Cyclopropane-derived peptidomimetics. Design, synthesis, evaluation, and structure of novel HIV-1 protease inhibitors

Article abstract of DOI:10.1021/jm980033d

Toward establishing the general efficacy of using trisubstituted cyclopropanes as peptide mimics to stabilize extended peptide structures, the cyclopropanes 20a-d were incorporated as replacements into 9-13, which are analogues of the known HIV-1 protease

Full text of DOI:10.1021/jm980033d

J . Med. Chem. 1998, 41, 1581-1597
1581
Cyclop r op a n e-Der ived P ep tid om im etics. Design , Syn th esis, Eva lu a tion , a n d
Str u ctu r e of Novel HIV-1 P r otea se In h ibitor s
Stephen F. Martin,* Gordon O. Dorsey, Todd Gane, and Michael C. Hillier
Department of Chemistry and Biochemistry, The University of Texas, Austin, Texas 78712
Horst Kessler,* Matthias Baur, and Barbara Matha¨
Institut fu¨r Organische Chemie und Biochemie, Technische Universita¨t Mu¨nchen, D-85747 Garching, Germany
J ohn W. Erickson,* T. Nagarajan Bhat, Sanjeev Munshi, Sergei V. Gulnik, and Igor A. Topol
Structural Biochemistry Program, Frederick Biomedical Supercomputing Center, NCI-FCRDC, Frederick, Maryland 21702
Received J anuary 15, 1998
Toward establishing the general efficacy of using trisubstituted cyclopropanes as peptide mimics
to stabilize extended peptide structures, the cyclopropanes 20a -d were incorporated as
replacements into 9-13, which are analogues of the known HIV-1 protease inhibitors 14 and
15. The syntheses of 20a -d commenced with the Rh2[5(S)-MEPY]4-catalyzed cyclization of
the allylic diazoesters 16a -d to give the cyclopropyl lactones 17a -d in high enantiomeric
excess. Opening of the lactone moiety using the Weinreb protocol and straightforward
refunctionalization of the intermediate amides 18a -d gave 20a -d . A similar sequence of
reactions was used to prepare the N-methyl-2-pyridyl analogue 28. Coupling of 20a -d and
28 with the known diamino diol 22 delivered 9-13. Pseudopeptides 9-12 were found to be
competitive inhibitors of wild-type HIV-1 protease in biological assays having Kis of 0.31-0.35
nM for 9, 0.16-0.21 nM for 10, 0.47 nM for 11, and 0.17 nM for 12; these inhibitors were thus
approximately equipotent to the known inhibitor 14 (IC50 ) 0.22 nM) from which they were
derived. On the other hand 13 (Ki ) 80 nM) was a weaker inhibitor than its analogue 15 (Ki
) 0.11 nM). The solution structures of 9 and 10 were analyzed by NMR spectroscopy and
simulated annealing procedures that included restraints derived from homo- and heteronuclear
coupling constants and NOEs; because of the molecular symmetry of 9 and 10, a special protocol
to treat the NOE data was used. The final structure was checked by restrained and free
molecular dynamic calculations using an explicit DMSO solvent box. The preferred solution
conformations of 9 and 10 are extended structures that closely resemble the three-dimensional
structure of 10 bound to HIV-1 protease as determined by X-ray crystallographic analysis of
the complex. This work convincingly demonstrates that extended structures of peptides may
be stabilized by the presence of substituted cyclopropanes that serve as peptide replacements.
Moreover, the linear structure enforced in solution by the two cyclopropane rings in the
pseudopeptides 9-12 appears to correspond closely to the biologically active conformation of
the more flexible inhibitors 14 and 15. The present work, which is a combination of medicinal,
structural, and quantum chemistry, thus clearly establishes that cyclopropanes may be used
as structural constraints to reduce the flexibility of linear pseudopeptides and to help enforce
the biologically active conformation of such ligands in solution.
In tr od u ction
energy, and the knowledge of the preferred solution
conformation(s) of an oligopeptide does not necessarily
provide information regarding the conformation of that
ligand when bound to its respective host.1 In the
absence of three-dimensional structural data for the
ligand-receptor complex itself, information relating to
the biologically active conformation of oligopeptide
ligands has been generally obtained by introducing
conformational restraints by cyclization or by incorpora-
tion of peptide mimics at selected sites on the peptide
backbone.2-5 The resultant effects upon binding and
biological activity of the resulting pseudopeptides are
then evaluated in connection with available reduced
conformational space to provide insights into the bio-
logically active conformation of the ligand.
One of the exciting challenges in contemporary bioor-
ganic chemistry is determining the three-dimensional
structural details of the complexes of biologically active
oligopeptides and their derivatives with enzyme active
sites or receptors. Such studies provide critical infor-
mation that may be exploited to design new ligands with
even higher binding affinities for the macromolecular
binding site. When a direct investigation of the complex
is not possible, one common approach to gain insights
into the topological requirements within the host-guest
complex involves the elucidation of the biologically
active conformation of the peptide ligand. However,
linear oligopeptides interconvert rapidly in solution
between a number of conformations differing little in
S0022-2623(98)00033-8 CCC: $15.00 © 1998 American Chemical Society
Published on Web 04/17/1998

1582 J ournal of Medicinal Chemistry, 1998, Vol. 41, No. 10
Martin et al.
Sch em e 1
F igu r e 1. A rigid superimposition of -PheΨ[COcpCO]Phe-
surrogate 5 with a Phe-Phe dipeptide in which the backbone
is in an idealized â-strand and the two phenyl groups are fixed
in gauche(-) orientations.
relative to the backbone that it would if the correspond-
ing amino acid residue in 1 were locked in the gauche-
(-) conformation in which the ø1-angle is about -60°.
The geometric properties of replacements of type 2 may
be better illustrated by examining the superimposition
of the -PheΨ[COcpCO]Phe- replacement 5, in which
the N-terminal Phe is also anchored in the gauche(-)
conformation, with the corresponding Phe-Phe dipeptide
in which the backbone is in an idealized â-strand and
both phenyl groups are fixed in gauche(-) orientations
(Figure 1). The root-mean-square (rms) fit for this rigid
superimposition is approximately 0.35 Å. In a similar
fashion, the spatial orientation of the R2 group in the
epimeric replacement 3 mimics that found when the side
chain of the C-terminal residue in 1 is constrained in
the gauche(+) conformation wherein the ø1-angle is
about +60°.
Although some hydrogen-bonding capability of an
amide carbonyl group is maintained by the keto func-
tions in the rigid isosteres 2 and 3, deletion of a
backbone N-H in these surrogates eliminates one
possible intermolecular hydrogen bond to a receptor or
enzyme. However, the actual importance of losing this
hydrogen bond is difficult to assess a priori since other
peptide replacements that lack hydrogen-bonding ability
(e.g., double bonds) are known to be efficacious in certain
cases. The entropic advantage that arises from restrict-
ing the rotors (0.8-1.2 kcal/rotor) associated with the
φ- and the ø1-angles in the constrained pseudopeptide
could approximately compensate the energetic cost of
losing one intermolecular hydrogen bond, (0.5-1.8 kcal/
H-bond) provided the receptor bound conformation is
well-matched.11,12 It must also be recognized that
consequent to the mutation of the dipeptide 1 into 2 or
3 amide resonance and the associated hindered rotation
about the amide -C(dO)N bond is lost, so there is not
a net restriction of two full rotors.
Most of the conformationally restricted replacements
of peptide secondary structure reported to date imitate
turns, or helices. Inasmuch as peptide substrates
generally appear to bind to proteases in extended
conformations, the design of structural elements mim-
icking â-strand arrays is an important, albeit more
difficult, challenge. Recently there have been several
reports of peptide mimics that enforce extended struc-
ture although they make no provision for orienting the
side chains.6 The importance of controlling the side
chain orientations of pseudopeptides has been recog-
nized because these appendages provide crucial sites for
recognition, specificity, binding, and consequent trans-
duction.7 Consequently, we initiated a research pro-
gram to design and develop dipeptide replacements that
would enforce an extended (â-strand) conformation on
the backbone of oligopeptides while projecting the amino
acid side chains in specific orientations.
Design . On the basis of a series of molecular model-
ing studies, we postulated that the 1,2,3-trisubstituted
cyclopropanes 2 and 3, which shall be generally denoted
as -XaaΨ[COcpCO]Yaa-, would constitute novel rigid,
isosteric replacements of an extended conformation of
the dipeptide moiety 1 (Scheme 1). Operationally, 2 and
3, which differ from the more common 1-aminocyclo-
propanecarboxylic acids,8 are derived from 1 by replac-
ing the amide nitrogen with a carbon and forming a
single bond between this atom and C(â) on the amino
acid side chain.9 The peptide backbone in both 2 and 3
is locally rigidified in a â-strand conformation by locking
the φ angle at about -120°. Another factor that
contributes to the overall rigidity induced by this
surrogate is the preferred conformation, which is a
consequence of steric and electronic effects, about the
carbon-carbon bond between the carbonyl group and
the cyclopropane.10
Preliminary modeling studies suggested that trisub-
stituted cyclopropanes related to 2 and 3 would serve
as rigid mimics of localized â-strand structure. To
evaluate this hypothesis we first incorporated N-termi-
nal truncated analogues of 2 and 3 at the P3 position
of conformationally restricted analogues of the aspartic
proteinase renin.13 We discovered that 6, which bears
an N-terminal amide group rather than the keto func-
tion of the parent 2, was a subnanomolar inhibitor of
renin.13a Since 6 was approximately equipotent with
The portion of the amino acid side chain designated
as R2 in the cyclopropane-derived dipeptide surrogate
2 occupies approximately the same region of space

Cyclopropane-Derived Peptidomimetics
J ournal of Medicinal Chemistry, 1998, Vol. 41, No. 10 1583
the corresponding flexible analogue 7, it seems probable
that the two inhibitors bind to the active site of renin
in closely similar conformations and that the cyclopro-
pane ring in 6 orients the aromatic side chain at the P3
subsite in the biologically active conformation of 7. We
then began to exploit cyclopropane-containing inhibitors
as topographical probes of the bound conformation of
collagenase inhibitors and other biologically active
pseudopeptides.14 Others have recently incorporated
replacements related to 2 and 3 in non-peptide fibrino-
gen receptor antagonists.15
NMR techniques and molecular dynamics (MD) calcula-
tions, with the biologically active conformation of 10,
which was determined by X-ray analysis of the complex
of 10 bound to the active site of HIV-1 protease.
While these early investigations began to establish
the efficacy of trisubstituted cyclopropanes as novel
peptide mimics, a number of important questions re-
main. For example, can more than one cyclopropane
replacement be incorporated into a flexible enzyme
inhibitor to further rigidify the structure and stabilize
the biologically active conformation? Can the cyclopro-
pane subunit(s) be positioned closer to the site of the
“scissile amide bond” (i.e., the P1-P1′ juncture) without
loss of potency? To address these critical issues, we
queried whether cyclopropane-derived peptide isosteres
might be introduced into conformationally constrained
inhibitors of HIV-1 protease, an aspartic proteinase that
is responsible for the release of protease, reverse tran-
scriptase, and integrase from the gag-pol fusion protein
of the HIV-1 virus.16,17 Numerous X-ray structures of
HIV-1 protease complexed with inhibitors reveal that
the inhibitors typically bind to the enzyme active site
clefts in an extended â-strand conformation in which
the three-dimensional structure of the bound inhibitors
is very similar for the P2-P2′ segment of the backbone.
It thus occurred to us that a biological and structural
study of the pseudopeptides 9-13, which are derivatives
of the known inhibitors 14 (A-75925, IC50 ) 0.22 nM)16d
and 15 (A-76889, Ki ) 0.11 nM),17d would elicit some
additional insights regarding the scope and limitations
of -XaaΨ[COcpCO]Yaa- replacements. Each of the
pseudopeptides 9-13 contain two N-terminal truncated
-XaaΨ[COcpCO]Yaa- replacements, and these re-
placements are incorporated one residue closer (i.e., the
P2 and P2′ subsites) to the scissile bond than in the
renin inhibitor 6 (i.e., the P3 subsite). We now report
the syntheses and biological evaluation of the pseudopep-
tides 9-13 together with a comparison of the solution
structures of 9 and 10, which were determined using
Resu lts
Syn th esis of In h ibitor s. We envisioned that 9-12
could be prepared by coupling the enantiomerically pure
carboxylic acids 20a -d with the known diamino diol
22 (Scheme 2).18 On the basis of prior experience, the
simplest approach to these acids would involve applica-
tion of our general procedures for asymmetric intramo-
lecular cyclopropanations.19 In the event, the allylic
diazoacetates 16a -d , which were prepared from the
corresponding alcohols,20 were heated in the presence
of the chiral rhodium catalyst Rh2[5(S)-MEPY]4 to give
the cyclopropyl lactones 17a -d in 84-92% yield and
87-98% enantioselectivity. Subsequent opening of the
lactone rings of 17a -d with N-benzyl-N-(p-methox-
yphenylmethyl)amine (MPMNHBn) was accomplished
in 73-79% yield using the Weinreb protocol to give the
amides 18a -d , respectively.21 Oxidation of the primary
alcohol function in 18a -d with pyridinium chlorochro-
mate (PCC) gave the corresponding aldehydes (77-
93%), which were epimerized to deliver 19a -d . J ones
oxidation of 19a -d followed by removal of the methox-
yphenylmethyl (MPM) protecting group with trifluoro-
acetic acid (TFA) furnished the requisite carboxylic acids
20a -d in 60-69% yield. Reaction of 20a -d with the
diamino diol 22 under standard peptide coupling condi-
tions using N-ethyl-N-(dimethylaminopropyl)carbodi-

1584 J ournal of Medicinal Chemistry, 1998, Vol. 41, No. 10
Martin et al.
Sch em e 2
Sch em e 3
alcohol 26 in 74% overall yield. Oxidation of the
primary alcohol with the Dess-Martin periodinane
followed by epimerization of the resulting aldehyde
afforded 27 in 78% overall yield. Oxidation of 27 with
J ones reagent gave the acid 28 (55%), which was then
coupled to the diamino diol 21 as previously described
to furnish 14.
Biologica l Activity. Compounds 9-12, which are
analogues of the known inhibitor 14, were each found
to be fast-binding, competitive inhibitors of wild-type
HIV-1 protease; there was no evidence for slow binding.
The potencies of these pseudopeptides were virtually
identical with the respective Kis being 0.31-0.35 nM
for 9, 0.16-0.21 nM for 10, 0.47 nM for 11, and 0.17
nM for 12. On the other hand, the inhibitor 13, which
is an analogue of 15, was a significantly weaker inhibi-
tor having a Ki of 80 nM.
NMR Str u ctu r a l Stu d ies. Because of the sym-
metrical nature of the inhibitors 9 and 10, only a single
set of signals corresponding to one molecular half is
1
imide hydrochloride (EDC) and 1-hydroxybenzotriazole
(HOBT) afforded the pseudopeptides 9-12 in 77-98%
yield.
observed in the H NMR spectrum; the notation used
to refer to the different signals is shown in Figure 2.
The diastereotopic methylene Hâ protons of the Phe
residues at P1/P1′ were assigned using vicinal homo-
nuclear and heteronuclear coupling constants. The
Although the epimerization of the aldehydes obtained
by oxidation of 18a ,b,d proceeded to completion, only
partial epimerization of 21 occurred, and an equilibrium
mixture (40:60) of 21 and the desired aldehyde 19c was
obtained.22 To address this inefficiency an alternate
route to 20c was developed (Scheme 3). In the event,
the allylic diazo ester 16b was cyclized in 78% yield and
92% ee upon heating in the presence of the enantiomeric
dirhodium catalyst Rh2[5(R)-MEPY]4 to give 23. Open-
ing of the lactone ring with N-benzyl-N-(p-methoxyphe-
nylmethyl)amine under Weinreb conditions gave 24,
which was subjected to the standard sequence of oxida-
tion, epimerization, and deprotection to give 20c in 60%
overall yield from 23.
1
detailed H and 13C NMR assignments are listed in the
Supporting Information.
Solu tion Str u ctu r e of 9. No positive cross peaks
were detected in the ROESY spectrum of 9, thereby
indicating that it adopted a single major conformation
in solution on the NMR shift time scale and that any
other conformers were present at concentrations below
the limits of detection. A total of 23 distance restraints
were obtained from a NOESY spectrum with an S/N of
about 400 for geminal peaks, so it was possible to
observe H-H distances up to approximately 480 pm.
Furthermore, a restraint for the Phe HR-Phe HR′
separation was obtained from the 13C satellites of the
Phe HR diagonal peak in a ω1-13C filtered ROESY
spectrum. These two R protons are separated by five
bonds, and the chemical shift of Phe HR (δ, 4.60 ppm)
was very close to the applied spinlock frequency of 4.83
ppm, so any TOCSY and NOESY artifact contributions
in this restraint are negligible.23 Because the number
of long-range NOEs is small, a reliable conformational
analysis is difficult, and special emphasis was therefore
placed upon determining a large set of homo- and
The N-methyl-2-pyridyl analogue 13 was prepared by
a sequence of reactions analogous to those presented in
Scheme 4. Thus, amidation of the lactone 17a with
2-aminomethyl pyridine in the presence of a catalytic
amount of NaCN in methanol gave 25 in 98% yield.
Protection of the primary alcohol as the tert-butyldim-
ethylsilyl (TBDMS) ether, followed by methylation of
the amide nitrogen using NaH and MeI and removal of
the hydroxyl protecting group by the action of tetra-n-
butylammonium fluoride (TBAF) afforded the N-methyl

Cyclopropane-Derived Peptidomimetics
J ournal of Medicinal Chemistry, 1998, Vol. 41, No. 10 1585
F igu r e 2. Constitution and conformational labeling in 9 and
10.
heteronuclear J coupling constants.24 A total of 18
homo- and heteronuclear vicinal J values were mea-
sured, and the upper limits for two further vicinal J
values were established.
The orientation of the symmetry center between the
1,2-diol array was determined by the HF,HF′ coupling
constant of 10.0 Hz, which was obtained from 13C
satellites via a nondecoupled HMQC spectrum (see the
Experimental Section); the magnitude of this coupling
constant suggests an antiperiplanar arrangement of
these protons. Furthermore, the distance between the
protons Phe HR-Phe HR′ is 2.5 Å, thus indicating that
the dihedral angle between the two central CR carbons
is about +60°. Taken together, these observations
reliably define the conformation about the carbon-
carbon bond of the diol core.
F igu r e 3. Stereoplot of the structure of 9 obtained by
averaging the last 50 ps rMD and subsequent minimization.
The terminal Bzl groups do not display a defined orientation
in the MD run and were orientated to the two main conformers
indicated by J coupling data before minimization.
procedure converged for 9, and the conformations of the
12 low-energy structures were very similar having a
root-mean-square deviation (rmsd) of <0.6 Å. The
lowest energy structure resulting from simulated an-
nealing of 9 was then taken as the starting structure
for subsequent molecular dynamic (MD) simulations in
DMSO. After 100 ps MD with NOE restraints ((10%,
standard pseudoatom correction for upper bounds),29 the
restraints were removed during the following 100 ps MD
calculations to determine the stability of the resulting
structure. The structure did not change significantly
during free MD, suggesting that the resulting rMD
structure is energetically favorable. No explicit sym-
metry forcing was performed during the MD calcula-
tions, so the structure displays only approximate C2
symmetry on this time scale. Figure 3 shows the
extended solution structure of 9 that was obtained by
All calculated distance limits were then incorporated
as restraints for simulated annealing calculations. The
coupling constants for which parameters for the Karplus
curves exist were used directly as restraints.25-27
A
total of 21 experimental NOESY distances were set up
as 42 symmetry ambiguous restraints of the Nilges
type.28 Because the J coupling data rendered two
internal NOESY restraints within the Phe residues
redundant, these NOESY restraints were not utilized.
To maintain the conformation at the symmetry center,
the central bond was locked so that the two protons were
fixed in an antiperiplanar orientation with the two
hydroxyl groups and the two CR carbons, respectively,
being gauche to one another. The simulated annealing
Sch em e 4

1586 J ournal of Medicinal Chemistry, 1998, Vol. 41, No. 10
Martin et al.
Ta ble 1. Interproton Distances in 9 Determined by ISPA Approximation from a NOESY Spectrum Compared to Distances from the
MD Structure (Average during 50 ps rMD)a
b
H1
H2
reff,NOESY (pm)
reff,MD (pm)
rintra,MD(H1,H2)c (pm)
rinter,MD(H1,H2′)c (pm)
Bzl HN
Bzl H1/Bzl H2
Cp H1
265
229
387
406
341
374
439
332
224
419
458
417
350
349
275
259
326
303
303
270
244
236d
307
257
221
239
422
412
266
317
443
264
242
444
428
499
635
598
298
287
379
310
300
236
244
221
239
422
412
266
317
452
264
242
462
598
523
677
757
302
288
388
382
303
238
258
-
>1000
>1000
>1000
>1000
>1000
905
Bzl HN
Bzl HN
Cp H2
Cp H1
Bzl H1/Bzl H2
Cp Mepro-S
Phe HN
Phe HR
Cp Mepro-R
Phe HN
Phe HR
HF
Cp H1
Cp H1
Cp H1
635
Cp H2
>1000
Cp H2
802
Cp H2
573
438
629
769
626
455
599
529
328
482
402
303
219
461
477
Cp H2
Cp Mepro-S
Cp Mepro-S
Cp Mepro-S
Phe HN
Phe HN
Phe HN
Phe HN
Phe HR
Phe HR
Phe HR
Phe HR
Phe Hâ,pro-R
Phe Hâ,pro-S
Phe HR
Phe Hâ,pro-S
HF
Phe HR
Phe Hâ,pro-R
Phe Hâ,pro-S
HF
Phe Hâ,pro-R
Phe Hâ,pro-S
HF
Phe HR′
HF
263
249
265
249
HF
a
For easy comparison with the NOE data, values in column 4 are the sum of intra and inter symmetric contributions in the MD
6
6
b
structure according to (1/reff
) ) (1/rintra)
+ (1/rinter).6 Multiple entries denote integration of an overlapped or degenerate peak; CH3
peak intensities were divided by 3 before transformation into distances. c In cases of degenerate or overlapped peaks the distances were
calculated to the corresponding carbon. From ω1-13C filtered ROESY.
d
Ta ble 2. Coupling Constants J in 9 in DMSO-d6 at 300 Ka
the NMR data. Major violations are recognized only for
distances between pseudoatoms. All backbone angles
were experimentally determined by at least one distance
and at least one three-bond coupling, J , whose value
was maximal based upon the Karplus curve; thus, these
dihedral angles are well-defined. The extended confor-
mation of the backbone is also in agreement with the
observed temperature gradients, which show that the
amide protons (Bzl HN: dδ/dT ) -5.2 ppb/K; Phe HN:
dδ/dT ) -6.0 ppb/K) are exposed to solvent.30
dihedralMD
(deg)
coupling
J exp (Hz)
J MD (Hz)g
3
Bzl 3J (HN, H1) +
3J (HN, H2)
12.0 ( 0.4b
-
-
Cp
3J (Bzl N,H1)
3J (CO1,H2)
3J (CO2,H1)
3J (H1,H2)
0.0 ( 0.4d J (HR,15N(i+1)) ) 0.2 54
4.2 ( 1.0c
4.3 ( 1.0c
5.5 ( 0.4b
5.0 ( 0.4d
5.2 ( 0.4d
-
-
-
-
-
9
13
148
-6
-9
3J (H1,CMe,pro-S
)
)
3J (H2,CMe,pro-R
3J (Phe N,H2)
0.2 ( 0.4d J (HR,15N(i+1)) ) 0.2 -53
The large sums of the Bzl HN Bzl H1 and Bzl HN Bzl
H2 couplings in the NMR data suggest that the terminal
Bzl groups at P3 and P3′ are directed predominantly in
the two orientations shown in Figure 3. However, the
NMR data also clearly show that these Bzl groups are
freely rotating because the Bzl H1 and Bzl H2 protons
do not show the chemical shift dispersion that would
be observed if they resided in a single preferred orienta-
tion relative to the preceding Cp CO1. In contrast, the
experimental values of 11 and 3 Hz that were deter-
Phe 3J (HN,HR)
3J (HN,Câ)
9.5 ( 0.4b J (HN,HR) ) 9.9
Φ ) -125
0.8 ( 0.4d J (HN,Câ) ) 0.2
Φ ) -125
Φ ) -125
3J (HN,CF)
<0.8e
J (HN,CO) ) 0.8
3J (Cp CO2,HR)
3.8 ( 1.0c J (HR,CO(i-1)) ) 2.8 Φ ) -125
3J (HR,Hâ,pro-S
3J (HR,Hâ,pro-R
)
)
)
3.0 ( 0.4b J (HR,Hâ) ) 4.9
11.0 ( 0.4b J (HR,Hâ) ) 11.8
1.7 ( 1.5e J (Hâ,CO) ) 0.8
1.1 ( 1.5e J (Hâ,CO) ) 2.1
ø1 ) -167
ø1 ) -167
ø1 ) -167
ø1 ) -167
ø1 ) -167
68
3J (CF,Hâ,pro-S
3J (CF,Hâ,pro R
3J (HR,Ci)
)
3.3 ( 1.0c
<1.0b
-
3J (HR,HF)
J (HR,Hâ) ) 2.8
3J (HF,OH)
3J (HF,HF′)
5.6 ( 0.4b
-
-
-174
9.9 ( 0.4b J (HR,Hâ) ) 12.9
a
3
mined for the two J (Phe HR, Phe Hâ) couplings imply
The dihedral J values are compared to J values calculated
from the angles in the MD structure (average during 50 ps rMD)
using peptidic Karplus curves.54 b From absortive in phase signals.
c From peak fitting according to Keeler on HMBC and 1H-1D.50
that the side chains of Phe and Phe′ have a preferred
orientation. Quantitative analysis according to the
Pachler equations31 indicates that the rotamer having
ø1 ) -60° is populated to about 76%, which projects the
Phe side chain below the center of the cyclopropane unit.
The rotamer ø1 ) +60° is populated to about 21%.
Because of the relatively short simulation time, no
interconversion between these rotamers was observed
in the MD. The proximity of the phenyl rings of the
Phe and Phe′ residues at P1 and P1′ and the cyclopro-
pane units is also suggested by the spatial contacts
between the Cp H1 and H2 protons and the phenyl
protons that were observed in the ROESY spectrum.
d
From heteronuclear E.COSY. e From peak intensity simulation-
(see Supporting Information). f From antiphase signals according
to Kim and Prestegaard.49 Karplus curve parametrization is
g
indicated in parentheses.
averaging the last 50 ps of the rMD followed by
minimization; because the terminal Bzl groups did not
prefer a defined orientation in the MD run, they are
oriented in the two conformers indicated by the J
coupling data before minimization.
The structure depicted in Figure 3 is consistent with
all of the experimental data (Tables 1 and 2). For
example, the observed syn orientations of Phe HN to Cp
H1 and Cp H2 to Phe HN are consonant with the short
distances between these protons that were derived from
Solu tion Str u ctu r e of 10. The same protocol used
to determine the solution conformation of 9 was applied
to the pseudopeptide 10, but the analysis was compro-

Cyclopropane-Derived Peptidomimetics
J ournal of Medicinal Chemistry, 1998, Vol. 41, No. 10 1587
Ta ble 3. Interproton Distances in 10 as Determined by ISPA
Approximation from a ROESY Spectruma
reff,ROESY
(pm)
reff,NOESY
(9) (pm)
b
b
H1
H2
Bzl HN
Bzl H1/Bzl H2
Cp H1/Cp H2
Cp H3
246
216
514
511
420
311
216
404
402
265
Bzl HN
Bzl HN
Bzl HN
227c
bigd
big
Cp HMe
Cp H1/Cp H2
Cp H1/Cp H2
Cp H1/Cp H2
Cp H1/Cp H2
Cp H1/Cp H2
Bzl H1/Bzl H2
Cp HMe
406c
332c
222c
381c
458c
Phe HN
Phe HR
Phe HF
Cp H3
Bzl H1/Bzl H2
Phe HN
553
457
421
401
384
505
599
542
285
264
282
233
252
bigd
bigd
417d
350c,d
349d
big
Cp H3
Cp H3
Phe HR
Cp H3
Phe Hâ1/Phe Hâ2
HF
Cp H3
Cp HMe
Cp HMe
Cp HMe
Phe HN
Phe HN
Phe HN
Phe HR
Phe Hâ1/Phe Hâ2
Bzl H1/Bzl H2
Phe HN
big
big
HF
F igu r e 4. Binding site interactions made by compound 10
and HIV-1 protease observed in the X-ray structure. The
inhibitor makes asymmetric interactions with the active site
aspartates, Asp25 and 125, as well as with the S and S′ halves
of the binding site. Hydrogen bonds are shown as dashed lines
with distances in Å. Oxygen and nitrogen atoms are colored
red and blue, respectively. Carbon atoms are rust and gray
for inhibitor and enzyme, respectively.
Phe HR
275
252c
303
244
245c
Phe Hâ1/Phe Hâ2
Phe HF
Phe HF
Phe HF
a
For comparison experimental distances between corresponding
b
protons in compound 9 are listed in column 4. Multiple entries
denote integration of overlapped or degenerate peak; CH3 peak
intensities were divided by 3. c Overlapping in 10 was simulated
from corresponding nonoverlapped protons A and B in 9 according
Ta ble 5. Refinement Statistics for HIV-1 Protease Complexed
with Compound 10
6
6
d
to (1/reff
)
) (1/rPROTON
)
+ (1/rPROTON B).6
Distances from
A
resolution, Å
7.0-2.0
6154
49
nonequivalent protons are compared: CH3 versus H.
unique reflections (|F|/σ > 2.0)
completeness
Ta ble 4. Coupling Constants J in 10 in DMSO-d6 at 300 Ka
a
Rmerge
,
%
5.6
R factor, %
rms bonds, Å
rms angles, deg
no. of water molecules
18.5
0.013
3.1
coupling
J exp (Hz)
J exp (9) (Hz)
12.0 ( 0.4b
Bzl
Cp
3J (HN, H1) + 3J (HN, H2)
12.2 ( 0.4b
3J (CO1,H3)
3J (CO2,H3)
3J (H3,H1)
3J (H3,H2)
3.5 ( 1.0c
4.7 ( 1.0c
10.0 ( 0.4b
5.5 ( 0.4b
6.0 ( 0.4b
-
60
4.3 ( 1.0c
a
Rmerge ) ∑|I - I |/∑I where I is the observed intensity and
-
I
is the average intensity obtained from multiple observations.
5.5 ( 0.4b
Coordinates will be deposited with the Brookhaven Data Bank.
3J (H3,HMe
)
-
Phe
3J (HN,HR)
3J (HN,Câ)
3J (HN,CF)
9.7 ( 0.4b
9.5 ( 0.4b
0.8 ( 0.4
<0.8d
structure. The failure of these simulated annealing
calculations most likely arises from a lack of experi-
mental data, although failure to observe convergence
in such calculations may also indicate a flexible molecule
for which several conformations are energetically favor-
able. For example, the absence of distance restraints
for the Cp H1 and CpH2 protons result in an undefined
orientation of the neighboring atom Phe HN. Because
an unbiased starting structure was not obtained, no MD
simulations were performed. Nevertheless, the close
similarity of the NMR data that was obtained for 9 and
10 indicates that 10 adopts the extended conformation
as found for 9 at least part of the time.
Cr ysta llogr a p h ic An a lysis of a Com p lex of 10
w ith HIV-1 P r otea se. The 2.3 Å crystal structure of
HIV-1 protease complexed with compound 10 reveals
that the inhibitor binds in an extended conformation
similar to that found with other peptidomimetic inhibi-
tors of the protease17b and with the related C2 sym-
metry-based inhibitors upon which the design of 10 was
based16d (Figure 4, Table 5). The two active site aspartic
acids of HIV-1 protease are shared unequally by the two
OH groups of 10 in a manner characteristic of an
asymmetric mode of inhibitor binding in which the
molecular 2-fold axis of the enzyme passes close to one
of the two hydroxyl-bearing carbon atoms.16d One of the
hydroxyl groups of 10 is within hydrogen-bonding
<1.0d
<1.0d
3J (Cp CO2,HR)
3J (HR,Hâ1) + 3J (HR,Hâ2
3J (HR,Ci)
4.4 ( 1.0c
17.2 ( 0.4b
3.3 ( 1.0c
1.1 ( 0.4b
3.8 ( 1.0c
-
)
3.3 ( 1.0c
3J (HR,HF)
3J (HF,OH)
3J (HF,HF′)
<1.0b
5.9 ( 0.4b
5.6 ( 0.4b
9.9 ( 0.4b
9.8 ( 0.4b
a
In the last column, the corresponding experimental J values
in compound 9 are listed. b From absortive in-phase signals. c From
peak fitting according to Keeler on HMBC and 1H-1D.50 From
d
peak intensity simulation.
mised by two factors. First, there was virtually no
dispersion for the resonances associated with the cy-
clopropane protons H1 and H2 (1.95 and 1.93 ppm).
Second, the Phe Hâs shifts were nearly degenerate (2.68
and 2.64 ppm). A total of 23 ROESY distances were
identified that corresponded to H,H distances up to 480
pm, and 14 vicinal J couplings and 2 upper bounds for
vicinal J couplings were determined. Tables 3 and 4
list the experimentally determined values with a com-
parison to the corresponding data for 9. Examination
of these tables reveals the close similarity of the NMR
data of 10 and 9, except for the degeneracies noted above
in the spectrum of 10.
When a simulated annealing procedure was applied
to the data for 10, there was no convergence to a single

1588 J ournal of Medicinal Chemistry, 1998, Vol. 41, No. 10
Martin et al.
and X-ray crystal structures in the P2-P2′ portion of
the inhibitor. The HF- and DFT-optimized structures
for conf 1 are nearly identical with rmsd fit for all atoms
of 0.12 Å. The torsion angle value calculated for the
central dihedral angle CR-C-C-CR, Θ1, in conf 1 is
within 6° and 16° of the experimentally determined
values of 68.1° and 78.6° for the NMR and crystal
structures, respectively; these values are well within
experimental error. The calculated Θ1 angle for conf 1
is also nearly identical, 61.9-62.6° vs 62.0°, to that
found in the related inhibitor 15 with valine at P2 and
P2′.16d The other two calculated conformational minima,
conf 2 and conf 3, have Θ1 angles far outside the range
of observed experimental values.
The three optimized conformations also exhibit dif-
ferent hydrogen bonding patterns for the diol hydroxy
groups. The relative disposition of the two hydroxyls
is given by the torsion angle O1-C-C-O2, Θ2 in Table
6. In the lowest energy structure, conf 1, there is an
internal hydrogen between the two hydroxyl groups. In
conf 1, Θ2 is -52.6°, compared with -42.5° and -23.4°
for the NMR and crystal structures, respectively. In
conf 2, one hydroxy group hydrogen bonds to the
nitrogen of the closest amide, and the other hydroxy
group forms a hydrogen bond with the carbonyl oxygen
of the second amide. In conf 3, both hydroxy groups
interact with the nitrogen atoms of the nearest amide.
These calculations show that the gas phase energy
differences between the three structures are determined
mainly by differences in the energetics of the hydrogen-
bonding patterns. Inclusion of solvation effects reduces
these energy differences, but conf 1 is still predicted to
be the most stable structure.
F igu r e 5. Model of the P2-P2′ portion of compound 12 used
for conformational analysis by ab initio methods. Coordinates
of the NMR structure of 9 were used to create a starting model
which was optimized using DFT methods (see Methods). The
structure shown corresponds to the lowest energy conformation
(conf 1) listed in Table 6.
Ta ble 6. Relative Energies (kcal/mol) of Optimized
Conformations in the Gas Phase (∆Eg), Calculated Using HF
and DFT Methods, and in Water (∆Es)a
Θ1
Θ2
conformation ∆Eg-HF ∆Eg-DFT (HF/DFT)
(HF/DFT) ∆Es
conf 1
conf 2
conf 3
0.0
3.1
4.1
0.0
1.9
4.2
62.6/61.9 -52.6/-52.6 0.0
170.2/175.3 55.5/59.9
152.2/167.4 32.7/48.3
0.6
2.8
a
The values of torsional angles Θ1 (CR-C-C-CR) and Θ2
(O1-C-C-O2) are in degrees.
distance of three of the four aspartate oxygen atoms,
whereas that of the other hydroxyl can only hydrogen
bond to one aspartate oxygen (Figure 4). The P2 and
P2′ carbonyl oxygens make hydrogen bonds with the
flap water (Wat 301) that bridges the inhibitor with the
enzyme at Ile 50 and 150. However, the inhibitor-Wat
301 hydrogen bond distances are unusually long at 3.4
and 3.0 Å. Similarly the hydrogen bonds between the
NH atoms of the P3 and P3′ benzylamides and the
Gly48 and Gly148 backbone CO atoms of the enzyme
are rather long, being 3.5 and 2.9 Å, respectively.
Con for m a tion a l An a lysis of th e Cyclop r op a n e-
d iols. A computational model of the P2-P2′ region of
12 consisting of the 64 atoms (Figure 5) was built using
the conformation for the corresponding core region of 9
that was determined by NMR experiments (Figure 3).
Rotations of 30° about the central C-C diol bond were
performed, and the resulting structures were subjected
to complete optimization at an ab initio level of accuracy
using Hartree-Fock (HF) and density functional theory
(DFT) methods with the 6-31* basis set as implemented
in the Gaussian 94 program package.32-34 This analysis
revealed three different structures corresponding to
local energy minima in the subspace of torsion angle
Θ1 about the central carbon-carbon bond. The relative
energies of these conformations and their corresponding
torsion angles are shown in Table 6. While some of the
parameters listed in Table 6 are sensitive to correlation
effects, as shown by the differences between the HF and
DFT calculations, good qualitative agreement was ob-
tained using both ab initio methods. The lowest energy
conformation (conf 1) is close to the experimental NMR
Discu ssion
Numerous X-ray structures of HIV-1 protease com-
plexed with flexible, peptide-derived inhibitors have
shown that the inhibitors typically bind to the enzyme
active site cleft in an extended, â-strand conformation.17
On the basis of the working hypothesis that introducing
two cyclopropane subunits related to 2 and 3 at the P2
and P2′ subsites of the known HIV-1 protease inhibitors
14 and 15 would stabilize their biologically active
conformations, the pseudopeptides 9-13 were prepared.
The key step in the syntheses of 9-13 was an asym-
metric, intramolecular cyclopropanation that led to the
cyclopropanecarboxylic acids 20a -d and 28, which were
then coupled with the known diamino diol 22 to give
the targeted inhibitors 9-13. Each of the pseudopep-
tides 9-13 was found to be a potent, competitive
inhibitor of HIV-1 protease in a standard enzyme assay
with the Kis being determined as 0.31-0.35 nM for 9,
0.16-0.21 nM for 10, 0.47 nM for 11, 0.17 nM for 12,
and 80 nM for 13. Although the Kis for 9-12 compare
favorably with those observed for the related inhibitors
14 (IC50 ) 0.22 nM) and 15 (Ki ) 0.11 nM), 13 was a
significantly weaker inhibitor than anticipated.
The structures of the pseudopeptides 9 and 10 in
solution were studied to ascertain whether the presence
of two cyclopropanes in inhibitors related to 9-13
stabilized an extended conformation for the entire
ligand. Because of their highly flexible nature, it has
generally not been possible to determine the solution
structures of linear, peptidomimetic inhibitors for HIV-1

Cyclopropane-Derived Peptidomimetics
J ournal of Medicinal Chemistry, 1998, Vol. 41, No. 10 1589
protease. Indeed, prior to the present work, KNI-272,
which possesses a central norstatine moiety, was the
only such inhibitor whose tight binding was suggested
to be partly based on a favorable conformational pre-
organization of the backbone into the bioactive confor-
mation.35,36 It is therefore significant that the NMR and
associated molecular dynamics investigations conducted
with the linear pseudopeptide 9 in DMSO clearly
indicate that it adopts a well-defined, preferred confor-
mation in solution in which the backbone is extended
(Figure 3). Unfortunately, the lack of sufficient disper-
sion in the chemical shifts for several key protons in 10
reduced the number of NMR restraints that could be
used in simulated annealing calculations, and the
available data did not converge to a single structure that
could be used in further calculations. Nevertheless,
because the NMR parameters that were obtained for
10 correspond well to those measured for 9, there is good
reason to believe that 10 also prefers a similar extended
conformation.
F igu r e 6. Superposition of inhibitor conformations of com-
pounds 10 (rust-colored) and 15 (A-76889) (gray) complexed
with HIV-1 protease. Both compounds adopt very similar
conformations near the diol core, but show significant differ-
ences in P3-P2 and P2′-P3′. Superposition was performed
using only the CR coordinates for the enzymes from the two
complexes (rmsd ) 0.7 Å for 198 Ca atoms). Oxygen and
nitrogen atoms are colored red and blue, respectively.
minimizes the allylic strain along the extended peptide
backbone.41
The three-dimensional structure of 10 in its complex
with HIV-1 protease superimposes well onto the bound
conformation of the related inhibitor 15, particularly in
the segment defined by the P2-P2′ moieties (Figure 6).
The positions of hydrogen-bonding groups of both in-
hibitors are spatially conserved throughout their back-
bones. A comparison of the relative locations of the
cyclopropanyl and valinyl side chains of 10 and 15,
respectively, reveals that the methyl substituents of the
former nearly overlap with the Cγ methyl groups of
their valine counterparts in 15. Examination of the
structure of the complex of 10 with the protease also
suggests that both the monomethylated cyclopropane
in 11 and the dimethyl-substituted cyclopropane in 9
can fit nicely within the S2 and S2′ enzyme subsites
without any significant conformational changes, thereby
providing a structural basis for the close similarities in
the Kis for the series of inhibitors 9-12. These results
are consistent with the initial design principles of these
pseudopeptides and support the conclusion that the
restrictive cyclopropane moiety can be accommodated
within the context of a linear peptidomimetic inhibitor
to stabilize an extended conformation.
Despite the structural similarities of the inhibitors
10 and 15 when complexed to HIV-1 protease, there are
some notable differences in their specific interactions
with the enzyme. Because of the cyclopropane replace-
ments at P2 and P2′, compound 10 lacks the two NH
groups that are present in the valinyl residues of 15.
In the complex of 15 with HIV-1 protease, these NH
groups form short hydrogen bonds, 3.0 and 3.2 Å, with
the corresonding Gly48 and Gly148 backbone CO groups
of the enzyme. Interestingly, in the complex of 10 with
the protease, the energy lost by the lack of these specific
hydrogen bonds is compensated, at least in part, by the
formation of two different hydrogen bonds, 3.5 and 2.9
Å, between the NH atoms of the P3 and P3′ benzylamide
groups and the respective carbonyl groups of Gly48 and
Gly148 (Figure 4).
A number of factors appear to play roles in stabilizing
the extended structure of 9 in solution. Examination
of the conformation about the carbon-carbon bond at
the symmetry center of the 1,2-diol array reveals an
antiperiplanar arrangement of the protons HF and HF′,
which is often sterically favored in a tetrasubstituted
ethane fragment, and a consequent gauche orientation
of the two hydroxyl groups. The preferred gauche
orientation of two vicinal hydroxyl groups is known,37
and solvent seems to play a role in stabilizing this
conformation.38 Accordingly the measured temperature
gradient (dδ/dT ) -7.1 ppb/K) for the proton resonance
of the hydroxyl group indicates its exposure to solvent.
To gain further insights regarding the preferred
conformational properties about the central core of 9,
ab initio calculations were conducted. Preliminary
calculations for a simple model of a C2 symmetric diol
indicated the existence of a large number of intercon-
vertible conformational minima with low-energy barri-
ers.39 However, such calculations on the cyclopropane-
substituted diol segment found in 12 suggest that these
rigid backbone replacements at P2 and P2′ can lead to
further reduction in conformational flexibility about the
otherwise freely rotatable diol core. Thus, the experi-
mentally determined conformation about the central
carbon-carbon bond of 9 is the same as that predicted
by ab initio calculations and appears to be influenced
by the presence of the adjacent cyclopropane rings.
The two dihedral angles that are defined by rotations
about the CO1-Cp C1 and the Cp C2-CO2 bonds in 9
appear to be strongly constrained in an orientation in
which there is an approximate synperiplanar arrange-
ment between the Bzl HN and Cp H1 protons as well as
of the Cp H2 and Phe HN protons. This conformation is
presumably stabilized by the interaction of the π-orbital
of the carbonyl group with the carbon-carbon σ-bond
of the adjacent cyclopropane,40 an arrangement that is
found in the solid-state structures of several R,â-
cyclopropyl amides.10 The φ-angle of the Phe moiety,
which corresponds to rotation about the Phe NH-Phe
CR bond, is -125°. This conformation is commonly
observed in peptides because the syn relationship
between the amide carbonyl group and the R-proton
In the context of considering hydrogen-bonding in-
teractions between the different cyclopropane-derived
inhibitors and HIV-1 protease, it is significant that the
cyclopropane-derived pseudopeptide 13 lacks not only
the valinyl NH groups at the P2 and P2′ subsites of 15
but also the benzylamide NH groups at P3 and P3′ of

1590 J ournal of Medicinal Chemistry, 1998, Vol. 41, No. 10
Martin et al.
9-12. Thus, there are no hydrogen bond donors at the
P3/P2 and the P3′/P2′ subsites in 13 that can form
stabilizing hydrogen bonds with the Gly48 and Gly148
residues of the protease. A comparison of the binding
affinities of 13 and 15 suggests that there is an
approximately 3.9 kcal difference in their respective
binding energies, a value that corresponds approxi-
mately to the loss of two hydrogen bonds.11,12 However,
because the presence of N-methyl groups at the P3 and
P3′ subsites of 13 may also induce other conformational
changes arising from their possible interactions with the
proximal Cp H1 and Cp′ H1 protons, a precise analysis
of the basis for the observed difference in binding
affinities of 13 and 15 is problematic; additional struc-
tural information is required.
F igu r e 7. Comparison of crystal structure of compound 10
(rust-colored) complexed with HIV-1 protease with an aver-
aged, minimized structure of compound 9 (gray) in solution
as determined by NMR. In the solution structure of 9, the
benzyl groups at P3 and P3′ are disordered in the NMR
structure as indicated by the stippled dot surfaces. Coordinates
for six pairs of atoms in the diol core portions of the two
inhibitors were used for the superposition analysis (rmsd )
0.15 Å). Oxygen and nitrogen atoms are colored red and blue,
respectively.
Another significant contrast between the crystal
structures of the protease complexes of 10 and 15 is in
the conformations of the aromatic rings at the P3 and
P3′ subsites (Figure 6). In the complex with 15, the P3/
P3′ pyridine rings make electrostatically favorable
stacking interactions with the guanidinium groups of
Arg8 and Arg108, respectively.16d These P3/P3′ pyridine
rings in 15 also make hydrogen bonds via the ring
nitrogens with water molecules located within the S3
and S3′ pockets.42 Somewhat surprisingly, another
variance in these two complexes is found in the locations
of the two central hydroxy groups, which differ by about
1 Å, despite the identity of the P1-P1′ core regions of
10 and 15.
will require restriction of the torsional flexibility at the
P3 and P3′ subsites; however, the similar Ki values for
the two compounds suggest that the conformational
adjustments required for binding are probably compa-
rable.
Con clu sion
As part of a general program directed toward the de
novo design of molecules having defined topographical
and biological properties, the pseudopeptides 9-13 were
prepared as potential inhibitors of HIV-1 protease. The
design of these targets was based upon the hypothesis
that the presence of two cyclopropane rings, which are
N-terminal truncated analogues of the dipeptide mimics
2 and 3, at the P2 and P2′ subsites of 9-13 would
stabilize the biologically active conformations of the
related HIV-1 protease inhibitors 14 and 15. The
results presented herein provide compelling evidence
supporting this hypothesis. The similarity of Ki values
for the different cyclopropane analogues 9-12 suggests
that small substituents at the 2-position of the cyclo-
propane ring do not adversely affect the relative stability
of the bioactive conformation of the inhibitor. This
conclusion is supported both by the NMR data, which
are similar for 9 and 10, and by the model calculations
that were done using the unsubstituted analogue 12.
The magnitude of the reduced affinity of 13 was
unexpected and presumably results from a combination
of undetermined conformational factors at the P3/P3′
subsites coupled with a reduced hydrogen bonding
capability. The combination of NMR and X-ray struc-
tural work with the HIV-1 protease inhibitors 10 and
9, respectively, indicates that the preferred solution
conformation of these inhibitors corresponds closely to
their structure when bound to the protease active site.
These results further suggest that the cyclopropane-
derived peptidomimetics found in 9-12 stabilize and
enforce a three-dimensional structure at the P2/P1 and
P2′/P1′ positions that corresponds closely to the biologi-
cally active conformation at the same subsites of the
HIV-1 protease inhibitors 14 and 15. Despite the
similarities of the bound structures of 10 and 15 within
the protease active site, there are some significant
differences in the specific enzyme/inhibitor interactions
that result from the optimization of many interactions
The similarities and differences that are observed
between the complexes of 10 and 15 with HIV-1 pro-
tease underscore the variable nature of the interactions
between small molecule ligands and the active site
residues of the protease. The detailed binding mode for
a given inhibitor is a consequence of optimizing many
different interactions within the relatively large and
flexible binding site. Because 10 and 15 are virtually
equipotent inhibitors of HIV-1 protease, the overall
energetics of their binding to the enzyme are ap-
proximately equal, the changes in one specific interac-
tion being balanced by another factor.
The averaged and energy-minimized structure of 9
in DMSO as determined from rMD calculations using
NMR restraints is very similar to the three-dimensional
structure of the bound form of 10 as obtained from X-ray
analysis of its complex with HIV-1 protease (Figure 7).
This structural similarity suggests that there may little
entropy loss on binding of these and related ligands to
the protease. The structures of 9 and 10 are most
closely conserved in the conformations of their diol cores,
which superimpose to within 0.15 Å rmsd. The side
chain phenyl groups at P1 and P1′ in both the solution
and solid-state structures are oriented in the gauche-
(-) conformation with ø1 ) -60°. Although there is also
good similarity in the structures of the two inhibitors
in the P2 and P2′ regions containing the cyclopropane
replacements, there are significant differences in the
conformations of the P3 and P3′ subsites. In the
solution structure of compound 9, the P3/P3′ benzyl
groups are freely rotating, whereas in the bound struc-
ture of 10, the P3 and P3′ groups are oriented toward
the P2 and P2′ cyclopropyl groups. If the solution and
bound conformations of 9 and 10 are similar, then
binding of these pseudopeptides to the HIV-1 protease

Cyclopropane-Derived Peptidomimetics
J ournal of Medicinal Chemistry, 1998, Vol. 41, No. 10 1591
[1R-(1r,2r)]-(N-Ben zyl-N-p-m eth oxyben zyla m in ylca r -
bon yl)-2-for m yl-3,3-d im eth ylcyclop r op a n e-1-ca r boxa m -
id e. A solution of the alcohol 18a (166 mg, 0.47 mmol) and
pyridinium chlorochromate (PCC) (199 mg, 0.94 mmol) dis-
solved in dry CH2Cl2 (10 mL) was stirred for 12 h at room
temperature. The mixture was then diluted with Et2O (10 mL)
and filtered through a small plug of silica gel. The solvent
was concentrated under reduced pressure, and the crude
material was purified by flash chromatography eluting with
hexanes/EtOAc (1:1) to afford 148 mg (93%) of the aldehyde
as a clear oil: 1H NMR (rotamers) δ 9.70 (d, J ) 8.6 Hz, 1 H),
7.40-7.05 (comp, 7 H), 6.92 (dd, J ) 8.6 Hz, 1 H), 6.85 (d, J )
8.6 Hz, 1 H), 5.02 and 4.95 (rotameric d, J ) 14.4 Hz, 1 H),
4.63 and 4.60 (rotameric d, J ) 16.6 Hz, 1 H), 4.44 and 4.35
(rotameric d, J ) 14.4 Hz, 1 H), 4.20 and 4.15 (rotameric d, J
) 16.6 Hz, 1 H) 3.80 and 3.78 (rotameric s, 3 H), 2.32 and
2.27 (rotameric d, J ) 8.7 Hz, 1 H), 2.79 and 2.75 (rotameric
t, J ) 8.7 Hz, 1 H), 1.49 and 1.47 (rotameric s, 3 H), 1.18 and
1.11 (rotameric s, 3 H); 13C NMR (rotamers) δ 200.7, 169.6,
159.5, 159.4, 136.7, 136.5, 129.6, 129.0, 128.6, 128.2, 127.6,
126.3, 114.4, 114.0, 55.1, 55.0, 50.0, 49.0, 48.1, 48.0, 39.9, 39.5,
within a relatively large and flexible binding site, and
the results of such investigations will be reported in due
course. With this caveat in mind, substituted cyclopro-
panes might be used to probe structure-activity rela-
tionships of S2 and S2′ subsite preferences.
Exp er im en ta l Section
Unless otherwise noted, solvents and reagents were reagent
grade and used without purification. Tetrahydrofuran (THF)
was distilled from potassium/benzophenone ketyl under ni-
trogen, and dichloromethane (CH2Cl2) was distilled from
calcium hydride prior to use. Reactions involving air or
moisture sensitive reagents or intermediates were performed
under an inert atmosphere of argon in glassware that had been
oven or flame dried. Melting points are uncorrected. Infrared
(IR) spectra were recorded either neat on sodium chloride
plates or as solutions in CHCl3 as indicated and are reported
in wavenumbers (cm-1) referenced to the 1601.8 cm-1 absorp-
tion of a polystyrene film. 1H (300 MHz) and 13C (75.5 MHz)
NMR spectra were obtained as solutions in CDCl3 unless
otherwise indicated, and chemical shifts are reported in parts
per million (ppm, δ) downfield from internal standard Me4Si
(TMS). Coupling constants are reported in hertz (Hz). Spec-
tral splitting patterns are designated as s, singlet; br, broad;
d, doublet; t, triplet; q, quartet; m, multiplet; and comp,
complex multiplet. Flash chromatography was performed
using Merck silica gel 60 (230-400 mesh ASTM). Percent
yields are given for compounds that were g95% pure as judged
by NMR.
29.2, 28.6, 28.5, 15.6, 15.5; IR (CHCl3) ν 2955, 1701, 1647 cm-1
;
mass spectrum (CI) m/z 352.1922 (C22H25N1O3 + H requires
352.1912), 260, 121.
[1R-(1r,2â)-(N-Ben zyl-N-p -m et h oxyb en zyla m in ylca r -
bon yl)-2-for m yl-3,3-d im eth ylcyclop r op a n e-1-ca r boxa m -
id e (19a ). The aldehyde from the preceding experiment (100
mg, 0.30 mmol) was dissolved in degassed MeOH (10 mL)
containing K2CO3 (151 mg, 1.08 mmol), and the reaction was
stirred for 24 h at room temperature. Water (10 mL) was
added, and the mixture was extracted with CH2Cl2 (3 × 30
mL). The combined organic extracts were dried (MgSO4) and
concentrated under reduced pressure to afford 80 mg (80%) of
19a as a clear viscous oil; 1H NMR (rotamers) δ 9.70 and 9.69
(rotameric d, J ) 7.4 Hz, 1 H), 7.41-7.05 (comp, 7 H), 6.90 (d,
J ) 8.6 Hz, 1 H), 6.84 (d, J ) 8.6 Hz, 1 H), 4.85 and 4.83
(rotameric d, J ) 15.2 Hz, 1 H), 4.62 and 4.58 (rotameric d, J
) 15.2 Hz, 1 H), 4.46-4.28 (comp, 2 H), 3.81 and 3.79
(rotameric s, 3 H), 2.77-2.75 (m, 1 H), 2.61 and 2.55 (rotameric
d, J ) 5.0 Hz, 1 H), 1.26 and 1.23 (rotameric s, 3 H), 1.14 and
1.07 (rotameric s, 3 H); 13C NMR (rotamers) δ 199.4, 169.1,
158.8, 158.7, 137.8, 137.2, 129.6, 129.4, 129.0, 128.6, 128.2,
127.7, 127.5, 127.1, 126.4, 114.4, 113.9, 55.2, 55.1, 49.6, 49.5,
48.2, 48.1, 40.3, 35.3, 33.0, 21.0, 20.9, 20.0, 19.9; IR (CHCl3) ν
2970, 1710, 1646 cm-1; mass spectrum (CI) m/z 352.1922
(C22H25N1O3+H requires 352.1912), 260.
[1S -(1r,2r)]-(N -Be n zyl-N -p -m e t h oxyb e n zyl)-2-(h y-
d r oxym et h yl)-3,3-d im et h ylcyclop r op a n e-1-ca r b oxa m -
id e (18a ). A 2.0 M solution of Me3Al in hexanes (1.2 mL, 2.4
mmol) was slowly added to a suspension of N-benzyl-N-(4-
methoxybenzyl)amine (0.54 g, 2.4 mmol) in 1,2-dichloroethane
(DCE) (8 mL). The reaction mixture was stirred at room
temperature for 30 min, whereupon a solution of the cyclo-
propyl lactone 17a 19c (150 mg, 1.2 mmol) in DCE (0.5 mL) was
slowly added. The reaction was refluxed for 24 h, cooled to 0
°C, and carefully quenched with 1 N HCl (4 mL). The reaction
mixture was extracted with CH2Cl2 (3 × 10 mL), dried
(MgSO4), and concentrated under reduced pressure. The crude
product thus obtained was purified by flash chromatography
eluting with hexanes/EtOAc (1:1) to give 318 mg (75%) of 18a
as a colorless oil: 1H NMR (rotamers) δ 7.40-7.05 (comp, 7
H), 6.92 (d, J ) 8.6 Hz, 1 H), 6.84 (d, J ) 8.6 Hz, 1 H), 5.21
and 5.16 (rotameric d, J ) 14.6 Hz, 1 H), 4.75 and 4.70
(rotameric d, J ) 16.6 Hz, 1 H), 4.29 and 4.24 (rotameric d, J
) 16.6 Hz, 1 H), 4.16-3.87 (comp, 3 H), 3.83 and 3.80
(rotameric s, 3 H), 1.64 and 1.60 (rotameric d, J ) 8.5 Hz, 1
H), 1.41-1.31 (m, 1 H), 1.15 and 1.14 (rotameric s, 3 H), 1.12
and 1.11 (rotameric s, 3 H); 13C NMR (rotamers) δ 172.1, 159.0,
158.5, 133.5, 133.0, 129.5, 128.9, 128.5, 128.1, 127.8, 127.6,
127.3, 126.4, 114.3, 113.9, 59.4, 55.2, 55.1, 49.8, 49.4, 47.5, 47.1,
[1R-(1r,2r,3r)]-(N-Ben zyl-N-p-m eth oxyben zyla m in yl-
ca r bon yl)-2-for m yl-3-m eth ylcyclop r op a n e-1-ca r boxa m -
id e. Prepared from 18b as a clear glass in 88% yield using
the procedure described for the oxidation of 18a : 1H NMR
(rotamers) δ 9.85 (d, J ) 5.9 Hz, 1 H), 7.38-7.09 (comp, 7 H),
6.91 (d, J ) 8.5 Hz, 1 H), 6.85 (d, J ) 8.5 Hz, 1 H), 5.04 and
4.99 (rotameric d, J ) 14.7 Hz, 1 H), 4.72 and 4.68 (rotameric
d, J ) 16.7 Hz, 1 H), 4.43 and 3.38 (rotameric d, J ) 16.7 Hz,
1 H), 4.21-4.10 (m, 1 H), 3.82 and 3.80 (rotameric s, 3 H),
2.52 and 2.46 (rotameric t, J ) 8.9 Hz, 1 H), 1.92-1.77 (comp,
2 H), 1.51 and 1.48 (rotameric d, J ) 6.5 Hz, 3 H); 13C NMR
(rotamers) δ 200.4, 168.9, 159.3, 137.1, 136.2, 129.7, 129.1,
129.0, 128.6, 128.2, 127.8, 127.7, 127.6, 127.5, 126.3, 114.4,
114.0, 55.4, 55.3, 50.2, 50.0, 48.5, 48.1, 31.9, 31.1, 21.4, 21.3,
21.0, 14.2, 9.0; IR (CDCl3) ν 2934, 1696, 1636, 1512, 1248, 1204,
31.9, 30.4, 27.6, 23.7, 15.9; IR (CHCl3) ν 3457, 2948, 1628 cm-1
;
mass spectrum (CI) m/z 354.2060 (C22H29N1O3 + H requires
354.2069), 336.
[1R-(1r,2r,3r)]-(N-Ben zyl-N-p -m et h oxyb en zyl)-2-(h y-
d r oxym e t h yl)-3-m e t h ylcyclop r op a n e -1-ca r b oxa m id e
(18b). Prepared as a colorless oil in 75% yield from 17b after
purification flash chromatography eluting with hexanes/EtOAc
(1:1) according to procedure described for 18a : 1H NMR
(rotamers) δ 7.40-7.09 (comp, 7 H), 6.91 (d, J ) 8.7 Hz, 1 H),
6.83 (d, J ) 8.7 Hz, 1 H), 5.20 (d, J ) 14.4 Hz, 1 H), 4.99 and
4.83 (rotameric d, J ) 16.6 Hz, 1 H), 4.32 and 4.27 (rotameric
d, J ) 16.6 Hz, 1 H), 4.10-3.84 (comp, 3 H), 3.81 and 3.79
(rotameric s, 3 H), 1.87 and 1.82 (rotameric t, J ) 8.8 Hz, 1
H), 1.58-1.41 (comp, 2 H), 1.14 and 1.11 (rotameric d, J ) 7.0
Hz, 3 H); 13C NMR (rotamers) δ 171.8, 159.0, 158.9, 137.1,
136.2, 129.5, 129.2, 128.9, 128.5, 128.1, 127.8,127.6, 126.4,
114.4, 113.9, 58.7, 55.3, 55.2, 49.7, 49.4, 47.6, 47.2, 23.9, 22.6,
17.0, 16.9, 9.2, 9.1; IR (CDCl3) ν 3407, 2955, 1614, 1512, 1467,
1175, 1036 cm-1; mass spectrum (CI) m/z 338.1756 (C21H23
NO3 + H requires 338.1748), 324, 266, 230, 154, 137, 121.
-
[1R-(1r,2â,3r)]-(N-Ben zyl-N-p-m eth oxyben zyla m in yl-
ca r bon yl)-2-for m yl-3-m eth ylcyclop r op a n e-1-ca r boxa m -
id e (19b). Prepared from the preceding aldehyde as a clear
glass in 91% yield using the procedure described for the
preparation of 19a : 1H NMR (rotamers) δ 9.59 and 9.58
(rotameric d, J ) 8.2 Hz, 1 H), 7.39-7.08 (comp, 7 H), 6.90 (d,
J ) 8.5 Hz, 1 H), 6.84 (d, J ) 8.5 Hz, 1 H), 4.96 and 4.91
(rotameric d, J ) 14.7 Hz, 1 H), 4.74 and 4.69 (rotameric d, J
) 16.8 Hz, 1 H), 4.42 and 4.37 (rotameric d, J ) 16.3 Hz, 1
H), 4.25 and 4.19 (rotameric d, J ) 14.7 Hz, 1 H), 3.81 and
3.79 (rotameric s, 3 H), 2.75-7.71 (m, 1 H), 2.61 and 2.55
1248, 1032 cm-1; mass spectrum (CI) m/z 340.1913 (C21H25
NO3 + H requires 340.1922).
-

1592 J ournal of Medicinal Chemistry, 1998, Vol. 41, No. 10
Martin et al.
(rotameric dd, J ) 4.4, 9.6 Hz, 1 H), 1.89-1.81 (m, 1 H), 1.22
and 1.20 (rotameric d, J ) 6.3 Hz, 3 H); 13C NMR (rotamers)
δ 199.7, 167.9, 159.1, 137.1, 136.2, 129.6, 129.1, 128.9, 128.5,
128.1, 127.9, 127.8, 127.6, 127.4, 126.4, 114.2, 113.9, 55.3, 55.1,
49.6, 49.3, 48.3, 48.0, 35.7, 29.7, 24.8, 24.7, 11.7, 11.6; IR
(CDCl3) ν 2934, 1702, 1627, 1513, 1464, 1284, 1213, 1175, 1035
cm-1; mass spectrum m/z 338.1756 (C21H23NO3 + H requires
338.1752), 324, 233, 175.
5 H), 4.40 (d, J ) 15.0 Hz, 1 H), 4.34 (d, J ) 15.0 Hz, 1 H),
2.24 (dd, J ) 5.0, 9.6 Hz, 1 H), 2.01 (dd, J ) 5.0, 10.1 Hz, 1
H), 1.75-1.68 (m, 1 H), 1.20 (d, J ) 6.3 Hz, 3 H); 13C NMR
(CD3OD) δ 176.2, 170.9, 140.0, 129.5, 128.5, 128.2, 44.3, 30.7,
27.7, 24.2, 11.5; IR (Nujol) ν 3370, 2937, 2498, 1697, 1650,
1459, 1122, 977 cm-1; mass spectrum (CI) m/z 234.1130
(C13H15NO3 + H requires 234.1139), 151.
[1R,2S]-1-[4-(2′-Am in om eth ylpyr idin e)-car bon yl]-2-(h y-
d r oxym eth yl)-3,3-d im eth ylcyclop r op a n e (25). A solution
containing the lactone 17a (50 mg, 0.4 mmol), 2-aminomethyl
pyridine (0.25 mL, 2.4 mmol), catalytic NaCN (2 mg, 0.04
mmol), and MeOH (1 mL) was heated in a sealed flask at 70
°C for 72 h. The solvent was removed under reduced pressure,
and the crude orange-red oil was purified via flash chroma-
tography eluting with MeOH/CH2Cl2 (5:95) to give 92 mg (98%)
of 25 as an amber oil: 1H NMR δ 8.54 (d, J ) 4.7 Hz, 1 H),
7.74 (dt, J ) 1.7, 7.7 Hz, 1 H), 7.67 (dd, J ) 5.8, 16.3 Hz, 1 H),
7.50 (dd, J ) 4.9, 16.3 Hz), 4.67 (dd, J ) 5.8, 16.3 Hz, 1 H),
4.50 (dd, J ) 4.9, 16.3 Hz, 1 H), 4.12 (br s, 1 H), 3.98 (dd, J )
6.4, 12.1 Hz, 1 H), 3.82 (dd, J ) 9.8, 12.1 Hz, 1 H), 1.54 (d, J
) 8.7 Hz, 1 H), 1.39 (ddd, J ) 6.4, 8.7, 9.8 Hz, 1 H), 1.19 (s, 3
H), 1.18 (s, 3 H); 13C NMR δ 171.7, 157.3, 150.0, 136.9, 122.6,
59.4, 52.8, 36.1, 34.0, 32.1, 28.0, 23.5, 15.9; IR (CDCl3) ν 3440,
3054, 1648, 1512, 1422, 1265 cm-1; mass spectrum (CI) m/z
235.1447 (C13H18N2O2 + H requires 235.1438) 217, 203, 154,
135, 109.
[1R-(1r,2â)]-(N-Ben zyl-N-p-m eth oxyben zyl)-2-car boxyl-
3,3-d im eth ylcyclop r op a n e-1-ca r boxa m id e. A solution of
8 N J ones reagent (4 equiv) was added to a stirred solution of
19a (100 mg, 0.28 mmol) in acetone (2.8 mL) at 0 °C, and the
reaction was stirred for 3 h. Aqueous 1 N HCl (3 mL) was
added, and the mixture was extracted with CH2Cl2 (3 × 10
mL). The combined organic layers were dried (MgSO4) and
concentrated under reduced pressure, and the crude acid was
purified by flash chromatography eluting with hexanes/EtOAc
(1:1) containing 2% AcOH to afford 97 mg (93%) of the
carboxylic acid as a white solid: mp 165-167 °C; 1H NMR
(rotamers) δ 7.40-7.01 (comp, 7 H), 6.90 (d, J ) 8.6 Hz H),
6.81 (d, J ) 8.6 Hz, 1 H), 4.88 and 4.80 (rotameric d, J ) 16.3
Hz, 1 H), 4.62 and 4.55 (rotameric d, J ) 16.3 Hz, 1 H), 4.21-
4.15 (comp, 2 H), 3.80 and 3.78 (rotameric s, 3 H), 2.44 and
2.41 (rotameric d, J ) 6.0 Hz, 1 H), 2.38 and 2.32 (rotameric
d, J ) 6.0 Hz, 1 H), 1.15 (comp, 6 H); 13C NMR (rotamers) δ
176.5, 169.1, 159.2, 159.0, 137.2, 136.5, 129.6, 129.2, 128.9,
128.6, 128.5, 128.1, 127.9, 127.7, 129.4, 126.5, 114.4, 113.9,
55.3, 49.8, 49.5, 48.1, 48.0, 34.9, 31.6, 30.5, 20.8, 19.8, 19.7;
IR (CHCl3) ν 2957, 2622, 1731 cm-1; mass spectrum (CI) m/z
368.1858 (C22H25N1O4+H requires 368.1861), 349, 322, 136,
121.
[1R-(1r,2â)]-(N-Ben zyl)-2-ca r boxyl-3,3-d im eth ylcyclo-
p r op a n e-1-ca r boxa m id e (20a ). The (N-benzyl-N-p-meth-
oxybenzyl)-2-carboxyl-3,3-dimethylcyclopropane-1-carboxam-
ide from the preceding experiment (200 mg, 0.810 mmol) was
dissolved in trifluoroacetic acid (TFA) (10 mL), and the solution
was stirred for 24 h at room temperature. The reaction
mixture was concentrated under reduced pressure, and the
residue was dissolved in CH2Cl2 (50 mL). The organic layer
was washed with water (2 × 25 mL), dried (MgSO4), and
concentrated under reduced pressure to leave a yellow solid
that was purified by flash chromatography eluting with
hexanes/EtOAc (1:1) containing 2% AcOH to give 180 mg (90%)
of 20a as a white solid: mp 178-180 °C; 1H NMR δ 7.40-
7.29 (comp, 5 H), 6.40 (t, J ) 2.8 Hz, 1 H), 4.41 (d, J ) 5.7 Hz,
2 H), 2.32 (d, J ) 5.6 Hz, 1 H), 2.06 (d, J ) 5.6 Hz, 1 H), 1.32
(s, 3 H), 1.25 (s, 3 H); 13C NMR δ 176.0, 168.7, 138.0, 128.7,
127.5, 127.3, 126.7, 126.5, 43.8, 35.9, 32.1, 30.5, 20.5, 20.3; IR
(CHCl3) ν 2954, 2750, 1697, 1649 cm-1; mass spectrum (CI)
m/z 268.2528 (C14H17N1O3 + H requires 268.2532), 248(M+1),
230, 202, 184, 141, 91.
[1R-(1r,2â,3r)]-(N-Ben zyl-N-p-m eth oxyben zyla m in yl-
ca r bon yl)-2-ca r boxyl-3-m eth ylcyclop r op a n e-1-ca r boxa -
m id e. Prepared as a yellow oil in 94% yield from the aldehyde
19b using the procedure described for the oxidation of 19a :
1H NMR δ 10.00-9.40 (br s, 1 H), 7.39-7.08 (comp, 7 H), 6.90
(d, J ) 8.6 Hz, 1 H), 6.83 (d, J ) 8.6 Hz, 1 H), 4.93 and 4.90
(rotameric d, J ) 14.7 Hz, 1 H), 4.73 and 4.70 (rotameric d, J
) 16.4 Hz, 2 H), 4.42 and 4.36 (rotameric d, J ) 16.4 Hz, 1
H), 4.26 and 4.20 (rotameric d, J ) 14.7 Hz, 1 H), 3.80 and
3.79 (rotameric s, 3 H), 2.52 and 2.47 (rotameric dd, J ) 4.5,
9.7 Hz, 1 H), 2.40-2.36 (m, 1 H), 1.83-1.74 (m, 1 H), 1.19
and 1.17 (rotameric d, J ) 6.5 Hz, 3 H); 13C NMR (rotamers)
δ 178.2, 168.4, 159.1, 159.0, 137.1, 136.2, 129.6, 129.1, 128.9,
128.7, 128.2, 127.9, 127.7, 127.4, 126.6, 114.4, 114.0, 55.3, 55.2,
49.7, 49.4, 48.2, 47.9, 29.2, 26.7, 24.2, 24.1, 20.6, 11.7, 11.6;
IR (CDCl3) ν 3412, 3025, 1519, 1419, 1208 cm-1; mass
spectrum (CI) m/z 354.1705 (C21H23NO4 + H requires 354.1709),
262, 154, 136, 121, 107.
[1R,2S]-1-[4-(2′-Am in om eth ylpyr idin e)car bon yl]-2-(ter t-
bu tyldim eth ylsilyloxym eth yl)-3,3 dim eth ylcyclopr opan e.
A solution of the alcohol 25 (812 mg, 3.47 mmol) in CH2Cl2
(17 mL) was cooled to -20 °C, and TBDMS triflate (1.15 mL,
6.93 mmol) and 2,6-lutidene (0.92 mL, 5.21 mmol) were added
sequentially. The cooling bath was removed, and the reaction
was stirred for 1 h. Saturated NaHCO3 (10 mL) was added,
and the layers were separated. The aqueous phase was then
extracted with CH2Cl2 (3 × 10 mL), and the combined organics
were dried (MgSO4), concentrated, and purified via flash
chromatography, eluting with MeOH/CH2Cl2 (3:97) to give
1.083 g (90%) of the TBDMS protected alcohol as a tan oil:
1H NMR δ 8.52 (ddd, J ) 0.8, 1.7, 4.9 Hz, 1 H), 7.64 (dt, J )
1.7, 7.7 Hz, 1 H), 7.30-7.19 (comp, 2 H), 7.18-7.15 (m, 1 H),
4.54 (d, J ) 5.3 Hz, 1 H), 4.53 (d, J ) 5.3 Hz, 1 H), 4.10 (dd,
J ) 7.2, 11.3 Hz, 1 H), 3.88 (dd, J ) 7.7, 11.3 Hz, 1 H), 1.45
(d, J ) 8.9 Hz, 1 H), 1.27-1.22 (m, 1 H), 1.24 (s, 3 H), 1.17 (s,
3 H), 0.86 (s, 9 H), 0.05 (s, 3 H), 0.02 (s, 3 H); 13C NMR δ 170.5,
157.2, 148.9, 148.8, 136.6, 122.1, 121.8, 59.4, 44.6, 32.7, 31.2,
28.8, 25.9, 25.8, 22.9, 18.1, 14.9; IR (CDCl3) ν 3308, 2955, 2857,
1654, 1508, 1256, 1072, 836 cm-1; mass spectrum (CI) 349.2311
(C19H32N2O2Si + H requires 349.2305) 333, 291, 247, 231, 217.
[1R,2S]-1-[4-(2′-(N-Meth yl)a m in om eth ylp yr id in e)ca r -
bon yl]-2-(ter t-bu tyld im eth ylsilyloxym eth yl)-3,3-d im eth -
ylcyclop r op a n e. To a suspension of NaH (35% dispersion
in oil, 31 mg, 1.28 mmol) in THF (0.6 mL) at 0 °C was added
the preceding TBDMS protected alcohol (110 mg, 0.32 mmol)
in a small amount of THF (0.1 mL). This mixture was stirred
for 15 min, whereupon methyl iodide (0.08 mL, 1.28 mmol)
was slowly added and the heterogeneous mixture was stirred
for 2 h. Water (4 mL) and ether were then added, and the
layers were separated. The aqueous phase was extracted with
ether (3 × 2 mL), and the combined organics were dried
(MgSO4), concentrated, and purified via flash chromatography,
eluting with MeOH/CH2Cl2 (1:99) to give 96 mg (83%) of the
N-methyl TBDMS protected alcohol as an orange oil: 1H NMR
(rotamers) δ 8.59 & 8.52 (rotameric d, J ) 4.0 Hz, 1 H), 7.71
and 7.63 (rotameric dt, J ) 1.7, 7.7 Hz, 1 H), 7.30-7.14 (comp,
2 H), 4.77-4.60 (m, 2 H), 4.09-3.90 (m, 2H), 3.10 and 2.99
(rotameric s, 3 H), 1.58 and 1.51 (rotameric d, J ) 9.0 Hz, 1
H), 1.30-1.08 (comp, 7 H), 0.89 (s, 9 H), 0.05 (s, 6 H); 13C NMR
(rotamers) δ 171.6, 171.0, 157.7, 157.4, 149.7, 149.0, 136.7,
122.3, 122.0, 121.7, 120.1, 60.0, 59.8, 55.6, 52.7, 35.8, 34.2, 33.1,
32.5, 29.8, 28.9, 28.4, 26.4, 25.9, 25.4, 23.0, 22.3, 18.1, 15.1,
14.9; IR (CDCl3) ν 3424, 2955, 2857, 1636, 1472, 1436, 1403,
1257, 1075 cm-1; mass spectrum (CI) m/z 363.2467 (C20H34N2O2-
Si + H requires 363.2472) 347, 305, 231, 217, 149, 123.
[1R-(1r,2â,3r)]-(N-Ben zyl)-2-ca r b oxyl-3-m et h ylcyclo-
p r op a n e-1-ca r boxa m id e (20b). Prepared as a white solid
in 77% yield from the carboxylic acid in the preceding experi-
ment using the procedure described for the preparation of
1
20a : mp 150-152 °C; H NMR (CD3OD) δ 7.33-7.20 (comp,

Cyclopropane-Derived Peptidomimetics
J ournal of Medicinal Chemistry, 1998, Vol. 41, No. 10 1593
[1R,2S]-1-[4-(N-Meth yl)am in om eth ylpyr idin e)car bon yl]-
2-(h yd r oxym eth yl)-3,3-d im eth ylcyclop r op a n e (26). To a
solution of the preceding N-methyl TBDMS protected alcohol
(302 mg, 0.83 mmol) in THF (5 mL) at 0 °C was added TBAF
(1.5 mL, 1.50 mmol), and the mixture was stirred for 2.5 h at
0 °C. Saturated NaHCO3 (5 mL) was added, and the mixture
was extracted with Et2O (3 × 10 mL). The combined organic
layers were dried (MgSO4) and concentrated, and the residue
was purified by flash chromatography, eluting with MeOH/
CH2Cl2 (2:98) to give 209 mg (96%) of 26 as an amber oil: 1H
NMR (rotamers) δ 8.60 and 8.52 (rotameric d, J ) 4.8 Hz, 1H),
7.73 and 7.66 (rotameric dt, J ) 1.7, 7.7 Hz, 1 H), 7.28-7.17
(m, 2 H), 5.00 (rotamer, d, J ) 15.3 Hz, 0.5 H), 4.78 and 4.68
(rotameric d, J ) 17.0 Hz, 1 H), 4.50 (rotamer, d, J ) 15.3 Hz,
0.5 H), 3.94 (rotameric d, J ) 12.1 Hz, 1 H), 3.75 and 3.70
(rotameric dd, J ) 10.8, 12.1 Hz, 1 H), 3.15 and 3.01 (rotameric
s, 3 H), 1.61 and 1.60 (d, J ) 8.5 Hz, 1 H), 1.47-1.32 (m, 1 H),
1.24 and 1.15 (rotameric s, 3 H), 1.13 and 1.11 (rotameric s, 3
H); 13C NMR (rotamers) δ 172.0, 171.8, 157.1, 157.0, 149.9,
149.1, 137.0, 136.8, 122.6, 122.3, 122.0, 120.6, 59.4, 59.3, 55.5,
52.7, 36.0, 34.0, 31.9, 31.7, 30.9, 30.4, 27.9, 27.8, 23.5, 23.1,
16.0, 15.9; IR (CDCl3) ν 3406, 2950, 1622, 1476, 1435, 1405
cm-1; mass spectrum (CI) m/z 249.1603 (C14H20N2O2 + H
requires 249.1605) 231, 217, 123 (base).
with brine (1 mL) and extracted with EtOAc (3 × 2 mL). The
combined organics were dried (MgSO4) and concentrated, and
the residue was purified via flash chromatography, eluting
with MeOH/AcOH/CH2Cl2 (5:1:94) to give 2.8 mg (55%) of 28
as a colorless oil: 1H NMR (rotamers) (CD3OD) δ 8.55 and 8.47
(rotameric d, J ) 4.4 Hz, 1 H), 7.88-7.76 (m, 1 H), 7.36-7.28
(comp, 2 H), 4.85-4.70 (m, 2 H), 3.18 and 3.04 (rotameric s, 3
H), 2.43 and 2.32 (rotameric d, J ) 5.6 Hz, 1 H), 2.19 and 2.18
(rotameric d, J ) 5.6 Hz, 1 H), 1.16 and 1.11 and 1.10
(rotameric s, 6 H); 13C NMR (rotamers) (CD3OD) δ 174.2, 174.1,
171.9, 171.6, 158.2, 158.0, 150.6, 149.9, 146.0, 139.0, 138.9,
124.2, 123.9, 123.2, 122.6, 116.4, 56.0, 53.8, 36.5, 35.8, 35.3,
33.1, 32.9, 30.4, 30.0, 29.9, 21.4, 21.0, 20.7, 20.1; IR (CDCl3) ν
2930, 1703, 1640, 1406, 1260, 1112 cm-1; mass spectrum (CI)
m/z 263.1396 (C14H18N2O3 + H requires 263.1406) 245, 161,
123.
HIV-P R In h ibitor 9. N-Ethyl-N-(dimethylaminopropyl)-
carbodiimide hydrochloride (EDC) (25 mg, 0.13 mmol) was
added with stirring to a solution of the carboxylic acid 20a
(30 mg, 0.12 mmol), the diamino diol 22 (20 mg, 0.067 mmol),
and 1-hydroxybenzotriazole (HOBT) (36 mg, 0.27 mmol) in dry
DMF (1.7 mL) at -10 °C. The reaction was then stirred for
24 h at room temperature. A portion of EtOAc (2 mL) was
added, and the organic mixture was washed with brine (1 mL),
10% aqueous citric acid (1 mL), 10% aqueous NaHCO3 (1 mL),
and brine (1 mL). After drying (MgSO4), the solvents were
removed in vacuo to give 36 mg (77%) of 9 as a white solid:
[1R,2S]-1-[4-(2′-(N-Meth yl)a m in om eth ylp yr id in e)ca r -
bon yl]-3,3-d im eth ylcyclop r op a n e-2-ca r boxa ld eh yd e. To
suspension of Dess-Martin periodinane (157 mgs, 0.37 mmol)
in dry CH2Cl2 (2 mL) containing a catalytic amount of tert-
butyl alcohol (2 drops) was added the alcohol 26 (73 mg, 0.29
mmol) in dry CH2Cl2 (1 mL). The mixture was stirred for 2 h
whereupon saturated NaHCO3 (1 mL) and saturated K2S2O3
were added. The layers were separated, and the aqueous
phase was extracted with CH2Cl2 (3 × 2 mL). The combined
organics were dried (MgSO4) and concentrated in vacuo, and
the residue was purified via flash chromatography eluting with
MeOH/CH2Cl2 (1:19) to give 57 mg (80%) of the cis-aldehyde
as a pale yellow oil: 1H NMR δ 9.73 and 9.67 (rotameric d, J
) 6.4 Hz, 1 H), 8.61 and 8.53 (rotameric d, J ) 4.4 Hz, 1 H),
7.79-7.62 (m, 1 H), 7.29-7.18 (comp, 2 H), 4.87-4.28 (m, 2
H), 3.16 and 3.06 (rotameric s, 3 H), 2.32 (rotameric d, J )
8.7 Hz, 1 H), 1.84-1.72 (m, 1 H), 1.47-1.17 (m, 6 H); 13C NMR
(rotamers) δ 200.62, 200.5, 169.0, 168.3, 156.7, 156.1, 149.6,
148.8, 137.0, 136.7, 122.6, 122.2, 121.9, 120.5, 55.6, 52.8, 52.4,
39.6, 39.5, 39.4, 39.1, 36.1, 34.5, 28.5, 28.3, 27.2, 27.1, 15.8,
15.6, 15.1; IR (CDCl3) ν 2958, 1693, 1640, 1435 cm-1; mass
spectrum (CI) m/z 247.1447 (C14H19N2O2 + H requires
247.1446), 149, 123.
[1R,2R]-1-[4-(2′-(N-Meth yl)a m in om eth ylp yr id in e)ca r -
bon yl])-3,3-dim eth ylcyclopr opan e-2-car boxaldeh yde (27).
A solution of CsCO3 (586 mg, 1.8 mmol) in degassed MeOH (6
mL) and the cis-aldehyde from the previous experiment (111
mg, 0.45 mmol) was stirred overnight at room temperature.
Water (4 mL) and CH2Cl2 (10 mL) were added, and the layers
were separated. The aqueous phase was extracted with CH2-
Cl2 (3 × 5 mL), and the combined organics were dried (MgSO4)
and concentrated in vacuo. The crude oily product was
purified by filtration through a short plug of silica gel, eluting
with MeOH/CH2Cl2 (5:95) to afford 109 mg (98%) of 27 as an
oil: 1H NMR (rotamers) δ 9.78 and 9.68 (rotameric d, J ) 3.0
Hz, 1 H), 8.60 and 8.53 (rotameric d, J ) 4.5 Hz, 1 H), 7.73-
7.62 (m, 1 H), 7.29-7.16 (comp, 2 H), 4.87-4.63 (m, 2 H), 3.15
and 3.05 (rotameric s, 3 H), 2.71-2.68 (m, 1 H), 2.61 and 2.58
(rotameric d, J ) 5.4 Hz, 1 H), 1.33 and 1.22 (rotameric s, 3
H), 1.24 and 1.14 (rotameric s, 3 H); 13C NMR (rotamers) δ
200.3, 198.6, 198.2,168.7, 168.2, 157.0, 156.5, 149.7, 149.0,
136.7, 122.5, 122.2, 121.8, 120.3, 55.2, 53.0, 40.2, 39.8, 35.5,
35.0, 34.6, 32.9, 32.7, 21.0, 20.7, 19.9; IR (CDCl3) ν 2957, 1702,
1639, 1477, 1110 cm-1; mass spectrum (CI) m/z 247.1447
(C14H18N2O2 requires 247.1453), 229, 148, 123.
1
mp 264-267 °C; H NMR (DMSO-d6) δ 8.56 (t, J ) 6.4 Hz, 2
H), 7.81 (d, J ) 9.5 Hz, 2 H), 7.31-7.08 (comp, 20 H), 4.68 (s,
2 H), 4.60 (t, J ) 9.5 Hz, 2 H), 4.22 (d, J ) 6.4 Hz, 4 H), 3.17
(s, 2 H), 2.80 (dd, J ) 6.4, 9.5 Hz, 2 H), 2.09 (d, J ) 5.5 Hz, 2
H), 2.00 (d, J ) 5.5 Hz, 2 H), 1.15 (s, 6 H), 1.12 (s, 6 H); 13C
NMR (CD3OD) δ 172.5, 172.0, 140.2, 131.2, 131.0, 130.5, 130.2,
129.8, 129.4, 128.9, 128.3, 127.6, 74.3, 73.2, 52.5, 44.1, 39.2,
34.6, 34.2, 28.1, 21.2, 20.3; IR (CHCl3) ν 3050, 2952, 1685, 1678
cm-1; mass spectrum (FAB) m/z 759.4133 (C46H54N4O6 + H
requires 759.4121), 741.
HIV-P R In h ibitor 10. Isolated as a white solid in 91%
yield from 20b according to the procedure described for
1
preparing 9: mp 285-288 °C; H NMR (DMSO-d6) δ 8.57 (t,
J ) 6.1 Hz, 2 H), 7.78 (d, J ) 9.2 Hz, 2 H), 7.32-7.09 (comp,
20 H), 4.58 (s, 2 H), 4.34 (dd, J ) 9.2, 15.7 Hz, 2 H), 4.26 (d,
J ) 6.1 Hz, 4 H), 3.16 (s, 2 H), 2.71-2.61 (m, 4 H), 1.28 (dt, J
) 6.0, 9.3 Hz), 1.96-1.91 (comp, 4 H), 1.07 (d, J ) 6.0 Hz, 6
H); 13C NMR (CD3OD) δ 178.1, 173.2, 137.5, 136.2, 129.9,
128.4, 128.1, 127.9, 127.7, 127.5, 127.3, 126.9, 77.2, 73.2, 52.1,
43.2, 28.2, 26.3, 23.5, 12.6, 10.2; IR (CHCl3) ν 3050, 2929, 1729,
1602 cm-1; mass spectrum (CI) m/z 730.3818 (C44H50N4O6
H requires 731.3808), 613, 460, 307, 289, 248.
+
HIV-P R In h ibitor 13. To a solution of the acid 28 (26 mg,
0.099 mmol), HOBT (27 mg, 0.198 mmol), and diamino diol
22 (16.4 mg, 0.055 mmol) in DMF (1.4 mL) at -10 °C was
added EDC (21 mg, 0.11 mmol), and the solution was stirred
at room temperature for 24 h. EtOAc (2.8 mL) was added,
and the mixture was washed with saturated NaHCO3 (1 mL),
water (1 mL), and brine (1 mL). The organic layer was
separated, dried (Na2SO4), and concentrated in vacuo, and the
residue was purified via flash chromatography eluting with
MeOH/CHCl3 (10:90) to give 26 mg (60%) of product as a white
solid (mp 108-110 °C); 1H NMR (CD3OD) (rotamers) δ 8.57 &
4.47 (rotameric d, J ) 5.0 Hz, 2 H), 7.90-7.76 (m, 2 H), 7.37-
7.06 (comp, 14 H), 4.79-4.67 (m, 6 H), 3.12 and 3.00 (rotameric
s, 6 H), 2.90-2.70 (comp, 4 H), 2.30-2.25 (m, 2 H), 2.16-2.09
(m, 2 H), 1.71-0.83 (rotameric s, 12 H); 13C NMR (CD3OD) δ
173.0, 158.3, 150.6, 139.0, 138.9, 130.4, 129.2, 127.1, 123.9,
123.0, 122.7, 74.9, 53.5, 53.1, 39.8, 36.6, 35.2, 33.5, 29.2, 29.0,
21.3, 20.9, 20.6; IR (CHCl3) ν 3417, 2949, 1626, 1434, 1114
cm-1; mass spectrum (CI) m/z 789.4340 (C46H56N6O6 + H
requires 789.4328) 262, 245, 123, 120.
Kin etic Mea su r em en ts of 9-13 w ith HIV-1 P r otea se.
Recombinant HIV-1 protease was expressed and purified as
described.1 The Kis of 9-13 were determined using the
fluorogenic substrate Lys-Ala-Arg-Val-Tyr-Phe(NO2)-Glu-Ala-
Nle-NH2 using the excitation and emission maxima at 277 and
306 nm, respectively.2 Thus, the measured Kis were 0.31-
[1R,2R]-1-[4-(2′-(N-Meth yl)a m in om eth ylp yr id in e)ca r -
bon yl]-3,3-d im eth ylcyclop r op a n e-2-ca r boxylic a cid (28).
A solution of 27 (4.8 mg, 0.0195 mmol) in acetone (0.2 mL) at
0 °C was treated with 8 N J ones reagent (9 µL) and stirred
overnight. The resulting blue-green suspension was washed

1594 J ournal of Medicinal Chemistry, 1998, Vol. 41, No. 10
Martin et al.
0.35 nM for of 9, 0.16-0.21 nM for 10, 0.47 nM for 11, 0.17
nM for 12, and 80 nM for 13. Each of these pseudopeptides
was found to be a fast-binding, competitive inhibitor.
propane-1,2-dicarboxylic acid52 and from ab initio calculations
of substituted cyclopropanes;53 see Supporting Information for
the detailed parameter set.
The NOE-derived distances of the symmetric molecules 9
and 10 were included as symmetry-restraints using the
calculation strategy suggested by Nilges.28 In this approach
an “effective distance“ is defined, which takes into account that
every observed NOE cross peak in a symmetric dimer has
contributions from intra symmetry and inter symmetry cross
relaxation. The total NOE distance Rhij is described in eq 1.
In our case instead of a dimer a single symmetric small
molecule was analyzed. The observed signals carry contribu-
tions from distances within each half (intra) and between both
halves (inter) (Figure 2).
NMR Sp ectr oscop y. Acqu isition a n d An a lysis. All
spectra were recorded at 300 K on a BRUKER AMX500
spectrometer using a 18 mM natural abundance sample of 9
and a 30 mM natural abundance sample of 10 dissolved in
DMSO-d6, respectively. Temperature gradients of the 1H
chemical shifts were investigated by a series of 1D spectra
recorded between 300 and 320 K at a proton resonance
frequency of 250 MHz. Intraresidual proton and carbon
assignments were obtained from TOCSY and HMQC-COSY
spectra.45 Sequential assignment of the proton spin systems
was achieved by heteronuclear correlations (HMBC) supple-
mented by information from NOESY experiments. NOESY
and ROESY spectra with mixing times of 100 and 150 ms,
respectively, were recorded to generate quantitative restraints
for interproton distances. NOESY and ROESY peak integrals
were transformed into distance restraints using the isolated
two-spin approximation (ISPA).46 ROESY integrals were
transformed after correction for offset effects.47 To elucidate
distances between protons in opposite halfs of the symmetrical
molecule 9 Bzl-Cp-Phe to Phe′-Cp′-Bzl′, a ω1-13C filtered
ROESY spectrum with a mixing time of 150 ms was recorded.48
The intensitities in the ω1 13C satellites were summed,
corrected for the offset and transformed into distances using
the ISPA approximation.
intra-6
ij
inter-6 -1/6
Rhij ) {R
+ R
}
(1)
ij
Rhij: effective distance between atoms i and j
Rijintra: intra symmetry distance between atoms i and j
Rijinter: inter symmetry distance between atoms i and j
A harmonic flat bottomed potential is used for Enoe as given
in eq 2 and is included as additional potential term to the
DISCOVER 95.0 program.
Homonuclear J values were read from absorptive in-phase
signals in one-dimensional (1D) 1H spectra and two-dimen-
sional (2D) ROESY spectra using Gaussian apodization for
separation of the patterns. The dihedral proton coupling of
HF across the symmetry center to HF′ is a weak-regime
coupling, when one HF is attached to a 13C nucleus in the
natural abundance sample. To determine this dihedral cou-
pling and the other proton couplings at the 13C bound HF in 9
and 10, a 2D HMQC-COSY spectrum was recorded without
13C decoupling during the acquisition (8K real points). Using
the method of Kim and Prestegard,49 approximate J values
for the proton couplings were obtained from the proton
antiphase peaks observed in the row extracted at the carbon
chemical shift of CF. The proton chemical shift of the an-
tiphase peaks pointed to the coupling partner. Specifically,
the antiphase peak in the center of the HF 13C satellites yielded
an approximate value for the dihedral coupling HF-HF′. These
approximate J values were readily assigned to corresponding
correct J values determined via evaluation of the absorptive
inphase pattern at one HF 13C satellite at the carbon chemical
kij(rlower - Rhij)2
Rij < rlower
rlower e Rhij e rupper
rupper < Rhij
ENOE ) 0
kij(Rhij - rupper
(2)
2
)
Because of ambiguities of the NOE assignment, distance
geometry calculations for the generation of an unbiased
starting structure could not be performed. Therefore a three-
step simulated annealing protocol was applied. In the first
step the coordinates were randomized and then minimized
with nonbonded terms set to zero. The second step consisted
of molecular dynamics at 1000 K for 2 ps and cooling to 300 K
in seven steps with the nonbonded term scaled up to normal
values successively. In the third step, molecular dynamics
calculations were performed for additional 2 ps at 300 K with
the nonbonded term set to normal, and the resulting structure
was minimized. Chiral restraints, to maintain the correct
chirality, symmetry distance restraints as described above, and
restraints form coupling constants26,27 according to the Karplus
equations54 were included during the entire calculation. Since
these initial calculations are performed in vacuo, a dielectric
constant of 48.0 was applied. A total of 100 structures for each
compound 9 and 10 were generated. In the case of conver-
gence, a MD calculation in explicit solvent was performed for
further refinement.
The lowest energy structure resulting from simulated an-
nealing, was placed in a cubic box with 33 Å sides and soaked
with DMSO.55 A time step of 1 fs was maintained during all
MD simulations. The minimization was separated into two
steps for conjugate gradient energy minimization. In the first
step the solute was fixed, and in the second step all atoms
were allowed to move freely. The system was heated gradually
from 50 to 300 K in 2 ps steps, each by direct velocity scaling.
The system was equilibrated at 300 K for 10 ps with temper-
ature coupling (300 K), and then configurations were saved
every 100 fs for another 100 ps. During this simulation, the
shift of CF in
a 2D HMQC spectrum, recorded without
decoupling during the acquisition (8K real points).
1
Pairs of HMBC (mixing time 70 ms) and H 1D spectra were
analyzed for heteronuclear 3J (13C,1H) coupling constants em-
ploying the peak fitting procedure of Keeler.50 For some
signals, the signal-to-noise (S/N) of 3J -correlated peaks in the
HMBC was too weak for the Keeler fitting procedure. In these
cases, the value for the 3J coupling was obtained by numerical
simulation of the peak intensity in the HMBC. Input to this
numerical simulation were all (1H,1H) couplings of the proton
and one 2J (13C,1H) coupling of the proton along with the
2
integral of the J (13C,1H) HMBC peak. Details of the simula-
tion protocol are given in the Supporting Information. To
3
determine the small J (15N,1H) couplings of the cyclopropane
ring protons to the nitrogens in the adjoining amide plane in
1
9, a ω1-15N filtered H-1H ROESY24 with a mixing time of 300
ms was recorded. The measuring time was 4 days to ensure
a S/N of the important peaks better than 3. J values were
read from the resulting E.COSY patterns.
symmetric-distance restraints were applied with
a force
Str u ctu r e Ca lcu la tion s. The structure calculations for
9 and 10 were performed on Silicon Graphics Crimson R4000
and Indy R4600 computers. The MD calculations were carried
out with the program DISCOVER 95.0 (Biosym) using the
CVFF force field.51 For the parameters of the cyclopropane
ring, new atom types were defined, Cp and Hp. The geometric
parameters (angles, bond lengths, and torsions) were taken
from the X-ray structures of cis- and trans-3,3-dimethylcyclo-
constant of 1 kcal Å-2 mol-1 Additionally a harmonic 3J
.
coupling potential according to the Karplus equation was
included.26,27 The geometry of the symmetry center was
constrained adding the experimentally derived distance re-
straint between Phe HR and Phe HR′ and applying a torsion
force on the central dihedral angle HF-HF′ of 180° with a force
constant of 50 kcal/Å2mol to take account for the vicinal J
coupling constant.

Cyclopropane-Derived Peptidomimetics
J ournal of Medicinal Chemistry, 1998, Vol. 41, No. 10 1595
The averaged structure of the restraint MD of 9 was
minimized with steepest descent. The MD simulation was-
continued without restraints for another 100 ps, to examine
the stability of the structure defined by restrained MD.14 The
distance violations were calculated according a r-3 -1/3 averag-
ing. For the distance r an effective distance (formula 1) was
used to take account for symmetry effects.
of tradenames, commercial products, or organization
imply endorsement by the U. S. Government.
Su p p or tin g In for m a tion Ava ila ble: Experimental pro-
cedures and spectral data for compounds 11 and 12 and
intermediates in their synthesis, detailed NMR assignments
for 9 and 10, parameters for the cyclopropane used in the MD
calculations, and descriptions of techniques used to handle
ambiguous NOE data and to determine hetero coupling
constants from the HMBC spectrum (11 pages). Ordering
information is given on any current masthead page.
X-r a y Cr ysta llogr a p h y. Cr ysta lliza tion . Crystals of
HIV-1 protease/10 complex were grown by vapor diffusion in
hanging drops. HIV-1 protease, at 5 mg/mL in 50 mM sodium
acetate buffer, pH 5.6, containing 10 mM dithiothreitol, was
mixed with a stock solution of 10 in 25% DMSO to give a 1:5
molar ratio of enzyme:inhibitor at a final DMSO concentration
of 5%. The crystallization well buffer consisted of 75 mM
sodium citrate-150 mM sodium phosphate, pH 5.6-6.5, with
20-30% saturated ammonium sulfate. Drops were prepared
by mixing equal volumes of sample and well solution, with
subsequent addition of 10 µL of â-mercaptoethanol, 100 µL of
DMSO, and 40 µL of 2-propanol to the well before sealing.
Drops were sometimes seeded with microcrystals prior to
sealing. Small crystals were used to seed hanging drops using
a different well mix, 0.3 M ammonium phosphate, pH 6.2,
containing 20-30% saturated ammonim sulfate, with 2-5%
ethanol used as well additive prior to sealing. Crystals of
suitable size for diffraction data collection grew over a period
of several weeks at room temperature.
Refer en ces
(1) Notable cases are found in the structures of immunosupressants
and their complexes. Cyclosporin A. Solution and X-ray: (a)
Loosli, H. R.; Kessler, H.; Oschkinat, H.; Petcher, H. P.; Weber,
T. J .; Widmer A. 76. Peptide Conformations. The Conformation
of Cyclosporin A in the Crystal and in Solution. Helv. Chim. Acta
1985, 68, 682-704. (b) Kessler, H.; Ko¨ck, M.; Wein, T.; Gehrke,
M. 164. Reinvestigation of the Conformatin of Cyclosporin A in
Chloroform. Helv. Chim. Acta 1990, 73, 1818-1832. Complex:
(c) Weber, C.; Wider, G.; von Freyberg, B.; Traber, R.; Braun,
W.; Widmer, H.; Wu¨thrich, K. The NMR Structure of Cyclosporin
A Bound to Cyclophilin in Aqueous Solution. Biochemistry 1991,
30, 6563-6574. (d) Fesik, S. W.; Gampe, R. T.; Eaton, H. L.;
Gemmecker, G.; Olejniczak, E. T.; Neri, P.; Holzman, T. F.; Egan,
D. A.; Edalji, R.; Simmer, R.; Helfrich, R.; Hochloewski, J .;
J ackson, M. NMR Studies of [U-13C]Cyclosporin A Bound to
Cyclophilin: Bound Conformation and Portions of Cyclosporin
Involved in Binding. Biochemistry 1991, 30, 6574-6583. FK506.
Solution and X-ray: (e) Karuso, P.; Kessler, H.; Mierke, D. F.
Solution Structure of FK506 from Nuclear Magnetic Resonance
and Molecular Dynamics. J . Am. Chem. Soc. 1990, 112, 9434-
9436. (f) Siekerka, J . J .; Hung, S. H. Y.; Poe, M.; Lin, C. S.; Sigal,
N. H. A Cytosolic Binding Protein for the Immunosuppressant
FK506 has Peptidyl-Prolyl Isomerase Activity but is Distinct
from Cyclophilin. Nature 1989, 341, 755-757. Complex: (g) van
Duyne, G. D.; Standaert, R. F.; Karplus, P. A.; Schreiber, S. L.;
Clardy, J . Atomic Structure of FKBP-FK506, an Immunophilin
Immunosuppressant Complex. Science 1991, 252, 839-842.
(2) (a) Kessler, H. NMR-spektroskopishe Methoden Zur Konforma-
tionanayse von Peptiden. (Conformations and Biological Activity
of Cyclic Peptides.) Angew. Chem., Int. Ed. Engl. 1982, 21, 512-
523. (b) Hruby, V. J . Conformational Restrictions of Biologically
Active Peptides via Amino Acid Side Chain Groups. Life Sci.
1982, 31, 189-199. (c) Toniolo, C. Conformationally Restricted
Peptides Through Short-Range Cyclizations. Int. J . Pept. Protein
Res. 1990, 35, 287-300.
(3) (a) Spatola, A. F. Peptide Backbone Modifications: A Structure-
Activity Analysis of Peptides Containing Amide Bond Sur-
rogates. Conformational Constraints, and Related Backbone
Replacements. In Chemistry and Biochemistry of Amino Acids,
Peptides, and Proteins; Weinstein, B., Ed.; Marcel Dekker, New
York, 1983; Vol. 7, pp 267-357. (b) Hruby, B. J .; Al-Obeidi, F.;
Kazmierski, W., Emerging Approaches in the Molecular Design
of Receptor-Selective Peptide Ligands: Conformational, Topo-
graphical and Dynamic Considerations. Biochem. J . 1990, 268,
249-262. (c) Olson, G. L.; Bolin, D. R.; Bonner, M. P.; Bo¨s, M.;
Cook, C. M.; Fry, D. C.; Graves, B. J .; Hatada, M.; Hill, D. E.;
Kahn, M.; Madison, V. S.; Ruisiecki, V. K.; Sarabu, R.; Sepinwall,
J .; Vincent, G. P.; Voss, M. E. Concepts and Progress in the
Development of Peptides Mimics. J . Med. Chem. 1993, 36, 3039-
3049. (d) Giannis, A.; Kolter, T. Peptidomimetics for Receptor
Ligands. Discovery, Development, and Medical Perspectives.
Angew. Chem., Int. Ed. Eng. 1993, 32, 1244-1267. (e) Gante, J .
Peptidomimetics-Tailored Enzyme Inhibitors. Angew. Chem., Int.
Ed. Eng. 1994, 33, 1699-1720. (f) Wiley, R. A.; Rich, D. H.
Peptidomimetics Derived from Natural-Products. Med. Res. Rev.
1993, 13, 327-384. (g) Adang, A. E. P.; Hermkens, P. H. H.;
Linders, J . T. M.; Ottenheijm, H. C. J .; van Staveren, C. J . Case
Histories of Peptidomimetics. Progression from Peptides to
Drugs. Recl. Trav. Chim. Pays-Bas 1994, 113, 63-78.
Str u ctu r e Deter m in a tion a n d Refin em en t. Crystals
belong to the space group P61 with unit cell parameters a ) b
) 63.3 Å and c ) 83.5 Å. X-ray diffraction data were collected
from a single crystal using a MAR imaging plate area detector
system and an RU-200 rotating anode X-ray generator operat-
ing at 50 kV and 100 mA. Data were processed using the
program DENZO. The data are 49% complete to 2.0 Å and
70% complete to 2.5 Å with a merging R factor of 5.6% overall.
A total of 6154 unique reflections with |F|/σ > 2.0 in the
resolution range 7-2.0 Å were used in the structure determi-
nation and refinement. The crystal structure was solved by
molecular replacement using the protein coordinates of HIV-
PR complexed with PD9956057 and refined to a final R factor
of 18.5% using the program X-PLOR 3.1. The inhibitor was
initially built using QUANTA (Molecular Simulations, Inc.
Cambridge, MA) and fitted to the difference electron density
map calculated using the phases from the refined protein
structure. Refinement statistics are listed in Table 1.
Mod el Ca lcu la tion s for th e Cyclop r op a n e Diols. Ab
initio calculations were performed using Hartree-Fock (HF)
and density functional theory (DFT) methods with the 6-31G*
basis set as implemented in the Gaussian 94 program pack-
age.32 For all DFT calculations, hybrid B3LYP exchange
correlation potentials were included with the 6-31G* basis
set.33,34 Optimizations at the semiempirical level used AM1
and PM3 methods as implemented in the MNDO94 program
suite in UniChem 3.0. Solvation free energies were calculated
using a reaction field approach based on the boundary element
method (BEM).58 The effective atomic charges used for the
BEM calculations of solvation energies were obtained from
fitting the electrostatic potentials, calculated using DFT at the
B3LYP/6-31G* level, for each optimized structure in its gas-
phase geometry.
Ack n ow led gm en t. We are grateful for generous
financial support from the National Institutes of Health,
the Robert A. Welch Foundation, the Texas Advanced
Technology Program, Alexander von Humboldt Founda-
tion (S.F.M.), the Deutsche Forschungsgemeinschaft
and the Fonds der chemischen Industrie (H.K.), and the
National Cancer Institute (J .W.E.). We also thank Drs.
Dale J . Kempf and Norman E. Wideburg (Abbott
Laboratories) for preliminary biological testing of com-
pounds 9 and 10. The content of this publication does
not necessarily reflect the view or policies of the Depart-
ment of Health and Human Services, nor does mention
(4) (a) Ball, J . B.; Alewood, P. F. Conformational Constraints:
Nonpeptide â-Turn Mimics. J . Mol. Recognit. 1990, 3, 55-64.
(b) Kahn, M. Peptide Secondary Structure Mimetics: Recent
Advances and Future Challenges. SynLett. 1993, 821-826.
(5) Mueller, K.; Obrecht, D.; Knierzinger, A.; Stankovic, C.; Spiegler,
C.; Bannwarth, W.; Trzeciak, A; Englert, G.; Labhardt, A. M.;
Scho¨nholzer, P. Building Blocks for the Induction or Fixation of
Peptide Conformations. In Perspectives in Medicinal Chemistry;
Testa, B., Fuhrer, W., Kyburz, E., Giger, R., Eds.; VCH:
Weinheim, Germany, 1993; pp 513-532.
(6) For example, see: (a) Hagihara, M.; Anthony, N. J .; Stout, T.
J .; Clardy, J .; Schreiber, S. L. Vinylogous Polypeptides: An
Alternative Peptide Backbone. J . Am. Chem. Soc. 1992, 114,
6568-6570. (b) Smith, A. B., III; Keenan, T. P.; Holcomb, R. C.;

1596 J ournal of Medicinal Chemistry, 1998, Vol. 41, No. 10
Martin et al.
Sprengeler, P. A.; Guzman, M. C.; Wood, J . L.; Carroll, P. J .;
Hirschmann, R. Design, Synthesis, and Crystal Structure of a
Pyrrolinone-Based Peptidomimetic Possessing the Conformation
of a â-Strand: Potential Application to the Design of Novel
Inhibitors of Proteolytic Enzymes, J . Am. Chem. Soc. 1992, 114,
10672-10674. (c) Chen, L.; Trilles, R. V. Asymmetric Synthesis
of a Novel Phenylogous Amino Acid Mimicking an Extended
Dipeptide. Tetrahedron Lett. 1992, 36, 8715-8718. (d) Bou-
mendjel, A.; Roberts, J . C.; Hu, E.; Pallai, P. V.; Rebek, J ., J r.
Design and Asymmetric Synthesis of â-Strand Peptidomimetics.
J . Org. Chem. 1996, 61, 4434-4438. (e) J anetka, J . W.; Raman,
P.; Satyshur, K.; Flentke, G. R.; Rich, D. H. Novel Cyclic
Biphenyl Ether Peptide â-Strand Mimetics and HIV-Protease
Inhibitors. J . Am. Chem. Soc. 1997, 119, 441-442.
Miyano, M.; Feigen, L. P.; Adams, S. P. Design of Orally, Active,
Non-Peptide Fibrinogen Receptor Antagonists. An Evolutionary
Process from the RGD Sequence to Novel Anti-Platelet Aggrega-
tion Agents. Bioorg. Med. Chem. 1994, 2, 881-895.
(16) For a review HIV-1 protease inhibitors, see: (a) Huff, J . R. HIV
Protease: A Novel Chemotherapeutic Target for AIDS. J . Med.
Chem. 1991, 34, 2305-2314. See also: (b) Dreyer, G. B.;
Lambert, D. M.; Meek, T. D.; Carr, T. J .; Tomaszek, T. A., J r.;
Fernandez, A. V.; Bartus, H.; Cacciavillani, E.; Hassell, A. M.;
Minnich, M.; Petteway, S. R., J r.; Metcalf, B. W. Hydroxyeth-
ylene Isostere Inhibitors of Human Immunodeficiency Virus-1
Protease: Structure-Activity Analysis Using Enzyme Kinetics,
X-ray Crystallography, and Infected T-Cell Assays. Biochemistry
1992, 31, 6646-6659. (c) Thompson, W. J .; Fitzgerald, P. M. D.;
Holloway, M. K.; Emini, E. A.; Drake, P. L.; McKeever, B. M.;
Schleif, W. A.; Quintero, J . C.; Zugay, J . A.; Tucker, T. J .;
Schwering, J . E.; Homnick, C. F.; Nunberg, J .; Springer, J . P.;
Huff, J . R. Synthesis and Antiviral Activity of a Series of HIV-1
Protease Inhibitors with Functionality Tethered to the P1 or P1′
Phenyl Substituents: X-ray Crystal Structure Assisted Design.
J . Med. Chem. 1992, 35, 1685-1701. (d) Kempf, D. J .; Codacovi,
L.; Wang, X. C.; Kohlbrenner, W. E.; Wideburg, N. E.; Saldivar,
A.; Vasavanonda, S.; March, K. C.; Bryant, P.; Sham, H. L.;
Green, B. E.; Betebenner, D. A.; Erickson, J .; Norbeck, D. W.
Symmetry-Based Inhibitors of HIV Protease. Structure-activity
Studies of Acylated 2,4-Diamino-1,5-diphenyl-3-hydroxypentane
and 2,5-Diamino-1,6-diphenylheane-3,4-diol. J . Med. Chem.
1993, 36, 320-330. (e) Thaisrivongs, S.; Turner, S. R.; Strohbach,
J . W.; TenBrink, R. E.; Tarpley, W. G.; McQuade, T. J .;
Heinrikson, R. L.; Tomasselli, A. G.; Hui, J . O.; Howe, W. J .
Inhibitors of the Protease from Human Immunodeficiency
Virus: Synthesis, Enzyme Inhibition, and Antiviral Activity of
(7) For the fixation of the side chain orientation of Phe see (a) Zechel,
C.; Kessler, H.; Geiger, R. Synthesis and Conformational Studies
of Enkephalin-like Cyclic Peptides and Depsipeptides. In Pep-
tides, Structure and Function (Proceedings of the 9th American
Peptide Symposium); Deber, C. M., Hruby, V. J ., Kopple, K. D.,
Eds.; Pierce Chemical Co.: Rockford, IL, 1985; pp 507-510. (b)
Tourwe, D.; Verschueren, K.; Frycia, A.; Davis, P.; Porreca, F.;
Hruby, V. J .; Toth, G.; J aspers, H.; Verheyden, P.; van Binst,
G. Conformational Restriction of Tyr and Phe Side Chains in
Opioid Peptides: Information About Preferred and Bioactive
Side-Chain Topology. Biopolymers 1996, 38, 1-12.
(8) For reviews, see: (a) Stammer, C. H. Cyclopropane Amino Acids
(2,3- and 3,4-Methanoamino Acids). Tetrahedron 1990, 46,
2231-2254. (b) Burgess, K.; Ho,, K.-K.; Moye-Sherman, D.
Asymmetric Syntheses of 2,3-Methanoamino Acids. SynLett.
1994, 575-583.
(9) See also: Melnick, M. J .; Bisaha, S. N.; Gammill, R. B.,
Conformationally Restricted P1-P1′ Transition State Analogues.
Synthesis of 1(R), 3(R) [1(S), 2(S)] 3-[3-Cyclohexyl-2-[(Boc)-
amino]-1-hydroxypropyl]-2,2-dimethylcyclopropane Carboxylic
Acids. Tetrahedron Lett. 1990, 31, 961-964.
(10) (a) Chesnut, D. B.; Marsh, R. E. The Crystal Structure of
Cyclopropanecarbohydrazide. Acta Crystallogr. 1958, 11, 413-
419. (b) Long, R. E.; Maddox, H.; Trueblood, K. N. The Crystal
and Molecular Structure of Cyclopropanecarboxamide. Acta
Crystallogr. 1969, B25, 2083-2094. (c) Lynch, V. M.; Austin, R.
E.; Martin, S. F.; George, T. Determination of the Absolute
Configuration of a Novel Dipeptide Isostere. Acta Crystallogr.
1991, C47, 1345-1347. (d) Itai, A.; Toriumi, Y.; Saito, S.;
Kagechika, H.; Shudo, K. Preference of cis-Amide Structure in
N-Acyl-N-methylanilines. J . Am. Chem. Soc. 1992, 114, 10649-
100650.
(11) (a) Page, M. I.; J encks, W. P. Entropic Contributioins to Rate
Accelerations in Enzymatic and Intramolecular Reactions and
the Chelate Effect. Proc. Natl. Acad. Sci. U.S.A. 1971, 68, 1678-
1683. (b) J encks, W. P. On the Attribution and Additivity of
Binding Energies. Proc. Natl. Acad. Sci. U.S.A. 1981, 78, 4046-
4050. (c) Searle, M. S.; Williams, D. H. The Cost of Conforma-
tional Order: Entropy Changes in Molecular Associations. J .
Am. Chem. Soc. 1992, 114, 10690-10697. (d) Gerhard, U.;
Searle, M. S.; Williams, D. H. The Free Energy Change of
Restricting a Bond Rotation in the Binding of Peptide Analogues
to Vancomycin Group Antibiotics. BioMed. Chem. Lett. 1993, 3,
803-808 and references therein.
(12) (a) Fersht, A. R., The Hydrogen Bond in Molecular Recognition.
Trends Biol. Sci. 1987, 12, 301-304. (b) Searle, M. S.; Williams,
D. H.; Gerhard, U. Partitioning of Free Energy Contributions
in the Estimation of Binding Constants: Residual Motions and
Consequences for Amide-Amide Hydrogen Bond Strengths. J .
Am. Chem. Soc. 1992, 114, 10697-10704. (c) Kato, Y.; Conn,
M. M.; Rebek, J r., J . Hydrogen Bonding in Water Using
Synthetic Receptors. Proc. Natl. Acad. Sci. U.S.A. 1995, 92,
1208-1212.
(13) (a) Martin, S. F.; Austin, R. E.; Oalmann, C. J .; Baker, W. R.;
Condon, S. L.; DeLara, E.; Rosenberg, S. H.; Spina, K. P.; Stein,
H. H.; Cohen, J .; Kleinert, H. D. 1,2,3-Trisubstituted Cyclopro-
panes as Conformationally Restricted Peptide Isosteres: Ap-
plication to the Design and Synthesis of Novel Renin Inhibitors.
J . Med. Chem. 1992, 35, 1710-1721. (b) Baker, W. R.; J ae, H.-
S.; Martin, S. F.; Condon, S. L.; Stein, H. H.; Cohen, J .; Kleinert,
H. D. Conformationally Restricted Peptide Isosteres. 2. Synthesis
and In Vitro Potency of Dipeptide Renin Inhibitors Employing
a 2-Alkylsulfonyl-3-Phenylcyclopropane Carboxamide as a P3
Amino Acid Replacement. Bioorg. Med. Chem. Lett. 1992, 2,
1405.
a
Series of Compounds Containing the Dihydroxyethylene
Transition-State Isostere. J . Med. Chem. 1993, 36, 941-952.
(f) Waller, C. L.; Oprea, T. I.; Giolitti, A.; Marshall, G. R. Three-
Dimensional QSAR of Human Immunodeficiency Virus (I)
Protease Inhibitors. 1. A CoMFA Study Employing Experimen-
tally-Determined Alignment Rules. J . Med. Chem. 1993, 36,
4152-4160. (g) Lam, P. Y. S.; J adhav, P. K.; Eyermann, C. J .;
Hodge, C. N.; Ru, Y.; Bacheler, L. T.; Meek, J . L.; Otto, M. J .;
Rayner, M. M.; Wong, Y. N.; Chang, C.-H.; Weber, P. C.; J ackson,
D. A.; Sharpe, T. R.; Erickson-Viitanen, S. Rational Design of
Potent, Bioavailable, Nonpeptide Cyclic Ureas as HIV Protease
Inhibitors. Science 1994, 263, 380-384. (h) Smith, A. B., III;
Hirschmann, R.; Pasternak, A.; Guzman, M. C.; Yokoyama, A.;
Sprengeler, P. A.; Darke, P. L.; Emini, E. A.; Schleif, W. A.
Pyrrolinone-Based HIV Protease Inhibitors. Design, Synthesis
and Antiviral Activity: Evidence for Improved Transport. J . Am.
Chem. Soc. 1995, 117, 11113-11123.
(17) Leading references to structural studies of HIV-1 protease, see:
(a) Miller, M.; Swain, A. L.; J asko´lski, M.; Sathyanarayana, B.
K.; Marshall, G. R.; Rich, D. H.; Kent, S. B. H.; Wlodawer, A.
X-ray Analysis of HIV-1 Protease and Its Complexes with
Inhibitors. In Retroviral Proteases: Control of Maturation and
Morphogenesis; Pearl, L., Ed; Macmillan Press: New York, 1990,
p 93-106. (b) Wlodawer, A., Erickson, J . W. Structure-Based
Inhibitors of HIV-1 Protease. Annu. Rev. Biochem. 1993, 62,
543-585. (c) Dreyer, G. B.; Boehm, J . C.; Chenera, B.; DesJ ar-
lais, R. L.; Hassell, A. M.; Meek, T. D.; Tomaszek, T. A., J r. A
Symmetric Inhibitor Binds HIV-1 Protease Asymmetrically.
Biochemistry 1993, 32, 937-947. (d) Hosur, M. V.; Bhat, T. N.;
Kempf, D. J .; Baldwin, E. T.; Liu, B.; Gulnik, S.; Wideburg, N.
E.; Norbeck, D. W.; Appelt, K.; Erickson, J . W. Influence of
Stereochemistry on Activty and Binding Modes for C2 Symmetry-
Based Diol Inhibitors of HIV-1 Protease. J . Am. Chem. Soc. 1994,
116, 847-855. (e) Smith, A. B., III; Hirschmann, R.; Pasternak,
A.; Yao, W. Sprengeler, P. A.; Holloway, M. K.; Kuo, L. C.; Chen,
Z.; Darke, P. L.; Schleif, W. A. An Orally Bioavailable Pyrroli-
none Inhibitor of HIV-1 Protrease: Computational Analysis and
X-ray Crystal Structure of the Enzyme Complex. J . Med. Chem.
1997, 40, 2440-2444.
(18) Kempf, D. J .; Sowin, T. J .; Doherty, E. M.; Hannick, S. M.;
Codavoci, L.; Henry, R. F.; Green, B. E.; Spanton, S. G.; Norbeck,
D. W. Stereocontrolled Synthesis of C2-Symmetric and Pseudo-
C2-Symmetric Diamino Alcohols and Diols for Use in HIV
Protease Inhibitors. J . Org. Chem. 1992, 57, 5692-5700.
(19) (a) Doyle, M. P.; Pieters, R. J .; Martin, S. F.; Austin, R. E.;
Oalmann, C. J .; Mu¨ller, P. High Enantioselectivity in the
Intramolecular Cyclopropanation of Allyl Diazoacetates Using
a Novel Rhodium(II) Catalyst. J . Am. Chem. Soc. 1991, 113,
1423-1424. (b) Martin, S. F.; Spaller, M. R.; Liras, S.; Hart-
mann, B. Enantio- and Diastereoselectivity in the Intramolecular
Cyclopropanation of Secondary Allylic Diazoacetates. J . Am.
Chem. Soc. 1994, 116, 4493-4494. (c) Doyle, M. P.; Austin, R.
E.; Bailey, A. S.; Dwyer, M. P.; Dyatkin, A. B.; Kalinin, A. V.;
Kwan, M. M. Y.; Liras, S.; Oalmann, C. J .; Pieters, R. J .;
Protopopova, M. N.; Raab, C. E.; Roos, G. H. P.; Zhou, Q.-L.;
(14) Martin, S. F.; Oalmann, C. J .; Liras, S. Cyclopropanes as
Conformationally Restricted Peptide Isosteres. Design and
Synthesis of Novel Collagenase Inhibitors, Tetrahedron 1993,
49, 3521-3532.
(15) Bovy, P. R.; Tjoeng, F. S.; Rico, J . G.; Rogers, T. E.; Lindmark,
R. J .; Zablocki, J . A.; Garland, R. B.; McMackins, D. E.;
Dayringer, H.; Toth, M. V.; Zupec, M. E.; Rao, S.; Panzer-Knodle,
S. G.; Nicholson, N. S.; Salyers, A.; Taite, B. B.; Herin, M.;

Cyclopropane-Derived Peptidomimetics
J ournal of Medicinal Chemistry, 1998, Vol. 41, No. 10 1597
Martin, S. F. Enantioselective Intramolecular Cyclopropanations
of Allylic and Homoallylic Diazoacetates and Diazoacetamides
Using Chiral Dirhodium (II) Carboxamide Catalysts. J . Am.
Chem. Soc. 1995, 117, 5763-5775.
(42) Bhat, T. N.; Baldwin, E. T.; Liu, B.; Cheng, Y.-S. E.; Erickson,
J . W. Crystal-Structure of a Tethered Dimer of HIV-1 Proteinase
Complexed with an Inhibitor. Nat. Struct. Biol. 1994, 1, 552-
556.
(43) Gulnik, S. V.; Suvorov, L. I.; Liu, B.; Anderson, B.; Mitsuya, H.,
Erickson, J . W. Kinetic Characterization and Cross-Resistance
Patterns of HIV-1 Protease Mutants Selected under Drug
Pressure. Biochemistry 1995, 34, 9282-9287.
(44) Kageyama, S.; Mimoto, T.; Murakawa, Y.; Nomizu, M.; Ford,
H., J r.; Shirasaka, T.; Gulnik, S.; Erickson, J .; Takada, K.;
Hayashi, H.; Broder, S.; Kiso, Y.; Mitsuya, H. Invitro Anti-
Human-Immunodeficiency-Virus (HIV) Activities of Transition-
State HIV Protease Inhibitors Containing Allophenylnorstatine.
Antimicrob. Agents Chemother. 1993, 37, 810-817.
(45) Kessler, H.; Gehrke, M.; Griesinger, C. Two-Dimensional NMR
Spectroscopy: Background and Overview of the Experiments.
Angew. Chem., Int. Ed. Engl. 1988, 27, 490-536.
(46) Neuhaus, D.; Williamson, M. The Nuclear Overhauser Effect in
Structural and Conformational Analysis; VCH: Weinheim,
Germany, 1989.
(47) Griesinger, C.; Ernst, R. R. Frequency Offset Effects and Their
Elimination in NMR Rotating-Frame Cross-Relaxation Spec-
troscopy. J . Magn. Reson. 1987, 75, 261-271.
(48) Otting, G.; Wu¨thrich, K. Heteronuclear Filters in Two-Dimen-
sional [Proton-Proton]-NMR Spectroscopy: Combined Use with
Isotope Labelling for Studies of Macromolecular Conformation
and Intermolecular Interactions. Q. Rev. Biophys. 1990, 23, 39-
96.
(49) Kim, Y.; Prestegard, J . H. Measurement of Vicinal Couplings
from Cross Peaks in COSY Spectra. J . Magn. Reson. 1989, 84,
9-13.
(50) Keeler, J .; Neuhaus, D.; Titman J . J . A Convenient Technique
for the Measurement and Assignment of Long-Range Carbon-
13 Proton Coupling Constants. J . Chem. Phys. 1988, 146, 545-
548.
(51) (a) Hagler, A. T.; Lifson, S.; Dauber, P. Consistent Force Field
Studies on Intramolecular Forces in Hydrogen-Bonded Crystals.
2. A Benchmark for the Objective Comparison of Alternative
Force Fields. J . Am. Chem. Soc. 1979, 101, 5122-5130. (b)
DISCOVER Version 95.0, Biosym Technologies, 10065 Barnes
Canyon Road, San Diego, CA 92121, USA.
(52) J essen, S. M. Structure of Two Cyclopropane Derivatives: cis-
and trans-Caronic Acid. Acta Crystallogr., C 1992, 48, 106-116.
(53) (a) Cremer, D.; Gauss, J . Theoretical Determination of Molecular
Structure and Conformation. 20. Reevaluation of the Strain
Energies of Cyclopropane and Cyclobutane-CC and CH Bond
Energies, 1,3 Interactions, and σ-Aromticity. J . Am. Chem. Soc.
1986, 108, 7467-7477. (b) Cremer, D.; Gauss, J . Theoretical
Determination of Molecular Structure and Conformation. 16.
Substituted Cyclopropanes-An Electron Density Model of Sub-
stituent-Ring Interactions. J . Am. Chem. Soc. 1985, 107, 3811-
3819.
(54) (a) Bystrov, V. F. Spin-Spin Coupling and the Conformational
States of Peptide Systems. Prog. Nucl. Magn. Reson. Spectrosc.
1976, 10, 41-81. (b) Kessler, H.; Griesinger, C.; Wagner, K.
Peptide Conformations. 42. Conformation of Side Chains in
Peptides Using Heteronuclear Coupling Constants Obtained by
Two-Dimensional NMR Spectroscopy. J . Am. Chem. Soc. 1987,
109, 6927-6933. (c) Wang, A. C.; Bax, A. Reparametrization of
the Karplus Relation for 3J (HR-N) and 3J (HΝ-C1) in Peptides
Form Uniformly 13C/15N-Enriched Human Ubiquitin. J . Am.
Chem. Soc. 1995, 117, 1810-1813. (d) Wang, A. C.; Bax, A.
Determination of the Backbone Dihedral Angles φ in Human
Ubiquitin from Reparameterixed Emperical Karplus Equations.
J . Am. Chem. Soc. 1996, 118, 2483-2494.
(55) Mierke, D. F.; Kessler, H. Molecular Dynamics with Dimethyl
Sulfoxide as a Solvent. Conformation of a Cyclic Hexapeptide.
J . Am. Chem. Soc. 1991, 113, 9466-9470.
(56) Saulitis, J .; Mierke, D. F.; Byk, G.; Gilon, C.; Kessler, H.
Conformation of Cyclic Analogues of Substance P: NMR and
Molecular Dynamics in Dimethyl Sulfoxide. J . Am. Chem. Soc.
1992, 114, 4818-4827.
(57) Lunney, E. A.; Hagen, S. E. Domagala, J . M.; Humblet, C.;
Kosinski, J .; Tait, B. D.; Warmus, J . S.; Wilson, M.; Ferguson,
D.; Hupe, D.; Tummino, P. J .; Baldwin, E. T.; Bhat, T. N.; Liu,
B.; Erickson, J . W. A Novel Nonpeptide HIV-1 Protease Inhibi-
tor: Elucidation of the Binding Mode and Its Application in the
Design of Related Analogs. J . Med. Chem. 1994, 37, 2664-2677.
(58) Rashin, A. A. Aspects of Protein Energetics and Dynamics. Prog.
Biophys. Mol. Biol. 1993, 70, 73-200.
(20) Corey, E. J .; Myers, A. G. Efficient Synthesis and Intramolecular
Cyclopropanation of Unsaturated Diazoacetic Esters. Tetrahe-
dron Lett. 1984, 23, 3559-3562.
(21) (a) Basha, A.; Lipton, M.; Weinreb, S. M. A Mild, General Method
for Conversion of Esters to Amides. Tetrahedron Lett. 1977,
4171-4174. (b) Sibi, M. P. Chemistry of N-Methoxy-N-Methy-
lamides. Application in Synthesis. A Review. Org. Prep. Proc.
Int. 1993, 25, 15-40.
(22) This mixture of 18c and 20 could not be readily separated, but
the corresponding carboxylic acids could be easily separated by
flash chromatography.
(23) Schleucher, J .; Quant, J .; Glaser, S. J .; Griesinger, C. A Theorem
Relating Cross Relaxation and Hartmann-Hahn Transfer in
Multiple-Pulse Sequences. Optimal Suppression of TOCSY
Transfer in ROESY. J . Magn. Reson. A 1995, 112, 144-151.
(24) Eberstadt, M; Gemmecker, G.; Mierke, D. F.; Kessler, H. Scalar
Coupling Constants- Their Analysis and Their Application for
the Elucidation of Structures. Angew. Chem., Int. Ed. Engl. 1995,
34, 1671-1695.
(25) Kim Y.; Prestegard, J . H. Motional Properties of Acyl Carrier
Protein: Effects on NMR Structural Data. Proteins Struct.
Funct. Genet. 1990, 8, 377-385.
(26) Mierke, D. F.; Kessler, H. Improved Molecular Dynamics
Simulations for the Determination of Peptide Structures. Biopoly-
mers 1993, 33, 1003-1017.
(27) Mierke, D. F.; Kessler, H. Combined Use of Homo- and Hetero-
nuclear Coupling Constants as Restraints in Molecular Dynam-
ics Simulations. Biopolymers 1992, 32, 1277-1282.
(28) Nilges, M. A Calculation Strategy for the Structure Determina-
tion of Symmetrical Dimers by 1H NMR. Proteins: Struct. Funct.
Genet. 1993, 17, 297-309.
(29) Wu¨thrich, K.; Billeter, M.; Braun, W. Sequential Resonance
Assignments in Protein Proton Nuclear Magnetic Resonanace
Spectra. Computation of Sterically Allowed Proton-Proton
Distances and Statistical Analysis of Proton-Proton Distances
in Single-Crystal Protein Conformations. J . Mol. Biol. 1983, 169,
946-961.
(30) Kessler, H.; Schmitt, W. in Encyclopedia of Nuclear Magnetic
Resonance; Grant, D. M., Harris, R. K., Eds.; J ohn Wiley &
Sons: Chichester, England, 1996; Vol. 6, pp 3527-3537.
(31) (a) Pachler, K. G. R. Nuclear Magnetic Resonance (NMR) Study
of Some R-Amino Acids. I. Coupling Constants in Alkaline and
Acid Medium. Spektochim. Acta 1963, 19, 2085-2092. (b)
Pachler, K. G. R. Nuclear Magnetic Resonance Study of Some
R-Amino Acids. II. Rotational Isomerism. Spektochim. Acta 1964,
20, 581-587.
(32) Frisch, M. J .; et al. Gaussian 94 (Revision D.4); Gaussian, Inc.:
Pittsburgh, PA, 1995.
(33) Becke, A. D. Universality in the Peroxidase-Oxidase Reaction:
Period Doublings, Chaos, Period Three, and Unstable Limit
Cycles. J . Phys. Chem. 1993, 98, 5648-5642.
(34) Lee, C.; Yang, W.; Parr, R. G. Various Functionals for the Kinetic
Energy Density of an Atom or Molecule. Phys. Rev. 1988, B37,
785-789.
(35) Baldwin, E. T.; Bhat, T. N.; Gulnik, S. V.; Liu, B.; Topol, I. A.;
Kiso, Y.; Mimoto, T.; Mitsuya, H.; Erickson, J . W. Structure of
HIV-1 Protease with KN-272, a Tight Binding Allophenylnorsta-
tine. Structure 1995, 3, 581-590.
(36) Ohno, Y.; Kiso, Y.; Kobayashi, Y. Solution Conformations of KNI-
272, a Tripeptide HIV Protease Inhibitor Designed on the Basis
of Substrate Transition State: Determinated by NMR Spectros-
copy and Simulated Annealing Calculations. Bioorg. Med. Chem.
1996, 4, 1565-1572.
(37) Preuss, R.; Schmidt, R. R. C-Glucoside Durch Direkte 1-C-
Lithiierung von 2-Phenylsulfinyl Aktiviertem D-Glucal. (C-
Glycosides from Direct C(1)-Lithiation of 2-Phenylsulfinyl Ac-
tivated D-Glucals.) Liebigs Ann. Chem. 1989, 429-434.
(38) Tan, B.; Chia, L. H. L.; Huang, H.; Kuok, M.; Tang, S. Rotational
Isomerism in 2,3-Dinitro-2,3-dimethylbutane. J . Chem. Soc.,
Perkin Trans. 1984, 2, 1407-1415.
(39) Topol, I. A., unpublished results.
(40) (a) de Meijere, A. Bonding Properties of Cyclopropane and Their
Chemical Consequences. Angew. Chem., Int. Ed. Eng. 1979, 18,
809-826. (b) J orgenson, M. J .; Leung, T. Cyclopropyl Conjuga-
tion in Olefinic Esters. Conformational Effects on Ultraviolet
Absorption. J . Am. Chem. Soc. 1968, 90, 3769-3774.
(41) Hoffmann, R. W. Flexible Molecules with Defined Shape-
Conformational Design. Angew. Chem., Int. Ed. Engl. 1992, 31,
1124-1134.
J M980033D