In vitro evolution of a Friedel-Crafts deoxyribozyme

Article abstract of DOI:10.1039/c3ob40080h

We report the in vitro selection of a single-stranded 72-nucleotide DNA enzyme (deoxyribozyme) that catalyzes a Friedel-Crafts reaction between an indole and acyl imidazole in good yield and in aqueous solvent. Appreciable Friedel-Crafts product requires addition of copper nitrate and the deoxyribozyme. We observe deoxyribozyme-mediated bond formation for both in cis and in trans Friedel-Crafts reactions.

Full text of DOI:10.1039/c3ob40080h

Organic &
Biomolecular Chemistry
View Article Online
View Journal
COMMUNICATION
In vitro evolution of a Friedel–Crafts deoxyribozyme†
Utpal Mohan,a Ritwik Buraia and Brian R. McNaughton*a,b
Cite this: DOI: 10.1039/c3ob40080h
Received 15th January 2013,
Accepted 15th February 2013
DOI: 10.1039/c3ob40080h
www.rsc.org/obc
We report the in vitro selection of a single-stranded 72-nucleotide
DNA enzyme (deoxyribozyme) that catalyzes a Friedel–Crafts reac-
tion between an indole and acyl imidazole in good yield and in
aqueous solvent. Appreciable Friedel–Crafts product requires
addition of copper nitrate and the deoxyribozyme. We observe
deoxyribozyme-mediated bond formation for both in cis and
in trans Friedel–Crafts reactions.
Well-folded single-stranded RNAs and DNAs can achieve struc-
tural diversities and complexities that approach those observed
in globular proteins.1 Like proteins, single-stranded RNAs and
DNAs can have enzymatic activity. A relatively large number of
naturally occurring and unnatural RNA enzymes (ribozymes)
are known.2 In contrast, a much smaller number of DNA Fig. 1 Friedel–Crafts reactions via: (A) single-stranded deoxyribozyme-
enzymes (deoxyribozymes) have been reported.3–5 In compari-
son to ribozymes, deoxyribozymes are relatively stable and
mediated bond formation. A folded single-stranded DNA engages directly with
reactant(s) and stabilizes a reaction transition state, thereby facilitating bond
formation. (B) dsDNA chirality transfer catalysis. Double-stranded DNA, or
inexpensive to prepare, making these DNAs potentially useful
G-quadruplex DNA participates in the formation of a DNA – chelating metal
as catalysts. A modest number of deoxyribozymes have been
reagent – reactant complex, thereby facilitating transfer of DNA stereochemistry
found that catalyze bond-forming or bond-breaking reactions to the resulting product. M = metal.
involving nucleic acids, including ion-dependent RNA clea-
vage,6,7 DNA cleavage,8 DNA phosphorylation,9 DNA adenyla-
of, or combination of, general acid-, general base-, Lewis acid-,
or hydrophobic effect-dependent mechanisms (Fig. 1A).
tion,10 DNA deglycosylation,11 and thymine dimer
photoreversion12 have been reported. A smaller subset of deoxy-
ribozymes have been reported to facilitate intramolecular or
intermolecular reactions that expand beyond nucleic acid
chemistry.13
An alternative approach to facilitating organic reactions
using DNA is DNA-based chirality transfer catalysis.14,15 In this
approach, DNA alone doesn’t necessarily bind reactants, or
stabilize a transition state through the formation of a DNA-
reactant(s) complex. A hybrid catalyst is first generated by pre-
complexing double-stranded DNA (dsDNA) or G-quadruplex
Like a protein enzyme, a single-stranded DNA (ssDNA)
deoxyribozyme would be expected to fold into a precise three-
dimensional shape, directly engage with reactant(s), stabilize a
reaction transition state complex in a well-defined catalytic
pocket, and potentially participate in catalysis through the use
DNA with
a DNA-intercalating transition-metal reagent.
Together, these components permit catalysis (such as Lewis
acid-mediated catalysis), as well as the transfer of chirality
from the DNA double helix to the stereocenter(s) formed in the
reaction. DNA-based asymmetric catalysis has recently been
aDepartment of Chemistry, Colorado State University, 200 West Lake Street,
Fort Collins, Colorado, USA. E-mail: brian.mcnaughton@colostate.edu;
Fax: (+1) 970-491-1801; Tel: (+1) 970-491-7060
applied to
(Fig. 1B).16
a copper(II)-catalyzed Friedel–Crafts reaction
While DNA chirality transfer chemistry will undoubtedly
continue to be a useful method to impart stereocontrol, the
development of single-stranded deoxyribozyme catalysts that
directly participate in facilitating a chemical transformation is
bDepartment of Biochemistry and Molecular Biology, Colorado State University,
200 West Lake Street, Fort Collins, Colorado, USA
†Electronic supplementary information (ESI) available: Experimental details and
characterization of new compounds. See DOI: 10.1039/c3ob40080h
This journal is © The Royal Society of Chemistry 2013
Org. Biomol. Chem.

View Article Online
Communication
Organic & Biomolecular Chemistry
needed to expand the repertoire of deoxyribozyme catalysis,
and their study may lead to a deeper understanding of cataly-
sis mechanisms. Inspired by recent work using double-
stranded DNA16 or G-quadruplex DNA17 as a reagent to impart
stereocontrol for Friedel–Crafts reactions, and as part of an
effort to expand the synthetic scope of deoxyribozyme-
mediated synthesis, we set out to evolve a deoxyribozyme for
the Friedel–Crafts reaction between acyl imidazole 1 and
5-methoxy indole 2 (Fig. 1A).
Initially, we attempted to isolate deoxyribozymes using an
in vitro selection scheme based on a biotin pull-down. We envi-
saged that a single stranded DNA library equipped with a
5′-linked acyl imidazole could be incubated with biotinylated
indole 5, and DNAs that form a covalent bond with 5 could be
isolated via streptavidin-coated magnetic beads. DNAs cova-
lently bonded to biotin could then be amplified by polymerase
chain reaction (PCR). Toward this end, we synthesized bromo
amide acyl imidazole 3, and conjugated this compound to a
5′-thiol DNA primer (thiol-5′-GGAGCTCGCTTGTCG-3′) in quanti-
tative yield to generate 2, which was purified by PAGE gel elec-
trophoresis and ethanol precipitation (Fig. 2). The conjugation
product 4 was confirmed by matrix-assisted laser desorption/
ionization mass spectrometry (ESI†). A single stranded DNA
library was generated by polymerase chain reaction (PCR) on a
5′-CCTCGAGCGAACAGC-N40-AGCTGATCCTGTGATGG-3′ DNA
library template, using forward primer 4 and a reverse primer
Fig. 3 Biotinylated indole 5, which was initially used in streptavidin pull-down
selections, then used as a co-reactant during gel-shift-based selections.
fully remove bead-binding DNAs from the library, even when
negative selections to remove these DNAs were performed in
advance of positive selections for bond formation, and exten-
sive washing protocols. As a result, we were unable to selec-
tively enrich and amplify active DNA using a streptavidin
pull-down approach.
Since biotinylated indole 5 has a large molecular weight
(703 Da), we were able to employ an alternative selection strat-
egy to enrich active DNAs that is based on PAGE gel-shift of
active DNAs. Each selection round consisted of three inter-
acted steps: (1) incubating the 5′-linked acyl imidazole DNA
library with 5; (2) denaturing PAGE separation of active (higher
molecular weight) DNA strands; and (3) PCR amplification and
denaturing PAGE gel electrophoresis to amplify and isolate
acyl imidazole 5′-linked single-stranded DNA, respectively,
which is enriched with active species (Fig. 4A). In each round,
selection conditions consisted of 150 pmol (1.5 μM) DNA,
20 mM MOPS buffer (pH 6.5), 150 μM Cu(NO3)2, 400 nmol
(400 μM) biotinylated indole 5. Importantly, when we mixed
1.5 μM acyl imidazole 1, 400 μM 5-methoxy indole 2, and
150 μM Cu(NO3)2, but no DNA, in MOPS buffer for 24 hours at
25 °C, we did not observe any Friedel–Crafts product. There-
fore, in our selection, only DNAs that actively facilitate bond
formation between the DNA-linked acyl imidazole and biotin-
indole 5 are enriched. Before PAGE analysis, the mixture is
heated to 94 °C for 5 minutes, then analyzed following
containing
a
5′-(PEG)18 spacer and (AAC)4 terminus
((AAC)4(PEG)18-CCATCAGGATCAGCT-3′)). Since the reverse
primer contains a 5′-(AAC)4(PEG)18-terminus, the two DNA
strands of the PCR product have different molecular weights.
In a manner that is similar to previous reports,18 we were able
to separate the single-stranded DNA library containing the
5′-linked acyl imidazole reactant from its complementary strand
by denaturing PAGE gel electrophoresis.
We initially performed multiple rounds of selection for
DNA-catalyzed bond formation by incubating the 5′-linked acyl
imidazole DNA library with biotinylated indole 5 (Fig. 3), fol-
lowed by pull-down of biotin-linked DNA on streptavidin-
coated magnetic beads (Invitrogen) and PCR amplification of
enriched DNAs. However, we observed high levels of DNA
binding to the streptavidin-coated beads, and were unable to
Fig. 4 (A) Gel-shift selection strategy to evolve catalytically active DNAs.
(B) Gel-shift data from selections rounds 1–7. For each selection round, the ratio
of active : inactive (higher molecular weight : lower molecular weight) DNA is
provided in (C).
Fig. 2 Synthesis of bromoamide 3, and α/β-unsaturated ketone-linked DNA
primer 4.
Org. Biomol. Chem.
This journal is © The Royal Society of Chemistry 2013

View Article Online
Organic & Biomolecular Chemistry
Communication
denaturing PAGE electrophoresis. Taken together, the heat example, using the same conditions, but 10 mol% M14 DNA,
denaturation and denaturing PAGE electrophoresis likely dis- resulted in 18% yield of the Friedel–Crafts product (Table 1,
sociates tightly bound DNA – 5 complexes. As a result, higher entry 5). Performing the reaction in the absence of DNA, or in
molecular weight DNAs should be covalently bound to 5.
the absence of Cu(NO3)2, resulted in ≤10% yield of the
We performed three consecutive selection rounds with an Friedel–Crafts product (Table 1, entries 6 and 7, respectively),
incubation time of 24 h. This allowed us to enrich catalytic which is consistent with previous reports.16,17 Affinity con-
DNAs under relatively unstringent conditions (long reaction stants for the interaction between M14 DNA and reactants, as
time). These selections were followed by two selection rounds well as the molecular mechanisms of catalysis, are currently
with 6 hour incubation, and finally two rounds with a 2 hour unknown. However, the increase in product yield for the
incubation period. This allowed us to enrich those deoxyribo- in trans reaction, compared to the in cis reaction, is likely a result,
zymes with the most robust activities. Following the final selec- at least in part, of the increased concentration of reactants.
tion round, enriched DNAs were amplified by PCR, cloned into Since the selection is designed to enrich DNA that mediates a
a pUC plasmid, and transformed into E. coli. Single clones single bond forming reaction, and does not select for high cata-
were isolated and the enriched DNAs were sequenced using lyst turnover, it is unsurprising that relatively high amounts
standard methods. We screened 20 clones for in cis reactivity, of deoxyribozyme are needed to observe good product yield.
which was measured using a PAGE gel-shift assay that is iden- The precise sequence of M14 is required to achieve good reac-
tical to selection conditions. Interestingly, we observed very tion yield. An in trans reaction using a DNA sequence different
little sequence homology among the enriched DNAs we inde- from M14 does not generate appreciable levels of Friedel–
pendently screened.
Crafts product (Table 1, entry 8).
The most active DNA we isolated from our selection is
The enantioselectivity of M14-catalyzed reactions was not
referred to as M14 throughout. M14 is predicted to have a low determined experimentally. However, good stereoselectivity in
folding energy (−16.2 kcal mol−1), and contain an extended DNA-catalyzed13 and DNA-templated19 reactions has previously
centralized base-paired sequence, which is flanked by ordered been demonstrated.
regions that contain multiple stem-loops and higher ordered
In summary, we have used a gel-shift in vitro nucleic acid
structures (ESI†). As stated previously, running the Friedel– selection to identify a 72-nucleotide deoxyribozyme that cata-
Crafts reaction under selection conditions (25 °C, 150 μM lyzes a Friedel–Crafts reaction This deoxyribozyme functions
Cu(NO3)2, 1.5 μM acyl imidazole 1 and 400 μM biotin indole 5) well in the in cis reaction, wherein the acyl imidazole moiety is
did not yield appreciable product (Table 1, entry 1). However, tethered the 5′ end of DNA and the indole is liked to a biotin
the in cis reaction using acyl imidazole-linked M14 DNA gener- moiety at position 5. When the reaction is run in trans, at a
ated 32% product, as determined by gel-shift (Table 1, entry 2). higher concentration (to facilitate isolation and characteri-
Bond formation requires addition of Cu2+. Running the zation of product), the deoxyribozyme catalyzes formation of
reaction without Cu(NO3)2 did not generate appreciable levels the Friedel–Crafts product in 72% isolated yield. Deoxyribo-
of higher molecular weight DNA (Table 1, entry 3).
zyme activity is dependent upon addition of Cu2+; no appreci-
We next determined if clone M14 could facilitate a Friedel– able yield is observed in the absence of Cu(NO3)2. Since
Crafts reaction in trans, wherein neither reactant is tethered to appreciable yield of Friedel–Crafts product is dependent upon
the DNA. When we reacted 2 mM acyl imidazole 1, 10 mM M14 DNA, and appreciable levels of product is not observed
indole 2, 0.6 mM Cu(NO3)2, and 1 mM M14 DNA at 25 °C for when a random DNA sequence is used, the mechanism of
24 hours, we observed the formation of the expected Friedel– enzymatic action likely differs from DNA chirality transfer cata-
Crafts product in 72% yield (Table 1, entry 4). Lowering the lysis. Our ongoing efforts seek to understand the molecular
amount of M14 DNA dramatically lowered reaction yield. For mechanisms of M14 deoxyribozyme catalysis.
Table 1 In trans and in cis Friedel–Crafts reactions catalyzed by M14 DNA
Notes and references
Entry
In cis or in trans
Catalyst
Conversion (%)
1 K. Y. Chang and G. Varani, Nat. Struct. Biol., 1997, 4, 854–
858.
2 E. A. Doherty and J. A. Doudna, Annu. Rev. Biochem., 2000,
69, 597–615.
3 S. K. Silverman, Chem. Commun., 2008, 3467–3485.
4 Y. Li and R. R. Breaker, Curr. Opin. Struct. Biol., 1999, 9,
315–323.
1a
2
In cis
In cis
In cis
In trans
In trans
In trans
In trans
In trans
No
M14-Cu
M14
M14-Cu (50 mol%)
M14-Cu (10 mol%)
M14
0
32
0
72
18
10
7
3
4b
5
6
7
8
No
Random DNA
5
a Reaction conditions: MOPS buffer, pH = 6.5, 150 μM Cu(NO3)2,
1.5 μM acyl imidazole-linked DNA, 400 μM biotin indole 5, 25 °C, 24 h.
b Reaction conditions: MOPS buffer, pH = 6.5, 0.6 mM Cu(NO3)2, 2 mM
1, 10 mM 2, 1 mM M14 DNA, for entry 4, and 0.2 mM M14 DNA for
entry 5, 25 °C, 24 h. Cu = copper.
5 S. K. Silverman, Org. Biomol. Chem., 2004, 2, 2701–2706.
6 J. Li, W. Zheng, A. H. Kwon and Y. Lu, Nucleic Acids Res.,
2000, 28, 481–488.
7 S. K. Silverman, Nucleic Acids Res., 2005, 33, 6151–6163.
This journal is © The Royal Society of Chemistry 2013
Org. Biomol. Chem.

View Article Online
Communication
Organic & Biomolecular Chemistry
8 M. Chandra, A. Sachdeva and S. K. Silverman, Nat. Chem. 14 G. Roelfes, A. J. Boersma and B. L. Feringa, Chem.
Biol., 2009, 5, 718–720. Commun., 2006, 635–637.
9 Y. Li and R. R. Breaker, Proc. Natl. Acad. Sci. U. S. A., 1999, 15 G. Roelfes and B. L. Feringa, Angew. Chem., Int. Ed. Engl.,
96, 2746–2751. 2005, 44, 3230–3232.
10 Y. Li, Y. Liu and R. R. Breaker, Biochemistry, 2000, 39, 3106– 16 A. J. Boersma, B. L. Feringa and G. Roelfes, Angew. Chem.,
3114. Int. Ed. Engl., 2009, 48, 3346–3348.
11 T. L. Sheppard, P. Ordoukhanian and G. F. Joyce, Proc. Natl. 17 C. Wang, Y. Li, G. Jia, Y. Liu, S. Lu and C. Li, Chem.
Acad. Sci. U. S. A., 2000, 97, 7802–7807. Commun., 2012, 48, 6232–6234.
12 D. J. Chinnapen and D. Sen, Proc. Natl. Acad. Sci. U. S. A., 18 M. Chandra and S. K. Silverman, J. Am. Chem. Soc., 2008,
2004, 101, 65–69. 130, 2936–2937.
13 S. K. Silverman, Angew. Chem., Int. Ed. Engl., 2010, 49, 19 X. Y. Li and D. R. Liu, J. Am. Chem. Soc., 2003, 125, 10188–
7180–7201.
10189.
Org. Biomol. Chem.
This journal is © The Royal Society of Chemistry 2013