A coiled coil with a fluorous core

Article abstract of DOI:10.1021/ja002961j

The design, synthesis, and structural characterization of a highly fluorinated peptide system based on the coiled coil region of the yeast transcription factor GCN4 is described. All four leucine residues (a position) and three valine residues (d position) were replaced by the unnatural amino acids 5,5,5-trifluoroleucine and 4,4,4-trifluorovaline, respectively. The peptide is highly α-helical at low micromolar concentrations as judged by circular dichroism spectra, sediments as a dimeric species in the 5-30 μM concentration range, and exhibits a dimer melting temperature that is 15 °C higher than a control peptide with a hydrocarbon core. Furthermore, the apparent free energy of unfolding as calculated from guanidinium hydrochloride denaturation experiments is larger by ～1.0 kcal/mol for the fluorinated peptide than its hydrocarbon counterpart. We conclude that additional stability is derived from sequestering the more hydrophobic trifluoromethyl groups from aqueous solvent. These studies introduce a new paradigm in the design of molecular self-assembling systems, one based on orthogonal solubility properties of liquid phases.

Full text of DOI:10.1021/ja002961j

J. Am. Chem. Soc. 2001, 123, 4393-4399
4393
A Coiled Coil with a Fluorous Core
Basar Bilgic¸er, Alfio Fichera, and Krishna Kumar*
Contribution from the Department of Chemistry, Tufts UniVersity, Medford, Massachusetts 02155
ReceiVed August 9, 2000. ReVised Manuscript ReceiVed January 18, 2001
Abstract: The design, synthesis, and structural characterization of a highly fluorinated peptide system based
on the coiled coil region of the yeast transcription factor GCN4 is described. All four leucine residues (a
position) and three valine residues (d position) were replaced by the unnatural amino acids 5,5,5-trifluoroleucine
and 4,4,4-trifluorovaline, respectively. The peptide is highly R-helical at low micromolar concentrations as
judged by circular dichroism spectra, sediments as a dimeric species in the 5-30 µM concentration range, and
exhibits a dimer melting temperature that is 15 °C higher than a control peptide with a hydrocarbon core.
Furthermore, the apparent free energy of unfolding as calculated from guanidinium hydrochloride denaturation
experiments is larger by ∼1.0 kcal/mol for the fluorinated peptide than its hydrocarbon counterpart. We conclude
that additional stability is derived from sequestering the more hydrophobic trifluoromethyl groups from aqueous
solvent. These studies introduce a new paradigm in the design of molecular self-assembling systems, one
based on orthogonal solubility properties of liquid phases.
Introduction
bZIP transcriptional activator GCN4. We hypothesized that
immiscibility of fluorous phases in both water and many organic
solvents should provide enough driving force for fluorinated
side chains to collapse in a predetermined fashion to yield a
fluorous core.8 We expect such altered cores to have desirable
properties for use in high-temperature catalysis,9 stabilization
of native protein structures in organic solvents,9b selective
protein-protein interactions,10 and the design of membrane
spanning peptides.11
Formation of defined structures in biology is frequently
accompanied by separation of immiscible phases into distinct
domains.1,2 Indeed, the energetic advantage of getting nonpolar
substances to minimize their contacts with water is evident in
the structure of the plasma membrane and of most proteins,
which in essence form intramolecular micelles. Phase separation
is therefore a structural imperative in biology. Recently, the
orthogonal solubility of fluorous phases has been exploited in
catalysis,3a reaction acceleration,4 combinatorial chemistry,3b and
organic separation methodology.5 However, structural use of
this phase separation paradigm has been, at best, limited.6,7
Herein, we report the design, chemical synthesis, and
structural characterization of a self-assembling peptide with a
highly fluorinated core based on the coiled coil domain of yeast
Results and Discussion
Design Principles. The coiled coil domain is a wide-spread
structural pattern found in fibrous as well as globular proteins.12
It is estimated that greater than 5% of all open reading frames
in genomes that have been sequenced contain the coiled coil
motif.13 These motifs are well-suited to protein design studies,
as the interactions governing structure are relatively well
understood and represent the simplest case for which compu-
tational tools are sufficiently sophisticated to allow rational
design. Therefore, it is not surprising that coiled coils have been
extensively used in de novo peptide and protein design.14 Coiled
coils self-assemble into oligomers, the primary driving force
being hydrophobic interactions. The sequences exhibit a 4-3
* To whom correspondence should be addressed. Phone: (617) 627-
5651. Fax: (617) 627-3443. E-mail: kkumar01@tufts.edu.
(1) (a) Tanford, C. Science 1978, 200, 1012. (b) Kauzmann, W. AdV.
Protein Chem. 1959, 14, 1-63. (c) Tanford, C. J. Am. Chem. Soc. 1962,
84, 4240.
(2) (a) Makhatadze, G. I.; Privalov, P. L. AdV. Protein Chem. 1995, 47,
307-425. (b) Murphy, K. P.; Privalov, P. L.; Gill, S. J. Science 1990, 247,
559-561. (c) Privalov, P. L.; Tsalkova, T. N. Nature 1979, 280, 693-
696.
(3) (a) Horvath, I. T.; Rabai, J. Science 1994, 266, 72-75. (b) Horvath,
I. T. Acc. Chem. Res. 1998, 31, 641-650. (c) Richter, B.; Spek, A. L.; van
Koten, G.; Deelman, B. J. J. Am. Chem. Soc. 2000, 122, 3945-3951. (d)
Barthel-Rosa, L. P.; Gladysz, J. A. Coord. Chem. ReV. 1999, 192, 587-
605.
(8) (a) Scott, R. L. J. Am. Chem. Soc. 1948, 70, 4090. (b) Hildebrand,
J. H.; Cochran, D. H. F. J. Am. Chem. Soc. 1949, 71, 22. (c) Hildebrand,
J. H.; Scott, R. L. Solubility of Non-Electrolytes, 3rd ed.; Reinhold Publishing
Corporation: New York, NY, 1949.
(9) (a) Imanaka, T.; Shibazaki, M.; Takagi, M. Nature 1986, 324, 695.
(b) Zaks, A.; Klibanov, A. M. Science 1984, 224, 1249. (c) Matsumura,
M.; Yasumura, S.; Aiba, S. Nature 1986, 323, 356.
(4) Myers, K. E.; Kumar, K. J. Am. Chem. Soc. 2000, 122, 12025-
12026.
(5) (a) Studer, A.; Hadida, S.; Ferritto, R.; Kim, S. Y.; Jeger, P.; Wipf,
P.; Curran, D. P. Science 1997, 275, 823-826. (b) Gladysz, J. A. Science
1994, 266, 55. (c) Fish, R. H. Chem. Eur. J. 1999, 5, 1677-80. (d) Curran,
D. P.; Luo, Z. Y. J. Am. Chem. Soc. 1999, 121, 9069-9072. (e) Linclau,
B.; Sing, A. K.; Curran, D. P. J. Org. Chem. 1999, 64, 2835-2842.
(6) (a) Ishikawa, Y.; Kuwahara, H.; Kunitake, T. Chem. Lett. 1989,
1737-1740. (b) Ishikawa, Y.; Kuwahara, H.; Kunitake, T. J. Am. Chem.
Soc. 1994, 116, 5579-5591. (c) Johansson, G.; Percec, V.; Ungar, G.; Zhou,
J. P. Macromolecules 1996, 29, 646-660. (d) Percec, V.; Johansson, G.;
Ungar, G.; Zhou, J. P. J. Am. Chem. Soc. 1996, 118, 9855-9866.
(7) Tang, Y.; Tirrell D. A. Abstracts of Papers; 218th National Meeting
of the American Chemical Society, August 22, 1999; American Chemical
Society: Washington, DC, 1999; Part 2, 416-ORGN.
(10) Marsh, E. N. G. Chem. Biol. 2000, 7, R153-R157.
(11) (a) Marsh, D. Biochem. J. 1996, 315, 345-361. (b) Mingarro, I.;
vonHeijne, G.; Whitley, P. Trends Biotechnol. 1997, 15, 432-437. (c)
Kienker, P. K.; Degrado, W. F.; Lear, J. D. Proc. Natl. Acad. Sci. U.S.A.
1994, 91, 4859-4863. (d) Tuchscherer, G.; Scheibler, L.; Dumy, P.; Mutter,
M. Biopolymers 1998, 47, 63-73. (e) Futaki, S. Biopolymers 1998, 47,
75-81. (f) Lee, S.; Kiyota, T.; Kunitake, T.; Matsumoto, E.; Yamashita,
S.; Anzai, K.; Sugihara, G. Biochemistry 1997, 36, 3782-3791.
(12) Lupas, A. Trends Biochem. Sci. 1996, 21, 375-382.
(13) Newman, J. R. S.; Wolf, E.; Kim, P. S. Proc. Natl. Acad. Sci. U.S.A.
2000, 97, 13203-13208.
10.1021/ja002961j CCC: $20.00 © 2001 American Chemical Society
Published on Web 04/12/2001

4394 J. Am. Chem. Soc., Vol. 123, No. 19, 2001
Bilgic¸er et al.
Scheme 1
hydrophobic repeat (abcdefg)n, with the a and d positions
predominantly occupied by amino acid residues with hydro-
phobic side chains and charged residues occurring frequently
at the e and g positions. Residues at these four positions
participate in interhelical hydrophobic and electrostatic interac-
tions. The solvent-exposed positions b, c, and f are not directly
involved in the self-assembly process and can tolerate a wide
variety of amino acid substitutions. Studies of peptide models
have suggested that four heptad repeats in the sequence are
sufficient to yield a stable dimer.15,23a We hypothesized that
extensive fluorination of the core found in these small proteins
would provide greater driving force for structure formation. The
coiled coil domain of GCN4 is a 33 residue peptide that forms
a parallel homodimer.23b It contains leucine residues in d
positions and valine in a positions with the exception of an
asparagine at position 16. We sought to replace the core residues
by trifluoromethyl-containing amino acid residues. It is well-
known that replacement of CH3 groups with CF3 groups results
in an increase in hydrophobicity due to the low polarizibility
of the fluorine atoms.16,17 This is also evident when one
compares the estimated partial molal heat capacities for the
process of (aqueous) solution18,19 (98 cal mol-1 K-1 for CF4
and 173 cal mol-1 K-1 for C2F6 which are considerably more
positive than the values of 50 cal mol-1 K-1 for CH4 and 72
cal mol-1 K-1 for C2H6). The higher values for the fluorocarbon
gases are consistent with much larger perturbations of water
structure,20 based on the classical analysis of Frank and Evans.21
Furthermore, although the trifluoromethyl group has twice the
molar volume of a methyl group, it can statistically replace the
latter in crystals16b,22 (solid solutions). The designed peptides 1
and 2 (Figure 1) are similar to the coiled coil region of GCN4
with minor modifications.23 Four leucine and three valine
residues in the hydrophobic core of peptide 1 were replaced by
5,5,5-trifluoroleucine and 4,4,4-trifluorovaline respectively to
yield peptide 2.24 In this manner, an R-helical coiled coil dimer
(14) (a) Kohn, W. D.; Kay, C. M.; Sykes, B. D.; Hodges, R. S. J. Am.
Chem. Soc. 1998, 120, 1124-1132. (b) Schneider, J. P.; Lear, J. D.;
DeGrado, W. F. J. Am. Chem. Soc. 1997, 119, 5742-5743. (c) Chao, H.
M.; Houston, M. E.; Grothe, S.; Kay, C. M.; O’Connor-McCourt, M.; Irvin,
R. T.; Hodges, R. S. Biochemistry 1996, 35, 12175-12185. (d) Betz, S.;
Fairman, R.; O’Neil, K.; Lear, J.; Degrado, W. Philos. Trans. R. Soc. London
Ser. B-Biol. Sci. 1995, 348, 81-88.
(15) (a) O’Neil, K. T.; Degrado, W. F. Science 1990, 250, 646-651.
(b) Lau, S. Y. M.; Taneja, A. K.; Hodges, R. S. J. Biol. Chem. 1984, 259,
3253-3261.
(16) (a) Seebach, D. Angew. Chem., Int. Ed. Engl. 1990, 29, 1320. (b)
Seebach, D.; Renaud, P.; Schweizer, W. B.; Zueger, M. F. HelV. Chim.
Acta 1984, 67, 1843.
(23) (a) O’Shea, E. K.; Rutkowski, R.; Kim, P. S. Science 1989, 243,
538-542. (b) O’Shea, E. K.; Klemm, J. D.; Kim, P. S.; Alber, T. Science
1991, 254, 539-544. (c) Harbury, P. B.; Zhang, T.; Kim, P. S.; Alber, T.
Science 1993, 262, 1401-1407. (d) Kenar, K. T.; Garciamoreno, B.; Freire,
E. Protein Sci. 1995, 4, 1934-1938. (e) Hodges, R. S.; Semchuk, P. D.;
Taneja, A. K.; Kay, C. M.; Parker, J. M. R.; Mant, C. T. Pept. Res. 1988,
1, 19-30.
(24) Amino acids were mixtures of N-Boc-(2S,4S)- and (2S,4R)-5,5,5-
trifluoroleucine (d position) or N-Boc-(2S,3S)- and (2S,3R)-4,4,4-trifluo-
rovaline (a position).
(17) Resnati, G. Tetrahedron 1993, 49, 9385-9445.
(18) Wen, W.-Y.; Muccitelli, J. A. J. Solution Chem. 1979, 8, 225.
(19) Wilhelm, E.; Battino, R.; Wilcock, R. J. Chem. ReV. 1977, 77, 219-
62.
(20) (a) Mukerjee, P.; Korematsu, M.; Okawauchi, M.; Sugihara, G. J.
Phys. Chem. 1985, 89, 5308. (b) Sugihara, G.; Mukerjee, P. J. Phys. Chem.
1981, 85, 1612.
(21) Frank, H. S.; Evans, M. W. J. Chem. Phys. 1945, 13, 507.
(22) Hagan, D. O.; Rzepa, H. S. Chem. Commun. 1997, 645.

A Coiled Coil with a Fluorous Core
J. Am. Chem. Soc., Vol. 123, No. 19, 2001 4395
Figure 1. Helical wheel diagram and peptide sequences 1 and 2.
Abbreviations, Ar ) 4-acetamidobenzoic acid, Ac ) acetyl, C†)
acetamidocysteine. The asterisk indicates unresolved stereochemistry.
was obtained with a very highly fluorinated core shielding 14
trifluoromethyl groups from aqueous solvent. Since the popula-
tion of multiple oligomerization states by self-assembly is
possible in these systems, a design feature that restricts the
oligomerization to a dimeric state was needed to facilitate
biophysical studies. Earlier studies have established that a single
polar residue in the hydrophobic core can impart structural
uniqueness in designed coiled coil systems by formation of an
inter-strand hydrogen bond.14d,25 Therefore, we retained the
single conserved asparagine residue (a position) in the core to
dictate oligomerization and orientation specificity.23c,26
Figure 2. [A] Reversed phase HPLC trace (Vydac C18 column; 22 ×
250 mm, 300 Å size, 10-15 µm) showing the fluoro peptide before
final purification. Both peaks correspond to fluoro peptide 2. The peaks
were separated and the studies reported here were conducted exclusively
with peak 2. [B] Analytical run of the purified peak 2 eluting at 16.65
min used in all structural characterization experiments. Gradient for
[B]: 32% to 50% acetonitrile/H2O/0.1% TFA at 2.5 mL/min (over 20
min). Detection: absorbance at 230 nm.
Synthesis. Fluorine-containing building blocks for peptide
synthesis were resolved by enzymatic deacylation of trifluo-
romethyl-containing N-acetyl amino acids with porcine kidney
acylase I to deliver S stereochemistry at the CR position.27 The
stereochemical purity of the enzyme-resolved amino acids was
ascertained by chiral HPLC.28 All four diastereomers of 4,4,4-
trifluorovaline and 5,5,5-trifluoroleucine were separable. Only
those diastereomers that had an S configuration at the CR
position could be detected after enzymatic resolution, judged
by comparison to the mixture of all four diastereomers in each
case. Peptides were assembled by manual solid-phase peptide
synthesis by using the in situ neutralization protocol29 on
commercially available methylbenzhydrylamine [MBHA, co-
poly(styrene, 1% DVB)] resin, using t-Boc chemistry, and
purified by a combination of reversed-phase (C4 and C18) HPLC
(Figure 2) and ion exchange chromatography. While the CR
position was exclusively S stereochemistry for all residues, the
secondary stereocenters at Câ (trifluorovaline) and Cγ (trifluo-
roleucine) positions were left unresolved. Given the presence
of two diastereomeric amino acids at each fluorinated core
residue that was substituted, peptide 2 is a mixture of several
diastereomers.30
Structural and Biophysical Characterization. The MALDI-
TOF mass spectrum of the purified fluorinated peptide 2 showed
a single peak at 4167.3 Da (calculated, 4165.9 Da) consistent
with the proposed structure (Figure 3). Circular dichroism
spectra of purified peptides 1 and 2 at 5 °C showed that both
peptides are extremely R-helical in aqueous solution at pH 7.40
in the low micromolar concentration regime with characteristic
minima at 222 and 208 nm. The magnitude of the minima at
222 and 208 nm for 2 is much larger than those for 1, indicating
a much higher level of helicity in the peptide. Both the
hydrocarbon peptide 1 and fluoro peptide 2 exhibit cooperative
unfolding transitions as a function of temperature (Figure 4).
Thus, it would appear that both trifluoroleucine and trifluo-
rovaline residues are incorporated in the hydrophobic core
without large structural distortions. Apparent molecular masses
of the self-assembled coiled coils in solution were determined
by sedimentation equilibrium experiments. Data obtained were
fit to an ideal single species model plot of ln(absorbance) versus
radial distance squared. Both peptides 1 and 2 sediment as
dimers in the 5-30 µM concentration range (Figure 5), thus
validating the comparison of melting temperatures. Dimer
melting temperatures obtained by monitoring the molar ellip-
(25) Lumb, K. J.; Kim, P. S. Biochemistry 1995, 34, 8642-8648.
(26) (a) O’Shea, E. K.; Lumb, K. J.; Kim, P. S. Curr. Biol. 1993, 3,
658-667. (b) Gonzalez, L.; Plecs, J. J.; Alber, T. Nat. Struct. Biol. 1996,
3, 510-515.
(27) (a) Chenault, H. K.; Dahmer, J.; Whitesides, G. M. J. Am. Chem.
Soc. 1989, 111, 6354-64. (b) Houng, J.-Y.; Wu, M.-L.; Chen, S.-T.
Chirality 1996, 8, 418-422.
(28) CROWNPAK(+)-CR column, Daicel Chemical Industries.
(29) Schnolzer, M.; Alewood, P.; Jones, A.; Alewood, D.; Kent, S. B.
Int. J. Pept. Protein Res. 1992, 40, 180-193.
(30) A maximum of 27 ) 128 diastereomers could have been produced
as seven of the core residues have been replaced by fluorinated amino acids.

4396 J. Am. Chem. Soc., Vol. 123, No. 19, 2001
Bilgic¸er et al.
Figure 3. MALDI-TOF MS of purified fluorinated peptide 2 (MWcalc
) 4165.9 Da, found ) 4167.3 Da).
Figure 5. Sedimentation equilibrium trace for peptide 2 (14 µM in
PBS at 10 °C at pH 7.40) confirming a dimeric species in the
concentration regime of the CD experiments. Likewise, control peptide
1 sediments as a dimer (Supporting Information).
The stability of the peptides toward chaotropic reagents was
determined by monitoring the molar ellipticity at 222 nm as a
function of Gdn‚HCl concentration (Figure 6). The apparent
free energy of unfolding as calculated from these experiments
is modestly larger for peptide 2 than peptide 1 (∼1.0 kcal/mol
at 5 °C, Table 1). The sensitivity of peptides to the denaturant
value m, given by d(∆G)/d[Gdn‚HCl], is believed to be
proportional to the amount of hydrophobic surface area exposed
upon denaturation.31 The m value for fluoro peptide 2 extracted
from nonlinear fitting is 3.08 ( 0.22 kcal mol-1 M-1, which is
much larger than the 2.40 ( 0.24 kcal mol-1 M-1 obtained for
1. This is consistent with the burial of 14 voluminous trifluo-
romethyl instead of methyl groups. Once some of the structural
features of packing in the fluorinated cores are elucidated in
greater detail, optimization of the packing interactions in
fluorinated cores should provide additional stability. On the basis
of the denaturation and thermal stability data, we conclude that
the additional stability of the fluorinated peptide is derived from
sequestering the more hydrophobic trifluoromethyl groups from
aqueous solvent.
The helix stabilizing effects of certain alcohol solvents,
especially 2,2,2-trifluoroethanol (TFE), have been employed
frequently in the study of equilibrium32 and kinetic33 studies of
protein folding. Given the large number of trifluoromethyl
groups buried in the hydrophobic core of peptide 2, we sought
to examine the effect of TFE on peptide 2 in relation to the
effects observable on 1. The data are shown in Figure 7 where
Figure 4. [A] Circular dichroism spectra of peptides 1 ([) and 2 (O)
in PBS (10 mM phosphate, 137 mM NaCl, 2.7 mM KCl, pH 7.40) at
5 °C ([1] ) [2] ) 5 µM). The minima at 222 and 208 nm and the
large negative [θ]222 indicate that both peptides are highly R-helical.
[B] Thermal denaturation of peptides 1 and 2 monitored by CD ([θ]222
)
yielded a Tm of 47 (1) and 62 °C (2). The Tm values were determined
from the minima of the first derivative of [θ]222 with T-1, where T is
in K. Melting curves were reversible (folding and unfolding curves
were superimposable, with g97% of the initial signal regained upon
cooling).
the percentage decrease in molar ellipticity at 222 nm ([θ]222
)
(31) (a) Schellman, J. A. Biopolymers 1978, 17, 1305-22. (b) Shortle,
D. FASEB J. 1996, 10, 27-34.
(32) (a) Jasanoff, A.; Fersht, A. R. Biochemistry 1994, 33, 2129-35.
(b) Schonbrunner, N.; Wey, J.; Engels, J.; Georg, H.; Kiefhaber, T. J. Mol.
Biol. 1996, 260, 432-35.
ticity at 222 nm as a function of temperature show that peptide
1
2 is more stable than peptide 1 by ∼15 °C (Tm ) 47 ( 1 °C,
(33) (a) Hamada, D.; Goto, Y. J. Mol. Biol. 1997, 269, 479-87. (b) Lu,
H.; Buck, M.; Radford, S. E.; Dobson, C. M. J. Mol. Biol. 1997, 265, 112-
7.
2
Tm ) 62 ( 1 °C), establishing the higher thermal stability of
the fluorinated core in aqueous solution (Figure 4).

A Coiled Coil with a Fluorous Core
J. Am. Chem. Soc., Vol. 123, No. 19, 2001 4397
Figure 6. Guanidinium denaturation of control peptide 1 (5 µM, [Panel A]) and fluorinated peptide 2 (5 µM, [Panel B]) at 5 °C (pH 7.40)
monitored by the circular dichroism at 222 nm. The solid line is a curve generated from the parameters listed in Table 1.
Table 1. Fitting Parameters for Gdn‚HCl Denaturation
sufficient to explain our observations given the higher intrinsic
helical content of peptide 2. A second possibility is that TFE
may effectively solvate the fluorinated residues, thereby disrupt-
ing dimerization. In this case, the competing effects of dena-
turation of the fluorous core (decrease in helicity) and amide
backbone desolvation (increase in helicity) with increasing TFE
are responsible for the smaller increase observed for 2. Our data
do not allow us to distinguish between these mechanisms. It
would be informative to explore the effects of fluorinated alcohol
cosolvents in systems with even more highly fluorinated cores.
Experimentsa
peptide
∆GH° O (5 °C) (kcal mol-1
)
m (kcal mol-1 M-1
)
2
1
2
10.50 ( 0.42
11.43 ( 0.34
2.40 ( 0.24
3.08 ( 0.22
a The parameters were obtained by nonlinear least-squares fitting of
the data in Figure 6.
Conclusions
We have outlined a strategy for the design and construction
of peptides with highly fluorinated interiors by solid-phase
peptide synthesis. Characterization of such peptides by a wide
variety of biophysical techniques reveals higher stability
compared to hydrocarbon counterparts. The intriguing possibility
of creating a binary patterning scheme for protein design based
on hydrocarbon/fluorocarbon phase separation seems within
reach. With the recent introduction of protein ligation techniques,
the size range of proteins accessible by chemical synthesis and
semisynthesis has been greatly expanded.35 The site-specific
incorporation of fluorinated amino acids onto hydrophobic folds
of larger proteins is therefore feasible with current technology.
In our laboratories, further biophysical studies aimed at probing
the nature of fluorous interactions in the core, synthesis, and
characterization of coiled coils with hexafluoroleucine and
fluorinated membrane spanning peptides are currently in
progress. With this study, we introduce a new paradigm for
molecular design, one based on orthogonal solubility properties
of liquid phases.
Figure 7. Effect of 2,2,2-trifluoroethanol (TFE) on the helical content
of peptide 1 (5 µM, b) and peptide 2 (5 µM, [) monitored by circular
dichroism at 222 nm.
in circular dichroism experiments is plotted with TFE concen-
tration. The percentage increase in helicity is much less
pronounced for 2 than for control peptide 1. There are two
possible explanations. The first possibility is that this difference
is due to the intrinsically higher helical content of the fluoro
peptide 2 at 0% TFE. At low concentrations, the helix-enhancing
ability of TFE in coiled coil systems has been attributed to an
increase in the structure of the binary alcohol/water solvent and
a concomitant increase in the energetic cost of polypeptide
backbone solvation.34 This view is consistent with minimal
involvement of the side chains of the hydrophobic core residues
in the mechanism of helical induction by TFE, and could be
Experimental Methods
Chemicals. Acetonitrile (optima grade), dichloromethane (ACS
grade), and dimethylformamide (DMF, sequencing grade) were pur-
chased from Fisher Scientific and used without further purification.
N,N-Diisopropylethylamine (DIEA, Aldrich, biotech grade), trifluoro-
acetic acid (TFA, New Jersey Halocarbon), guanidine hydrochloride
(35) (a) Dawson, P. E.; Muir, T. W.; Clarklewis, I.; Kent, S. B. H. Science
1994, 266, 776-779. (b) Blaschke, U. K.; Silberstein, J.; Muir, T. W.
Methods Enzymol. 2000, 328, 478-496. (c) Ayers, B.; Blaschke, U. K.;
Camarero, J. A.; Cotton, G. J.; Holford, M.; Muir, T. W. Biopolymers 1999,
51, 343-354.
(34) Kentsis, A.; Sosnick, T. R. Biochemistry 1998, 37, 14613-14622.

4398 J. Am. Chem. Soc., Vol. 123, No. 19, 2001
Bilgic¸er et al.
(Gdn‚HCl, Fisher Scientific, electrophoresis grade), 2-(1H-benzotriazol-
1-yl)-1,1,3,3-tetramethyluronium hexafluorophosphate (HBTU, Quan-
tum Biotechnologies), 4-methylbenzhydrylamine resin (MBHA, Nov-
abiochem), anisole (anhydrous, Aldrich), and trifluoroethanol (99%,
Acros) were used without further purification. N-Boc-R-amino acids
were used as obtained from Novabiochem, Advanced Chemtech or
American Peptide Company. Hydrogen fluoride was purchased from
Matheson Gas.
Preparation of Fluorinated Amino Acids for Peptide Synthesis.
DL-Trifluoroleucine and DL-trifluorovaline were purchased from Oak-
wood Chemicals or prepared by literature procedures.36 The N-acetyl
derivatives of both 5,5,5-trifluoroleucine36c and 4,4,4-trifluorovaline36b
are known. The N-acetyl amino acids were then deacylated by porcine
kidney acylase I, using a modification of a known literature procedure.27a
Details of acetylation and deacylation reactions are given in the
Supporting Information. The hydrochloride salts of the amino acids
with S stereochemistry at CR were t-Boc protected on the amino
terminus by using a mild protection method37 and after purification
were used directly in solid-phase peptide synthesis.
similar Tm values. However, in sedimentation equilibrium experiments,
the residuals were large and non-random for material obtained from
this peak (even after repeated purification), preventing us from assigning
a unique oligomerization state.
Circular Dichroism. CD spectra were obtained on a JASCO J-715
spectropolarimeter fitted with a PTC-423S single position Peltier
temperature controller. Buffer conditions were 10 mM phosphate (pH
7.40), 137 mM NaCl, 2.7 mM KCl unless otherwise noted. The
spectrometer was calibrated with an aqueous solution of recrystallized
d10-(+)-camphorsulfonic acid at 290.5 nm. The concentrations of the
peptide stock solutions were determined by amino acid analysis (average
of 3 runs) or by integration of the tyrosine absorbance on analytical
HPLC relative to standard N-acetyltyrosinamide (ꢀ ) 1490 M-1 cm-1).
Mean residue ellipticities (deg cm2 dmol-1) were calculated by using
the relation:
[θ] ) θobs × MRW/10‚l‚c
(1)
where θobs is the measured signal (ellipticity) in millidegrees, l is the
optical path length of the cell in cm, c is concentration of the peptide
in mg/mL, and MRW is the mean residue molecular weight (molecular
weight of the peptide divided by the number of residues).
Thermal denaturation studies were carried out at the concentrations
indicated by monitoring the change in [θ]222 as a function of temper-
ature. Temperature was increased in steps of 0.5 °C with an intervening
equilibration time of 120 s. Data were collected over 16 s per point.
The Tm was determined from the minima of the first derivative of [θ]222
with T-1, where T is in K. All thermal melts were reversible with g97%
of the starting signal regained upon cooling.
N-(t-Boc)-(2S,4S),(2S,4R)-trifluoroleucine (5). (2S,4S),(2S,4R)-
5,5,5-trifluoroleucine HCl salt (1.00 g, 4.5 mmol), NaHCO3 (1.14 g,
13.5 mmol), and (Boc)2O (1.08 g, 5.0 mmol) were suspended in 22
mL of MeOH and sonicated for 1 h. The reaction mixture was filtered.
Rotary evaporation of the filtrate gave a white solid. Further purification
of this material by C18 reversed phase column chromatography (30%
CH3CN/H2O, 0.25% HOAc eluant) provided pure N-t-Boc-5,5,5-
1
(2S,4S),(2S,4R)-trifluoroleucine in 80% yield (1.05 g). H NMR (300
MHz, D2O) [mixture of two diastereomers] δ 1.13 (d, 3H, J ) 6.8
Hz), 1.42 (s, 9H), 1.58-2.13 (m, 2H), 2.30-2.48 (m, 1H), 3.90-4.02
(m, 1H). GC-MS (CI, CH4): calcd for C11H18F3NO4 285.28, found 286
([M + 1]+).
Gdn‚HCl Denaturation. Denaturation experiments were carried out
assuming a two-state unfolding transition.39
N-(t-Boc)-(2S,3S),(2S,3R)-trifluorovaline (6). (2S,3S),(2S,3R)-Tri-
fluorovaline hydrochloride salt (1.00 g, 4.8 mmol) was t-Boc protected
following the protocol given for trifluoroleucine above to yield the
desired product in 82% yield (1.07 g). 1H NMR (300 MHz, D2O) [23%
(2S,3S); 77% (2S,3R)] δ [1.06 (d, J ) 7.2 Hz), (2S,3R), 1.16 (d, J )
7.1 Hz), (2S,3S)] (3H), 1.41 (s, 9H), 2.90-2.95 (m, 1H), [4.12 (d, J )
5.1 Hz), (2S,3S), 4.43 (d, J ) 2.6 Hz), (2S,3R)] (1H). GC-MS (CI,
CH4): calcd for C10H16F3NO4 271.25, found 272 ([M + 1]+).
Peptide Synthesis. Peptides were prepared by using the N-tert-
butyloxycarbonyl (t-Boc) amino acid derivatives for Merrifield manual
solid-phase synthesis (MBHA resin), using the in situ neutralization/
HBTU protocol typically on a 0.5 mmol scale.29 N-R-Boc-R-S-amino
acids were used with the following side chain protecting groups: Arg-
(Tos), Asp(OBzl), Asn(Xan), Gln(Xan), Cys(Acm), Glu(OBzl), Lys-
(2-Cl-Z), Ser(Bzl), and Tyr(2-Br-Z). Peptide coupling reactions were
carried out with a 4-fold excess (2.0 mmol) of activated amino acid
for at least 15 min. Peptides were cleaved from the resin by using high
HF conditions (90% anhydrous HF/10% anisole at 0 °C for 1.5 h).38
Purification. Peptides were desalted by reversed phase HPLC
[Vydac C4 column with a 30 min linear gradient of 20-40%
acetonitrile/H2O/0.1% TFA at 8.0 mL/min]. Further purification of 2
was carried out by ion exchange chromatography by using reverse
organic conditions [PolyWAX A column from PolyLC, Inc., Columbia,
MD (200 × 9.4 mm, 5 µm, 300 Å)]. Following purification by ion-
exchange chromatography, 2 was again subjected to purification and
desalting by reversed phase HPLC (Vydac C18 column, 22 mm × 250
mm, 300 Å size, 10-15 µm). There were two peaks in the reversed
phase that could be separated (Figure 2). The studies reported in this
paper were all carried out with the peak eluting at 16.65 min [Figure
2B]. The other peak eluted at 15.69 min (under the conditions of Figure
2B), gave similar CD spectra at 5 µM concentration, and also shows
D h 2M
where Kd ) [M]2/[D] (M ) monomer and D ) dimer). Given that the
total peptide concentration, Pt, in terms of monomers is Pt ) 2[D] +
[M], the CD signal Yobs can be described in terms of folded and unfolded
baselines, Yfol and Yunfol, respectively, by the following expression.40
K
x
2 + 8KdPt - K
4Pt
d
Yobs ) (Yunfol - Yfol)
d + Yfol
(2)
Furthermore, Kd can be expressed in terms of the free energy of
unfolding,
Kd ) exp(-∆G°/RT)
(3)
and assuming that the apparent free energy difference between folded
dimer and unfolded monomer states is linearly dependent on the Gdn‚
HCl concentration, ∆G° can be written as:
∆G° ) ∆GH° O - m[Gdn‚HCl]
(4)
2
where ∆G° is the free energy difference in the absence of denatur-
H2O
ant and m is the dependency of the unfolding transition with respect to
the concentration of Gdn‚HCl.41 The data were fit for two parameters,
namely ∆G° and m, by nonlinear least-squares fitting (Kalieda-
H2O
Graph v3.5).
Analytical Ultracentrifugation. Apparent molecular masses were
determined by sedimentation equilibrium on a Beckman XL-A ultra-
centrifuge. Loading peptide concentrations were 5-30 µM in 10 mM
phosphate (pH 7.40), 137 mM NaCl, 2.7 mM KCl. The samples were
centrifuged at 32 000 and 26 000 rpm for 18 h at 10 °C before
absorbance scans were performed. Data obtained at 10 °C were fit
globally to an equation that describes the sedimentation of a homoge-
(36) (a) Rennert, O. M.; Anker, H. S. Biochemistry 1963, 2, 471. (b)
Ojima, I.; Kato, K.; Nakahashi, K. J. Org. Chem. 1989, 54, 4511-22. (c)
Tsushima, T.; Kawada, K.; Ishihara, S.; Uchida, N.; Shiratori, O.; Higaki,
J.; Hirata, M. Tetrahedron 1988, 44, 5375. (d) Weinges, K.; Kromm, E.
Liebigs Ann. Chem. 1985, 90-102.
(37) Einhorn, J.; Einhorn, C.; Luche, J.-L. Synlett 1991, 37-38.
(38) Tam J. P.; Merrifield, R. B. In The Peptides; Udenfriend, S.,
Meienhofer, J., Eds.; Academic Press Inc.: New York, NY, 1987; Vol. 9,
p 185.
(39) (a) Tanford, C. AdV. Protein Chem. 1970, 24, 1. (b) Pace, C. N.
Methods Enzymol. 1986, 131, 266.
(40) Zhu, H.; Celinski, S. A.; Scholtz, J. M.; Hu, J. C. J. Mol. Biol.
2000, 300, 1379-1389.
(41) (a) Becktel, W. J.; Schellman, J. A. Biopolymers 1987, 26, 1859-
1877. (b) Matthews, B. W. Biochemistry 1987, 26, 6885-6888.

A Coiled Coil with a Fluorous Core
J. Am. Chem. Soc., Vol. 123, No. 19, 2001 4399
neous species.
Committee (Tufts), and Tufts University for support. We thank
Dr. David H. Lee for help with the analytical ultracentrifugation
experiments and for helpful discussions.
2
Abs ) A′ exp(H × M[x2 - x0 ]) + B
(5)
Supporting Information Available: Experimental details,
HPLC data, MALDI mass spectrum, and sedimentation equi-
librium trace for peptide 1 (PDF). This material is available
free of charge via the Internet at http://pubs.acs.org.
where Abs ) absorbance at radius x, A′ ) absorbance at reference
radius x0, H ) (1 - VhF)ω2/2RT, Vh ) partial specific volume ) 0.75531
mL/g, F ) density of solvent ) 1.0017 g/mL, ω ) angular velocity in
radians/s, M ) apparent molecular weight, and B ) solvent absorbance
(blank). We estimated partial specific volume of the peptides using
amino acid composition, substituting leucine and valine for 5,5,5-
trifluoroleucine and 4,4,4-trifluorovaline respectively in peptide 2 for
lack of available data.42
JA002961J
(42) (a) Cohn, E. J.; Edsall, J. T. Proteins, Amino Acids and Peptides as
Ions and Dipolar Ions; Reinhold: New York, 1943. (b) Laue, T. M.; Shah,
B. D.; Ridgeway, T. M.; Pelletier, S. L. In Analytical Ultracentrifugation
in Biochemistry and Polymer Science; Harding, S. E., Rowe, A. J., Horton,
H. C., Eds.; The Royal Society of Chemistry: Cambridge, England, 1992;
p 90.
(43) Note Added in Proof: While this manuscript was under review,
another study detailing a similar approach to ours has been reported: Tang,
Y.; Ghirlanda, G.; Vaidehi, N.; Kua, J.; Mainz, D. T.; Goddard III, W. A.;
DeGrado, W. F.; Tirrell, D. A. Biochemistry 2001, 40, 2790-2796.
Acknowledgment. The authors thank the Petroleum Re-
search Fund, administered by the American Chemical Society
(ACS-PRF No. 33807-G4), the Faculty Research Awards