Discrete assembly of synthetic peptide-DNA triplex structures from polyvalent melamine-thymine bifacial recognition

Article abstract of DOI:10.1021/ja2099326

We have designed a 21-residue α-peptide that simultaneously recognizes two decadeoxyoligothymidine (dT₁₀) tracts to form triplexes with a peptide-DNA strand ratio of 1:2. The synthetic peptide side chain displays 10 melamine rings, which provide a bifacial thymine-recognition interface along the length of the 21-residue peptide. Recognition is selective for thymine over other nucleobases and drives the formation of ternary peptide· [dT₁₀]₂ complexes as well as heterodimeric peptide· [dT₁₀C₁₀T₁₀] hairpin structures with triplex stems.

Full text of DOI:10.1021/ja2099326

Communication
pubs.acs.org/JACS
Discrete Assembly of Synthetic Peptide−DNA Triplex Structures from
Polyvalent Melamine−Thymine Bifacial Recognition
Yingying Zeng, Yaowalak Pratumyot, Xijun Piao, and Dennis Bong*
Department of Chemistry, The Ohio State University, 100 West 18th Avenue, Columbus, Ohio 43210, United States
S
* Supporting Information
ABSTRACT: We have designed a 21-residue α-peptide
that simultaneously recognizes two decadeoxyoligothymi-
dine (dT10) tracts to form triplexes with a peptide−DNA
strand ratio of 1:2. The synthetic peptide side chain
displays 10 melamine rings, which provide a bifacial
thymine-recognition interface along the length of the 21-
residue peptide. Recognition is selective for thymine over
other nucleobases and drives the formation of ternary
peptide·[dT10]2 complexes as well as heterodimeric
peptide·[dT10C10T10] hairpin structures with triplex stems.
Figure 1. Bifacial melamine−thymine recognition. Synthetic peptides
1−4 present melamine and methylated melamine on derivatized lysine
side chains. Peptide 2 is N-terminated with (5,6)-carboxyfluorescein
ucleic acid triplex structures formed from native
Noligonucleotides are known to occur via purine
Hoogsteen base-pairing of a third strand in the major groove
of a Watson−Crick base-paired duplex.1 This has been
extensively developed by Dervan2 and others3 as a targeting
concept using both native and artificial nucleobase recognition
elements and backbones. McLaughlin4 and Tor5 have both
reported elegant “Janus-wedge” 6 recognition of Watson−Crick
interfaces using synthetic nucleobases on PNA and sugar−
phosphate backbones, respectively. While these prior methods
have largely sought to develop general strategies for sequence
targeting of preformed oligonucleotide structures, there are
fewer synthetic approaches to generate structure in single-
stranded oligos or effect triplex formation without prior
oligonucleotide duplex assembly; such structure-inducing
recognition may be useful in design of synthetic regulators of
transcription7 or translation.8 We have explored this notion
using a Janus-wedge approach to address two identical
interfaces: two oligothymidine (dT10) DNA tracts were
assembled on a peptide template via bifacial melamine
recognition to form peptide−DNA triplex structures. Com-
pounds closely related to melamine and its canonical hydrogen-
bonding partner, cyanuric acid,9 have been used to site-
substitute for native nucleobases in PNA−DNA duplex
recognition. Melamine itself is a well-known molecular
recognition module in a number of contexts,10 and Baranger
and Zimmerman11 have recently reported melamine targeting
of thymine−thymine or uracil−uracil (T-T, U-U) mismatch
sites in d(CTG) and r(CUG) repeats, assisted by acridine
intercalation.12 Indeed, the hydrogen-bonding pattern of
melamine precisely complements the Watson−Crick face of
thymine/uracil (Figure 1). We recently characterized the
recognition of cyanuric acid and melamine derivatives at
aqueous interfaces13 as well as bulk solution14 and found that
binding was strongly dependent on the number of heterocycles
(Cbf) and β-alanine (β-Ala). Putative peptide−DNA complex
structures are shown: triplex 5 and hairpin 6.
per scaffold: a trivalent system yielded robust binding while
monovalent recognition was undetectable. Based on this prior
work, we hypothesized that multivalent presentation of
melamine heterocycles on a peptide backbone would bind
two strands of dT10 into a peptide−DNA heterotrimeric
bundle. Indeed, though dT10 has no detectable homooligome-
rization behavior, synthetic melamine-displaying peptide 1
induced assembly of a peptide−DNA triplex structures in
heterotrimeric stem and heterodimeric hairpin systems with a
peptide−DNA ratio of 1:2 and 1:1, respectively (Figure 1).
Prior studies from Eschenmoser and Krishnamurthy9a,c and
Ghadiri15 demonstrated the use of α-peptide backbones,
instead of the traditional PNA backbone,4,16 to recognize
DNA. This method forms a recognition interface from alternate
residues and was more synthetically convenient for our
purposes. We introduced melamine through side-chain
functionalization of Boc-Lysine with chlorodiaminotriazine, to
yield a derivative we term melaminolysine (M*). Boc
deprotection with TFA followed by reaction with Fmoc-OSu
yielded Fmoc-M*, which was used in standard solid-phase
peptide synthesis (SPPS) with a C-terminal glycinamide
(Figure 1). Alternate residues were glutamic acid to provide
water solubility and to avoid nonspecific electrostatic binding to
DNA, yielding peptide 1, (EM*)10G. This sequence was N-
terminally capped with carboxyfluorescein (peptide 2) to
permit fluorescence-based binding analysis. Control peptides
were also synthesized in which hydrogen-bonding sites were
systematically blocked by methylation of the exocyclic amines
on the melamine ring. Stepwise chloride displacement of
Received: October 21, 2011
Published: December 27, 2011
© 2011 American Chemical Society
832
dx.doi.org/10.1021/ja2099326 | J. Am. Chem.Soc. 2012, 134, 832−835

Journal of the American Chemical Society
Communication
trichlorotriazine17 with ammonia and/or dimethylamine and
Boc-lysine yielded the dimethylated and tetramethylated
melaminolysine derivatives, which were used in SPPS to
provide peptides 3 and 4, respectively.
Nanoparticle assembly from melamine and cyanuric acid
derivatives results from noncovalent polymerization of the two-
fold symmetric heterocycle recognition faces; thymine has only
one recognition face complementary to melamine, and thus
discrete assembly was anticipated. Gratifyingly, no large
peptide−DNA aggregates were detectable by dynamic light
scattering, consistent with the model of discrete triplex
formation. Peptide-triggered base-stacking signatures were
observed by UV absorbance changes, with a 1:2 stoichiometry
between 1 and dT10, consistent with bivalent melamine−
thymine recognition and formation of triplex 5; peptide alone
did not induce such a signal (Figures 1 and 2, Supporting
Figure 3. Circular dichroism spectra in DPBS, pH 7.4, of (A) 1
□
complexed with dT10 ( ) vs dT10 alone (--) and (B) 1 complexed with
○
dT10C10T10 ( ) vs dT10C10T10 alone (--). Peptide 1 in both is at 5 μM
concentration (), while dT10 and dT10C10T10 are maintained at 10
and 5 μM, respectively. Electrophoretic mobility shift assays imaged by
Cy5 fluorescence for (C) Cy5-dT10 (DNA1) and (D) Cy5-dT10C10T10
(DNA2) at 20 nM in each lane, with increasing peptide 1
concentration from left to right. (E) Relative electrophoretic mobilities
of the free DNA oligos and their peptide complexes, a mixture of
complex 5 and DNA1 shown in the central lane. See SI for further
details.
Figure 2. Peptide 1 titrated into (A) dT10 and (B) Flc-dT10C10T10-
Dabcyl, followed by UV absorbance (260 nm) and fluorescein
emission (521 nm), respectively. Saturation is observed at 33 and 50
mol% peptide, indicating 1:2 and 1:1 peptide:DNA binding
stoichiometries in (A) and (B), respectively.
Cooperative melting was observed for triplex 5 and hairpin 6
by both UV and fluorescence dequenching and differential
scanning calorimetry (DSC) (Figure 4, SI). The hairpin
Information (SI)). Unlike DNA and peptide−DNA4 triplex
structures which optimally form at high salt with divalent metal
ions (1−2 M NaCl, 50 mM MgCl2), robust assembly was
observed under standard salt conditions (Dulbecco’s phosphate
buffered saline, DPBS), akin to the conditions used by
Eschenmoser and Krishnamurthy9c for PNA−DNA duplex
formation. Similarly, UV absorbance signatures indicated a 1:1
binding stoichiometry between 1 and dT10C10T10, as would be
expected if an intramolecular peptide−DNA triplex structure
formed from the dT10 termini of dT10C10T10. Indeed, binding
of 1 to Flc-dT10C10T10-Dabcyl resulted in maximal fluorescein
quenching at a 1:1 peptide−DNA ratio, supportive of the
formation of heterodimeric hairpin structure 6, which would
bring the 3′ and 5′ ends of the oligo in close proximity,9c
resulting in efficient dabcyl quenching of fluorescein (Figure 2).
Triplex and hairpin formation was further corroborated by
circular dichroism, which indicated significant structuring of
DNA upon addition of peptide, signified by the development of
a negative CD signal at 260 nm at the expense of a positive CD
signal at 280 nm, which we assign to the peptide complex and
free DNA, respectively (Figure 3). Notably, while the peak at
280 nm is completely ablated in the triplex, there is a residual
peak in the hairpin; this is consistent with the presence of an
unstructured C10 loop found in hairpin 6 but not triplex 5.
Clean transformation of DNA to peptide−DNA complex bands
on native polyacrylamide electrophoresis indicated discrete
peptide−DNA recognition in both triplex and hairpin contexts
(Figure 3).
Figure 4. (A) First-derivative plot of melting transitions of triplex 5
(--) and hairpin 6 () followed by UV absorbance (260 nm).
Normalized absorbance change is shown inset. (B) DSC upscan traces
of triplex 5 () and hairpin 6 (), with downscan traces shown as
dashed regular and bold lines. Peptide−DNA ratios used in triplex 5
and hairpin 6 experiments were 1:2 and 1:1, respectively. DSC and UV
experiments were performed in DPBS, pH 7.4 at peptide
concentrations of 25 μM (DSC) and in UV experiments, 1 μM (5)
and 2.5 μM (6).
heterodimer structure 6 was more thermally stable (Tm = 54
°C) than the heterotrimeric triplex 5 (Tm = 43 °C). For
comparison, a dA10-T10 duplex has Tm = 35 °C, and a dA10-
(T10)2 triplex has Tm = 17 °C in the presence of 50 mM
MgCl2.18 It is calculated that a dA10-T10 duplex will have Tm =
22.5 °C under similar salt conditions. Reversible peptide−DNA
complexation was supported by observation of endothermic
melting and exothermic cooling peaks by DSC; though triplex 5
833
dx.doi.org/10.1021/ja2099326 | J. Am. Chem.Soc. 2012, 134, 832−835

Journal of the American Chemical Society
Communication
transition temperature was elevated (48 °C) relative to the
other methods, the hairpin Tm was very similar (55 °C).
Assembly of complexes 5 and 6 was highly exothermic and
reproducible over several thermal cycles (ΔHtriplex = −285 kcal/
mol peptide, ΔHhairpin = −314 kcal/mol), consistent with base-
stacking-driven assembly.19 The large enthalpy values (−28.5 to
−31.4 kcal/mol per peptide:DNA triplet layer) are somewhat
higher than, but similar to, enthalpies observed for (TAT)n
intramolecular DNA triplexes,20 which exhibit −21 kcal/mol
per triplet stack. Peptide-displayed melamine−DNA binding
appears to have exothermic assembly profiles and Tm values
similar to those of DNA−DNA and melamine−cyanuric acid
recognition, suggestive of similar driving forces, perhaps to be
expected given the similarity of melamine and native
nucleobases. Given that peptides 1 and 2 and DNA are all
anionic polyelectrolytes, we anticipate negligible nonspecific
electrostatic interactions. Indeed, partial and full methylation of
each melamine ring on the peptide to respectively yield
peptides 3 and 4 abolished all detectable binding to dT10 or
dT10C10T10, indicating the importance of the melamine
recognition interface (SI). As methylation increases hydro-
phobicity, it is clear that assembly depends on recognition of
donor−acceptor patterning and not simply nonspecific hydro-
phobic collapse. Notably, peptides 1 and 2 did not show any
binding signatures when annealed with dA10, dC10, and dG5A10,
indicating a selectivity for thymine over the other native
nucleobases (SI). Thus, melamine peptide recognition is
relatively specific for the hydrogen-bonding pattern of
oligothymidine, as predicted. Taken together with the observed
stoichiometry of binding with dT10 and dT10C10T10, this is
strongly supportive of the expected bivalent thymine
recognition by melamine (Figure 1).
rapidly with another strand of dT10 to form the heterotrimer.
Indeed, the coefficient derived from a Hill-type binding
isotherm was essentially 2, supportive of a cooperative 1:2
peptide−DNA assembly process. A hypothetical equivalent
two-step dissociation process with an overall apparent Kd =
4000 nM2 would have Kd = 63 nM for both trimer−dimer and
dimer−monomer dissociation steps, which is the observed free
DNA concentration at half-saturation (Figure 5). A good fit to
a 1:1 binding model was obtained from the binding isotherm of
peptide 2 to dT10C10T10, revealing an approximate Kd = 2.7 nM
for dissociation of hairpin 6.
Overall, these data support the model of bivalent melamine−
thymine recognition yielding formation of novel discrete
peptide−DNA triplex structures with robust binding affinities.
Notably, prior work from Eschenmoser and Krishnamurthy
using similar diaminotriazine nucleobase mimics derived from
aspartic and glutamic acids resulted in duplex formation with
oligothymidines rather than triplex.9a Diaminotriazine has two
exocyclic hydrogen bond donor sites compared to three on
melamine, which yields two potential thymine recognition
interfaces in melaminolysine and just one in diaminotriazine
derivatives. Thus, it appears that the assembly state pivots from
dimer to trimer based on the addition of a single hydrogen
bond donor, with the increased entropic cost likely paid for by
increased base stacking. Polyvalent melamine−thymine recog-
nition between peptides 1 and 2 with dT10 tracts is unique in
that structure is induced in unstructured single-stranded
oligothymidines to yield novel triplex and hairpin peptide−
DNA hybrid structures, thus expanding the range of non-native
nucleobase-derived structures already known.21 Though
melamine recognition of preformed T-T/U-U mismatch sites
has been previously reported,11 the ability of polyvalent
melamine peptides to broker assembly with two non-interacting
oligothymidine strands to form peptide−DNA triplex structures
is non-obvious. We anticipate that this design element may be
used to manipulate the structure and function of thymine- or
uracil-rich targets.
We quantified association using fluorescein-tagged peptide 2
binding to unlabeled DNA (Figure 5). Complex formation was
ASSOCIATED CONTENT
* Supporting Information
■
S
Additional CD, fluorescence, UV, and fluorescence anisotropy
data, fitting procedures, experimental details, and detailed
synthetic procedures and compound characterization. This
material is available free of charge via the Internet at http://
pubs.acs.org.
AUTHOR INFORMATION
Corresponding Author
bong@chem.osu.edu
Figure 5. Binding isotherms in DPBS, pH 7.4, of (A) peptide 2
binding to dT10 followed by fluorescein quenching upon binding and
(B) 2 binding to dT10C10T10 followed by fluorescence anisotropy.
Solid lines show fits to (A) 1:2 binding (fraction bound peptide =
[DNA]2/Kd + [DNA]2, R2 ≥ 0.96) and (B) 1:1 binding model,
corrected for fluorescence quenching, R2 ≥ 0.98. Peptide concen-
tration is constant at 25 nM.
■
ACKNOWLEDGMENTS
This work was supported in part by the NSF and The Institute
of Materials Research at The Ohio State University.
■
followed by both fluorescence anisotropy and quenching, which
accompanies assembly; similar results were obtained with the
two signatures. Using the stoichiometries established by
titration (Figure 2), we fit the binding curves to a trimer−
monomer model (SI) to obtain an apparent Kd ≈ 4000 nM2
(by quenching or anisotropy) for dissociation of triplex 5. A
second transition was undetectable by fitting to a two-step
model or by observation of assembly processes, suggesting that
the peptide−dT10 heterodimer that forms initially must react
REFERENCES
■
(1) (a) Duca, M.; Vekhoff, P.; Oussedik, K.; Halby, L.; Arimondo, P.
B. Nucleic Acids Res. 2008, 36, 5123. (b) Francois, J. C.; Saison-
́ ̀
Behmoaras, T.; Helene, C. Nucleic Acids Res. 1988, 16, 11431.
(c) Moser, H. E.; Dervan, P. B. Science 1987, 238, 645.
(2) Dervan, P. B.; Doss, R. M.; Marques, M. A. Curr. Med. Chem.
Anticancer Agents 2005, 5, 373.
(3) Simon, P.; Cannata, F.; Concordet, J.-P.; Giovannangeli, C.
Biochimie 2008, 90, 1109.
834
dx.doi.org/10.1021/ja2099326 | J. Am. Chem.Soc. 2012, 134, 832−835

Journal of the American Chemical Society
Communication
(4) Chen, H.; Meena.; McLaughlin, L. W. J. Am. Chem. Soc. 2008,
130, 13190.
(5) Shin, D.; Tor, Y. J. Am. Chem. Soc. 2011, 133, 6926.
(6) Branda, N.; Kurz, G.; Lehn, J.-M. Chem. Commun. 1996, 1996,
2443.
(7) (a) Mapp, A. K.; Ansari, A. Z. ACS Chem. Biol. 2007, 2, 62.
(b) Payankaulam, S.; Li, L. M.; Arnosti, D. N. Curr. Biol. 2010, 20,
R764.
(8) (a) DeiganK. E.Ferre-
wa, A. RNA 2011, 17, 1. (c) Zhang, J.; Lau, M. W.; Ferre-
R. Biochemistry 2010, 49, 9123. (d) Suess, B.; Weigand, J. E. RNA Biol.
2008, 5, 24.
́
D’AmareA
́
. R.Acc. Chem. Res.2011(b) Oga-
́
D’Amare, A.
́
(9) (a) Mittapalli, G. K.; Reddy, K. R.; Xiong, H.; Munoz, O.; Han,
B.; De Riccardis, F.; Krishnamurthy, R.; Eschenmoser, A. Angew.
Chem., Int. Ed. 2007, 46, 2470. (b) Vysabhattar, R.; Ganesh, K. N.
Tetrahedron Lett. 2008, 49, 1314. (c) Mittapalli, G. K.; Osornio, Y. M.;
Guerrero, M. A.; Reddy, K. R.; Krishnamurthy, R.; Eschenmoser, A.
Angew. Chem., Int. Ed. 2007, 46, 2478.
(10) (a) Ai, K.; Liu, Y.; Lu, L. J. Am. Chem. Soc. 2009, 131, 9496.
(b) Kawasaki, T.; Tokuhiro, M.; Kimizuka, N.; Kunitake, T. J. Am.
Chem. Soc. 2001, 123, 6792. (c) Prins, L. J.; De Jong, F.; Timmerman,
P.; Reinhoudt, D. N. Nature 2000, 408, 181. (d) Prins, L. J.;
Reinhoudt, D. N.; Timmerman, P. Angew. Chem., Int. Ed. 2001, 40,
2382. (e) Kimizuka, N.; Kawasaki, T.; Hirata, K.; Kunitake, T. J. Am.
Chem. Soc. 1998, 120, 4094. (f) Mathias, J. P.; Simanek, E. E.; Seto, C.
T.; Whitesides, G. M. Macromol. Symp. 1994, 77, 157. (g) Zerkowski,
J. A.; MacDonald, J. C.; Seto, C. T.; Wierda, D. A.; Whitesides, G. M. J.
Am. Chem. Soc. 1994, 116, 2382. (h) Seto, C. T.; Whitesides, G. M. J.
Am. Chem. Soc. 1993, 115, 905. (i) Zerkowski, J. A.; Seto, C. T.;
Whitesides, G. M. J. Am. Chem. Soc. 1992, 114, 5473. (j) Seto, C. T.;
Whitesides, G. M. J. Am. Chem. Soc. 1990, 112, 6409.
(11) Arambula, J. F.; Ramisetty, S. R.; Baranger, A. M.; Zimmerman,
S. C. Proc. Natl. Acad. Sci. U.S.A. 2009, 106, 16068.
(12) (a) Rapireddy, S.; He, G.; Roy, S.; Armitage, B. A.; Ly, D. H. J.
Am. Chem. Soc. 2007, 129, 15596. (b) Fechter, E. J.; Olenyuk, B.;
Dervan, P. B. Angew. Chem., Int. Ed. 2004, 43, 3591. (c) Fechter, E. J.;
Dervan, P. B. J. Am. Chem. Soc. 2003, 125, 8476. (d) Durand, M.;
́ ̀
Maurizot, J. C.; Asseline, U.; Thuong, N. T.; Helene, C. Bioconjugate
Chem. 1993, 4, 206. (e) Bentin, T.; Nielsen, P. E. J. Am. Chem. Soc.
2003, 125, 6378.
(13) (a) Ma, M.; Bong, D. Langmuir 2011, 27, 1480. (b) Ma, M.;
Gong, Y.; Bong, D. J. Am. Chem. Soc. 2009, 131, 16919. (c) Ma, M.;
Paredes, A.; Bong, D. J. Am. Chem. Soc. 2008, 130, 14456. (d) Ma, M.;
Bong, D. Org. Biomol. Chem. 2011, 9, 7296.
(14) Ma, M.; Bong, D. Langmuir 2011, 27, 8841.
(15) Ura, Y.; Beierle, J. M.; Leman, L. J.; Orgel, L. E.; Ghadiri, M. R.
Science 2009, 325, 73.
(16) (a) Nielsen, P. E. Chem. Biodiversity 2007, 4, 1996.
(b) Porcheddu, A.; Giacomelli, G. Curr. Med. Chem. 2005, 12, 2561.
(17) Zhang, W.; Nowlan, D. T.; Thomson, L. M.; Lackowski, W. M.;
Simanek, E. E. J. Am. Chem. Soc. 2001, 123, 8914.
(18) (a) Pilch, D. S.; Brousseau, R.; Shafer, R. H. Nucleic Acids Res.
1990, 18, 5743. (b) Pilch, D. S.; Levenson, C.; Shafer, R. H. Proc. Natl.
Acad. Sci. U.S.A. 1990, 87, 1942.
(19) (a) Kool, E. T. Annu. Rev. Biophys. Biomol. Struct. 2003, 30, 1.
(b) Guckian, K. M.; Schweitzer, B. A.; Ren, R. X. F.; Sheils, C. J.;
Tahmassebi, D. C.; Kool, E. T. J. Am. Chem. Soc. 2000, 122, 2213.
(c) McKay, S. L.; Haptonstall, B.; Gellman, S. H. J. Am. Chem. Soc.
2001, 123, 1244. (d) Yakovchuk, P.; Protozanova, E.; Frank-
Kamenetskii, M. D. Nucleic Acids Res. 2006, 34, 564. (e) SantaLucia,
J. Jr.; Hicks, D. Annu. Rev. Biophys. Biomol. Struct. 2004, 33, 415.
(20) Soto, A. M.; Loo, J.; Marky, L. A. J. Am. Chem. Soc. 2002, 124,
14355.
(21) (a) Krueger, A. T.; Kool, E. T. Chem. Biol. 2009, 16, 242.
(b) Krueger, A. T.; Kool, E. T. Curr. Opin. Chem. Biol. 2007, 11, 588.
835
dx.doi.org/10.1021/ja2099326 | J. Am. Chem.Soc. 2012, 134, 832−835