.
Angewandte
Communications
DOI: 10.1002/anie.201205201
Synthetic Methods
Structure-Selective Catalytic Alkylation of DNA and RNA**
Kiril Tishinov, Kristina Schmidt, Daniel Hꢀussinger, and Dennis G. Gillingham*
Deciphering the complex puzzle of nucleic acid (NA)
chemistry and biology has proven incredibly difficult[1] and
continues unabated after more than a century.[2] Chemists
have begun to recognize that the extraordinary informational
and structural properties that have made NAs so fundamental
in biology could be exploited in other contexts.[3] Simple,
selective chemical strategies to modify NAs would therefore
have a significant impact on studying their varied functions as
well as repurposing their special abilities, yet most extant
tools to target them are unselective or labor intensive.[4]
Although a number of reports exploring the potential of
organometallic catalysis in biological environments have
recently appeared,[5] such strategies have not yet been applied
towards selective NA alkylation. But untapped potential is
suggested by the ability to tune the reactivity of metal
complexes through judicious ligand selection and a number of
studies demonstrating the feasibility of metal catalysis with
DNA as a ligand.[6] We establish herein that rhodium-
catalyzed carbene transfer is a viable means of achieving
the selective alkylation of a variety of NAs.
Although there are a range of chemical methods to
modify NAs,[7] researchers typically resort to the effort of
building a new phosphoramidite for solid-phase synthesis
whenever an unnatural NA is required. In addition, access to
unnatural NAs beyond the size limit of solid-phase synthesis
is especially difficult.[7c] A handful of methods to target native
NAs have been described and employ a reactive molecule
linked to a guiding motif to direct specificity.[8] While
important in certain contexts, the complexity of these systems
precludes their use as general synthetic strategies. A promis-
ing enzymatic method has recently been developed for NA
modification which co-opts the natural S-adenosylmethionine
(SAM)/methyltransferase system by employing SAM ana-
logues bearing functional groups which can undergo subse-
quent chemoselective transformations.[9] In the same spirit as
enzymatic catalysis, organometallic approaches could bring
unrivalled levels of efficiency and selectivity to the chemistry
of NA modification. Indeed, as we show herein, despite the
myriad of possible reaction pathways of a rhodium carbenoid
with a NA, N-H insertion into exocyclic amines is preferred
(see equation for Table 1). We also find that the secondary
structure of the NA can be exploited to guide selectivity in the
alkylation. Taken together the recently developed enzymatic
method and the organometallic approach we describe should
offer researchers a considerable increase in flexibility for
tailored NA synthesis.
Recent reports on employing rhodium carbenoids,[10] and
rhodium/peptide conjugates[11] in alkylating various proteins,
even in cell lysates,[12] as well as structural studies of a number
of RhII/DNA complexes,[13] suggested dimeric rhodium com-
plexes as a promising candidate catalyst system for NA
alkylation. Indeed, in an initial proof-of-concept experiment
the simple tetradeoxynucleotide d(ATGC) was treated with
10 mol% [Rh2(OAc)4] and the diazo substrate Dz1 in
aqueous buffer (Scheme 1). Under these conditions the Dz1
substrate was completely consumed after 24 hours and
consumption of the tetradeoxynucleotide (56% conversion)
corresponded to the appearance of a number of new products
whose masses indicate singly and doubly modified d(ATGC).
Tandem MS analysis of the monoalkylation products allowed
unambiguous assignment of the purine bases as the sites of
modification (see Figure S1 in the Supporting Information).
Scheme 1. Proof-of-concept for rhodium-catalyzed NA alkylation.
With the feasibility of targeting NAs through rhodium-
carbenoid catalysis established, we undertook a more com-
prehensive investigation of the reaction with a series of
hairpin sequences (Table 1). Hairpins were chosen because
they contain a number of common NA structural elements in
a single molecule. The first hairpin we tested contained
thymidine (T) rings in its turn region and was otherwise
double-stranded (entry 1). We knew that T was unreactive
from the experiment with d(ATGC), so this molecule would
allow an assessment of the susceptibility of double-stranded
stretches to the alkylation. Interestingly this hairpin was
completely unreactive, thus revealing the prospect of exploit-
ing double strands as a type of shielding motif to enable the
targeting of specific unpaired bases in a given NA. To test this
possibility alkylation was attempted on hairpins containing
adenine as an overhang base (entries 2, 3, and 5) or in the turn
region (entry 4). As expected both of these motifs were viable
substrates. The specificity of the process for unpaired NA
sequences offers a new strategic tool for post-synthetic NA
[*] K. Tishinov, K. Schmidt, Priv.-Doz. D. Hꢀussinger,
Prof. Dr. D. G. Gillingham
Department of Chemistry, University of Basel
St. Johanns-Ring 19, CH-4056, Basel (Switzerland)
E-mail: dennis.gillingham@unibas.ch
[**] This work was supported by start-up funds from the University of
Basel. Karl Gademann and Florian Seebeck are gratefully acknowl-
edged for valuable discussions. Raphael Wyss is thanked for his
help in interpreting some of the NMR spectra.
Supporting information for this article is available on the WWW
12000
ꢀ 2012 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Angew. Chem. Int. Ed. 2012, 51, 12000 –12004