Efficient and Sequence-Specific DNA-Templated Polymerization of Peptide Nucleic Acid Aldehydes

Article abstract of DOI:10.1021/ja038058b

On the basis of the distance-dependence of DNA-templated reductive amination reactions and of recent findings of D. Lynn and co-workers, we developed DNA-templated polymerizations of synthetic peptide nucleic acid (PNA) aldehydes. The coupling reactions proceed in a highly efficient and sequence-specific manner, even in the presence of mixtures of PNA aldehydes of different sequence. Synthetic peptide nucleic acid polymers containing as many as 40 PNA units (representing 10 consecutive coupling reactions) were formed efficiently. The ease of preparing PNAs containing tailor-made functional groups together with these findings raises the possibility of evolving synthetic sequence-defined polymers by iterated cycles of translation, selection, PCR amplification, and diversification previously available only to biological macromolecules. Copyright

Full text of DOI:10.1021/ja038058b

Published on Web 10/28/2003
Efficient and Sequence-Specific DNA-Templated Polymerization of Peptide
Nucleic Acid Aldehydes
Daniel M. Rosenbaum and David R. Liu*
Department of Chemistry and Chemical Biology, HarVard UniVersity, Cambridge, Massachusetts 02138
Received August 21, 2003; E-mail: drliu@fas.harvard.edu
The highly efficient and sequence-specific nucleic acid-templated
polymerization of proteins and nucleic acids is a fundamental feature
of living systems that enables biopolymers to evolve.1 Over the
past few decades, several efforts to adapt templated polymerization
to synthetic systems have been reported.2 These systems typically
use oligonucleotides (or analogues) as a template for coupling DNA-
or RNA-based monomers. Despite significant advances,3 the
generality and sequence specificity of template-directed polymer-
ization is still largely unexplored. For example, the efficient and
sequence-specific templated polymerization of easily functionalized
synthetic monomers lacking a ribose backbone has not been
reported. Such a system may enable polymers comprised of these
synthetic monomers to evolve through iterated cycles of translation
(polymerization), selection, and amplification presently available
only to DNA, RNA, and proteins.
The minimal requirements of a system for synthetic polymer
evolution are (i) efficient nucleic acid-templated oligomerization
to provide sufficient quantities of full-length products for in vitro
selections;4 (ii) distance-dependent nucleic acid-templated monomer
coupling reactions5 to ensure that oligomerization proceeds exclu-
sively between adjacently annealed monomers; (iii) stable linkage
of each synthetic polymer to its encoding nucleic acid template to
ensure the survival of the appropriate template during polymer
selection; and (iv) a readily functionalized synthetic monomer
backbone to introduce chemical functionality into the polymer.
Peptide nucleic acids (PNAs) are attractive candidates for
synthetic polymer evolution because of their known ability to bind
DNA sequence specifically and their simple preparation from
synthetically accessible amino acids. Previous efforts to oligomerize
PNAs on DNA or RNA templates used amine acylation as the
coupling reaction and proceeded with modest efficiency and
sequence specificity.6 Consistent with these findings, when we
combined five PNA tetramers using a variety of aqueous amine
acylation conditions5 in the presence of DNA templates containing
complementary 20-base annealing regions, we observed only modest
formation (e20% yield) of full-length PNAs representing five
successive coupling reactions (Supporting Information). Even more
problematic, the formation of higher molecular weight products was
observed independent of the position of a mismatched 4-base
annealing region in the template (Supporting Information). These
observations indicate that PNAs are able to couple using amine
acylation chemistry even when not adjacently annealed, leading to
an unpredictable mixture of products.
Figure 1. DNA-templated PNA aldehyde oligomerization. A 5′-amine-
terminated DNA template (0.4 µM) was mixed with PNA aldehyde 1 (8
µM), heated to 95 °C, and cooled to 25 °C over 60 min. NaCNBH3 was
added to 80 mM. After 30 min, reactions were subjected to gel filtration
and denaturing PAGE (bottom). Band doubling in lanes 13 and 14 is
attributed to open versus closed hairpin structures in the products.
results with reductive amination-mediated step-growth oligomer-
ization of DNA analogues.3b
A thymine-containing PNA monomer amino aldehyde was
synthesized and coupled to threonine-linked resin.7 Standard FMOC
peptide synthesis was used to extend the peptide by three PNA
monomers (final sequence: NH2-gact-CHO), and aqueous acidic
cleavage from the resin yielded the desired tetrameric peptide
aldehyde (1, Figure 1). A DNA template containing a 5′-amine-
terminated hairpin and five successive repeats of the 5′-AGTC-3′
“codon” complementary to 1 was combined with 8 µM 1 in aqueous
pH 8.5 buffer. The reactants were annealed, and NaCNBH3 was
added. The starting template was almost entirely consumed, and a
higher molecular weight product was formed in >90% yield (Figure
1). Gel purification of the product followed by removal of the DNA
template with DNase I and MALDI-TOF mass spectrometry
confirmed a full-length pentamer of the gact PNA aldehyde
(expected mass ) 5366; observed mass ) 5375 ( 15). No product
was observed in the absence of NaCNBH3 (Supporting Information).
These results indicate that DNA-templated reductive amination can
mediate the highly efficient oligomerization of PNA aldehydes.
To examine the regio- and sequence specificity of this reaction,
we repeated the above oligomerization using a variety of template
sequences. Remarkably, when a mismatched DNA template codon
(5′-ATGC-3′) was introduced at the second, third, fourth, or fifth
4-base coding region of the template, highly efficient (>90% yield)
formation of products corresponding to the coupling of one, two,
three, or four copies of 1, respectively, was observed (Figure 1).
When the mismatched codon was placed at only the first coding
position, or at all five coding positions, no product formation was
observed (Figure 1). The termination of oligomerization at the first
mismatched codon in every case indicates that this DNA-templated
We hypothesized that the distance independence we previously
reported in DNA-templated amine acylation reactions5b,c was the
origin of the poor regiospecificity of amine acylation-mediated PNA
couplings. A distance-dependent DNA-templated reaction such as
reductive amination5b,c should address this problem and serve as a
more promising chemistry for DNA-templated PNA polymeriza-
tions. Indeed, Lynn and co-workers previously reported excellent
9
13924
J. AM. CHEM. SOC. 2003, 125, 13924-13925
10.1021/ja038058b CCC: $25.00 © 2003 American Chemical Society

C O M M U N I C A T I O N S
Figure 2. Polymerization on long (40-base) templates. Polymerizations
were carried out as in Figure 1, with the following exceptions: PNA
aldehyde concentration was 16 µM, and reaction time with NaCNBH3 was
15 min. Gel lanes alternate between template only (with mismatch at
indicated position) and reactions (template + gact monomer).
Figure 3. Polymerization in the presence of multiple PNA aldehydes.
Conditions are the same as those in Figure 2, except that the reaction time
was 5 min at 37 °C. Lane 1 (A) contains 16 µM 1 + 2 µM of each of the
eight other gvvt PNA aldehydes. Each set of four lanes after lane 3
contains: (T) template only; (O) reaction with 16 µM 1; (C) reaction with
14.4 µM 1 + 1.6 µM PNA aldehyde complementary to the orange codon;
(E) reaction with 14.4 µM 1 + 0.2 µM of each PNA aldehyde of sequence
gvvt except the PNA complementary to the orange codon (eight PNAs total).
PNA aldehyde coupling indeed requires functional group adjacency
(i.e., is highly distance dependent5) and therefore is ideally suited
for templated polymerizations.3b
We further probed the sequence specificity of this system by
performing oligomerization experiments using DNA templates
containing eight different mismatched codons (ATTC, ATGC,
ATCC, AGGC, AGCC, ACTC, ACGC, or ACCC) in the third
coding region. Even though four of these codons differ from the
matched sequence (AGTC) in only a single base, in each case only
two copies of 1 were coupled to the template (Figure 1). This high
degree of sequence specificity raises the possibility that libraries
of different DNA sequences may be faithfully translated into
libraries of corresponding polymers using this system, analogous
to our ongoing DNA-templated small-molecule studies.5
aldehyde tetramers were included in the reaction except the PNA
complementary to the fifth coding region (eight in total), again only
the truncated product was predominantly generated (Figure 3, E
lanes). Taken together, these experiments reveal that DNA-
templated PNA aldehyde polymerizations maintain efficiency and
sequence specificity even in the presence of a mixture of different
PNA building blocks closely related in sequence.
In summary, we have developed an efficient and sequence-
specific translation of DNA templates containing as many as 40
bases of mixed coding sequence into corresponding synthetic
peptide nucleic acid polymers. These findings are a key step toward
the evolution of sequence-defined synthetic heteropolymers, rather
than biological macromolecules, using iterated cycles of DNA-
templated translation, in vitro selection, PCR amplification, and
sequence diversification.
Synthetic polymers with desired properties may require lengths
beyond those previously achieved efficiently using nucleic acid-
templated synthesis. To test the ability of the above system to
generate longer polymers efficiently and sequence specifically, we
translated DNA templates with 40-base coding regions encoding
10 repeats of the above matched or mismatched codon into
corresponding PNA aldehyde polymers. Gratifyingly, both denatur-
ing PAGE and MALDI-TOF mass spectrometry revealed a single
predominant product corresponding to the polymerization of a full
length 40-mer PNA after 15 min (expected mass ) 10 719; ob-
served mass ) 10 729 ( 30) (Figure 2 and Supporting Information).
Introducing a mismatched codon in the first, third, fifth, seventh,
or ninth coding positions on the template again resulted in truncation
(Figure 2). This efficient translation of DNA sequences into 40
PNA bases (10 couplings) provides a polymer of length similar to
that of DNA and RNA oligonucleotides with binding or catalytic
properties,4a but made entirely of synthetic building blocks.
A challenging requirement of creating libraries of sequence-
defined synthetic polymers in this manner is maintaining sequence
specificity in the presence of multiple monomers of closely related
sequence. To study the specificity of DNA-templated polymeriza-
tion using multiple PNA building blocks in a single solution, we
synthesized the nine PNA aldehyde tetramers of the sequence NH2-
gvvt-CHO (v ) g, a, or c). We also prepared nine DNA templates
containing one of nine codons complementary to gvvt at codon 5,
and containing AGTC at codons 1-4 and 6-10. Each of the eight
templates not containing AGTC at position 5 was translated into a
predominant truncated product of one apparent length when 1 was
the only PNA aldehyde included in the reaction (Figure 3, 0 lanes).
Full-length polymer was the major product for all nine templates,
however, when the PNA aldehyde complementary to the fifth codon
was included in addition to 1 (C lanes), or when all nine PNA
aldehydes were included (A lane). Importantly, when all PNA
Acknowledgment. This work was supported by the Office of
Naval Research (N00014-03-1-0749), the Beckman Foundation, the
Searle Scholars Program (00-C-101), and the Sloan Foundation
(BR-4141). D.M.R. is a Hertz Foundation Graduate Fellow. We
thank Chieh-Ting Yeh for experimental assistance.
Supporting Information Available: Experimental details, oligo-
nucleotide sequences, and additional results (PDF). This material is
available free of charge via the Internet at http://pubs.acs.org.
References
(1) (a) Kuchta, R. D.; Benkovic, P.; Benkovic, S. J. Biochemistry 1988, 27,
6716-6725. (b) Steitz, T. A. HarVey Lect. 1997-98, 93, 75. (c) Yonath,
A. Annu. ReV. Biophys. Biomol. Struct. 2002, 31, 257.
(2) (a) Joyce, G. F. Nonezymatic Template-Directed Synthesis of Informational
Macromolecules; Cold Spring Harbor Laboratory Press: Plainview, NY,
1987; LII, pp 41-51. (b) Orgel, L. E. Nature 1992, 358, 203-209. (c)
Bag, B. G.; von Keidrowski, G. Pure Appl. Chem. 1996, 68, 2145-2152.
(d) Gat, Y.; Lynn, D. G. Templated Organic Synthesis; Wiley: New York,
2000; p 133.
(3) (a) Kozlov, I. A.; De Bouvere, B.; Aerschot, A. V.; Herdewijn, P.; Orgel,
L. E. J. Am. Chem. Soc. 1999, 121, 5856-5859. (b) Li, X.; Zhan, Z. J.;
Knipe, R.; Lynn, D. G. J. Am. Chem. Soc. 2002, 124, 746-747.
(4) (a) Wilson, D. S.; Szostak, J. W. Annu. ReV. Biochem. 1999, 68, 611-
647. (b) Doyon, J. B.; Snyder, T. M.; Liu, D. R. J. Am. Chem. Soc. 2003,
125, 12372-12373.
(5) (a) Gartner, Z. J.; Liu, D. R. J. Am. Chem. Soc. 2001, 123, 6961. (b)
Gartner, Z. J.; Kanan, M. W.; Liu, D. R. Angew. Chem., Int. Ed. 2002,
41, 1796. (c) Gartner, Z. J.; Grubina, R.; Calderone, C. T.; Liu, D. R.
Angew. Chem., Int. Ed. 2003, 42, 1370-1375.
(6) (a) Bohler, C.; Nielsen, P. E.; Orgel, L. E. Nature 1995, 376, 578-581.
(b) Schmidt, J. G.; Christensen, L.; Nielsen, P. E.; Orgel, L. E. Nucleic
Acids Res. 1997, 25, 4792-4796.
(7) Ede, N. J.; Bray, A. M. Tetrahedron Lett. 1997, 38, 7119-7122.
JA038058B
9
J. AM. CHEM. SOC. VOL. 125, NO. 46, 2003 13925