DNA display for drug discovery

Article abstract of DOI:10.1039/c3ra42440e

A novel DNA display strategy, based on a new puromycin modifier, has been developed. The 5′-end puromycin-tethered oligonucleotide was synthesized to hybridize with mRNA, so that it could attack the nascent polypeptides during in vitro translation. The DNA-peptide fusion molecule can tolerate more harsh and stringent selection conditions, therefore, this DNA display may become a useful tool for in vitro display technologies for the selection of peptide drug candidates. The Royal Society of Chemistry 2013.

Full text of DOI:10.1039/c3ra42440e

RSC Advances
View Article Online
View Journal | View Issue
COMMUNICATION
DNA display for drug discovery3
Haodong Chen,abc Gangyi Chen,a Feng Du,a Qingquan Fu,a Yun Zhaob
and Zhuo Tang*a
Cite this: RSC Advances, 2013, 3, 16251
Received 16th May 2013,
Accepted 15th July 2013
DOI: 10.1039/c3ra42440e
www.rsc.org/advances
A novel DNA display strategy, based on a new puromycin
modifier, has been developed. The 59-end puromycin-tethered
oligonucleotide was synthesized to hybridize with mRNA, so that
it could attack the nascent polypeptides during in vitro transla-
tion. The DNA–peptide fusion molecule can tolerate more harsh
and stringent selection conditions, therefore, this DNA display
may become a useful tool for in vitro display technologies for the
selection of peptide drug candidates.
109 independent molecules. Ribosome display is an in vitro
system, which has the potential to screen peptide libraries of three
to six orders of magnitude more than the in vivo display.10 Despite,
the ribosome–mRNA–peptide complex is extremely labile in cell-
free conditions and it is difficult to preserve its integrity during the
selection process. The complete in vitro selection strategy relies on
the covalent linkage between mRNA and peptides, known as
mRNA display or in vitro virus, and was proposed by Szostak and
Nemoto independently in 1997.11,12 Puromycin, which has a
similar structure to the aminoacyl end of tRNA, was tethered on
the 39-end of mRNA to capture the nascent peptides after
translation by the ribosome. However, the use of RNA as the
library-encoding nucleic acid is less convenient compared to the
use of DNA due to its susceptibility towards RNase degradation,
which limits the application of mRNA display in drug discovery.
Recently, covalent display technologies have been developed to
allow direct binding between peptides and encoding DNA
sequences without the help of any enzymes.13,14 Therefore, a
branched DNA primer tethered with puromycin can be applied to
accept the nascent peptide during in vitro translation, and the
code of the linked peptide can be obtained by reverse transcription
of this primer on an mRNA template. The branched DNA primer
with the puromycin modifier can only be synthesized in a
professional laboratory, and is not available on a commercial basis
for ordinary use.
Herein, we report a novel DNA display strategy based on a new
puromycin modifier, that can be attached to the 59-end of a DNA
primer (Fig. 1). As this primer is complimentary to the 39-end
sequence of mRNA, it can hybridize with mRNA to form a duplex
tail. During the in vitro translation of mRNA, when the ribosome
reaches the duplex domain, puromycin can enter the ribosomal
A-site and form a stable amide linkage with the nascent
polypeptide. Afterwards, the 39-end of the DNA primer can be re-
transcribed to obtain the DNA–peptide complex (Fig. 1). Such a
system has the following features: (1) the displayed polypeptides
and the nucleic acids encoding them are covalently linked. (2)
Unlike RNA, which is subjected to RNases, the more robust and
stable DNA is employed to encode the peptide. Therefore, the
DNA–peptide fusion molecules can tolerate more harsh and
A major issue in pharmaceutical science is the identification of
small organic molecules capable of specific binding to target
proteins of interest. Conventional technologies for drug discovery
are based on screening individual molecules against enzymatic
activity of the target protein or by displacement of labeled ligands.
High-throughput screening allows a researcher to quickly conduct
millions of chemical, genetic or pharmacological tests.1 However,
the success of screening depends on the amount and structural
diversity of the libraries of compounds. The use of ‘‘virtual
screening’’ based on the analysis of structural information may
reduce the number of compounds that need to be evaluated.2,3
Alternatively, a fragment-based approach may be used for the
identification of multidentate ligands with suitable binding
affinities and pharmacological properties.4–6 However, the display
of ligands such as peptides and proteins on the surface of
bacteriophage (Phage display) has revolutionized the field of
ligand isolation from a large library.7,8 Billions of polypeptides,
physically linked to their encoding DNA and target proteins can be
immobilized on a solid support. The enrichment of ligands from
these libraries is achieved by the co-selection of target-binding
peptides. The transfection efficiency is one of the main limitations
of phage display and other in vivo display technologies such as
plasmid display,9 which leads the library diversity to approximately
aNatural Products Research Center, Chengdu Institute of Biology, Chinese Academy
of Sciences, Chengdu 610041, P. R. China. E-mail: tangzhuo@cib.ac.cn; Fax: +86 28
85243250; Tel: +86 28 85243250
bCollege of Life Sciences, Sichuan University, Chengdu 610064, P. R. China
cUniversity of Chinese Academy of Sciences, Beijing 100049, P. R. China
Electronic supplementary information (ESI) available: details of the experimental
3
section and NMR spectra. See DOI: 10.1039/c3ra42440e
This journal is ß The Royal Society of Chemistry 2013
RSC Adv., 2013, 3, 16251–16254 | 16251

View Article Online
Communication
RSC Advances
Fig. 1 In vitro selection of peptides. First, the puromycin–oligonucleotide is annealed to the DNA near the 39-end, then the hybridization mixture is added into the in vitro
translation system. At the end of translation, the ribosome stops when it comes to the junction of single stranded RNA and the duplex DNA region. Here, puromycin enters
the A site of the ribosome to capture the nascent peptide. On reverse transcription, the cDNA–peptide fusion is formed and affinity selection is executed. The target cDNA–
peptide fusion is amplified by PCR after enrichment and the second round of selection is carried out.
stringent selection conditions. (3) A library with more than 1012
DNA–peptide molecules can be constructed for drug discovery.
In this study, first we have synthesized the puromycin
phosphoramidite that could be coupled to the 59-end of
oligonucleotides directly by standard DNA synthetic protocols,
which is different from the way developed by Szostak and other
groups in which puromycin was tethered at the 39-end of
oligonucleotides.11,15,16 As the puromycin was labeled at the
59-end, the modified DNA primer could be extended by re-
transcription from the 39-end. To mimic the aminoacyl end of
tRNA, puromycin was linked through a phosphate ester bond
from the 59-hydroxyl group.
(Fmoc) to obtain P1. As the Fmoc group can easily be deprotected
at high temperature, all the following reactions were carried out
below 45 uC. In the next step, the bulky dimethoxytrityl (DMT)
group could only be incorporated to the 59-position of the sugar
moiety due to steric hindrance and thus provided exclusively the
product P2. On the other hand, acetic anhydride was used to
protect the 29-hydroxyl group of the ribose sugar to get P3 due to
its high reactivity and the small size of the acetyl group. After the
DMT group was removed by mild protic acids, the 59-hydroxyl
moiety on P4 was the only reactive nucleophile capable of
participating in the subsequent derivatization reaction. To obtain
the final phosphoramidite product, pyridine was used as solvent
with a catalytic amount of DIPEA (N,N-diisopropylethylamine).
Finally, the 59-phosphoramidite puromycin P5 was synthesized
after five simple steps with the yield of 66%.
Therefore, synthetic work was carried out to obtain 59-phos-
phoramidite puromycin P5 (Scheme 1). We initiated our research
by protecting the amino group with fluorenylmethoxycarbonyl
With the puromycin phosphoramidite in hand, we have carried
out the synthesis of the modified primer on the DNA synthesizer
(Fig. S1, ESI ). Before coupling the final puromycin modifier, one
3
PEG modifier (Spacer-18) and two reverse deoxy-cytidines (dC-59–
CE phosphoramidite) were added in between. The puromycin
modified DNA primer was purified by reverse phase HPLC and
verified with mass spectrometry (Fig. S2 and S3, ESI ). The full
3
primer (Pu16-L1) was constructed by the ligation of the puromycin
modified DNA primer and the sequence that is complimentary to
the 39-end sequence of mRNA (Fig. 2A). Further investigations by
using the 59-end puromycin-tethered oligonucleotide as a captured
primer for the in vitro selection of peptides were carried out. After
PCR (Polymerase Chain Reaction) and in vitro transcription, a full-
length mRNA including a RBS (Ribosome Binding Site) and a ORF
(Open Reading Frame) was used as the template for peptide
expression. The fusion between peptides and Pu16-L1 was formed
Scheme 1 Synthesis of the puromycin modifier. Reagents and conditions: a.
Fmoc–OSu, Et3N–CH3CN; b. DMTrCl, Et3N–pyridine; c. (Ac)2O, DMAP–pyridine; d.
Cl2CHCOOH–CH2Cl2; e. Phosphorus reagent, DIPEA–pyridine.
16252 | RSC Adv., 2013, 3, 16251–16254
This journal is ß The Royal Society of Chemistry 2013

View Article Online
RSC Advances
Communication
Fig. 2 A) From top to bottom: hybridization of the DNA primer (16-L1) without puromycin with mRNA–DNA (16aa-RNA-Tail); peptide displayed on the puromycin modified
primer (Pu16-L1); mRNA display based on Szostak’s strategy. The asterisk represents 32P. B) In vitro translation and proteinase K degradation. Lane 1: in vitro translation with
primer 16-L1; lane 2: in vitro translation with primer Pu16-L1; lane 3: mRNA display with RNA-30P; lanes 4–5: the same samples from lanes 2 and 3 were treated with
proteinase K; lanes 6–8: markers of 16-L1, Pu16-L1 and RNA-30P. C) In vitro translation with different lengths of the mRNA template (encoding peptide with 7 and 16 amino
acids). D) Lane 1: in vitro translation without any mRNA template; lane 2: in vitro translation containing 16aa-RNA-Tail; lane 3: in vitro translation containing 7aa-RNA-Tail;
lane 4: marker of Pu16-L1.
initially (Fig. S4, ESI ), but this result was not consistent with
under optimized conditions (Lane 2, Fig. 2B). This result is
comparable to the yield of mRNA display (Lane 3, Fig. 2B) based
on Szostak’s strategy (RNA-30P, Fig. 2A). However, no DNA–
peptide fusion was detected under the same conditions by using
the similar primer 16-L1 that was devoid of puromycin on the
59-end compared to primer Pu16-L1. This result indicates that the
puromycin is indispensable in our DNA display strategy. To
confirm that the higher molecular weight product (upper band in
Lane 2, Fig. 2B) was the DNA–peptide fusion, proteinase K was
applied to digest the peptides in the fusion product. As expected,
the products of the DNA display and mRNA display disappeared
upon treatment with proteinase K (Lane 4 and 5, Fig. 2B).
Furthermore, when we used different lengths of mRNA for
translation (Fig. 2C), different bands of DNA–peptide products
formed with migration rates according to their corresponding
molecular size (lanes 2 and 3, Fig. 2D). All these results indicate
that the peptides can be captured efficiently by the 59-end
puromycin-tethered oligonucleotides in our novel system.
3
following repeated experiments. The situation did not improve
even when a stop codon was inserted between the ORF and the
duplex domain (data now shown). After excluding the possibility
that the DNA primer was not long enough to get into the
ribosome, we suspected that the ribsome was not completely
stopped by the RNA and DNA hybridized region when it reached
the junction of single stranded RNA and double stranded DNA
because of ribosome helicase activity. Therefore, the 39-end region
of the RNA tail was replaced by a DNA sequence (Fig. 2A), and
more than 10% of a DNA–peptide fusion product was formed.
Afterwards, it was proved that only a DNA sequence can reduce the
helicase activity of the ribosome to support the 59-tethered
puromycin on the primer Pu16-L1 to attack nascent polypeptides.
The 59-end puromycin-tethered with a short oligonucleotide (30
nt) was also designed for translation. The autoradiography results
show that the tether was still long enough for puromycin to
capture resultant peptides in our system (data not shown).
Furthermore, the fraction of DNA–peptide fusion increased
significantly to 20% when potassium chloride and sodium
chloride17 were added into the in vitro translation reaction mixture
To encode nascent polypeptides, the DNA primer must be re-
transcribed on the mRNA template after in vitro translation. Three
different DNA primers (perfect matched primer L1, full primer
This journal is ß The Royal Society of Chemistry 2013
RSC Adv., 2013, 3, 16251–16254 | 16253

View Article Online
Communication
RSC Advances
systems. Moreover, the DNA sequences encoding displayed
peptides in our strategy are hybridized with the corresponding
mRNAs to form duplex molecules that avoid the undesired ssDNA
aptamer in the drug selection process. Due to its feasible
application, together with the highly stable nature of the
protein–DNA complexes, our procedure proves to be a useful tool
for in vitro display technology for the selection of peptide drug
candidates. Follow-up screening experiments of specific target
proteins by using this technology are currently under way in our
laboratory.
Acknowledgements
This study was supported by the Chinese Academy of Science
(Hundreds of Talents Program) and the National Sciences
Foundation of China (Grant No. 21172215 and 21102139), the
Innovation Program of the Chinese Academy of Sciences
(Grant No. KSCX2-EW-J-22) and a grant (GZ632) from the Sino-
German Center for Research Promotion.
Notes and references
1 L. M. Mayr and D. Bojanic, Curr. Opin. Pharmacol., 2009, 9,
580–588.
2 D. B. Kitchen, H. Decornez, J. R. Furr and J. Bajorath, Nat. Rev.
Drug Discovery, 2004, 3, 935–949.
3 J. Bajorath, Nat. Rev. Drug Discovery, 2002, 1, 882–894.
4 P. J. Hajduk and J. Greer, Nat. Rev. Drug Discovery, 2007, 6,
211–219.
5 S. L. Schreiber, Science, 2000, 287, 1964–1969.
6 D. C. Rees, M. Congreve, C. W. Murray and R. Carr, Nat. Rev.
Drug Discovery, 2004, 3, 660–672.
7 G. P. Smith, Science, 1985, 228, 1315–1317.
8 J. McCafferty, A. D. Griffiths, G. Winter and D. J. Chiswell,
Nature, 1990, 348, 552–554.
Fig. 3 A) Reverse transcription on mRNA templates with different primers. From top
to bottom: perfect matched primer L2, full primer Pu16-L1 and full primer Pu16-L1
with peptide. B) lane 1: reverse transcription with perfect matched DNA primer L2;
lane 2: reverse transcription with primer Pu16-L1; lane 3: reverse transcription with
peptide tethered primer Pu16-L1; lane 4: reverse transcription with peptide
tethered primer Pu16-L1, but without mRNA template; lanes 5–7: markers of L2,
Pu16-L1 and Pu16-L1 with a peptide of 16 amino acids.
9 M. G. Cull, J. F. Miller and P. J. Schatz, Proc. Natl. Acad. Sci. U. S.
A., 1992, 89, 1865–1869.
10 L. C. Mattheakis, R. R. Bhatt and W. J. Dower, Proc. Natl. Acad.
Sci. U. S. A., 1994, 91, 9022–9026.
11 R. W. Roberts and J. W. Szostak, Proc. Natl. Acad. Sci. U. S. A.,
1997, 94, 12297–12302.
12 N. Nemoto, E. Miyamoto-Sato, Y. Husimi and H. Yanagawa,
FEBS Lett., 1997, 414, 405–408.
13 I. Tabuchi, S. Soramoto, N. Nemoto and Y. Husimi, FEBS Lett.,
2001, 508, 309–312.
14 M. Kurz, K. Gu, A. Al-Gawari and P. A. Lohse, ChemBioChem,
2001, 2, 666–672.
15 S. Ikeda, I. Saito and H. Sugiyama, Tetrahedron Lett., 1998, 39,
5975–5978.
16 K. Okuda, A. C. Seila and S. A. Strobel, Tetrahedron, 2004, 60,
12101–12112.
Pu16-L1, and full primer with peptide, Fig. 3B) were used in the
reverse transcription experiment (Fig. 3). As illustrated in the
autoradiographic results, all those primers have elongated well
(lanes 1, 2 and 3, Fig. 3), which implies that the extra section,
either the DNA tail or the tethered peptide, do not affect the re-
transcription process. Based on our DNA display method, a
peptide library containing as many as 1012–1013 unique sequences
could be constructed in one 1.5 ml Eppendorf tube in one day.
In summary, we have developed a new DNA display strategy
based on a puromycin modifer that can be tethered on the 59-end
of a DNA primer by using standard DNA synthetic protocols. The
DNA primer can hybridize with the mRNA to attack the nascent
polypeptide to obtain a DNA–peptide fusion during the translation
process. Due to the covalently bound DNA–peptide sequence, the
fusion molecule is rather robust compared to an mRNA–peptide
molecule, and may tolerate harsher in vitro drug selection
processes than other display strategies that rely on biological
17 R. Liu, J. E. Barrick, J. W. Szostak and R. W. Roberts, Methods
Enzymol., 2000, 318, 268–293.
16254 | RSC Adv., 2013, 3, 16251–16254
This journal is ß The Royal Society of Chemistry 2013