Synthesis of P1′-functionalized macrocyclic transition-state mimicking HIV-1 protease inhibitors encompassing a tertiary alcohol

Article abstract of DOI:10.1021/jm500434q

Seven novel tertiary alcohol containing linear HIV-1 protease inhibitors (PIs), decorated at the para position of the benzyl group in the P1′ side with (hetero)aromatic moieties, were synthesized and biologically evaluated. To study the inhibition and antiviral activity effect of P1-P3 macrocyclization, 14-and 15-membered macrocyclic PIs were prepared by ring-closing metathesis of the corresponding linear PIs. The macrocycles were more active than the linear precursors and compound 10f, with a 2-thiazolyl group in the P1′ position, was the most potent PI of this new series (K_i 2.2 nM, EC ₅₀ 0.2 μM). Co-crystallized complexes of both linear and macrocyclic PIs with the HIV-1 protease enzyme were prepared and analyzed.

Full text of DOI:10.1021/jm500434q

Article
pubs.acs.org/jmc
Synthesis of P1′-Functionalized Macrocyclic Transition-State
Mimicking HIV‑1 Protease Inhibitors Encompassing a Tertiary
Alcohol
Maria De Rosa,† Johan Unge,‡ Hitesh V. Motwani,† Åsa Rosenquist,§ Lotta Vrang,§ Hans Wallberg,§
and Mats Larhed*,∥
†Department of Medicinal Chemistry, Organic Pharmaceutical Chemistry, BMC, Uppsala University, P.O. Box 574, SE-751 23
Uppsala, Sweden
‡MAX IV-Laboratory, Lund University, P.O. Box 118, SE-221 00 Lund, Sweden
§Medivir AB, P.O. Box 1086, SE-141 22 Huddinge, Sweden
∥Department of Medicinal Chemistry, Science for Life Laboratory, BMC, Uppsala University, P.O. Box 574, SE-751 23 Uppsala,
Sweden
S
* Supporting Information
ABSTRACT: Seven novel tertiary alcohol containing linear
HIV-1 protease inhibitors (PIs), decorated at the para position
of the benzyl group in the P1′ side with (hetero)aromatic
moieties, were synthesized and biologically evaluated. To study
the inhibition and antiviral activity effect of P1−P3 macro-
cyclization, 14- and 15-membered macrocyclic PIs were
prepared by ring-closing metathesis of the corresponding linear
PIs. The macrocycles were more active than the linear
precursors and compound 10f, with a 2-thiazolyl group in the
P1′ position, was the most potent PI of this new series (Ki 2.2 nM, EC50 0.2 μM). Co-crystallized complexes of both linear and
macrocyclic PIs with the HIV-1 protease enzyme were prepared and analyzed.
improved affinity for protein targets by reducing the entropy
loss upon binding.11 Furthermore, cyclizations may provide
better resistance to proteolytic digestion and facilitate internal
hydrogen bonding, thus helping to improve bioavailability and
enable the compound to reach its intracellular targets.12
Recently, we have applied the macrocyclization strategy for
preparing a new series13 of HIV-1 PIs related to atazanavir14,15
but containing a tertiary alcohol as part of the transition-state
mimicking core structure (Figure 1).13,16
Cyclic PIs with bromine in the para-position of the P1′ were
up to 15 times more potent in the cell assay as compared to the
linear precursors, exhibiting EC50 values down to 370 nM.13
Although the best macrocycles had lower EC50 values than the
corresponding linear PIs, the differences at the enzyme level
(Ki) were not impressive. X-ray crystallography data suggested
that the P1−P3 sites of both linear and macrocylic PIs
accommodate the lipophilic pockets of the enzyme in the same
fashion.13 Therefore, we decided to explore deeply the
structure−activity-relationships (SARs) of the linear mole-
cule−enzyme interactions to optimize the PIs in the P1′
position and to investigate the macrocyclization strategy further
to obtain new potential anti-HIV compounds with both
INTRODUCTION
■
Antiretroviral therapy (ART) has markedly changed the
outcome for HIV-infected patients.1 Gratifyingly, there were
more than 700000 fewer new HIV infections globally in 2011
than in 2001. However, the emergence of resistance strains
continues to impact the efficacy of current antiretroviral
regimens.2 The road from 2.5 million new HIV infections in
2011 to nearly zero new HIV infections is still very long, and a
significant effort is required to further develop the existing HIV
prevention programmes.3
HIV-1 protease inhibitors (PIs) are essential components in
the therapeutic treatment against HIV, but their use is
somewhat limited by severe side effects such as dyslipidaemia
and hypersensitivity.4,5 Peptide-derived drugs are also known to
suffer from metabolic instability6 and poor membrane
permeation,7 which mainly arise due to the presence of peptide
bonds, high conformational flexibility and the exposure of polar
functionalities in linear molecules. Macrocyclization of linear
peptides represents an important approach to overcome this
problem.8
There has been large interest in both naturally occurring and
synthetic cyclic peptides as scaffolds in drug discovery
efforts.9,10 A macrocyclic peptide provides diverse functionality,
increased structural rigidity, and large stereochemical complex-
ity in a preorganized ring structure.8 Preorganization will favor
Received: March 19, 2014
Published: July 23, 2014
© 2014 American Chemical Society
6444
dx.doi.org/10.1021/jm500434q | J. Med. Chem. 2014, 57, 6444−6457

Journal of Medicinal Chemistry
Article
Figure 1. (a) General structure of reported linear PIs encompassing a tertiary alcohol,16 (b) general structure of reported macrocyclic PIs,13 (c)
general structure of the new series of linear and macrocyclic PIs functionalized in the P1′ position, (d) atazanavir.
improved potency (EC50, Ki) and pharmacokinetic profile. We
herein report the synthesis, biological evaluation, and
crystallography data of two new series, linear and 14−15-
membered P1−P3 macrocyclic PIs, encompassing a tertiary
alcohol in the transition-state mimicking scaffold and various
(hetero)aromatic substituents in the para-position of the P1′
side chain (Figure 1c).
palladacycle (as a palladium precatalyst) together with
potassium carbonate as base and the preligand tri(tert-
butyl)phosphonium tetrafluoroborate in dimethoxyethane/
water21−23 to obtain 8a−b in moderate to high isolated yields
(54−78%). Unfortunately, this protocol was not successful
when phenyl-, thienyl-, and thiazolylboronic acids were used.
Hence, an alternative protocol using palladium(II) acetate as
precatalyst, 2 M aqueous solution of sodium carbonate as base,
and triphenylphosphine as ligand in dimethoxyethane/ethanol
and irradiating the reaction vial in a microwave reactor was
chosen.24 The desired products (8c−e) were obtained in high
isolated yields (70−85%).
RESULTS
■
Chemistry. Compound 1 was synthesized as previously
reported13 and used as the starting material for the preparation
of the new PIs. The lactone moiety of compound 1 was
reduced using lithium borohydride, and the newly formed
primary alcohol was selectively protected as a pivalate ester to
give 2 in good yield of 70% over two steps. The tertiary alcohol
was thereafter protected using tert-butyldimethylsilyl triflate
affording intermediate 3. The bromine in 4-position of the aryl
bromide 3 was substituted with an allyl group using a Stille
coupling reaction17,18 to provide compound 4 in high yield of
75% (Scheme 1). Reaction of ester 4 with lithium borohydride
yielded intermediate primary alcohol 5 (85%), which was
oxidized using Dess−Martin reagent to the corresponding
aldehyde. This was followed by a reductive amination with the
hydrazide 619 using a catalytic amount of acetic acid to activate
the aldehyde and sodium triacetoxyborohydride as the
reductive agent in a one-pot procedure to obtain 7 (44%
over three steps). Compound 7 was thereafter used as
precursor for the preparation of the linear series of new PIs
diversly substituted at the P1′ side chain. The 4-position of the
benzyl moiety was decorated with phenyl-, pyridyl-, thienyl-,
and thiazolyl groups.
The functionalization of precursor 7 via Suzuki cross-
coupling reaction with 2-(pyridyl/thiazolyl)boronic acids was
not successful, possibly due to rapid protodeboronation,
oxidation, and/or polymerization.25−27 In an effort to combat
this problem, several attempts were made using the
corresponding N-methyliminodiacetic (MIDA) esters28 of the
above-mentioned boronic acids. The reactions were performed
using temperatures in the range of 85−130 °C and investigating
different amount of base for the in situ hydrolysis of the MIDA
esters but without any positive outcome. In all attempts, an
isocyanate byproduct was generated by a β-elimination reaction
at the carbamate function in P2′ position, according to LC-MS
29
analysis and H NMR. Stille coupling reactions16b were also
tried as a possible alternative using 2-(tributylstannyl)pyridine,
but unfortunately only the starting material 7 was recovered.
Thus, to introduce the 2-pyridyl and 2-thiazolyl moieties in P1′
position, a different approach was chosen (Scheme 3a).
Commercially available benzaldehydes 15 and 16 were reacted
with hydrazide 17 to obtain the hydrazides 18a−b.30
Subsequent reaction of hydrazides 18a−b with the primary
alcohol 5 via a reductive amination process afforded the desired
tert-butyldimethylsilyl (TBS)-protected compounds 8f−g.
1
The 4- and 3-pyridyl moieties were introduced via microwave
assisted20 Suzuki−Miyaura cross-coupling reactions (Scheme
2a) using the corresponding boronic acids, Herrmann’s
6445
dx.doi.org/10.1021/jm500434q | J. Med. Chem. 2014, 57, 6444−6457

Journal of Medicinal Chemistry
Article
a
Scheme 1. Synthesis of Precursor 7
a
Reagents and conditions: (a) (i) 2 M LiBH4 in THF, rt, 15 h, (ii) pivaloyl chloride, anhydrous pyridine, anhydrous DCM, rt, 17 h, 70% (over two
steps); (b) TBDMS triflate, TEA, anhydrous DCM, rt, 16 h, 92%; (c) Pd(PPh3)2Cl2, CuO, allyltributylstannane, anhydrous DMF, MW, 150 °C, 30
min, 75%; (d) 2 M LiBH4 in THF, 0 °C to rt, 16 h, 85%; (e) (i) Dess−Martin periodinane, anhydrous DCM, rt, 2 h, (ii) cat. AcOH, 6, DCE, 4 h, rt,
(iii) NaBH(OAc)3, rt, 18 h, 44% (over three steps).
Removal of the TBS group in compounds 8a−g (Schemes
2b and 3c), using 1 M tetrabutylammoniumfluoride solution in
THF at room temperature, furnished the desired linear PIs 9a−
g in high yields (55−95%).
contraction (13g) products were obtained. These results lead
us to believe that the nature of the substituent at P1′ somehow
influences the outcome of the metathesis toward different
degrees of ring closure, double-bond migration, and ring
contraction. The position of the double bond in the ring and its
Z configuration were assigned by NMR data and X-ray data.
Biological Evaluation: HIV Protease Inhibition, MT4
Cell-Based Anti-HIV Activity, Permeability, and Meta-
bolic Stability in Vitro. HIV-1 protease inhibition and cell-
based antiviral activity of the newly synthesized linear and
macrocyclic PIs were evaluated (Tables 1 and 2, respectively),
and the corresponding Ki and EC50 values are presented.
All PIs displayed good inhibition profile having Ki in the low
nanomolar range (2.2 nM ≤ Ki ≤ 120 nM) and EC50 down to
0.13 μM. We hypothesized that the introduction of varying
moieties at the P1′ position, with different polarity and
hydrogen bonding potential, could improve the protease−
inhibitor interaction in the lipophilic S1/S1′ enzyme region and
thus enhance the inhibition potency of the respective PIs.16c
Interestingly, out of the seven linear PIs (Table 1), four were
equally or more active than the corresponding bromide
analogue A (7.8 nM ≤ Ki ≤ 18 nM vs Ki 30 nM for bromide
analogue A), meaning that the P1′ functionalization was well
tolerated and fitted nicely into the S1′ hydrophobic pocket.
Gratifyingly, an improvement was also observed in anti-HIV
Macrocyclization of linear PIs (9a−e, Scheme 2c; 9f−g,
Scheme 3d) was thereafter performed via a microwave-induced
ring-closing metathesis (RCM) reaction using second-gener-
ation Hoveyda−Grubbs catalyst (20 mmol %) in anhydrous
dicholoromethane at 130 °C.31,32 Lowering the temperature
(60−110 °C) and the catalyst loading (5−10 mmol %) resulted
in no or only partial conversion of starting materials 9a−g,
indicating the need for higher temperature and amount of
catalyst. In some cases, double-bond migration and ring
contraction reactions also occurred.33−35 Figure 2 depicts the
outcome of RCM for all linear PIs (9a−g). Macrocyclization of
compounds 9a−b bearing a 4- and 3-pyridyl moiety at P1′
provided the cyclic 15-membered products 10a−b with high
selectivity in good yields (57% and 54%, respectively) after 30
min of irradiation at 130 °C. For the linear compound 9c with a
phenyl group at P1′, three products were isolated under
identical reaction conditions, namely a ring closed product 10c,
a double-bond migrated cyclic structure 11c, and a contracted
14-membered cyclic product 13c. Macrocyclization of 9g was
the only reaction that did not give the expected 15-membered
ring product, instead double-bond migration (11g) and ring
6446
dx.doi.org/10.1021/jm500434q | J. Med. Chem. 2014, 57, 6444−6457

Journal of Medicinal Chemistry
Article
9a with a 4-pyridyl moiety in P1′ position was the best linear PI
of this new series, having a Ki value of 7.8 nM (Table 1) and
EC50 value of 0.81 μM. Introduction of a phenyl group (9c) or
a 3-thiophene group (9d) was not as beneficial because less
potent PIs were obtained (Ki 65 and 64 nM, respectively)
although with slightly better antiviral activities (3.7 μM for both
PIs vs 6.0 μM for A). Activity of PI 9e carrying a 2-thiophene
substituent on the P1′ side chain seemed to be quite similar to
the bromide analogue (Ki 35 nM vs 30 nM and EC50 2.4 μM vs
6.0 μM, respectively, of 9e vs A). The thiazole structure (9f, Ki
8.0 nM) was almost as potent as 9a in the enzyme assay but
with a somewhat poor antiviral activity (EC50 2.2 μM).
Scheme 2. Synthesis of Linear PIs 9a−e and Macrocyclic PIs
a
10a−e, 11c−d, 12e, and 13c
The P1−P3 macrocyclization approach furnished in almost
all cases improved PIs in terms of inhibition potency and
antiviral activity (Table 2), providing Ki values in the range of
2.2−31.5 nM and EC50 in the range of 0.13−3.05 μM.
Macrocycle 12e was the only exception; it was the least potent
PI in this series with (Ki 120 nM and EC50 7.35 μM). Table 2
also includes corresponding values of the bromide precursor
B13 and FDA approved atazanavir for comparison. Because our
previous results showed that 14-membered macrocyclic PIs are
less potent,13 15-membered macrocyclic rings were synthesized
in the present study by using the allyl group at 4-position of the
P1 nonprime side. In addition to the desired 15-membered
cyclic compounds, double-bond migrated isomeric structures
and ring-contracted 14-membered compounds were also
formed during the metathesis reaction (Figure 2 and Table
2).33−35 As expected, 14-membered ring macrocycles were in all
cases less potent than the corresponding 15-membered cyclic
PIs (comparing 13c to 10c/11c, 14f to 10f/11f, and 13g to
11g). Double-bond migration from 2,3- to the 1,2- position of
the 15-membered ring macrocycles did not strongly affect the
inhibition profile, as PIs with similar Ki values were observed
(comparing 10c to 11c, 10d to 11d, and 10f to 11f). It might,
though, be worth noting that when double-bond migration
occurred to the 3,4-position of the 15-membered ring (12e, Ki
120 nM), the inhibiting potency dramatically dropped nearly 7-
fold (10e vs 12e). Molecules 10f and 11g, with 2-thiazolyl and
2-pyridyl groups, respectively, were highly potent PIs of the
macrocyclic series and slightly more active than the bromide
precursor B; 10f was the best PI (Ki 2.2 nM), with very good
antiretroviral activity (EC50 0.2 μM). A few PIs were also
evaluated in selected in vitro resistant HIV-1 isolates and the
EC50 values are shown in the Tables 1 and 2. Saquinavir
resistant virus containing A71V, I84V and L90M mutations and
ritonavir resistant virus containing M46I, V82F, and I84V
mutations were isolated and used for cross resistance studies
(cf. Supporting Information). Antiviral activity of atazanavir
against the same mutants is also shown for comparison. The
tested linear PIs were moderately active against mutations
A71V, I84V, L90M, and inactive against mutations M46I,
V82F, and I84V. The tested cyclic PIs exhibited equipotent
activity in these mutant MT4/HIV-1 protease assays with a
better inhibition profile against A71V, I84V, and L90M
mutations. In general, macrocyclization improved inhibition
a
Reagents and conditions: (a) Herrmann’s palladacycle, K2CO3, aryl
boronic acid, [HP(tBu)3]BF4, DME/water, MW, 140 °C, 30 min, 54−
78% or Pd(OAc)2, PPh3, aryl boronic acid, 2 M Na2CO3, DME/
EtOH, MW, 130 °C, 20 min/1 h, 70−85%; (b) 1 M TBAF in THF, 0
°C to rt, 2 h, 70−95%; (c) second-generation Hoveyda−Grubbs
catalyst, anhydrous DCM, MW, 130 °C, 1 h, 13−57%.
Scheme 3. Synthesis of Linear PIs 9f−g and Macrocyclic PIs
10f, 11f−g, 13ge, and 14f
a
a
Reagents and conditions: (a) (i) ethanol, reflux, 5 h, 70−80%, (ii)
potency on mutated HIV-1 protease. The cytotoxicity (CC50)
H2/Pd−C, methanol, 48 h, 55−60%; (b) (i) Dess−Martin period-
inane, anhydrous DCM, rt, 2 h, (ii) cat. AcOH, 5, DCE, 4 h, rt, (iii)
NaBH(OAc)3, rt, 18 h, 44−49% (over three steps); (c) 1 M TBAF in
THF, 0 °C to rt, 2 h, 55−65%; (d) second-generation Hoveyda−
Grubbs catalyst, anhydrous DCM, MW, 130 °C, 1 h, 15−30%.
36
and Caco-2 permeation (Papp
)
and metabolic stability (Clint)
were also evaluated (cf. Supporting Information), and the
results are summarized in Tables 1 and 2. Rewardingly, all new
21 PIs showed lower MT4 cell toxicity than atazanavir. All
three Caco-2 tested linear PIs (9a, 9f, and 9g, Table 1) showed
high permeation (Papp of 22−47 × 10−6 cm/s). When the linear
PIs were cyclized, unfortunately, somehow the permeation
cellular activity, as all new linear PIs exhibited EC50 lower (EC50
≤ 3.7 μM) than the bromide analogue A (EC50 6.0 μM).13 PI
6447
dx.doi.org/10.1021/jm500434q | J. Med. Chem. 2014, 57, 6444−6457

Journal of Medicinal Chemistry
Article
Figure 2. Outcome of ring closing metathesis reaction.
decreased (Table 2). Nevertheless, all measured cyclic PIs
showed single- or double-digit Papp values similar to or higher
than those of atazanavir, except for the 4-pyridine derivative
10a. Further, all four PIs tested for the metabolic stability
showed very good clearance. As expected, the macrocyclic PI
10a (Clint > 300 μg/min/mg) was more stable than the
corresponding linear PI 9a (Clint 73 μg/min/mg). Both linear
(9f) and macrocyclic (10f) PIs, having a thiazole moiety at the
P1′ position, showed similar metabolic stability (Clint > 300 μg/
min/mg). In general, the macrocycles showed a better intrinsic
clearance when compared to atazanavir (Clint 90 μg/min/mg),
indicating that the macrocyclization approach can improve the
metabolic stability of peptide-like molecules.
In summary, macrocyclization resulted in a consistent
improvement of antiretroviral activity, in several cases similar
to atazanavir. Further, the cytotoxicity property for most
macrocyclic PIs was better than for the linear PIs, while the
permeability was reduced from high/very high to moderate/
high; however, the metabolic stability was very good.
complexes, had a limited density map. Nonetheless, the
electron density maps suggested a good fit to the data and
the orientation could be easily sorted out for all complexes. In
our previously published work, the linear PIs carrying a vinylic
or allylic P1 substituent failed to give good crystallography
data.13 With the set of nine structures presented in this study, a
detailed comparison of linear and macrocyclic scaffolds was
carried out. Current data set has helped to analyze the global
effect of the P1−P3 macrocyclization approach on the binding
profile of the PIs. The two different scaffolds (linear and
macrocyclic) bind similarly to the enzyme, with only minor
differences (Figure 3). The main effect of the macrocyclization
seems to be that the introduced constraints make the
macrocycles slightly smaller than the corresponding linear
PIs, which allowed small adjustments and a better fit of the
macrocycles in the P1−P3 pockets. Interestingly, this better
accommodation of the macrocyclic structures at the nonprime
site of the enzyme affects the hydrogen bonding distances
between the tertiary hydroxyl group of the PIs and the aspartic
residues Asp25/Asp125 and Gly27 of the active site of the
enzyme (Figure 4). The strengths of these bonds are of
particular interest because they mimic the bonds formed during
the catalytic reaction.
X-ray Structure Analysis. X-ray crystallography data were
obtained for nine new HIV-1 PI−enzyme complexes. The
resolution for all complex structures were in the range of 1.5−
2.45 Å. Refinement were straightforward for all structures and
R
free around 20% for most structures with a few exceptions (cf.
All macrocyclic PIs exhibited hydrogen bonds shorter than
3.0 Å (strong interaction) between the tertiary hydroxyl group
and the carboxylate oxygen of Asp125. The distances between
the tertiary hydroxyl group and the carbonyl oxygen Gly27
were found to be longer than 3.0 Å, indicating a less favorable
Supporting Information). The density was continuous in most
areas of the HIV-1 PIs, which made the modeling smooth.
Exceptions were observed at the P1′ moiety, where the atoms
of the heterosubstituted outermost ring, in most of the
6448
dx.doi.org/10.1021/jm500434q | J. Med. Chem. 2014, 57, 6444−6457

Journal of Medicinal Chemistry
Article
Figure 3. Comparison of the noncyclic PI 9e (PDB ID 4cpr, turquoise) and the corresponding macrocycle 10e (PDB ID 4cpx, green) complexes.
The macrocyclic compound mimicked the binding of the linear PI very well, differing significantly only at the point of macrocyclization at the S3 site
and at the thiophene substituent at the S1′ site, the latter having partial rotational freedom. The macrocycle in nonprime side has a somewhat smaller
ring structure which allows small shifts of the arrangement between several macrocyclic PIs and their linear counterparts.
bonding. The PIs 10b, 10d, and 10e (PDB IDs 4cpt, 4cpu, and
4cpx, respectively) had the most favored hydrogen bond
lengths, 2.7 Å, to the catalytic aspartic residues Asp125. The
bond lengths for the linear scaffold on the other hand are
shifted. In the 9e complex, the bonds to Asp125 are all greater
than 3.0 Å and less than 2.7 Å for the bond to the Gly27
carbonyl oxygen. The two exceptions are 9b (PDB ID 4cpq)
and 10a (PDB ID 4cps), both having intermediate distances to
the Asp125 carboxylate group (2.9 and 3.1 Å, respectively, cf.
Supporting Information). The adjustment of the macrocyclic
scaffold in the P1−P3 region is also reflected in the rotation of
the isopropyl group at the S2 pocket (Figure 4), which includes
the amino acid residues Ala28, Asp30, and Val184. In the linear
PIs, the isopropyl group has a clear (9a, PDB ID 4cp7) or
possible (9b, PDB ID 4cpq and 9e, PDB ID 4cpr) orientation
where the two methyl groups are situated closer to the Ala28
side chain. For the P1−P3 macrocyclic PIs, the isopropyl group
appears to be forced to point away from the nitrogen (10a,
PDB ID 4cps; 10d, PDB 4cpu; and 10e, PDB ID 4cpx), as the
isopropyl group is now closer to the backbone of the protein
(cf. Supporting Information).
The terminal allylic groups in P1 and P3 position of the
linear PIs do not seem to interact closely with the protein
(Figure 3). Instead, the P3-allyl group is directed away from the
protein surface and is assumed to be flexible. The same is true
for the allylic extension in para position on the phenyl alanine
mimicking group at P1 site, which has little or no defined
density for those atoms. As for previously examined groups in
P1′ position with this scaffold, the rather large thiophene side
chain is directed toward the surface of the protein. In almost all
complexes examined so far, the conserved aromatic structure of
the P1′ side chain has a well-defined density, whereas the para
substituted (hetero)aromatic substituent was only partially
defined.
6449
dx.doi.org/10.1021/jm500434q | J. Med. Chem. 2014, 57, 6444−6457

Journal of Medicinal Chemistry
Article
Figure 4. A close-up of the hydrogen bonding distances between the tertiary hydroxyl alcohol of the linear PI 9e (top, pink, PDB ID 4cpr) and the
macrocyclic PI 10e (bottom, green, PDB ID 4cpx) and the catalytic residues Asp25/125, as well as the carbonyl oxygen of Gly27. The macrocyclic PI
10e had hydrogen bond distances between the tertiary hydroxyl group and Asp125/25 of 2.7 and 3.4 Å, respectively. A contrasting scene is observed
for the linear compounds 9e, where the bonds to Asp125/25 are longer than 3.4 Å (not denoted in figure) and the bond to Gly27 is only 2.5 Å. The
distance between the tertiary hydroxyl alcohol of the cyclic PI 10e and the carbonyl oxygen of Gly27 is instead 3.4 Å.
6450
dx.doi.org/10.1021/jm500434q | J. Med. Chem. 2014, 57, 6444−6457

Journal of Medicinal Chemistry
Article
Figure 5. Structure of the complex protease enzyme/10d PI (blue, PDB ID 4cpu) is superimposed on the 11d complex (yellow, PDB ID 4cpw).
Crystal structures of macrocycles having the same ring size
(15 members) but with different double-bond position in the
ring as a result of the double-bond migration during the RCM
reaction, as mentioned in the Chemistry section, were also
obtained. The restriction of the angles from the double bond
causes this region to adapt different conformations. The
electron density for PIs 10d (PDB ID 4cpu) and 11d (PDB
ID 4cpw, Figure 5) could be confidently interpreted and shows
that the region of the ring closure in PI 10d is directed away
from Arg108. In PI 11d, on the other hand, it is pointing
toward Arg108, causing the side chain to shift its position. For
large size images and overview of all X-ray structures obtained,
see Supporting Information.
(Ki 30 nM and EC50 6.0 μM), with 9a (Ki 7.8 nM, EC50 0.81
μM) as the most potent linear PI. Fourteen macrocyclic PIs
were obtained by a P1−P3 RCM reaction of the new linear
HIV-1 PIs. Macrocyclic PIs exhibited EC50 values down to 0.13
μM and Ki in the low nanomolar range (2.2 nM ≤ Ki ≤ 120
nM) combined with low cell toxicity and sufficient Caco-2
permeation. Overall, macrocyclization furnished a consistent
improvement of antiretroviral activity, in many cases better
than the precursor bromide B, with 10f (Ki 2.2 nM and EC50
0.2 μM) as the most potent and promising macrocyclic PI of
this new series.
EXPERIMENTAL SECTION
■
1
General Methods. H and 13C NMR spectra were recorded on
1
Varian Mercury Plus instruments: H at 399.9 MHz and 13C at 100.6
CONCLUSION
■
MHz at 25 °C. Analytical HPLC-MS was performed on a Gilson
HPLC system with a Finnigan AQA quadropole low-resolution mass
spectrometer in positive or negative ESI mode using a Onyx
Monolithic C18 4.6 mm × 50 mm 5 μm (Phenomenex) or ACT C4
4.6 mm × 50 mm, 5 μm column with MeCN in 0.05% aqueous
HCOOH as mobile phase at a flow rate of 4 mL/min. Preparative
HPLC-MS was performed on a Gilson HPLC equipped with a Zorbax
SB-C8 (21.2 mm × 150 mm, Agilent Technologies) column or a 10
μm Vydac C18 column (250 mm × 22 mm), in both cases with UV
In summary, 21 novel HIV-1 PIs have been synthesized. A set
of seven linear new PIs encompassing a tertiary alcohol as part
of the transition-state mimicking scaffold were obtained by
decoration of the P1′ site with (hetero)aromatic moieties. The
P1′ functionalization was in general beneficial as very good
linear PIs were obtained. Almost all of them showed improved
inhibition potency (Ki ≤ 35 nM) and better antiviral activity
(EC50 ≤ 3.7 μM) compared to the precursor phenyl bromide A
6451
dx.doi.org/10.1021/jm500434q | J. Med. Chem. 2014, 57, 6444−6457

Journal of Medicinal Chemistry
Article
are reported in deg/dm, and the concentration (c) is given in g/100
mL in the specified solvent. The microwave reactions were performed
in a Biotage Initiator producing controlled irradiation at 2450 MHz
with a power of 0−300 W. Reaction temperatures were determined
using the built-in online IR-sensor.
Table 1. Enzyme Inhibition Data, Antiviral Activity, MT4
Cell Toxicity, Caco-2 Permeation, and Metabolic Stability of
Linear PIs 9a−g
1
All PIs were >95% pure according to H NMR, LC-UV (254 nm
and 214 or 220 nm), and LC-MS (TIC).
General Procedure of Ring Closing Metathesis for the
Synthesis of Macrocyclic PIs 10a−f, 11c−d, 11f−g, 12e, 13c,
13g, and 14f and Spectroscopic Data. In a 2−5 mL process vial,
to a solution of linear PI 9a−g (1.0 equiv) in anhydrous DCM (3 mL),
second-generation Hoveyda−Grubbs catalyst (20 mmol %) was added
under nitrogen flow. The vial was sealed then irradiated in the
microwave reactor cavity at 130 °C for 1 h. After cooling to room
temperature, the reaction mixture was filtered off through a plug of
Celite, which was rinsed with ethyl acetate. The filtrate was evaporated
under reduced pressure, giving the crude product, which was purified
by preparative HPLC.
Methyl ((S)-1-(2-(3-((3R,6S,Z)-3-Hydroxy-6-isopropyl-4,7-
dioxo-5,8-diaza-1(1,4)-benzenacyclododecaphan-10-en-3-yl)-
propyl)-2-(4-(pyridin-4-yl)benzyl)hydrazinyl)-3,3-dimethyl-1-
oxobutan-2-yl)carbamate (10a). PI 10a was prepared according to
procedure E from 9a (14.5 mg, 0.0192 mmol) and second-generation
Hoveyda−Grubbs catalyst (2.4 mg, 0.0038 mmol) in anhydrous DCM
(3 mL). The crude product was purified by preparative HPLC, to give
8 mg of 10a as white solid, in 57% isolated yield. 1H NMR (400 MHz,
CD3OD): δ 8.59−8.57 (m, 2H), 7.71−7.67 (m, 4H), 7.56 (d, J = 8.3
Hz, 2H), 7.31 (dd, J = 7.9, 1.9 Hz, 1H), 7.11 (dd, J = 7.9, 1.9 Hz, 1H),
7.04 (dd, J = 7.9, 1.9 Hz, 1H), 6.88 (dd, J = 7.8, 1.9 Hz, 1H), 6.32 (dt,
J = 10.3, 7.4 Hz, 1H) 5.69 (dt, J = 10.3, 7.2 Hz, 1H), 3.96 (s, 2H),
3.88−3.84 (m, 1H), 3.75−3.71 (m, 2H), 3.54 (s, 3H), 3.42 (t, J = 6.4
Hz, 1H), 2.87−2.82 (m, 4H), 2.10−1.98 (m, 2H), 1.95−1.83 (m, 3H),
1.80−1.68 (m, 2H), 0.80 (s, 9H), 0.78 (d, J = 6.7 Hz, 3H), 0.75 (d, J =
6.7 Hz, 3H). 13C NMR (100 MHz, CD3OD): δ 176.03, 172.11,
171.89, 170.55, 150.56, 142.35, 137.85, 136.24, 135.34, 131.46, 131.41,
131.10, 129.33, 129.24, 128.15, 127.84, 123.06, 80.43, 62.43, 59.39,
58.76, 52.69, 38.12, 37.47, 34.94, 33.07, 32.83, 32.14, 30.77, 30.46,
26.91, 23.73, 19.89, 18.44. MS (ESI): m/z 727 [M + H]+ (100).
+
HRMS m/z 727.4186 [(M + H)+ calcd for C41H55N5O6 727.4183].
Methyl ((S)-1-(2-(3-((3R,6S,Z)-3-Hydroxy-6-isopropyl-4,7-
dioxo-5,8-diaza-1(1,4)-benzenacyclododecaphan-10-en-3-yl)-
propyl)-2-(4-(pyridin-3-yl)benzyl)hydrazinyl)-3,3-dimethyl-1-
oxobutan-2-yl)carbamate (10b). PI 10b was prepared according to
procedure E from 9b (14 mg, 0.0185 mmol) and second-generation
Hoveyda−Grubbs catalyst (2.3 mg, 0.0037 mmol) in anhydrous DCM
(2 mL). The crude product was purified by preparative HPLC to give
1
7.3 mg of 10b as white solid in 54% isolated yield. H NMR (400
MHz, CD3OD): δ 8.78−8.76 (m, 1H), 8.09−8.07 (m, 1H), 7.59−7.51
(m, 6H), 7.32 (dd, J = 7.9, 1.9 Hz, 1H), 7.12 (dd, J = 7.8, 1.9 Hz, 1H),
7.04 (dd, J = 7.9, 1.9 Hz, 1H), 6.88 (dd, J = 7.8, 1.9 Hz, 1H), 6.32 (dt,
J = 10.4, 7.3 Hz, 1H) 5.74−5.67 (m, 1H), 3.96 (s, 2H), 3.87 (dd, J =
13.3, 9.6 Hz, 1H), 3.74−3.72 (m, 2H), 3.54 (s, 3H), 3.44−3.40 (m,
1H), 2.87−2.83 (m, 4H), 2.05−2.03 (m, 2H), 1.91−1.86 (m, 3H),
1.75−1.74 (m, 2H), 0.81 (s, 9H), 0.79 (d, J = 6.8 Hz, 3H), 0.75 (d, J =
6.7 Hz, 3H). 13C NMR (100 MHz, CD3OD): δ 174.28, 171.80,
170.22, 169.98, 151.22, 140.72, 135.81, 134.62, 134.54, 133.54, 129.78,
129.60, 129.39, 127.64, 127.57, 126.38, 126.10, 105.55, 78.58, 57.90,
57.28, 56.67, 54.22, 36.28, 35.64, 33.23, 31.21, 29.74, 28.87, 28.61,
25.25, 21.90, 20.67, 18.19, 16.27. MS (ESI): m/z 727 [M + H]+ (100).
a
Bromide analogue of the linear PIs (PI 16a in an earlier paper).13
b
c
Atazanavir was used as reference compound.37 Selected linear PIs
d
+
HRMS m/z 727.4175 [(M + H)+ calcd for C41H55N6O6 727.4183].
were tested for Caco-2 cell permeation. Selected linear PIs were
tested for the metabolic stability. Mutations in MT4/HIV-1 protease:
A71V, I84V, L90M. Mutations in MT4/HIV-1 protease: M46I, V82F,
I84V.
e
Methyl ((S)-1-(2-([1,1′-Biphenyl]-4-ylmethyl)-2-(3-((3R,6S,Z)-
3-hydroxy-6-isopropyl-4,7-dioxo-5,8-diaza-1(1,4)-benzenacy-
clododecaphan-10-en-3-yl)propyl)hydrazinyl)-3,3-dimethyl-1-
oxobutan-2-yl)carbamate (10c). PI 10c was prepared according to
procedure E from 9c (29 mg, 0.0385 mmol) and second-generation
Hoveyda−Grubbs catalyst (4.8 mg, 0.0077 mmol) in anhydrous DCM
(3.5 mL). The crude product was purified by preparative HPLC to
give 3.6 mg (13%, isolated yield) of 10c as white solid, along with PIs
11c (10.0 mg, 36%, white solid) and 13c (2.9 mg, 10.5%, white solid),
respectively.
f
detection at 220 nm, using MeCN in 0.05% aqueous HCOOH as
mobile phase at a flow rate of 5 mL/min or MeCN in 0.1% aqueous
TFA as mobile phase at a flow rate of 5 mL/min. Exact molecular
masses were determined on Micromass Q-Tof2 mass spectrometer
equipped with an electrospray ion source. Optical rotations were
obtained on a PerkinElmer 241 polarimeter, specific rotations ([α]D)
6452
dx.doi.org/10.1021/jm500434q | J. Med. Chem. 2014, 57, 6444−6457

Journal of Medicinal Chemistry
Article
Table 2. Enzyme Inhibition Data, Antiviral Activity, MT4 Cell Toxicity, Caco-2 Permeability, and Metabolic Stability of
Macrocyclic PIs 10a−f, 11c−d, 11f−g, 12e, 13c, 13g, and 14f
6453
dx.doi.org/10.1021/jm500434q | J. Med. Chem. 2014, 57, 6444−6457

Journal of Medicinal Chemistry
Article
Table 2. continued
a
b
c
Bromide analogue of the macrocyclic PIs (PI (R)-19a in an earlier paper).13 Atazanavir was used as reference compound.37 Selected macrocyclic
d
e
PIs were tested for Caco-2 cell permeation. Selected macrocyclic PIs were tested for the metabolic stability. Mutations in MT4/HIV-1 protease:
A71V, I84V, L90M. Mutations in MT4/HIV-1 protease: M46I, V82F, I84V.
f
10c: 1H NMR (400 MHz, CD3OD): δ 7.59−7.30 (m, 10H), 7.13−
7.11 (m, 1H), 7.04−7.03 (m, 1H), 6.88 (dd, J = 7.8, 1.9 Hz, 1H),
6.35−6.26 (m, 1H), 5.74−5.57 (m, 1H), 3.93 (s, 2H), 3.75−3.73 (m,
2H), 3.54 (s, 3H), 3.43−3.40 (m, 2H), 2.85−2.83 (m, 4H), 2.07−2.01
(m, 2H), 1.94−1.85 (m, 3H), 1.77−1.71 (m, 1H), 1.54−1.51 (m, 1H),
0.81 (s, 9H), 0.79 (d, J = 6.7 Hz, 3H), 0.75 (d, J = 6.7 Hz, 3H). 13C
NMR (100 MHz, CD3OD): δ 176.59, 172.24, 171.87, 166.38, 158.92,
142.37, 142.27, 141.66, 137.36, 136.26, 135.36, 131.42, 131.12, 129.81,
129.32, 129.25, 128.26, 127.90, 127.75, 87.11, 80.43, 63.07, 62.56,
58.61, 52.69, 38.13, 37.60, 34.96, 33.06, 32.83, 32.12, 30.74, 26.92,
23.72, 22.57, 19.89, 18.41. MS (ESI): m/z 726 [M + H]+ (100).
13.1 Hz, 1H), 2.89−2.80 (m, 3H), 2.35−2.31 (m, 1H), 2.05−2.00 (m,
2H), 1.88−1.85 (m, 2H), 0.81 (s, 9H), 0.77 (d, J = 6.8 Hz, 3H), 0.64
(d, J = 6.8 Hz, 3H). 13C NMR (100 MHz, CD3OD): δ 131.14, 129.81,
128.26, 127.89, 127.76, 121.70, 87.19, 80.98, 58.58, 52.68, 38.20,
36.12, 34.97, 30.45, 26.90, 20.10, 16.59. MS (ESI): m/z 712 [M + H]+
+
(100). HRMS m/z 712.4072 [(M + H)+ calcd for C41H54N5O6
712.4072].
Methyl ((S)-1-(2-(3-((3R,6S,Z)-3-Hydroxy-6-isopropyl-4,7-
dioxo-5,8-diaza-1(1,4)-benzenacyclododecaphan-10-en-3-yl)-
propyl)-2-(4-(thiophen-3-yl)benzyl)hydrazinyl)-3,3-dimethyl-
1-oxobutan-2-yl)carbamate (10d). PI 10d was prepared according
to procedure E from 9d (11.3 mg, 0.0149 mmol) and second-
generation Hoveyda−Grubbs catalyst (1.9 mg, 0.0030 mmol) in
anhydrous DCM (2 mL). The crude product was purified by
preparative HPLC to give 2.6 mg (24%) of 10d as white solid, along
with the double-bond migration compound 11d, obtained as white
solid in 4.4 mg, 40% isolated yield.
+
HRMS m/z 726.4238 [(M + H)+ calcd for C42H56N5O6 726.4231].
Methyl ((S)-1-(2-([1,1′-Biphenyl]-4-ylmethyl)-2-(3-((3R,6S,Z)-
3-hydroxy-6-isopropyl-4,7-dioxo-5,8-diaza-1(1,4)-benzenacy-
clododecaphan-9-en-3-yl)propyl)hydrazinyl)-3,3-dimethyl-1-
oxobutan-2-yl)carbamate (11c). 1H NMR (400 MHz, CD3OD): δ
7.59−7.40 (m, 9H), 7.34−7.30 (m, 2H), 7.27−7.19 (m, 1H), 6.95−
6.90 (m, 1H), 6.45 (d, J = 9.2 Hz, 1H), 4.96−4.93 (m, 1H), 3.95 (s,
2H), 3.73−3.70 (m, 1H), 3.54 (s, 3H), 3.49−3.45 (m, 1H), 2.88−2.79
(m, 4H), 2.48−2.40 (m, 1H), 2.19−2.16 (m, 1H), 2.09−1.84 (m, 5H),
1.60−1.52 (m, 2H), 0.84 (d, J = 6.8 Hz, 3H), 0.81 (s, 9H), 0.79 (d, J =
6.6 Hz, 3H). 13C NMR (100 MHz, CD3OD): δ 176.26, 171.85,
171.24, 164.27, 158.92, 142.25, 141.73, 141.15, 135.78, 131.17, 129.81,
129.39, 129.29, 128.27, 127.91, 127.77, 123.17, 113.98, 110.53, 89.92,
80.87, 63.05, 62.64, 61.10, 58.68, 52.71, 47.22, 38.34, 35.98, 34.96,
30.74, 26.93, 22.53, 19.51, 19.44. MS (ESI): m/z 726 [M + H]+ (100).
1
10d: H NMR (400 MHz, CD3OD): δ 7.58−7.56 (m, 3H), 7.46−
7.30 (m, 4H), 7.32 (dd, J = 7.9, 1.9 Hz, 1H), 7.13−7.11 (m, 1H), 7.04
(dd, J = 7.8, 1.9 Hz, 1H), 6.88 (dd, J = 7.8, 1.9 Hz, 1H), 6.36−6.27 (m,
1H), 5.74−5.67 (m, 1H), 3.90 (s, 2H), 3.75−3.73 (m, 3H), 3.54 (s,
3H), 3.44−3.40 (m, 1H), 2.82 (m, 4H), 2.04−2.02 (m, 2H), 1.89−
1.87 (m, 2H), 1.74−1.65 (m, 2H), 1.53−1.49 (m, 1H), 0.82 (s, 9H),
0.78 (d, J = 6.8 Hz, 3H), 0.75 (d, J = 6.8 Hz, 3H). 13C NMR (100
MHz, CD3OD): δ 176.79, 174.38, 169.42, 164.55, 159.57, 157.14,
154.70, 139.96, 131.11, 127.10, 80.42, 68.80, 30.76, 26.93. MS (ESI):
m/z 732 [M + H]+ (100). HRMS m/z 732.3803 [(M + H)+ calcd for
+
HRMS m/z 726.4230 [(M + H)+ calcd for C42H56N5O6 726.4231].
+
Methyl ((S)-1-(2-([1,1′-Biphenyl]-4-ylmethyl)-2-(3-((3R,6S,Z)-
3-hydroxy-6-isopropyl-4,7-dioxo-5,8-diaza-1(1,4)-benzenacy-
cloundecaphan-9-en-3-yl)propyl)hydrazinyl)-3,3-dimethyl-1-
oxobutan-2-yl)carbamate (13c). 1H NMR (400 MHz, CD3OD): δ
7.58−7.38 (m, 10H), 7.33−7.25 (m, 2H), 7.18 (dd, J = 7.8, 1.9 Hz,
1H), 7.03−7.01 (m, 1H), 6.45 (d, J = 10.2 Hz, 1H), 4.03 (d, J = 3.9
Hz, 1H), 3.93 (s, 2H), 3.75−3.37 (m, 3H), 3.53 (s, 3H), 2.98 (d, J =
C40H54N5O6 732.3795].
Methyl ((S)-1-(2-(3-((3R,6S,Z)-3-Hydroxy-6-isopropyl-4,7-
dioxo-5,8-diaza-1(1,4)-benzenacyclododecaphan-9-en-3-yl)-
propyl)-2-(4-(thiophen-3-yl)benzyl)hydrazinyl)-3,3-dimethyl-
1-oxobutan-2-yl)carbamate (11d). 1H NMR (400 MHz, CD3OD):
δ 7.58−7.56 (m, 3H), 7.46−7.24 (m, 4H), 7.25 (dd, J = 7.6, 1.8 Hz,
1H), 7.08 (dd, J = 7.8, 1.6 Hz, 1H), 6.95−6.90 (m, 2H), 6.45 (d, J =
6454
dx.doi.org/10.1021/jm500434q | J. Med. Chem. 2014, 57, 6444−6457

Journal of Medicinal Chemistry
Article
1
9.2 Hz, 1H), 4.96−4.94 (m, 1H), 3.90 (s, 2H), 3.72−3.70 (m, 1H),
3.55 (s, 3H), 3.47 (d, J = 10.1 Hz, 1H), 2.81−2.79 (m, 4H), 2.48−2.30
(m, 3H), 2.19−2.18 (m, 1H), 1.92−1.88 (m, 3H), 1.87−1.58 (m, 2H),
0.85−0.82 (s, 12H), 0.78 (d, J = 6.7 Hz, 3H). 13C NMR (100 MHz,
CD3OD): δ 174.38, 171.88, 170.65, 164.58, 159.63, 157.11, 154.69,
152.24, 147.34, 143.37, 136.41, 131.12, 129.28, 127.22, 127.10, 127.06,
127.07, 110.58, 80.86, 63.06, 59.40, 58.59, 52.70, 38.38, 35.38, 34.97,
33.06, 30.74, 30.46, 26.93, 23.72, 19.50, 19.44. MS (ESI): m/z 732 [M
oxobutan-2-yl)carbamate (11f). H NMR (400 MHz, CD3OD): δ
7.88−7.85 (m, 3H), 7.59−7.58 (m, 1H), 7.53 (d, J = 7.9 Hz, 2H),
7.27−7.25 (m, 1H), 7.09−7.07 (m, 2H), 6.95−6.91 (m, 2H), 6.45 (d, J
= 9.1 Hz, 1H), 3.95 (s, 2H), 3.73−3.72 (m, 1H), 3.55 (s, 3H), 3.47 (d,
J = 9.4 Hz, 1H), 2.86−2.84 (m, 4H), 2.49−2.43 (m, 1H), 2.21−2.17
(m, 2H), 2.06−2.03 (m, 2H), 1.88−1.81 (m, 2H), 1.60−1.68 (m, 2H),
0.84 (d, J = 6.8 Hz, 3H), 0.81−0.78 (m, 12H). 13C NMR (100 MHz,
CD3OD): δ 176.23, 144.37, 131.34, 127.37, 120.64, 111.41, 80.88,
63.05, 52.74, 47.25, 38.36, 35.99, 34.93, 30.78, 26.91, 19. MS (ESI):
m/z 733 [M + H]+ (100). HRMS m/z 733.3742 [(M + H)+ calcd for
C39H53N5O6S+ 733.3747].
+
+ H]+ (100). HRMS m/z 732.3791 [(M + H)+ calcd for C40H54N5O6
732.3795].
Methyl ((S)-1-(2-(3-((3R,6S,Z)-3-Hydroxy-6-isopropyl-4,7-
dioxo-5,8-diaza-1(1,4)-benzenacyclododecaphan-10-en-3-yl)-
propyl)-2-(4-(thiophen-2-yl)benzyl)hydrazinyl)-3,3-dimethyl-
1-oxobutan-2-yl)carbamate (10e). PI 10e was prepared according
to procedure E from 9e (20 mg, 0.0263 mmol) and second-generation
Hoveyda−Grubbs catalyst (3.3 mg, 0.0053 mmol) in anhydrous DCM
(3 mL). The crude product was purified by preparative HPLC to give
7.3 mg (34%) of 10e as white solid, along with the PI 12e, obtained as
white solid in 2.5 mg, 13% isolated yield.
Methyl ((S)-1-(2-(3-((3R,6S,E)-3-Hydroxy-6-isopropyl-4,7-
dioxo-5,8-diaza-1(1,4)-benzenacycloundecaphan-9-en-3-yl)-
propyl)-2-(4-(thiazol-2-yl)benzyl)hydrazinyl)-3,3-dimethyl-1-
1
oxobutan-2-yl)carbamate (14f). H NMR (400 MHz, CD3OD): δ
7.88−7.85 (m, 4H), 7.59 (d, J = 3.3 Hz, 1H), 7.55−7.53 (m, 3H), 7.27
(dd, J = 7.9, 1.9 Hz, 1H), 7.19 (dd, J = 7.8, 2.0 Hz, 1H), 7.03 (dd, J =
7.7, 1.9 Hz, 1H), 6.45 (dt, J = 10.2, 2.4 Hz, 1H), 4.03 (d, J = 3.9 Hz,
1H), 3.96 (s, 2H), 3.55 (s, 3H), 3.49−3.47 (m, 1H), 3.00−2.91 (m,
2H), 2.88−2.82 (m, 4H), 2.36−2.31 (m, 2H), 2.05−2.00 (m, 3H),
0.81−0.77 (m, 12H), 0.65 (d, J = 6.9 Hz, 3H). 13C NMR (100 MHz,
CD3OD): δ 177.75, 144.36, 138.91, 137.16, 133.71, 132.27, 131.31,
130.47, 129.46, 127.37, 121.72, 120.65, 113.19, 112.66, 80.88, 66.99,
62.54, 61.05, 58.59, 52.71, 38.21, 36.14, 33.07, 31.07, 30.75, 26.91,
23.73, 22.67, 20.10, 16.59. MS (ESI): m/z 719 [M + H]+ (100).
HRMS m/z 719.3594 [(M + H)+ calcd for C38H51N6O6S+ 719.3591].
Methyl ((S)-1-(2-(3-((3R,6S,Z)-3-Hydroxy-6-isopropyl-4,7-
dioxo-5,8-diaza-1(1,4)-benzenacycloundecaphan-10-en-3-yl)-
propyl)-2-(4-(pyridin-2-yl)benzyl)hydrazinyl)-3,3-dimethyl-1-
oxobutan-2-yl)carbamate (11g). PI 11g was obtained according to
procedure E from 9g (38.5 mg, 0.0510 mmol) and second-generation
Hoveyda−Grubbs catalyst (6.4 mg, 0.0100 mmol) in anhydrous DCM
(3.5 mL). The crude product was purified by preparative HPLC to
give 4.7 mg (13%) of 11g as white solid, along with PI 13g, obtained
as white solid in 7.4 mg, 20% isolated yield.
1
10e: H NMR (400 MHz, CD3OD): δ 7.55 (d, J = 8.2 Hz, 2H),
7.42 (d, J = 8.2 Hz, 2H), 7.43−7.35 (m, 3H), 7.14−7.07 (m, 3H), 6.89
(dd, J = 7.8, 1.9 Hz, 1H), 6.33 (dt, J = 10.5, 7.3 Hz, 1H), 5.75−5.68
(m, 1H), 3.91 (s, 2H), 3.76−3.73 (m, 2H), 3.57 (s, 3H), 3.44−3.41
(m, 2H), 2.85−2.82 (m, 4H), 2.06−2.02 (m, 1H), 1.95−1.85 (m, 2H),
1.77−1.72 (m, 1H), 1.62−1.48 (m, 3H), 0.83 (s, 9H), 0.80 (d, J = 6.8
Hz, 3H), 0.76 (d, J = 6.8 Hz, 3H). 13C NMR (100 MHz, CD3OD): δ
174.63, 170.71, 170.45, 162.84, 140.94, 134.84, 133.93, 133.55, 129.98,
129.79, 129.70, 127.90, 127.82, 127.65, 126.71, 125.07, 124.26, 122.66,
109.10, 88.50, 79.00, 61.64, 61.07, 57.96, 57.18, 51.30, 45.55, 36.71,
36.05, 33.53, 31.41, 30.68, 29.03, 25.51, 18.27, 16.98. MS (ESI): m/z
732 [M + H]+ (100). HRMS m/z 732.3796 [(M + H)+ calcd for
C40H54N5O6S+ 732.3795].
Methyl ((S)-1-(2-(3-((3R,6S,Z)-3-Hydroxy-6-isopropyl-4,7-
dioxo-5,8-diaza-1(1,4)-benzenacyclododecaphan-11-en-3-yl)-
propyl)-2-(4-(thiophen-2-yl)benzyl)hydrazinyl)-3,3-dimethyl-
1-oxobutan-2-yl)carbamate (12e). 1H NMR (400 MHz, CD3OD):
δ 7.55 (d, J = 7.9 Hz, 2H), 7.42 (d, J = 7.9 Hz, 2H), 7.36−7.26 (m,
3H), 7.10−7.07 (m, 3H), 6.94−6.91 (m, 1H), 6.46 (d, J = 9.1 Hz,
1H), 5.36−5.34 (m, 1H), 3.68 (s, 5H), 3.58−3.48 (m, 4H), 2.38−2.30
(m, 2H), 2.25−2.20 (m, 2H), 2.07−2.03 (m, 2H), 1.65−1.59 (m, 5H),
0.91−0.83 (s, 15H). 13C NMR (100 MHz, CD3OD): δ 170.39, 164.26,
155.70, 149.16, 110.52, 89.91, 69.94, 64.36, 52.67, 34.94, 33.06, 30.75,
30.45, 30.45, 28.10, 26.93, 25.30, 23.72, 14.43. MS (ESI): m/z 732 [M
+ H]+ (100). HRMS m/z 732.3790 [(M + H)+ calcd for
C40H54N5O6S+ 732.3795].
1
11g: H NMR (400 MHz, CD3OD): δ 8.69−8.68 (m, 1H), 8.19−
8.15 (m, 1H), 8.04−8.02 (m, 1H), 7.88 (d, J = 8.0 Hz, 2H), 7.63−7.53
(m, 4H), 7.27 (dd, J = 7.8, 1.8 Hz, 1H), 7.19 (dd, J = 7.8, 1.8 Hz, 1H),
7.03 (dd, J = 7.7, 1.8 Hz, 1H), 6.88 (d, J = 9.9 Hz, 1H), 6.47−6.43 (m,
1H), 4.06−4.00 (m, 3H), 3.75−3.73 (m, 1H), 3.53 (s, 3H), 2.99 (d, J
= 13.1 Hz, 1H), 2.87−2.83 (m, 3H), 2.36−2.31 (m, 2H), 2.05−2.02
(m, 2H), 1.88−1.85 (m, 1H), 1.55−1.49 (m, 2H), 0.82 (s, 9H), 0.78
(d, J = 6.9 Hz, 3H), 0.65 (d, J = 6.8 Hz, 3H). 13C NMR (100 MHz,
CD3OD): δ 178.86, 177.33, 171.89, 158.89, 157.20, 147.99, 138.88,
137.12, 132.24, 131.78, 131.29, 130.46, 129.44, 128.24, 124.61, 121.69,
112.84, 80.87, 68.85, 63.09, 62.53, 58.58, 52.67, 46.98, 38.20, 36.12,
34.95, 31.04, 30.72, 26.92, 23.71, 22.65, 20.10, 16.58. MS (ESI): m/z
713 [M + H]+ (100). HRMS m/z 713.4033 [(M + H)+ calcd for
Methyl ((S)-1-(2-(3-((3R,6S,Z)-3-Hydroxy-6-isopropyl-4,7-
dioxo-5,8-diaza-1(1,4)-benzenacyclododecaphan-10-en-3-yl)-
propyl)-2-(4-(thiazol-2-yl)benzyl)hydrazinyl)-3,3-dimethyl-1-
oxobutan-2-yl)carbamate (10f). PI 10f was prepared according to
procedure E from 9f (43 mg, 0.0565 mmol) and second-generation
Hoveyda−Grubbs catalyst (7.0 mg, 0.0113 mmol) in anhydrous DCM
(4 mL). The crude product was purified by preparative HPLC to give
7.7 mg (19% isolated yield) of 10f as white solid, along with two more
PIs, 11f obtained as white solid in 2.0 mg (5% isolated yield) and PI
14f (2.6 mg, 6.5%, white solid), respectively.
+
C40H53N6O6 713.4027].
Methyl ((S)-1-(2-(3-((3R,6S,Z)-3-Hydroxy-6-isopropyl-4,7-
dioxo-5,8-diaza-1(1,4)-benzenacyclododecaphan-9-en-3-yl)-
propyl)-2-(4-(pyridin-2-yl)benzyl)hydrazinyl)-3,3-dimethyl-1-
oxobutan-2-yl)carbamate (13g). 1H NMR (400 MHz, CD3OD): δ
8.79−8.77 (m, 1H), 8.50−8.46 (m, 1H), 8.25−8.24 (m, 1H), 7.90−
7.85 (m, 3H), 7.71−7.66 (m, 2H), 7.23 (dd, J = 7.8, 1.8 Hz, 1H),
7.13−7.04 (m, 2H), 6.95−6.91 (m, 1H), 6.45 (d, J = 9.3 Hz, 1H),
4.97−4.95 (m, 1H), 4.08−4.01 (m, 2H), 3.75−3.73 (m, 2H), 3.53 (s,
3H), 2.89−2.79 (m, 4H), 2.49−2.40 (m, 1H), 2.04−2.02 (m, 2H),
1.89−1.75 (m, 3H), 1.60−1.58 (m, 3H), 0.85−0.83 (m, 12H), 0.79 (d,
J = 6.8 Hz, 3H). 13C NMR (100 MHz, CD3OD): δ 176.16, 171.90,
171.18, 158.88, 155.77, 145.72, 145.12, 141.16, 135.72, 135.18, 131.68,
131.63, 131.39, 129.38, 129.02, 128.67, 125.75, 123.14, 114.00, 80.87,
68.84, 63.09, 62.40, 58.80, 52.70, 47.22, 38.30, 35.96, 34.91, 33.12,
32.08, 30.72, 26.93, 22.59, 19.81, 19.42. MS (ESI): m/z 727 [M + H]+
10f: 1H NMR (400 MHz, CD3OD): δ 7.88−7.84 (m, 3H), 7.57 (d,
J = 3.3 Hz, 1H), 7.52 (d, J = 8.1 Hz, 2H), 7.32 (dd, J = 8.0, 1.9 Hz,
1H), 7.12 (dd, J = 7.9, 1.9 Hz, 1H), 7.04 (dd, J = 7.9, 1.9 Hz, 1H), 6.89
(dd, J = 8.0, 1.9 Hz, 1H), 6.32 (dt, J = 10.6, 7.4 Hz, 1H), 5.73−5.66
(m, 1H), 3.95 (s, 2H), 3.75−3.71 (m, 3H), 3.55 (s, 3H), 3.43−3.40
(m, 1H), 2.87−2.83 (m, 4H), 2.08−2.01 (m, 1H), 1.96−1.85 (m, 4H),
1.68−1.57 (m, 2H), 0.81−0.78 (m, 12H), 0.74 (d, J = 6.8 Hz, 3H).
13C NMR (100 MHz, CD3OD): δ 176.07, 172.11, 171.89, 170.08,
158.88, 144.34, 142.36, 136.26, 135.34, 133.67, 131.41, 131.31, 131.11,
129.32, 129.24, 128.12, 127.35, 120.64, 80.43, 63.04, 62.38, 59.39,
58.80, 52.74, 46.97, 38.13, 37.48, 34.92, 33.06, 32.83, 32.11, 30.74,
26.91, 19.69, 18.41. MS (ESI): m/z 733 [M + H]+ (100). HRMS m/z
733.3743 [(M + H)+ calcd for C39H53N5O6S+ 733.3747].
Methyl ((S)-1-(2-(3-((3R,6S,Z)-3-Hydroxy-6-isopropyl-4,7-
dioxo-5,8-diaza-1(1,4)-benzenacyclododecaphan-9-en-3-yl)-
propyl)-2-(4-(thiazol-2-yl)benzyl)hydrazinyl)-3,3-dimethyl-1-
+
(100). HRMS m/z 727.4177 [(M + H)+ calcd for C41H55N6O6
727.4183].
6455
dx.doi.org/10.1021/jm500434q | J. Med. Chem. 2014, 57, 6444−6457

Journal of Medicinal Chemistry
Article
(15) Wang, F.; Ross, J. Atazanavir: a novel azapeptide inhibitor of
HIV-1 protease. Formulary 2003, 38 (12), 691−702.
ASSOCIATED CONTENT
* Supporting Information
Experimental details and spectroscopic data for all PIs, X-ray
pictures, and structure determination details. This material is
available free of charge via the Internet at http://pubs.acs.org.
■
S
(16) (a) Ekegren, J. K.; Ginman, N.; Johansson, A.; Wallberg, H.;
Larhed, M.; Samuelsson, B.; Unge, T.; Hallberg, A. Microwave-
accelerated synthesis of P1′-extended HIV-1 protease inhibitors
encompassing a tertiary alcohol in the transition-state mimicking
scaffold. J. Med. Chem. 2006, 49 (5), 1828−1832. (b) Ekegren, J. K.;
Unge, T.; Safa, M. Z.; Wallberg, H.; Samuelsson, B.; Hallberg, A. A
new class of HIV-1 protease inhibitors containing a tertiary alcohol in
the transition-state mimicking scaffold. J. Med. Chem. 2005, 48 (25),
8098−8102. (c) Wu, X.; Ohrngren, P.; Ekegren, J. K.; Unge, J.; Unge,
T.; Wallberg, H.; Samuelsson, B.; Hallberg, A.; Larhed, M. Two-
carbon-elongated HIV-1 protease inhibitors with a tertiary-alcohol-
containing transition-state mimic. J. Med. Chem. 2008, 51 (4), 1053−
1057. (d) Mahalingam, A. K.; Axelsson, L.; Ekegren, J. K.; Wannberg,
J.; Kihlstrom, J.; Unge, T.; Wallberg, H.; Samuelsson, B.; Larhed, M.;
Hallberg, A. HIV-1 protease inhibitors with a transition-state mimic
comprising a tertiary alcohol: improved antiviral activity in cells. J.
Med. Chem. 2010, 53 (2), 607−615. (e) Ohrngren, P.; Wu, X. Y.;
Persson, M.; Ekegren, J. K.; Wallberg, H.; Vrang, L.; Rosenquist, A.;
Samuelsson, B.; Unge, T.; Larhed, M. HIV-1 protease inhibitors with a
tertiary alcohol containing transition-state mimic and various P2 and
P1′ substituents. MedChemComm 2011, 2 (8), 701−709. (f) Wu, X.
Y.; Ronn, R.; Gossas, T.; Larhed, M. Easy-to-execute carbonylations:
microwave synthesis of acyl sulfonamides using Mo(CO)6 as a solid
carbon monoxide source. J. Org. Chem. 2005, 70 (8), 3094−3098.
(17) Larhed, M.; Hoshino, M.; Hadida, S.; Curran, D. P.; Hallberg, A.
Rapid fluorous Stille coupling reactions conducted under microwave
irradiation. J. Org. Chem. 1997, 62 (16), 5583−5587.
AUTHOR INFORMATION
Corresponding Author
*Phone: +46 18 4714667. Fax: +46 18 4714474. E-mail: Mats.
Larhed@orgfarm.uu.se.
■
Notes
The authors declare no competing financial interest.
ACKNOWLEDGMENTS
MDR thanks Olaf van der Veen for his contribution to the
synthetic work and Jean-Baptiste Ver
■
́
on for helpful discussions.
ABBREVIATIONS USED
■
AIDS, acquired immune deficiency syndrome; ART, antire-
troviral therapy; HIV, human immunodeficiency virus; MIDA,
N-methyliminodiacetic acid; PDB, Protein Data Bank; PI,
protease inhibitor; MW, microwaves; RCM, ring-closing
metathesis; SARs, structure−activity relationships
REFERENCES
■
(1) Brun-Vezinet, F.; Charpentier, C. Update on the human
immunodeficiency virus. Med. Mal. Infect. 2013, 43 (5), 177−184.
(2) Clavel, F.; Hance, A. J. HIV drug resistance. N. Engl. J. Med. 2004,
350 (10), 1023−1035.
(3) Sidibe, M. We need you beside us on the road to zero. AIDS
Care: Psychol. Socio-Med. Aspects AIDS/HIV 2012, 24 (8), 944.
(4) Rusconi, S.; Scozzafava, A.; Mastrolorenzo, A.; Supuran, C. T. An
update in the development of HIV entry inhibitors. Curr. Top. Med.
Chem. 2007, 7 (13), 1273−1289.
(18) Floistrup, E.; Goede, P.; Stromberg, R.; Malm, J. Synthesis of
estradiol backbone mimics via the Stille reaction using copper(II)
oxide as co-reagent. Tetrahedron Lett. 2011, 52 (2), 209−211.
(19) Wu, X.; Ohrngren, P.; Joshi, A. A.; Trejos, A.; Persson, M.;
Arvela, R. K.; Wallberg, H.; Vrang, L.; Rosenquist, A.; Samuelsson, B.
B.; Unge, J.; Larhed, M. Synthesis, X-ray analysis, and biological
evaluation of a new class of stereopure lactam-based HIV-1 protease
inhibitors. J. Med. Chem. 2012, 55 (6), 2724−2736.
(20) Nilsson, P.; Gold, H.; Larhed, M.; Hallberg, A. Microwave-
assisted enantioselective Heck reactions: expediting high reaction
speed and preparative convenience. Synthesis 2002, 11, 1611−1614.
(21) Larhed, M.; Hallberg, A. Microwave-promoted palladium-
catalyzed coupling reactions. J. Org. Chem. 1996, 61 (26), 9582−9584.
(22) Larhed, M.; Moberg, C.; Hallberg, A. Microwave-accelerated
homogeneous catalysis in organic chemistry. Acc. Chem. Res. 2002, 35
(9), 717−727.
(23) Kappe, C. O.; Dallinger, D. The impact of microwave synthesis
on drug discovery. Nature Rev. Drug Discovery 2006, 5 (1), 51−63.
(24) Callam, C. S.; Lowary, T. L. Suzuki cross-coupling reactions:
synthesis of unsymmetrical biaryls in the organic laboratory. J. Chem.
Educ. 2001, 78 (7), 947−948.
(25) Miyaura, N.; Suzuki, A. Palladium-catalyzed cross-coupling
reactions of organoboron compounds. Chem. Rev. 1995, 95 (7), 2457−
2483.
(26) Billingsley, K.; Buchwald, S. L. Highly efficient monophosphine-
based catalyst for the palladium-catalyzed Suzuki−Miyaura reaction of
heteroaryl halides and heteroaryl boronic acids and esters. J. Am. Chem.
Soc. 2007, 129 (11), 3358−3366.
(27) Tyrrell, E.; Brookes, P. The synthesis and applications of
heterocyclic boronic acids. Synthesis 2003, 4, 469−483.
(28) Knapp, D. M.; Gillis, E. P.; Burke, M. D. A general solution for
unstable boronic acids: slow-release cross-coupling from air-stable
MIDA boronates. J. Am. Chem. Soc. 2009, 131 (20), 6961−6963.
(29) Wu, X. Y.; Ekegren, J. K.; Larhed, M. Microwave-promoted
aminocarbonylation of aryl iodides, aryl bromides, and aryl chlorides in
water. Organometallics 2006, 25 (6), 1434−1439.
(30) Bold, G.; Fassler, A.; Capraro, H. G.; Cozens, R.; Klimkait, T.;
Lazdins, J.; Mestan, J.; Poncioni, B.; Rosel, J.; Stover, D.; Tintelnot-
Blomley, M.; Acemoglu, F.; Beck, W.; Boss, E.; Eschbach, M.;
Hurlimann, T.; Masso, E.; Roussel, S.; Ucci-Stoll, K.; Wyss, D.; Lang,
(5) Hawkins, T. Understanding and managing the adverse effects of
antiretroviral therapy. Antiviral Res. 2010, 85 (1), 201−209.
(6) Lin, J. H. Pharmacokinetics of biotech drugs: peptides, proteins
and monoclonal antibodies. Curr. Drug. Metab. 2009, 10 (7), 661−
691.
(7) Craik, D. J.; Fairlie, D. P.; Liras, S.; Price, D. The future of
peptide-based drugs. Chem. Biol. Drug. Des. 2013, 81 (1), 136−147.
(8) Rezai, T.; Yu, B.; Millhauser, G. L.; Jacobson, M. P.; Lokey, R. S.
Testing the conformational hypothesis of passive membrane
permeability using synthetic cyclic peptide diastereomers. J. Am.
Chem. Soc. 2006, 128 (8), 2510−2511.
(9) Meyer, F. M.; Collins, J. C.; Borin, B.; Bradow, J.; Liras, S.;
Limberakis, C.; Mathiowetz, A. M.; Philippe, L.; Price, D.; Song, K.;
James, K. Biaryl-bridged macrocyclic peptides: conformational
constraint via carbogenic fusion of natural amino acid side chains. J.
Org. Chem. 2012, 77 (7), 3099−3114.
(10) Pitsinos, E. N.; Vidali, V. P.; Couladouros, E. A. Diaryl ether
formation in the synthesis of natural products. Eur. J. Org. Chem. 2011,
2011 (7), 1207−1222.
(11) Driggers, E. M.; Hale, S. P.; Lee, J.; Terrett, N. K. The
exploration of macrocycles for drug discoveryan underexploited
structural class. Nature Rev. Dru. Discovery 2008, 7 (7), 608−624.
(12) Marsault, E.; Peterson, M. L. Macrocycles are great cycles:
applications, opportunities, and challenges of synthetic macrocycles in
drug discovery. J. Med. Chem. 2011, 54 (7), 1961−2004.
́
(13) Joshi, A.; Veron, J.-B.; Unge, J.; Rosenquist, Å.; Wallberg, H.;
Samuelsson, B.; Hallberg, A.; Larhed, M. Design and synthesis of P1-
P3 macrocyclic tertiary alcohol comprising HIV-1 protease inhibitors.
J. Med. Chem. 2013, 56 (22), 8999−9007.
(14) Croom, K. F.; Dhillon, S.; Keam, S. J. Atazanavir: a review of its
use in the management of HIV-1 infection. Drugs 2009, 69 (8), 1107−
1140.
6456
dx.doi.org/10.1021/jm500434q | J. Med. Chem. 2014, 57, 6444−6457

Journal of Medicinal Chemistry
Article
R. New aza-dipeptide analogues as potent and orally absorbed HIV-1
protease inhibitors: candidates for clinical development. J. Med. Chem.
1998, 41 (18), 3387−3401.
(31) Clavier, H.; Urbina-Blanco, C. A.; Nolan, S. P. Indenylidene
ruthenium complex bearing a sterically demanding NHC ligand: an
efficient catalyst for olefin metathesis at room temperature. Organo-
metallics 2009, 28 (9), 2848−2854.
(32) Rix, D.; Caijo, F.; Laurent, I.; Boeda, F.; Clavier, H.; Nolan, S.
P.; Mauduit, M. Aminocarbonyl group containing Hoveyda−Grubbs-
type complexes: synthesis and activity in olefin metathesis trans-
formations. J. Org. Chem. 2008, 73 (11), 4225−4228.
(33) Courchay, F. C.; Sworen, J. C.; Ghiviriga, I.; Abboud, K. A.;
Wagener, K. B. Understanding structural isomerization during
ruthenium-catalyzed olefin metathesis: a deuterium labeling study.
Organometallics 2006, 25 (26), 6074−6086.
(34) Schmidt, B. An olefin metathesis/double-bond isomerization
sequence catalyzed by an in situ generated ruthenium hydride species.
Eur. J. Org. Chem. 2003, 5, 816−819.
(35) Furstner, A.; Thiel, O. R.; Ackermann, L.; Schanz, H. J.; Nolan,
S. P. Ruthenium carbene complexes with N,N′-bis(mesityl) imidazol-
2-ylidene ligands: RCM catalysts of extended scope. J. Org. Chem.
2000, 65 (7), 2204−2207.
(36) Lazorova, L.; Hubatsch, I.; Ekegren, J. K.; Gising, J.; Nakai, D.;
Zaki, N. M.; Bergstrom, C. A. S.; Norinder, U.; Larhed, M.; Artursson,
P. Structural features determining the intestinal epithelial permeability
and efflux of novel HIV-1 protease inhibitors. J. Pharm. Sci. 2011, 100
(9), 3763−3772.
(37) Robinson, B. S.; Riccardi, K. A.; Gong, Y. F.; Guo, Q.; Stock, D.
A.; Blair, W. S.; Terry, B. J.; Deminie, C. A.; Djang, F.; Colonno, R. J.;
Lin, P. F. BMS-232632, a highly potent human immunodeficiency
virus protease inhibitor that can be used in combination with other
available antiretroviral agents. Antimicrob. Agents Chem. 2000, 44 (8),
2093−2099.
6457
dx.doi.org/10.1021/jm500434q | J. Med. Chem. 2014, 57, 6444−6457