4-Hydroxy-5,6-dihydropyrones as inhibitors of HIV protease: The effect of heterocyclic substituents at C-6 on antiviral potency and pharmacokinetic parameters

Article abstract of DOI:10.1021/jm0003844

Due largely to the emergence of multi-drug-resistant HIV strains, the development of new HIV protease inhibitors remains a high priority for the pharmaceutical industry. Toward this end, we previously identified a 4-hydroxy-5,6-dihydropyrone lead compound

Full text of DOI:10.1021/jm0003844

J . Med. Chem. 2001, 44, 2319-2332
2319
4-Hyd r oxy-5,6-d ih yd r op yr on es a s In h ibitor s of HIV P r otea se: Th e Effect of
Heter ocyclic Su bstitu en ts a t C-6 on An tivir a l P oten cy a n d P h a r m a cok in etic
P a r a m eter s
Susan E. Hagen,*,† J ohn Domagala,† Christopher Gajda,† Michael Lovdahl,† Bradley D. Tait,† Eric Wise,†
Tod Holler,‡ Donald Hupe,‡ Carolyn Nouhan,‡ Andrej Urumov,‡ Greg Zeikus,‡ Eric Zeikus,‡
Elizabeth A. Lunney,§ Alexander Pavlovsky,§ Stephen J . Gracheck,| J ames Saunders,| Steve VanderRoest,| and
J oanne Brodfuehrer
Departments of Chemistry, Biochemistry, Biomolecular Structure and Drug Design, Infectious Diseases, and
PDM, Pfizer Global Research and Development, 2800 Plymouth Road, Ann Arbor, Michigan 48105
Received September 6, 2000
Due largely to the emergence of multi-drug-resistant HIV strains, the development of new
HIV protease inhibitors remains a high priority for the pharmaceutical industry. Toward this
end, we previously identified a 4-hydroxy-5,6-dihydropyrone lead compound (CI-1029, 1) which
possesses excellent activity against the protease enzyme, good antiviral efficacy in cellular
assays, and promising bioavailability in several animal species. The search for a suitable back-
up candidate centered on the replacement of the aniline moiety at C-6 with an appropriately
substituted heterocyle. In general, this series of heterocyclic inhibitors displayed good activity
(in both enzymatic and cellular tests) and low cellular toxicity; furthermore, several analogues
exhibited improved pharmacokinetic parameters in animal models. The compound with the
best combination of high potency, low toxicity, and favorable bioavailabilty was (S)-3-(2-tert-
butyl-4-hydroxymethyl-5-methyl-phenylsulfanyl)-4-hydroxy-6-isopropyl-6-(2-thiophen-3-yl-
ethyl)-5,6-dihydro-pyran-2-one (13-(S)). This thiophene derivative also exhibited excellent
antiviral efficacy against mutant HIV protease and resistant HIV strains. For these reasons,
compound 13-(S) was chosen for further preclinical evaluation.
It has become fashionable to think of HIV infection
as a chronic condition, much like diabetes, rather than
the inevitably fatal scourge depicted in the lay press
only five years ago. The introduction of potent antiret-
roviral chemotherapies has revolutionized HIV/AIDS
care and dramatically impacted the life span and
lifestyle of the HIV patient. In particular, the combina-
tion of protease inhibitors with reverse transcriptase
inhibitors has become a standard of front-line therapy
in the developed world.1 Despite the hope and promise
these agents have engendered, very real problems exist
with the current antiviral armamentarium, especially
with the protease inhibitors: low bioavailability,2 side
effects such as lipodystrophy3 and toxicity,4 and drug
interactions.5 More significantly, the serious threat
posed by resistant strains of HIV6 necessitates contin-
F igu r e 1.
ued research for novel inhibitors of viral replication.
Indeed, it has been estimated that only half of those
Previous reports from our laboratory10 and others11
patients starting a regimen containing protease inhibi-
have identified the 4-hydroxy-5,6-dihydropyrones as a
tors actually achieve a sustained suppression of the
novel class of non-peptidic protease inhibitors. Optimi-
zation of the enzymatic potency of pyrone I (Figure 1),
identified from a mass screen, was achieved via exten-
sive synthetic modification and concomitant X-ray crys-
tallographic studies. X-ray crystal structures of the
resulting dihydropyrones II bound to the protein (Figure
2) revealed that the enolic hydroxyl group of the
dihydropyrone core forms hydrogen bond(s) with the
aspartate residues at the cleavage site while the lactone
moiety interacts with the Ile residues in the flap region.
Moreover, the groups appended to the 6-position occupy
the S1 and S2 pockets while the phenyl substituents on
virus;7 the outlook is particularly grim for treatment-
experienced patients considering salvage therapy.8,9 In
addition, the high cost of drug has made these regimens
prohibitively expensive for the vast majority of AIDS
sufferers worldwide.
* Corresponding author: Susan Hagen, Pfizer Global Research and
Development, 2800 Plymouth Road, Ann Arbor, MI 48105. Tel: (734)
622-7351. Fax: 734-622-1407. E-mail: Susan.hagen@pfizer.com.
† Department of Chemistry.
‡ Department of Biochemistry.
§ Department of Biomolecular Structure and Drug Design.
| Department of Infectious Diseases.
Department of PDM.
10.1021/jm0003844 CCC: $20.00 © 2001 American Chemical Society
Published on Web 06/08/2001

2320 J ournal of Medicinal Chemistry, 2001, Vol. 44, No. 14
Hagen et al.
mentary approach (also Scheme 1), the halogenated
heterocycle C was reacted with the allylic alcohol in the
presence of palladium acetate and various cofactors.
This particular strategy obviates the need for a reduc-
tion step in that the desired ketone is produced in
moderate to good yield in one step.14 Conversion of these
heterocyclic ketones to the penultimate dihydropyrones
D was accomplished smoothly with little deviation from
the previously reported methods15 and is summarized
in Table 2.
Several analogues containing a methylene alcohol
(CH2OH) in the heterocyclic ring were also prepared.
In general, the alcohol moiety was protected as a tert-
butyl-dimethylsilyl ether during the early steps of the
synthesis and then removed just prior to the ring closure
to dihydropyrone D. Deprotection was easily effected via
treatment with fluoride ion, and subsequent purification
was not necessary.
The majority of the 4-hydroxy-5,6-dihydropyrones
synthesized for this study were racemic mixtures at the
C-6 position. For those compounds of particular interest,
the enantiomerically pure isomers were prepared as
shown in Scheme 2. In this route, the intermediate
ketone B is reacted with the anion of tert-butyl acetate
to give intermediate ester E. Resolution of the mixture
of stereoisomers was accomplished via chromatographic
separation on a Chiralpak column to give esters E-(S)
and E-(R). These esters were hydrolyzed to the requisite
acids (F -(S) and F -(R)) which were then elaborated to
the â-ketoesters. Treatment of â-ketoesters G-(S) and
G-(R) with dilute base gave the desired 4-hydroxy-5,6-
dihydropyrones (D) as before.
F igu r e 2. X-ray crystal structure of prototypical dihydro-
pyrone.
the S-phenyl moiety at C-3 filled the S1′ and S2′ pockets.
Although the prototypic dihydropyrone II displayed
excellent efficacy in enzyme assays, no significant
antiviral activity was observed in cell culture. Attempts
to enhance the antiviral potency by manipulating the
lipophilicity and polarity of the target compounds
resulted in the discovery of dihydropyrone 1 (Figure 1)
This agent, CI-1029, demonstrated excellent enzyme
activity against the HIV protease, but more importantly
displayed an EC50 of 200 nM with a therapeutic index
of >1000. In addition, CI-1029 also exhibited good
pharmacokinetic parameters, favorable activity against
mutant HIV PR and against resistant HIV strains, and
little interaction with cytochrome P-450 (30% inhibition
at 100 µM for 3A4)sall characteristics of a promising
preclinical candidate.
With one development candidate in hand, efforts were
concentrated on the discovery of an appropriate second-
generation agent. Ideally, this back-up candidate would
display enhanced cellular potency and improved phar-
macokinetic parameters when compared to CI-1029,
while retaining a favorable side effect profile. Of par-
ticular importance was the identification of a new agent
without the aniline moiety at C-6, since N-acetylation
of the aniline group presents a likely avenue for
metabolic instability. Toward this end, a series of
dihydropyrones was synthesized in which a heterocyclic
group was introduced into the S2 pocket via appropriate
substitution at the C-6 position. In this paper we will
discuss the synthesis, antiviral activity, and preliminary
pharmacokinetics for this series and the discovery of
another preclinical candidate, derivative 13-(S).
The requisite thiotosylate contains the benzyl alcohol
moiety in the phenyl ring at C-3. Its preparation from
the phenol has been described in our earlier reports.12,15
Reaction of the dihydropyrone nucleus with excess
thiotosylate and several equivalents of potassium car-
bonate in DMF gave, after purification, target com-
pounds 2-35 (Table 3.)
Biologica l Meth od s
The compounds in this study were tested for their
inhibition of HIV protease using an HPLC assay at pH
6.2 as reported previously16 and as outlined in the
Experimental Section. Cellular anti-HIV activity was
measured at Southern Research Institute in an assay
using HIV-IIIB-infected human lymphocyte-derived
CEM cells.17 EC50 refers to the concentration of drug at
which 50% of the cells are protected against the cyto-
pathic effects of HIV; TC50 refers to the concentration
of drug which elicits cytotoxicity in 50% of CEM cells
uninfected with HIV. Enzymatic Ki’s and cellular EC50’s
and TC50’s are reported in Table 3.
Ch em istr y
The synthesis of the target compounds requires the
preparation of the thiotosylate side chain12 and the
dihydropyrone nucleus,10 both of which have been
reported previously. In general, the dihydropyrone core
was prepared from the requisite ketone and the dianion
of methyl acetoacetate. These heterocyclic ketones B
(listed in Table 1) were synthesized via two complemen-
tary routes as summarized in Scheme 1. In the first
approach, the starting aldehyde was condensed with a
triphenyl arsenate salt, prepared from the reaction of
1-chloro-3-methyl-2-butanone and triphenyl arsine,13 to
give the enone A; alternatively, the aldehyde could be
reacted with 3-methyl-2-butanone and barium hydrox-
ide under standard Claisen-Schmidt conditions. In
either case, the resulting enones A were then reduced
catalytically to give the target ketones B. In a comple-
Resu lts a n d Discu ssion
CI-1029 (compound 1 in Table 3) possessed all the
desirable attributes of a promising preclinical candidate
but also a potentially serious liability. The amino group
in the phenethyl chain at C-6 presents a likely site for
metabolism and unwanted derivatization. Indeed, meta-
bolic profiling of this compound in several species
revealed that 20-40% of the compound was acylated
(unpublished results.) Therefore, one of the primary
goals in the search for a second-generation agent was

4-Hydroxy-5,6-dihydropyrones as HIV Protease Inhibitors
J ournal of Medicinal Chemistry, 2001, Vol. 44, No. 14 2321
Ta ble 1. Analytical Data for Intermediate Ketones B

2322 J ournal of Medicinal Chemistry, 2001, Vol. 44, No. 14
Ta ble 1 (Continued)
Hagen et al.

4-Hydroxy-5,6-dihydropyrones as HIV Protease Inhibitors
J ournal of Medicinal Chemistry, 2001, Vol. 44, No. 14 2323
Ta ble 1 (Continued)
a
Comm. refers to commercially available material; Expt. denotes that the preparation of the starting material can be found in the
b
Experimental Section. Letter refers to the general method found in the Experimental Section, Figure 1.
Sch em e 1. Preparation of Racemic Dihydropyronesa
a
(a) AsPh3CH2COCH(Me)2, K2CO3, aq CH3CN; (b) CH3COCH(Me)2, Ba(OH)2; (c) H2, Pd/BaSO4; (d) NaI, NaHCO3, PPh3, Pd(OAc)2,
DMF; (e) (nBu)4NCl, NaHCO3, pyrrolidine, DMF, Pd(OAc)2; (f) NaH; (g) nBuLi; (h) OH-, then H+; (i) thiotosylate, K2CO3, DMF.
the replacement of the aniline moiety with a more
metabolically stable functionality.
position has also been shown to confer good activity.
However, since our intent was to remove any potential
metabolic liability associated with an aniline group,
introduction of yet another aniline into the 3-position
would be undesirable. The current lead compound,
compound 1 (CI-1029), was used as a benchmark for all
potential back-up candidates. Of note is the chirality
issue: 1 exists as a single enantiomer whereas all of
the newly synthesized comparitors were tested as ra-
cemic mixtures. For the sake of comparison, data for
racemic 1 (designated as 1-(()) are also included in
Table 3.
Compounds 2 through 8 contain a nitrogen hetero-
cycle (pyridine, pyrimidine, imidazole, and pyrazole) in
the region previously occupied by the aminophenyl
group. In the case of pyrazole 7, the enzyme activity
remained essentially equipotent with that of racemic
1-((); in all other cases, the potency of the nitrogen
analogues decreased by a least 2-fold (and by as much
as 700-fold in the case of 8). More important was the
significant loss in antiviral potency: although pyridine
analogues 2 (EC50 ) 1.5 µM), 3 (EC50 ) 0.9 µM), and 4
(EC50 ) 1.5 µM) and pyrimidine 5 (EC50 ) 1.2 µM) were
modestly active, several derivatives showed more dra-
Previous SAR studies and X-ray analyses indicated
that the S2 pocket of the protease enzyme could tolerate
a wide variety of substituents without subsequent loss
of enzyme potency. Indeed, past increases in enzyme
potency did not necessarily correlate with increased
antiviral potency until substantial changes in polarity
and lipophilicity were introduced.15,18 For this reason,
we theorized that an appropriate heterocyclic replace-
ment of the 4-aminophenyl group at S2 could effect the
same change of polarity and/or lipophilicity without the
use of an aniline. In other words, we hoped to modulate
the physical properties of these analogues without
affecting the enzyme potency or the antiviral efficacy
already achieved by CI-1029.
Therefore, a series of dihydropyrones containing a
heterocycle in the S2 pocket was synthesized and is
summarized in Table 3. For these analogues, the
substitution at C-3 was restricted to the 2′-tert-butyl-
4′-(hydroxymethyl)-5′-methylphenyl moiety previously
used for lead compound CI-1029. Prior work in our
laboratory had established the superiority of this par-
ticular group at C-3, although an amino group at the 4′

2324 J ournal of Medicinal Chemistry, 2001, Vol. 44, No. 14
Ta ble 2. Analytical Data for Intermediate Dihydropyrones D
Hagen et al.
a
Yield calculated from the appropriate ketone precursor in Table 1; includes protection/deprotection steps where necessary. b Trituration
(trit) refers to grinding of the solids under solvent to produce a fine powder. c Another 25% of the N-trityl product (mp 66-70 °C) was also
obtained.

4-Hydroxy-5,6-dihydropyrones as HIV Protease Inhibitors
J ournal of Medicinal Chemistry, 2001, Vol. 44, No. 14 2325
Sch em e 2. Preparation of Chiral Dihydropyronesa
general, these thiazoles were less toxic (TC50 > 200 µM)
than the thiophene analogues and similar to the original
lead.
Prior experience in the SAR of the 6-position sug-
gested that further elaboration of the heterocycle ring
might prove advantageous in fully occupying the S2
pocket (results not shown.) For this reason, a number
of thiophene and thiazole derivatives were made in
which alkyl groups were appended to various positions
of the heterocycle. In the case of the thiophene, these
new methylated derivatives (17 and 18) were essentially
equipotent to the parent unmethylated compounds;
however, in both cases the TC50 dropped significantly.
Again, the presence of a single methyl group in the
thiazole series engendered a more dramatic effect on
the antiviral potency than it did in the thiophene series.
The EC50’s of the five various methylthiazole isomers
varied from 0.37 to 4.5 with a similar range of enzymatic
Ki’s, all dependent on the positions of the ethylene linker
and the methyl group. The most potent of these thiazole
derivatives, the 4-methyl compound 26, possessed an
EC50 of 0.37 µM and an excellent TC50 of 164 µM. For
this reason, the 4-methyl-5-thiazole derivative emerged
as a potential back-up to CI-1029 along with the simple
thiophene analogues mentioned earlier.
Several additional substituted thiazole derivatives
were also synthesized. Enlarging the alkyl group found
in the methylthiazole series to an isopropyl moiety
resulted in compounds 27-30. Unfortunately the larger
group proved deleterious to cellular activity, since the
isopropyl analogues were much less active and more
toxic than the methyl comparitors; apparently, the steric
requirements of the S2 pocket do not allow for such a
large hydrophilic group. Also of note was the influence
of a 2-amino group on the activity of the thiazole
analogue. While the 2-acetylamino analogue 31 (EC50
) 23 µM) displayed a 10-fold decrease in antiviral
potency when compared to the 2-hydrogen parent 15
(EC50 ) 2.8 µM), the 2-amino derivative 32 (EC50 ) 3.5
µM) was essentially equipotent with the unsubstituted
parent. More importantly, the presence of the amino
group did not engender any appreciable improvement
in pharmacokinetic parameters (results not shown.)
a
(a) LDA, tert-butyl acetate; (b) separation via Chiralpak
AD; (c) NaOH; (d) CDI, then Mg(O2CCH2CO2Et)2; (e) OH-,
then H+.
matic 100-fold losses in activity (for example, imidazole
8, EC50 ) 89 µM). Overall, none of these nitrogen
heterocycles were equipotent with parent compound 1
although all derivatives displayed low toxicity.
Simple heterocyclesssuch as furan (9-10), tetrahy-
drofuran (11), thiophene (12-13), and thiazole (14-
16)swere also prepared, with more encouraging results.
With one exception (compound 11), these analogues
displayed excellent activity against the HIV-1 protease
enzyme, with a Ki of 1 nM or lower. Several compounds
also showed encouraging antiviral activity: thiophene
12 (EC50 ) 0.50 µM) and its regioisomer 13 (EC50 ) 0.30
µM) were of particular promise. Both derivatives dis-
played good antiviral activity albeit with a slight
increase in toxicity (TC50 ) 81 µM and 110 µM,
respectively) when compared to CI-1029 and its racemic
parent. It is interesting to note that regiochemistry
exerted little influence over the antiviral activity in both
the furan and the thiophene series, in that the activity
of the 2-furan 9 (EC50 ) 0.75 µM) did not vary signifi-
cantly from the 3-furan 10 (EC50 ) 1.0 µM). However,
greater variation in activity was observed for the
Earlier work in our laboratories identified the benzyl
alcohol moiety as a group that imparted good activity
when placed on the phenyl group at C-3. When this
same benzyl alcohol group was affixed to a phenyl ring
at C-6susually ortho to the ethylene linkersa boost in
activity was also seen, although not of the magnitude
observed for the 3-position isomer (data not shown). For
this reason, a series of thiophene analogues (compounds
19-21, Table 3) was prepared in which a hydroxy-
methyl substituent was appended to various positions
of the thiophene ring. The most potent of these hy-
droxymethyl thiophenes was compound 20, with an
EC50 of 0.92 µM and a TC50 of 200 µM, but in general
these analogues showed no advantages over the unsub-
stituted thophenes. It is interesting to note that the
analogue with the best potency (compound 20) contains
the hydroxymethyl adjacent to the ethylene spacer -
the same orientation seen in the series containing a
phenethyl group at C-6.
thiazole regioisomers in which the 2-isomer 14 (EC50
)
0.7 µM) was two times more active than the 5-isomer
16 (EC50 ) 1.9 µM) in the cellular assay and 4 times
more active than the 4-isomer 15 (EC50 ) 2.8 µM). In
A limited number of fused heterocyclic analoguess
namely indoles and benzthiazolesswere also prepared.

2326 J ournal of Medicinal Chemistry, 2001, Vol. 44, No. 14
Hagen et al.
Ta ble 3. Chemical and Biological Data for 4-Hydroxy-1,4-dihydropyrones

4-Hydroxy-5,6-dihydropyrones as HIV Protease Inhibitors
J ournal of Medicinal Chemistry, 2001, Vol. 44, No. 14 2327
Ta ble 3 (Continued)

2328 J ournal of Medicinal Chemistry, 2001, Vol. 44, No. 14
Ta ble 3 (Continued)
Hagen et al.
a
b
All compounds were analyzed for C, H, and N and had results +0.4% of theoretical values. Enzyme inhibition was determined as
described in the Experimental Section. c Antiviral activity in HIV infected CEM cells. EC50 refers to the effective concentration at which
50% of the cells are protected from cytopathic effects. Toxicity measured in CEM cells in the absence of virus.
d
Ta ble 4. Pharmacokinetic Data for Heterocyclic
Ta ble 5. Pharmacokinetic Parameters for Heterocyclic
Dihydropyrones in Micea
Dihydropyrones in Ratsa
cmpd
Cmax (µM)
Tmax (h)
T1/2 (h)
AUC (µM h/mL)
cmpd
Cmax (µM/mL)
Tmax (h)
T1/2 (h)
AUC (µg h/mL)
1-(S)
1(()
3
10
13
26.5
23
45.8
54.4
39.6
33
0.5
0.5
0.5
0.5
0.5
0.5
2.2
2.2
7.3
5.6
3.5
3.3
82
42
143
241
221
132
1-(S)
13
13-(S)
14
7.2
11.2
14.7
2.7
0.5
0.5
0.5
0.25
0.25
3.7
9.4
11.5
12.2
ND
11.4
21
22.5
10
26
0.77
ND
14
a
All compounds were dosed at 10 mg/kg using a capsule or a
solution buffered to pH 7.4. Values are an average of two
animals.
a
All compounds were dosed at 25 mg/kg in a vehicle consisting
of 20% 0.1 N NaOH/80% 0.5% methylcellulose. Values are from a
pool of five mice.
with an especially dramatic improvement in Cmax (0.77
µM for thiazole 26 versus 11.2 µM for thiophene 13.)
Interesting to note is the effect of a single methyl group
on the activity of thiazole derivatives: 2-thiazole deriva-
tive 14 has a Cmax that is 3.5 times higher than that of
the 4-methyl-5-thiazole 26. Thiophene 13 also exhibits
a significant improvement in pharmacokinetics when
compared to CI-1029 (compound 1). Most notable were
the changes in T1/2 (9.4 for 13 vs 3.7 for 1) and AUC.
This enhancement in pharmacokinetic properties, along
with its excellent antiviral efficacy and low toxicity, led
us to choose thiophene 13 for further study.
The enantiomers of racemate 13 were synthesized as
outlined in Scheme 2; the activities of the individual
stereoisomers are listed in Table 3. As expected from
analogy with previous series, the more active isomer
possesses the S stereochemistry: compound 13-(S) was
100 times more potent in both the HIV-protease assay
and the cellular antiviral screen than its enantiomer
13-(R). The more active isomer was assayed for phar-
macokinetic properties in the rat model and, as ex-
pected, demonstrated comparable properties to that of
the racemate 13. However, the supremacy of the chiral
thiophene 13-(S) over the prior lead compound can best
be seen in the dog model, for which the data is
summarized in Table 6. By every measure, the optically
pure thiophene displayed superior pharmacokinetic
properties with an increased half-life, excellent bioavail-
ability, and an encouraging time above the EC90.
While all these derivatives possessed moderate to good
activity (for example, indole 34 with an EC50 of 0.37 µM),
they also displayed increased toxicities of 40-60 µM.
The racemic analogue of our lead compound, CI-1029,
exhibited excellent cellular potency (EC50 ) 0.5 µM) and
low toxicity in cell culture (TC50 > 100 µM.)17 Using
these numbers as a benchmark, several derivatives
containing a heterocycle at C-6 also displayed excellent
enzyme activities, good antiviral efficacy, and low toxic-
ity. These analoguessboth racemicsincluded the 3-thio-
phene 13 (EC50 ) 0.3 µM, TC50 ) 110 µM) and the
4-methyl-5-thiazole 26 (EC50 ) 0.37 µM, TC50 )164 µM).
Although other analogues also possessed excellent anti-
viral activity (for example, thiophene 17 and indole 34),
these compounds also exhibited unacceptable toxicity
(TC50 ) 80 µM and 46 µM, respectively.) Therefore, fur-
ther assessment of analogues 13 and 26 was warranted.
Pharmacokinetic parameters for a variety of hetero-
cyclic derivatives (including thiophene 13 and thiazole
26) were measured in mice and rats and were compared
to the racemic and chiral lead CI-1029, where possible.
Mice were dosed po at 25 mg/kg using a vehicle of 20%
0.1 N NaOH in 0.5% methylcellulose, and plasma levels
were measured via an HPLC assay. Similarly, rats were
dosed po at 10 mg/kg using a capsule or solution
buffered to pH 7.4. The results are summarized in
Tables 4 (mice) and 5 (rats).When assayed in mice, the
heterocyclic derivatives all displayed improvements in
C
max and a concomitant improvement in AUC. Although
Further preclinical profiling of thiophene 13-(S) was
undertaken by measuring cytochrome P-450 inhibition
and antiviral activity in the presence of serum. Incuba-
tion of compound 1-(S) with human liver microsomes
and a series of major P-450 isoenzyme selective probes
revealed no inhibition of CYP3A4 at 1 and 10 µM and
pyridine 3 and furan 10 showed promising improve-
ments in half-life, most of the new analogues were
essentially equivalent in the mouse model.
In the rat, however, the thiophene compound 13
appears substantially better than the other comparitors,

4-Hydroxy-5,6-dihydropyrones as HIV Protease Inhibitors
J ournal of Medicinal Chemistry, 2001, Vol. 44, No. 14 2329
Ta ble 6. Pharmacokinetic Parameters for Selected
at the 6-position. The chiral form (13-(S)) was then
synthesized and assayed for potency against the HIV
protease enzyme, antiviral activity in cell culture, and
pharmacokinetic parameters in mice, rats, and dogs. In
all cases, the thiophene compound proved superior to
the prior lead compound, displaying excellent antiviral
efficacy (EC50 ) 0.12 µM), low toxicity (TC50 ) 92 µM),
and promising bioavailablity across all species tested.
Also encouraging was the activity against protease-
resistant strains of HIV. The ease of synthesis, low
molecular weight, and presence of a single chiral center
of this agent also bode well for future success. For these
reasons, compound 13-(S) was chosen for further pre-
clinical studies.
Dihydropyrones in Dogsa
Cmax
AUC
hours above bioavailability
(µg/mL) T1/2 (µg h/mL)
the EC90
(%)
1-(S)
13-(S)
77
152
1.2
5.4
113
592
4
9
33
88
a
Animals were dosed at 10mg/kg, and the vehicle was the
sodium salt dissolved in normal saline for 1 and free acid dissolved
in 0.1 NaOH (buffered to pH 10) for 13-(S). Values are an average
from two animals.
Ta ble 7. Antiviral Activity of Selected Dihydropyrones against
HIV Protease-Resistant Strains
fold increase in EC50 for
inhibitors
protease amino acid substitutions
IDVa RTVb 1-(S) 13-(S)
Exp er im en ta l Section
V31I, L10I, I15V, K20R, M36I,
S37N, R41K, G48V, I54T/I, L63T,
A71V, T74A, V82A/V, 193L
V3I, L10I, K20R, E35D, M36I,
S37D, R41K, G48V, L63C, A71V,
I72T, V82A/V, 193L
V31, L10I, L19I, S37Q, M46L,
I54V, R57K, L63P, A71V, V82A,
L90M
V3I, L10I, I15V, K20R, E35D,
M36I, S37K, R41N, K43T/K, M46I,
L63P, H69K, A71V, T74S, V82F,
N88E, L89M, L90M, I93L
V3I, L10I, 115V, K20I, L24I,
M36I, S37N, I54V, R57K, L63P,
A71V, V82A
19
3
38
9
1
1
2
1
2
2
En zym e In h ibition Assa y. For determination of IC50
values, affinity-purified HIV-1 protease (Bachem Bioscience;
1.1 nM) was added to a solution of the following: the inhibitor,
40 µM peptide substrate (His-Lys-Ala-Arg-Val-Leu-(p-NO2-
Phe)-Glu-Ala-Nle-Ser-NH2; Bachem Bioscience), and 1.0%
DMSO in assay buffer (1.0 mM dithiothreitol, 80 mM MES,
160 mM NaCl, 1.0 mM EDTA, 0. 1% poly(ethylene glycol) (MW
8000), pH 6.2 at 25 °C). The final volume was 100 µL; final
concentration of HIV protease was 1.5 nM. The solution was
mixed, incubated for 60 min at 37 °C, and then quenched
with trifluoroacetic acid (2% final). In this assay, the Leu-(p-
NO2-Phe) bond of the substrate was cleaved by the enzyme,
and the substrate and cleavage products were separated by
reverse-phase HPLC. Absorbance was measured at 220 nm,
peak areas were determined, and percent conversion to product
was used to calculate percent control () [%conversion (+in-
hibitor)/%conversion (-inhibitor)] × 100).
5
39
23
1
40
<1
17
44
<1
<1
a
b
Indinavir. Ritonavir.
very little inhibition (30%) at 100 µM. Results for
thiophene 13-(S) were strikingly similar: no inhibition
at 1 and 10 µM with slight inhibition (34%) at 100 µM.
This dearth of activity even at high concentrations bodes
well for 13-(S) in the area of cytochrome inhibition and
potential drug interactions. In addition, dihydropyrones
1-(S) and 13-(S) were assayed for antiviral activity in
the presence of 30% human serum in order to ascertain
the effect of protein binding on efficacy. Both analogues
did lose activity such that the two were essentially
equipotent in the presence of serum (EC50 for 1-(S) )
6.31 µM and EC50 for 13-(S) ) 10.4 µM.)
The problems stemming from drug-resistant strains
of HIV have emerged as the most serious dilemma
facing clinicians and patients today. To help ameliorate
this problem, many pharmaceutical companies are
developing structurally dissimilar protease inhibitors in
the hopes of producing dissimilar resistance patterns.
Toward this goal, dihydropyrones 1-(S) and 13-(S) were
assayed against a panel of protease inhibitor-resistant
HIV strains. The antiviral potency was measured in an
in vitro assay in HIV-infected PBMC cells, and the
results are summarized in Table 7; the data are reported
as fold increases in activity. In general, both CI-1029
(1-(S)) and thiophene 13-(S) displayed excellent activi-
ties against strains of virus already resistant to the
clinical agents indinavir and ritonavir. In all cases the
dihydropyrone derivatives show a reduction in suscep-
tibility that is 2-fold or less.
Ch em istr y. Melting points (mp) were obtained on a Thomas-
Hoover capillary melting point apparatus and are uncorrected.
Proton magnetic resonance spectra (1H NMR) were obtained
on a Varian Unity 400 MHz NMR using TMS as an internal
standard. Chemical shifts are reported in ppm. Mass spectral
data were obtained on a Micromass PlatformLC spectropho-
tometer. Elemental analyses were performed by Robertson
Labs and are within 0.4% of theoretical values unless other-
wise indicated. Column chromatography was performed on
silica gel 60, 230-400 mesh, purchased from Mallinckrodt.
Solutions were dried over magnesium sulfate. Yields are of
purified product (except where noted), and reaction conditions
were not optimized.
P r epar ation of Star tin g Mater ials. (5-Br om o-th ioph en -
2-yl)-m eth a n ol. A solution of 5-bromo-2-thiophenecarbalde-
hyde (26.2 g, 137 mmol) in MeOH (500 mL) was treated with
NaBH4 (5.2 g, 137 mmol). The reaction was stirred for 2 h at
0 °C and 2 h at room temperature. The MeOH was evaporated,
and saturated NH4Cl was added followed by 2 N HCl. The
aqueous layers were extracted with EtOAc, dried, and con-
centrated. Flash chromatography using 100% CH2Cl2 as eluent
afforded 23.3 g (92%) of the title compound. 1H NMR (CDCl3):
δ 4.74 (d, 2 H), 6.75 (m, 1 H), 6.91 (d, 1 H).
(3-Br om o-t h iop h en -2-yl)-m et h a n ol. A mixture of 3-
bromothiophene-2-carboxylic acid methyl ester (9.9 g, 45
mmol), LAH (1.7 g, 45 mmol), and THF (150 mL) was stirred
at 0 °C for 1 h and then overnight at room temperature. The
reaction was worked up by addition of 1 mL of H2O, 1 mL of
15% NaOH, and 3 mL of H2O followed by filtration through
Celite. Concentration of the filtrate gave 6.5 g (75%) of the
title compound. 1H NMR (CDCl3): δ 4.80 (s, 2 H), 6.96 (d, 1
H), 7.26 (d, 1 H).
2-Meth yl-th ia zole-4-ca r ba ld eh yd e. A mixture of 5.52 g
(30 mmol) of 4-chloromethyl-2-methylthiazole hydrochloride
and 0.5 M NaOH (180 mL) was refluxed for 8 h and then
stirred for 18 h at room temperature. EtOAc and H2O were
added. The organic phase was separated, washed with brine,
and dried. Concentration gave 2.2 g (57%) of 4-(hydroxy-
methyl)-2-methylthiazole which was clean enough to carry on
In conclusion, a series of dihydropyrones was synthe-
sized in which a heterocycle was substituted for the
aminophenyl substituent found in lead compound CI-
1029. From this series, a new preclinical candidate,
compound 13, was identified which contains a thiophene

2330 J ournal of Medicinal Chemistry, 2001, Vol. 44, No. 14
Hagen et al.
1
without purification. Yield: 2.2 g (57%). H NMR (CDCl3): δ
were filtered, and the filtrate was concentrated. The residue
was taken up in EtOAc and filtered; the filtrate was chro-
matographed, eluting with 1:2 EtOAc:hexane, to give 4-methyl-
1-(4-methyl-thiazol-2-yl)-pent-1-en-3-one (2.5 g, 81%). 1H NMR
(CDCl3): δ 1.13 (d, J ) 6.8. Hz, 6 H), 2.45 (s, 3 H), 2.82-2.89
(m, 1 H), 6.96 (s, 1 H), 7.01 (d, J ) 16 Hz, 1 H), 7.57 (d, J )
16 Hz, 1 H).
The enone was then dissolved in THF (100 mL), treated with
0.4 g of 5%Pd/BaSO4, and shaken in a hydrogen atmosphere
of 32 psi for 26 h. The catalyst was filtered, and the filtrate
was concentrated. The crude product was chromatographed,
eluting with 2:1 hexane:EtOAc, to give 1.3 g (55%) of the title
ketone. 1H NMR (CDCl3): δ 1.05 (d, J ) 6.8. Hz, 6 H), 2.34 (s,
3 H), 2.55-2.62 (m, 1 H), 2.93 (t, J ) 7.1 Hz, 2 H), 3.18 (t, J
) 7.2 Hz, 2 H), 6.65 (s, 1 H).
Ketones B-1, B-3, B-6, B-14, B-15, B-16, B-26, B-27, B-28,
B-29, B-30, B-31, B-32, B-33, B-34, B-36, B-37, and B-38 were
prepared in similar fashion from the appropriate starting
aldehyde as listed in Table 1. The yields reported are over two
steps, including condensation and subsequent catalytic reduc-
tion.
Gen er a l Meth od B. Syn th esis of 1-F u r a n -2-yl-4-m eth -
yl-p en ta n -3-on e (B-10). To a reaction flask was added
2-furancarbaldehyde (11.3 g, 117 mmol), 3-methyl-2-butanone
(10.1 g, 117 mmol), 95% EtOH (200 mL), and anhydrous Ba-
(OH)2 (2.2 g). The reaction was heated at 80 °C for 3 h and
stirred at room temperature overnight. The EtOH was evapo-
rated, and the residue was partitioned between EtOAc and 1
N HCl. The combined organic extract was dried and concen-
trated; the crude product was purified by flash chromatogra-
phy, eluting with 50:50 hexane:CH2Cl2, to give 1-furan-2-yl-
4-methyl-pent-1-en-3-one (6.1 g, 32%). 1H NMR (CDCl3): δ 1.15
(d, J ) 7.1 Hz, 6 H), 2.82-2.90 (m, 1 H), 6.45 (m, 1 H), 6.65
(m, 1 H), 6.72 (d, J ) 14 Hz, 1 H), 7.40 (d, J ) 14 Hz, 1 H),
7.52 (m, 1 H).
The titled compound was prepared from the reduction of
the enone (1.0 g, 6.2 mmol) with (Ph3P)3RhCl (0.1 g) in THF
(50 mL) under hydrogenation conditions similar to those used
in general method A. The crude product was chromatographed,
eluting with 98:2 hexane:EtOAc, to give 0.83 g (82%; 26% over
two steps) of the title ketone. 1H NMR (CDCl3): δ 1.09 (d, J )
7.1 Hz, 6 H), 2.57-2.64 (m, 1 H), 2.80 (m, 2 H), 2.85 (m, 2 H),
5.99 (m, 1 H), 6.26 (m, 2 H), 7.28(m, 1 H).
Gen er a l Meth od C. Syn th esis of 4-Meth yl-l-p yr id in -3-
yl-p en ta n -3-on e (B-2). 3-Iodo-pyridine (7.50 g, 36.6 mmol),
4-methyl-1-penten-3-ol (5.45 g, 54.9 mmol), tetrabutylammo-
nium chloride (10.2 g, 36.6 mmol), NaHCO3 (6.15 g, 73.2
mmol), pyrrolidine (2 mL), DMF (25 mL), and Pd(OAc)2 (0.01
equiv) were added to a reaction vessel.14 The solution was
heated to 80 °C for 18 h and then cooled to room temperature.
The mixture was diluted with H2O and CH2Cl2 and filtered
through Celite. The organic extract was washed with brine,
dried, and concentrated. The title compound was flash chro-
matographed, eluting with 98:2 CH2Cl2:MeOH, to give 5.15 g
(79%) of the title ketone. 1H NMR (CDCl3): δ 1.15 (d, 6 H),
2.5-2.7 (m, 1 H), 2.7-2.9 (m, 2 H), 2.85-2.95 (m, 2 H), 7.2-
7.3 (m, 1 H), 7.5-7.6 (m, 1 H), 8.4-8.5 (m, 2 H).
2.63 (s, 3 H), 4.66 (s, 2 H), 6.96 (s, 1 H). The crude product
(2.2 g, 17 mmol) was oxidized with MnO2 (20 g) in 75 mL of
CHCl3 for 16 h. The suspension was filtered, and the filtrate
was concentrated. The residue was chromatographed on silica
gel, eluting with 2:1 hexane:EtOAc, to give 1.4 g (36% over
two steps) of the title compound as an oil which solidified upon
1
standing. H NMR (CDCl3): δ 2.71 (s, 3 H), 7.98 (s, 1 H), 9.89
(s, 1 H).
5-Meth yl-th ia zole-4-ca r ba ld eh yd e. A solution of 2.48 g
(14.4 mmol) of 5-methyl-thiazole-4-carboxylic acid ethyl ester19
in CH2Cl2 (50 mL) was cooled to -78 °C under nitrogen.
DIBAL (1.0 M in CH2Cl2, 15 mmol) was added dropwise, and
the solution was stirred at low temperature for 45 min.
Another 10 mmol of DIBAL was then added dropwise, and the
mixture was stirred for another 45 min. A solution of MeOH:
CH3CO2H (10 mL:5 mL) was added slowly, followed by H2O.
The organic layer was separated, washed with brine, and
dried. Concentration gave a residue which was chromato-
graphed, eluting with 2:1 hexane:EtOAc, to give 1.02 g (56%)
1
of the title compound. H NMR (CDCl3): δ 2.76 (s, 3 H), 8.58
(s, 1 H), 10.14 (s, 1 H).
4-Isop r op yl-t h ia zole-2-ca r b a ld eh yd e. The title com-
pound was prepared from 2.1 g (11 mmol) of 4-isopropyl-
thiazole-2-carboxylic acid ethyl ester20 and DIBAL (a total of
19 mmol of 1.0 M in CH2Cl2) in a fashion similar to the
preparation of 5-methyl-thiazole-4-carbaldehyde. Chromatog-
raphy of the crude product, eluting with 3:1 hexane:EtOAc,
gave the title compound in 51% yield. 1H NMR (CDCl3): δ 1.35
(d, 6 H), 3.16-3.23 (m, 1 H), 7.33 (s, 1 H), 9.96 (s, 1 H).
2-Isop r op yl-t h ia zole-4-ca r b a ld eh yd e. The title com-
pound was prepared from 2.7 g (13 mmol) of 2-isopropyl-
thiazole-4-carboxylic acid ethyl ester20 and DIBAL (a total of
30 mmol of 1.0 M in CH2Cl2) in a fashion similar to the
preparation of 5-methyl-thiazole-4-carbaldehyde. Chromatog-
raphy of the crude product, eluting with 3:1 hexane:EtOAc,
gave 1.45 g (69%) of the title compound as a clear colorless
oil. 1H NMR (CDCl3): δ 1.37 (d, 6 H), 3.30-3.35 (m, 1 H), 8.01
(s, 1 H), 9.93 (s, 1 H).
5-Isop r op yl-t h ia zole-4-ca r b a ld eh yd e. The title com-
pound was prepared from 6.8 g (37 mmol) of 5-isopropyl-
thiazole-4-carboxylic acid methyl ester19 in CH2Cl2 (120 mL)
and DIBAL (a total of 57 mmol of 1.0 M in CH2Cl2) in a fashion
similar to the preparation of 5-methyl-thiazole-4-carbaldehyde.
Chromatography of the crude product, eluting with 2:1 hexane:
EtOAc, gave 4.65 g (81%) of the title compound. 1H NMR
(CDCl3): δ 1.34 (d, 6 H), 4.09-4.16 (m, 1 H), 8.63 (s, 1 H),
10.20 (s, 1 H).
4-Isop r op yl-th ia zole-5-ca r ba ld eh yd e. A solution of 5.9
g (30 mmol) of 4-isopropyl-thiazole-5-carboxylic acid ethyl
ester19 in toluene (100 mL) was cooled in an ice bath under
nitrogen and treated dropwise with DIBAL (150 mL of 1.0 M;
150 mmol). The mixture was stirred at low temperature for
45 min and then allowed to warm to room temperature
overnight. H2O was added cautiously, and the mixture was
extracted with EtOAc (3 × 200 mL). The combined extracts
were washed with brine, dried, and concentrated. The crude
5-(hydroxymethyl)-4-isopropyl-thiazole thus obtained was used
as is in the next step. 1H NMR (CDCl3): δ 1.30 (d, 6 H), 3.13-
3.20 (m, 1 H), 4.85 (s, 2 H), 8.67 (s, 1 H).
Similar procedures were used to prepare ketones B-4, B-5,
B-8, and B-9 from the corresponding bromides; the results are
summarized in Table 1.
A mixture of 4.65 g (30 mmol) of crude 5-(hydroxymethyl)-
4-isopropyl-thiazole in 200 mL of CDCl3 was treated with 45
g of MnO2 and stirred at room temperature for 2.5 h. The
suspension was filtered, and the filtrate was concentrated. The
residue was chromatographed, eluting with 3:1 hexane:EtOAc,
to give 2.35 g (51% over 2 steps) of the title compound. 1H NMR
(CDCl3): δ 1.40 (d, 6 H), 3.62-3.69 (m, 1 H), 8.97 (s, 1 H),
10.17 (s, 1H).
P r ep a r a tion of Keton es. Gen er a l Meth od A. Syn th esis
of 4-Meth yl-1-(4-m eth yl-th ia zol-2-yl)-p en ta n -3-on e (B-25).
A mixture of 2.0 g (16 mmol) of 4-methyl-thiazole-2-carbalde-
hyde,24 8.9 g (19 mmol) of 3-methyl-2-oxobutyltriphenylarso-
nium bromide,13 2.6 g (19 mmol) of K2CO3, and 1% H2O in
CH3CN was stirred for 18 h at room temperature. The solids
Gen er a l Meth od D. Syn th esis of 4-Meth yl-1-th iop h en -
2-yl-p en ta n -3-on e (B-12). A solution of 2-bromothiophene
(14.7 g, 100 mmol), 4-methyl-1-penten-3-ol (15.0 g, 150
mmol), NaI (0.52 g, 3.5 mmol), NaHCO3 (10.1 g, 120 mmol),
PPh3 (0.78 g, 3.0 mmol), Pd(OAc)2 (0.22 g, 1.0 mmol), and DMF
(0.1-1 mL per mmol of halide) was heated to 130 °C for 20
h.14 The reaction was cooled to room temperature and parti-
tioned between H2O and ether. The organic extract was
washed with brine, dried, and concentrated. The residue was
chromatographed, eluting with hexane:EtOAc 95:5, to give
4.3 g (24%) of the title compound. 1H NMR (CDCl3): δ 1.09 (d,
6 H), 2.54-2.64 (m, 1 H), 2.83 (t, J ) 7.4 Hz, 2 H), 3.11 (t, J
) 7.4 Hz, 2 H), 6.79 (m, 1 H), 6.88-6.92 (m, 1 H), 7.11 (m, 1
H).

4-Hydroxy-5,6-dihydropyrones as HIV Protease Inhibitors
J ournal of Medicinal Chemistry, 2001, Vol. 44, No. 14 2331
Ketones B-13, B-17, B-18, B-19, B-21, and B-23 were
synthesized in a similar fashion from the appropriate halide,
and the results are summarized in Table 1.
of THF, treated with 200 mL of 1.0 N NaOH, and stirred for
1 h. The solution was washed with Et2O. The aqueous phase
was cooled in an ice bath, acidified to pH 4.5 with 1.0 N HCl,
and extracted with EtOAc. The extract was washed with brine,
dried, and concentrated. Chromatography of this residue,
eluting with 95:5 CH2Cl2:MeOH, gave 1.57 g (72%) of the title
Gen er a l Meth od E. Silyla tion of In ter m ed ia te Hy-
d r oxym et h yl Ket on es. Syn t h esis of 1-[5-(ter t-Bu t yl-d i-
m eth yl-silan yloxym eth yl)-th ioph en -2-yl]-4-m eth yl-pen tan -
3-on e (B-20). A mixture of B-19 (19.5 g, 73.0 mmol) in
CH2Cl2 (300 mL) was treated with imidazole (5.47 g, 80.3
mmol) and cooled to 0 °C. tert-Butyldimethylsilyl chloride (12.1
g, 80.3 mmol) was added, and the cold bath was removed. The
reaction mixture was stirred at room temperature for 6 h. The
suspension was filtered; the filtrate was washed with brine,
dried, and concentrated. The product was flash chromato-
graphed, eluting with hexane:EtOAc 95:5 to 92.5:7.5, to give
1
compound. H NMR (DMSO-d6): δ 0.87-0.90 (m, 6 H), 1.82-
1.96 (m, 2 H), 2.07-2.18 (m, 1 H), 2.25 (s, 3 H), 2.31 (d of ABX
q, J ) 17.8 Hz, 1 H), 2.60 (d of ABX q, J ) 18 Hz, 1 H), 2.74-
2.81 (m, 2 H), 4.96 (s, 1 H), 8.78 (s, 1 H), 11.39 (s, 1 H). MS
(APCI): AP+ ) 282.
Compounds D(1-6), D(8-17), and D(21-34) were pre-
pared in the same fashion from the analogous ketones. The
results are summarized in Table 2. During the workup of the
pyrazole analogue, the trityl protecting group was inadvert-
ently cleaved to give the free pyrazole D-7 in Table 2. For
compounds D(18-20), a slightly modified procedure was used
as follows: The crude aldol product isolated after the first step
was dissolved in THF and treated with Bu4NF (1.5 equiv) The
reaction mixture was stirred at room temperature for 3 h, then
concentrated, and dissolved in Et2O. The solution was washed
with dilute HCl, dried, and concentrated. This material was
then dissolved in THF, diluted with 1.0 N NaOH, and stirred
for 1 to 3 h. The reaction mixture was worked up as before to
give the desired compounds.
1
16.1 g (68%) of the title compound. H NMR (CDCl3): δ 0.09
(s, 6 H), 0.92 (s, 9 H), 1.09 (d, 6 H), 2.59 (m, 1 H), 2.81 (t, 2 H),
3.06 (t, 2 H), 4.78 (s, 2 H), 6.61 (d, 1 H), 6.69 (d, 1 H).
Silylated intermediates B-22 and B-24 were prepared from
compounds B-21 and B-23, respectively, in a similar fashion.
In these examples, the final product was not purified but used
crude in the next step.
4-Meth yl-1-(1-tr ityl-1H-p yr a zol-3-yl)-p en ta n -3-on e (B-
7). Compound B-6 (1.96 g, 12 mmol) was dissolved in DMF
(10 mL) and treated with trityl chloride (3.3 g, 12 mmol). The
mixture was stirred for 5 min, treated with NEt3, and stirred
overnight at room temperature. The solution was poured into
water and extracted with ether; the organic phase was dried
and concentrated. Chromatography of the residue, eluting with
9:1 hexane:EtOAc, gave 4.8 g (99%) of the title compound.
4-Met h yl-1-(t et r a h yd r o-fu r a n -2-yl)-p en t a n -3-on e (B-
11). A mixture of B-10 (1.14 g, 7.00 mmol), 5% Pd/CaCO3 (0.2
g), and THF (100 mL) was shaken in a hydrogen atmosphere
of 42 psi at 25 °C for 1.5 h. At that time, 20% Pd/C (0.4 g) was
added, and the reaction mixture was shaken in a hydrogen
atmosphere for another 50 h. The catalysts were filtered, and
the filtrate was concentrated. Chromatography of the residue,
eluting with 99:1 CH2Cl2:MeOH, yielded 0.88 g (74%) of the
title compound as a mixture of stereoisomers. 1H NMR
(CDCl3): δ 1.10 (d, J ) 6.8 Hz, 6 H), 1.46-2.04 (m, 6 H), 2.47-
2.66 (m, 3 H), 3.67-3.75 (m, 1 H), 3.77-3.87 (m, 2 H).
2,2,2-Tr iflu or o-N-[4-(4-m eth yl-3-oxo-p en tyl)-th ia zol-2-
yl]-a ceta m id e (B-35). A solution of B-34 (1.21 g, 5.03 mmol)
in 6 N HCl (50 mL) and THF (5 mL) was refluxed for 4 h and
then cooled to room temperature. Solid NaHCO3 was added
portionwise with caution until pH 7.2 was achieved. The
suspension was extracted with EtOAc. The combined extracts
were washed with brine, dried, and concentrated. The residue
was chromatographed, eluting with EtOAc, to give the depro-
Gen er a l Meth od for th e P r ep a r a tion of F in a l Dih y-
d r op yr on es. 3-(2-ter t-Bu tyl-4-h yd r oxym eth yl-5-m eth yl-
p h en ylsu lfa n yl)-4-h yd r oxy-6-isop r op yl-6-[2-(4-m et h yl-
th iazol-5-yl)-eth yl]-5,6-dih ydr o-pyr an -2-on e (26). A mixture
of D-25 (0.22 g, 0.78 mmol) in DMF (4 mL) was treated with
toluene-4-thiosulfonic acid S-(2-tert-butyl-4-hydroxymethyl-5-
methyl-phenyl) ester (0.33 g, 0.90 mmol) and K2CO3 (1.0 g,
7.2 mmol). The suspension was stirred overnight at room tem-
perature and then poured into a mixture of EtOAc and either
1 N HCl or saturated NH4Cl. The organic phase was separated,
washed with brine, dried and concentrated. Chromatography
(10% MeOH in CH2Cl2) afforded the title compound, mp 138-
1
141 °C. H NMR (DMSO-d6): δ 0.90-0.96 (m, 6 H), 1.46 (s, 9
H), 1.85 (s, 3 H), 1.95-2.00 (m, 2 H), 2.15-2.23 (m, 4 H), 2.70-
2.82 (m, 3 H), 2.93 (d of ABX q, 1 H), 4.34 (s, 2 H), 4.92 (br s,
1 H), 6.66 (s, 1 H), 7.24 (s, 1 H), 8.79 (s, 1 H).
Compounds 1-31 and 33-36 were prepared in similar
fashion from the thiotosylate and the dihydropyrone from
Table 2. For compound 32, the thiotosylate was reacted with
the trifluroacetamide D-33 under the general conditions listed
above, and the protecting group was removed using dilute
NaOH. For compound 8, the thiotosylate was reacted with the
trityl-protected compound D-6 under the same general condi-
tions, and the protecting group was removed with acetic acid
in methanol.
1
tected intermediate. H NMR (CDCl3): δ 1.04 (d, 6 H), 2.52-
2.59 (m, 1 H), 2.75 (m, 4 H), 6.06 (s, 1 H).
P r ep a r a tion of Ch ir a l Dih yd r op yr on es. P r ep a r a tion
of â-Hyd r oxyester fr om Keton e B-13. Diisopropylamine (43
mmol) was cooled to -10 °C and treated with nBuLi (43 mmol)
over 20 min. The solution was stirred for 30 min at -10 °C
and then cooled to -78 °C. tert-Butyl acetate (42.6 mmol) was
dissolved in THF and added dropwise to the LDA solution over
30 min. When addition was complete, the reaction mixture was
stirred at -78 °C to -40 °C for another 60 min. A solution of
ketone B-13 (21 mmol) in THF was added over 15 min, and
the reaction mixture was warmed to room temperature. The
solution was poured into 1 N HCl and ice; the product was
extracted into EtOAc, dried, and concentrated. Purification via
chromatography, eluting with hexane:EtOAc 97:3, afforded the
title compound as the tert-butyl ester. 1H NMR (CDCl3): δ 0.95
(dd, 6 H), 1.47 (s, 9 H), 1.7-2.0 (m, 3 H), 2.36-2.54 (AB q, 2
H), 2.6-2.8 (m, 2 H), 6.93-6.95 (m, 2 H), 7.23-7.25 (m, 1 H).
Separ ation of â-Hydr oxy Ester En an tiom er s via Ch ir al
HP LC: (S)-3-Hyd r oxy-4-m eth yl-3-(2-th iop h en -3-yl-eth yl)-
p en ta n oic Acid ter t-Bu tyl Ester (E-(S)). The title com-
pound was prepared by resolution on a Chiralpak AD column
eluting with 1:99 2-propanol:hexane to afford both enantiomers
of 3-hydroxy-4-methyl-3-(2-thiophen-3-yl-ethyl)-pentanoic acid
tert-butyl ester. The S enantiomer eluted first. 1H NMR
(CDCl3): δ 0.95 (dd, 6 H), 1.47 (s, 9 H), 1.7-2.00 (m, 3 H),
A solution of the ketone prepared above (0.75 g, 3.8 mmol)
in CH2Cl2 (50 mL) was cooled in an ice bath, treated with NEt3
(0.6 mL, 4.3 mmol) and trifluoromethyl acetic anhydride (0.6
mL, 4.3 mmol), and allowed to warm to room temperature.
H2O was added. The organic layer was separated, washed with
brine, and dried. Concentration gave an oil which was chro-
matographed over silica gel, eluting with EtOAc, to give 0.70
g (55% over two steps) of the title compound. 1H NMR (CDCl3)
δ 1.10 (d, 6 H), 2.58-2.65 (m, 1 H), 2.83-2.86 (m, 2 H), 2.91-
2.95 (m, 2 H), 6.61 (s, 1 H).
Gen er a l Meth od for th e P r ep a r a tion of 4-Hyd r oxy-5,6-
d ih yd r op yr on es. 4-Hyd r oxy-6-isop r op yl-6-[2-(4-m eth yl-
th ia zol-5-yl)-eth yl]-5,6-d ih yd r o-p yr a n -2-on e (D-25). Meth-
yl acetoacetate (1.00 g, 8.61 mmol) was added dropwise to a
slurry of NaH (0.38 g, 9.5 mmol) in 100 mL of anhydrous THF
at 0 °C, and the mixture was stirred at 0 °C for 20 min.
n-Butyllithium (4.5 mL of 2.1 M; 9.45 mmol) was then added,
and the reaction mixture was stirred at 0 °C for 30 min. A
solution of B-29 (1.53 g, 142 mmol) in THF (50 mL) was added
all at once. The solution was stirred at 0 °C to room temper-
ature for 90 min, then treated with 0.1 N HCl with stirring.
The THF was evaporated; the residue was partitioned between
H2O and EtOAc. The organic layer was separated, dried, and
concentrated. The crude aldol product was dissolved in 20 mL

2332 J ournal of Medicinal Chemistry, 2001, Vol. 44, No. 14
Hagen et al.
Thaisrivongs, S. Tipranavir (PNU-140690): A potent, orally
bioavailable nonpeptidic HIV protease inhibitor of the 5,6-
dihydro-4-hydroxy-2-pyrone sulfonamide class. J . Med. Chem.
1998, 41 (18), 3467-3476 and references therein.
2.36-2.54 (AB q, 2 H), 2.6-2.8 (m, 2 H), 6.93-6.95 (m, 2 H),
7.23-7.25 (m, I H).
(R)-3-Hyd r oxy-4-m eth yl-3-(2-th iop h en -3-yl-eth yl)-p en -
ta n oic Acid ter t-Bu tyl Ester (E-(R)). The R enantiomer
eluted second: 1H NMR (CDCl3): δ 0.95 (dd, 6 H), 1.47 (s, 9
H), 1.7-2.00 (m, 3 H), 2.36-2.54 (AB q, 2 H), 2.6-2.8 (m, 2
H), 6.93-6.95 (m, 2 H), 7.23-7.25 (m, 1 H).
(S)-3-Hyd r oxy-4-m eth yl-3-(2-th iop h en -3-yl-eth yl)-p en -
t a n oic Acid (F -(S)). Ester (S)-3-hydroxy-4-methyl-3-(2-
thiophen-3-yl-ethyl)-pentanoic acid tert-butyl ester (8 mmol)
was dissolved in EtOH and treated with LiOH (16 mmol), H2O
(5 mL), and MeOH (15 mL). The mixture was stirred at room
temperature for 18 h and concentrated. The residue was
partitioned between H2O and Et2O. The aqueous layer was
separated, acidified with 1 N HCl, and extracted with Et2O.
The solution was dried and concentrated to give the title
compound: 1H NMR (CDCl3): δ 0.98 (t, 6 H), 1.8-2.0 (m, 3
H), 2.50-2.70 (AB q, 2 H), 2.6-2.8 (m, 2 H), 6.93-6.95 (m, 2
H), 7.23-7.25 (m, 1 H).
(12) Boyer, F. E.; Domagala, J . M.; Ellsworth, E. L.; Gajda, C. A.;
Hagen, S. E.; Hamilton, H. W.; Lunney, E. A.; Markoski, L. J .;
Vara Prasad, J . V. N.; Tait, B. D. Dihydropyrones with improved
antiviral activity. U.S. Patent 6,046,355, April 4, 2000.
(13) Huang, Y.-Z.; Shi, L.-L.; Li, S.-W. An efficient and stereoselective
synthesis of E)-enones via arsonium salts. Preparation of key
intermediates for the synthesis of brassinosteroid and prosta-
glandin. Synthesis 1988, 975-977.
(14) (a) Tamaru, Y.; Yamada, Y.; Yoshida, Z. The palladium catalyzed
thienylation of allylic alcohols with 2- and 3-bromothiophenes
and their derivatives. Tetrahedron 1979, 29 35(3), 329-40. (b)
J effrey, T. Palladium-catalyzed arylation of allylic alcohols:
highly selective synthesis of â-aromatic carbonyl compounds or
â-aromatic R,â-unsaturated alcohols. Tetrahedron Lett. 1991, 32
(19), 2121-2124.
(15) Hagen, S. E.; Vara Prasad, J . V. N.; Boyer, F. E.; Domagala, J .
M.; Ellsworth, E. L.; Gajda, C.; Hamilton, H. W.; Markoski, L.
J .; Steinbaugh, B. A.; Tait, B. D.; Lunney, E. A.; Tummino, P.
J .; Ferguson, D.; Hupe, D.; Nouhan, C.; Graceck, S. J .; Saun-
ders: J . M.; VanderRoest, S. Synthesis of 5,6-dihydro-4-hydroxy-
2-pyrones as HIV-1 protease inhibitors: the profound effect of
polarity on antiviral activity. J . Med. Chem. 1997, 40 (23), 3707-
3711.
(16) Tummino, P. J .; Vara Prasad, J . V. N.; Ferguson, D.; Nouhan,
C.; Graham, N.; Domagala, J . M.; Ellsworth, E.; Gajda, C.;
Hagen, S. E.; Lunney, E. A.; Para, K. S.; Tait, B. D.; Pavlovsky,
A.; Erickson, J . W.; Gracheck, S. J .; McQuade, T. J .; Hupe, D. J .
Discovery and optimization of non-peptide HIV-1 protease
inhibitors. Bioorg. Med. Chem. 1996, 4 (9), 1401-1410.
(17) Buckheit, R. W., J r.; Hollingshead, M. G.; Germany-Decker, J .;
White, E. L.; McMahon, J . B.; Allen, L. B.; Ross, L. J .; Decker,
W. D.; Westbrook, L.; Shannon, W. M.; Weislow, O.; Bader, J .
P.; Boyd, M. R. Thiazolobenzimidazole: biological and biochemi-
cal anti-retroviral activity of a new nonnucleoside reverse
transcriptase inhibitor. Antiviral Res. 1993, 21, 247-265.
(18) Vara Prasad, J . V. N.; Boyer, F. E.; Domagala, J . M.; Ellsworth,
E. L.; Gajda, C.; Hamilton, H. W.; Hagen, S. E.; Markoski, L. J .;
Steinbaugh, B. A.; Tait, B. D.; Humblet, C.; Lunney, E. A.;
Pavlovsky, A.; Rubin, J . R.; Ferguson, D.; Graham, N.; Holler,
T.; Hupe, D.; Nouhan, C.; Tummino, P. J .; Uromov, A.; Zeikus,
E.; Zeikus, G.; Gracheck, S. J .; Saunders, J . M.; VanderRoest,
S.; Brodfuehrer, J .; Iyer, K.; Sinz, M.; Gulnik, S. V.; Erickson,
J . W. Nonpeptidic HIV protease inhibitors possessing excellent
antiviral activities and therapeutic indices. PD 178390: A lead
HIV protease inhibitor. Bioorg. Med. Chem. 1999, 7, 2775-2800.
(19) Barton, A.; Breukelman, S. P.; Kaye, P. K.; Meakins, G. D.;
Morgan, D. J . The preparation of thiazole-4- and -5-carboxylates,
and an infrared study of their rotational isomers. J . Chem. Soc.,
Perkin Trans. 1 1982, 159-164.
(20) Kempf, D. J .; Sham, H. L.; Marsh, K. C.; Flentge, C. A.;
Betebenner, D.; Green, B. E.; McDonald, E.; Vasavanonda, S.;
Saldivar, A.; Wideburg, N. E.; Kati, W. M.; Ruiz, L.; Zhao, C.;
Fino, L.; Patterson, J .; Molla, A.; Plattner, J . J .; Norbeck, D. W.
Discovery of Ritonavir, a potent inhibitor of HIV protease with
high oral bioavailability and clinical efficacy. J . Med. Chem.
1998, 41 (4), 602-617.
(21) Dondoni, A.; Fantin, G.; Fogagnolo, M.; Medici, A.; Pedrini, P.
A new convenient preparation of 2-, 4-, and 5-thiazolecarboxal-
dehydes and their conversion into the corresponding carbonitrile
N-oxides: synthesis of 3-thiazolyl-isoxazoles and 3-thiazolyl-
isoxazolines. Synthesis 1987, 11, 998-1001.
The R isomer was prepared in identical fashion from the R
ester.
P r ep a r a tion of th e Ch ir a l â-Ketoester fr om â-Hyd r oxy
Acid : (S)-5-Hyd r oxy-6-m eth yl-3-oxo-5-(2-th iop h en -3-yl-
eth yl)-h ep ta n oic Acid Eth yl Ester (G-(S)). A solution of
(S)-3-hydroxy-4-methyl-3-(2-thiophen-3-yl-ethyl)-pentanoic acid
(7.81 mmol) as isolated above in THF (30 mL) was treated with
CDI (8.6 mmol) and stirred for 18 h at room temperature. Bis
[3-methoxy-3-oxopropanoato (1-)-O,O′] magnesate (15.6 mmol)
was added, and the reaction was stirred for 6 h at room
temperature. The reaction was concentrated and the residue
partitioned between EtOAc and 1 N HCl. The organic layer
was washed with aqueous NaHCO3 and brine, dried, and
concentrated. Purification was accomplished using silica
gel chromatography, eluting with 100% CH2Cl2: 1H NMR
(CDCl3): δ 0.94 (dd, 6 H), 1.27 (t, 3 H), 1.75-1.90 (m, 2 H),
1.9-2.0 (m, 1 H), 2.65-2.75 (m, 3 H), 2.80 (d, 1 H), 3.48 (s, 2
H), 4.15-4.25 (m, 2 H), 6.93 (m, 2 H), 7.2-7.3 (m, 1 H).
The R isomer was prepared in identical fashion from the R
acid.
Refer en ces
(1) Carpenter, C. C.; Fischl, M. A.; Hammer, S. M.; Hirsch, M. S.;
J acobsen, D. M.; Katzenstein, D. A. Antiretroviral therapy for
HIV infection in 1998: updated recommendations of the inter-
national AIDS Society - USA panel. J . Am. Med. Assoc. 1998,
280, 78-86.
(2) Barry, M.; Gibbons, S.; Back, D.; Mulcahy, F. Protease inhibitors
in patients with HIV disease. Clin. Pharmacokinet. 1997, 32,
194-209.
(3) Vigoroux, C.; Gharakhanian, S.; Salhi, Y.; Nguyen, T. H.; Adda,
N.; Rozenbaum, W.; Capeau, J . Adverse metabolic disorders
during highly active retroviral treatments (HAART) of HIV
disease. Diabetes Metab. 1999, 25 (5), 383-392.
(4) Sommadossi, J . P. HIV protease inhibitors: pharmacologic and
metabolic distinctions. AIDS (London) 1999, 13 (Suppl. 1), S29-
S40.
(5) Malaty, L. I.; Kuper, J . J . Drug interactions of HIV protease
inhibitors. Drug Saf. 1999, 20 (2), 147-169.
(6) Erickson, J . W.; Gulnick, S. V.; Markowitz, M. Protease inhibi-
tors: resistance, cross-resistance, fitness and the choice of initial
and salvage therapies. AIDS (London) 1999, 13 (Suppl. A),
S189-S204.
(7) Valdez, H. N.; Lederman, M. M.; Wooley, I.; Walker, C. J .;
Vernon, L. T. Hise, A.; Gripshover, B. M. Human immunodefi-
ciency virus 1 protease inhibitors in clinical practice. Predictors
of clinical outcome. Arch. Intern. Med. 1999, 159, 1771-1776.
(8) Battegay, M.; Harr, T.; Sponagel, L. Salvage treatment against
human immunodeficiency virus. Ann. Med. 1999, 31 (4), 253-
260.
(22) Comins, D. L.; Killpack, M. O. Lithiation of heterocycles directed
by amino alkoxides. J . Org. Chem. 1987, 52 (1), 104-109.
(23) Dowle, M. D.; Hayes, R.; J udd, D. B.; Williams, C. N. A
convenient synthesis of Mannich bases of thiophene and sub-
stituted thiophene. Synthesis 1983, 1, 73-75.
(24) Iversen, P. E.; Lund, H. Preparation of 2-imidazole and 2-thia-
zolecarboxaldehydes. Acta Chem. Scand. 1966, 20, 2649-2657.
(25) White, R. L.; Spenser, I. D. Thiamin biosynthesis in yeast. Origin
of the 5-carbon unit of the thiazole moiety. J . Am. Chem. Soc.
1982, 104 (18), 4934-4943.
(26) Silberg, A.; Frenkel, Z. Contributions a l’etude des thiazols. XVI-
Syntheses de quelques aldehydes thiazoliques et de leurs derives
d’un interet therapeutique presomptif. Bull. Soc. Chim. Fr. 1967,
6, 2235-2238.
(27) Moyer, M. P.; Shiube, J . F.; Rapoport, H. Metal-halogen
exchange of bromoindoles. A route to substituted indoles. J . Org.
Chem. 1986, 51 (26), 5106-5110.
(9) Deeks, S. G. Failure of HIV-1 protease inhibitors to fully
suppress viral replication: Implications for salvage therapy. Adv.
Exp. Med. Biol. 1999, 458, 175-182.
(10) For a review of our work, see Hagen, S.; Vara Prasad, J . V. N.;
Tait, B. D. Nonpeptide inhibitors of HIV protease. Adv. Med.
Chem. 2000, 5, 159-195.
(11) Turner, S. R.; Strohbach, J . W.; Tommasi, R. A.; Aristoff, P. A.;
J ohnson, P. D.; Skulnick, H. I.; Dolak, L. A.; Seest, E. P.; Tomich,
P. K.; Bohanon, M. J .; Horng, M.-M.; Lynn, J . C.; Chong, K.-T.;
Hinshaw, R. R.; Watenpaugh, K. D.; J anakiraman, M. N.;
(28) Vetelino, M. G.; Coe, J . W. A mild method for the conversion of
activated aryl methyl groups to carboxaldehydes via the un-
catalyzed periodic cleavage of enamines. Tetrahedron Lett. 1994,
35(2), 219-222.
J M0003844