performed
in solution and requires equimolar amounts of coupling partners.
Additionally, the solution-based method allows for synthesis of
initiator molecules on higher scales.
finally preparation of the 2ꢀ-phosphoramidite. As shown in Fig. 3,
selective 3ꢀ,5ꢀ-OH protection of 8 was accomplished with the
Markiewicz silyl group to yield 9. The BOC group was removed
from the amino linker using 50% CF3COOH in CHCl3, and
the resulting primary amino group was coupled to biotin that
was functionalised as an active succinimidyl ester. The resulting
nucleoside derivative 10 was then reacted with (2-cyanoethyl-N,N-
diisopropyl)chlorophosphoramidite to yield building block 3. This
was then coupled to immobilised derivatised guanosine 4 on the
solid phase to generate initiator molecule 1 (Fig. 2). First, CPG
loaded with 1 lmol guanosine was reacted with hexaethylene
glycol phosphoramidite in an automated synthesiser using the
standard protocol,20 followed by coupling of 3 without changes.
After deprotection and cleavage from the solid phase, 1 was
purified by reverse phase chromatography (Fig. 4A). The product
was assigned to the large peak with a retention time of 40 min (peak
II) in the HPLC diagram. Further mass spectrometric analysis
(MALDI) clearly identified the product as the desired adenosine–
guanosine conjugate 1. A smaller peak (peak I, retention time 20
min) presumably represents the guanosine–hexyl adduct resulting
Strategy 2: Coupling of functionalised adenosine to guanosine in
solution
As an alternative to solid phase coupling as described above,
the synthetic route in solution involves coupling of a guanosine
This journal is
The Royal Society of Chemistry 2008
Org. Biomol. Chem., 2008, 6, 899–907 | 901
©

View Article Online
5ꢀ-phosphoramidite to the 5ꢀ-OH group of the biotinylated adeno-
sine derivative via a hexaethylene glycol bridge (product 2 in
Fig. 1).
First, we aimed to couple an equimolar amount of mono-
DMT-protected hexaethylene glycol to the protected adenosine
phosphoramidite 3. However, we did not observe any coupling
product. Possibly, steric hindrance due to the bulky silyl groups
at the 5ꢀ- and 3ꢀ-OH hampers reaction at the neighbouring 2ꢀ-
O-phosphoramidite in solution. Also, problems concerned with
loss of the rather labile cyanoethyl group at the phosphorus in
solution chemistry have been reported.21 Therefore, we changed
the synthetic strategy for the adenosine part, as depicted in Fig. 5.
Instead of the labile cyanoethyl group, we chose the more stable
methyl group for protection of the phosphite, and decided to
couple the adenosine and guanosine moieties via the primary 5ꢀ-
OH groups of both sugar residues.
Fig. 6 Synthesis of the guanosine–hexaethylene glycol conjugate 15.
Reagents and conditions: I) 8 eq. TMSCl, 10 eq. isobutyric anhy-
dride, 25% NH4OH, RT, 26%. II) 1.3 eq. DMTCl, NEt3, pyridine,
overnight, RT, 41%. III) 10 eq. isobutyric anhydride, pyridine, 78%.
IV) 5% CCl3COOH in CH2Cl2, 20 min, pyridine, RT, 49%. V) 1.8 eq.
(methyl-N,N-diisopropyl)chlorophosphoramidite, 7.5 eq. diisopropy-
lethylamine, 1 h, CH2Cl2, used as crude product, quantitative yield
assumed from TLC analysis. VI) 1.0 eq. DMT-hexaethylene glycol,
acetonitrile, 0.45 M tetrazole solution; 5% CCl3COOH in CHCl3, 20 min,
RT, 35%.
Fig. 5 Synthesis of the adenosine derivative phosphoramidite 14 for
coupling in solution. Reagents and conditions: I) 1.4 eq. DMTCl, 3.6 eq.
NEt3, 4-DMAP, pyridine, 24 h, RT, 76%. II) 13 eq. isobutyric anhydride,
pyridine, 24 h, RT, 93%. III) 5% CCl3COOH in CHCl3, 20 min, RT,
73%. IV) 1.8 eq. (methyl-N,N-diisopropyl)chlorophosphoramidite, 7.5 eq.
diisopropylethylamine, 1 h, used as crude product, quantitative yield
assumed from TLC analysis.
Thus, 8 was selectively protected with a DMT-group at
the 5ꢀ-OH to provide 11, followed by isobutyrylation of the
remaining free 2ꢀ- and 3ꢀ-OH functionalities to result in 12.
After removal of the DMT-group (13), the 5ꢀ-methyl-N,N-
diisopropylphosphoramidite 14 was prepared and used without
further purification for the coupling reaction with the guanosine
moiety 15. The guanosine moiety 15 was synthesised as shown in
Fig. 6.
Guanosine 16 was isobutyrylated at the N2-position
and crystallised from water to give sufficiently pure N2-
isobutyrylguanosine 17. The 5ꢀ-position was protected with a
DMT-group (18) followed by isobutyrylation of the 2ꢀ- and 3ꢀ-
OH groups (19). The following removal of the DMT-group by
treatment with 5% trichloroacetic acid delivered 20, which was
converted into the 5ꢀ-O-methylphosphoramidite 21 and subse-
quently coupled with a mono-DMT-protected hexaethylene glycol
linker. Final detritylation delivered building block 15 in reasonably
good yield. This was then coupled to the adenosine derivative 14
as shown in Fig. 7.
The coupling product 22 was BOC-deprotected to give 23 and
biotinylated to provide 24 in a manner similar to that described for
the synthesis of compound 3 (Fig. 3). Removal of the methyl group
from the phosphate was performed using disodium 2-carbamoyl-
2-cyanoethylen-1,1-dithiolate (“Na2S2”) followed by treatment
with a 1 : 1 mixture of 32% aqueous ammonium hydroxide and
methylamine to deprotect the 2ꢀ- and 3ꢀ-OH functionalities.22 The
obtained initiator molecule 2 was purified by HPLC on a reverse
phase column. Fig. 8 shows the reverse phase HPLC diagram
and the MALDI spectra. Peak III was identified as the desired
compound 2 by mass spectrometry and NMR. Peak II presumably
results from uncoupled, deprotected compound 15, which is in
analogy to observed peak II in Fig. 5. The intense peak I results
from the UV-visible reagent Na2S2, which was used in excess during
the deprotection procedure.
902 | Org. Biomol. Chem., 2008, 6, 899–907
This journal is
The Royal Society of Chemistry 2008
©

View Article Online
Fig. 7 Reagents and conditions: I) 0.45 M tetrazole, acetonitrile, 83%. II)
30% CF3COOH in CH2Cl2, quantitative yield. III) TSTU, biotin, 52%.
IV) Na2S2, 32% NH4OH, MeNH2, quantitative yield assumed from TLC
analysis.
Fig. 8 A) HPLC reverse phase diagram of biotinylated adenosine–guano-
sine conjugate 2. A254 represents the absorption at 254 nm. B)
MALDI-spectrum of the pooled HPLC-fractions from peak III; r.I.
represents the relative intensity. Peaks at 1315.0, 1337.0, 1359.1 and 1381.0
correspond to the molecular ion and the molecule with 1, 2, or 3 bound
Na+ ions, respectively.
Transcription priming with initiator molecules 1 and 2
After extensive qualitative analysis of both compounds 1 and
2 to ensure their identity, we carried out initial experiments on
transcription priming. To this end, derivatives 1 or 2, respectively,
were added in different molar ratios to a transcription mixture
containing the four natural nucleoside triphosphates, a model
DNA template and T7 RNA polymerase. After reaction, product
RNAs were separated by electrophoresis through 10% denaturing
polyacrylamide gels and subsequently blotted onto a nylon mem-
brane. The membrane was incubated with streptavidin-conjugated
alkaline phosphatase. Thus, after several washing steps, alkaline
phosphatase remains bound to the membrane only at the sites
containing biotinylated RNA. Upon incubation with a solution
of a pro-luminescent compound (CDP-Star), biotinylated RNA
can be identified due to alkaline phosphatase-induced conversion
of CDP-Star into a chemiluminescence marker.
Fig. 9 shows examples of transcription priming reactions of a
55-mer RNA (HP-WTL) with initiator molecule 2. In individual
experiments, different concentrations of 2 were applied, whereas
the concentrations of NTPs were kept constant. Fig. 9A shows
the effect of augmentation of the initiator concentration on total
RNA yield: the higher the initiator concentration, the smaller the
total yield of transcription product. This is in agreement with
Fig. 9 Transcription reactions in presence of different concentrations of
2 (RNA: HPWTL 55-mer). Figures represent the ratio of initiator concen-
tration to GTP concentration (2 mM) in standard transcription reactions
(concentrations of ATP, CTP, and UTP were also 2 mM). A) Standard
reaction analysed by denaturing polyacrylamide gel electrophoresis. Bands
are visualised by UV shadowing. RNA yield diminishes with growing
initiator concentration. B) Chemiluminescence proof of incorporation of
the biotin-carrying initiator molecule. Note that there is no spot for the
transcription reaction without initiator molecule.
This journal is
The Royal Society of Chemistry 2008
Org. Biomol. Chem., 2008, 6, 899–907 | 903
©

View Article Online
similar results obtained for other derivatives.14 Fig. 9B shows the
incorporation of the initiator molecule by detecting the present
biotin with a chemiluminescence assay. Obviously, an excess of
initiator molecule 2 over natural GTP does not significantly
enhance the yield of primed RNA. Thus, it can be concluded
that a 1 : 1 molar ratio of initiator 2 to natural GTP is most
suitable for both overall yield of the transcript and incorporation
of the initiator. Similar results have been obtained with
initiator 1.
In order to better quantify the efficiency of incorporation
of the synthesised initiator molecule, we chemically synthesised
a 39-mer RNA HHR6-St. The phosphoramidite 3 of biotiny-
lated adenosine was coupled via a hexaethylene glycol unit
to the 5ꢀ-end of this RNA by solid phase chemistry. Thus,
the chemically synthesised RNA HHR6-St is identical to the
RNA HHR6 that was functionalised statistically by transcription
priming experiments with initiator 1, and nearly identical to
RNA HHR6 that was prepared and primed with initiator 2
(Fig. 10A).
The chemically synthesised 39-mer was used as a standard
to estimate the extent of incorporation of initiator in a tran-
scription priming reaction. To this end, defined amounts of
statistically labelled 39-mer HHR6 obtained by transcription
priming with 2 were blotted together with defined amounts of
the fully labelled synthetic standard RNA HHR6-St. The chemi-
luminescence that appeared upon incubation with streptavidin-
linked alkaline phosphatase and the pro-luminescent compound
CDP-Star as described above was analysed in a photo-system,
and quantified by analysis of the chemiluminescence intensities.
By comparison between the standard and the priming product
chemiluminescence, the efficiency could be determined to be ∼2%.
Changes in the ratio of concentrations of initiator 2 and GTP
during the priming reactions did not significantly improve this
efficiency. The rather low incorporation efficiency could be due
to a comparably low hydrophobicity of the initiator molecule,
which has been shown to reduce incorporation.11,23 Additionally,
the steric hindrance of both 1 and 2 might also diminish
efficiency.
Even though transcription priming with the initiator molecules
1 and 2 has so far delivered only about 2% of the RNA population
as primed products, the strategy has the potential to be used for
functionalization of RNA libraries. The sequence complexity of
RNA libraries is supposed to be about 1015 to 1016 molecules; 2% of
this library still corresponds to 1013 to 1014 molecules. Therefore,
initial transcription priming with 2% yield decreases the size of
the library by two orders of magnitude, and even though the
major part of the initial population is lost, the remaining part
is still well in the range of what can be termed a library.24 Protein
libraries for example may involve only 107 to 1010 members and yet
have been extensively used in molecular evolution with impressive
results.
Fig. 10 A) Chemical structure of the 5ꢀ-end of RNA obtained by
transcription priming experiments with initiator 2. B) Chemical structure
of the 5ꢀ-end of a standard RNA (HHR6-St) synthesised on the solid phase
using phosphoramidite 3 in the last coupling step. Note that in A) this label
is incorporated depending on the incorporation efficiency, whereas in B) it
is introduced chemically during synthesis and therefore conjugated to all
RNA molecules.
Experimental
General
To the best of our knowledge, the molecules presented here
belong to the largest initiator molecules that have been used so
far for transcription priming. We have developed an efficient
synthetic route to biotinylated adenosine–guanosine conjugates
that are accepted by T7 RNA polymerase and thus incorporated
at the 5ꢀ-end of RNAs prepared by run-off transcription. Fur-
ther studies to improve the incorporation efficiency are under
way.
Dry methylene chloride and dry methanol were obtained from
Fluka. Pyridine was dried overnight over KOH, heated to reflux
for 2 h, distilled off and stored over molecular sieves. Diiso-
propylethylamine was stored over calcium hydride and distilled
before use. All other reagents, chemicals, buffers and solvents were
obtained as the highest commercially available grade and used
without further purification. Silica gel for column chromatography
(0.063–0.2 mm) was obtained from Sigma-Aldrich. Reactions were
904 | Org. Biomol. Chem., 2008, 6, 899–907
This journal is
The Royal Society of Chemistry 2008
©

View Article Online
carried out at room temperature unless stated otherwise, and in
the case of moisture-sensitive compounds under an atmosphere
of argon. HPLC spectra were performed on an Akta Purifier
3ꢀ-O-,5ꢀ-O-(1,1ꢀ,3,3ꢀ-Tetraisopropyldisiloxan-1,3-diyl)-N6-(3,6-
dioxa-8-Nꢀ-biotinylaminooctyl)adenosine 10
¨
1.3 g 8 (2.6 mmol) was coevaporated three times with 10 mL
of dry pyridine and finally dissolved in 15 mL of dry pyridine.
The mixture was cooled and 0.8 mL of 1,3-dichloro-1,1ꢀ,3,3ꢀ-
tetraisopropyldisiloxane (0.8 g, 2.5 mmol) in 2 mL of dry DMF
were added. The mixture was stirred overnight and quenched
by adding 5 mL of water. The solvent was removed in vacuo
and the residue worked up as described for 7. Chromatographic
purification (CHCl3–MeOH = 92 : 8) yielded 1.6 g of 9 as a
yellow oil (2.1 mmol, 84%). This oil was dissolved in 6 mL of
50% CF3COOH in CHCl3, stirred for 20 min, diluted with 80 mL
of diethyl ether, and neutralised with a saturated solution of
sodium hydrogen carbonate. The aqueous phase was extracted
with diethyl ether (3 × 60 mL) and the organic phase removed.
0.4 g biotin (1.6 mmol), 0.5 g TSTU (1.7 mmol) and 0.7 mL
diisopropylethylamine (0.5 g, 4 mmol) were dissolved in 5 mL of
DMF and 5 mL of dioxane. This mixture was added to the yellow
oil and stirred overnight at room temperature. After removal
of the solvent, the residue was directly subjected to silica gel
chromatography with CHCl3–MeOH = 9 : 1 as solvent. After
removal of the solvent, 0.9 g of a colourless oil 10 was obtained
(1.0 mmol, 67%). dH(300 MHz; DMSO-d6; Me4Si) 1.05 [24 H,
m, Si–CH–(CH3)2], 1.26–1.61 (6 H, m, biotin-H-4ꢀ, biotin-H-5ꢀ,
biotin-H-3ꢀ), 2.06 (2 H, t, J 7.2, biotin-H-2ꢀ), 2.81 (1 H, m, biotin-
H-6x), 3.08 (1 H, m, biotin-H-6y), 3.17 (1 H, m, biotin-H-4),
3.36-3.60 (12 H, m, ether-H), 3.92-4.14 (3 H, m, 4ꢀ-H, 2 × H-5ꢀ),
4.28 (1 H, m, biotin-H-3a), 4.50 (1 H, m, biotin-H-6a), 4.79 (1 H,
m, H-2ꢀ), 5.10 (1 H, m, H-3ꢀ), 5.88 (1 H, m, H-1ꢀ), 6.36 (1 H, s,
biotin-N-H1), 6.42 (1 H, s, biotin-N-H3), 8.16 (1 H, s, H-2), and
8.23 (1 H, s, H-8); dC(400 MHz; DMSO-d6; Me4Si) 12.0, 12.2, 12.4,
12.7 [Si–CH–(CH3)], 16.8, 16.9, 16.9, 17.0, 17.2, 17.2, 17.3, 17.4
[Si–CH–(CH3)], 25.2 (biotin-C-3), 28.0, 28.2 (biotin-C-5, biotin-
C-4), 35.1 (biotin-C-2), 38.4 (biotin-C-9), 55.4 (biotin-C-6), 59.2
(biotin-C-8), 60.8 (C-5ꢀ), 61.0 (biotin-C-7), 68.7, 69.2, 69.5, 69.5,
69.8 (ether-C, C-3ꢀ), 73.7 (C-2ꢀ), 81.9 (C-4ꢀ), 89.4 (C-1ꢀ), 119.7 (C-
5), 139.1 (C-8), 147.9 (C-4), 152.3 (C-2), 154.5 (C-6), 162.7 (biotin-
C-10), 172.1, and 172.8 (biotin-C-1); MALDI+-MS: calculated for
C38H66N8O9Si2 [MH+]: 867.2; found: 867.9.
(Amersham Biosciences) with the columns described. NMR
spectra were recorded with TMS as the internal standard on
the following machines: 300 MHz: ARX-300; 400 MHz: Bruker
DRX-400; 600 MHz: Bruker DRX-600. J values are given in Hz.
Mass spectra were recorded on a Bruker Autoflex MALDI-TOF,
VG Autospec (FAB), and ESI Bruker Esquire LC (Ion Trap).
UV measurements were obtained using an Ultrospec 2100 pro
R
(Amersham Biosciences) Cary 1E or a Varian Nano Dropꢀ ND-
1000 spectrophotometer.
2ꢀ-O-,3ꢀ-O-,5ꢀ-O-Triacetyl-6-iodopurine-9-b-D-riboside 7
17 g of 2ꢀ-O-,3ꢀ-O-,5ꢀ-O-triacetyladenosine 6 (43 mmol) were
suspended in 110 mL methylene iodide (1.3 mol) and 110 mL
isopentyl nitrite (816 mmol). The suspension was stirred at 60 ◦C
for 5 h, then at room temperature overnight. The solvent was
removed in vacuo and the residue dissolved in CHCl3 and worked
up as described for 6 in the ESI†. The residue was purified
chromatographically (silica gel, hexane to remove iodine, then
CHCl3 → CHCl3–MeOH = 95 : 5). 7 was obtained as a colourless
foam (9.3 g, 18.4 mmol, 43%); dH(400 MHz; DMSO-d6; Me4Si)
=
2.01, 2.04, 2.12 [9 H, 3 × s, 3 × (C O)–CH3], 4.29 (1 H, dd, J 12.9
and 6.4, H-5ꢀ), 4.40–4.44 (2 H, m, H-4ꢀ, H-5ꢀ), 5.63–5.66 (1 H, m,
H-3ꢀ), 6.02–6.05 (1 H, m, H-2ꢀ), 6.33 (1 H, d, J 5.1, H-1ꢀ), 8.69
(1 H, s, H-2), and 8.85 (1 H, s, H-8); dC(400 MHz; DMSO-d6;
ꢀ
=
Me4Si) 20.0, 20.2, 20.3 [3 × (C O)–CH3], 62.5 (C-5 ), 69.8 (C-
3ꢀ), 71.9 (C-2ꢀ), 79.6 (C-4ꢀ), 86.1 (C-1ꢀ), 123.2 (C-5), 138.6 (C-8),
145.2 (C-6), 147.3 (C-4), 151.9 (C-2), 169.1, 169.2, and 169.8 [3 ×
+
+
=
(C O)–CH3]; FAB -MS: calculated for C16H17IN4O7 [M + H ]:
505.2; found: 505.1.
N6-(3,6-Dioxa-8-Nꢀ-BOC-aminooctyl)adenosine 8
22 g N-BOC-2,2ꢀ-(ethylenedioxy)bis(ethylamine) (88 mmol) were
added to a solution of 9.3 g 2ꢀ-O-3ꢀ-O-,5ꢀ-O-triacetyl-6-iodopurine-
9-b-D-riboside 7 (18.4 mmol) in 50 mL ethanol. The mixture was
heated to 78 ◦C and stirred for 6 h under reflux. The solvent
was removed in vacuo, the residue dissolved in CHCl3, and the
organic layer was worked up as described for 7. 30 mL of aqueous
ammonium hydroxide (25%) were added and the solution was
stirred for 1 h. Afterwards, the solvent was removed in vacuo, and
the residue dissolved in CHCl3 and worked up as described for
7. The crude product 8 was purified by chromatography (silica
gel, CHCl3–MeOH = 9 : 1) and obtained as a colourless foam
(5.7 g, 11 mmol, 64%); dH(400 MHz; DMSO-d6; Me4Si) 1.36
(9 H, s, BOC), 2.66–2.69, 3.05–3.09, 3.37–3.40 (12 H, 3 × m,
ether-H), 3.60–3.69 (2 H, m, H-5ꢀ), 3.96-3.99 (1 H, m, H-4ꢀ),
4.15–4.17 (1 H, m, H-3ꢀ), 4.59-4.62 (1 H, m, H-2ꢀ), 5.90 (1 H,
d, J 6.1, H-1ꢀ), 8.31 (1 H, s, H-2), and 8.35 (1 H, s, H-8);
Solid phase synthesis of initiator 1 and RNA HHR6
For the synthesis of 1, phosphoramidite 3 was synthesised from
10 as described in the general procedure (ESI†). Initiator 1 and
RNA HHR6 were prepared by solid phase phosphoramidite
chemistry as described previously.26 In the case of initiator 1, two
syntheses (2 × 1 lmol scale) were performed, in which solid bound
guanosine [ChemGenes G (N-PAC) 3ꢀ-tBDSilyl 2ꢀ-lcaa CPG 1000
˚
A] was first coupled to a hexaethylene glycol phosphoramidite
(ChemGenes DMT-hexaethyloxy-glycol phosphoramidite) and
phosphoramidite 3 in the second step. Afterwards, the solid phase
was treated overnight with 2 mL of ammoniacal methanol and
separated from the solid support. The solvent was removed and
the residue treated with 400 lL triethylamine tris(hydrofluoride)–
DMF = 3 : 1 (v/v) for 90 min at 55 ◦C. The solvent was removed
in vacuo and the residue was dissolved in 1 mL saturated aqueous
NaHCO3. 1 was purified by reverse phase chromatography [col-
umn: Macherey Nagel VP 250/10 Nucleodur 100-5 C18 ec; buffers
=
dC(400 MHz; DMSO-d6; Me4Si) 28.1 [BOC–(C O)–O–(CH3)3],
61.5 (C-5ꢀ), 69.0, 69.4 (2 × ether), 70.5 (C-3ꢀ), 72.3 (C-2ꢀ), 79.0
(Cq-BOC), 85.8 (C-4ꢀ), 87.9 (C-1ꢀ), 122.5 (C-5), 139.7 (C-8),
=
148.5 (C-4), 152.2 (C-2), 155.5 (C-6), and 157.9 [BOC–(C O)–O–
(CH3)3]; FAB+-MS: calculated for C21H34N6O8 [M + H+]: 499.5;
found: 499.3.
This journal is
The Royal Society of Chemistry 2008
Org. Biomol. Chem., 2008, 6, 899–907 | 905
©

View Article Online
used were (A) 0.1 M triethylammonium acetate (pH 7.0) and
(B) 0.1 M triethylammonium acetate (pH 7.0), 50% acetonitrile;
flow rate 4 mL min−1; gradient: 0% B for 5 min 30 s, linear
augmentation to 100% B over 50 min, 100% for 5 min, 0%
for 5 min]. The retention time of 1 was 40 min. The fractions
containing 1 were pooled, the solvent was removed, the residue
dissolved in 1 mL of water, filtered (0.2 lm), and the yield was
determined by UV (e = 27 100 cm−1 M−1) to be 240 nmol; ESI−-
MS: calculated for C48H77N13O24P2S [M−] 1313.2; found: 1312.5.
HHR6St (5ꢀ-UGGG CAG CUG AUG AGC UCC AAA UAG
AGC GAA AGU UAC ACC-3ꢀ, where U denotes a modification
as shown in Fig. 10) was synthesised by solid phase syn-
thesis on a Gene Assembler Special (Pharmacia) with PAC-
phosphoramidites, using hexaethylene glycol phosphoramidite
(ChemGenes DMT-hexaethyloxy-glycol phosphoramidite) and
phosphoramidite 3 in the second-to-last and final coupling steps,
respectively. The obtained RNA was deprotected as described26
and purified by electrophoresis using a 10% denaturing gel for
RNA purification. Elution was carried out using 0.5 M LiOAc
followed by acetone precipitation.
ramidite 14 as described in the ESI†. 0.23 mmol of 15 were
coupled to 14 in the same way as described for the coupling of
21 and mono-DMT-hexaethylene glycol. The resulting coupling
product 22 was purified chromatographically (CHCl3–MeOH =
95 : 5; Rf = 0.4) and gave 0.19 mmol of 22 (83%); MALDI+-MS:
calculated for C65H105N11O29P2 [MH+]: 1567.5; found: 1566.6. 22
(1.0 g, 63 lmol) was treated with 30% CF3COOH in CH2Cl2 for
20 min. The mixture was diluted with 50 mL CH2Cl2, neutralised
with saturated NaHCO3 solution (50 mL), and washed with brine
(50 mL). The aqueous solution was extracted with CH2Cl2 (3 ×
50 mL). The organic solvent was removed and the crude product
23 was used in the following steps without further purification. A
solution of 14 mg biotin (55 lmol), 17 mg TSTU (55 lmol), and
28 lL (20 mg, 160 lmol) diisopropylethylamine in a mixture of
0.5 mL dioxane–DMF = 1 : 1 was stirred for 10 min and added
to 23. The mixture was stirred overnight at room temperature,
and worked up as described for 10. Chromatographic purification
was performed with CHCl3–MeOH = 9 : 1. Compound 24 was
obtained as a colourless oil (48 mg, 28 lmol, 52%); dH(600 MHz;
DMSO-d6; Me4Si) 1.01, 1.05, 1.15 [30 H, 3 × m, 5 × –CH–(CH3)2],
1.31 (2 H, m, biotin-H-4ꢀ), 1.50, 1.62 (4 H, m, biotin-H-3ꢀ, biotin-
H-5ꢀ), 2.07 (2 H, m, biotin-H-2ꢀ), 2.52–2.67 [5 H, 3 × m, –CH–
(CH3)2], 2.74 (1 H, m, biotin-H-6x), 2.90 (1 H, m, biotin-H-6y),
3.10 (1 H, m, biotin-H-4), 3.19, 3.40 (4 H, 2 × m, ether-H), 3.48–
3.67 (6 H, m, P–O–CH3; 28 H, m, ether-H), 4.06, 4.13, 4.31 (10 H,
3 × m, 1 biotin-H-6a, 1 biotin-H-3a, 4 ether-H, 2 guanosine-H-5ꢀ,
2 adenosine-H-5ꢀ), 4.41 (2 H, m, guanosine-H-4ꢀ, adenosine-H-4ꢀ),
5.53 (1 H, m, guanosine-H-3ꢀ), 5.67 (1 H, m, adenosine-H-3ꢀ), 5.85
(1 H, m, guanosine-H-2ꢀ), 6.00 (1 H, m, adenosine-H-2ꢀ), 6.10 (1 H,
d, J 6.8, guanosine-H-1ꢀ), 6.23 (1 H, d, J 5.5, adenosine-H-1ꢀ), 6.33
(1 H, s, biotin-NH-3), 6.39 (1 H, s, biotin-NH-1), 8.23 (1 H, s,
adenosine-H-2), 8.26 (1 H, s, guanosine-H-8), and 8.36 (1 H, s,
adenosine-H-8); dC(600 MHz; DMSO-d6; Me4Si) 18.9, 19.0, 19.1,
19.3 (6 × isobutyryl-CH3), 25.4, 28.6, 28.7, 31.1, 31.2, 33.4, 33.6
Hexaethylene glycol-2ꢀ,3ꢀ,N2-triisobutyrylguanosine conjugate 15
Mono-BOC-protected DMT-hexaethylene glycol (4 mmol) was
coevaporated twice with 10 mL of dry pyridine and stored in vacuo
overnight. Then it was dissolved in 5 mL of dry acetonitrile and
mixed with 4 mmol of dried 2ꢀ-O-,3ꢀ-O-,N2-triisobutyrylguanosine
(methyl-N,N-diisopropyl)phosphoramidite 21, which was also
dissolved in 1 mL of dry acetonitrile (the synthesis of 21 is
described in the ESI†). The mixture was cooled in an ice bath
and 8.5 mL of a 0.45 M solution of tetrazole (4 mmol) in
acetonitrile was added. The mixture was stirred for 30 min
and 5.5 mL of a 0.1 M solution of iodine in collidine–THF–
H2O = 2 : 2 : 1 was added dropwise. After a further 5 min,
excess iodine was reduced with 1 M Na2S2O3. The solvent was
removed in vacuo, the residue was dissolved in CHCl3, worked up
and purified chromatographically as described for 6 (CHCl3 →
CHCl3–MeOH = 95 : 5). The purified coupling product was
DMT-deprotected with 5% CCl3COOH in CH2Cl2 and purified
chromatographically as described for 12. Compound 15 was
obtained as a colourless oil (1.3 g, 1.4 mmol, 35%); dH(400 MHz;
DMSO-d6; Me4Si) 1.01, 1.04 [2 × 3 H, 2 × d, J 7.0, 3ꢀ-O–CH–
(CH3)2], 1.13-1.17 [12 H, m, 2ꢀ-O–CH–(CH3)2, NH–CH–(CH3)2],
2.50 [1 H, sept., J 7.0, 3ꢀ-O–CH–(CH3)2], 2.64 [1 H, sept., J 7.0, 2ꢀ-
=
[6 × (C O)–CH–(CH3)2], 35.3, 35.6, 36.2, 38.7, 38.9, 54.4, 54.6,
55.4, 55.9, 59.7, 61.5, 66.4, 66.7, 66.9, 67.0, 67.1, 67.2, 67.3, 69.6,
69.7, 70.0, 70.2, 70.5, 70.7 (ether, C-3ꢀ, C-5ꢀ, biotin), 72.5, 72.6 (2 ×
C-2ꢀ), 80.8, 81.5 (2 × C-4ꢀ), 85.1, 86.1 (2 × C-1ꢀ), 120, 121 (2 × C-
5), 138.2, 140.0 (C-8), 148.8, 149.1 (C-2, C-4), 155.1, 155.2 (C-6),
162.8, 163.3 (biotin-C-10), 172.6 (biotin-C-1), 175.2 [isobutyryl-
=
=
O–(C O)], 175.4 [isobutyryl-O–(C O)], and 180.6 [isobutyryl-
=
N-(C O)]. For the deprotection of 24, 200 mg of disodium-2-
carbamoyl-2-cyanoethylen-1,1-dithiolate was dissolved in 2 mL
of dry DMF and added to 48 mg (28 lmol) of 24. The solution
was stirred for 10 min and the solvent removed in vacuo. The
residue was dissolved in 5 mL 32% NH4OH–CH3NH2 = 1 :
1 (v/v) and stirred at 50 ◦C for 90 min. The solvent was removed
in vacuo, and the residue dissolved in water and purified by
HPLC. A reverse phase column was used (Knauer, Nucleosil-
120, C18, 10 lm; buffers used were (A) 0.1 M triethylammonium
acetate, pH 7.0 and (B) 0.1 M triethylammonium acetate, pH 7.0,
50% acetonitrile; flow rate 3 mL min−1; gradient: 0% B for 8 min
30 s, linear augmentation to 100% B over 24 min, 100% for 9 min,
0% for 6 min). The retention time was 32 min. The fractions
containing 2 were pooled, the solvent was removed, the residue
was dissolved in 1 mL of water, filtrated (0.2 lm) and the yield
was determined by UV (e = 27 100 cm−1M−1) to be 9.2 lmol;
dH(600 MHz; DMSO-d6; Me4Si) 1.35 (2 H, m, biotin-H-4ꢀ), 1.56
(4 H, m, biotin-H-3ꢀ, biotin-H-5ꢀ), 2.13 (2 H, m, biotin-H-2ꢀ), 2.63,
=
O–CH–(CH3)2], 2.79 [1 H, sept., J 6.8, NH–(C O)–CH–(CH3)2],
3.32-3.72 (m, 20 H, ether-H), 3.57-3.62 (2 H, m, H-5ꢀ), 3.67 (3 H, s,
O–CH3), 4.05–4.10, 4.31–4.35 (2 × 2 H, 2 × m, ether-H), 4.51-
4.53 (1 H, m, H-4ꢀ), 5.51-5.54 (1 H, m, H-3ꢀ), 5.83–5.87 (1 H,
m, H-2ꢀ), 6.10 (1 H, d, J 6.8, H-1ꢀ), and 8.22 (1 H, s, H-8);
dC(400 MHz; DMSO-d6; Me4Si) 18.3–18.7 (6 × isobutyryl-CH3),
32.7 [3ꢀ-O–CH–(CH3)2], 33.0 [2ꢀ-O–CH–(CH3)2], 34.7 [NH–CH–
(CH3)2], 54.1 (P–O–CH3), 60.0, 66.6, 69.5 (ether-C), 71.9 (C-2ꢀ),
70.2 (C-3ꢀ), 80.9 (C-4ꢀ), 84.6 (C-1ꢀ), 120.4 (C-5), 137.6 (C-8), 148.2,
=
148.5 (C-2, C-4), 154.6 (C-6) 174.6, 174.8 [O–(C O)], and 180.0
+
+
=
[N–(C O)]; MALDI -MS: calculated for C35H58N5O17P [MH ]:
852.8; found: 852.9.
Initiator 2
13 (0.5 mmol) was reacted with (methyl-N,N-diisopropyl)-
chlorophosphoramidite to yield the respective methyl phospho-
906 | Org. Biomol. Chem., 2008, 6, 899–907
This journal is
The Royal Society of Chemistry 2008
©

View Article Online
2.66 (2 H, 2 × m, biotin-H-6x, -H-6y), 3.15 (1 H, m, biotin-H-4),
3.24 (2 H, m, ether-H), 3.45, 3.54, 3.67 (30 H, m, ether-H), 3.82
(4 H, m, ether-H), 3.87-3.99 (4 H, m, guanosine-H-5ꢀ, adenosine-
H-5ꢀ), 4.05, 4.11 (2 × 1 H, 2 × m, adenosine-H-4ꢀ, guanosine-
H-4ꢀ), 4.21 (2 H, m, adenosine-H-3ꢀ, biotin-H-3a), 4.26 (1 H, m,
guanosine-H-3ꢀ), 4.36 (1 H, m, biotin-H-6a), 4.67, 4.56 (2 × 1 H,
2 × m, guanosine/adenosine-H-2ꢀ), 5.77 (1 H, d, J 9.1, guanosine-
H-1ꢀ), 6.00 (1 H, d, J 8.8, adenosine-H-1ꢀ), 6.42 (1 H, s, biotin-
NH-3), 6.47 (1 H, s, biotin-NH-1), 7.98, 8.29, and 8.50 (3 H,
3 × s, adenosine-H-8 and -H-2, guanosine-H-8); MALDI+-MS:
calculated for C48H77N13O24P2S [MH+]: 1315.2; found: 1315.0.
protocol used was as follows: 2 × WP 5 min; 2 × BP 5 min, 1 × BP
30 min, 1 × BP 30 min + 1 lL streptavidin-alkaline phosphatase
(Promega); 1 × 15 min BP, 3 × 10 min WP, 2 × 2 min AP. After
further washing steps, the membrane was incubated with a 0.25 M
solution of 2-chloro-5-{5ꢀ-chloro-4-methoxyspiro[1,2-dioxetane-
3,2ꢀ-tricyclo(3.3.1.13,7)decan]-4-yl} dihydrogenphosphate (“CDP-
Star”, Roche). For qualitative detection, chemiluminescence was
detected with an X-ray film. For quantitative detection, chemi-
luminescence was detected in a photo-system (Vilber Lourmat
Darkroom-CN-3000) and integrated using the program “Im-
age J”.
Transcription priming reactions
References
HHR6 (5ꢀ-GGG CAG CUG AUG AGC UCC AAA UAG AGC
GAA AGU UAC ACC-3ꢀ) and HPWTL (5ꢀ-G GGA GAA AGA
GAG AAG UGA ACC AGA GAA ACA CAC GUU GUG GUA
UAU UAC CUG GUA-3ꢀ) were labelled with initiator molecules
1 or 2 by in vitro transcription from a double-stranded DNA
template. The DNA templates were obtained fromtwo overlapping
synthetic DNA primers (Purimex) by enzymatic strand extension
using Klenow Fragment exo− (MBI Fermentas).25 In a standard
transcription reaction (50 ll), the concentrations of the
components were as follows: 2 mM NTPs, 50 pmol ds DNA, 1 ×
HEPES, 5 U ll−1 T7 RNA polymerase. Initiator concentrations
(1 or 2) depended on the specific reaction and varied between
0 and 8 mM. Reaction components were mixed and incubated
for 2 h at 37 ◦C. The mixture was subjected to a phenol–
chloroform extraction, the transcription product obtained was
precipitated with ethanol, purified by denaturing polyacrylamide
gel electrophoresis (10%), visualised by UV-shadow, excised,
eluted with 0.5 M LiOAc, and precipitated with acetone.
Determination of yield was accomplished by UV measurement.
1 W. Gilbert, Nature, 1986, 319, 618.
2 B. Seelig and A. Ja¨schke, Chem. Biol., 1999, 6, 167–176.
3 G. Sengle, A. Eisenfu¨hr, P. S. Arora, J. S. Nowick and M. Famulok,
Chem. Biol., 2001, 8, 459–473.
4 S. Fusz, A. Eisenfu¨hr, S. G. Srivatsan, A. Heckel and M. Famulok,
Chem. Biol., 2005, 12, 941–950.
5 B. K. Oh and N. R. Pace, Nucleic Acids Res., 1994, 22, 4087–4094.
6 M. E. Harris and E. L. Christian, Methods, 1999, 18, 51–59.
7 A. B. Burgin and N. R. Pace, EMBO J., 1990, 9, 4111–4118.
8 J. R. Sampson and O. C. Uhlenbeck, Proc. Natl. Acad. Sci. U. S. A.,
1988, 85, 1033–1037.
9 C. Pitulle, R. G. Kleineidam, B. Sproat and G. Krupp, Gene, 1992, 112,
101–105.
10 F Huang, Nucleic Acids Res., 2003, 31, e8/1–e8/8.
11 R. Fiammengo, K. Musilek and A. Ja¨schke, J. Am. Chem. Soc., 2005,
127, 9271–9276.
12 L. Zhang, L. Sun, Z. Cui, R. L. Gottlieb and B. Zhang, Bioconjugate
Chem., 2001, 12, 939–948.
13 J. C. Schlatterer and A. Ja¨schke, Biochem. Biophys. Res. Commun.,
2006, 344, 887–892.
14 B. Seelig and A. Ja¨schke, Bioconjugate Chem., 1999, 10, 371–378.
15 V. Nair and S. G. Richardson, J. Org. Chem., 1980, 45, 3969–3974.
16 S. Bae and M. K. Lakshman, J. Am. Chem. Soc., 2007, 129, 782–789.
17 E. A. Ve´liz and P. A. Beal, Tetrahedron Lett., 2000, 41, 1695–1697.
18 E. M. Van Der Wenden, J. K. von Frijtag Drabbe Ku¨nzel, R. A. Mathoˆt,
M. Danhof, A. P. Ijzerman and W. Soudijn, J. Med. Chem., 1995, 38,
4000–4006.
Blotting of biotin-containing samples
19 G. B. Sigal, M. Mammen, G. Dahmann and G. M. Whitesides, J. Am.
The biotinylated RNA probe was purified on a denaturing poly-
acrylamide gel (10%, 8 × 10 cm) and transferred onto a same-sized
nylon membrane by electrophoresis. The membrane was heated to
75 ◦C for 5 min and washed using the following solutions: washing
buffer (WP): 130 mM NaCl, 3 mM KCl, 12 mM Na2HPO4,
2 mM KH2PO4, 0.5% SDS, pH 7.4; blocking buffer (BP): same
as washing buffer, plus 2 g l−1 coffee whitener; assay buffer (AP):
0.1 M Tris, 0.1 M NaCl, pH 9.5. The membrane was covered with
the respective buffer solution in a container and moved gently. The
Chem. Soc., 1996, 118, 3798–3799.
20 S. Mu¨ller, J. Wolf and S. A. Ivanov, Curr. Org. Synth., 2004, 1, 293–307.
21 C. B. Reese, Org. Biomol. Chem., 2005, 3, 3851–3868.
22 B. J. Dahl, K. Bjerga˚rde, L. Henriksen and O. Dahl, Acta Chem. Scand.,
1990, 44, 639–641.
23 F. Huang, G. Wang, T. Coleman and N. Li, RNA, 2003, 9, 1562–1570.
24 C. Wilson and J. W. Szostak, Nature, 1995, 374, 777–782.
25 C. Schmidt, R. Welz and S. Mu¨ller, Nucleic Acids Res., 2000, 28, 886–
894.
26 R. Welz, K. Bossmann, C. Klug, C. Schmidt, H. J. Fritz and S. Mu¨ller,
Angew. Chem., Int. Ed., 2003, 42, 2424–2427.
This journal is
The Royal Society of Chemistry 2008
Org. Biomol. Chem., 2008, 6, 899–907 | 907
©