Creation and investigation of protein-core mimetics with parallel and antiparallel aligned amino acids

Article abstract of DOI:10.1021/jo0479563

Mimetic protein cores were created that align a set of L-Phe, D-Phe, or L-Leu residues in a parallel or an antiparallel arrangement in chloroform. Not all cores show a single conformation at room temperature. Stable structures require a synergistic relati

Full text of DOI:10.1021/jo0479563

Creation and Investigation of Protein-Core Mimetics with Parallel
and Antiparallel Aligned Amino Acids
Juris Fotins and David B. Smithrud*
Department of Chemistry, University of Cincinnati, Cincinnati, Ohio 45221-0172
david.smithrud@uc.edu
Received November 17, 2004
Mimetic protein cores were created that align a set of L-Phe, D-Phe, or L-Leu residues in a parallel
or an antiparallel arrangement in chloroform. Not all cores show a single conformation at room
temperature. Stable structures require a synergistic relationship between the H-bonding groups
and the residues within the core. The spatial arrangement of the side chains dictates whether a
zippered or a crossed pattern of H-bonds is observed for these cores. Variable-temperature 1H NMR
experiments were used to determine the strengths of the H-bonds. The existence of H-bonds was
verified through FTIR spectroscopic analysis. Large temperature coefficients exist for some protons
of aromatic rings that are held in a T-shaped arrangement. A comparison of these temperature
coefficients shows that a more stable core is obtained by combining benzenoid and nitrobenzenoid
rings as compared to benzenoid rings. Structures were determined using a combination of 2D NMR
analysis and molecular modeling.
Introduction
synthetic scaffold to align aromatic rings in a T-shaped
arrangement (Figure 1). Formation of the core would
theoretically not require a large entropic penalty for
folding. A T-shaped (or edge to face) arrangement of
aromatic rings has been observed in the X-ray structures
Creating compounds with protein-like secondary struc-
tures remains a challenging endeavor. Impressive results
have been obtained using nonnatural oligomers that fold
into well-defined three-dimensional structures referred
to as foldamers1 and â-turn mimetics2 that align peptide
or peptidomimetic chains to form â-sheet structures. Our
approach in creating compounds with protein-like sec-
ondary structures is to use an aromatic or aliphatic
assembly to stabilize the interactions between amino
acids or peptides. These protein-core mimetics (PCMs)
were inspired by Kelly’s isoquinoline-based â-sheet mi-
metics.3 In this system, hydrophobic side chains fold back
onto the isoquinoline ring, providing additional stabiliza-
tion energy for â-sheet formation. We postulated that a
more stabilized core would be obtained by using a
(2) (a) Blomberg, D.; Hedenstrom, M.; Kreye, P.; Sethson, I.;
Brickmann, K.; Kihlberg, J. J. Org. Chem. 2004, 69, 3500-3508. (b)
Tatko, C. D.; Waters, M. L. J. Am. Chem. Soc. 2004, 126, 2028-2034.
(c) Nowick, J. S.; Brower, J. O. J. Am. Chem. Soc. 2003, 125, 876-
877. (d) Nowick, J. S.; Lam, K. S.; Khasanova, T. V.; Kemnitzer, W.
E.; Maitra, S.; Mee, H. T.; Liu, R. W. J. Am. Chem. Soc. 2002, 124,
4972-4973. (e) Kaul, R.; Deechongkit, S.; Kelly, J. W. J. Am. Chem.
Soc. 2002, 124, 11900-11907. (f) Wang, W.; Yang, J.; Ying, J.; Xiong,
C.; Zhang, J.; Cai, C.; Hruby, V. J. J. Org. Chem. 2002, 67, 6353-
6360. (g) Fernandez, M. M.; Diez, A.; Rubiralta, M.; Montenegro, E.;
Casamitjana, N.; Kogan, M. J.; Giralt, E. J. Org. Chem. 2002, 67,
7587-7599. (h) Eguchi, M.; Shen, R. Y. W.; Shea, J. P.; Lee, M. S.;
Kahn, M. J. Med. Chem. 2002, 45, 1395-1398. (i) Hoffmann, T.; Lanig,
H.; Waibel, R.; Gmeiner, P. Angew. Chem., Int. Ed. 2001, 40, 3361-
3364. (j) Lacroix, E.; Kortemme, T.; de la Paz, M. L.; Serrano, L. Curr.
Opin. Struct. Biol. 1999, 9, 487-493. (k) Blanco, F.; Ram´ırez-Alvarado;
M.; Serrano, L. Curr. Opin. Struct. Biol. 1998, 8, 107-111. (l) Gellman,
S. H. Curr. Opin. Chem. Biol. 1998, 2, 717-725. (m) Smith, C. K.;
Regan, L. Acc. Chem. Res. 1997, 30, 153-161.
(1) (a) Cheng R. P. Curr. Opin. Struct. Biol. 2004, 14, 512-520. (b)
Hill, D. J.; Mio, M. J.; Prince, R. B.; Hughes, T. S.; Moore, J. S. Chem.
Rev. 2001, 101, 3893-4011. (c) Cubberley, M. S.; Iverson, B. L. Curr.
Opin. Chem. Biol. 2001, 5, 650-653. (d) Gellman, S. H. Acc. Chem.
Res. 1998, 31, 173-180.
(3) Tsang, K. Y.; Diaz, H.; Graciani, N.; Kelly, J. W. J. Am. Chem.
Soc. 1994, 116, 3988-4005.
10.1021/jo0479563 CCC: $30.25 © 2005 American Chemical Society
Published on Web 04/23/2005
4452
J. Org. Chem. 2005, 70, 4452-4459

Investigation of Protein Core Mimetics
FIGURE 2. (A) The original PCM that displays interactions
between both aromatic side chains and one of the scaffold’s
aromatic rings and two isoenergetic H-bonds. (B) The new
PCM also contains a scaffold and H-bonding plane, but its
internal amines are not methylated.
FIGURE 1. The number of aromatic rings held in a T-shaped
arrangement determines whether a zippered or a crossed
pattern of H-bonds is formed. A nonsynergistic relationship
between the core and H-bonding residues, however, results in
multiple structures.
In the initial study of our PCMs,11 we also observed
the importance of aromatic rings in the formation of core
structures. Aromatic rings provide for more structure and
H-bonds between Phe residues, when positioned in a
T-shaped arrangement, as compared to scaffolds linked
to Leu or to a scaffold without an aromatic core. The
internal amides of these compounds were methylated
(Figure 2A) to ensure that a 7-membered H-bonded ring
would not form between one of these amides and a
carbonyl group. We were concerned that this H-bonded
ring would control structure formation and not the
interactions between aromatic or aliphatic groups. In this
report, we describe the properties of PCMs without these
Me groups. We wanted to determine whether a set of core
residues could work synergistically with H-bonding resi-
dues to form a single stable structure. Nitrated com-
pounds were created and investigated to determine
whether a change in the electrostatic forces of a core could
enhance the strength of aromatic-aromatic interactions.
We found that the nature of the side chain and its
chirality determine core structures with either a zippered
or crossed pattern of H-bonded rings (Figure 1). In some
cases, however, strong aromatic-aromatic interactions
can work against the H-bonds, resulting in multiple
structures. Investigation of these flexible PCMs will lead
to a better understanding of the dynamic instability
observed in some peptides, such as in the mutated zinc
finger peptides.
of aromatic compounds,4 observed in proteins,5 studied
using model systems,6 and investigated through theoreti-
cal calculations.6e,7
The design of the PCM is based on the small hydro-
phobic cores (typically containing Tyr, Phe, and Leu)
found in the zinc finger peptides, which are often used
as a model system for protein stability studies. Although
small, these cores provide for a substantial amount of
structural stability. Imperiali showed that modified
fingers without a metal binding site can give similar
structures as the native finger.8 Mutating the highly
conserved Phe of zinc fingers to Leu in a model of the
Xfin-31 finger peptide9 and in a model finger of the
human Y-encoded protein ZFY10 reduces their stability.
Interestingly, in the latter study, the instability is caused
by an increase in the finger’s dynamics.
(4) Desiraju, G. R.; Gavezzotti, A. J. Chem. Soc., Chem. Commun.
1989, 621-623.
(5) (a) Thomas, A.; Meurisse, R.; Charloteaux, B.; Brasseur, R.
Proteins 2002, 48, 628-634. (b) Thomas, A.; Meurisse, R.; Brasseur,
R. Proteins 2002, 48, 635-644. (c) McGaughey, G. B.; Gagne, M.;
Rappe, A. K. J. Biol. Chem. 1998, 273, 15458-15463. (d) Hunter, C.
A.; Singh, J.; Thornton, J. M. J. Mol. Biol. 1991, 218, 837. (e) Burley,
S. K.; Petsko, G. A. J. Am. Chem. Soc. 1986, 108, 7995-8001. (g)
Burley, S. K.; Petsko, G. A. Science 1985, 229, 23-28.
(6) (a) Waters, M. L. Biopolymers 2004, 76, 435-445. (b) Meyer, E.
A.; Castellano, R. K.; Diederich, F. Angew. Chem., Int. Ed. 2003, 42,
1210-1250. (c) Waters, M. L. Curr. Opin. Chem. Biol. 2002, 6, 736-
741. (d) Rashkin, M. J.; Waters, M. L. J. Am. Chem. Soc. 2002, 124,
1860-1861. (e) Hunter, C. A.; Lawson, K. R.; Perkins, J.; Urch, C. J.
J. Chem. Soc., Perkin Trans. 2 2001, 651-669. (f) Carver, F. J.; Hunter,
C. A.; Jones, P. S.; Livingstone, D. J.; McCabe, J. F.; Seward, E. M.;
Tiger, P.; Spey, S. E. Chem. Eur. J. 2001, 7, 4854-4862. (g) Adams,
H.; Hunter, C. A.; Lawson, K. R.; Perkins, J.; Spey, S. E.; Urch, C. J.;
Sanderson, J. M. Chem. Eur. J. 2001, 7, 4863-4877. (h) Sindkhedkar,
M. D.; Mulla, H. R.; Cammers-Goodwin, A. J. Am. Chem. Soc. 2000,
122, 9271-9277. (i) Liu, D. H.; Williamson, D. A.; Kennedy, M. L.;
Williams, T. D.; Morton, M. M.; Benson, D. R. J. Am. Chem. Soc. 1999,
121, 11798-11812. (j) Heaton, N. J.; Bello, P.; Herradon, B.; del Campo,
A.; Jimenez-Barbero, J. J. Am. Chem. Soc. 1998, 120, 12371-12384.
(k) Kim, E.; Paliwal, S.; Wilcox, C. S. J. Am. Chem. Soc. 1998, 120,
11192-1193.
Results
Physical Properties of Mimetics with Parallel
Aligned Amino Acids. Attaching L-Phe, D-Phe, or L-Leu
to an aromatic scaffold gives a series of PCMs that align
the amino acids in a parallel fashion (Figure 2B). Parallel
means that both amino acids are attached to the scaffold
through their amino terminus. The physical properties
of the compounds were investigated using 1D NMR
analysis to determine H-bond stabilities and 2D NMR
analysis for structural information. To verify the exist-
ence of H-bonds, the FTIR spectrum of each compound
was examined for a H-bonded N-H stretching band,12
(7) (a) Sinnokrot, M. O.; Sherrill, C. D. J. Phys. Chem. A 2004, 108,
10200-10207. (b) Sinnokrot, M. O.; Sherrill, C. D. J. Am. Chem. Soc.
2004, 126, 7690-7697. (c) Morozov, A. V.; Misura, K. M. S.; Tsemekh-
man, K.; Baker, D. J. Phys. Chem. B 2004, 108, 8489-8496. (d)
Gervasio, F. L.; Chelli, R.; Procacci, P.; Schettino V. Proteins 2002,
48, 117-125. (e) Tsuzuki, S.; Honda, K.; Uchimaru, T.; Mikami, M.;
Tanabe, K. J. Am. Chem. Soc. 2002, 124, 104-112 and references
therein.
which is generally observed between 3300 and 3360 cm-1
A free N-H band is observed between 3410 and 3450
cm-1
.
(8) Struthers; M. D.; Cheng, R. P.; Imperiali, B. J. Am. Chem. Soc.
1996, 118, 3073-3081.
.
(9) Mortishiresmith, R. J.; Lee M. S.; Bolinger, L.; Wright P. E.
FEBS Lett. 1992, 296, 11-15.
(10) Lachenmann, M. J.; Ladbury, J. E.; Phillips, N. B.; Narayana,
N.; Qian, X. Q.; Stern, A. S.; Weiss, M. A. J. Mol. Biol. 2002, 316, 969-
989.
(11) Turk, J. A.; Smithrud, D. B. J. Org. Chem. 2001, 66, 8328-
8335.
J. Org. Chem, Vol. 70, No. 11, 2005 4453

Fotins and Smithrud
FIGURE 3. An overlay of the 1H NMR spectra of (L,D)-
Pheparallel and (L,L)-Pheparallel, showing that a correct match
ing of core with H-bonding residues will lead to a stable
structure.
The first compound constructed contains one L-Phe and
one D-Phe. This combination of amino acids, when
methylated and attached to the original scaffold (Figure
2A), produces a stable, unique PCM that displays two
stable isoenergetic H-bonds. Both aromatic side chains
make contact with the scaffold’s aromatic ring. The
addition of one L-and one D-Phe to the new scaffold gave
diastereomers, which were separated by HPLC. 2D NMR
analysis of one diastereomer revealed the predicted
multiple NOE cross-peaks between the scaffold’s aro-
matic ring and the aromatic side chains and a cross-peak
between the protons of the external Me groups. The FTIR
spectrum of the (L,D)-Pheparallel compound contains ab-
sorbance bands that are consistent with H-bonded N-H
groups (data not shown). This compound, however, exists
as multiple conformers at room temperature (Figure 3),
confirming our concerns that the formation of a 7-mem-
bered H-bonded ring would disrupt structure formation.
H-bond stability was not determined because multiple
conformations exist from room temperature to 240 K.
In an attempt to obtain stable and thus useful PCMs,
only L-amino acids were attached to the scaffold. The
(L,L)-Pheparallel compound shows a single conformation on
the NMR time scale from 240 to 300 K (300 K spectrum
is shown in Figure 3). Plotting the chemical shifts of the
amide protons against the temperature of the experi-
ments produced straight lines. Temperature coefficients
(TCs) derived from the slope of these lines provide a
measure of H-bond stability in CDCl3.12a,b,f,13 According
to the literature, two N-H’s of (L,L)-Pheparallel (TC ) -1.3
and -2.2 ppb/K, Figure 4) would be considered as being
either ring-locked (H-bonds in cyclic peptides) or shel-
tered from the solvent. The other two coefficients of -3.3
and -3.7 ppb/K are in the range found for non-H-bonded
N-H’s. Supporting evidence for the existence of H-bonds
is the observation of an intense H-bonded N-H band in
the FTIR spectrum of (L,L)-Pheparallel (Figure 5). N-Ac-Phe-
CONHMe, dissolved in CHCl3 to give the same concen-
tration of 3 mM, shows only a single non-H-bonded N-H
band.
FIGURE 4. Schematic drawings of the compounds showing
key NOE cross-peaks (double-headed arrows). Temperature
coefficients (standard deviations are less than (0.1) for amide
N-H and aromatic Ar-H protons are given. Highlighted in
(L,L)-PheNO2parallel are the observed diastereomeric differences.
Analysis of NOESY and ROESY spectra of (L,L)-
Pheparallel revealed key NOE cross-peaks: two between
one aromatic side chain and the scaffold’s aromatic ring
and one between the methyl groups of the external
amides. The chemical shift of the ortho proton of the
scaffold that is correlated to the side chain is shielded
(δAr-H ) 5.48 ppm) and greatly affected by temperature
(TC ) 5.7 ppb/K). These results are consistent with a
T-shaped arrangement of aromatic rings. As the solution
was cooled, the population of compounds with a scaffold
proton in the shielding cone of the side chain’s aromatic
ring increased. To further demonstrate the importance
of the aromatic-aromatic interactions for structure,
compound 10 (Figure 4; for formation see Supporting
Information), which does not have an aromatic ring, was
constructed and investigated. In chloroform, a single
H-bonded ring possibly forms (TC ) -1.9 ppb/K), and
the other N-H’s are not H-bonded. Only a few NOE
cross-peaks exist in its ROESY and NOESY spectra and
none occur between the amino acid protons. Although a
single conformer exists, this scaffold does not promote
an alignment of the amino acids or extend the H-bonded
network.
Although these findings suggest that strong aromatic-
aromatic interactions in (L,L)-Pheparallel contribute to
structural stability, another possibility is that the aro-
matic side chain is in a T-shaped arrangement because
of restricted conformational freedom. In this case, stabil-
ity would be a result of steric constraints. To explore this
possibility, the electronic property of the scaffold was
changed by the addition of a nitro group. Favorable
aromatic-aromatic interactions should increase7b with
(12) (a) Luppi, G.; Lanci, D.; Trigari, V.; Garavelli, M.; Garelli, A.;
Tomasini, C. J. Org. Chem. 2003, 68, 1982-1993. (b) Trabocchi, A.;
Occhiato, E. G.; Potenza, D.; Guarna, A. J. Org. Chem. 2002, 68, 1982-
1993. (c) Yang, J. H.; Christianson, L. A.; Gellman, S. H. Org. Lett.
1999, 1, 11-14. (d) Nowick, J. S.; Mahrus, S.; Smith, E. M.; Ziller, J.
W. J. Am. Chem. Soc. 1996, 118, 1066-1072. (e) Gardner, R. R.; Liang,
G.-B.; Gellman, S. H. J. Am. Chem. Soc. 1995, 117, 3280-3281. (f)
Gellman, S. H.; Dado, G. P.; Liang, G.-B.; Adams, B. R. J. Am. Chem.
Soc. 1991, 113, 1164-1173.
4454 J. Org. Chem., Vol. 70, No. 11, 2005

Investigation of Protein Core Mimetics
FIGURE 5. FTIR spectral data from the N-H stretching region of the PCMs and N-Ac-phenylalanine methyl amide. All samples
were 3 mM in CHCl3 at 298 K. Each spectrum had its baseline corrected, and the absorbance of CHCl3 was subtracted.
(L,L)-PheNO2parallel. The nitro group pulls electron density
out of the ring, making its Ar-H more acidic and
attracted to the electron-rich face of a stacked aromatic
ring. (L,L)-PheNO2parallel was isolated and investigated as
an approximately 50:50 diastereomeric mixture. Each
diastereomer exists as a single conformer on the NMR
time scale from 240 to 300 K. Both external amides are
strongly H-bonded (TC ) -6.2 and -6.6 ppb/K, Figure
4). One internal amide forms a moderately stable H-bond,
and the other internal amide is ring-locked or shielded
(TC ) -4.6 and -1.0 ppb/K, respectively). As seen with
the (L,L)-Pheparallel compound, one Ar-H of the scaffold
is significantly shielded (δAr-H ) 5.62 ppm) at room
temperature. The chemical shift of this proton, however,
is more temperature-sensitive (TC ) 7.1 ppb/K) than the
one observed for the (L,L)-Pheparallel compound (TC ) 5.7
ppb/K; ∆TC ) 1.4 ppb/K). This larger temperature
coefficient suggests that the nitrated benzenoid ring
forms a more favorable aromatic-aromatic interaction
than the benzenoid ring. Further evidence for the exist-
ence of more favorable aromatic-aromatic interactions
was the detection of other temperature-sensitive chemical
shifts for more than one scaffold Ar-H (TC ) 2.6, 1.4,
and 0.5 ppb/K) and NOE cross-peaks between the scaf-
fold’s aromatic ring and both aromatic side chains.
Another different feature observed for (L,L)-PheNO2parallel
as compared to (L,L)-Pheparallel is the existence of an NOE
cross-peak between each external Me group and the
R-proton of the opposite amino acid (Figure 4).
To further demonstrate the importance of aromatic-
aromatic interactions, the properties of the compounds
that contain phenylalanine residues are compared to a
compound that contains leucine residues. In our previous
study,11 we found that aromatic-aromatic interactions
provided for a different structure and H-bond stability
of PCMs when compared to aromatic-aliphatic interac-
tions. (L,L)-Leuparallel (Figure 4) shows a single conformer
on the NMR time scale from 240 to 300 K. Three strong
H-bonds exist (TC ) -8.0, -6.4, and -5.1 ppb/K), and
the fourth N-H is not H-bonded (TC ) -2.9 ppb/K). The
properties of (L,L)-Leuparallel are remarkably similar to the
properties observed for (L,L)-PheNO2parallel. Both com-
pounds show NOE cross-peaks between the external Me
groups and between each external Me group and the
R-C-H of the opposite amino acid. One major difference
is that only one side chain of (L,L)-Leuparallel is positioned
close to the scaffold’s aromatic ring. There are also
differences in the stability of the H-bonded rings. (L,L)-
Leuparallel contains a very stable 7-membered H-bonded
ring (TC ) -8.0 ppb/K), and a close to zero temperature
coefficient is not observed.
Physical Properties of Mimetics with Antiparal-
lel Aligned Amino Acids. PCMs with antiparal-
lel aligned L-Phe or D-Phe were constructed (Figure 6)
to further test the ability of core interactions to control
structure. Antiparallel means that one phenylalanine
is attached through its carboxylate and the other is
attached through its amine. The N-H’s are more directly
aligned with the carbonyl oxygen atom of the op-
posite amino acid, which could have produced more stable
H-bonded rings as compared to the parallel com-
pounds. Separation of diastereomers 6a and 6b (for
formation see Supporting Information) led to two dia-
stereomeric sets called 1 and 2 (Figure 7). The addition
of a L-Phe or D-Phe to each diastereomer produced four
antiparallel aligned PCMs, which will be referred to as
(L,L)1-Pheantiparallel, (L,D)1-Pheantiparallel, (L,L)2-Pheantiparallel
,
and (L,D)2-Pheantiparallel
.
One compound from each diastereomeric set (1 and 2)
shows a single conformation on the NMR time scale
from 240 to 300 K. The other two compounds dis-
play multiple conformations. For the structurally stable
(L,L)-Pheantiparallel compound (Figure 6), a stable H-bond
and a very stable H-bond exist for one external N-H (TC
) -6.3 ppb/K) and the internal N-H of the other
phenylalanine residue (TC ) -7.6 ppb/K), respectively.
For the structurally stable (L,D)-Pheantiparallel compound,
both external amides are involved in moderately stable
H-bonds (TC ) -4.5 and -4.3 ppb/K) and one internal
N-H forms a very stable H-bond (TC ) -7.0 ppb/K). For
both compounds, the chemical shift of one internal N-H
is temperature-insensitive, which suggests that it was
shielded or existed in a ring-locked H-bond. The FTIR
spectra for the anitparallel compounds (Figure 6) are
J. Org. Chem, Vol. 70, No. 11, 2005 4455

Fotins and Smithrud
FIGURE 6. Key NOE cross-peaks (double-headed arrows), temperature coefficients for amide N-H and aromatic Ar-H protons,
and FTIR spectral data from the N-H stretching region are given for the antiparallel compounds.
FIGURE 7. The diastereomeric sets of the antiparallel compounds, a cartoon depiction of their calculated stable structures
highlighting the orientation of the aromatic rings (side chain rings are given as rectangles), and whether a single conformer is
observed.
significantly different than those observed for the parallel
compounds. Multiple peaks are observed for these com-
pounds at a lower frequency than normally found for
H-bonded amides (low-frequency bands have been ob-
served for â-sheet mimetics).12d The same IR cell was
used for all spectra shown in Figures 5 and 6, and the
maximum absorbances were within 0.07-0.1 absorbance
units. Therefore, the low-frequency bands observed for
the antiparallel cores are not an artifact of the IR cell or
the experiment. Multiple NOE cross-peaks between both
4456 J. Org. Chem., Vol. 70, No. 11, 2005

Investigation of Protein Core Mimetics
The necessity of this pattern explains the lack of an
observed single conformer for (L,D)-Pheparallel (Figure 3).
One low-energy conformer observed in the model study
shows a zippered pattern of H-bonds. As seen in this
structure (Figure 9), one aromatic ring forms a T-shaped
arrangement with the scaffold’s aromatic ring and the
other aromatic ring is positioned to also interact with the
core. For one conformer, we observed NOE cross-peaks
between the scaffold aromatic ring and an aromatic side
chain and the existence of a shielded scaffold aromatic
proton (5.82 ppm, Figure 3), which are consistent with
T-shaped aromatic rings. We speculate that the formation
of the second T-shaped arrangement gives the observed
multiple structures. According to the molecular modeling
results, when the second aromatic side chain forms a
T-shaped arrangement with the scaffold’s aromatic ring,
its internal carbonyl shifts up and away from the face of
this aromatic ring. Both internal carbonyls are now
positioned up, and the 7-membered H-bonded ring breaks.
Apparently, the stabilizing energies provided by the
7-membered H-bonded ring and the second T-shaped
aromatic interaction are similar. Because these two
interactions promote different structures, a nonsyner-
gistic relationship exists for structure formation.
The position of the amino acid side chains relative to
the scaffold’s aromatic ring dictates which pattern of
H-bonds forms. This statement is readily evident when
comparing the properties of (L,L)-Pheparallel to (L,L)-
PheNO2parallel. In the latter compound, both aromatic side
chains are positioned close to the scaffold’s aromatic ring.
According to the molecular modeling studies, when the
second aromatic ring is placed near the scaffold’s ring
the external amide of this amino acid rotates toward the
internal carbonyls (Figure 9). This rotation gives (L,L)-
PheNO2parallel a crossed pattern of H-bonds, which is
consistent with its three stable H-bonds and the NOE
cross-peaks between each external methyl group to the
opposite amino acid’s R-proton. The lack of this cross-
peak and the existence of an NOE cross-peak between
the external methyl groups suggests that the (L,L)-
Pheparallel compound forms a zippered pattern. Although
the temperature coefficients of the possible H-bonded
N-H’s observed for (L,L)-Pheparallel are surprisingly close
to zero, the observation of a strong H-bonded N-H
stretching band in the FTIR spectrum supports the
existence of a H-bond. Apparently, the 7- and 12-
membered H-bonded rings are the preferred pattern
when one set of benzenoid rings interacts. The less stable
crossed pattern is forced to form when both aromatic side
chains interact with the scaffold’s aromatic ring. This
result shows that nitrobenzenoid-benzeniod interactions
are more stabilizing than benzenzoid-benzenoid interac-
tions. The enhanced stability is most likely a result of a
more stable H-bond15 or a greater π-σ interaction16 and
consistent with the results obtained by Sherrill,7b who
performed a computation analysis of the interaction
energies between substituted aromatic rings. Any pos-
sible steric hindrance imposed by the NO2 group would
have only lessened the favorable aromatic interaction
energies.
FIGURE 8. A zippered (dashed double-headed arrow) or a
crossed (solid double-headed arrow) pattern of H-bonds exists
between the amino acids of the parallel aligned compounds
and the antiparallel aligned compounds. The numbers indicate
the length of a H-bonded ring. R is a set of L-Phe, D-Phe, or
L-Leu and X is H or NO2.
aromatic side chains and the scaffold’s aromatic ring are
observed for both compounds. For (L,D)-Pheantiparallel, a
unique cross-peak exists between its R-protons and no
cross-peak is found between its external Me groups. For
(L,L)-Pheantiparallel, a NOE cross-peak is observed between
the external Me groups. As seen with the parallel
compounds, the spatial arrangement of the aromatic side
chains of the antiparallel compounds has a pronounced
affect on their structures.
Discussion
The experimental results show that not all PCMs have
a single conformation in chloroform. The structural stable
compounds show a different number of core interactions
and H-bonding patterns depending on whether they
contained phenylalanine residues, leucine residues, or a
substituted aromatic ring. To understand the synergistic
relationship between the interactions, a Monte Carlo
method was used to search the conformational space of
the compounds (MMFF94 force field, package procedure
of SpartanPro).14 A series of low-energy conformers were
obtained that contained a zippered or a crossed pattern
of H-bonds (Figures 1 and 8). One or the other pattern is
consistent with the observed properties of the PCMs. For
the parallel compounds, the zippered pattern contains
7- and 12-membered H-bonded rings, and the crossed
pattern contains one 7-membered and two 10-membered
H-bonded rings. For the antiparallel compounds, the
zippered pattern contains 8- and 12-membered H-bonded
rings, and the crossed pattern contains 8-, 11-, and 12-
membered H-bonded rings. For the antiparallel com-
pounds, the length of each H-bonded ring was considered
independent of the other rings. Both patterns of H-bonds
require the internal carbonyls to form an up-down
pattern.
(13) (a) de la Paz, M. L.; Ellis, G.; Perez, M.; Perkins, J.; Jimenez-
Barbero, J.; Vicent, C. Eur. J. Org. Chem. 2002, 840-855. (b) Crich,
D.; Dudkin, V. J. Am. Chem. Soc. 2001, 123, 6819-6825. (c) Belvisi,
L.; Gennari, C.; Madder, A.; Mielgo, A.; Potenza, D.; Scolastico, C. Eur.
J. Org. Chem. 2000, 695-699. (d) Soth, M. J.; Nowick, J. S. J. Org.
Chem. 1999, 64, 276-281. (e) Smith, A. B.; Favor, D. A.; Sprengeler,
P. A.; Guzman, M. C.; Carroll, P. J.; Furst, G. T.; Hirschmann, R.
Bioorg. Med. Chem. 1999, 7, 9-22. (f) Yang, J.; Gellman, S. H. J. Am.
Chem. Soc. 1890, 12, 9090-9091. (g) Jones, I. G.; Jones, W.; North,
M. J. Org. Chem. 1998, 63, 1505-1513. (h) Fowler, P.; Bernet, B.;
Vasella, A. Helv. Chim. Acta 1996, 79, 269-287. (i) Vasella, A.; Witzig,
C. Helv. Chim. Acta 1995, 78, 1971-1982.
(15) Burley, S. K.; Petsko, G. A. Adv. Protein Chem. 1988, 39, 125-
189.
(16) Hunter, C. A.; Sanders, J. K. M. J. Am. Chem. Soc. 1990, 112,
5525-5534.
(14) Wavefunction, Inc.: Irvine, CA.
J. Org. Chem, Vol. 70, No. 11, 2005 4457

Fotins and Smithrud
FIGURE 9. Low-energy structures obtained from molecular modeling studies that are consistent with the observed properties
of the compounds. For (L,D)-Pheparallel, moving the second aromatic side chain to interact with the scaffold’s aromatic ring forces
the internal carbonyl to rotate up (indicated by arrows), which breaks the 7-membered H-bonded ring. The R-proton is shown to
indicate that the aromatic interactions do not require bond rotation around the amide bond. For (L,L)-PheNO2parallel as compared
to (L,L)-Pheparallel, the more favorable aromatic interactions drives the second aromatic side chain to the scaffolds ring, which
rotates the amide bond (indicated by arrows), giving the crossed pattern of H-bonds. This phenomenon also occurs for (L,D)2-
Pheantiparallel. T indicates a T-shaped arrangement of rings, and H-atoms not involved in H-bonds are removed for clarity.
Differences in aromatic-aromatic interactions as com-
pared to aromatic-aliphatic interactions are readily
apparent when comparing the properties of (L,L)-Pheparallel
and (L,L)-Leuparallel. As predicted, there is a closer associa-
tion of the aromatic core residues as evident by the
greater number of NOE cross-peaks for (L,L)-Pheparallel as
compared to (L,L)-Leuparallel. The patterns of H-bonds are
also different. Surprisingly, the structures and properties
of (L,L)-Leuparallel are very similar to those of (L,L)-
PheNO2parallel. Both compounds display the required three
stable H-bonds and the key NOE cross-peaks between
each external methyl group to the opposite amino acid’s
R-proton. Does this mean that benzenoid-aliphatic in-
teractions are more stable than the interactions between
benzenoid rings and equivalent in energy to interactions
between nitrobenzenoid-benzenoid rings? The answer is
probably no. Only a single NOE cross-peak is observed
between one isobutyl side chain and the scaffold’s aro-
matic ring for (L,L)-Leuparallel, which suggests that exten-
sive aliphatic-aromatic interactions do not occur. Most
likely, the very stable 7-membered H-bonded ring of (L,L)-
Leuparallel forced the crossed pattern of H-bonds. The N-H
in this ring has a temperature coefficient (-8.0 ppb/K)
that is substantially smaller than the ones observed for
the other parallel compounds and even smaller than the
ones observed for the antiparallel compounds. Addition-
ally, steric interactions between the side chains most
likely kept them from both residing above the H-bonding
plane.
To determine whether the observed structures of the
PCMs with antiparallel aligned amino acids depend on
the spatial arrangement of the side chains, we had to
assign the stable diastereomers. One diastereomer con-
tains two L-Phe and the other contains one L-Phe and
one D-Phe. According to the synthetic route, all com-
pounds contain a L-Phe that is attached to the scaffold
through its amine.
Because these diastereomers were separated before the
attachment of the second amino acid, we know that a
compound belongs to one of the two sets of diastereomers,
which are called set 1 and 2 (Figure 7). For example, one
compound containing two (L)-Phe’s could either be (L,L)1-
Pheantiparallel or (L,L)2-Pheantiparallel. Results obtained from
the 2D NMR experiments and molecular modeling stud-
4458 J. Org. Chem., Vol. 70, No. 11, 2005

Investigation of Protein Core Mimetics
ies were used to assign the conformationally stable
antiparallel compounds. Molecular modeling results show
that a stable zippered pattern of H-bonds should be
observed for one diastereomer of the (L,L)1-Pheantiparallel
and (L,L)2-Pheantiparallel pair with both aromatic side chains
in contact with the scaffold’s aromatic ring (Figure 9 and
diastereomer A of Figure 7). The other compound (dia-
stereomer C) would have had both side chains above the
H-bonding plane. (L,L)1-Pheantiparallel displays a zippered
pattern and NOE cross-peaks between both aromatic side
chains and the scaffold’s aromatic ring and thus is
assigned as the stable diastereomer A. The assignment
of (L,L)1-Pheantiparallel removes the possibility that (L,D)1-
Pheantiparallel is a stable PCM because only one compound
of this diastereomic set displays a single conformer. Thus,
the other stable diastereomer has to be the other com-
pound that contains one L-Phe and one D-Phe, which we
call (L,D)2-Pheantiparallel (diastereomer D). For this com-
pound, we observe multiple NOE cross-peaks between
both aromatic side chains and the scaffold’s aromatic
ring, a cross-peak between the R-protons, and no cross-
peak between the external Me groups. A low-energy
structure consistent with these results was obtained in
a molecular modeling study (Figure 9) by constraining
the aromatic rings at a distance observed for T-shaped
aromatic interactions (5 Å).5b This structure nicely shows
both external amides being H-bonded, a close proximity
of the R-protons, and far-separated external Me groups.
The modeling results also demonstrates that when the
second aromatic ring is positioned near the scaffold’s
aromatic ring, its external carbonyl rotates back toward
the scaffold, giving the crossed pattern of H-bonds. This
structure is formed. Most likely, two sets of T-shaped
arrangements of rings is formed for diastereomer B as
well. Unlike with diastereomer D, molecular modeling
results show that favorable H-bonds are disrupted when
the second aromatic side chain is placed near the scaf-
fold’s aromatic ring of diastereomer B. One major differ-
ence between diastereomers B and D is the nature of the
H-bonding atoms in the 7-membered ring for the amino
acid side chain that is drawn close to the scaffold in
Figure 7. For diastereomer D, its carbonyl is involved in
the 7-membered ring. For diastereomer B, its N-H would
be involved in the 7-membered ring. Apparently, this
subtle difference is enough to disrupt a synergistic
relationship between the core and H-bonding residues for
diastereomer B.
Conclusion
A series of protein-core mimetics were constructed to
investigate the synergistic relationship between an as-
semblage of aliphatic or aromatic side chains and H-bond
forming groups of the amide backbone that is required
to produce stable protein cores. Although H-bonds are
very stable in chloroform (used to represent the core
environment) and could have dominated the PCM prop-
erties, interactions between aromatic or aliphatic groups
either stabilize the H-bonds, alter the pattern of H-bonds,
or disrupt the H-bonds. A greater number of NOE cross-
peaks are observed between core residues for a PCM with
a more acidic Ar-H, which is most likely caused by more
favorable T-shaped aromatic interactions. Besides the
observation of NOE cross-peaks, experimental evidence
for aromatic interactions is obtained through the obser-
vation of temperature-sensitive chemical shifts of aro-
matic protons. Our PCMs could potentially provide an
experimental measure of T-shaped aromatic interaction
energies. Aromatic interactions are not required to obtain
a stable core. The structural stability of the Leu-contain-
ing compound, however, most likely arose through a very
stable H-bond and steric constraints and not through core
interactions. The study of our PCMs demonstrates that
subtle differences in the spatial arrangement of the
interacting functional groups of amino acids have a great
impact on the properties of the mimetics. Considering
that protein cores are much more complex than our model
systems, it is not surprising that the a priori creation of
proteins remains a challenging endeavor.
phenomenon also occurs for (L,L)-PheNO2parallel
.
One consistent pattern observed for the Phe-containing
compounds is that at least one aromatic side chain has
to interact with the scaffold’s aromatic ring to obtain a
single, stable structure. In the cases of (L,L)-PheNO2parallel
and (L,D)-Phe2antiparallel, the second aromatic side chain
also makes contact with the scaffold’s aromatic ring,
resulting in a crossed pattern of H-bonds. In the case of
(L,L)-Phe1antiparallel, both aromatic side chains interact with
the scaffold’s aromatic ring, but a zippered pattern is
formed. The lack of a single conformer for (L,D)-Pheparallel
(Figure 3) demonstrates that a synergistic relationship
will not always occur. Another nonsynergistic relation-
ship appears to occur for (L,D)-Phe1antiparallel (diastereomer
B). Its instability was unexpected. A comparison of the
cartoon structures of diastereomer B with D as drawn
in Figure 7 suggests that both compounds should have
had a stable structure with one pair of interacting rings.
Molecular modeling results bolstered that expectation by
showing both diastereomers existing in a stable zippered
pattern of H-bonds with one T-shaped, aromatic arrange-
ment. Although both aromatic side chains of diastereo-
mer D interact with the scaffold’s aromatic ring, a stable
Acknowledgment. The authors thank the Univer-
sity of Cincinnati for funding this project
1
Supporting Information Available: 1-Dimensional H
NMR, ROESY and TOCSY spectra of the compounds and
synthetic procedures. This material is available free of charge
via the Internet at http://pubs.acs.org.
JO0479563
J. Org. Chem, Vol. 70, No. 11, 2005 4459