10246 J. Am. Chem. Soc., Vol. 123, No. 42, 2001
Maison et al.
pair could permit construction of designer proteins5 containing
helices that are confined to particular regions within the peptide
sequence. In this report we demonstrate examples of X-Hel
functions that meet two of these three goals.
Background and Design Plan
The large helix stabilizing effects reported for simple peptide
N-caps set a high standard for our goal of optimizing X-Hel
helix stabilizers. They also provide clues to the mechanisms of
cap stabilization. In 1987 Shoemaker et al. noted helix stabiliza-
tion for succinyl capped analogues of the C-peptide of RNAse
A, which they attributed to an interaction between the backbone
amide helix dipole and the negative charge of the N-terminal
succinate.6 Drawing from the protein X-ray crystallographic
database, in 1988 Presta and Rose noted sequence correlations
within the end regions of helices found in native structures of
globular proteins and argued that helix start and stop signals
are likely to exert more control over peptide helicity than
stabilizing effects attributable to helix propagation.7 In 1994,
using a 15 residue helix-prone model peptide, Forood et al. noted
that replacement of its N-terminal acetyl by a negatively charged
succinyl cap doubled the value of -[θ]222, implying an
equivalent increase in fractional helicity.8 They also noted little
change in -[θ]222 when the carboxylate anion of the N-succinyl
was replaced by a sulfonate anion or by an uncharged methyl
sulfone. All these capped peptides are exceptionally helical.
Studies by Nambiar et al.9 and by Doig and Baldwin10 with
peptides containing 12 to 17 residues further showed that,
relative to an uncharged alanine residue, the N-acetyl cap
increases -[θ]222 by an average of 70%. For the entire peptide
series the span of increase resulting from N-acetylation was 40-
120%, with larger increases observed for shorter peptides or
for weakly helical peptide sequences. Strongly helix stabilizing
N-capping effects were also seen for natural amino acids
normally classified as helix breakers; for example, a 40 to 60%
increase of -[θ]222 was observed when N-terminal unprotonated
alanine residues were replaced by asparagine, glycine, or serine
residues, or by negatively charged aspartates.
Figure 2. Comparison of templated and nontemplated N-capped
peptides. (a) The N-terminus of an N-acyl capped R-helical peptide
presents three NH functions that lack H-bonding capacity within the
helix (labeled H-1, H-2, and H-3). (b) The corresponding N-terminus
of an N-acyl-Hel peptide, in which the constrained Pro-Pro sequence
of Hel corresponds to the first two peptide residues of 2a. The potential
single H-bonding donor site H-3 is labeled.
can generate an H-bond with a single NH, but the remaining
three NH residues must find H-bond donors within a complex
of hydrating water molecules. The additional stabilization noted
by Forood et al. when the monovalent H-bonding acetyl N-cap
is replaced with a potentially bivalent succinyl can be explained
if its charged carboxylate stabilizes the N-terminus by assuming
the helix stabilizing roles of one or more water molecules.
Binding of anions by a helix N-terminus, usually attributed to
charge-dipole interactions, is well documented by protein X-ray
crystallographic data.11 However, the nearly equivalent helix
stabilizing effects of uncharged and negatively charged bivalent
caps reported for N-capped peptides by Forood et al. suggest
that the helix-stabilizing capacity of the second polar functions
of their N-caps may correlate better with H-bonding acceptor
affinity.
In certain respects the picture of Figure 2a is simplistic, but
a more incisive view of the roles of N-caps requires a more
detailed and rigorous model for the structure of the highly
functionalized and solvated N-terminus of a normal helical
peptide. Such models are unlikely to be forthcoming, given the
limitations of currently available molecular modeling protocols
and the degree to which conformational averaging complicates
interpretation of CD-derived peptide helicity. To design a new
generation of N-caps one would like answers to a series of
simple questions. Does a strongly helix stabilizing polar or
charged H-bond acceptor that is part of a peptide N-cap act by
interacting directly with a specific backbone NH residue of the
peptide? Does it interact directly with more than one NH via
bifurcated H-bonds? Does it stabilize the helix indirectly by
local charge-dipole effects or by modifying the structure or
affinities of the water molecules within the N-capping region?
These questions cannot be answered rigorously, but as a working
model for design of new experiments, we have used heuristic
comparisons of models for the structures of normal N-capped
helical peptides with models for analogous N-capped Hel-
peptide conjugates. This comparison leads us to focus attention
on H-bond formation by a particular NH of the peptide sequence.
The structure of the interface between a peptide and its Ac-
Hel cap can be viewed in two ways. The Ac-Hel cap can be
Three significant generalizations follow from these studies.
Large increases in peptide helicity can result from N-capping,
even by functions as simple as an acetyl. Negatively charged
N-caps are very strong helix stabilizers. Uncharged caps
containing a variety of polar functions can match the effective-
ness of negatively charged N-caps.
These generalizations suggest key features of mechanisms
by which highly efficient N-caps may stabilize helices. At an
N-terminus of an R-helix formed by an uncapped peptide, the
NH functions of the first four peptide amide residues lack
intramolecular H-bonds. As represented in Figure 2a, N-capping
by a simple acyl-derived H-bond donor such as an acetamide
(4) Deechongkit, S.; Kennedy, R. J.; Tsang, K. Y.; Renold, P.; Kemp,
D. S. Tetrahedron Lett. 2000, 41, 9679-9683.
(5) (a) Tuscherer, G.; Scheibler, L.; Dumy, P.; Mutter, M. Biopolymers
1998, 47, 63-73. (b) Imperiali, B.; Ottesen, J. J. Biopolymers 1998, 47,
23-29. (c) Kohn, W. D.; Hodges, R. S. Trends Biotechnol. 1998, 16, 379-
389. (d) Kemp, D. S. TIBTECH 1990, 8, 249-255.
(6) Shoemaker, K. R.; Kim, P. S.; York, E. J.; Stewart, J. M.; Baldwin,
R. L. Nature 1987, 326, 563-567.
(7) (a) Presta, L. G.; Rose, G. D. Science 1988, 240, 1632-1641. (b)
Aurora, R.; Rose, G. D. Protein Sci. 1998, 7, 21-38.
(8) Forood, B.; Reddy, H. K.; Nambiar, K. P. J. Am. Chem. Soc. 1994,
116, 6935-6936.
(9) Forood, B.; Feliciano, E. J.; Nambiar, K. P. Proc. Natl. Acad. Sci.
U.S.A. 1993, 90, 838-842.
(10) Doig, A. J.; Baldwin, R. L. Protein Sci. 1995, 4, 1325-1336. See
also N-Ac examples reported in: Chakrabartty, A.; Kortemme, T.; Baldwin,
R. L. Protein Sci. 1994, 3, 843-952.
(11) Copley, R. R.; Barton, G. J. J. Mol. Biol. 1994, 242, 321-329.