A DNA-templated synthesis of encoded small molecules by DNA self-assembly

Article abstract of DOI:10.1039/c4cc03380a

We report a novel method for the synthesis of DNA-encoded libraries without the need for discrete DNA template. Reactant DNAs self-assemble to enable chemical reactions and photo-cleavage transfers the product to the DNA terminus, making it suitable for the subsequent affinity-based selection and hit deconvolution. This journal is the Partner Organisations 2014.

Full text of DOI:10.1039/c4cc03380a

ChemComm
View Article Online
View Journal | View Issue
COMMUNICATION
A DNA-templated synthesis of encoded small
molecules by DNA self-assembly†
Cite this: Chem. Commun., 2014,
50, 10997
Cheng Cao,a Peng Zhao,a Ze Li,a Zitian Chen,c Yanyi Huang,c Yu Bai*a and
Xiaoyu Li*ab
Received 6th May 2014,
Accepted 28th July 2014
DOI: 10.1039/c4cc03380a
www.rsc.org/chemcomm
We report a novel method for the synthesis of DNA-encoded libraries ESAC by Neri and co-workers, which is specific for fragment-based
without the need for discrete DNA template. Reactant DNAs self- libraries,13 and the YoctoReactor by Vipergen, which confines DTS
assemble to enable chemical reactions and photo-cleavage transfers within the tight spaces of multi-way DNA junctions.20 Recently we
the product to the DNA terminus, making it suitable for the subsequent reported a ‘‘universal template’’ strategy capable of synthesizing
affinity-based selection and hit deconvolution.
DELs with just a single template regardless the size of the
library.27 Here we report a new library synthesis method that does
Originally in 1992, Brenner and Lerner proposed a visionary concept not require any discrete DNA template.
of using DNA to record chemical reactions in combinatorial
Our approach is shown in Fig. 1. We use a simple ‘‘T’’
libraries.1 During the past two decades, many strategies on architecture to assemble reactant DNAs to proximity but also with
DNA-encoded library (DEL) synthesis and selection have been flexibility to support DTS with various building blocks.20,28,29 After
developed.2–13 Today DELs can be prepared with thousands of chemical reactions, a ligase connects DNA strands so that the DTS
millions of compounds,3,14 and library selection and hit decoding product is encoded by the original reagent DNAs. Typically, DTS
can be feasibly accomplished14–19 to discover novel binders against synthesizes compounds at the DNA terminus, presumably more
biological targets.8,9 Pharmaceutical companies have also employed suitable for selections against protein targets (i.e. less interference
DELs in drug discovery programs.14,20–23
from the DNA tags). In order to achieve this, we incorporated
DNA-templated synthesis (DTS), developed by Liu and photo-cleavable o-nitrobenzyl linkers in R2 and R3 (Fig. 1 and
co-workers,8,24–26 is an important DEL synthesis strategy. Based Fig. S1, ESI†),30 while R1 connects with DNA via a non-cleavable
on sequence complementarity, DTS uses many DNA templates amide linkage; after chemical reactions and enzymatic ligation,
of different sequences to direct chemical reactions with multiple without the need for additional cleavage reagents, irradiation
sets of ‘‘reagent DNAs’’ carrying library building blocks. A DEL can readily cleaves both o-nitrobenzyl linkers and the final DTS product
be prepared using a large pool of templates through a multi-step is obtained at the 50-terminus of the encoding DNA.
DTS. However, this ‘‘one-template, one-compound’’ method
We initiated our study with an N-acylthiazolidine synthesis,
requires a unique DNA template for each library compound, which can be assembled with an aldehyde, an aminothiol, and
resulting in challenges and difficulties in sequence design and
template preparation, especially for large libraries. Previously, two
DEL strategies that do not need a template pool were reported:
a Key Laboratory of Bioorganic Chemistry and Molecular Engineering of the Ministry
of Education, Beijing National Laboratory of Molecular Sciences (BNLMS),
College of Chemistry and Molecular Engineering, Peking University, Beijing,
China 100871. E-mail: xiaoyuli@pku.edu.cn, yu.bai@pku.edu.cn
b Key Laboratory of Chemical Genomics, School of Chemical Biology and
Biotechnology, Peking University Shenzhen Graduate School, Shenzhen,
China 518055
c Biodynamic Optical Imaging Center (BIOPIC) and College of Engineering,
Fig. 1 Scheme for DEL synthesis without DNA template. A typical 3-step
synthesis with 3 sets of reagent DNAs is shown. Red dots indicate photo-
cleavable linkers (structure is shown on the lower left). Red arrows indicate
enzyme ligation sites.
Peking University, Beijing, China 100871
† Electronic supplementary information (ESI) available: Materials and general
methods, experimental details, library selection and sequencing methods and
enrichment fold calculations. See DOI: 10.1039/c4cc03380a
This journal is ©The Royal Society of Chemistry 2014
Chem. Commun., 2014, 50, 10997--10999 | 10997

View Article Online
Communication
ChemComm
Fig. 3 (a) Scheme for the dihydropyran decoration. GP: gel purification; EP:
ethanol precipitation. (b) Denaturing PAGE analysis. Lane 1: DNA standards;
lane 2: reaction of RD4 and RD5 with Cu(OAc)2/sodium ascorbate; lane 3:
reaction of RD4, RD5, and RD6 with Cu(OAc)2/sodium ascorbate and
NaBH3CN; lane 4: lane 3 after irradiation; lane 5–9: same as lane 3 but
with mismatched RD4, with mismatched RD6, with a RD4 without the
aldehyde appendage (Fig. S6, ESI†), no NaBH3CN, with a RD4 without the
scaffold (Fig. S6, ESI†). Reagent DNAs: 1 mM each; Click reaction: 500 mM
Cu(OAc)2, 500 mM sodium ascorbate, 25 1C, 3 hours; reductive amination:
50 mM NaBH3CN, 25 1C, 4 hours; irradiation: 365 nm, 0 1C, 15 min.
(c) MALDI-MS characterization of the final product after AflII digestion.
Fig. 2 (a) Synthesis scheme of an N-acylthiazolidine. GP: gel purification; EP:
ethanol precipitation. PNK: T4 polynucleotide kinase. (b) Reactions analyzed by
denaturing PAGE. Lane 1: DNA standards; lane 2: incubation of RD2 and RD3;
lane 3: reaction of RD1, RD2, RD3 and DMT-MM; lane 4: lane 3 after irradiation;
lane 5–9: same as lane 3 but with a control RD2 (no aminothiol), with
mismatched RD3, with mismatched RD1, without DMT-MM, or with Ac2O
capping before amidation. RD1-3: 1 mM each; amidation: 50 mM DMT-MM,
25 1C, 12 hours; irradiation: 365 nm, 0 1C, 15 min. (c) MALDI-MS characteriza-
tion of the final product after AflII digestion. See ESI† for details.
a carboxylic acid.29,31 We first prepared reagent DNAs (RD1, 41-nt;
RD2, 26-nt; RD3, 47-nt; Fig. 2a and Fig. S3, ESI†); RD1/RD2 contains generated the product in the presence of RD6 and NaBH3CN
an AflII digestion site so that the majority of DNA can be removed (114-nt band; lane 3). As expected, this band can be cleaved under
afterwards to facilitate MS characterization. These reagent DNAs irradiation (lane 4). Negative control reactions again indicate that the
were mixed and subjected to a typical DMT-MM-mediated acylation product formation requires all three reagent DNAs (lanes 5–9). After
condition, along with a series of control experiments to validate gel purification, enzymatic ligation and AflII digestion (Fig. S5, ESI†),
the reaction specificity. As shown in Fig. 2b, the incubation of RD2 we characterized the product by MALDI (Fig. 3c). In addition, since
and RD3 gave a lower motility band, which presumably is the the final product is encoded by the ligated reagent DNAs, which
unacylated thiazolidine (lane 2).29 Upon the addition of RD1 and retains the duplex structure and has the ‘‘scar’’ from photo-cleavage,
DMT-MM, a band matching the expected length of the N-acylthiazol- we also validated that the encoding DNA is still compatible with PCR
idine product (114-nt) was observed (lane 3), which can be cleaved amplification (Fig. S4 and S5, ESI†).
after irradiation (lane 4). Importantly, negative control reactions
Next, we prepared a model library (Fig. 4a). In this library,
did not generate the same band (lanes 5–9), proving that the RD7-1 has a 3-base codon (‘‘CTCC’’; T is the small molecule
114-nt product formation requires all three reagent DNAs with conjugation site) encoding a biotinylated L-propargylglycine.
complementary sequences. The 114-nt band in lane 3 was gel- RD7-2 contains mixed sequences (‘‘DTDD’’, D = A, G, or T); for
purified and then subjected to PNK/T4 DNA ligase-mediated simplicity, they only encode one amino acid (5-hexynoic acid). We
ligation to afford the N-acylthiazolidine product (Fig. S4, ESI†), mixed RD7-1 with excess of RD7-2 (1 : 100) as the ‘‘RD7’’. RD7 was
which was characterized by MALDI-MS after AflII digestion combined with RD4/RD5 and then subjected to the same reaction
(Fig. 2c).
conditions as in Fig. 3. After the synthesis, the library was selected
Next, we performed a DNA-templated decoration of a dihydro- against immobilized streptavidin,27 which should capture library
pyran scaffold using this method (Fig. 3). This scaffold is based on members containing the biotinylated L-propargylglycine. Selected
the core structure of a class of approved anti-influenza drugs such as compounds were eluted, PCR-amplified and sequenced (Fig. 4b).
oseltamivir, zanamivir, and laninamivir.32 We attached four chemi- The results show that the biotin-encoding ‘‘CTCC’’ was distinctly
cally orthogonal appendages on the scaffold reasoning that either of enriched after selection. This result is corroborated by the
these appendages can be conjugated to DNA while the other three sequencing results of the opposite strand (Fig. S8, ESI†).
could be diversified with library building blocks. We prepared the
Furthermore, we prepared another library containing 18
scaffold following a report by Smith, Itzstein, and co-workers (Fig. S2, N-acylthiazolidines (Fig. 5), in which a biotin-containing building
ESI†)33 and then conjugated it to DNA via the carboxylic acid handle block is encoded by ‘‘TAG’’ at the RD1 position (RD1-3). The ratio
to afford RD4. RD5 and RD6 were prepared bearing a photo- of RD1-3/(RD1-1 + RD1-2) is 1 : 400 in the library synthesis. This
cleavable alkyne and amine, respectively. These reagent DNAs were library was prepared using the same procedure as the Fig. 2 and
assembled and subjected to the typical Click reaction and reductive then was also selected against streptavidin. Selection results were
amination conditions. As shown in Fig. 3b, RD4’s azide reacted decoded by high throughput sequencing (Fig. 5b) and again
with RD5’s alkyne, catalyzed by copper (lane 2), and then further sequences containing the biotin-encoding ‘‘TAG’’ have been
10998 | Chem. Commun., 2014, 50, 10997--10999
This journal is ©The Royal Society of Chemistry 2014

View Article Online
ChemComm
Communication
libraries and accelerate the interrogation of biological targets in
drug discovery with DELs.
This work was supported by the Ministry of Science and
Technology Basic Research Program (2011CB809100), NSFC
(21272016, 91013003, 21175008, and J1030413), the Beijing
Nova Program (2010B002), and the Doctoral Fund of Ministry
of Education of China (20120001110083).
Notes and references
1 S. Brenner and R. A. Lerner, Proc. Natl. Acad. Sci. U. S. A., 1992, 89, 5381.
2 J. Nielsen, S. Brenner and K. D. Janda, J. Am. Chem. Soc., 1993,
115, 9812.
3 M. C. Needels, D. G. Jones, E. H. Tate, G. L. Heinkel, L. M.
Kochersperger, W. J. Dower, R. W. Barrett and M. A. Gallop, Proc.
Natl. Acad. Sci. U. S. A., 1993, 90, 10700.
4 D. R. Halpin and P. B. Harbury, PLoS Biol., 2004, 2, 1015–1021.
5 J. Scheuermann, C. E. Dumelin, S. Melkko and D. Neri, J. Biotechnol.,
2006, 126, 568.
6 S. K. Silverman, Angew. Chem., Int. Ed., 2010, 49, 7180.
7 M. A. Clark, Curr. Opin. Chem. Biol., 2010, 14, 396.
8 R. E. Kleiner, C. E. Dumelin and D. R. Liu, Chem. Soc. Rev., 2011, 40, 5707.
9 L. Mannocci, M. Leimbacher, M. Wichert, J. Scheuermann and
D. Neri, Chem. Commun., 2011, 47, 12747.
Fig. 4 (a) Model library synthesis scheme. RD7 contains mixed sequences.
Reaction conditions are the same as in Fig. 3. (b) Sequencing results of the
model library before and after the selection against immobilized streptavidin.
Encoding sites before and after the selection were compared as marked in the
figure. See ESI† for details.
10 J. P. Daguer, C. Ciobanu, S. Alvarez, S. Barluenga and N. Winssinger,
Chem. Sci., 2011, 2, 625.
11 S. J. Wrenn, R. M. Weisinger, D. R. Halpin and P. B. Harbury, J. Am.
Chem. Soc., 2007, 129, 13137.
12 N. Winssinger, K. Gorska, M. Ciobanu, J. P. Daguer and S. Barluenga,
Methods Mol. Biol., 2014, 1050, 95.
13 S. Melkko, J. Scheuermann, C. E. Dumelin and D. Neri, Nat.
Biotechnol., 2004, 22, 568.
14 M. A. Clark, et al., Nat. Chem. Biol., 2009, 5, 647.
15 L. Mannocci, Y. X. Zhang, J. Scheuermann, M. Leimbacher, G. De
Bellis, E. Rizzi, C. Dumelin, S. Melkko and D. Neri, Proc. Natl. Acad.
Sci. U. S. A., 2008, 105, 17670.
16 F. Buller, L. Mannocci, Y. Zhang, C. E. Dumelin, J. Scheuermann
and D. Neri, Bioorg. Med. Chem. Lett., 2008, 18, 5926.
17 F. Buller, Y. Zhang, J. Scheuermann, J. Schafer, P. Buhlmann and
D. Neri, Chem. Biol., 2009, 16, 1075.
18 R. E. Kleiner, C. E. Dumelin, G. C. Tiu, K. Sakurai and D. R. Liu,
J. Am. Chem. Soc., 2010, 132, 11779–11791.
19 F. Buller, M. Steiner, K. Frey, D. Mircsof, J. Scheuermann,
M. Kalisch, P. Buhlmann, C. T. Supuran and D. Neri, ACS Chem.
Biol., 2011, 6, 336.
20 M. H. Hansen, P. Blakskjaer, L. K. Petersen, T. H. Hansen, J. W. Hojfeldt,
K. V. Gothelf and N. J. V. Hansen, J. Am. Chem. Soc., 2009, 131, 1322.
21 J. S. Disch, G. Evindar, C. H. Chiu, C. A. Blum, H. Dai, L. Jin,
E. Schuman, K. E. Lind, S. L. Belyanskaya, J. Deng, F. Coppo,
L. Aquilani, T. L. Graybill, J. W. Cuozzo, S. Lavu, C. Mao,
G. P. Vlasuk and R. B. Perni, J. Med. Chem., 2013, 56, 3666.
22 R. K. Thalji, et al., Bioorg. Med. Chem. Lett., 2013, 23, 3584.
23 H. Encinas, et al., J. Med. Chem., 2014, 57, 1276.
24 Z. J. Gartner and D. R. Liu, J. Am. Chem. Soc., 2001, 123, 6961.
25 Z. J. Gartner, B. N. Tse, R. Grubina, J. B. Doyon, T. M. Snyder and
D. R. Liu, Science, 2004, 305, 1601.
Fig. 5 (a) Preparation and selection of an 18-member N-acylthiazolidine
library, which was prepared with the same procedure as in Fig. 2. The library
was selected against immobilized streptavidin. Selected compounds were
eluted, PCR-amplified and decoded by high throughput sequencing and the
result is shown in (b). Enrichment fold = (post-selection fraction)/(pre-selection
fraction) based on sequence counts before and after the selection. See the ESI†
for details.
26 B. N. Tse, T. M. Snyder, Y. H. Shen and D. R. Liu, J. Am. Chem. Soc.,
2008, 130, 15611.
27 Y. Li, P. Zhao, M. Zhang, X. Zhao and X. Li, J. Am. Chem. Soc., 2013,
135, 17727.
28 Z. J. Gartner, R. Grubina, C. T. Calderone and D. R. Liu, Angew.
Chem., Int. Ed., 2003, 42, 1370.
29 X. Y. Li, Z. J. Gartner, B. N. Tse and D. R. Liu, J. Am. Chem. Soc., 2004,
126, 5090.
significantly enriched (336-fold on average, see ESI† for details).
In addition, we have also characterized a portion of the library
compounds by mass spectrometry (ESI-MS; Fig. S9, ESI†).
Collectively, these results have demonstrated the applicability
of our method in the synthesis and selection DNA-encoded
libraries with structural and sequence diversities.
30 C. P. Holmes, J. Org. Chem., 1997, 62, 2370.
In conclusion, we have developed a novel DNA-templated 31 C. T. Calderone and D. R. Liu, Angew. Chem., Int. Ed., 2005, 44, 7383.
32 M. von Itzstein, Nat. Rev. Drug Discovery, 2007, 6, 967.
33 P. W. Smith, I. D. Starkey, P. D. Howes, S. L. Sollis, S. P. Keeling,
synthesis method capable of synthesizing DNA-encoded small
molecules without the need for discrete DNA template. This
P. C. Cheery, M. von Itzstein, W. Y. Wu and B. Jin, Eur. J. Med. Chem.,
method may further streamline the development of DNA-encoded
1996, 31, 143.
This journal is ©The Royal Society of Chemistry 2014
Chem. Commun., 2014, 50, 10997--10999 | 10999