Introduction of cyclically constrained γ-residues stabilizes an α-peptide hairpin in aqueous solution

Article abstract of DOI:10.1021/ol4001125

The synthesis and structural characterization of hybrid α/γ-peptides resulting from a 1:1 α→γ residue substitution at cross-strand positions in a hairpin-forming α-peptide sequence are described. Cyclically constrained γ-residues based on 1,3-substituted

Full text of DOI:10.1021/ol4001125

ORGANIC
LETTERS
2013
Vol. 15, No. 4
944–947
Introduction of Cyclically Constrained
γ‑Residues Stabilizes an r‑Peptide
Hairpin in Aqueous Solution
George A. Lengyel, Geoffrey A. Eddinger, and W. Seth Horne*
Department of Chemistry, University of Pittsburgh, Pittsburgh, Pennsylvania 15260,
United States
horne@pitt.edu
Received January 14, 2013
ABSTRACT
The synthesis and structural characterization of hybrid R/γ-peptides resulting from a 1:1 Rfγ residue substitution at cross-strand positions in a
hairpin-forming R-peptide sequence are described. Cyclically constrained γ-residues based on 1,3-substituted cyclohexane or benzene scaffolds
support a native-like hairpin fold in aqueous solution, and the unnatural residues stabilize the folded state by ∼0.2 kcal/mol per Rfγ substitution.
Unnatural-backbone oligomers that fold like proteins,
coined “foldamers,”1 are of interest as therapeutic agents
due to their ability to combine protein-like biological
effects with stability to degradation by proteases.2 One
method for the design of a desired folding pattern and
function in an unnatural oligomer is to systematically alter
the backbone of a natural peptide or protein sequence to
generate an analogue with similar folding behavior.
To date, this sequence-based approach has been applied
to generate unnatural-backbone mimics of peptides with
helical,3 sheet,4 and nonregular5 secondary structure as
well as peptide/polymer chimeras based on a larger tertiary
fold.6
As part of a program aimed at the mimicry of a protein
tertiary structure by unnatural backbones, we have
focused efforts recently on hairpin and sheet secondary
structures.4 β-Sheets and hairpins are essential compo-
nents of natural tertiary folds,7 and their mimics have
potential use in diverse biomedical applications.8 One
method to generate unnatural-backbone sheet mimics is
to introduce β-amino acid residues in an R-peptide context
to produce heterogeneous-backbone R/β-peptides.9 Appli-
cation of 1:1 Rfβ residue substitution at matched posi-
tions in each strand of an R-peptide hairpin can lead to
R/β-peptide analogues that fold in organic solvent10 or
aqueous solution;4a however, β-residue incorporation
leads to an altered display of side chains relative to a
(1) Gellman, S. H. Acc. Chem. Res. 1998, 31, 173–180.
(2) (a) Bautista, A. D.; Craig, C. J.; Harker, E. A.; Schepartz, A. Curr.
Opin. Chem. Biol. 2007, 11, 685–692. (b) Goodman, C. M.; Choi, S.;
Shandler, S.; DeGrado, W. F. Nat. Chem. Biol. 2007, 3, 252–262. (c)
Guichard, G.; Huc, I. Chem. Commun. 2011, 47, 5933–5941. (d) Horne,
W. S. Expert Opin. Drug Discov. 2011, 6, 1247–1262.
(3) (a) Horne, W. S.; Boersma, M. D.; Windsor, M. A.; Gellman,
S. H. Angew. Chem., Int. Ed. 2008, 47, 2853–2856. (b) Horne, W. S.;
Price, J. L.; Gellman, S. H. Proc. Natl. Acad. Sci. U.S.A. 2008, 105,
9151–9156. (c) Horne, W. S.; Johnson, L. M.; Ketas, T. J.; Klasse, P. J.;
Lu, M.; Moore, J. P.; Gellman, S. H. Proc. Natl. Acad. Sci. U.S.A. 2009,
106, 14751–14756.
(4) (a) Lengyel, G. A.; Frank, R. C.; Horne, W. S. J. Am. Chem. Soc.
2011, 133, 4246–4249. (b) Lengyel, G. A.; Horne, W. S. J. Am. Chem.
Soc. 2012, 134, 15906–15913.
(5) Haase, H. S.; Peterson-Kaufman, K. J.; Lan Levengood, S. K.;
Checco, J. W.; Murphy, W. L.; Gellman, S. H. J. Am. Chem. Soc. 2012,
134, 7652–7655.
(7) (a) Searle, M. S.; Ciani, B. Curr. Opin. Struct. Biol. 2004, 14, 458–
464. (b) Nowick, J. S. Acc. Chem. Res. 2008, 41, 1319–1330.
(8) (a) Robinson, J. A. Acc. Chem. Res. 2008, 41, 1278–1288. (b)
Srinivas, N.; Jetter, P.; Ueberbacher, B. J.; Werneburg, M.; Zerbe, K.;
Steinmann, J.; Van der Meijden, B.; Bernardini, F.; Lederer, A.; Dias,
R. L. A.; Misson, P. E.; Henze, H.; Zumbrunn, J.; Gombert, F. O.;
Obrecht, D.; Hunziker, P.; Schauer, S.; Ziegler, U.; Kach, A.; Eberl, L.;
Riedel, K.; DeMarco, S. J.; Robinson, J. A. Science 2010, 327, 1010–
1013.
(6) Reinert, Z. E.; Musselman, E. D.; Elcock, A. H.; Horne, W. S.
ChemBioChem 2012, 13, 1107–1111.
(9) (a) Horne, W. S.; Gellman, S. H. Acc. Chem. Res. 2008, 41, 1399–
1408. (b) Pilsl, L. K. A.; Reiser, O. Amino Acids 2011, 41, 709–718.
r
10.1021/ol4001125
Published on Web 02/07/2013
2013 American Chemical Society

natural backbone. Sequence-guided RRfβ or RRfββ
residue replacement can restore a native-like side-chain
display, but at a thermodynamic cost to folded stability.4b
We considered the possibility that a strategy for mod-
ification of an R-peptide sheet that made use of alternate
unnatural building blocks besides β-residues might lead to
an oligomer with both a native-like fold and enhanced
aqueous folded stability relative to the natural prototype.
Here, we show that the incorporation of cyclically con-
strained γ-amino acid residues11 in each strand of a
protein-derived hairpin generates an R/γ-peptide analogue
with native-like folding behavior and improved aqueous
folded stability relative to the natural sequence.
Figure 1. (A) Structures of γ-amino acid residues mABA (X) and
ACC (Z). (B) Sequences of R-peptide 1 and R/γ-peptide analogues
2 and 3; hydrophobic residues key to hairpin folding are shown in
bold.
Simple acyclic γ-residues have been shown to be too
flexible to be accommodated into a folded hairpin or
sheet.12 Replacement of one of the backbone CꢀC single
bonds in a γ-residue with a double bond restricts torsional
freedom and can lead to hairpin formation in organic
solvent.13 An alternate method to impart backbone con-
formational rigidity is to employ cyclically constrained
γ-residues (Figure 1A).14ꢀ17 Designed hairpins containing
m-aminobenzoic acid (mABA, X) can fold in organic
solvent,15 and substituted derivatives have been incorpo-
rated into protein-like sheets that fold in water, blocking
interchain association in the process.16 A saturated analogue
of mABA, (1R,3S)-3-aminocyclohexanecarboxylic acid
(ACC, Z), has also found use as a component of sheet-like
structures. Cyclic oligomers in which ACC is alternated
with D-R-residues can stack via intermolecular backbone
hydrogen bonding to form nanotubular assemblies.17
These nanotubes, analogous to those that form from the
self-assembly of cyclic D,L-R-peptides,18 are essentially
cylindrical sheets.
with a high propensity to adopt an extended strand con-
formation. We further reasoned that such a strand might
beused asa component of a proteinsheet mimetic. Inorder
to test these hypotheses, we introduced ACC residues into
anR-peptide hairpinmodel systemderivedfroma bacterial
protein. We incorporated the related aromatic γ-amino
acid mABA into the same sequence as a point of comparison
and determined the folded structure and thermodynamic
folded stability of the resulting R/γ-peptides by multidimen-
sional NMR in aqueous buffer. Collectively, the results
obtained show that both mABA and ACC stabilize the hairpin
fold of the parent R-peptide without compromising the native-
like display of side chains near the unnatural residues.
We chose R-peptide 1 (Figure 1B),4b,19 a mutant of the
C-terminal hairpin from the Streptococcal protein GB1,20
as a model system to examine the folding propensity of ACC
and mABA in a protein sheet context. Two R/γ-peptide
variants of 1 were synthesized, incorporating mABA
(peptide 2) or ACC (peptide 3) in place of Ala residues at
cross-strand positions 4 and 13. This substitution pattern
maintains the hydrophobic side-chain cluster of Trp3, Tyr5,
Phe12, and Val14, which is essential for hairpin folding of 1
in water.
We synthesized the Fmoc-protected derivatives of ACC
and mABA for use in solid-phase peptide synthesis (SPPS).
Commercially available m-aminobenzoic acid was Fmoc
protected using standard methods. Enantiomerically pure
Boc-ACC was obtained by recrystallization of racemic
material derivatized as the (R)-1-phenethylamine salt21 and
converted to Fmoc-ACC by treatment with TFA followed
by Fmoc protection. Peptides 2 and 3 were prepared using
standard SPPS techniques, purified by reversed-phase
HPLC, and their identities were confirmed using MALDI-
TOF MS. All samples used for biophysical analysis were
>95% pure by analytical HPLC. Full experimental details
can be found in the Supporting Information (SI).
Given its behavior in a cyclic peptide context,17 we
hypothesized that pairing the (1R,3S) enantiomer of
ACC with L-R-residues would generate a peptide chain
(10) (a) Krauthauser, S.; Christianson, L. A.; Powell, D. R.; Gellman,
S. H. J. Am. Chem. Soc. 1997, 119, 11719–11720. (b) Karle, I. L.;
Gopi, H. N.; Balaram, P. Proc. Natl. Acad. Sci. U.S.A. 2001, 98,
3716–3719. (c) van der Knaap, M.; Otero, J. M.; Llamas-Saiz, A.;
van Raaij, M. J.; Lageveen, L. I.; Busscher, H. J.; Grotenbreg, G. M.;
van der Marel, G. A.; Overkleeft, H. S.; Overhand, M. Tetrahedron
2012, 68, 2391–2400.
(11) (a) Seebach, D.; Beck, A. K.; Bierbaum, D. J. Chem. Biodiversity
2004, 1, 1111–1239. (b) Vasudev, P. G.; Chatterjee, S.; Shamala, N.;
Balaram, P. Chem. Rev. 2011, 111, 657–687.
(12) (a) Roy, R. S.; Gopi, H. N.; Raghothama, S.; Karle, I. L.;
Balaram, P. Chem.;Eur. J. 2006, 12, 3295–3302. (b) Bandyopadhyay,
A.; Mali, S. M.; Lunawat, P.; Raja, K. M. P.; Gopi, H. N. Org. Lett.
2011, 13, 4482–4485.
(13) Hagihara, M.; Anthony, N. J.; Stout, T. J.; Clardy, J.; Schreiber,
S. L. J. Am. Chem. Soc. 1992, 114, 6568–6570.
(14) (a) Woll, M. G.; Lai, J. R.; Guzei, I. A.; Taylor, S. J. C.; Smith,
M. E. B.; Gellman, S. H. J. Am. Chem. Soc. 2001, 123, 11077–11078.
(b) Qureshi, M. K. N.; Smith, M. D. Chem. Commun. 2006, 5006–5008.
(15) Srinivasulu, G.; Rao, M. H. V. R.; Kumar, S. K.; Kunwar, A. C.
ARKIVOC 2004, 69–86.
(16) (a) Nowick, J. S.; Holmes, D. L.; Mackin, G.; Noronha, G.;
Shaka, A. J.; Smith, E. M. J. Am. Chem. Soc. 1996, 118, 2764–2765. (b)
Khakshoor, O.; Lin, A. J.; Korman, T. P.; Sawaya, M. R.; Tsai, S.-C.;
Eisenberg, D.; Nowick, J. S. J. Am. Chem. Soc. 2010, 132, 11622–11628.
(19) Fesinmeyer, R. M.; Hudson, F. M.; Andersen, N. H. J. Am.
Chem. Soc. 2004, 126, 7238–7243.
(17) (a) Amorı
´
n, M.; Castedo, L.; Granja, J. R. J. Am. Chem. Soc.
2003, 125, 2844–2845. (b) Brea, R. J.; Amorı
´
n, M.; Castedo, L.; Granja,
(20) Blanco, F. J. R., G.; Serrano, L. Nat. Struct. Mol. Biol. 1994, 1,
584–590.
(21) Badland, M.; Bains, C. A.; Howard, R.; Laity, D.; Newman,
S. D. Tetrahedron: Asymmetry 2010, 21, 864–866.
J. R. Angew. Chem., Int. Ed. 2005, 44, 5710–5713.
(18) Ghadiri, M. R.; Granja, J. R.; Milligan, R. A.; McRee, D. E.;
Khazanovich, N. Nature 1993, 366, 324–327.
Org. Lett., Vol. 15, No. 4, 2013
945

Peptides 2 and 3 were analyzed by NMR at 278 K in
50 mM phosphate buffer at pH 6.3 (9:1 H2O/D2O, pH
uncorrected for the presence of D2O). Structural analysis
in the demanding environment of aqueous buffer at near-
physiological pH provides an important test as to the
potential applicability of the Rfγ substitutions examined
in a biological context.22 Moreover, these specific condi-
tions allow direct comparison withdata wehave previously
obtained for R-peptide 1 and R/β-peptide analogues de-
rived from sequence-guided Rfβ residue replacement.4b
Homonuclear Hꢀ H TOCSY, NOESY, and COSY
experiments enabled unambiguous assignment of all back-
bone NꢀH and CꢀH resonances as well as most side chain
resonances in R/γ-peptides 2 and 3 (Tables S1, S2). Com-
parison of these data with results reported for parent R-
peptide 1 under identical conditions4b allowed us to ascer-
tain the impact of γ-residue incorporation on hairpin
thermodynamic stability.
Table 1. Thermodynamic Analysis of Folding Equilibria for
R-Peptide 1 and R/γ-Peptides 2 and 3a
0
Δδ Gly10 HR/HR
fraction
ΔGfold
ΔΔGfold
peptide
(ppm)b
folded (%) (kcal/mol)c (kcal/mol)d
1
2
3
0.21 ( 0.01
0.25 ( 0.01
0.26 ( 0.01
67 ( 6
79 ( 6
83 ( 6
ꢀ0.4 ( 0.1
ꢀ0.7 ( 0.2
ꢀ0.9 ( 0.2
ꢀ
ꢀ0.3
ꢀ0.5
a NMR experiments were performed in 50 mM phosphate, 9:1 H2O/
D2O, pH 6.3 at 278 K. b Chemical shift separation between diastereo-
topic HR protons of Gly10. c Free energy of folding. d Difference in ΔGfold
compared to R-peptide 1.
1
1
We used NMR chemical shift data for Gly10 to quantify
the folded population of R/γ-peptides 2 and 3, by analogy
to methods previously reported for R-peptide 1.4b The
chemical shift separation between diastereotopic Gly HR
protons (Δδ) is diagnostic of folding in hairpin peptides;
the magnitude of Δδ increases with folded population.23
We used data from a variant of 1 cyclized via a disulfide
bridge between N- and C-terminal Cys residues to deter-
mine Δδ for the fully folded state.4b In prior work, we
showed the Δδ value for the fully folded state does not vary
among this cyclic variant of 1 and several analogues with
different unnatural backbone compositions.4b
Folded populations determined by NMR were used to
calculate the equilibrium constant for hairpin formation
and, in turn, the folding free energy (ΔGfold) of each
peptide. Comparison of the thermodynamic data for the
folding equilibria of 1ꢀ3 (Table 1) yields two noteworthy
observations. First, the folded populations of the two
R/γ-peptides are similar, although the oligomer with the
ACC residues appears to be slightly more stable. Second,
both R/γ-peptides are actually more folded than parent
R-peptide 1 under the conditions of the experiment. This
gain in thermodynamic folded stability is a surprising
finding, since analogous sequence-based Rfβ residue
substitution is accompanied by a loss in stability to the
folded hairpin. The R/γ-peptide hairpins are stabilized by
∼0.2 kcal/mol per Rfγ residue replacement. This value
is comparable in magnitude to that observed for mutation
of Ala to Ile at a hydrophilic position in a β-sheet.24
Considering that isoleucine has the highest sheet pro-
pensity among ribosomally encoded R-residues, a similar
stabilization from incorporation of an unnatural residue is
a significant finding.
Figure 2. Select interstrand NOEs for R/γ-peptides 2 and 3.
mABA and ACC residues are shown in green. Dashed lines
indicate ambiguous NOE assignments. NMR experiments were
performed in 50 mM phosphate, 9:1 H2O/D2O, pH 6.3 at 278 K.
showed interstrand contacts consistent with hairpin for-
mation along the entire length of the backbone (Figure 2).
Additionally, cross-strand NOEs involving backbone pro-
tons indicated a hydrogen-bond register where side chains
flanking the γ-residue were juxtaposed on the same face
of the sheet. This observation is in contrast to the side-
chain inversion resulting from analogous 1:1 Rfβ residue
replacement4 and suggests the 1:1 Rfγ residue substitution
leads to a native-like sheet fold.
We employed NOE-derived distance restraints and
simulated annealing to generate high-resolution NMR
solution structures of 2 and 3. The ensemble of low-energy
structures obtained for each peptide (Figure S1) was
consistent with the expected hairpin fold. Minimized
average coordinates derived from each ensemble provide
a snapshot of the solution structures (Figure 3A,B). In
the case of 2, a minor conformer with a horseshoe shape
(Figure S1B) was observed in 2 of the 10 low energy
structures; only the structures from the major conformer
family were used when calculating the minimized average
coordinates.
With thermodynamic evidence of folding in hand, we
next analyzed long-range NOEs in R/γ-peptides 2 and 3 to
determine their folded structure in solution. Both 2 and 3
(22) Maynard, A. J.; Sharman, G. J.; Searle, M. S. J. Am. Chem. Soc.
1998, 120, 1996–2007.
(23) Searle, M. S.; Griffiths-Jones, S. R.; Maynard, A. J. J. Mol. Biol.
1999, 292, 1051–1069.
(24) Kim, C. A.; Berg, J. M. Nature 1993, 362, 267–270.
946
Org. Lett., Vol. 15, No. 4, 2013

Figure 3. (A,B) Minimized average coordinates from the NMR solution structure of R/γ-peptide 2 (A) and 3 (B). (C,D) Side-on views of
an mABA residue from 2 (C) and an ACC residue from 3 (D). (E,F) Overlay of R/γ-peptide 2 (E) or 3 (F) with the C-terminal hairpin
from protein GB1. The R/γ-peptides are shown in yellow, and the natural protein is in gray; the C, N, O, CR, and Cβ atoms of the four
hydrophobic residues used in pair-fitting are shown as sticks. γ-Residues are colored green. NMR experiments were performed in
50 mM phosphate, 9:1 H2O/D2O, pH 6.3 at 278 K.
R/γ-Peptides 2 and 3 both show hairpin folds with the
four key hydrophobic residues from 1 on the same face of
the sheet (Figure 3A,B). Close inspection of the γ-residues
in the structures of 2 and 3 show that the amides flanking
the unnatural residues are poised to stack in an extended
sheet without interference by the rings of the unnatural
residues (Figure 3C,D).
they display side chains flanking the points of Rfγ
substitution.
In summary, we have shown in this work that a 1:1 Rfγ
substitution at cross-strand positions in a prototype
R-peptide hairpin sequence leads to an R/γ-peptide analo-
gue that folds in neutral aqueous buffer. Thermodynamic
analysis of folding equilibria reveals that introduction
of the cyclically constrained γ-residue mABA or ACC
stabilizes the folded state on the order of 0.2 kcal/mol
per residue replaced. NMR solution structures of the
R/γ-peptides show that the display of hydrophobic core
residues from the R-peptide prototype is maintained and
that the γ-residues are oriented in a way that would allow
stacking in a larger sheet. Both constrained γ-residues
examined give a similar fold when introduced into the
parent sequence; however, ACC showed a slightly more
stable folded structure and a greater degree of structural
similarity to the natural protein. These observations sug-
gest ACC may be particularly amenable for use in the
mimicry of multistranded sheets in larger protein tertiary
structures.
In order to quantify the ability of R/γ-peptides 2 and 3
to display key hydrophobic side chains in a native-like
arrangement, we compared their NMR structures with the
corresponding hairpin segment from a crystal structure
of full length GB1,25 on which prototype sequence 1 is
based (Figure 3E,F). The overlays and resulting root-
mean-square deviation (RMSD) values were based on
structural alignment of CR and Cβ atoms from residues
Trp3, Tyr5, Phe12, and Val14. R/γ-Peptide 2 shows high
structural homology to the full-length GB1 protein in its
display of hydrophobic core residues. The RMSD for the
overlay of 2 with the native protein (1.12 A) is almost
identical to that for R-peptide 1 (1.18 A). R/γ-Peptide 3
had an even better agreement, with an RMSD of just
0.68 A for the same overlay, suggesting that ACC residues
may be a better mimic of natural structure than mABA
residues. In total, the above data suggest that the
folded structures resulting from introduction of either
mABA or ACC γ-residues at cross-strand positions of an
R-peptide are very similar to the natural backbone in how
Acknowledgment. We thank the University of Pittsburgh
for funding.
Supporting Information Available. Figure S1, Tables
S1ꢀS4, experimental methods, and NMR data for pep-
tides 1ꢀ3. This material is available free of charge via the
Internet at http://pubs.acs.org.
(25) Frericks Schmidt, H. L.; Sperling, L. J.; Gao, Y. G.; Wylie, B. J.;
Boettcher, J. M.; Wilson, S. R.; Rienstra, C. M. J. Phys. Chem. B 2007,
111, 14362–14369.
The authors declare no competing financial interest.
Org. Lett., Vol. 15, No. 4, 2013
947