β-strand mimicking macrocyclic amino acids: Templates for protease inhibitors with antiviral activity

Article abstract of DOI:10.1021/jm010414i

New amino acids are reported in which component macrocycles are constrained to mimic tripeptides locked in a β-strand conformation. The novel amino acids involve macrocycles functionalized with both an N- and a C-terminus enabling addition of appendages a

Full text of DOI:10.1021/jm010414i

J . Med. Chem. 2002, 45, 371-381
371
â-Str a n d Mim ick in g Ma cr ocyclic Am in o Acid s: Tem p la tes for P r otea se
In h ibitor s w ith An tivir a l Activity
Matthew P. Glenn,† Leonard K. Pattenden,† Robert C. Reid,† David P. Tyssen,‡ J oel D. A. Tyndall,†
Christopher J . Birch,*,‡ and David P. Fairlie*,†
Centre for Drug Design and Development, Institute for Molecular Bioscience, University of Queensland, Brisbane Qld 4072,
Australia, and Victorian Infectious Diseases Reference Laboratory, Private Bag 815, Carlton South, Victoria 3053, Australia
Received September 5, 2001
New amino acids are reported in which component macrocycles are constrained to mimic
tripeptides locked in a â-strand conformation. The novel amino acids involve macrocycles
functionalized with both an N- and a C-terminus enabling addition of appendages at either
end to modify receptor affinity, selectivity, or membrane permeability. We show that the cycles
herein are effective templates within inhibitors of HIV-1 protease. Eleven compounds originating
from such bifunctionalized cyclic templates are potent inhibitors of HIV-1 protease (Ki 0.3-50
nM; pH 6.5, I ) 0.1 M). Unlike normal peptides comprising amino acids, five of these macrocycle-
containing compounds are potent antiviral agents with sub-micromolar potencies (IC50 170-
900 nM) against HIV-1 replication in human MT2 cells. The most active antiviral agents are
the most lipophilic, with calculated values of LogD6.5 g 4. All molecules have a conformationally
constrained 17-membered macrocyclic ring that has been shown to structurally mimic a
tripeptide segment (Xaa)-(Val/Ile)-(Phe/Tyr) of a peptide substrate in the extended conformation.
The presence of two trans amide bonds and a para-substituted aromatic ring prevents
intramolecular hydrogen bonds and fixes the macrocycle in the extended conformation. Similarly
constrained macrocycles may be useful templates for the creation of inhibitors for the many
other proteins and proteases that recognize peptide â-strands.
In tr od u ction
â-strand structures.14,15 Surprisingly, almost all known
protease inhibitors are conformationally flexible mol-
ecules. Yet templates that fix peptides or other small
molecules in a â-strand conformation could conceivably
be used to build potent inhibitors that are preorganized
for protease binding, leading to higher affinity for a
protease due to significant entropic advantages.
We have been investigating the potential of macro-
cyclic peptidomimetic protease inhibitors that use a
cyclic replacement for a three amino acid segment of a
peptide substrate that is in a â-strand conformation. We
have demonstrated the use of certain conformationally
constrained macrocycles (e.g., 1 and 2) as excellent
structural mimics of the â-strand of a protease-binding
peptide, reproducing both H-bond donor/acceptor prop-
erties of a peptide as well as pocket-binding properties
of peptide side chains.16-20 Such macrocycles are highly
resistant to degradation by gastric, serum, and cellular
proteases and are lipophilic structural components of
agents with potent activity against HIV in vitro.21 This
approach to inhibitors has the potential for numerous
other proteases of the aspartic, serine, and metallo
classes.22
To date, antiviral macrocyclic peptidomimetics have
been constructed by adding appendages to either the
C-terminus (e.g., of 1) or the N-terminus (e.g., of 2) of a
macrocycle.21 In this paper, we significantly extend their
utility by modifying the macrocyclic templates to ac-
commodate both N- and C-terminal functional groups,
leading to macrocyclic amino acids (e.g., 3) that can be
readily inserted into longer peptide sequences or elabo-
rated by adding various N- and/or C-terminal nonpep-
tidic appendages to the cycle. These new macrocyclic
Accelerated access during the past decade to struc-
tures of macromolecular receptors is enabling more
effective and faster structure-based drug design,1-3
which is finally realizing its long held promise as a
powerful method for the rational design and develop-
ment of bioactive compounds and drug leads. This has
been particularly exemplified in the generation of
inhibitors of proteolytic enzymes in general,4,5 with
HIV-1 protease inhibitors6-13 being important examples
of structure-based designed compounds that have shown
clinical utility.
On the other hand, there is little or no structural
information available for most proteases, necessitating
the design of inhibitors based upon the sequence of
peptide substrates that are recognized and processed
by the enzyme.5 This has led to a more circuitous route
to inhibitor development, since substrates are by defini-
tion not optimized for high affinity binding to proteases.
Instead, substrates need to have low (micro- to milli-
molar) affinities for proteases in order for there to be
rapid turnover. Methods are consequently still needed
to improve the process of development of protease
inhibitors from the limited information provided by
protease substrates. One little studied approach stems
from the now established paradigm that proteases
commonly, if not universally, recognize their peptide
substrates/inhibitors in the form of extended “saw-tooth”
* To whom correspondence should be addressed. D.P.F.: E-mail,
d.fairlie@imb.uq.edu.au; Fax, +61733651990. C.J .B.: E-mail,
Chris.Birch@mh.org.au.
† University of Queensland.
‡ Victorian Infectious Diseases Reference Laboratory.
10.1021/jm010414i CCC: $22.00 © 2002 American Chemical Society
Published on Web 12/12/2001

372 J ournal of Medicinal Chemistry, 2002, Vol. 45, No. 2
Glenn et al.
F igu r e 1. Top: Superimposition of macrocyclic inhibitor (9g, white) on acyclic peptide J G365 (yellow) bound within the active
site of HIV-1 protease (oxygen, red; nitrogen, blue; carbon, white and yellow, respectively). The surface of the HIV-1 protease
active site (pale blue) is from the protease-J G365 structure.23 Bottom: Hydrogen bonding (dashed lines) of acyclic peptidic inhibitor
J G365 (left)23 and macrocyclic inhibitors (right)20 to HIV-1 protease.
amino acids (e.g., 3) are representative of an important
new class of molecular building blocks or templates that
rigidly maintain their structural integrity, chemical
stability, and receptor-binding properties typical of the
peptides that they were designed to mimic.
in 1 .21 We consequently chose the 17-membered mac-
rocyclic templates 3-5 as target macrocyclic amino
acids for synthesis, allowing ready derivatization to
HIV-1 protease inhibitors.
An illustration of how strand-mimicking cycles such
as 3-5 are able to reproduce the same interactions with
HIV-1 protease as made by corresponding peptides23
from acyclic inhibitors of HIV-1 protease is shown in
Figure 1. The structure for a potential HIV-1 protease
inhibitor (Figure 1, white, 9g) incorporating macrocycle
4 was derived from the coordinates of a previously
published HIV-1 protease bound macrocyclic inhibitor.21b
This was energy minimized with template forcing24 onto
the structure23 of a heptapeptide inhibitor, the substrate
analogue J G365, bound to HIV-1 protease (Figure 1,
yellow). A total of twelve carbon and nitrogen atoms of
Resu lts
Design of Ma cr ocyclic Am in o Acid s. In previous
work, we had incorporated 14-17-membered rings as
in 1 and 2 into inhibitors of HIV-1 protease and found
their three-dimensional structures to be identical to the
extended â-strand of acyclic peptides complexed within
the active site of HIV-1 protease.16,17,20,21 We also found
that these cycles were unlike peptides in being pro-
teolytically stable and conferring antiviral activity,
particularly when a pentamethylene linker was used as

Templates for Protease Inhibitors
J ournal of Medicinal Chemistry, 2002, Vol. 45, No. 2 373
the peptide backbone of J G365 between the amide
nitrogen of serine and the hydroxyl-bearing carbon of
the hydroxyethylamine transition state isostere (P4 to
P1) were used as the template. The template forcing led
to an energy minimized structure for the proposed
macrocyclic inhibitor that was 4.0 kcal mol-1 lower in
energy than the non-template forced energy minimized
structure (144.1 vs 140.0 kcal mol-1), clearly indicating
this to be an accessible conformation.
chains are theoretically able to stretch out into a bigger
enzyme pocket or curl up to accommodate a smaller
binding site. We envisage that this, and other limited
flexibility in the cycle, may be key features in enabling
inhibitors to “breathe” and thereby still maintain po-
tency and binding to resistant mutants by varying their
shape in response to changes in the substrate-binding
pocket. In this paper, we restrict our assessment of these
templates to whether they confer protease inhibition
and to what extent they are lipophilic enough to confer
antiviral properties in cell culture. Detailed studies on
the viral resistance of similar complexes will be reported
subsequently.
Superimposition of the template-forced structure
(white) onto J G365 (yellow) using the same atoms
reveals excellent structural mimicry (RMSD of 0.22 Å).
The macrocycle fits neatly into the active site of HIV-1
protease, with carbonyl oxygens and amide nitrogens
located in the appropriate positions to make the same
hydrogen-bonding interactions as J G365 does with the
enzyme. The serine and the amine nitrogen at the
N-terminus of the cycle are also positioned and orien-
tated appropiately to make the same H-bonds as the
corresponding region of J G365. There is also ample
space in the enzyme to accommodate the flexible nor-
leucine side chain at P2, the aliphatic linker at P3, and
the tyrosinyl aromatic ring at P1.
Syn th esis. The macrocycles 3-5, designed to mimic
P1 to P3 residues of peptide substrates for proteases,
were created from tripeptide 6 assembled using stan-
dard solution phase peptide coupling procedures. The
P3 residue, (2S)-amino-6-hydroxyhexanoic acid, was
derived from diazotization of N-Boc-L-lysine with sodium
nitroprusside under basic aqueous conditions (pH 9).28
Macrocyclization was effected by stepwise halogenation
of the primary hydroxyl group of 6 (OHfBrfI), followed
by nucleophilic displacement of halide by the phenol of
the tyrosine at P1. Hydrolysis of the C-terminal ester
provided the macrocyclic acids 7a -c (Scheme 1).
To investigate the utility of the macrocyclic amino
acids 3-5, we decided to test them as inhibitors of
HIV-1 protease. Like the tripeptides they mimic, these
compounds are not very potent inhibitors of this enzyme
in their own rite (IC50 5-50 µM). We therefore created
a limited set of derivatives as putative inhibitors featur-
ing one of the cycles 3-5, a hydroxyethylamine transi-
tion state isostere, a series of N-alkyl benzenesulfon-
amides at the C-terminus, and either an amine, an
amide, or an amino acid at the N-terminus.
On the C-terminal side of the macrocycle, the hy-
droxyethylamine isotere is identical to that in J G365.
The N-alkylbenzene sulfonamide, while occupying dif-
ferent space to the Pro-Ile-Val portion of J G365, occupies
the same space as this same moiety in other inhibitors
of HIV-1 protease.21 In summary, the macrocyclic amino
acid component of the inhibitors 9a -k not only mimics
the H-bonding and space-filling properties of the P3 to
P1 section of J G365 but also, importantly, is an effective
template or scaffold for N- and C-terminal appendages
at P4, P1′, and P2′ and does not interfere with their
binding in the enzyme.
Each of the three macrocycles 3-5 is formed by
linking together P1 and P3 side chains derived from
amino acids. Each cycle is significantly constrained by
the presence of two trans amide bonds and the aromatic
side chain of tyrosine. As a result, the amide backbone
exists in the extended â-strand rather than folding into
a turn conformation. Elsewhere, we show that the 17-
membered constrained macrocycle is flexible enough to
permit rotation of the component aromatic ring, whereas
the 16-membered ring with one less methylene causes
restricted rotation of the aromatic ring (Reid et al.,
submitted for publication). Cycle 3 contains the same
valine side chain at P2 as do cycles 1 and 2, thereby
allowing a close comparison. On the other hand, com-
pounds 4 and 5 do not contain branched side chains at
P2 like 3 but rather have more flexible unbranched
aliphatic substituents.
Conversion of the acids 7a -c to their mixed anhy-
drides with isobutylchloroformate and subsequent ad-
dition of diazomethane allowed one carbon homologation
at the C-terminus. The resulting diazomethyl ketones
were treated with 1 equiv of dry HBr, and the resulting
bromomethyl ketones were reduced with sodium boro-
hydride (d.e. > 50%). The 13(R)-diastereomers (major
product) were separated by reverse phase high perfor-
mance liquid chromatography (rpHPLC),29 before con-
version to the key epoxides 8a -c. Ring opening of the
epoxide with isoamylamine affected formation of the
hydroxyethylamine transition state isostere, which was
condensed with 4-acetamidobenzenesulfonyl chloride.
Various combinations of deprotection (N-terminal depro-
tection with TFA or removal of both the N-terminal Boc
group and the phenyl acetamido group with 2 M HCl)
and N-terminal coupling with amino acids afforded the
inhibitors 9a -k . Inhibitors 9a -c employ macrocycle 3
containing the small branched substituent of valine at
P2 within the 17-membered cycle formed through con-
nection of amino acid side chain substituents at P1 and
P3. Inhibitors 9d -i contain macrocycle 4 with norleu-
cine at P2, while 9j,k have macrocycle 5 and propyl
serine at P2.
The rationale for these choices stems from the knowl-
edge that viral resistance can originate from mutations
in the protease that directly or indirectly influence the
sizes of the “pockets” or indentations in the lining of
the substrate-binding groove of HIV-1 protease.25-27
Regardless of where mutations occur within the pro-
tease structure, they can potentially reduce the comple-
mentary fit between enzyme and inhibitor/substrate.
The norleucine and propyl serine side chains in 4 and
5, respectively, were considered to offer more flexibility
in how they could bind at S2 in response to enzyme
mutations that alter this binding site. The linear side
P r ot ea se In h ib it ion . Table 1 reports the in vitro
inhibitor potencies against HIV-1 protease for a range
of compounds (9a -k ) containing macrocyclic amino
acids 3-5. All of the compounds have Ki < 50 nM under
the assay conditions employed (pH 6.5, I ) 0.1 M30,31).

374 J ournal of Medicinal Chemistry, 2002, Vol. 45, No. 2
Glenn et al.
Sch em e 1a
a
Reagents: (a) TPP, CBr4. (b) (1) Nal, acetone, heat; (2) K2CO3, DMF. (c) NaOH, MeOH. (d) IBCF, THF, CH2N2. (e) HBr, THF. (f)
NaBH4, EtOH. (g) NaOMe, THF. (h) Isoamylamine, EtOH. (i) Benzenesulfonyl chloride, THF:H2O. (j) 2 M HCl, MeOH, reflux. (k) TFA.
(l) N-Boc-Serine, HBTU, DIPEA, DMF. (m) Pentanoic acid, HBTU, DIPEA, DMF. * Where necessary.
In the series 9a , 9d , and 9j, the respective Ki values
0.9, 0.3, and 45 nM indicate that there is little difference
between the isopropyl and the n-butyl side chains at
P2, although the longer propyl methyl ether side chain
of 9j is less favored. The purpose of this latter side chain
is related to its capacity to fit the S2 pocket in mutant
proteases, and thus, 9j may still be important in
resistance studies not reported here.
An tivir a l Activity. None of the compounds possess-
ing free amino termini (9a , 9d , and 9j), which are likely
protonated under the assay conditions and thus charged
and highly water soluble, show antiviral activity at sub-
micromolar concentrations against HIV-1-infected MT2
cells (Table 1). These compounds have calculated33 LogD
values at pH 6.5 of 1.0, 1.7, and 0.5, respectively.
Similarly, compounds with Ser or AcSer added at P4
(9c and 9g) are too polar (LogD6.5 ) 0.4, 0.6) to penetrate
membranes and did not inhibit HIV replication in cell
culture at 10 µM, although some activity was seen when
an additional Boc group was added to the serine to
increase lipophilicity (9h , LogD6.5 ) 3.0).
When the LogD6.5 increases to 1.5-3.6, some activity
against HIV-1 is observed (9d , 9f, 9h , and 9k ) at
inhibitor concentrations of 500 nM-5 µM. When the
lipophilicity is further increased to LogD6.5 ) 4-5 (e.g.,
9b, 9e, and 9i) more substantial antiviral activity was
observed at inhibitor concentrations of 100-600 nM,
with values for the selective index (SI) lying between
17 and 40 (Table 1). The selectivity index (defined in
the Experimental Section) is a comparitive measure of
selective toxicity for the virus over MT2 cells. While
overall these compounds are not as active against HIV-1
as clinically proven inhibitors (e.g., ritonavir and am-
prenavir), Table 1 shows that the SI value is roughly
proportional to LogD6.5 (r2 ) 0.56, not shown) recogniz-
ing that some of the compounds are inactive. This
indicates that LogD has some value in guiding further
inhibitor optimization.
At the N-terminus, the unsubstituted free amino
group is surprisingly better (Table 1) than when capped
by acetyl, Boc, or Ser (e.g., 9a > 9b, 9f, and 9g) although
not as good as AcSer (9c). We had expected that the
acetyl, serine, and acetylserine N-termini would have
contributed additional hydrogen-bonding donors/accep-
tors, as would the serine side chain, that are known in
peptides to interact with the enzyme (see P4-P3 region
in Figure 1). However, a possible explanation for the
observed higher potency for the free amino N-terminus
is that the latter is protonated (RNH3+) under the assay
conditions and in the enzyme and therefore may be able
to H-bond more extensively with the enzyme. We have
previously reported that amines are indeed protonated
within the active site groove of HIV-1 protease.17 The
Boc group does not appear to be suited at P4 (9e, 9h ,
and 9k ) since it does not increase potency of the
inhibitors, and this is likely because it is exposed to
solvent and because the tert-butyl substitutent does not
fit P4 well. The N-terminal substituent at P4 in 9i was
added to increase lipophilicity for enhanced cell uptake
and demonstrates a 2-fold improvement in activity
against HIV-infected cells as compared to 9f.
Overall, the results indicate that the macrocyclic
amino acids 3-5 can indeed function as effective
templates for the creation of potent protease inhibitors
with excellent antiviral properties. Although they con-
sist of several H-bond donors and acceptors that con-
tribute to the water solubility of the inhibitors, com-
pounds containing the cycles are still membrane
At the C-terminus, the N-alkyl benezenesulfonamide
substituent is able to effectively occupy P1′ and P2′ and
confer protease activity as well as antiviral activity, as
shown elsewhere by VX47832 and our own compounds.21
Acetylation of the p-amino substituent on the benzene-
sulfonamide (9b, 9e, 9h , and 9k ) leads to 2-10-fold
reduced inhibitory potency (Table 1).

Templates for Protease Inhibitors
J ournal of Medicinal Chemistry, 2002, Vol. 45, No. 2 375
Ta ble 1.
permeable and bioavailable depending upon what is
appended to the cyclic template.
conformation and rely on a combination of interactions
with the enzyme for high affinity. When enzyme muta-
tions occur that have an impact on the shape of the
active site, significant attenuation of inhibitor activity
may result, without comparable retardation of substrate
processing.
Discu ssion
A significant problem with current antiviral treat-
ments based on inhibition of HIV-1 protease is the rapid
onset of viral resistance and the limited number of
options available after initial failure with mono or
combination drug therapy. Reported inhibitors of HIV-1
protease are in general flexible molecules that are
severely restricted in their mode of high affinity binding.
They must bind to the protease in only a single
Our approach to peptidomimetic inhibitors, which
very closely mimic the structures of protease substrates,
is based on the hypothesis that drug resistance will be
more difficult for substrate mimics since the protease
must still recognize its substrates to form a functional
and infectious virion. This is in contrast to current

376 J ournal of Medicinal Chemistry, 2002, Vol. 45, No. 2
Glenn et al.
Varian 300 MHz NMR spectrometer. Proton assignments were
made using 2D COSY and TOCSY experimental data. Pre-
parative scale rpHPLC separations were performed on a
Waters Delta-Pak C18 40 × 100 mm cartridges (100 Å) using
gradient mixtures of water/0.1% TFA and water (10%)/aceto-
nitrile (90%)/0.1% TFA. Pure fractions were lyophilized under
high vacuum (0.1 mm Hg) and were determined free from
solvent/TFA by 1H and 13C NMR. Analytical scale rpHPLC was
performed on a Phenomenex Luna 5µ C18(2) 250 × 4.6 mm
column. Accurate mass determinations were performed on an
API QSTAR mass spectrometer using electron impact ioniza-
tion. Water-octanol partition coefficients (Log D) were calcu-
lated using PALLAS prolog D 2.1. Molecular modeling was
performed on an SGI Octane R12000, with minimization
calculations performed with the cff91 force field using the
Discover Module within InsightII.24b Volumes given as mil-
liliters per millimole in general procedures refer to the limiting
reagent.
therapeutics that have pronounced structural differ-
ences from substrates. Mutations in response to sub-
strate analogues will obviously lead to an inability to
process the corresponding substrates and thus stop viral
replication. Our first step toward the goal of low
resistance inhibitors that closely resemble substrates
has been to develop macrocyclic amino acids for incor-
poration to protease inhibitors that consequently main-
tain a substrate-like composition. Unlike peptide sub-
strates, the macrocyclic tripeptide analogues have been
shown to be resistant to hydrolytic/proteolytic degrada-
tion17,21 and have much higher affinities for HIV-1
protease than substrates.15-21 Some analogues (e.g., 9b,
9d -f, 9i, and 9k ) are cell permeable (unlike substrates)
and, except for 9d , also exhibit potent antiviral activity
being active at sub-micromolar concentrations in cell
culture.21
The results in Table 1 demonstrate that macrocyclic
amino acids such as 3-5 can successfully be used as
templates to create potent inhibitors of both HIV-1
protease and replication of HIV-1 in cell culture. The
results for inhibition of HIV-1 protease are consistent
with the macrocycles being pre-organized in protease-
binding â-strand mimetics as demonstrated by modeling
in Figure 1 and by crystallographic studies of analogues
incorporating similar cycles 1 and 2 bound to HIV-1
protease.20 Both linear and branched substituents are
tolerated at P2, and this suggests that the strategy of
making the P2 substituent more flexible to accom-
modate variable sizes of S2 in mutant proteases may
be feasible. Additional flexibility within a reasonably
constrained macrocycle comes from the pentamethylene
linker between P1 and P3. The free amino terminus is
best masked by lipophilic substituents to prevent its
protonation and polarization and to enhance its mem-
brane permeability over water solubility. The C-termi-
nal N-alkyl benzensulfonamide substituent is effective
at P1′ and P2′ in conferring high protease-binding
affinity to the macrocycles.
En zym e Assa y. Synthetic HIV-1 PR (SF2 isolate)34 with
mutations (C67B, C95B, Q7K, L33I, where B ) L-R-amino-n-
butyric acid) was solubilized in 6 M Gu‚HCl (0.05 mg mL-1
)
and then refolded for 60 min in buffer A (20 mM phosphate,
pH 7.0, 20% v/v glycerol, 10 mg mL-1 bovine serum albumin).
At time zero, the buffer A/protease solution was added to an
excess of the fluorogenic substrate Abz-NF-6*,30 buffer B (MES,
pH 6.5, 37 °C, 100 mM NaCl, 10% v/v glycerol), and varying
concentrations of the inhibitor. Km values for HIV-1 PR were
calculated using a Hanes plot of [s]/v vs [s], and the Ki value
was then calculated from either Dixon plots of 1/v vs [s] or
Henderson plots of [I]/{1-(Vi/Vo)} vs Vo/Vi in cases where the
Ki value was found to approach the concentration of the
enzyme. Under these assay conditions, the cyclic compounds
were confirmed to be competitive inhibitors.
Vir ology. Anti-HIV activity was assessed in acutely in-
fected MT2 cells by measuring the extent to which the
compounds reduced HIV specific cytopathic effects at noncyto-
toxic concentrations.35 The medium used for passage and
maintenance of MT2 cells was RF10, an RPMI-1640-based
medium.36 Inhibition of HIV-1 replication was confirmed by
measurement of virion-associated reverse transcriptase (RT)
activity in cell supernatants.34 All tests were performed in 48
well tissue culture plates (Costar). Each drug concentration
and all controls were tested in duplicate. Virus inoculum was
added immediately after the addition of the drugs to the cells.
MT2 cells were counted and resuspended at a concentration
of 5 × 104 per 0.25 mL of RF10 medium. Compounds at twice
the final concentration required in the test were prepared in
RF10 medium from a 20 mM stock of the drugs prepared in
60% v/v ethanol. A total of 0.5 mL of each dilution was pipetted
into duplicate test and cytotoxicity control wells, followed by
the addition of 0.25 mL of cell suspension to all wells. A total
of 0.5 mL of RF10 medium was added to the virus-only
controls, 0.75 mL to the cell-only controls, and 0.25 mL of
HIV-1 strain to the cytotoxicity controls. A volume of 0.25 mL
of strain #23728836 diluted in the appropriate medium to
contain 200 TCID50 was then immediately added to virus-
only control wells and the test wells. The plates were then
incubated at 37 °C in 5% CO2. All wells in the test were
examined by light microscopy and harvested after 6-7 days
incubation, at which time viable cell numbers were determined
in cell-only and cytotoxicity control wells using trypan blue
dye exclusion counting.
Inhibitory potency against HIV-1 protease correlates
roughly with the number of H-bonding and pocket-filling
substituents in the inhibitor. Antiviral potency requires
potent protease inhibition but also correlates roughly
with the degree of hydrophobicity for compounds of
similar protease inhibitory potency. We are currently
investigating both the capacity of HIV-1 to develop
resistance to these and similar macrocyclic compounds
in cell culture as well as the activities of such substrate
mimetics against drug-resistant mutant proteases. Pre-
liminary data from cellular studies supports the concept
that certain substrate analogues, which are constrained
to mimic a â-strand but have some limited elasticity in
the projection of its side chains, are less sensitive to
resistance than known inhibitors of this enzyme.
The cell-only, virus-only, and test cultures were harvested
by removing 0.8 mL of supernatant from each well to separate
microtubes. The virus present in each supernate was then
precipitated using poly(ethylene glycol) and assayed for virion-
associated RT activity expressed as counts per minute (cpm)
on a liquid scintillation counter as previously described.35 The
percentage (%) inhibition was calculated by expressing the RT
activity at each dilution as a percent of the cpm in virus-only
controls subtracted from 100. The SI, which enables direct
comparison of the antiviral activity of the test compounds, was
calculated using the formula SI ) CC10/IC50, where CC10 was
the concentration of drug permitting at least 90% viability of
Exp er im en ta l Meth od s
Abbr evia tion s. Ac ) acetyl; Boc ) tert-butoxycarbonyl;
DIPEA ) diisopropylamine; DMF ) dimethylformamide; BOP
) (benzotriazol-1-yloxy)-tris(dimethylamino)phosphonium hexa-
fluorophosphate; FCC ) flash column chromatography; HBTU
) [(benzotriazolyl)oxy]-N′,N′,N′,N′-tatramethyluronium hexaflu-
orophosphate; IBCF ) isobutylchloroformate; Nle ) norleu-
cine; Ser ) serine; TFA ) trifluoroacetic acid; THF ) tetrahy-
drofuran; Tyr ) tyrosine; Val ) valine.
Gen er a l Meth od s. 1H nuclear magnetic resonance (NMR)
spectra were recorded on either a Bruker ARX 500 MHz or a

Templates for Protease Inhibitors
J ournal of Medicinal Chemistry, 2002, Vol. 45, No. 2 377
cells in the absence of infecting HIV-1 and IC50 was the
concentration of drug producing a 50% reduction in virion-
associated RT activity in HIV-1-infected MT2.
sulfate, and the solvent was removed under reduced pressure
to yield the bromomethyl ketone.
Con ver sion of Br om om eth yl Keton es to Br om oh y-
d r in s (P r oced u r e H). A solution of the bromomethyl ketone
(1 equiv) in ethanol (50 mL/mmol) was cooled to -15 °C, and
then, sodium borohydride (1.2 equiv) was added. After the
solution was stirred for 5 min, glacial acetic acid was added
(0.2 mL/mmol), and the solvent was removed under reduced
pressure. The residue was dissolved in DMF, and the dia-
sterisomers were separated and purified by rpHPLC.
Con ver sion of Br om oh yd r in s to Ep oxid es (P r oced u r e
I). A solution of sodium methoxide (2 equiv) and bromohydrin
(1 equiv) in tetrahydrofuran (25 mL/mmol) and methanol (75
mL/mmol) was stirred at room temperature for 15 min, and
then, water (10 mL/mmol) was added. The solution was
concentrated under reduced pressure, and the precipitate of
the epoxide was collected on a frit. The precipitate was washed
with 30% aqueous methanol and then dried under vacuum.
Con d en sa tion of Am in es w ith Ep oxid es (P r oced u r e J ).
A solution of the epoxide (1 equiv), the required amine (2-10
equiv), and DIPEA (2 equiv) in ethanol (50 mL/mmol) was
refluxed overnight. The solvent and, where possible, the excess
amine were removed under reduced pressure to provide the
condensed product in all cases as a solid, which was purified
where necessary by rpHPLC.
Gen er a l Syn t h et ic P r oced u r es. N-Boc Dep r ot ect ion
(P r oced u r e A). The Boc-protected amine (1 equiv) was stirred
for 1 h in TFA (1 mL/mmol) at room temperature. The TFA
was removed under reduced pressure, and the residue was
triturated with diethyl ether. The resulting semisolid was then
used directly in the next coupling without further purification.
P ep tid e Cou p lin g (P r oced u r e B). To a stirred solution
of the acid (1 equiv) in THF (2 mL/mmol) at room temperature
was added HBTU (1 equiv) and DIPEA (2 equiv). After 5 min,
a solution of the amine (1 equiv) and DIPEA (2 equiv) in THF
(1 mL/mmol) was added, and stirring was continued for a
further 30 min. After this time, the THF was removed under
reduced pressure, and the residue was taken up into ethyl
acetate (10 mL/mmol) and washed with equal portions of 1 M
HCl, NaHCO3, and brine. The organic layer was dried over
magnesium sulfate, and the solvent was removed under
reduced pressure to yield the peptide.
Br om in a tion of Alcoh ols (P r oced u r e C). Triphenylphos-
phine (1.5 equiv) was added to a cold solution (0 °C) of the
alcohol (1 equiv) in THF (10 mL/mmol). Once all of the
phosphine had dissolved, tetrabromomethane (1.2 equiv) was
added in small portions over 10 min. After the bromomethane
was completely added, the solution was stirred for a further
30 min at room temperature, and then, the solvent was
removed under reduced pressure. The resulting semisolid was
purified by column chromatography on silica gel.
Con d en sa tion of Am in es w ith Su lfon yl Ch lor id es
(P r oced u r e K). The sulfonyl chloride (3 equiv) was added to
a solution of the amine (1 equiv) in 4:1 THF:water (50 mL/
mmol). DIPEA (4 equiv) was added, and the solution was
stirred at room temperature for 30 min. The organics were
extracted into ethyl acetate, and the solvent was removed
under reduced pressure. The resulting sulfonamides were
purified by rpHPLC where necessary.
Ma cr ocycliza tion (P r oced u r e D). The bromide (1 equiv)
was converted to the iodide by refluxing with sodium iodide
(1.5 equiv) in acetone (10 mL/mmol) for 4 h. The solution was
cooled to room temperature, and the precipitate of sodium
bromide was removed by filtration. The filtrate was reduced
to dryness, and the iodide was dissolved in DMF (20 mL/
mmol). Finely ground anhydrous potassium carbonate (5
equiv) was then added, and the resulting suspension was
stirred overnight. The solvent volume was reduced under
vacuum (∼5 mL), and the residue was partitioned between
ethyl acetate and water. The aqueous layer was discarded, and
the organic layer was washed successively with dilute sodium
thiosulfate solution, 2 M HCl, saturated NaHCO3, and brine.
The organic layer was dried over magnesium sulfate, and the
solvent was removed under reduced pressure to yield the crude
macrocycle, which can be purified by flash column chroma-
tography on silica gel.
(2S)-ter t-Bu t oxyca r b on yla m in o-6-h yd r oxyh exa n oic
Acid . To a suspension of N-Boc-L-lysine (31.8 g, 129.3 mmol)
in water (500 mL) at 60 °C was added 4 M NaOH (50 mL).
Sodium nitroprusside (58.8 g, 197.3 mmol) was then added in
portions over 1 h, while maintaining the pH at ∼9.5 by the
addition of a further 50 mL of 4 M NaOH as required. After
the nitroprusside was completely added, the resulting red
suspension was stirred for a further 6 h at 60 °C. The resulting
suspension was cooled to ∼10 °C, and 1 M HCl (∼500 mL)
was added until a pH of ∼1 was achieved. The resulting
solution was extracted with ethyl acetate (3 × 500 mL), the
combined organic fractions were dried over magnesium sulfate,
and solvent was removed under reduced pressure. The crude
product was contaminated with ∼30% elimination products,
which were removed by flash column chromatography (silica
gel, 70% EtOAc, 29% petroleum spirit, 1% HOAc) to provide
(2S)-tert-butoxycarbonylamino-6-hydroxyhexanoic acid (14.6 g,
Ester Hyd r olysis (P r oced u r e E). A solution of 2 M NaOH
(2.5 equiv) was added to a solution of the ester (1 equiv) in
methanol (5 mL/mmol) at 0 °C, and the solution was stirred
for 1 h. The solution was neutralized with citric acid, and the
methanol was removed under reduced pressure. Water (5 mL/
mmol) was added, and the acid was extracted into ethyl
acetate. The organic phase was dried over magnesium sulfate,
and the solvent was removed under reduced pressure to yield
the crude acid, which could be purified by rpHPLC.
1
45.7%) as a white solid. H NMR (CDCl3, 300 MHz): 5.21 (br
d, 1H); 4.28 (m, 1H); 3.66 (t (6.0 Hz), 2H); 1.41 (s, 9H); 1.90 to
1.37 (m, 6H). 13C NMR (CDCl3, 75 MHz): 175.3, 156.5, 80.3,
62.1, 53.7, 32.4, 32.3, 28.5, 22.0.
(2S)-[(2S)-((2S)-ter t-Bu toxyca r bon yla m in o-6-h yd r oxy-
h exa n oyla m in o)-3-m eth yl-bu tyr yla m in o]-3-(4-h yd r oxy-
p h en yl)p r op ion ic Acid Meth yl Ester (6a ). This tripeptide
was synthesized sequentially from tyrosine, valine, and (2S)-
tert-butoxycarbonylamino-6-hydroxyhexanoic acid using the
standard Boc deprotection (procedure A) and coupling (proce-
Con ver sion of Acid s to Dia zom eth ylk eton se (P r oce-
d u r e F ). N-Methylmorpholine (1.5 equiv) was added to a
solution of the acid (1 equiv) in THF (20 mL/mmol) at -15 °C.
Iso-butyl chloroformate (1.25 equiv) was added, and the
solution was stirred for 15 min at -15 °C before an ethereal
solution of diazomethane (∼20 equiv) was added. The reaction
was stirred for a further 30 min before a stream of argon was
introduced to remove the excess diazomethane. Ethyl actetate
was added, and the solution was washed with brine and dried
over magnesium sulfate. Removal of the solvent under reduced
pressure yielded the diazomethyl ketone as a pale yellow solid.
1
dure B) procedures above. H NMR (CDCl3, 300 MHz): 7.09,
(d (8.0 Hz), 1H); 6.89 (d (8.1 Hz), 2H); 6.69 (d (7.3 Hz), 2H);
5.40 (br d, 1H); 4.80 (m, 1H); 4.25 (m, 1H); 4.07 (m, 1H); 3.61
(s, 3H); 3.57 (t (5.5 Hz), 2H); 2.98 (d (6.3 Hz), 2H); 2.1 to 1.3
(m, 7H); 1.40 (s, 9H); 0.85 (apparent t (7.1 Hz), 6H). 13C NMR
(CDCl3, 75 MHz): 172.4, 171.7, 171.0, 155.7, 155.6, 130.2,
126.6, 115.5, 80.1, 69.3, 58.2, 54.2, 53.4, 52.2, 36.9, 33.3, 32.1,
31.1, 28.2, 26.4, 19.0, 18.0.
(7S)-ter t-Bu t oxyca r b on yla m in o-(10S)-isop r op yl-8,11-
d ioxo-2-oxa -9,12-d ia za -bicyclo[13.2.2]n on a d eca -1(18),15-
(19),16-tr ien e-(13S)-ca r boxylic Acid (7a ). Tripeptide 6a (2.7
g, 5.2 mmol) was brominated (procedure C, FCC, 60%),
followed by cyclization (procedure D, FCC, 61%), and the
resulting macrocyclic ester was hydrolyzed (procedure E, 81%)
Con ver sion of Dia zom eth yl Keton es to Br om om eth yl
Keton es (P r oced u r e G). A solution of dry HBr in ethyl
acetate (∼0.2 M) was added in small portions to the diazo-
methyl ketone (1 equiv) in THF (25 mL/mmol) at -10 °C. The
reaction was monitored by TLC and EIMS, and when complete,
ethyl acetate (20 mL) was added, and the solution was washed
with brine. The organic layer was dried over magnesium

378 J ournal of Medicinal Chemistry, 2002, Vol. 45, No. 2
Glenn et al.
1
to afford the title acid. H NMR (CD3OD, 300 MHz): 6.86 (d
acetic acid (procedure B, rpHPLC, 62%) to afford the title
amide. 1H NMR (CD3OD, 500 MHz): 7.48 (d (6.8 Hz), 2H);
7.06 (d (7.4 Hz), 2H); 6.75 (d (7.8 Hz), 2H); 6.69 (d (6.83 Hz),
2H); 4.40 (t (5.5), 1H); 4.26 (m, 2H); 4.09 (m, 3H); 3.70 (m,
3H); 3.20 (m, 3H); 3.10 (m, 1H); 2.92 (dd (14.5 Hz, 8.6 Hz),
1H); 2.39 (t (12.8 Hz), 1H); 2.00 (s, 3H); 1.90 (m, 1H); 1.72 (m,
1H); 1.55 to 1.35 (m, 7H); 1.19 (m, 2H); 0.87 (d (6.5 Hz), 6H);
0.86 (d (6.8 Hz), 3H); 0.75 (d (6.8 Hz), 3H). HRMS (EI) m/z
calcd for C37H57N6O9S (MH+), 761.3903; found, 761.3892.
Analytical rpHPLC: isocratic 60A:40B rt ) 5.9 min; gradient
100A to 100B 20 min rt ) 18.6 min.
(7.6 Hz), 2H); 6.56 (d (7.8 Hz), 2H); 4.54 (m, 1H); 3.92 (m, 1H);
3.82 (m, 1H); 3.64 (m, 1H); 3.08 (m, 2H); 2.40 (t (12.2 Hz), 1H);
1.8 to 1.3 (m, 5H); 1.35 (s, 9H); 1.01 (m, 2H); 0.72 (d (6.8 Hz),
3H); 0.63 (d (6.8 Hz), 3H). 13C NMR (CD3OD, 75 MHz): 174.2,
174.1, 172.1, 159.2, 158.2, 134.3, 131.5 (br), 118.1 (br), 69.5,
68.5, 59.7, 55.1, 54.8, 38.8, 34.1, 33.1, 31.4, 30.6, 29.3, 20.2,
19.1.
((10S)-Isop r op yl-(13S)-[(R)-oxir a n yl]-8,11-d ioxo-2-oxa -
9,12-d ia za -bicyclo[13.2.2]n on a d eca -1(18),15(19),16-tr ien -
(7S)-yl)-ca r ba m ic Acid ter t-Bu tyl Ester (8a ). Macrocyclic
acid 7a (460 mg, 0.93 mmol) was converted to the diazomethyl
ketone (procedure F, 85%), which was subsequently converted
to the bromomethyl ketone (procedure G, 69%), reduced
(procedure H, rpHPLC (diastereomeric separation), 36%), and
finally converted to the title epoxide (procedure I, 85%). 1H
NMR (CDCl3, 300 MHz): 7.00 (d (8.2 Hz), 2H); 6.79 (d (8.7
Hz), 2H); 6.26 (d (8.4 Hz), 1H); 5.42 (d (9.5 Hz), 1H); 5.02 (d
(8.3 Hz), 1H); 4.28 (m, 2H); 4.11 (m, 1H); 3.90 (m, 2H); 3.08
(m, 2H); 2.81 (m, 2H); 2.42 (t (12.6 Hz), 1H); 1.9 to 1.1 (m,
7H); 1.38 (s, 9H); 0.85 (d (6.9 Hz), 3H); 0.82 (d (7.02 Hz), 3H).
13C NMR (CDCl3, 75 MHz): 171.8, 170.1, 157.3, 155.9, 130.1,
128.6, 116.6, 79.8, 67.8, 58.2, 54.2, 53.6, 50.9, 45.7, 35.4, 32.0,
31.9, 29.3, 28.3, 22.1, 18.9, 18.0.
(2S)-[(2S)-((2S)-ter t-Bu toxyca r bon yla m in o-6-h yd r oxy-
h exa n oyla m in o)h exa n oyla m in o]-3-(4-h yd r oxy-p h en yl)-
p r op ion ic Acid Meth yl Ester (6b). This tripeptide was
synthesized sequentially from (2S)-tert-butoxycarbonylamino-
6-hydroxyhexanoic acid, norleucine, and tyrosine using the
standard Boc deprotection (procedure A) and coupling (proce-
1
dure B) procedures above. H NMR (CDCl3, 300 MHz): 7.05
(m, 2H); 6.88 (d (8.5 Hz), 2H); 6.68 (d (8.5 Hz), 2H); 5.39 (br d,
1H); 4.72 (m, 1H); 4.39 (m, 1H); 4.08 (m, 1H); 3.67 (s, 3H);
3.51 (t (4.8 Hz), 1H); 3.1 to 2.6 (m, 3H); 1.8 to 1.1 (m, 12H);
1.38 (s, 9H); 0.80 (t (7.6 Hz), 3H). 13C NMR (CDCl3, 75 MHz):
172.6, 171.8, 171.6, 155.9, 155.5, 130.3, 126.6, 115.6, 80.2, 62.0,
54.3, 53.4, 53.2, 52.4, 38.6, 36.8, 31.9, 31.7, 28.3, 27.5, 22.3,
21.8, 13.8.
((13S)-{2-[(4-Acetyla m in o-ben zen esu lfon yl)-(3-m eth yl-
bu tyl)-a m in o]-(1R)-h yd r oxy-eth yl}-(10S)-isop r op yl-8,11-
d ioxo-2-oxa -9,12-d ia za -bicyclo[13.2.2]n on a d eca -1(18),15-
(19),16-tr ien -(7S)-yl)-ca r ba m ic Acid ter t-Bu tyl Ester (9b).
Epoxide 8a (46 mg, 94 µmol) was condensed with isoamyl-
amine (procedure J , 94%), followed by addition of 4-acetamido-
benzenesulfonyl chloride (procedure K, rpHPLC, 44%) to afford
the title compound. 1H NMR (CD3OD, 500 MHz): 7.76, (s, 4H);
7.01 (br d, 2H); 6.70 (br d, 2H); 4.25 (m, 1H); 4.08 (m, 3H);
3.88 (m, 1H); 3.75 (m, 1H); 3.46 (m, 1H); 3.37 (m, 1H); 3.18
(m, 2H); 2.98 (m, 1H); 2.38 (t (12.5 Hz), 1H); 2.14 (s, 3H); 1.89
(m, 1H); 1.72 (m, 1H); 1.63 (m, 1H); 1.50 (m, 5H); 1.40 (s, 9H);
1.18 (m, 2H); 0.88 (m, 9H); 0.76 (d (6.5 Hz), 3H). 13C NMR
(CD3OD, 125 MHz): 173.3, 172.3, 171.6, 158.6, 158.3, 144.2,
132.2, 132.1, 130.0 (br), 129.4, 120.5, 117.7 (br), 74.3, 69.0, 58.6,
55.5, 55.3, 55.2, 53.1, 38.3, 36.4, 33.9, 33.0, 30.7, 28.7, 27.2,
24.0, 23.5, 23.0, 22.8, 20.1, 17.9. HRMS (EI) m/z calcd for
(7S)-ter t-Bu toxyca r bon yla m in o-(10S)-bu tyl-8,11-d ioxo-
2-oxa -9,12-d ia za bicyclo[13.2.2]n on a d eca -1(18),15(19),16-
tr ien e-(13S)-ca r boxylic Acid (7b). Tripeptide 6b (1.17 g, 2.2
mmol) was brominated (procedure C, FCC, 64%), followed by
cyclization (procedure D, FCC, 72%), and the resulting mac-
rocyclic ester was hydrolyzed (procedure E, rpHPLC, 55%) to
1
afford the title acid. H NMR (CDCl3, 500 MHz): 7.14 (d (8.5
Hz), 1H); 7.03 (br s, 2H); 6.90 (d (8.5 Hz), 1H); 6.79 (d (8.0
Hz), 2H); 5.35 (d (8.0 Hz), 1H); 5.02 (m, 1H); 4.38 (m, 1H);
4.19 (m, 2H); 4.02 (m, 1H); 3.8 (br s, 1H); 3.40 (dd (4.5 Hz, 14
Hz), 1H); 2.65 (t (12.5 Hz), 1H); 1.8 to 1.4 (m, 6H); 1.41 (s,
9H); 1.3 to 1.1 (m, 6H); 0.81 (t (6.5 Hz), 3H). 13C NMR (CD3-
OD, 125 MHz): 174.2, 171.4, 171.1, 157.7, 155.6, 131.0 (br),
128.1, 116.0 (br), 80.3, 67.8, 54.2, 52.8, 52.5, 38.1, 33.3, 32.4,
29.6, 28.3, 27.0, 22.4, 13.7.
((10S)-Bu tyl-(13S)-[(R)-oxir a n yl]-8,11-d ioxo-2-oxa -9,12-
d ia za bicyclo[13.2.2]n on a d eca -1(18),15(19),16-tr ien -(7S)-
yl)-ca r ba m ic Acid ter t-Bu tyl Ester (8b). Macrocyclic acid
7b (120 mg, 0.24 mmol) was converted to the diazomethyl
ketone (procedure F, 91%), which was subsequently converted
to the bromomethyl ketone (procedure G, 90%), reduced
(procedure H, rpHPLC (diastereomeric separation), 51%), and
finally converted to the title epoxide (procedure I, 93%). 1H
NMR (CDCl3, 300 MHz): 7.01 (br s, 2H); 6.85 (d (8.0 Hz), 2H);
6.48 (d (8.5 Hz), 1H); 6.18 (d (8.0 Hz), 1H); 5.40 (d (8.0 Hz),
1H); 4.28 (m, 1H); 4.17 (m, 3H); 3.95 (m, 1H); 3.1 (m, 1H); 2.95
(dd (4.4 Hz, 14 Hz), 1H); 2.8 (m, 2H); 2.40 (t (14 Hz), 1H); 1.8
to 1.4 (m, 6H); 1.41 (s, 9H); 1.3 to 1.1 (m, 6H); 0.81 (t (6.5 Hz),
3H). 13C NMR (CD3OD, 125 MHz): 170.7, 170.7, 157.5, 154.8,
130.5, 128.7, 118.1, 80.1, 68.1, 54.3, 53.7, 52.8, 50.5, 44.7, 34.6,
33.4, 32.2, 29.6, 28.3, 27.1, 22.5, 22.1, 13.8.
((13S)-{2-[(4-Acetyla m in o-ben zen esu lfon yl)-(3-m eth yl-
bu tyl)-am in o]-(1R)-h ydr oxy-eth yl}-(10S)-bu tyl-8,11-dioxo-
2-oxa -9,12-d ia za -bicyclo[13.2.2]n on a d eca -1(18),15(19),16-
tr ien -(7S)-yl)-car bam ic Acid ter t-Bu tyl Ester (9e). Isoamyl-
amine was condensed with epoxide 8b (20 mg, 40 µmol)
(procedure J , 85%), followed by addition of 4-acetamidoben-
zenesulfonyl chloride (procedure K, rpHPLC, 55%) to afford
the title compound. 1H NMR (CD3OD, 500 MHz): 7.84 (d (13.5
Hz), 1H); 7.65 (m, 4H); 7.28 (d (8.5 Hz), 1H); 6.95 (br s, 2H);
6.65 (br d, 2H); 4.18 (m, 1H); 4.04 (m, 2H); 3.91 (m, 1H); 3.77
(m, 1H); 3.60 (m, 1H); 3.32 (m, 1H); 3.20 (m, 2H); 3.04 (m,
2H), 2.86 (dd (6.5 Hz, 14.5 Hz), 1H); 2.27 (t (13 Hz), 1H); 2.04
(s, 3H); 1.8 to 1.4 (m, 7H); 1.30 (s, 9H); 1.2 to 1.1 (m, 8H); 0.71
(d (6.5 Hz), 6H); 0.69 (t (6.5 Hz), 3H). 13C NMR (CD3OD, 125
MHz): 173.0, 171.9, 161.5, 158.1, 157.3, 144.1, 135.2, 131.84,
131.8 (br), 129.1, 120.4, 116.0 (br), 81.0, 74.5, 68.6, 55.5, 55.0,
53.5, 52.9, 37.9, 36.4, 34.8, 33.0, 30.7, 28.6, 28.3, 26.9, 24.0,
23.3, 23.1, 22.8, 22.6, 14.2. HRMS (EI) m/z calcd for C40H62N5O9S
C
39H60N5O9S (MH+), 774.4107; found, 774.4106. Analytical
rpHPLC: isocratic 30A:70B rt ) 6.5 min; gradient 100A to
100B 20 min rt ) 23.4 min.
4-Am in o-N-[2-((7S)-a m in o-(10S)-isop r op yl-8,11-d ioxo-
2-oxa -9,12-d ia za -bicyclo[13.2.2]n on a d eca -1(18),15(19),16-
t r ien -(13S)-yl)-(2R)-h yd r oxy-et h yl]-N-(3-m et h yl-b u t yl)-
ben zen esu lfon a m id e (9a ). Amide 9b (20.0 mg, 26 µmol) was
refluxed for 2 h in methanol (1 mL) with 3 drops of 2 M HCl.
After the solution was cooled to room temperature, water (1
mL) was added, and the solution was purified by rpHPLC to
provide the amine as a white powder (9.7 mg, 60%) after
1
lyophilization. H NMR (CD3OD, 500 MHz): 7.50, (d (9.8 Hz,
2H); 7.10 (br s, 2H); 6.72 (br s, 2H); 6.67 (d (9.8 Hz), 2H); 4.34
(m, 1H); 1.19 (d (5.7 Hz), 1H); 4.07 (m, 2H); 3.89 (m, 1H); 3.73
(m, 1H); 3.42 (dd (14.4 Hz, 4.7 Hz), 1H); 3.32 (obscured, 1H);
3.21 (dd (13.8 Hz, 3.38 Hz), 1H); 3.07 (m, 1H); 2.87 (dd (14.7
Hz, 8.4 Hz), 1H); 2.38 (t (12.6 Hz), 1H); 1.98 (m, 1H); 1.77 (m,
2H); 1.64 (m, 2H); 1.49 (m, 2H); 1.38 (m, 1H); 1.19 (m, 2H);
0.91 (d (6.8 Hz), 3H); 0.86 (d (6.5 Hz), 6H); 0.77 (d (6.8 Hz),
3H). 13C NMR (CD3OD, 125 MHz): 171.7, 171.4, 157.3, 154.3,
132.5, 130.4 (br), 130.3, 126.4, 116.5 (br), 114.5, 74.2, 67.8, 58.6,
55.4, 54.2, 53.1, 38.3, 36.1, 33.7, 32.4, 29.8, 27.1, 23.0, 22.8,
20.9, 20.0, 17.6. HRMS (EI) m/z calcd for C32H50N5O9S (MH+),
632.3477; found, 632.3473. Analytical rpHPLC: isocratic 50A:
50B rt ) 3.5 min; gradient 100A to 100B 20 min rt ) 17.4
min.
(2S)-Acetyla m in o-N-((13S)-{2-[(4-a m in o-ben zen esu lfo-
n yl)-(3-m eth yl-bu tyl)a m in o]-(1R)-h yd r oxy-eth yl}-(10S)-
isop r op yl-8,11-d ioxo-2-oxa -9,12-d ia za -b icyclo[13.2.2]-
n on adeca-1(18),15(19),16-tr ien -(7S)-yl)-3-h ydr oxy-pr opion -
a m id e (9c). Amine 9a (3.94 mg, 6.2 µmol) was coupled with
N-Boc-L-serine (procedure B), subsequently deprotected (pro-
cedure A, rpHPLC, 50% over two steps), and then coupled with

Templates for Protease Inhibitors
J ournal of Medicinal Chemistry, 2002, Vol. 45, No. 2 379
1
(MH+), 788.4263; found, 788.4245. Analytical rpHPLC: iso-
cratic 20A:80B rt ) 7.6 min; gradient 100A to 100B 20 min rt
) 24.1 min.
a white powder (3.76 mg, 48%) after lyophilization. H NMR
(CD3OD, 500 MHz): 7.49 (d (10.7 Hz), 2H); 7.06 (br s, 2H);
6.75 (br s, 2H); 6.68 (d (9.0 Hz), 2H); 4.27 (m, 2H); 4.19 (m,
1H); 4.13 (m, 1H); 4.06 (m, 1H); 3.91 (m, 2H); 3.87 (m, 1H);
3.76 (dd (11.3 Hz, 6.9 Hz), 1H); 3.71 (m, 1H); 3.43 (dd (14.0
Hz, 3.84 Hz), 1H); 3.23 (dd (13.4 Hz, 3.4 Hz), 1H); 3.08 (m,
1H); 2.90 (dd (14.6 Hz, 7.9 Hz), 1H); 2.39 (t (12.8 Hz), 1H);
1.75 (m, 1H); 1.66 (m, 2H); 1.48 (m, 5H); 1.37 (m, 1H); 1.21
(m, 6H); 0.87 (d (7.3 Hz), 6H); 0.83 (t (7.0 Hz), 3H). 13C NMR
(CD3OD, 125 MHz): 172.9, 172.1, 167.9, 158.1, 154.5, 132.3,
131.2 (br), 130.2, 126.5, 117.5 (br), 114.5, 74.6, 68.6, 61.9, 56.2,
55.2, 54.9, 53.2, 48.2, 38.3, 36.3, 35.0, 33.3, 30.7, 28.6, 27.2,
23.6, 23.0, 22.8, 14.4. HRMS (EI) m/z calcd for C36H57N6O8S
(MH+), 733.3953; found, 733.3933. Analytical rpHPLC: iso-
cratic 60A:40B rt ) 7.3 min; gradient 100A to 100B 20 min rt
) 17.6 min.
4-Am in o-N-[2-((7S)-a m in o-(10S)-bu tyl-8,11-d ioxo-2-oxa -
9,12-d ia za -bicyclo[13.2.2]n on a d eca -1(18),15(19),16- tr ien -
(13S)-yl)-(2R)-h ydr oxy-eth yl]-N-(3-m eth yl-bu tyl)ben zen e-
su lfon a m id e (9d ). Acetamide (9e) (11.5 mg, 16.7 µmol) was
refluxed for 2 h in methanol (1 mL) with 6 drops of 2 M HCl.
After the solution was cooled to room temperature, water (1
mL) was added, and the solution was purified by rpHPLC.
Lyophilization provided the title compound as a white powder
(8.7 mg, 81%). 1H NMR (CD3OD, 500 MHz): 7.91 (d (13.3 Hz),
1H); 7.49 (d (9.10 Hz), 2H); 7.35 (d (8.33 Hz), 1H); 7.07 (br s,
2H); 6.78 (br s, 2H); 6.70 (d (9.10 Hz), 2H); 4.25 (m, 1H); 4.15
(m, 1H); 4.10 (m, 1H); 4.05 (m, 1H); 3.73 (m, 1H); 3.40 (dd
(3.14, 14.30 Hz), 1H); 3.27 (m, 1H); 3.22 (m, 2H), 3.11 (m, 1H);
2.94 (dd (8.53 Hz, 12.8 Hz), 1H); 2.39 (t (12.5 Hz), 1H); 1.8 to
P en ta n oic Acid (13(S)-{2-[(4-Am in o-ben zen esu lfon yl)-
(3-m eth yl-bu tyl)a m in o]-(1R)-h yd r oxy-eth yl}-(10S)-bu tyl-
8,11-d ioxo-2-oxa -9,12-d ia za -b icyclo[13.2.2]n on a d e ca -
1(18),15(19),16-tr ien -(7S)-yl)a m id e (9i). Amine (9d ) (5.0 mg,
7.8 µmol) was coupled with pentanoic acid (procedure B,
rpHPLC, 57%). 1H NMR (CD3OD, 500 MHz): 7.87 (br, 2H);
7.75 (br, 1H); 7.44 (d (8.2 Hz), 2H); 7.04 (br d, 2H); 6.75 (br d,
2H); 6.66 (d (7.5 Hz), 2H); 4.23 (m, 1H); 4.15 (m, 3H); 4.05 (m,
1H); 3.69 (m, 1H); 3.36 (m, 1H); 3.22 (m, 1H); 3.09 (m, 2H);
2.91 (dd (14.5 Hz, 8.5 Hz), 1H); 2.38 (t (13.9 Hz), 1H); 2.16 (t
(7.47 Hz), 2H); 1.8 to 1.1 (m, 21H); 1.0 to 0.8 (m, 12H). HRMS
1.1 (m, 15H); 0.88 (d (6.52 Hz), 6H); 0.84 (t (6.96 Hz), 3H). 13
C
NMR (CD3OD, 125 MHz): 175.1, 173.2, 158.1, 154.1, 131.9,
130.1, 114.4, 74.4, 68.5, 55.8, 55.0, 53.5, 53.0, 38.0, 36.4, 36.2,
34.8, 31.1, 28.4, 27.0, 23.4, 22.8(2C), 22.7. The four aromatic
13C of the tyrosine residue were extremely broadened and are
unassigned. HRMS (EI) m/z calcd for C33H52N5O6S (MH+),
646.3633; found, 646.3634. Analytical rpHPLC: isocratic 50A:
50B rt ) 3.9 min; gradient 100A to 100B 20 min rt ) 17.9
min.
N-((13S)-{2-[(4-Am in o-ben zen esu lfon yl)-(3-m eth yl-bu -
tyl)a m in o]-(1R)-h yd r oxy-eth yl}-(10S)-bu tyl-8,11-d ioxo-2-
oxa -9,12-d ia za -b icyclo[13.2.2]n on a d eca -1(18),15(19),16-
tr ien -(7S)-yl)a ceta m id e (9f). Amine (9d ) (11.0 mg, 16.7
µmol) was coupled with acetic acid (procedure B, rpHPLC,
64%). 1H NMR (CD3OD, 500 MHz): 7.46 (d (7.1 Hz), 2H); 7.05
(br s, 2H); 6.75 (br s, 2H); 6.67 (d (8.1 Hz), 2H); 4.24 (m, 1H);
4.12 (m, 3H); 4.03 (m, 1H); 3.72 (m, 1H); 3.37 (m, 1H); 3.25
(m, 1H); 3.20 (m, 1H); 3.08 (m, 1H); 2.90 (dd (14.2 Hz, 7.8 Hz),
1H); 2.37 (t (12.3 Hz), 1H); 1.85 (s, 3H); 1.72 (m, 1H); 1.63 (m,
2H); 1.5 to 1.3 (m, 8H); 1.20 (m, 4H); 0.85 (d (6.3 Hz), 6H);
0.81 (t (7.5 Hz), 3H). 13C NMR (CD3OD, 125 MHz): 173.0,
172.8, 172.5, 158.2, 154.3, 132.0, 130.1, 128.7 (br), 126.3, 117.3
(br), 114.4, 74.6, 68.6, 55.1, 54.8, 53.7, 53.1, 38.1, 36.4, 34.8,
33.0. 28.4, 27.0, 23.4, 23.2, 22.8, 22.7, 22.2, 14.2. HRMS (EI)
m/z calcd for C35H54N5O7S (MH+), 688.3738; found, 688.3743.
Analytical rpHPLC: isocratic 40A:60B rt ) 5.3 min; gradient
100A to 100B 20 min rt ) 20.5 min.
[1-((13S)-{2-[(4-Acetylam in o-ben zen esu lfon yl)-(3-m eth yl-
bu tyl)a m in o]-(1R)-h yd r oxy-eth yl}-(10S)-bu tyl-8,11-d ioxo-
2-oxa -9,12-d ia za -bicyclo[13.2.2]n on a d eca -1(18),15(19),16-
tr ien -(7S)-ylcar bam oyl)-(2S)-h ydr oxy-eth yl]car bam ic Acid
ter t-Bu tyl Ester (9h ). Acetamide (9e) (17.2 mg, 25.0 µmol)
was deprotected (procedure A) and coupled to N-Boc-serine
(procedure B, rpHPLC, 52%). 1H NMR (CD3OD, 500 MHz):
10.18 (s, 1H); 7.88 (d (12.0 Hz), 1H); 7.79 (d (10 Hz), 1H); 7.72
(m, 4H); 7.54 (d (8.8 Hz), 1H); 7.03 (br s, 2H); 6.72 (br d, 2H);
4.28 (m, 2H); 4.17 (m, 2H); 4.13 (m, 2H); 4.05 (m, 1H); 3.72
(m, 2H); 3.43 (dd (14.0 Hz, 3 Hz), 1H); 3.34 (m, 1H); 3.23 (dd
(13.2 Hz, 3.2 Hz), 1H); 3.17 (m, 1H), 3.00 (dd (8.1 Hz, 14.1
Hz), 1H); 2.39 (t (12.5 Hz), 1H); 2.15 (s, 3H); 1.8 to 1.4 (m,
7H); 1.43 (s, 9H); 1.2 to 1.1 (m, 8H); 0.88 (d (6.3 Hz), 6H); 0.81
(t (6.9 Hz), 3H). 13C NMR (CD3OD, 125 MHz): 173.1, 172.6,
172.1(2C), 158.0, 157.8, 144.3, 135.3, 132.1, 130.1(br), 129.4,
120.7, 117.3(br), 74.4, 68.3, 63.2, 58.1, 55.2, 54.3, 53.6, 52.9,
37.9, 36.4, 34.7, 33.1, 30.3, 28.6, 28.4, 26.9, 24.0, 23.3, 22.8,
22.7, 22.2, 14.2. HRMS (EI) m/z calcd for C43H67N6O11S (MH+),
875.4583; found, 875.4582. Analytical rpHPLC: isocratic 40A:
60B rt ) 6.9 min; gradient 100A to 100B 20 min rt ) 21.0
min.
(EI) m/z calcd for
C
39H61N5O7S (MH+), 730.4205; found,
730.4213. Analytical rpHPLC: isocratic 20A:80B rt ) 5.8 min;
gradient 100A to 100B 20 min rt ) 23.8 min.
3-Allyloxy-(2S)-ter t-b u t oxyca r b on yla m in o-p r op ion ic
Acid . A solution of N-Boc-L-serine (5.0 g, 24 mmol) in DMF
(80 mL) was added slowly to a suspension of 60% NaH
dispersion (2.4 g, 60 mmol) in DMF (20 mL) at 0 °C. After the
amino acid was completely added, allylbromide was introduced
(2.6 mL, 30 mmol), and stirring was continued for 3 h at room
temperature. Excess sodium hydride was quenched by the
careful addition of water (5 mL), and the solvent was removed
under reduced pressure. The residue was then dissolved in
water, acidified to pH ∼ 2 with 4 M HCl, and subsequently
extracted into ethyl acetate. The organic layer was dried over
magnesium sulfate, and the solvent was removed to provide
1
the title allyl ether (5.4 g, 92%). H NMR (CDCl3, 300 MHz):
5.81 (m, 1H); 5.38 (br d, 1H); 5.22 (m, 3H); 3.91 (m, 2H); 3.82
(m, 2H); 1.39 (s, 9H). 13C NMR (CDCl3, 75 MHz): 172.9, 155.5,
134.0, 117.1, 79.7, 72.1, 69.8, 53.7, 28.1.
(2S)-ter t-Bu t oxyca r bon yla m in o-3-p r op oxy-p r op ion ic
Acid . 3-Allyloxy-(2S)-tert-butoxycarbonylamino-propionic acid
(5.48 g, 22.4 mmol) was dissolved in methanol (100 mL) and
hydrogenated over Pd/C in a Parr hydrogenation vessel under
30 psi of hydrogen for 3 h. The catalyst was removed from the
methanolic solution by filtration through super cell, and the
1
solvent was removed to yield the title ether (4.42 g, 80%). H
NMR (CDCl3, 300 MHz): 5.38 (d (8.6 Hz), 1H); 4.41 (m, 1H);
3.84 (m, 1H); 3.62 (dd (9.4 Hz, 3.8 Hz), 1H); 3.36 (t (6.6 Hz),
2H); 1.52 (m, 2H); 1.40 (s, 9H); 0.83 (t (7.7 Hz), 3H). 13C NMR
(CDCl3, 75 MHz): 175.2, 155.7, 80.3, 73.2, 70.1, 53.7, 28.3, 22.5,
10.4.
(2S)-[(2S)-((2S)-ter t-Bu toxyca r bon yla m in o-6-h yd r oxy-
h e xa n oyla m in o)-3-p r op oxy-p r op ion yla m in o]-3-(4-h y-
d r oxy-p h en yl)p r op ion ic Acid Meth yl Ester (6c). This
tripeptide was synthesized sequentially from (2S)-tert-butoxy-
carbonylamino-6-hydroxyhexanoic acid, (2S)-tert-butoxycar-
bonylamino-3-propoxy-propionic acid, and tyrosine methyl
ester using the standard Boc deprotection (procedure A) and
coupling (procedure B) procedures above. 1H NMR (CDCl3, 300
MHz): 7.10 (d (8.5 Hz), 2H); 6.92 (d (8.4 Hz), 2H); 6.90 (d (7.7
Hz), 1H); 6.72 (d (7.8 Hz), 1H); 5.25 (m, 1H); 4.78 (m, 1H);
4.47 (m, 1H); 4.08 (m, 1H); 3.79 (m, 1H); 3.68 (s, 3H); 3.62 (m,
1H); 3.53 (m, 1H), 3.42 (m, 1H); 3.37 (t (7.1 Hz), 2H); 3.07 (m,
2H); 1.8 to 1.4 (m, 8H); 1.40 (s, 9H); 0.83 (t (7.1 Hz), 3H). 13C
NMR (CDCl3, 75 MHz): 171.6, 171.5, 169.9, 155.8, 149.5,
133.7, 130.4, 121.4, 80.2, 73.1, 62.1, 53.6, 53.3, 52.6, 52.4, 38.6,
32.2, 31.9, 31.8, 28.3, 22.6, 21.7, 10.4.
(2S)-Am in o-N-((13S)-{2-[(4-a m in o-ben zen esu lfon yl)-(3-
m et h yl-b u t yl)a m in o]-(1R)-h yd r oxy-et h yl}-(10S)-b u t yl-
8,11-d ioxo-2-oxa -9,12-d ia za -b icyclo[13.2.2]n on a d e ca -
1(18),15(19),16-tr ien -(7S)-yl)-3-h ydr oxy-pr opion am ide (9g).
Compound 9h (9.48 mg, 10.8 µmol) was refluxed for 2 h in
methanol (2 mL) with 6 drops of 2 M HCl. After the solution
was cooled to room temperature, water (1 mL) was added, and
the solution was purified by rpHPLC to provide the amine as

380 J ournal of Medicinal Chemistry, 2002, Vol. 45, No. 2
Glenn et al.
(7S)-ter t-Bu toxyca r bon yla m in o-8,11-d ioxo-(10S)-p r o-
p oxym et h yl-2-oxa -9,12-d ia za -b icyclo[13.2.2]n on a d eca -
1(18),15(19),16-tr ien e-(13S)ca r boxylic Acid (7c). Tripep-
tide 6c (4.5 g, 8.1 mmol) was brominated (procedure C, FCC,
72%), followed by cyclization (procedure D, FCC, 36%), and
the resulting macrocyclic ester was hydrolyzed (procedure E,
94%) to afford the title acid. 1H NMR (CD3OD, 300 MHz): 6.86
(br s, 2H); 6.57 (br s, 2H); 4.56 (m, 1H); 4.18 (m, 1H), 4.02 (m,
1H); 3.95 (m, 1H); 3.73 (m, 1H); 3.29 (m, 2H); 3.16 (m, 2H);
3.06 (m, 1H); 2.44 (t (12.7 Hz), 1H); 1.6 to 1.4 (m, 6H); 1.32 (s,
9H); 0.99 (m, 2H); 0.75 (t (7.4 Hz), 3H). 13C NMR (CD3OD, 75
MHz): 174.9, 173.9, 171.2, 160.2, 159.3, 133.7, 131.9, 117.9,
74.7, 72.7, 72.4, 69.4, 55.9, 55.2, 54.9, 39.8, 34.1, 31.2, 29.3,
27.0, 23.6, 11.4.
Refer en ces
(1) Goodwill, K. E.; Tennant, M. G.; Stevens, R. C. High-throughput
X-ray crystallography for structure-based drug design. Drug
Discovery Today 2001, 6, S113-S118.
(2) Wade, R. C. “Flu” and structure-based drug design. Structure
1997, 5, 1139-1145.
(3) Walkinshaw, M. D. Protein targets for structure-based drug
design. Med. Res. Rev. 1992, 12, 317-372.
(4) Babine, R. E.; Bender, S. L. Molecular recognition of protein-
ligand complexes: Applications to drug design. Chem. Rev. 1997,
97, 1359-1472.
(5) Leung, D.; Abbenante, G.; Fairlie, D. P. Protease Inhibitors:
Current Status and Future Prospects. J . Med. Chem.. 2000, 43,
305-341.
(6) Darke, P. L.; Huff, J . R. HIV protease as an inhibitor target for
the treatment of AIDS. Adv. Pharmacol. 1994, 25, 399-454.
(7) West, M. L.; Fairlie, D. P. Targeting HIV-1 protease: a test of
drug-design methodologies. Trends Pharmacol. Sci. 1995, 16,
67-75.
(8) Kempf, D. J .; Sham, H. L. HIV protease inhibitors Curr. Pharm.
Des. 1996, 2, 225-246.
((13S)-[(R)-Oxir a n yl]-8,11-d ioxo-(10S)-p r op oxym eth yl-
2-oxa -9,12-d ia za -bicyclo[13.2.2]n on a d eca -1(18),15(19),16-
tr ien -(7S)-yl)ca r ba m ic Acid ter t-Bu tyl Ester (8c). Macro-
cyclic acid 7c (220 mg, 0.42 mmol) was converted to the
diazomethyl ketone (procedure F, 85%), which was subse-
quently converted to the bromomethyl ketone (procedure G,
72%), reduced (procedure H, rpHPLC (diastereomeric separa-
tion), 35%), and finally converted to the title epoxide (procedure
(9) March, D. R.; Fairlie, D. P. Designing New Antiviral Drugs for
AIDS: HIV-1 Protease and its Inhibitors; R. G. Landes Co.:
Austin, Texas, 1996.
(10) Misson, J .; Clark, W.; Kendall, M. J . Therapeutic advances:
protease inhibitors for the treatment of HIV-1 infection. J . Clin.
Pharm. Ther. 1997, 22, 109-117.
1
I, 92%). H NMR (CDCl3, 300 MHz): 7.01 (br s, 2H); 6.78 (d
(8.0 Hz), 2H); 6.32 (d, 7.2 Hz), 1H); 6.05 (d (9.4 Hz), 1H); 5.18
(d (8.1 Hz), 1H); 4.3 to 3.9 (m, 5H); 3.62 (dd (8.9 Hz, 4.3 Hz),
1H); 3.40 (m, 2H); 3.23 (t (9.1 Hz), 1H); 3.03 (m, 1H); 2.89 (dd
(13.2 Hz, 4.3 Hz), 1H); 2.78 (m, 2H); 2.40 (m, 1H); 1.8 to 1.6
(m, 6H); 1.40 (s, 9H); 1.31 (m, 2H); 0.82 (t (8.0 Hz), 3H). 13C
NMR (CDCl3, 75 MHz): 171.2, 169.0, 157.9, 155.4, 130.4,
128.8, 115.8, 73.2, 69.7, 68.2, 66.1, 54.1, 53.2, 51.9, 51.2, 46.2,
36.0, 32.3, 29.1, 28.3, 22.7, 21.2, 10.6.
(11) Wlodawer, A.; Vondrasek, J . Inhibitors of HIV-1 protease:
a
major success of structure-assisted drug design. Annu. Rev.
Biophys. Biomol. Struct. 1998, 27, 249-284.
(12) Lebon, F.; Ledecq, M. Approaches to the design of effective HIV-1
protease inhibitors. Curr. Med. Chem. 2000, 7, 455-477.
(13) Furfine, E. S. The next generation of human immunodeficiency
virus protease uinhibitors: targeting viral resistance. Handb.
Exp. Pharmacol. 2000, 140, 49-72.
(14) Tyndall, J . D. A.; Fairlie, D. P. Conformational Homogeneity in
Molecular Recognition by Proteolytic Enzymes. J . Mol. Recog.
1999, 12, 1-8.
(15) Fairlie, D. P.; Tyndall, J . D. A.; Reid, R. C.; Wong, A. K.;
Abbenante, G.; Scanlon, M. J .; March, D. R.; Bergman, D. A.;
Chai, C. L. L.; Burkett, B. A. Conformational Selection Of
Inhibitors and Substrates By Proteolytic Enzymes: Implications
for Drug Design and Polypeptide Processing. J . Med. Chem.
2000, 43, 1271-1281.
((13S)-{2-[(4-Acetyla m in o-ben zen esu lfon yl)-(3-m eth yl-
bu tyl)a m in o]-(1R)-h yd r oxy-eth yl}-8,11-d ioxo-(10S)-p r o-
p oxym et h yl-2-oxa -9,12-d ia za -b icyclo[13.2.2]n on a d eca -
1(18),15(19),16-tr ien -(7S)-yl)car bam ic Acid ter t-Bu tyl Ester
(9k ). Isoamylamine was condensed with epoxide 8c (9.5 mg,
18 µmol) (procedure J , 90%), followed by addition of 4-acet-
amidobenzenesulfonyl chloride (procedure K, rpHPLC, 36%)
to afford the title compound. 1H NMR (CD3OD, 500 MHz): 7.78
(d (9.6 Hz), 1H); 7.77 (s, 4H); 7.75 (d (8.6 Hz), 1H); 7.06 (br s,
2H); 6.76 (br s, 2H); 4.28 (m, 2H); 4.13 (m, 1H); 4.05 (m, 1H);
3.90 (m, 1H); 3.72 (m, 1H), 3.45 (m, 3H); 3.25 (m, 4H); 3.18
(m, 1H); 3.03 (dd (13.7 Hz, 8.4 Hz), 1H); 2.42 (t (12.1 Hz), 1H);
2.14 (s, 3H); 1.72 (m, 1H); 1.6 to 1.4 (m, 8H); 1.40 (s, 9H); 1.18
(m, 2H); 0.87 (d (6.7 Hz), 6H); 0.76 (t (7.0 Hz), 3H). 13C NMR
(CD3OD, 125 MHz): 173.5, 172.1, 171.1, 158.7, 157.6, 144.4,
135.6, 132.2, 131.8, 129.6, 120.7, 117.4, 74.3, 74.2, 72.3, 72.2,
68.9, 55.7, 54.1, 53.3, 47.8, 38.6, 37.0, 33.7, 30.9, 28.8, 27.3,
24.2, 23.9, 23.4, 23.1, 22.9, 11.0. HRMS (EI) m/z calcd for
(16) Abbenante, G.; March, D.; Bergman, D.; Hunt, P. A.; Garnham,
B.; Dancer, R. J .; Martin, J . L.; Fairlie, D. P. Regioselective
Structural and Functional Mimicry Of Peptides: Design Of
Hydrolytically Stable Cyclic Peptidomimetic Inhibitors Of HIV
1 Protease. J . Am. Chem. Soc. 1995, 117, 10, 220-10, 226.
(17) March, D.; Abbenante, G.; Bergman, D.; Brinkworth, R. I.;
Wickramasinghe, W.; Begun, J .; Martin, J . L.; Fairlie, D. P.
Substrate Based Cyclic Peptidomimetics Of PheIleVal That
Inhibit HIIV-1 Protease Using a Novel Enzyme Binding Mode.
J . Am. Chem. Soc. 1996, 118, 3375-3379.
(18) Abbenante, G.; March, D. R.; Bergman, D. A.; Hunt, P.; Fairlie,
D. P. Structure-Activity Relationships for Macrocyclic Inhibitors
of HIV-1 Protease. Bioorg. Med. Chem. Lett. 1996, 6, 2531-2536.
(19) Reid, R. C.; March, D. R.; Dooley, M. J .; Bergman, D. A.;
Abbenante, G.; Fairlie, D. P. A novel bicyclic enzyme inhibitor
as a consensus peptidomimetic for the receptor-bound conforma-
tions of 12 peptidic inhibitors of HIV-1 protease. J . Am. Chem.
Soc. 1996, 118, 8511-8517.
(20) Martin, J . L.; Begun, J .; Schindeler, A.; Wickramasinghe, W.
A.; Alewood, D.; Alewood, P. F.; Bergman, D. A.; Brinkworth,
R. I.; Abbenante, G.; March, D.; Reid, R. C.; Hunt, P. A.; Fairlie,
D. P. Molecular Recognition of Macrocyclic Peptidomimetic
inhibitors by HIV-1 Protease. Biochemistry 1999, 38, 7978-7988.
(21) Tyndall, J . D. A.; Reid, R. C.; Tyssen, D. P.; J ardine, D. K.; Todd,
B.; Passmore, M.; March, D. R.; Pattenden, L. K.; Bergman, D.
A.; Alewood, D.; Hu, S.-H.; Alewood, P. F.; Birch, C. J .; Martin,
J . L.; Fairlie, D. P. Synthesis, Stability, Antiviral Activity, and
Protease-Bound Structures of Substrate-Mimicking Constrained
Macrocyclic Inhibitors of HIV-1 Protease. J . Med. Chem. 2000,
43, 3495-3504. (b) pdb file: 1D4L.
(22) Tyndall, J . D. A.; Fairlie, D. P. Macrocycles Mimic The Extended
Peptide Conformation Recognized By Aspartic, Serine, Cysteine
and Metallo Proteases. Curr. Med. Chem. 2001, 8, 893-907.
(23) Swain, A. L.; Miller, M. M.; Green, J .; Rich, D. H.; Schneider,
J .; Kent, S. B.; Wlodawer, A. X-ray crystallographic structure
of a complex between a synthetic protease of human immuno-
deficiency virus 1 and a substrate-based hydroxyethylamine
inhibitor. Proc. Natl. Acad. Sci. U.S.A. 1990, 87, 8805-8809.
(24) (a) Leach, A. R. Template Forcing minimises the RMSD using a
specific minimisation algorithm that allows rotation around
single bonds as well as translation and rotation in space.
Molecular Modelling: Principles and Applications; Longman:
C
40H62N5O10S (MH+), 804.4212; found, 804.4189. Analytical
rpHPLC: isocratic 30A:70B rt ) 7.3 min; gradient 100A to
100B 20 min rt ) 23.9 min.
4-Am in o-N-[2-((7S)-a m in o-8,11-d ioxo-(10S)-p r op oxy-
methyl-2-oxa-9,12-diazabicyclo[13.2.2]nonadeca-1(18),15(19),16-
tr ien -13-yl)-(2R)-h yd r oxy-eth yl]-N-(3-m eth yl-bu tyl)ben -
zen esu lfon a m id e (9j). Amide (9k ) (3.0 mg, 3.7 mmol) was
refluxed for 2 h in methanol (2 mL) with 3 drops of 2 M HCl.
After the solution was cooled to room temperature, water (1
mL) was added, and the solution was purified by rpHPLC to
provide the amine as a white powder (1.6 mg, 65%) after
1
lyophilization. H NMR (CD3OD, 500 MHz): 7.50 (d (8.3 Hz),
2H); 7.15 (br s, 2H); 7.05 (br s, 2H); 6.70 (d (8.3 Hz), 2H); 4.39
(m, 1H); 4.30 (m, 1H); 4.11 (m, 2H); 3.85 (m, 1H); 3.72 (m,
3H); 3.52 (d (5.6 Hz), 2H); 3.45 (dd (14.3 Hz, 3.4 Hz), 1H); 3.3
(m, 1H); 3.22 (m, 1H); 3.09 (m, 1H); 2.90 (dd (14.7 Hz, 8.7 Hz),
1H); 2.36 (t (14.4 Hz), 1H); 1.86 (m, 3H); 1.75 (m, 2H); 1.63
(m, 1H); 1.5 to 1.3 (m, 4H); 1.20 (m, 1H);, 0.87 (d (6.8 Hz),
6H); 0.81 (t (7.9 Hz), 3H). HRMS (EI) m/z calcd for C33H52N5O7S
(MH+), 662.3582; found, 662.3563. Analytical rpHPLC: iso-
cratic 60A:40B rt ) 7.8 min; gradient 100A to 100B 20 min rt
) 17.8 min.
Ack n ow led gm en t. We thank the National Health
and Medical Research Council and the Australian
Research Council for partial financial support.

Templates for Protease Inhibitors
J ournal of Medicinal Chemistry, 2002, Vol. 45, No. 2 381
Harlow, England, 1996. (b) InsightII Modeling Envorionment,
Release 2000; Accelrys Inc.: San Diego, 2001.
(31) Toth, M. V.; Marshall, G. R. A simple, continuous fluorometric
assay for HIV protease. Int. J . Pept. Protein Res. 1990, 36, 544-
550.
(32) Kim, E. E.; Baker, C. T.; Dwyer, M. D.; Murcho, M. A.; Rao, B.
G.; Tung, R. D.; Navia, M. A. Crystal structure of HIV-1 protease
in complex with VX478, a potent and orally bioavailable inhibitor
of the enzyme. J . Am. Chem. Soc. 1995, 117, 1181-1182.
(33) Calculated using PALLAS prologD 2.1 from Compudrug Chem-
istry Ltd., Hungary.
(34) Schneider, J .; Kent, S. B. H. Enzymatic activity of a synthetic
99 residue protein corresponding to the putative HIV-1 protease.
Cell 1988, 54, 363-368.
(35) Neate, E. V.; Pringle, R. C.; J owett, J . B. M.; Healey, D. S.; Gust,
I. D. Isolation of HIV from Australian patients with AIDS, AIDS
related conditions and healthy antibody positive individuals.
Aust. N. Z. J . Med. 1987, 17, 461-466.
(25) Condra, J .; Schlief, W.; Blahy, O.; Gabryelski, L.; Graham, D.;
Quintero, J .; Rhodes, A.; Robbins, H.; Roth, E.; Shivaprakash,
M.; Titus, E.; Yang, T.; Teppler, H.; Squires, K.; Deutsch, P.;
Emini, E. In vivo emergence of HIV-1 variants resistant to
multiple protease inhibitors. Nature 1995, 375, 569-571.
(26) Erickson, J . W. The Not So Great Escape. Nat. Struct. Biol. 1995,
2, 523-529.
(27) Palmer, S.; Schafer, R. W.; Merigan, T. C. Highly drug-resistant
HIV-1 clinical isolates are cross-resistant to many antiretroviral
compounds in current clinical development. AIDS 1999, 13, 661-
667.
(28) Adamczyk, M.; J ohnson, D. D.; Reddy, R. E. Collagen Cross-
Links: Synthesis of Pyridinonline, Deoxypyridinoline and their
Analogues. Tetrahedron 1999, 55, 63-88.
(29) For a discussion of stereoisomer preference and binding confor-
mation, see ref 20.
(36) Tachedjian, G.; Tyssen, D.; Locarnini, S.; Gust, I.; Birch, C.
Investigation of topoisomerase inhibitors for activity against
human immunodeficiency virus: inhibition by coumermycin A1.
Antiviral Chem. Chemother. 1990, 1, 131-138.
(30) Bergman, D. A.; Alewood, D.; Alewood, P. F.; Andrews, J . L.;
Brinkworth, R. I.; Englebretsen, D. R.; Kent, S. B. H. Kinetic-
properties of HIV-1 Protease Produced by Total Chemical
Synthesis with Cysteine Residues Replaced by Isosteric L-alpha-
amino-n-butyric Acid Lett. Pept. Sci. 1995, 2, 99-107.
J M010414I