Biotin-conjugated N-methylisatoic anhydride: A chemical tool for nucleic acid separation by selective 2′-hydroxyl acylation of RNA

Article abstract of DOI:10.1039/c4cc01134a

An isatoic anhydride derivative conjugated to a biotin and a disulfide linker was specifically designed for the separation of nucleic acids. Starting from a DNA-RNA mixture, a selective 2′-hydroxyl acylation of RNAs followed by capture with streptavidin-coated magnetic beads and cleavage of the disulfide led to elution of RNAs. the Partner Organisations 2014.

Full text of DOI:10.1039/c4cc01134a

ChemComm
View Article Online
View Journal | View Issue
COMMUNICATION
Biotin-conjugated N-methylisatoic anhydride:
a chemical tool for nucleic acid separation by
selective 20-hydroxyl acylation of RNA†
Cite this: Chem. Commun., 2014,
50, 5748
Received 12th February 2014,
Accepted 7th April 2014
S. Ursuegui,ab N. Chivot,c S. Moutin,c A. Burr,c C. Fossey,ab T. Cailly,ab A. Laayoun,c
F. Fabis*ab and A. Laurent*c
DOI: 10.1039/c4cc01134a
www.rsc.org/chemcomm
An isatoic anhydride derivative conjugated to a biotin and a disulfide leaves DNA contamination. The use of DNAses remains very
linker was specifically designed for the separation of nucleic acids. expensive and does not avoid a further denaturation step to
Starting from a DNA–RNA mixture, a selective 20-hydroxyl acylation inactivate DNAses before any amplification. Other methods based
of RNAs followed by capture with streptavidin-coated magnetic on chromatographic separation such as affinity or anion exchange
beads and cleavage of the disulfide led to elution of RNAs.
chromatography can be used but lead to contaminated samples or
are not easy to perform in most laboratories.6 Traces of amplifica-
Since RNAs have been evidenced as playing fundamental roles tion inhibitors such as proteins or small molecules, cost, reagent
in life such as in coding or regulating important biological toxicity and poor automation are also major drawbacks of these
processes, they have been the focus of intense research efforts processes. Therefore, developing new methodologies for specific
in order to establish their structure and functions. Advances in RNA extraction devoid of these disadvantages would be a signifi-
understanding roles and structures of RNA have been enabled by cant step forward in the field of molecular biology including
altering structures of RNA using simple chemical modifications, diagnostic applications. Among the chemical methods for bio-
site-specific labeling or conjugation to other biomolecules. The logical RNA labeling, the 20-OH selective modifications are of
manipulation of RNA to introduce labeling groups includes interest because the 20-OH group constitutes a specific chemical
chemical RNA synthesis, post-synthetic RNA conjugation, enzymatic signature of RNAs. Acylation of oligonucleotides using acetic
incorporation of tags into non-synthetic RNA strands, or chemical anhydride has been described for a long time,7 whereas the
and enzymatic ligation of RNA strands.1,2 Besides these manipula- selective acetylation of the 20-OH of mRNA was later described
tions of modified RNAs for nanotechnology applications, there as exhibiting high stability against ribonucleases.8 Isatoic
is great interest in the detection of natural RNAs as biomarkers anhydride (IA) derivatives and their N-methylated counterparts
for diagnostic applications or following treatment efficacy. For N-Methyl Isatoic Anhydrides (NMIAs) were recently described as
this detection to be efficient, the RNAs need to be extracted 20-OH selective acylation agents of RNAs, and widely used for
from a biological sample in an intact and pure form before being resolving secondary RNA structures using the SHAPE (Selective
amplified.3 A number of methods already exist for extraction and 20-Hydroxyl Acylation Analyzed by Primer Extension) technology.9
purification of RNAs.4 Most of them are based on the biophysical
Owing to the selective reactivity of NMIA toward the 20-OH
properties of nucleic acids, and the most currently used method is group of ribose, we reasoned that a modified NMIA could be
the guanidinium–thiocyanate–phenol-extraction in which RNA designed for the selective tagging of RNAs and used to discri-
and DNA partition, respectively, between the aqueous and organic minate DNA and RNA. The two nucleic acids could then be
phases.5 This time-consuming and manual method requires to be separated, and purified before being used in different applica-
carefully carried out, uses many centrifugation steps and mostly tions where their detection is required (Fig. 1).
As a proof of concept, we designed a NMIA conjugated to a
biotin through a linker containing a disulfide bond and a pegylated
moiety which was used to improve water solubility (Fig. 2). The
biotin was used as a recognition group for specific solid phase
support capture of acylated RNA with streptavidin-coated mag-
netic beads. Captured RNA should be released from its support
by the chemical cleavage of the disulfide bond. The RNA tag 7
was synthesized and first evaluated for the selective acylation
a Normandie Univ, France
b
´
Universite de Caen Basse-Normandie, Centre d’Etudes et de Recherche sur le
´
Medicament de Normandie (EA 4258 - FR CNRS 3038 INC3M - SF 4206 ICORE)
UFR des Sciences Pharmaceutiques, Bd Becquerel, F-14032 Caen, France.
E-mail: frederic.fabis@unicaen.fr
c
´
´
BioMerieux SA, Centre Christophe Merieux, Parc Polytech, 5 rue des berges,
38024 Grenoble, France. E-mail: alain.laurent@biomerieux.com
† Electronic supplementary information (ESI) available: Detailed experimental
procedures and characterization of new compounds. See DOI: 10.1039/c4cc01134a of a 27 nt oligoribonucleotide vs. oligodesoxy-nucleotide.
5748 | Chem. Commun., 2014, 50, 5748--5751
This journal is ©The Royal Society of Chemistry 2014

View Article Online
Communication
ChemComm
oligomers were separately incubated with 7 in a DMSO–TEAAc
buffer at 65 1C for 1 h. Excess tag was then removed after
precipitation, and oligomers were subjected to an enzymatic
digestion using a mixture of nuclease P1 and alkaline phos-
phatase. The resulting nucleosides were then analyzed by LC/
MS and showed, for RNA oligomers, the presence of both non-
modified and modified nucleosides. The ratio of tagged nucleo-
sides was determined after enzymatic digestion by integrating
peaks corresponding to the four free ribonucleosides and those
corresponding to acylated nucleosides (Table 1). For the RNA
oligomers, the ratio of tagged nucleosides vs. free-tagged was
found to be 7.5/92.5, while the same experiment, using DNA
oligomers, led to a ratio of 0.7/99.3. These two experiments
demonstrated the ability of 7 to selectively acylate RNA vs. DNA
with a selectivity ratio of 10.7.
Fig. 1 Principle of RNA/DNA separation.
We then turned our attention to the separation of nucleic
acids using biological models: HIV transcript for RNA and calf
The separation process was then performed with biological genomic DNA. First, the extraction process was studied with a
models of nucleic acids. HIV-transcript RNA (1500 nt) and a calf genomic DNA (20 kb)
An isatoic anhydride derivative bearing an amino group in treated separately. The separation process was then applied to a
position 6 was chosen to attach the biotin-peg4-SS-COOH which is mixture of both nucleic acids. For these experiments, nucleic
commercially available as an activated ester. However, due to the acids were incubated with 7 (1.5 or 30 mM) and the excess tag
high reactivity of IA towards nucleophiles, it was necessary to use was removed using magnetic silica particles. The biotinylated
a precursor able to give the IA in the last step of the synthesis. nucleic acids were then captured with streptavidin-coated
According to the procedure developed by Sowell et al. for the magnetic beads and eluted after the disulfide cleavage with
generation of IA derivatives, we decided to use a tert-butyl amino dithiothreitol (DTT). The quantities of RNA and DNA extracted
ester as a precursor.10 The synthesis of the tert-butyl 5-amino-2- by this method were determined by fluorescence. The selectivity
(methylamino)benzoate 5 was performed starting from 2-amino- of the capture was calculated as the RNA/DNA ratio obtained
5-nitrobenzoic acid 1 through a four-step sequence (Scheme 1). from these extraction yields (Table 2). When RNA and DNA were
6-Nitroisatoic anhydride was first generated by treatment with treated separately with 7 (1.5 mM), the extraction yields were
phosgene and subsequently N-methylated with iodomethane. The evaluated to be 27 and 0.5%, respectively, leading to a selectivity
resulting N-methyl isatoic anhydride 3 was then opened with ratio of 57. Increasing the concentration of 7 (30 mM) led to
t-BuONa leading to the tert-butyl aminoester 4. The quantitative enhancement of both RNA and DNA extraction yields to 66 and
reduction of the nitro group followed by coupling with Biot-peg4- 6% respectively. However, in this case a decreased selectivity ratio
SS-CONHS led to the conjugated aminoester 6. The latter was then of 10.5, consistent with previous results on oligonucleotides at
reacted with phosgene to afford the IA 7 in a 36% overall yield.
30 mM, was observed (Table 2). Although, a decrease in the
The selective acylation of RNA with 7 was first examined extraction yield was obtained for the lower concentration, it led to
using 27 nt RNA and DNA oligomers. Therefore, RNA or DNA a higher selectivity of RNA extraction. This selectivity associated
Fig. 2 RNA purification process.
This journal is ©The Royal Society of Chemistry 2014
Chem. Commun., 2014, 50, 5748--5751 | 5749

View Article Online
ChemComm
Communication
Scheme 1 Synthesis of RNA tag 7. Isolated yields. Reagents and conditions: (a) COCl2, dioxane, 0.5 h, rt. (b) MeI, DIPEA, DMF, 1 h, rt. (c) t-BuONa, DCM,
0.25 h, rt. (d) H2, Pd/C, AcOEt, 16 h, rt. (e) Biot-peg4-SS-CONHS, TEA, DCM, 1 h, rt. (f) COCl2, DCM, 0.25 h, rt.
Table 1 Synthetic 27 nt RNA and 27 nt DNA tagging with 7
acids leading to the co-elution of DNA with the specifically
captured RNA and not to a decrease in the selectivity of the
acylation step in the mixture.
RNA tagginga
Tagged
DNA tagginga
Tagged
Free
Free
Tagged
Owing to the low concentration of RNA obtained from
biological samples, their amplification is a prerequisite to their
efficient detection and analysis. Selective acylation of 20-OH of
RNAs has been used to stop the RT process and therefore
analyze the whole RNA structure using the SHAPE technology.9
For detection applications using amplification processes, only
short sequences (50–200 nt) of RNAs are transcribed by the reverse
transcriptase (RT), and the cDNA amplicons are then amplified.
Since the acylation process occurs randomly on the RNA sequence,
the reverse transcription of the extracted RNAs should be effective
if the RT-target sequence remains unmodified after acylation.
In order to check the compatibility of our extraction protocol
with the amplification techniques, we performed RT-PCR
experiments. RNA HIV transcripts were reacted with 15 mM
of 7, captured using streptavidin coated magnetic beads and
eluted after disulfide cleavage by DTT. The extracted RNAs were
then amplified by RT-PCR. Different experiments were per-
formed by varying the number of tagged RNA copies per sample
(50, 100 and 1000 copies, Fig. 3b) and compared to amplifica-
tion results using non-modified RNA (Fig. 3a). The results show
that HIV transcripts extracted after acylation can successfully be
amplified, even with low RNA concentrations such as 50 copies
and using a high concentration of tag (15 mM). Compared to
non-acylated RNA, we can notice a slight decrease of fluores-
cence at all concentrations. This result can either be due to the
acylation of RNA inside the amplification target sequence,
leading to a decrease in the RNA concentration available for
amplification, or to the quenching of fluorescence caused by
the presence of anthranilate residues on the RNA sequence. We
then checked the efficiency of the amplification by determination
of the cycle thresholds (CT) for both acylated and non-acylated
RNA (Fig. 3C). No significant differences were observed for the
two lower RNA template concentrations (50 and 100 copies),
whereas at 1000 copies, only one additional cycle was necessary
to reach the CT. Taken together, these results clearly demonstrate
nucleosides nucleosides nucleosides nucleosides RNA/DNA ratio
7.5 ꢀ 0.2
92.5 ꢀ 0.2
0.7 ꢀ 0.1
99.3 ꢀ 0.1
10.7 ꢀ 0.3
a
Ratio determined by LC/MS (n = 2), see experimental data for details.
Table 2 Extraction of biological RNA or DNA
Extracted
Extracted
Extracted
NA [7] (mM) RNA yielda (%) DNA yielda (%) RNA/DNA yield ratio
RNA 1.5
DNA
RNA 30
DNA
27
—
66
—
—
0.5
—
6
57
10.5
a
Determined using fluorescence spectroscopy, see experimental data
for details.
with the quality of RNAs remains the key parameter for RNA-
based diagnostic applications.
The separation process was then applied to a 10/90 RNA/DNA
mixture, considering the predominance of DNA in biological
samples. For this experiment, 7 was used at a lower concentration
of 1.5 mM in order to minimize the DNA co-extraction. Using the
same protocol, the RNA/DNA ratio after elution was calculated to
be 86/14, demonstrating thus the selectivity of the RNA acylation
and capture in a mixture of nucleic acids (Table 3). Although we
were able to reverse the initial ratio of nucleic acids, the residual
presence of DNA was observed after the separation process. This
result could be attributed to interactions between the two nucleic
Table 3 Extraction of the biological RNA–DNA mixture
[7] (mM)
NA initial ratio (RNA/DNA)
10/90
NA final ratio (RNA/DNA)a
1.5
86/14 ꢀ 1.2
a
Determined using fluorescence spectroscopy (n = 4), see experimental
data for details.
5750 | Chem. Commun., 2014, 50, 5748--5751
This journal is ©The Royal Society of Chemistry 2014

View Article Online
Communication
ChemComm
Fig. 3 RT-PCR amplification of extracted HIV RNA.
that RNA amplification remains consistent with this new extrac- Notes and references
tion process, even at low concentrations of template, allowing its
use in diagnostic applications.
1 E. Paredes, M. Evans and S. R. Das, Methods, 2011, 54, 251.
2 K. Phelps, A. Morris and P. A. Beal, ACS Chem. Biol., 2012, 7, 100.
3 S. Fleige and M. W. Pfaffl, Mol. Aspects Med., 2006, 27, 126.
4 S. C. Tan and B. C. Yiap, J. Biomed. Biotechnol., 2009, 1.
5 P. Chomczynski and S. Sacchi, Anal. Biochem., 1987, 156; P. Chomczynski
and S. Sacchi, Nat. Protoc., 2006, 1, 581.
6 R. Martins, C. J. Maia, J. A. Queiroz and F. Sousa, J. Sep. Sci., 2012, 35, 3217.
7 D. G. Knorre, I. M. Pustoshilova, I. M. Teplova and G. G. Shamovsky,
Biokhimiya, 1965, 30, 1218.
8 S. Y. Ovodov and Y. B. Alakhov, FEBS Lett., 1990, 270, 111.
9 E. J. Merino, K. A. Wilkinson, J. L. Coughlan and K. M. Weeks, J. Am.
Chem. Soc., 2005, 127, 4223; K. A. Wilkinson, E. J. Merino and K. M.
Weeks, Nat. Protoc., 2006, 1, 1610.
10 J. R. Ross, J. A. S. Laks, D. L. C. Wang and J. W. Sowell Sr., Synthesis,
1985, 796; M. R. Player, J. W. Sowell Sr., G. R. Williams and G. T.
Cowley, J. Heterocycl. Chem., 1994, 31, 209; M. R. Player and J. W.
Sowell Sr., Bioorg. Med. Chem. Lett., 1994, 4, 949; M. R. Player and
J. W. Sowell Sr., J. Heterocycl. Chem., 1995, 32, 1537.
In summary, an isatoic anhydride derivative has been designed
for the specific 20-OH acylation of RNAs and has been successfully
used in RNA selective capture from a mixture of biological RNA
and DNA. We have demonstrated that this method can be used for
extraction, purification and amplification of RNAs. The combi-
nation of the chemical selectivity of isatoic anhydrides toward RNA
with their adequate functionalization could find other applications
where selective RNA chemical labeling is needed.11
This work was supported by the French public agency OSEO
(Advanced Diagnostics for New Therapeutic Approaches, a
French government-funded program dedicated to personalized
´
´
medicine), bioMerieux Laboratories and the ‘‘Conseil Regional
de Basse-Normandie’’.
11 A. Burr, A. Laayoun and A. Laurent, WO 2012076794, 2012.
This journal is ©The Royal Society of Chemistry 2014
Chem. Commun., 2014, 50, 5748--5751 | 5751