Conformational preferences of RNase A C-peptide derivatives containing a highly constrained analogue of phenylalanine

Article abstract of DOI:10.1021/ja981153d

Both enantiomers of a highly constrained derivative of phenylalanine, FiFi, were prepared in optically pure form. Studies were performed to elucidate the effects of substituting this amino acid for phenylalanine in RN-24, a derivative of the RNase A C-pep

Full text of DOI:10.1021/ja981153d

J. Am. Chem. Soc. 1998, 120, 9435-9443
9435
Conformational Preferences of RNase A C-Peptide Derivatives
Containing a Highly Constrained Analogue of Phenylalanine
Destardi Moye-Sherman,† Song Jin,† Inhye Ham,† Dongyeol Lim,† J. Martin Scholtz,‡ and
Kevin Burgess*,†
Contribution from the Chemistry Department and Department of Medicinal Biochemistry and Genetics,
Texas A & M UniVersity, College Station, Texas 77843-1114
ReceiVed April 6, 1998
Abstract: Both enantiomers of a highly constrained derivative of phenylalanine, FiFi, were prepared in optically
pure form. Studies were performed to elucidate the effects of substituting this amino acid for phenylalanine
in RN-24, a derivative of the RNase A C-peptide. Thus RN-24, and the analogues 9 and 10, in which Phe-8
was replaced by each of the enantiomers of FiFi, were prepared. Comparative circular dichroism (CD)
experiments indicated relative tendencies to adopt helical structures, and variable-temperature CD studies showed
the relative ease with which these conformations were lost at elevated temperatures. These observations were
rationalized via computer-assisted molecular modeling, which showed that the phenyl groups of (R,R)-FiFi in
the peptidomimetic 9 can be accommodated via distortion of the helical conformations and that such distorted
conformations persist as the temperature is increased. Conversely, intolerable contacts occur in an analogous
conformation of the (S,S)-FiFi peptidomimetic 10 and these preclude helicity. Consistent with these observations,
molecular dynamics studies of these peptidomimetics at 276 K indicate that helical conformations of 9 and
RN-24 are observable under conditions in which the analogous conformation of 10 is lost. Overall, these
studies demonstrate that cyclopropane amino acids can be used to enforce elements of secondary structure
(albeit distorted) or to preclude them altogether.
Introduction
formation of helical secondary structures for small peptides in
solution.4 First, the S-peptide,5 the C-peptide,6 and RN-242 have
significantly populated helical conformations at temperatures
near 3 °C. At the time when these observations were originally
made, it was unusual to find evidence of helicity in such small
peptides, though subsequently these findings have inspired the
design of other helical small-peptide systems.7-10 Second, solid-
state crystal structures have been determined for RNase A11 and
the S-protein/S-peptide complex;12 for RN-24, a solution-state
structure has been deduced from NMR.13 Consequently,
accurate physical models exist for structures of RNase A and
its derivatives. For the S-peptide, these static models have been
supported by computer-aided molecular simulations illustrating
the dynamic conformational behavior of this system.14 Finally,
association complexes of the S-protein and the S-peptide have
catalytic activities that are comparable with those of RNase A.15
This is remarkable because the S-protein alone is inactive.
Bovine pancreatic RNase A (Figure 1a) is a protein that
mediates the hydrolysis of RNA. Fragments of RNase A are
available via several enzymatic and chemical routes. Proteolytic
degradation of RNase A with subtilisin, for instance, gives a
C-terminal protein fragment, “the S-protein” (RNase A residues
21-124), and a N-terminal peptide called “the S-peptide”
(RNase A residues 1-20).1 The intermolecular complex formed
between the S-protein and the S-peptide is “RNase S.” A shorter
peptide, “the C-peptide lactone,” can also be formed from the
S-peptide via cleavage with CNBr. Moreover, synthetic peptides
based on the C-peptide lactone sequence have also been prepared
and studied, notably the system named RN-24.2
(4) Degrado, W. F. In AdVances in Protein Chemistry: Design of Peptides
and Proteins; Academic Press: New York, 1988; pp 51-124.
(5) Kim, P. S.; Baldwin, R. L. Nature 1984, 307, 329.
(6) Brown, J. E.; Klee, W. A. Biochemistry 1971, 10, 470.
(7) Marqusee, S.; Robbins, V. H.; Baldwin, R. L. Proc. Natl. Acad. Sci.
U.S.A. 1989, 86, 5286.
(8) Lyu, P. C.; Liff, M. I.; Marky, L. A.; Kallenbach, N. R. Science
1990, 250, 669.
(9) Merutka, G.; Stellwagen, E. Biochemistry 1990, 29, 894.
(10) Scholtz, J. M.; Qian, H.; York, E. J.; Stewart, J. M.; Baldwin, R. L.
Biopolymers 1991, 31, 1463.
(11) Birdsall, D. L.; McPherson, A. J. Biol. Chem. 1992, 267, 22230.
(12) Kim, E. E.; Varadarajan, R.; Wyckoff, H. W.; Richards, F. M.
Biochemistry 1992, 31, 12304.
(13) Osterhout, J. J.; Baldwin, R. L.; York, E. J.; Stewart, J. M.; Dyson,
H. J.; Wright, P. E. Biochemistry 1989, 28, 7059.
(14) Tirado-Rives, J.; Jorgensen, W. L. Biochemistry 1991, 30, 3864.
(15) Hofmann, K.; Finn, F.; Haas, W.; Smithers, M. J.; Wolman, Y.;
Yanaihara, N. J. Am. Chem. Soc. 1963, 85, 833.
Several characteristics of RNase A derivatives make them
prototypical models for studies in protein-folding3 and for
* Address correspondence to this author at Chemistry Department, Texas
A&M University, P.O. Box 300012, College Station, TX 77842. Tel: 409
845 4345 (or 1847). Fax: 409 845 8839. E-mail: burgess@chemvx.tamu.edu
or burgess@chemvx.chem.tamu.edu.
† Chemistry Department.
‡ Department of Medicinal Chemistry and Genetics.
(1) Richards, F. M.; Vithayathil, P. J. J. Biol. Chem. 1959, 234, 1459.
(2) Shoemaker, K. R.; Kim, P. S.; York, E. J.; Stewart, J. M.; Baldwin,
R. L. Nature 1987, 326, 563.
(3) Baldwin, R. L. TIBS 1989, 14, 291.
S0002-7863(98)01153-6 CCC: $15.00 © 1998 American Chemical Society
Published on Web 09/02/1998

9436 J. Am. Chem. Soc., Vol. 120, No. 37, 1998
Moye-Sherman et al.
Figure 1. (a) RNase A; (b) RN-24 after 300-ps molecular dynamics simulation; (c) peptidomimetic 9 before molecular dynamics simulation; (d)
side view of peptidomimetic 9 after molecular dynamics simulation; (e) top view of peptidomimetic 9 after molecular dynamics simulation; (f)
peptidomimetic 10 before molecular dynamics simulation; (g) peptidomimetic 10 after molecular dynamics simulation.
However, the S-protein enhances the disposition of the S-peptide
catalytic activities is that RNase A and the S-protein have similar
tertiary structures.
to adopt helical conformations,16-18 so the inference of the

Conformations of RNase C-Peptide DeriVatiVes
J. Am. Chem. Soc., Vol. 120, No. 37, 1998 9437
One of the consequences of research on the S-peptide, the
C-peptide, and RN-24 is that the interactions that stabilize helical
conformations in these systems are well understood. The most
pertinent observation with respect to the work to be described
here is that considerable evidence indicates a perpendicular
π-interaction between Phe-8 and His+-12.19-22 Similar interac-
tions have also been observed in some other systems.23 Another
crucial helix-stabilizing interaction in C-peptide analogues is a
salt bridge between Glu-2 and Arg-10; the importance of this
has been confirmed via experiments in media of variable
pH.19,24,25 Finally, the presence of a Glu- residue near the
N-terminus and a His+ residue near the C-terminus has been
postulated to be significant,2,23,26-28 consistent with the generally
accepted notion that helix dipoles are stabilized by negative
charges at the N-terminus and positive charges at the C-
terminus.29
rapeptides, for instance, but it is more amenable to conforma-
tional perturbations than is a cyclic hexapeptide.
The specific approach described in this paper is to substitute
the two enantiomers of the novel cyclopropane amino acid FiFi
for Phe-8 in RN-24. The relevance of Phe-8 substitutions in
RN-24 is evident from the postulated importance of the
perpendicular π-stacking Phe-8/His+-12 interaction, but the
choice of the new cyclopropyl amino acid FiFi warrants
explanation. Our rationale for this approach is as follows. The
purpose of using sterically constrained amino acid analogues
is that restricted rotation must enhance the probability that some
conformations will be populated while others are less accessible.
The nature of the critical interactions depends on the structures
studied. For example, compared with amino acid I, the
R-methyl substituent of II would be expected to exert influences
of similar magnitudes on rotations about the φ, ψ, and ø1 bonds.
Conversely, a cis-substituent on a cyclopropane ring, as in III,
locks ø1, which has a very significant influence on φ but
relatively minor effects on ψ. For the isomer IV, however,
locking ø1 impacts ψ most and φ least. A trans-disubstituted
cyclopropane amino acid of type V imposes relatively severe
constraints on φ and ψ simultaneously. Amino acids of this
type, therefore, should have the most pronounced effects on
secondary structure in the series and may provide clues to the
possible influence of the monosubstituted derivatives III and
IV.
Here we describe the first set of data on the effects of
substituting cyclopropyl amino acids into an RN-24 analogue.
This study has two goals: specifically, further elucidation of
the effects of cyclopropane amino acids on secondary structures,
and understanding how the secondary structure of C-peptide
analogues may be deliberately and systematically perturbed.
With respect to the first goal, RN-24 is significantly different
from other model systems that have been used.30-36 This is
important because there is no ideal system for studying
conformational biases of cyclopropane amino acids, so conclu-
sions must be drawn from several different systems. Short linear
peptides (e.g., 3-5 residues) yield useful data, but interpretation
is complicated by ill-defined, random-coil conformational states
that are hard to identify, even after substitution with a restricted
amino acid.30-35 Conversely, cyclic peptides tend to have better-
defined conformational biases, but the effects of replacing a
natural amino acid with a more-hindered one are obscured by
the constraints of peptide-cyclization.36 RN-24, a 13-mer, has
characteristics that are midway between these extremes. It has
more-pronounced conformational preferences than linear tet-
(16) Labhardt, A. M. Biopolymers 1981, 20, 1459.
(17) Labhardt, A. M. J. Mol. Biol. 1982, 157, 331.
(18) Connelly, P. R.; Varadarajan, R.; Sturtevant, J. M.; Richards, F.
M. Biochemistry 1990, 29, 6108.
(19) Rico, M.; Santoro, J.; Bermejo, F. J.; Herranz, J.; Nieto, J. L.;
Gallego, E.; Jiminez, M. A. Biopolymers 1986, 25, 1031.
(20) Shoemaker, K. R.; Fairman, R.; Kim, P. S.; York, E. J.; Stewart, J.
M.; Baldwin, R. L. Cold Spring Harbor Symp. Quant. Biol. 1987, 52, 391.
(21) Shoemaker, K. R.; Fairman, R.; Schultz, D. A.; Robertson, A. D.;
York, E. J.; Stewart, J. M.; Baldwin, R. L. Biopolymers 1990, 29, 1.
(22) Dadlez, M.; Bierzynski, A.; Godzik, A.; Sobocinska, M.; Kuprysze-
wski, G. Biophys. Chem. 1988, 31, 175.
To the best of our knowledge, there has been no previous
synthesis of the structure we call FiFi, even as a racemate.
Syntheses of dimethyl-substituted cyclopropane amino acids
have been reported, via routes that do37-41 and do not42 involve
elements of stereocontrol. An asymmetric synthesis of a gem-
dimethyl cyclopropane amino acid has also been reported.39,43
(23) Loewenthal, R.; Sancho, J.; Fersht, A. R. J. Mol. Biol. 1992, 224,
759.
Results and Discussion
(24) Rico, M.; Santoro, J.; Bermejo, F. J.; Herranz, J.; Nieto, J. L.;
Gallego, E.; Jimenez, M. A. 1986, 25, 1031.
(25) Rico, M.; Gallego, E.; Santoro, J.; Bermejo, F. J.; Nieto, J. L.;
Horranz, J. Biochem. Biophys. Res. Commun. 1984, 123, 757.
(26) Shoemaker, K. R.; Kim, P. S.; Brems, D. N.; Marqusee, S.; York,
E. J.; Chaiken, I. M.; Stewart, J. M.; Baldwin, R. L. Proc. Natl. Acad. Sci.
U.S.A. 1985, 82, 2349.
(27) Scheraga, H. A. Proc. Natl. Acad. Sci. U.S.A. 1985, 82, 5585.
(28) Armstrong, K. M.; Baldwin, R. l. Proc. Natl. Acad. Sci. U.S.A. 1993,
90, 11337.
(29) Blagdon, D. E.; Goodman, M. Biopolymers 1975, 14, 241.
(30) Burgess, K.; Ho, K.-K.; Pettitt, B. M. J. Am. Chem. Soc. 1994, 116,
799.
Preparation of FiFi and Incorporation into Peptidomi-
metics via Solid-Phase Syntheses. Several synthetic routes
to FiFi were explored before a satisfactory one was found. In a
first attempt, rhodium-catalyzed cyclopropanation of trans-1,2-
diphenylethene with an alkenyl diazoacetate was investigated,
but the desired product was not formed under the conditions
(37) Baldwin, J. E.; Adlington, R. M.; Rawlings, B. J.; Jones, R. H.
Tetrahedron Lett. 1985, 26, 485.
(38) Baldwin, J. E.; Adlington, R. M.; Rawlings, B. J. Tetrahedron Lett.
1985, 26, 481.
(31) Burgess, K.; Ho, K.-K.; Pettitt, B. M. J. Am. Chem. Soc. 1995, 117,
54.
(32) Burgess, K.; Ho, K.-K.; Pal, B. J. Am. Chem. Soc. 1995, 117, 3808.
(33) Burgess, K.; Li, W.; Lim, D.; Moye-Sherman, D. Biopolymers 1996,
42, 439.
(34) Burgess, K.; Ke, C.-Y. Int. J. Peptide Protein Res. 1997, 49, 201.
(35) Burgess, K.; Ke, C.-Y. J. Org. Chem. 1996, 61, 8627.
(36) Burgess, K.; Lim, D. J. Am. Chem. Soc. 1997, 119, 9632.
(39) Kirihata, M.; Sakamoto, A.; Ichimoto, I.; Ueda, H.; Honma, M.
Agric. Biol. Chem. 1990, 54, 1845.
(40) Koskinen, A. M. P.; Munoz, L. J. Org. Chem. 1993, 58, 879.
(41) Fadel, A. Synlett 1993, 503.
(42) Zhu, Y. F.; Yamazaki, T.; Tsang, J. W.; Lok, S.; Goodman, M. J.
Org. Chem. 1992, 57, 1074.
(43) Gaucher, A.; Ollivier, J.; Marguerite, J.; Paugam, R.; Salau¨n, J. Can.
J. Chem. 1994, 72, 1312.

9438 J. Am. Chem. Soc., Vol. 120, No. 37, 1998
Moye-Sherman et al.
Scheme 1. Asymmetric synthesis of FiFi
we used.44 Later, other work from our group implied that the
opening of cyclic sulfates with malonate derivatives would be
appropriate, but these reactions proved unsuitable in this
particular case.45 For instance, the ring-opened product 2
predominated over cyclopropane 3 when dimethyl malonate was
reacted with the cyclic sulfate 1. Other nucleophiles such as
4-phenyl-3-butenoate and benzylidene glycinate46 gave the
desired material with no detectable elimination on a small scale,
but failed to give the cyclopropane products cleanly when the
reaction was scaled up to multigram amounts. Byproducts
identified from these reactions included diphenyl acetylene and
others derived from ester hydrolysis. Thus, the spectrum of
chemical characteristics (e.g., nucleophilicity and basicity) that
are tolerable for the nucleophilic component in the formation
of cyclopropanes from cyclic sulfates seems to be limited.
Ultimately, a practical route to a usable cyclopropane was
developed, involving reaction of diethyl glutaconate with 1.
Consequently, a preparation of (2R,3R)-3-phenyl-cyclo-Phe [i.e.,
(R,R)-FiFi] from trans-1,2-diphenylethene was developed ac-
cording to Scheme 1. Asymmetric dihydroxylation47 gave diol
4 with an optical purity of >99% ee as assayed with a chiral
aryl boronic acid.48 This diol was converted first to the cyclic
sulfate 1,49 then to cyclopropane 5 by a double displacement
with diethyl glutaconate. Oxidation of alkene 550 provided the
acid ester 6. Curtius rearrangement of this acid via treatment
with diphenylphosphoryl azide (DPPA) in tert-butyl alcohol,
followed by basic hydrolysis, afforded the N-butoxycarbonyl
(BOC)-protected (2R,3R)-3-phenyl-cyclo-Phe. This protected
form is appropriate for solid-phase syntheses of peptidomimetics.
The enantiomeric product, N-BOC-protected (2S,3S)-3-phenyl-
cyclo-Phe, was obtained in exactly the same way, except that
AD-mix-R (Aldrich) was used in the first step.
For solid-phase syntheses of peptidomimetics from FiFi, we
selected the fluorenyl methoxycarbonyl (FMOC) approach51 to
avoid problems that could arise because of the inconvenience
and possible complications in HF cleavage procedures. FMOC-
N-protected forms of the two FiFi enantiomers were therefore
prepared via trifluoroacetic acid (TFA)-mediated removal of the
N-BOC group, and reaction with FMOC-OSu (Su ) succin-
imide). The FMOC-protected amino acids were purified via
flash chromatography on silica gel. Yields of materials isolated
after such purifications depend inversely on the residence time
on the support since these (and incidentally some other hindered
FMOC-protected amino acids that we have prepared) are not
particularly robust with respect to this form of chromatography.
We synthesized a sample of RN-24 on Rink’s amide resin,52
using benzotriazoyloxy-tris(pyrrolidino)-phosphonium hexafluo-
rophosphate/1-hydroxybenztriazole (PyBOP/HOBt)53 for the
couplings, performed for 1-h reaction times, and repeated once
or twice as required for complete coupling (indicated by negative
ninhydrin test results).54 Peptidomimetics 9 and 10 containing
(R,R)- and (S,S)-FiFi, respectively, were prepared via the same
procedure except for the coupling to incorporate the FiFi
residues, and for the coupling after that. For these couplings,
acid fluorides were produced in situ with use of the reagent
tetramethylfluoroformadinium hexafluorophosphate (TFFH)55
with 1-hydroxy-7-azabenztriazole (HOAt)56 as an activating
agent. The coupling to incorporate the FiFi moieties required
only 1 h, whereas the subsequent coupling was more difficult
and was run for 12 h. In this work, the quantities of these
relatively high-molecular-mass peptidomimetics obtained (∼6
mg) were sufficient for circular dichroism (CD) studies but not
for NMR, although the route described above could be used to
prepare enough for NMR studies if more time and resources
were available.
(44) Davies, H. M. L.; Bruzinski, P. R.; Lake, D. H.; Kong, N.; Fall, M.
J. J. Am. Chem. Soc. 1996, 118, 6897.
(45) Burgess, K.; Ho, K.-K.; Ke, C.-Y. J. Org. Chem. 1993, 58, 3767.
(46) Hercouet, A.; Bessieres, B.; Corre, M. L. Tetrahedron: Asymmetry
1996, 7, 283.
(47) Sharpless, K. B.; Amberg, W.; Bennani, Y. L.; Crispino, G. A.;
Hartung, J.; Jeong, K.-S.; Kwong, H.-L.; Morikawa, K.; Wang, Z.-M.; Xu,
D.; Zhang, X.-L. J. Org. Chem. 1992, 57, 2768.
CD Studies. Qualitative analysis of the CD spectra of the
peptidomimetics in pH 5.2 buffer indicated that RN-24 itself
(48) Burgess, K.; Porte, A. M. Angew. Chem., Int. Ed. Engl. 1994, 33,
1182.
(49) Gao, Y.; Sharpless, K. B. J. Am. Chem. Soc. 1988, 110, 7538.
(50) Carlsen, P. H. J.; Katsuki, T.; Martin, V. S.; Sharpless, K. B. J.
Org. Chem. 1981, 46, 3936.
(52) Rink, H. Tetrahedron Lett. 1987, 28, 3787.
(53) Coste, J.; Le-Nguyen, D.; Castro, B. Tetrahedron Lett. 1990, 31,
205.
(54) Kaiser, E.; Colescott, R. L.; Bossinger, C. D.; Cook, P. I. Anal.
Biochem. 1970, 34, 595.
(51) Atherton, E.; Sheppard, R. C. In Solid-Phase Peptide Synthesis, A
Practical Approach; IRL Press: Oxford, 1989.
(55) Carpino, L. A.; El-Faham, A. J. Am. Chem. Soc. 1995, 117, 5401.
(56) Carpino, L. A. J. Am. Chem. Soc. 1993, 115, 4397.

Conformations of RNase C-Peptide DeriVatiVes
J. Am. Chem. Soc., Vol. 120, No. 37, 1998 9439
difficult to correct for. The overall consequence of these
uncertainties in concentration determinations is that our com-
parisons between the molar ellipticities obtained for different
samples are tenuous. Differences between the concentration
data obtained from UV and amino acid analyses indicate that
the percentage errors in the ellipticities are on the order of 10%.
and the (R,R)-FiFi derivative 9 had the same shape, indicative
of helical character (Figure 2a). The RN-24 sample had molar
ellipticity minima near 206 and 222 nm, and 9 had a very similar
CD spectrum but with minima slightly shifted to ∼204 and 220
nm. However, the CD spectrum of the (S,S)-FiFi derivative
10 was an unusual shape that was reminiscent neither of a
random-coil situation nor any other common secondary structure
element. Because the compound gave no distinct minimum at
220 nm, this peptidomimetic does not appear to have any bias
to a helical structure that is identifiable via this technique.
Comparisons of variable temperature CD data obtained for
RN-24 with that for the (R,R)-FiFi derivative 9 were interesting
and informative (Figures 2b and 2c). As RN-24 was warmed
from 3 to 23 °C in 5 °C intervals, its molar ellipticity at 222
nm steadily decreased to 65% of its original value. However,
the molar ellipticity for peptidomimetic 9 in an analogous
experiment decreased to 90% of its original value; that is, the
decrease was only 10%. Thus, the CD spectrum of this FiFi
derivative, and therefore its secondary structure, are much less
affected by a 20 °C increase in temperature than is RN-24. This
appears to be a good example of stabilization of a secondary
structure by substitution with a 2,3-methanoamino acid.
The secondary structure of peptidomimetic 9 was also less
affected by pH than RN-24. Thus, whereas the molar ellipticity
of RN-24 changes quite dramatically over a pH range of 7.3
units, the ellipticity of 9 is relatively insensitive to this
perturbation (Figures 2d and 2e, respectively). Figure 2f shows
the molar ellipticity at 222 nm as a function of pH for RN-24
and for 9. The bell-shaped curve seen for RN-24 corresponds
to the forming and breaking of the interactions between charged
groups (i.e., Glu--2 to Arg+-10 and Phe-8 to His+-12).
However, the CD data show that those same factors cannot be
as prevalent in stabilizing the presumed helical structure of 9.
For this peptidomimetic, the ellipticity increases in the pH range
2 to 4, corresponding to progressively more-deprotonated Glu-
2, and decreases as Lys+-7 is deprotonated, in the pH 9.5 to
10.5 region. The latter may be taken as evidence for formation
of a Glu--2 to Lys+-7 salt bridge. This type of salt bridge was
also observed in some of the molecular dynamics studies
described below.
Apparently, compound 9 has less ellipticity at 222 nm and is
therefore less helical than RN-24, though numerical differences
between the two [θ]-values are hard to compare because of
experimental errors. Accurate quantitative analyses of the CD
data was not possible because of the problems with quantifica-
tion that have arisen frequently in studies of the C-peptide.21
Generally, the best method for measuring the concentration of
RN-24 derivatives is to replace Phe-8 with a tyrosine residue;
the molar UV absorbance at 280 nm is then relatively intense
and easily calibrated.21 This method is clearly not applicable
in this work, however, where our purpose was to replace Phe-8
by another phenylalanine surrogate. We therefore used two
other methods for concentration measurement in our studies:
calibration of the UV-absorbance at 205 nm57 and quantitative
amino acid analyses. Reasonable agreement was obtained with
both methods; however, we used the amino acid analysis results
to obtain the molar ellipticities given in Figure 2. Calibration
of concentrations by measuring the UV absorption at 205 nm
was less accurate than amino acid analyses because of the slight
absorbance of the FiFi residue at that wavelength, which is
Molecular Modeling. The structure of RN-24 was built in
CHARMm by using the coordinates of S-peptide as a starting
point, minimized via a molecular mechanics routine, and then
subjected to molecular dynamics studies at a simulated tem-
perature of 276 K for 300 ps. This study parallels the analysis
of an S-peptide previously performed by Jorgensen et al. and
yielded very similar results (Figure 1b).14 In the Jorgensen
work, the helical structure of an S-peptide analogue was
conserved throughout a 300-ps molecular dynamics run at a
simulated temperature of 278 K. In our study, the simulated
helical structure of RN-24 also persisted (276 K, 300 ps).
Peptidomimetic 9 was constructed by using modified coor-
dinates from the crystal structure of the S-peptide as a starting
point. A parameter set and topology file (see Supplementary
Information) were constructed for the FiFi component by use
of the approach previously outlined for other cyclopropane
amino acids.30,31 A simple molecular mechanics minimization
gave the conformer shown in Figure 1c. This has seven
hydrogen bonds from residues 3 to 13 that impart a helical
character, but some of the key features that stabilize the RN-24
helix are not present. For instance, the Glu--2 to Arg+-10 salt
bridge is disrupted, and the phenyl rings of the FiFi component
are too far from His+-12 for there to be a significant interaction
(4.7 Å, compared with 3.9 Å in RN-24, for the distances
between an NH proton of the His imidazole and the nearest
phenyl ring carbon). However, when this structure was
subjected to a molecular dynamics run, as outlined above, all
the backbone hydrogen bonds were conserved and those factors
that stabilize RN-24 were reinstated. Features of the simulation
that led to that conclusion are as follows. First, conformers
having the Glu--2 to Arg+-10 salt bridge predominated.
Second, and most interesting, the protonated His-12 side chain
tended to adopt conformations in which the imidazole interacted
with both phenyl rings of the FiFi residue, one in a perpendicular
edge-to-face orientation (3.5 Å separation) and the other in a
parallel face-to-face arrangement (3.5 Å separation). The edge-
to-face orientation is quite similar to that found in the S-peptide
crystal structure. Overall, the conformations that prevail toward
the end of the simulation (150-300 ps) are quite helical (Figures
1d and 1e), the phenyl rings of the FiFi residue being almost
perpendicular to the helix axis. Finally, a Glu-2 to Lys+-7
electrostatic interaction (not present in RN-24) was also
observed.
The protocol used to simulate the conformational biases of
the (S,S)-FiFi isomer 10 was identical to that used for its
diastereomer 9, but the results were very different. After
minimization, the helical conformation of this molecule is
distorted (Figure 1f). There is no salt bridge between Glu--2
and Arg+-10 in the starting structure. One of the FiFi phenyl
groups appears to interact with the protonated His-12 (2.6 Å
separation), but the angle between these two groups is nonideal
for a perpendicular or a face-to-face orientation. In fact, the
FiFi phenyl groups are approximately aligned with the helical
twist. This observation contrasts with the proposed conforma-
tional bias of peptidomimetic 9, wherein the two phenyl rings
are perpendicular to the helix. Overall, the region of the helix
around the FiFi residue is crowded, implying that the molecule
would relax away from this conformation if the energy were
(57) Scopes, R. K. Anal. Biochem. 1974, 59, 277.

9440 J. Am. Chem. Soc., Vol. 120, No. 37, 1998
Moye-Sherman et al.
Figure 2. CD spectra of samples in pH 5.2 phosphate/citrate/borate buffer. (a) RNase A, 9, and 10; (b) RN-24 at various temperatures; (c) 9 at
various temperatures; (d) RN-24 at various pH’s; (e) 9 at various pH’s; (f) the molar ellipticity of RN-24 and 9 at 222 nm as a function of pH.
Conclusions
Difficulties were encountered in our previous attempts to
study conformational biases imposed by cyclopropane amino
acids. Conformational ensembles rather than distinctly preferred
conformers were generated when small linear tetrapeptides were
used as model systems. These were difficult to characterize
since so few long-range rotating-frame Overhauser enhancement
(ROE) cross-peaks were observed in NMR studies. Conversely,
incorporation of cyclopropane amino acids into model cyclic
peptides gave more readily identified conformational biases, but
these were clearly a compromise between the effects of the
Figure 3. Deviation of the N, CR, C, and O backbone atoms of
unusual amino acid and the cyclic structure. Neither approach
conformers from their respective starting structures directly after the
was therefore ideal.
initial molecular mechanics minimization.
The model system described in this paper is large enough to
available to it. Indeed, upon molecular dynamics simulations
(276 K for 300 ps), the helical conformation is lost in all the
latter conformations sampled. The conformers sampled in the
latter half of this simulation tended to have the protonated His+-
12 side chain and the FiFi phenyl rings on opposite sides of
the structure (e.g., Figure 1g).
Figure 3 shows the root-mean-square (RMS) deviation of the
backbone atoms (N, CR, C, and O) of sampled conformers from
their respective starting structures directly after the initial
molecular mechanics minimization. Overall, peptidomimetic
9 shows less variance than RN-24 from its starting conformer,
whereas compound 10 undergoes a more drastic relaxation.
have a recognizable secondary structure, but not so rigid that
the effects of cyclopropane amino acid substitutions are
negligible. It is a useful compromise between the short linear
peptide and the cyclic system that have been studied previously
in these labs. The synthesis and application of FiFi are also
innovations, insofar as large steric influences on the φ and ψ
bond vectors can be probed simultaneously in this disubstituted
system. The CD data described indicates that the RN-24
derivative from (R,R)-FiFi, compound 9, is helical; this con-
formation was prevalent in molecular simulations of that
molecule, lending support to that hypothesis. Moreover, and
perhaps most interesting of all, the stability of the helical

Conformations of RNase C-Peptide DeriVatiVes
J. Am. Chem. Soc., Vol. 120, No. 37, 1998 9441
extracted with ethyl acetate (EtOAc) (2 × 200 mL). The combined
organics were dried over Na2SO4, filtered, evaporated, and dried. The
crude material was purified by silica gel chromatography using 10%
EtOAc/hexanes as eluent, Rf ) 0.60, collected, and dried to give 2.8 g
(72%) of 5 as a yellow oil: 1H NMR (300 MHz, CDCl3) δ 0.92 (t,
3H, J ) 7.2 Hz), 1.22 (t, 3H, J ) 7.2 Hz), 3.36 (d, 1H, J ) 8.4 Hz),
3.92 (m, 3H), 4.06 (m, 2H), 5.87 (d, 1H, J ) 15.9 Hz), 6.77 (d, 1H, J
) 15.9 Hz), 7.28-7.35 (bm, 10H); 13C NMR (75 MHz, CDCl3) δ 13.77,
14.15, 37.23, 39.50, 40.79, 60.23, 61.23, 121.38, 130.80, 127.39, 127.42,
128.22, 128.54, 128.90, 129.33, 134.89, 135.10, 143.91, 166.18, 168.48.
IR (CH2Cl2) 1684, 1602, 1228 cm-1. HRFABH+: calcd 365.1753,
found 365.1738. [R]D +69.7, (c )1.4, CHCl3). Anal. Calcd for
C23H24O4: C, 75.79; H, 6.64; O, 17.57. Found: C, 74.93; H, 6.53.
(2′S,3′S)-[1-(2′,3′-Diphenyl)-cyclopropyl] diethyl glutaconate was
prepared in the same way as 5, affording 3.3 g of a yellow oil (62%).
[R]D ) -78.7, (c ) 1.2 CHCl3).
(2R,3R)-1-Carboxy-1-(ethyloxycarbonyl)-2,3-diphenyl cyclopro-
pane (6). To a rapidly stirred mixture of compound 5 (7.72 mmol, 1
equiv, 2.80 g) and NaIO4 (61.7 mmol, 8 equiv) in acetonitrile and CCl4
(15 mL each) and H2O (23 mL) was added RuCl3‚H2O (0.154 mmol,
0.02 equiv). The resulting slurry was stirred at room temperature for
15 h. The mixture was filtered, and the filtrate was washed with H2O
and CH2Cl2. The aqueous layer was extracted with CH2Cl2 (2 × 150
mL). The combined organics were dried (MgSO4), filtered, and
evaporated to give 6, 1.6 g (70%), as a brown oil. The crude acid was
used in the next step without further purification. A small amount of
6 was purified by silica gel chromatography (25-40% EtOAc/hexanes)
to give the following analytical data: 1H NMR (300 MHz, D6-acetone)
δ 0.93 (t, 3H, J ) 7.2 Hz), 3.77 (q, 2H, J ) 6.3, 8.4 Hz), 3.91-4.00
(m, 2H), 7.31-7.44 (m, 10H); 13C NMR (75 MHz, D6-acetone) δ 13.21,
34.10, 34.21, 60.82, 127.12, 121.16, 128.12, 128.69, 128.77, 135.139,
166.66; IR (CH2Cl2) 1700, 1265 cm-1. HRFABH+: calcd 311.1283,
found 311.1302. (c ) 1.04, CHCl3). Anal. Calcd for C19H18O4: C,
73.52; H, 5.85; O, 20.63. Found: C, 72.71; H, 5.92.
conformation for 9 is less perturbed by temperature increases
and by pH changes than that of RN-24. We infer that this
greater stability can be attributed to the enhanced conformational
rigidity imparted by FiFi. Conversely, the effect of the (S,S)-
isomer of FiFi was large enough to override the helical
preference of the RN-24 sequence. We conclude this isomer
is not conducive to right-handed helical secondary structures
but would enhance populations of left-handed helices in an
appropriate sequence (i.e., one formed predominantly from
D-amino acids).
Experimental Section
General Procedures. Melting points were uncorrected. High-field
NMR spectra were recorded on a Unity+ 300 spectrometer (1H at 300
MHz, 13C at 75 MHz); 1H chemical shifts are reported in δ relative to
CHCl3 (7.24 ppm) as internal standard, and 13C chemical shifts are
reported in ppm relative to CDCl3 (77.0 ppm) unless otherwise
1
specified. Multiplicities in H NMR are reported as (br) broad, (s)
singlet, (d) doublet, (t) triplet, (q) quartet, and (m) multiplet. Thin-
layer chromatography was performed on silica gel 60 F254 plates from
Whatman. Flash chromatography was performed on SP silica gel 60
(230-600-mesh ASTM). tert-Butyl alcohol was distilled from CaH2;
dimethoxyethane (DME) was distilled from a sodium and benzophenone
ketyl. Other chemicals were purchased from commercial suppliers and
used as received.
(1R,2R)-1,2-Diphenyl-1,2-ethanediol (4). Prepared from trans-1,2-
diphenylethene according to method of Sharpless et al.47 the yield after
one recrystallization was 80%, ee >99% [HPLC chiral column (S,S)-
Whelk-O1, Regis Technologies]. Spectroscopic data for 1:1H NMR
(200 MHz, CDCl3) δ 2.91 (s, 2H), 4.69 (s, 2H), 7.08-7.25 (m, 10H);
13C NMR (50 MHz, CDCl3) δ 79.09, 126.92, 127.92, 128.12, 139.81;
mp 146-147 °C.
(2S,3S)-1-Carboxy-1-(ethyloxycarbonyl)-2,3-diphenyl cyclopropane
was prepared the same way as 6, providing 5.1 g of a brown oil (67%).
(2R,3R)-1-[N-[(tert-Butyloxycarbonyl)]amino]2,3-diphenyl-cyclo-
propane-1-carboxylic Acid (7). Compound 6 (4.19 mmol, 1 equiv),
triethylamine (5.03 mmol, 1.2 equiv), and dry tert-butyl alcohol (60
mL) were added to a 200-mL two-neck round-bottom flask with reflux
condenser and a N2 inlet. Diphenylphosphoryl azide (9.22 mmol, 2.2
equiv) was added, and the resulting mixture was refluxed for 15 h.
The solution was concentrated and the crude material was extracted
with EtOAc (2 × 200 mL), washed with H2O, brine (1 × 20 mL each),
and dried over Na2SO4. After evaporation of the solvent, the crude
residue was purified via flash chromatography (10-25% EtOAc/
hexanes to give 0.89 g of white crystals (55%): Rf ) 0.55 (20% EtOAc/
hexanes); 1H NMR (30 MHz, D6-acetone) δ 0.90 (t, 3H, J ) 7.0 Hz),
1.33 (s, 9H), 3.39 (d, 1H, J ) 8.7 Hz), 3.74 (d, 1H, J ) 8.7 Hz), 3.84
(dq, 2H, J ) 1.8, 5.1 Hz), 6.68 (s, 1H), 7.27-7.52 (m, 10H); 13C NMR
(75 MHz, D6-acetone) δ 14.20, 14.43, 27.96, 28.36, 35.41, 35.81, 37.78,
38.04, 48.01, 61.05, 61.22, 79.01, 79.29, 127.29, 127.44, 127.61, 128.61,
128.70, 129.50, 129.87, 130.02, 136.59, 136.90, 156.46, 169.51. IR
The (S,S)-enantiomer was prepared in an identical fashion; the yield
after one recrystallization was 84%, ee >98%.
(4R,5R)-4,5-Diphenyl-(1,3,2)-dioxathiolane-(2,2)-dioxide (1). A
250-mL two-neck round-bottom flask was fitted with a reflux condenser
carrying a drying tube with an HCl trap, and the sidearm was stoppered.
1,2-Diphenyl-1,2-ethanediol 4 (23.3 mmol, 5 g, 1 equiv) and CCl4 (25
mL) were placed in the flask, then thionyl chloride (1.2 equiv, 2.1 mL)
was added via syringe over ∼10 min. The resulting mixture was
refluxed for 30 min. The solution was cooled in an ice-water bath
and then diluted with acetonitrile (25 mL). Solid NaIO4 (7.5 g, 1.5
equiv) and RuCl3‚H2O (4 mg, 0.001 equiv) were added, followed by
H2O (35 mL), and the resulting biphasic mixture was rapidly stirred at
room temperature for 60 min. The mixture was then diluted with
diethyl ether (Et2O) (200 mL), and the phases were separated. After
the organic layer was washed with H2O (10 mL), saturated aqueous
NaHCO3 (2 × 10 mL), and brine (10 mL) and was dried over MgSO4,
the solution was filtered through a small pad of silica to remove the
dark color, concentrated, and dried under a high vacuum to give 5.2 g
(81%) of 1 as white crystals. This product decomposes at room
temperature but is stable for several months in the freezer. The crude
product49 was used without further purification: 1H NMR (200 MHz,
CDCl3) δ 5.76 (s, 2H), 7.29-7.34 (m, 5H), 7.41-7.51 (m, 5H); 13C
NMR (50 MHz, CDCl3) δ 89.72. 127.15, 129.19, 130.60.
The enantiomer of this product, (S,S)-4,5-diphenyl-(1,3,2)-dioxa-
thiolane-(2,2)-dioxide (ent-1), was prepared by an identical procedure,
starting with the enantiomeric diol to give 11 g of the product (81%).
(2′R,3′R)-[1-(2′,3′-Diphenyl)-cyclopropyl] diethyl glutaconate (5).
To a 250-mL two-neck round-bottom flask equipped with a reflux
condenser, N2 inlet, and a magnetic stirrer was added NaH (2.5 equiv,
26.7 mmol, 0.304 g). The sidearm was stoppered and freshly distilled
DME was added by syringe. Diethyl glutaconate (1.1 equiv, 10.7 mmol,
2.19 g) was added dropwise to the NaH suspension. (4R,5R)-4,5-
diphenyl-(1,3,2)-dioxathiolane-(2,2)-dioxide, 1 (1 equiv, 2.95 g), was
dissolved in 50 mL of DME and added dropwise to the anion mixture.
The resulting orange solution was stirred at room temperature for 2 h.
The solution was diluted with saturated aqueous NH4Cl (60 mL) and
(CH2Cl2) 3425, 1724, 1269, cm-1
. HRFABH+ calcd: 382.2018,
found: 382.2026. [R]D ) +56.41 (c ) 1.01, CHCl3). Mp ) 88-89
°C. Anal. Calcd for C23H27NO4: C, 72.4; H, 7.14; N, 3.67; O, 16.78.
Found: C, 72.5; H, 7.13; N, 3.71.
The enantiomeric product, (2S,3S)-ethyl-1-[N-[(tert-butyloxy-
carbonyl)]amino]-2,3-diphenyl cyclopropane-1-carboxylate, was pre-
pared by a procedure identical to that described above, with ent-6 as
substrate (50%). [R]D ) -60.2, (c ) 1.28, CHCl3).
In a 150-mL round-bottom flask equipped with a reflux condenser
and stir bar, (2R,3R)-ethyl-1-[N-[(tert-butyloxycarbonyl)]amino]-2,3-
diphenyl cyclopropane-1-carboxylate (1.83 mmol, 1 equiv) was mixed
with NaOH (5.49 mmol, 3 equiv) and aqueous methanol (50 mL) and
refluxed for 48 h. The methanol was evaporated and the residue was
extracted with Et2O, 100 mL. The aqueous was acidified to pH∼2
with 1M HCl and extracted with EtOAc (3 × 150 mL). The combined
organic layers were washed with H2O and brine (1 × 20 mL, each)
and dried (Na2SO4). After evaporation of the solvent, the crude residue
was purified via flash chromatography (20-40% EtOAc/hexane, Rf )

9442 J. Am. Chem. Soc., Vol. 120, No. 37, 1998
Moye-Sherman et al.
0.23), collected, and dried to give 0.6 g (98%) of BOC-FiFi 7 as a
white solid: 1H NMR (300 MHz) δ 1.18 (s, 3H), 1.31 (s, 6H), 3.38-
3.33 (m, 1H), 3.72 (d,1H, J ) 8.7 Hz), 6.67 (s, 1H), 7.22-7.53 (bm,
10H); 13C NMR (75 MHz, CDCl3) δ 27.74, 28.15, 35.32, 35.60, 37.80,
38.07, 47.76, 78.77, 79.14, 127.01, 127.17, 127.31, 128.39, 128.43,
129.23, 129.74. 129.82, 136.51, 136.51, 136.66, 155.29, 156.29,
TOF [M + H2O] calcd: 1455.73, found: 1455.91. Amino acid analysis
revealed Ala:Thr of 19.716:3.310 (6:1).
Succ-Ala-Glu-Thr-Ala-Ala-Ala-Lys-[(R,R)FiFi]-Leu-Arg-Ala-
His-Ala-NH2 (Peptidomimetic 9). This peptide was synthesized by
a protocol similar to the one used for RN-24 except that FMOC-(R,R)-
FiFi-OH was substituted for FMOC-Phe in the sequence. Additionally,
to incorporate this hindered amino acid, the coupling conditions for
this and the next residue were slightly modified. The acid fluoride of
FiFi was prepared in situ by mixing FMOC-FiFi (1.1 equiv) with
TFFH (2 equiv) EtNiPr2 (2 equiv), HOAt (2 equiv), and DMF for 7
min, then adding to the resin. The coupling was complete in 1 h
(negative ninhydrin test result). The next residue, Lys-7, was also
coupled by using TFFH, diisopropylethylamine (DIEA), and HOAt
(1:2:2:2 equiv) for 12 h at 25 °C. The remainder of the peptide was
prepared by using the standard conditions outlined above. The desired
peptide was collected and dried to give 9 (6 mg, 3%) as a white powder.
MALDI-TOF [M + H2O] calcd: 1543.76, found: 1543.90. Amino
acid analysis revealed Ala:Thr of 38.946:6.124 (6:1).
Succ-Ala-Glu-Thr-Ala-Ala-Ala-Lys-[(S,S)FiFi]-Leu-Arg-Ala-His-
Ala-NH2, (Peptidomimetic 10). Synthesis of this analogue was
performed by a protocol identical to that for 9 except that FMOC-
(S,S)-FiFi-OH (1.1 equiv) was used for the coupling at the eighth
position. The desired peptide was collected and dried to give 8 as a
white powder (5 mg, 3%). MALDI-TOF [M + H2O] calcd: 1543.76
found: 1543.90. Amino acid analysis revealed Ala:Thr of 7.433:1.058
(7:1).
Molecular Modeling. A Silicon Graphics IRIX-O2 workstation
was used for the molecular simulations performed in this work. All
calculations were performed with QUANTA97/CHARMm version 23.2
software (Molecular Simulations Inc.) with extended representations
of the nonpolar hydrogen atoms.
Since CHARMm does not have parameters for the cyclopropane
amino acids, additional atom types were assigned and a parameter set
was built, based on crystallographic data and CHARMm default
parameters, and then was appended to the CHARMm standard
parameter file. The residue topology files (RTFs) of the two FiFi amino
acids were built according to the standard geometry of these two
unnatural amino acids and then appended to the CHARMm standard
RTF. The combined CHARMm parameters and RTFs were imported
to QUANTA/CHARMm for the following calculations.
170.47. IR (CH2Cl2) 3426, 1696, 1264 cm-1
. HRFABH+ calcd:
354.1705, found: 354.1696. [R]D ) +70.1 (c ) 1.03, CHCl3). Anal.
Calcd for C21H23NO4: C, 71.4; H, 6.56; N, 3.97. Found: C, 71.3; H,
6.58 N, 3.93.
(2S,3S)1-[N-[(tert-Butyloxycarbonyl)]amino]2,3-diphenyl-cyclopro-
pane-1-carboxylic acid was prepared by a procedure identical to that
as described above but using the (S,S)-BOC amino ester as substrate
(95%). [R]D ) -66.8 (c ) 1.28, CHCl3).
(2R,3R)-1-[N-[(9-Fluorenylmethoxy)carbonyl]amino]-2,3-di-
phenyl-cyclopropane-1-carboxylic Acid (8). A 1:1 mixture of TFA/
CCl4 at 0 °C was added dropwise to 0.5 g of 7. This mixture was
stirred at 25 °C for 30 min, concentrated, neutralized with 2 M NaOH,
and lyophilized to dryness. The crude amine was then dissolved in a
mixture of 1 M Na2CO3 (5.1 mL), N,N-dimethylformamide (DMF) (7
mL), and H2O (1.8 mL) and cooled to 0 °C. FMOC-OSu was added
in one portion and the mixture was stirred for 8 h at 25 °C. The mixture
was then diluted with H2O (80 mL) and washed with Et2O (40 mL 5).
The aqueous layer was acidified to pH ∼2 with citric acid (4.5 g) and
extracted with EtOAc (150 mL). The organic layer was washed with
H2O and brine and then was dried over MgSO4. After evaporation of
the solvent, the crude residue was purified via flash chromatography
with 10-40% EtOAc/hexanes eluent to obtain 0.42 g (62%) of the
target molecule 8 as a white solid: Rf ) 0.29 (40% EtOAc/hexanes);
1H NMR (300 MHz) δ 3.45 (d, 1H, J ) 8.7 Hz), 3.73 (d, 1H, J ) 8.4
Hz), 4.00-4.31 (3H, m), 7.14-7.84 (m, 18 H), 7.99 (s, 1H); 13C NMR
(75 MHz, CDCl3) δ 35.04, 37.73, 47.05, 66.15, 78.32, 119.85, 125.29,
125.37, 126.81, 127.02, 127.57, 127.62, 127.88, 128.00, 128.69, 129.02,
129.32, 135.56, 135.94, 141.13, 143.90, 144.28, 156.50, 169.29, 169.73.
IR (CH2Cl2) 3416, 1736, 1268 cm-1. HRFABH+ calcd: 476.1862,
found: 476.1869. [R]D ) +47.1, (c ) 0.73, CHCl3).
(2S,3S)1-[N-[(9-Fluorenylmethoxy)carbonyl]amino]-2,3-diphenyl-cy-
clopropane-1-carboxylic acid was prepared in an identical fashion to 8
but with ent-7 as substrate and yielded 0.63 g (48%) of product as a
white solid. [R]D ) -46.4, (c ) 0.74, CHCl3).
Initial coordinates for RN-24 and the FiFi derivatives were obtained
from the X-ray structures of ribonuclease A (Protein Data Bank entry
1RTB.11 The first 13 residues were selected and then modified to form
the structure of RN-24 and the (R,R)-FiFi and (S,S)-FiFi peptidomi-
metics. These initial structures were first minimized by 500 steepest
decent cycles followed by conjugate gradient minimization until the
Succ-Ala-Glu-Thr-Ala-Ala-Ala-Lys-Phe-Leu-Arg-Ala-His-Ala-
NH2, (Peptide RN-24). Stepwise couplings of FMOC-amino acid
derivatives on Rink’s amide resin (Advanced Chemtech) were used to
prepare the 13-residue peptide. Manual peptide synthesis was carried
out in a 30-mL vessel fitted with a coarse glass frit and using a wrist-
action shaker (Burrel, model 75). The reagents were added manually.
All reactions were carried out at 25 °C unless otherwise specified.
FMOC deprotection was performed by shaking the resin twice with
20% piperidine in DMF (5 mL for 3 min and 5 mL for 7 min); DMF
washing cycles (10 × 1 min, ∼10 mL) were performed after each
coupling and deprotection. The coupling conditions were as follows.
Each FMOC-amino acid coupling was accomplished by premixing the
amino acid (2 equiv) with 4-methylmorpholine (NMM) (3 equiv), HOBt
(2 equiv), and PyBOP (2 equiv) in DMF (7 mL); this solution was
added to the resin (200 mg, 1 equiv, 0.6 mmol/g loading) and the
mixture was shaken for 1 h. A negative ninhydrin test result was
observed after this stage. Deprotection of the side chains, cleavage
from the resin, and purification of the peptide were performed as
follows. A mixture of phenol (1 mL), 1,2-ethanedithiol (0.5 mL),
thioanisole (1 mL), deionized water (1 mL), and TFA (16.5 mL) was
cooled to 0 °C and added to the crude peptide. The reaction mixture
was stirred for 12 h at 25 °C and then concentrated to dryness. Et2O
(30 mL) was added to precipitate the product, after which the ethereal
solution was decanted away from the solid residue. The crude peptide
was further purified by preparative reversed-phase HPLC (Vydac C18
column, 22 mm × 25 cm, 10 mm) with a linear gradient obtained by
mixing solvent A (0.1% TFA in water) and solvent B (0.1% TFA in
acetonitrile). The gradient was programmed to increase from 5% to
20% B over 60 min with a flow rate of 6 mL min-1. The desired
peptide RN-24 was obtained as a white powder (13 mg, 7%). MALDI-
RMS energy deviation was <0.01 kcal mol-1 Å-1
. The resulting
conformations were used as the starting structures in the following
molecular dynamics simulations.
Free dynamics simulations were performed on the three peptidomi-
metics at 276 K. The Verlet algorithm was used during the dynamics,
with a time step of 1 fs and SHAKE to constrain all bond lengths
containing polar hydrogens. A dielectric continuum with ꢀ ) 80 was
used to model the aqueous environment. A nonbonded pair list was
generated by using a 12-Å cutoff distance and was updated every 10
steps. For each peptide, the system was heated from 150 to 276 K
over 5 ps; assignment of the starting random velocities was according
to a Gaussian distribution. The resulting structure was equilibrated at
276 K for an additional 5 ps. The molecular dynamics calculations
were then performed for a total time of 300 ps, with coordinates,
velocities, and energies saved every 1 ps for further analysis.
CD Studies. CD measurements were obtained on an Aviv (model
62DS) spectrometer. The peptide/peptidomimetics were dissolved in
a buffer solution formed from 1 mM each of sodium phosphate, sodium
citrate, and boric acid. The pH of the samples was adjusted at room
temperature with HCl or NaOH. Peptide concentrations were deter-
mined by UV absorption assays57 and by quantitative amino acid
analysis. The concentration of peptide/peptidomimetics studied was
in the range of 2-13 mM. The data presented represent an average of
4 scans per sample with a time constant of 2 s, bandwidth of 1 nm,
and sampling every 0.5 nm from 260 to 195 nm. For temperature-

Conformations of RNase C-Peptide DeriVatiVes
J. Am. Chem. Soc., Vol. 120, No. 37, 1998 9443
dependence studies, the samples were equilibrated for at least 10 min
before data acquisition. Likewise, for the pH-dependence experiments
conducted at 3 °C, the samples were allowed to equilibrate for at least
Awardee (JFRA-577). We thank Larry Dangott and Jenny
Johnson for amino acid analysis. We also thank Jian Zhang for
MALDI analysis and Yangbo Feng for helpful discussions.
10 min prior to measurement of θ222
.
Supporting Information Available: Comparison of peptide
concentration determination by UV and AAA and parameter
sets and topology files for FiFi (7 pages, print/PDF). See any
current masthead page for ordering information and Web access
instructions.
Acknowledgment. K.B. gratefully acknowledges support
from NIH and The Robert A. Welch Foundation; an NIH
Research Career Development Award; and The Alfred P. Sloan
Foundation for a fellowship. D.M.S. thanks NIH for a predoc-
toral fellowship and TAMU for a Minority Merit Fellowship.
J.M.S. is an American Cancer Society Junior Faculty Research
JA981153D