Synthesis of new thienyl ring containing HIV-1 protease inhibitors: Promising preliminary pharmacological evaluation against recombinant HIV-1 proteases

Article abstract of DOI:10.1021/jm900846f

A series of new thienyl ring containing analogues of nelfinavir and saquinavir with different, substitution patterns were synthesized from suitable enantiopure diols. Their inhibitory activity against wild type recombinant HIV-1 protease was evaluated. In general thienyl groups spaced from the core by a methylene group gave products showing IC₅₀ in the nanomolar range, irrespective of the type and the substitution pattern of the heterocycle. The range of activity of the two most, active compounds is substantially maintained or even increased against two commonly selected mutants, under drug pressure, such as V32I and V82A.

Full text of DOI:10.1021/jm900846f

J. Med. Chem. 2010, 53, 1451–1457 1451
DOI: 10.1021/jm900846f
Synthesis of New Thienyl Ring Containing HIV-1 Protease Inhibitors: Promising Preliminary
Pharmacological Evaluation against Recombinant HIV-1 Proteases§
Carlo Bonini,*,† Lucia Chiummiento,† Margherita De Bonis,† Nadia Di Blasio,† Maria Funicello,† Paolo Lupattelli,†
Rocco Pandolfo,† Francesco Tramutola,† and Federico Berti‡
†Dipartimento di Chimica, Universitaꢀ degli Studi della Basilicata, Via Nazario Sauro 85, 85100 Potenza, Italy, and
‡Dipartimento di Scienze Chimiche, Universitaꢀ degli Studi di Trieste, Via Giorgeri 1, 34127 Trieste, Italy
Received May 25, 2009
A series of new thienyl ring containing analogues of nelfinavir and saquinavir with different substitution
patterns were synthesized from suitable enantiopure diols. Their inhibitory activity against wild type
recombinant HIV-1 protease was evaluated. In general thienyl groups spaced from the core by a
methylene group gave products showing IC50 in the nanomolar range, irrespective of the type and the
substitution pattern of the heterocycle. The range of activity of the two most active compounds is
substantially maintained or even increased against two commonly selected mutants, under drug
pressure, such as V32I and V82A.
Introduction
HIV protease has a C2-symmetric homodimeric structure.
As such, it selectively cleaves the Phe-Pro (Tyr-Pro) moiety of
the virus homoprotein6 and the rational design of inhibitors
based on substrate models is possible. Increasing the structur-
al diversity of inhibitor scaffolds may at least diminish the
accelerated development of multidrug-resistant variants.
Most of commercially available anti HIV protease drugs
have a peptidomimetic structure based on the tetrahedral
transition state mimetic concept, in which a nonhydrolyzable
hydroxyethylene or dihydroxyethylene or hydroxyethylamine
moiety is used as the central core of the molecule.
With structural diversity in mind, we explored the possibility
of introducing a thienyl ring in the core of several commercial
anti HIV protease drugs. It is well-known that a thienyl ring
mimics the phenyl group of phenylalanine in peptidomimetics7
and in many drugs.8 Furthermore, a recent study on C2-
symmetric diol-based HIV protease inhibitors has shown that
the introduction of a thienyl ring instead of a phenyl group
greatly enhances the antiviral activity of the compound.9
Thus, nelfinavir and saquinavir were chosen as model
structures into which a thienyl fragment was introduced. In
particular, the two cores were modified in the residues con-
taining the phenyl moiety, as shown in structures A and B in
Figure 1. We replaced them with a series of homo- or
polycyclic thienyl ring-containing fragments and kept the
same (R) absolute configuration of the carbon bearing the
central hydroxyl group and the same relative configuration of
the two vicinal stereocenters. To investigate the effect of chain
length on enzyme inhibition, we also introduced a series of
CH2 spaced thienyl fragments.
The global AIDS epidemic is one of the greatest medical
challenges of our time. Drug discovery for anti HIV-1a agents
has greatly accelerated in recent decades.1 Combination ther-
apy or highly active antiretroviral therapy (HAART) has
dramatically changed the history of the infection as well as
its morbidity and mortality, becoming the major current
treatment for AIDS.2 HIV/AIDS has become a manageable
disease in developed countries, where novel approaches such
as structured treatment interruptions (STIs) may be em-
ployed. On the basis of the evidence that viral protease is
essential to the life cycle of HIV-1, protease inhibitors have
beenintroduced into combination-therapyregimens, resulting
in a marked improvement in clinical outcomes.3 Nevertheless,
strains of HIV-1 become resistant to the currently available
drugs at an alarming rate, which accounts for the urgent need
for new, broad-spectrum protease inhibitors, which are effec-
tive against both new mutants and the wild-type virus.4
In recent years, several pharmaceutical companies have
developed, or are in the process of developing, an increasing
number of second-generation protease inhibitors, either with
or without a flexible structure.5 To date, nine inhibitors
against the HIV protease have been approved by the FDA.
However, the virus has shown a very high degree of adapt-
ability. Its high mutation rate is caused by the error-prone
viral reverse transcriptase and its fast replication rate. This
leads, especially under the additional selection pressure of
HAART, to the development of resistant strains of the virus.
The pronounced cross-resistance of all approved inhibitors
explains the ongoing need to seek new inhibitors.
*To whom correspondence should be addressed. Phone: 39
0971202254. Fax: 39 0971 202223. E-mail: bonini@unibas.it.
§Dedicated to the memory of Prof. A. M. Tamburro.
a Abbreviations: HIV-1, human immunodeficiency virus type 1;
HAART, highly active antiretroviral therapy; STI, structured treatment
interruption; FDA, Food and Drug Administration; WT, wild type;
PHIQ, perhydroisoquinoline; HOBt, 1-hydroxybenzotriazole; NMM,
4-methylmorpholine; EDC, 1-(3-dimethylaminopropyl)-3-ethylcarbo-
diimide.
Chemistry
Usually, the synthetic route to the commercial inhibitors
employsaconvergentsynthesisinwhichanelectrophilicchiral
amino alcohol moiety is the key building block for the central
core.10 Once formed, it is subsequently bonded to suitable side
chains. This general approach was also followed in our case,
r
2010 American Chemical Society
Published on Web 01/28/2010
pubs.acs.org/jmc

1452 Journal of Medicinal Chemistry, 2010, Vol. 53, No. 4
Bonini et al.
but we developed original synthetic routes for the prepara-
tion of the fragment bearing heterocyclic P1 substituent
(Scheme 1). Accordingly, the retrosynthesis required the
regio- and stereoselective introduction of the nitrogen group
fromthediol D and the Sharplessasymmetric dihydroxylation
of trans R,β-unsaturated esterE, which introduced stereogenic
centers with the right relative and absolute configuration.
Recently we successfully exploited such an approach to the
synthesis of two novel thienyl derivatives of nelfinavir and saqui-
navir 1 and 2 (Figure 2).11 This prompted us to widen its scope.
Because of the shortage of reports on asymmetric oxidation
of heteroaromatic acrylates and crotonates, we developed an
efficient Sharpless asymmetric dihydroxylation of thienyl and
benzothienyl derivatives, which afforded chiral 1,2-diols che-
moselectively in high ee and good chemical yield12 (Figure 3).
An azido group was introduced using two different ap-
proaches. According to our recently described procedure,11 3
was transformed into the corresponding cyclic sulfite. Then it
was submitted to regioselective and stereospecific nucleophilic
ring-opening reaction with sodium azide, affording the azido
hydroxy ester 7. Final reduction with BH3SMe2 and catalytic
NaBH4 gave the key intermediate azidodiol 813 (Scheme 2).
A different strategy was chosen for analogues containing an
additional methylene group. Thus, starting from the dihydrox-
ybutyl esters 4a-c, reduction by BH3SMe2 gave triols 10a-c
in almost quantitative yield (Scheme 3). Various reagents for
the formation of cyclic acetals were tested to perform selective
protection of hydroxyl groups on carbons 1 and 2. Among
them, freshly prepared 3,3-dimethoxypentane14 was the most
efficient. The free hydroxyl was then transformed into a
mesylate15 and substituted with the azido group. Final hydro-
lysis gave the azidodiols 13 in modest to good overall yield.
The common perhydroisoquinolinic residue was linked by
selective activation of primary hydroxyl group as a mesitylene
sulfonyl derivative and subsequent reaction with the commer-
cially available substituted perhydroisoquinoline (Scheme 4).
Finally, the azidoalcohols were transformed into amino alco-
hols 18 and 19 by Pd catalyzed hydrogenation in excellent yield.
Such amino alcohols are common intermediates for both
nelfinavir and saquinavir analogues, which are different only
in the P2 ligands. Thus, for the synthesis of nelfinavir analo-
gues, amino alcohols were condensed with 3-acetoxy-2-
methylbenzoic acid. Final deacetylation of the phenolic group
afforded the target compounds 22 and 23 (Scheme 5).
The characteristic dipeptide unit of saquinavir derivatives
26 was prepared by coupling of asparagine tert-butyl ester
25 and quinaldic acid 24 (Scheme 6). Hydrolysis with
TFA afforded the acid 27,11 which was coupled with amino
alcohols 18 and 19, affording the desired saquinavir deriva-
tives 28 and 29.
A noteworthy fact was that each synthetic sequence for
the most active compounds 23a and 29c, once optimized, has
been easily repeated in just 2 working weeks by a single operator.
Figure 1. Structures of nelfinavir and saquinavir and their thienyl
derivatives A and B.
Biological Evaluation
The inhibitory activities of these new thienyl derivatives of
nelfinavir and saquinavir on a recombinant wild-type HIV pro-
tease were evaluated, and the results are given in Tables 1 and 2.
From the reported data it is apparent that the presence of a
thienyl ring directly linked to the core generally causes a drop
in the activity, when compared with that of the original
commercial drugs. Indeed, it falls into the micromolar range
for compounds 1, 22 (Table 1) and 2, 28 (Table 2). The need of
a methylene spacer between the core and the heterocyclic
groups is proved by the general high activity of the corre-
sponding derivatives 23a-c and 29a-c, which falls in the
nanomolar range. In these cases the longer and more flexible
chain at P1 likely enters the S1 protease subsite more deeply
and the derivatives are better mimics of the model nelfinavir
and saquinavir. Small differences in the activity can also be
observed upon changing the type of heterocyclic moiety.
Among the nelfinavir analogues (Table 1), unsubstituted
Figure 2. Structures of thienyl derivatives of nelfinavir (1) and
saquinavir (2).
Figure 3. Starting thienyl and benzothienyl dihydroxy esters.
Scheme 1. Retrosynthetic Approach to the Thienyl Cores

Article
Journal of Medicinal Chemistry, 2010, Vol. 53, No. 4 1453
Scheme 6. Syntheses of Compounds 28 and 29a-c
Scheme 2. Synthesis of Azidodiol 8
Scheme 3. Syntheses of Azidodiols 13a-c
Table 1. Inhibitory Activity of Nelfinavir and Its Thienyl Derivatives
22 and 23a-c
Scheme 4. Syntheses of Amino Alcohols 18 and 19a-c
Scheme 5. Syntheses of Compounds 22 and 23a-c
a IC50 values were obtained by measuring the initial rates of hydro-
lysis of the fluorogenic substrate Abz-Thr-Ile-Nle-Phe(NO2)-Gln-Arg.
Results are the mean of at least three independent experiments.
IC50 = 0.6 nM. The data in Table 2 also show that an exten-
ded lipophilic interaction between the side chain and the
enzyme subsite S1 is mandatory for the proper binding of
these inhibitors.
2-thienyl derivative 23a reaches a promising IC50 = 2.9 nM,
while the more sterically demanding 4-phenyl-2-thienyl and
2-benzothienyl derivatives 23b and 23c show slightly higher
IC50 values.
We have also evaluated, in a preliminary test, the activity of
the most promising compounds 23a and 29c against two HIV
protease mutants that are commonly selected under drug pressure,
namely, V32I and V82A.16 The residual enzyme activities in
Saquinavir analogues take advantage of a more extended
heterocyclic moiety. Both 4-phenyl-2-thienyl 29b and 2-ben-
zothienyl derivative 29c show low IC50 values in the nanomo-
lar range, with the latter compound having a remarkable

1454 Journal of Medicinal Chemistry, 2010, Vol. 53, No. 4
Bonini et al.
Table 2. Inhibitory Activity of Saquinavir and Its Derivatives 28 and 29a-c
HIV-1 protease, and two of them, 23a and 29c, showed
activity in the same nanomolar range as the marketed drugs.
It is noteworthy that such a little change in the core structure
ofmarketedHIVprotease inhibitors nelfinavirand saquinavir
does not affect their activity against the wild type HIV
protease. Even more important and somewhat surprising is
the discovery that when 23a and 29c were preliminarily tested
against two commonly selected mutants under drug pressure,
such as V32I and V82A, the minor structural changes did not
substantially lead to cross-resistance to these two new com-
pounds. Inparticular, the activityof29ciscompletelyretained
or even enhanced with respect to that of saquinavir. Ongoing
workis in progress with more specific inhibition activity test of
the most promising compounds and with further research for
new and simplified potential HIV inhibitors.
Experimental Section
General. Unless otherwise specified, materials were pur-
chased from commercial suppliers and used without further
purification. THF, toluene, and diethyl ether were distilled from
sodium/benzophenone immediately before use. Dichloro-
methane was distilled from P2O5. DMF was freshly distilled
ꢁ
˚
and stored over 4 A sieves. Moisture-sensitive reactions were
conducted in oven- or flame-dried glassware under argon atmo-
sphere. All reactions were magnetically stirred and monitored
by thin-layer chromatography using precoated silica gel
(60 F254) plates. Column chromatography was carried out on
Merck silica gel (63-200 μm particle size) by progressive elution
with different solvents mixtures. Mass spectra were obtained by
GC/MS with electron impact ionization. 1H and 13C NMR
spectra were recorded in CDCl3 solutions at 500 and 125 MHz,
respectively. Chemical shifts (δ) were expressed in ppm and
coupling constants (J) in hertz. Optical rotations were deter-
mined at the sodium D line at 25 °C. HPLC analyses were
performed using Chiralcel OJ-H column with UV detection at
235 nm. Estimated purity of all compounds by combustional
analysis was always at least 95%.
(þ)-(2S,3R)-3-Azido-2-hydroxy-3-(4-phenylthiophen-2-yl)pro-
pionic acid ethyl ester (7) was obtained in 80% yield as a brown
oil. Rf = 0.5 (n-hexane/EtOAc 8:2). [R]D þ47.7 (c 1.24, CHCl3).
1H NMR (500 MHz, CDCl3) δ (ppm): 1.25 (t, J = 8.0 Hz, 3H),
3.32 (s, 1H), 4.25 (q, J = 8.0 Hz, 2H), 4.66 (d, J = 3.0 Hz, 1H),
5.12 (d, J = 3.0 Hz, 1H), 7.58 (d, J = 7.0 Hz, 2H), 7.45 (m, 5H).
13C NMR (125 MHz, CDCl3) δ (ppm): 14.4, 62.8, 62.9, 73.7,
122.0, 126.5, 127.7, 129.2, 135.6 136.3, 142.1, 171.2. Anal. Calcd
for C15H15N3O3S: C, 56.77; H, 4.76; N, 13.24; S, 10.10. Found:
C, 56.71; H, 4.86; N, 13.15; S, 10.10.
a IC50 values were obtained by measuring the initial rates of hydro-
lysis of the fluorogenic substrate Abz-Thr-Ile-Nle-Phe(NO2)-Gln-Arg.
Results are the mean of at least three independent experiments.
Table 3. Inhibitory Activity on HIV Protease Mutants
residual enzyme activitya (%)
WT
V32I
V82A
60
23a, 1 nM
23a, 10 nM
29c, 1 nM
34
5
97
32
54
53
37
a Residual enzyme activity values were obtained by measuring the
initial rates of hydrolysis of the fluorogenic substrate Abz-Thr-Ile-Nle-
Phe(NO2)-Gln-Arg. Results are the mean of at least three independent
experiments.
the presence of either 23a or 29c are reported in Table 3 and
compared with the residual activities of the WT enzyme
measured during the same kinetic run.17 The activity of
1 nM 23a is entirely lost on V32I, while it is 2-fold reduced
on V82A. When the concentration of 23a is increased to
10 nM, its activity is restored against V32I. Thus, 23a appears
more sensitive to the V32I mutation and its activity is seen to
decrease by 1 order of magnitude while the activity is largely
maintained against V82A. A similar behavior has also been
shown by nelfinavir against the V82A mutant.18
In contrast, 1 nM 29c fully maintains its activity against
V32I and is even somewhat more active against V82A than
against the WT enzyme. These are promising results, since
saquinavir is conversely about 20-fold less active against V32I
than against the WT protease.19
(þ)-(2S,3R)-3-Azido-3-(4-phenylthiophen-2-yl)propane-1,2-diol
(8) was isolated in 70% yield as a white solid. Mp = 96-98 °C.
Rf = 0.4 (petroleum ether/AcOEt 6:4). [R]D þ154.0 (c 1.20,
1
CHCl3). H NMR (500 MHz, CDCl3) δ (ppm): 2.24 (s, 1H),
2.70 (s, 1H), 3.77 (m, 2H), 3.94 (dd, J1 = 6.0 Hz, J2 = 4.0 Hz, 1H),
4.93 (d, J=6.0Hz, 1H), 7.34(t,J= 7.0 Hz, 1H), 7.43 (m, 3H), 7.49
(s, 1H), 7.60 (d, J = 7.0 Hz, 2H). 13C NMR (125 MHz, CDCl3) δ
(ppm): 62.9, 63.1, 74.2, 121.6, 121.8, 126.6, 127.1, 129.1, 135.5,
142.4, 148.2. Anal. Calcd for C13H13N3O2S: C, 56.71; H, 4.76; N,
15.26; S, 11.65. Found: C, 56.65; H, 4.81; N, 15.20; S, 11.75.
(þ)-(2R,3S)-4-(Thiophen-2-yl)butane-1,2,3-triol (10a) was iso-
lated in 98% yield as a colorless oil. Rf = 0.2 (CHCl3/MeOH
95:5). [R]D þ14.7 (c 0.55, CHCl3). 1H NMR (500 MHz, CD3OD)
Conclusion
δ (ppm): 3.00 (A part of an ABX system, JAB = 15 Hz, JAX
=
8.0 Hz, 1H), 3.15 (B part of an ABX system, JAB = 15 Hz,
We have synthesized, in a simple and easily reproducible
sequence, a series of thienyl containing analogues of HIV
protease reference inhibitors nelfinavir and saquinavir in
order to test the usefulness of such heteroaromatic residues
as mimics of P1 groups. Various derivatives possessing a
flexible side chain gave promising results against the wild type
JBX = 6.0 Hz, 1H), 3.55 (m, 1H), 3.60 (A part of an ABX system,
JAB = 11 Hz, JAX = 6.5 Hz, 1H), 3.65 (B part of an ABX system,
JAB= 11 Hz, JBX = 5.0 Hz, 1H), 3.84 (m, 1H), 6.92 (m, 2H), 7.20
(m, 1H). 13C NMR (125 MHz, CD3OD) δ (ppm): 33.7, 63.4,
72.5, 73.0, 123.5, 125.6, 126.5, 141.3. Anal. Calcd for C8H12O3S:
C, 51.04; H, 6.43; S, 17.03. Found: C, 51.10; H, 6.50; S, 17.05.

Article
General Procedure for the Synthesis of Cyclic Acetals 11. A
Journal of Medicinal Chemistry, 2010, Vol. 53, No. 4 1455
organic layer was dried over anhydrous Na2SO4 and concen-
trated, and the crude product was purified by chromatography
on silica gel (petroleum ether/EtOAc 8:2).
solution of the triol (10, 0.371 mmol) and 3,3-dimethoxypentane
(0.927 mmol, 122 mg) in dry THF (4 mL) under Ar atmosphere
was stirred for 15 min at room temperature. Then camphorsul-
fonic acid (0.148 mmol, 35 mg) was added at 0 °C, and the
mixture was allowed to warm to room temperature and was
stirred for 4 h. The reaction was quenched by adding Et3N
(0.15 mmol, 15 mg), and after solvent removal the crude
material was purified by chromatography on silica gel (petro-
leum ether/AcOEt 8:2), affording the desired product.
(2S,3S)-3-Azido-2-hydroxy-4-(thiophen-2-yl)butyl 2,4,6-tri-
methylbenzensulfonate (15a) was prepared running the reaction
at 50 °C for 24 h and was isolated in 64% yield as a colorless oil.
1
Rf = 0.8 (petroleum ether/AcOEt 6:4). H NMR (300 MHz,
CDCl3) δ (ppm): 2.34 (s, 3H), 2.65 (s, 6H), 3.07 (A part of an
ABX system, JAB = 15.0 Hz, JAX = 8.0 Hz, 1H), 3.30 (B part of
an ABX system, JAB = 15.0 Hz, JBX = 3.5 Hz, 1H), 3.74 (m,
4H), 4.07 (A part of an ABX system, JAB = 11.0 Hz, JAX
6.0 Hz, 1H), 4.17 (B part of an ABX system, JAB 11.0 Hz, JBX
=
=
(þ)-(R)-1-((R)-2,2-Diethyl-1,3-dioxolan-4-yl)-2-(thiophen-2-yl)-
ethanol (11a) was isolated in 70% yield as a colorless oil. Rf =
0.6 (petroleum ether/AcOEt 8:2). [R]D þ1.66 (c 1.75, CHCl3).
1H NMR (500 MHz, CD3OD) δ (ppm): 0.91 (t, J = 8.0 Hz,
3H), 0.93 (t, J = 8.0 Hz, 3H), 1.64 (q, J = 8.0 Hz, 2H), 1.71
(dq, J1 = 8.0 Hz, J2 = 3.0 Hz, 2H), 3.02 (A part of an ABX
system, JAB = 15.5 Hz, JAX = 5.5 Hz, 1H), 3.07 (B part of an
ABX system, JAB = 15.5 Hz, JBX = 7.5 Hz, 1H), 3.72 (A part
of an ABX system, JAB = 8.0 Hz, JAX = 8.0 Hz, 1H), 3.79 (m,
1H), 3.99 (B part of an ABX system JAB = 8.0 Hz, JBX = 7.0
Hz, 1H), 4.06 (m, 1H), 6.90 (d, J = 3.0 Hz, 1H), 6.97 (m, 1H),
7.20 (d, J = 5.0 Hz, 1H). 13C NMR (125 MHz, CD3OD) δ
(ppm): 8.3, 8.5, 29.3, 29.8, 34.7, 66.5, 73.1, 78.8, 113.5, 124.6,
126.4, 127.2, 139.8. EI-MS m/z:, 256 (1) [Mþ], 227 (100). Anal.
Calcd for C13H20O3S: C, 60.91; H, 7.86; S, 12.51. Found: C,
60.80; H, 7.91; S, 12.45.
Synthesis of Mesylates 12. To a solution of the alcohol (11,
0.25 mmol) and Et3N (0.38 mmol, 0.05 mL) in dry CH2Cl2
(2 mL) under argon atmosphere, methanesulfonyl chloride
(0.38 mmol, 0.03 mL) was added at 0 °C. The solution was
stirred for 1 h and then warmed to room temperature, dissolved
in CHCl3 (10 mL), and washed with saturated aqueous NH4Cl
(10 mL) and with saturated aqueous NaCl (10 mL). The organic
phase was dried on anhydrous Na2SO4, the solvent was evapo-
rated under reduced pressure, and the crude was purified on
silica gel, except for 19 which was used in the subsequent
reaction as crude product.
Synthesis of the Azidodiols 13. To a solution of the starting
mesylate (12, 0.12 mmol) in DMF (2 mL), solid NaN3
(0.24 mmol, 16 mg) was added at 0 °C, and the solution was
kept stirring at 100 °C. After 15 h the solution was cooled to
room temperature, EtOAc (10 mL) was added, and the mixture
was washed with H2O (10 mL ꢀ 2). The organic phase was dried
on anhydrous Na2SO4, and the solvent was evaporated under
reduced pressure. To the solution of the crude in THF (3 mL),
1 M aqueous HCl (3 mL) was added, and the reaction mixture
was warmed to 60 °C for 2 h. Then EtOAc (10 mL) and saturated
aqueous NaHCO3 (10 mL) were added until pH 7 was obtained.
The organic phase was separated, washed with saturated aqu-
eous NaCl, and dried on anhydrous Na2SO4. The solvent was
distilled under reduced pressure, and the crude was purified by
chromatography on silica gel.
3.0 Hz, 1H), 6.97 (m, 2H), 7.23 (d, J = 5.0 Hz, 1H). 13C NMR
(75 MHz, CDCl3) δ (ppm): 21.0, 22.6, 31.0, 64.2, 70.0, 70.6,
124.9, 126.9, 127.0, 129.7, 131.9, 140.0, 143.8. Anal. Calcd for
C17H21N3O4S2: C, 51.63; H, 5.35; N, 10.62; S, 16.22. Found: C,
51.55; H, 5.40; N, 10.56; S, 16.30.
Synthesis of Aminoazido Alcohols 16 and 17. To a solution of
mesitylen azidoalcohol (14 or 15, 0.32 mmol) in i-PrOH
(10.8 mL), PHIQ (132 mg, 0.55 mmol) and K2CO3 (90 mg,
0.65 mmol) were added. The mixture was stirred at 50 °C for
16-26 h, then i-PrOH was evaporated under reduced pressure
and the residue dissolved in EtOAc (20 mL). The organic layer
was washed with saturated aqueous NH4Cl and brine, dried
over anhydrous Na2SO4, and concentrated under reduced pres-
sure. The resulting residue was purified by chromatography on
silica gel (petroleum ether/EtOAc 7:3).
(-)-(3S,4aS,8aS)-2-((2R,3S)-3-Azido-2-hydroxy-4-(thiophen-2-yl)-
butyl)-N-tert-butyldecahydroisoquinoline-3-carboxamide (17a)
was isolated in 89% yield as white solid. Mp = 119-121 °C.
Rf = 0.2 (petroleum ether/AcOEt 8:2). [R]D -71.15 (c 1.3,
1
CHCl3). H NMR (500 MHz, CDCl3) δ (ppm): 1.35 (s, 9H),
1.53 (m, 14 H), 2.44 (m, 2H), 2.72 (m, 2H), 2.89 (m, 1H), 3.04 (A
part of an ABX system, JAB = 15.5 Hz, JAX = 9.5 Hz, 1H), 3.31
(B part of an ABX system, JAB = 15.5 Hz, JBX = 3.0 Hz, 1H),
3.63 (m, 2H), 5.86 (br s, 1H), 6.96 (m, 2H), 7.22 (d, J = 5.0 Hz,
1H). 13C NMR (125 MHz, CDCl3) δ (ppm): 21.0, 26.0, 26.3,
28.9, 30.5, 30.8, 31.3, 33.5, 36.3, 51.4, 58.5, 61.1, 66.8, 70.8, 71.0,
124.9, 126.7, 127.3, 140.0, 173.8. Anal. Calcd for C22H35N5O2S:
C, 60.94; H, 8.14; N, 16.15; S, 7.39. Found: C, 60.90; H, 8.18; N,
16.20; S, 7.35.
General Procedure for the Preparation of Amino Alcohols 18
and 19. A mixture of azido alcohol (16 or 17, 1 mmol) and Pd/C
10% (60 mg) in 20 mL of CH3OH was kept stirring at room
temperature under H2 (1 atm). After 2 h, the catalyst was filtered
off and the mixture was evaporated under reduced pressure. The
products were isolated by chromatographic purification on
silica gel of the crude product.
(-)-(3S,4aS,8aS)-2-((2R,3S)-3-Amino-2-hydroxy-4-(thiophen-2-yl)-
butyl)-N-tert-butyldecahydroisoquinoline-3-carboxamide (19a) was iso-
lated in 82% yield as a brown oil. Rf = 0.4 (CHCl3/CH3OH
9/1). [R]D -113.9 (c 0.8 CHCl3). 1H NMR (500 MHz, CDCl3) δ
(ppm): 1.35 (s, 9H), 1.60 (m, 12H), 2.32 (d, J = 11.0 Hz, 1H), 2.56
(A part of an ABX system, JAB = 13.0 Hz, JAX = 8.0 Hz, 1H),
2.79 (B part of an ABX system, JAB = 13.0 Hz, JBX = 3.0 Hz, 1H),
2.80 (dd, J1 = 3.0 Hz, J2 = 11.0 Hz, 1H), 3.15 (m, 3H), 3.32 (m,
1H), 3.98 (m, 1H), 4.73 (br s, 3H), 6.43 (br s, 1H), 6.93 (m, 1H),
6.97 (s, 1H), 7.16 (d, J = 5.0 Hz, 1H). 13C NMR (125 MHz,
CDCl3) δ (ppm): 20.7, 21.0, 25.9, 26.5, 29.0, 30.1, 30.5, 31.0, 32.8,
33.2, 51.5, 58.6, 60.3, 70.1, 124.9, 127.1, 127.5, 140.0, 173.8. Anal.
Calcd for C22H37N3O2S: C, 64.83; H, 9.15; N, 10.31; S, 7.87.
Found: C, 64.85; H, 9.10; N, 10.35; S, 7.90.
(þ)-(2S,3S)-3-Azido-4-(thiophen-2-yl)butane-1,2-diol (13a) was
isolated in 30% yield as a colorless oil. Rf = 0.25 (petroleum ether/
1
AcOEt 6/4). [R]D þ32.0 (c 0.5, CHCl3). H NMR (300 MHz,
CDCl3) δ (ppm): 2.93 (br s, 2H), 3.05 (A part of an ABX system,
JAB = 15.0 Hz, JAX = 8.0 Hz, 1H), 3.28 (B part of an ABX system,
J
AB = 15.0 Hz, JBX = 3.0 Hz, 1H), 3.74 (m, 4H), 6.97 (m, 2H),
7.23 (d, J = 5.0 Hz, 1H). 13C NMR (75 MHz, CDCl3) δ (ppm):
31.6, 63.4, 65.5, 72.9, 125.1, 126.9, 127.3, 139.3. Anal. Calcd for
C8H11N3O2S: C, 45.06; H, 5.20; N, 15.70; S, 15.04. Found: C,
45.10; H, 5.15; N, 15.72; S, 15.10.
Synthesis of Mesitylenes 14 and 15. To a cold (0 °C) stirred
solution of azidodiol (8 or 13, 1.43 mmol) in CH2Cl2 (3.7 mL),
pyridine (236 mg, 2.86 mmol) and mesitylenesulfonyl chloride
(344 mg, 1.57 mmol) were added. The reaction mixture was
stirred at 0 °C for 4 h and then at 20 °C for 20 h. The solvent was
evaporated under reduced pressure, and the crude mixture was
dissolved in EtOAc and washed with cold 2 M aqueous HCl
(5 mL), water, saturated aqueous NaHCO3, and brine. The
Synthesis of Nelfinavir and Saquinavir Analogues: Preparation
of Amides 22, 23, 28, and 29. To a solution of the amino alcohol
(18 or 19, 1.15 mmol) in anhydrous CH2Cl2 (10 mL), 2-methyl-
3-oxyacetylbenzoic acid (223 mg, 1.15 mmol, for nelfinavir
analogues) or (S)-4-amino-4-oxo-2-(quinoline-2-carboxamido)-
butanoic acid 27 (330 mg, 1.15 mmol, for saquinavir analogues),
HOBt (155 mg, 1.15 mmol), and NMM (0.25 mL, 2.3 mmol)
were added under argon atmosphere. The mixture was cooled to

1456 Journal of Medicinal Chemistry, 2010, Vol. 53, No. 4
Bonini et al.
0 °C. Then EDC (264 mg, 1.38 mmol) was slowly added. After
1 h at 0 °C, the mixture was allowed to reach room temperature.
After reaction completion, AcOEt (20 mL) was added and the
mixture was washed with H2O, saturated aqueous NaHCO3,
10% aqueous citric acid, saturated aqueous NaHCO3, and
finally brine. The organic phase was dried over anhydrous
Na2SO4 and filtered and the solvent evaporated under reduced
pressure. The products were isolated by chromatographic puri-
fication on silica gel of the crude product.
(-)-3-((2S,3R)-4-((3S,4aS,8aS)-3-(tert-Butylcarbamoyl)octahydro-
isoquinolin-2(1H)-yl)-3-hydroxy-1-(thiophen-2-yl)butan-2-ylcarba-
moyl)-2-methylphenyl acetate (21a) was isolated, after 4 h of
reaction, in 81% yield as a brown oil. Rf = 0.4 (CHCl3/CH3OH
95:5). [R]D -48.7 (c 1.14 CHCl3). 1H NMR (500 MHz, CDCl3) δ
(ppm): 1.19 (s, 9H), 1.20-2.15 (12H), 2.17 (s, 3H), 2.37 (s, 3H),
2.39 (m, 2H), 3.00 (m, 1H), 3.18 (m, 1H), 3.74 (m, 1H), 4.02 (m,
1H), 4.59 (m, 1H), 5.74 (br s, 1H), 6.92 (m, 1H), 6.97 (m, 2H),
7.04 (d, J = 8.0 Hz, 1H), 7.12 (d, J = 8.0 Hz, 1H), 7.18 (m, 1H).
13C NMR (125 MHz, CDCl3) δ (ppm): 12.8, 20.8, 25.9, 26.2,
28.4, 28.4, 30.4, 30.8, 30.8, 33.5, 36.0, 51.1, 56.2, 58.8, 59.5, 70.5,
71.0, 123.7, 123.9, 124.8, 125.8, 126.4, 126.9, 128.8, 137.8, 141.2,
149.6, 169.0, 170.4, 173.8. Anal. Calcd for C32H45N3O5S: C,
65.84; H, 7.77; N, 7.20; S, 5.49. Found: C, 65.75; H, 7.85; N,
7.22; S, 5.53.
Instruction, University and Research—Rome, Project FIRB:
RBNE017F8N), and University of Basilicata for financial
support.
Supporting Information Available: General experimental
procedures, spectroscopic data for compounds 5, 6a, 9, 10b,c,
11b,c, 12b,c, 13b,c, 14, 15b,c, 16, 17b,c, 18, 19b,c, 20, 21b,c,
22, 23a-c, 28, 29a,b, and enzyme assays procedure. This
material is available free of charge via the Internet at http://
pubs.acs.org.
References
(1) De Clerck, E. New approaches toward Anti-HIV chemotherapy.
J. Med. Chem. 2005, 48, 1297–1313.
(2) (a) Sepkowitz, K. A. AIDS—The first 20 years. N. Engl. J. Med.
2001, 344, 1764–1772. (b) Cihlar, T.; Bischofberger, N. Recent devel-
opments in antiretroviral therapies. Annu. Rep. Med. Chem. 2000, 35,
177–189. (c) Flexner, C. W. Pharmacology of Drug Interactions of
HIV Protease Inhibitors. In Protease Inhibitors in AIDS Therapy;
Ogden, R. C., Flexner, C. W., Eds.; Marcel Dekker: New York, 2001;
pp 139-160.
(3) (a) Hammer, S. M.; Squires, H. E.; Hughes, M. D.; Grimes, J. M.;
Demeter, L. M.; Currier, J. S.; Eron, J. J., Jr.; Feinberg, J. E.;
Balfour, H. H.; Deyton, L. R.; Chodakewitz, J. A.; Fischl, M. A.
A controlled trial of two nucleoside analogues plus indinavir in
persons with human immunodeficiency virus infection and CD4
cell counts of 200 per cubic millimetre or less. AIDS Clinical Trials
Group 320 Study Team. N. Engl. J. Med. 1997, 337, 725–733.
(b) Collier, A. C.; Coombs, R. W.; Schoenfeld, D. A.; Bassett, R. L.;
Timpone, J.; Baruch, A.; Jones, M.; Facey, K.; Whitcare, C.; McAuliffe,
V. J.; Friedman, H. M.; Merigan, T. C.; Reichman, R. C.; Hooper, C.;
Corey, L. Treatment of human immunodeficiency virus infection with
saquinavir, zidovudine and zalcitabine. N. Engl. J. Med. 1996, 334,
1011–1017. (c) Cameron, D. W.; Heat-Chiozzi, M.; Danner, S.; Cohen,
C.; Kravcik, S.; Maurath, C.; Sun, E.; Henry, D.; Rode, R.; Potthoff, A.;
Leonard, J. Randomised placebo-controlled trial of ritonavir an ad-
vanced HIV-1 disease. Lancet 1998, 351, 543–549. (d) Gulick, R. M.;
Mellors, J. W.; Havlir, D.; Eron, J. J.; Gonzales, C.; McMahon, D.;
Richman, D. D.; Valentine, F. T.; Jonas, L.; Meibohm, A.; Emini, E. A.;
Chodakewitz, J. A. Treatment with indinavir, zidovudine and lamivu-
dine in adults with human immunodeficiency virus infection and prior
antiretroviral therapy. N. Engl. J. Med. 1997, 337, 734–739.
(4) Freire, E. Overcoming HIV-1 resistance to protease inhibitors.
Drug Discovery Today: Dis. Mech. 2006, 3, 281–286.
Synthesis of Nelfinavir Analogues: Preparation of Final Pro-
ducts 22 and 23. To a stirring solution of amide (20 or 21, 1.0
mmol) in CH3OH (15 mL), Na (5 mg, 0.15 mmol) was carefully
added at room temperature. After 1 h the solvent was evapo-
rated under reduced pressure and the residue dissolved in AcOEt
(20 mL). The mixture was washed with H2O (2 ꢀ 20 mL), dried
over anhydrous Na2SO4, filtered, and evaporated under re-
duced pressure. The products were isolated by chromatography
on silica gel.
(-)-(3S,4aS,8aS)-N-tert-Butyl-2-((2R,3S)-2-hydroxy-3-(3-hydro-
xy-2-methylbenzamido)-4-(4-phenylthiophen-2-yl)butyl)decahydro-
isoquinoline-3-carboxamide (23b) was isolated in 70% yield as a
colorless oil. Rf = 0.5 (petroleum ether/AcOEt 4:6). [R]D -43.1
1
(c 1.5, CHCl3). H NMR (500 MHz, CDCl3) δ (ppm): 1.5 (m,
12H), 1.17 (s, 9H), 1.85 (s, 3H), 2.26 (m, 2H), 2.52 (m, 1H), 2.61
(m, 1H), 2.96 (m, 1H), 3.15 (m, 1H), 3.62 (m, 1H), 4.02 (m, 1H),
4.56 (bs, 1H), 5.90 (bs, 1H), 6.70 (d, J = 7.0 Hz, 1H), 6.77 (d,
J = 7.0 Hz, 1H), 6.84 (m, 2H), 7.28 (m, 2H), 7.35 (t, J = 8.0 Hz,
2H), 7.54 (d, J = 7.0 Hz, 2H). 13C NMR (125 MHz, CDCl3) δ
(ppm): 12.3, 20.5, 25.7, 26.2, 28.4, 30.3, 30.7, 33.4, 35.9, 51.2,
55.7, 58.8, 60.4, 70.3, 71.2, 116.8, 118.8, 122.5, 125.4, 126.2,
126.3, 127.0, 128.7, 135.9, 137.5, 141.9, 154.8, 171.7, 174.25.
Anal. Calcd for C36H47N3O4S: C, 69.98; H, 7.67; N, 6.80; S,
5.19. Found: C, 70.05; 7.70; N, 6.70; S, 5.20.
ꢀ
(5) Rusconi, S.; Vigano, O. New HIV protease inhibitors for drug-
resistant viruses. Therapy 2006, 3, 79–88.
(6) (a) Tomasselli, A. G.; Heinrikson, R. L. Targeting the HIV-
protease in AIDS therapy: a current clinical perspective. Biochim.
Biophys. Acta 2000, 1477, 189–214. (b) Wlodawer, A.; Gustchina, A.
Structural and biochemical studies of retroviral proteases. Biochim.
Biophys. Acta 2000, 1477, 16–34.
(7) Gante, J. Peptidomimetics—tailored enzyme inhibitors. Angew.
Chem., Int. Ed. Engl. 1994, 33, 1699–1720.
(-)-(R)-N1-((2S,3R)-1-(Benzo[b]thiophen-2-yl)-4-((3S,4aS,8aS)-
3-(tert-butylcarbamoyl)octahydroisoquinolin-2(1H)-yl)-3-hydr-
oxybutan-2-yl)-2-(quinoline-2-carboxamido)succinamide (29c)
was obtained after 24 h of reaction and isolated in 26% yield
as a brown oil. Rf = 0.3 (CHCl3/CH3OH 9:1). [R]D -1.02 (c
(8) (a) Corral, C.; Lissavetzky, J.; Manzanares, I.; Darias, V.;
Exposito-Orta, M. A.; Conde, J. A. M.; Sanchez-Mateo, C. C.
Synthesis and preliminary pharmacological evaluation of thio-
phene analogues of viloxazine as potential antidepressant drugs.
Bioorg. Med. Chem. 1999, 7, 1349–1359. (b) Zhang, A.; Zhou, G.;
Rong, S.-B.; Johnson, K. M.; Zang, M.; Kozikowsky, A. P. Thiophene
derivatives: a new series of potent norepinephrine and serotonin
reuptake inhibitors. Bioorg. Med. Chem. Lett. 2002, 12, 993–995.
(c) Uckun, F. M.; Pendergrass, S.; Maher, D.; Zhu, D.; Tuel-Ahlgren, L.;
Mao, C.; Venkatachalam, T. K. N0-[2-(2-Thiophene)ethyl]-N0-[2-(5-
bromopyridyl)]thiourea as a potent inhibitor of NNI-resistant and
multidrug-resistant human immunodeficiency virus-1. Bioorg. Med.
Chem. Lett. 1999, 9, 3411–3416.
1
0.5, CHCl3). H NMR (300 MHz, CDCl3) δ (ppm): 1.27 (s,
9H), 1.51 (13 H), 3.15 (m, 7H), 4.01 (m, 1H), 4.47 (m, 1H), 5.78
(bs, 1H), 6.15 (bs, 1H), 6.25 (bs, 1H), 6.97 (bs, 2H), 7.04
(bs, 1H), 7.41 (m, 3H), 7.77 (m, 3H), 8.11 (m, 2H), 9.12 (d,
J = 7.0 Hz, 1H). 13C NMR (125 MHz, CDCl3) δ (ppm): 20.6,
25.7, 26.0, 28.6, 28.6, 29.7, 30.6, 31.8, 35.8, 45.7, 50.1, 50.2,
50.9, 58.8, 70.5, 118.5, 121.8, 122.7, 123.3, 123.7, 127.6, 128.2,
129.4, 130.1, 130.1, 137.4, 139.4, 139.8, 142.0, 146.5, 148.6,
164.8, 170.5, 172.7. Anal. Calcd for C40H50N6O5S: C, 66.09;
H, 6.93; N, 11.56; S, 4.41. Found: C, 66.15; H, 6.85; N, 11.58;
S, 4.42.
€
(9) Muhlman, A.; Classon, B.; Hallberg, A.; Samuelsson, B. Synthesis
of potent C-symmetric, diol based HIV-1 protease inhibitors.
Investigation of thioalkyl and thioaryl P1/P10 substituents. J.
Med. Chem. 2001, 44, 3402–3406.
(10) Ghosh, A. K.; Bilcer, G.; Schiltz, G. Syntheses of FDA approved
HIV protease inhibitors. Synthesis 2001, 15, 2203–2229.
(11) (a) Bonini, C.; Chiummiento, L.; De Bonis, M.; Funicello, M.;
Lupattelli, P. Synthesis of a first thiophene analog of the HIV
protease inhibitor nelfinavir. Tetrahedron Lett. 2004, 45, 2797–
2799. (b) Bonini, C.; Chiummiento, L.; De Bonis, M.; Funicello, M.;
Lupattelli, P.; Suanno, G.; Berti, F.; Campaner, P. Synthesis, biological
activity and modeling studies of two novel anti HIV Pr inhibitors with a
thiophene containing hydroxyethylamino core. Tetrahedron 2005, 61,
Acknowledgment. The authors gratefully acknowledge
Prof. Silvano Geremia (University of Trieste) for the generous
gift of protease mutants, Dr. Adrian Ostric (University of
Trieste) for inhibition tests on mutants, MIUR (Ministry of

Article
6580–6589.(c) Bonini, C.; Funicello, M.; Lupattelli, P.; Chiummiento,
L.; De Bonis, M. Inibitori di HIV-1 Proteasi. Italian Patent Application
MI2008A000415, 2008.
(12) Bonini, C.; Chiummiento, L.; De Bonis, M.; Funicello, M.;
Lupattelli, P.; Pandolfo, R. Application of Sharpless asymmetric
dihydroxylation to thienyl- and benzothienyl acrylates and croto-
nates. Tetrahedron: Asymmetry 2006, 17, 2919–2924.
Journal of Medicinal Chemistry, 2010, Vol. 53, No. 4 1457
(15) Mesylate 12a was not isolated and was used in the subsequent
reaction as a crude.
(16) (a) Gulnik, S. V.; Suvorov, S. I.; Liu, B.; Yu, B.; Anderson, B.;
Mitsuya, H.; Erickson, J. B. Kinetic characterization and cross-
resistance patterns of HIV-1 protease mutants selected under drug
pressure. Biochemistry 1995, 34, 9282–9287. (b) Svicher, V.;
Ceccherini-Silberstein, F.; Erba, F.; Santoro, M.; Gori, C.; Bellocchi,
M. C.; Giannella, S.; Trotta, M. P.; Monforte, A. D.; Antinori, A.; Perno,
C. F. Novel human immunodeficiency virus type 1 protease mutations
potentially involved in resistance to protease inhibitors. Antimicrob.
Agents Chemother. 2005, 49 (5), 2015–2025.
(13) Unfortunately, although expected, the same approach gave the
wrong regioisomer 9 in the case of sulfite 6a, due to the higher
electrophilicity of R-carbon of sulfinyl esters.
(17) The WT HIV protease used for the IC50 measurements differs from
that used as reference for the mutant test, due to their different
commercial sources (Bioczech and Bachem, respectively).
(18) Mosebi, S.;Morris, L.;Dirr, H. W.;Sayed, Y. Active-sitemutationsin
the South African human immunodeficiency virus type 1 subtype C
protease have a significant impact on clinical inhibitor binding: kinetic
and thermodynamic study. J. Virol. 2008, 82, 11476–11479.
(19) Kovalevsky, A. Y.; Ghosh, A. K.; Weber, I. T. Solution kinetics
measurements suggest HIV-1 protease has two binding sites for
darunavir and amprenavir. J. Med. Chem. 2008, 51, 6599–6603.
(14) Napolitano, E.; Fiaschi, R.; Mastrorilli, E. Halogenative deoxy-
genation of ketones: vinyl bromides and/or gem-dibromides by
cleavage of 1,3-benzodioxoles (ketone phenylene acetals) with
boron tribromide. Synthesis 1986, 122–125.