Efficient method to prepare hydroxyethylamine-based aspartyl protease inhibitors with diverse P1 side chains

Article abstract of DOI:10.1016/S0040-4020(02)00629-4

An efficient procedure for the solid-phase synthesis of hydroxyethylamine-based aspartyl protease inhibitors is described. The 1,2-bromo alcohol 4 is the key intermediate and is prepared in four-steps in good overall yield from commercial available amino acids.

Full text of DOI:10.1016/S0040-4020(02)00629-4

Tetrahedron 58 (2002) 6305–6310
Efficient method to prepare hydroxyethylamine-based aspartyl
protease inhibitors with diverse P1 side chains
*
Masao Chino, Masahiro Wakao and Jonathan A. Ellman
Department of Chemistry, Center for New Directions in Organic Synthesis, University of California, Berkeley, CA 94720, USA
Received 20 February 2002; revised 4 April 2002; accepted 22 April 2002
Abstract—An efficient procedure for the solid-phase synthesis of hydroxyethylamine-based aspartyl protease inhibitors is described. The
1,2-bromo alcohol 4 is the key intermediate and is prepared in four-steps in good overall yield from commercial available amino
acids. q 2002 Published by Elsevier Science Ltd.
1. Introduction
D3,8 implicated in neurodegenerative disease9 and cancer
metastasis10 and to the plasmepsins I and II,11 which are key
proteases in the life cycle of the malaria parasite.
Combinatorial synthesis sequences designed to target
protein families are particularly important because a
sufficiently versatile and robust library synthesis approach
can potentially be used to identify ligands to any member of
that protein family.1,2 We have demonstrated the power of
this approach by developing a strategy for the solid-phase
synthesis of libraries of compounds3,4 to the aspartyl
protease family,5 which includes the extremely important
drug targets, HIV protease6 and b secretase implicated in
Alzheimer’s disease.7
To develop potent and selective aspartyl protease inhibitors
it is essential that an appropriate P1 side chain be identified.
In our previously reported method,3,4 the P1 side chain is
introduced by addition of Grignard reagents to a support-
bound amide followed by functional group manipulations
(Scheme 1). While this method is effective for introducing
hydrophobic side chains, Grignard reagents are not
compatible with a variety of functionality. For example,
yapsin, a recently identified aspartyl protease that is
implicated in peptide hormone processing, prefers basic
lysine and arginine side chains at the P1 position in peptide
substrates.12 Our previously reported method cannot be used
to incorporate these side chains. Here we describe a practical
and efficient sequence for the preparation of hydroxyethyl-
amine-based inhibitors that is amenable to the incorporation
of a wide range of functionality at the P1 position.
We designed our library synthesis sequence to provide
inhibitors based upon the hydroxyethylamine isostere (Fig.
1), which is a key pharmacophore for targeting aspartyl
proteases and is present in several of the approved drugs for
the treatment of AIDs.5,6 This library synthesis approach has
resulted in the rapid identification of single digit nanomolar
and high picomolar small molecule inhibitors to cathepsin
2. Results and discussion
The key intermediate, 1,2-bromo alcohol 4, is prepared in a
straightforward four-step sequence from side-chain
Figure 1. Hydroxyethylamine inhibitors.
Keywords: combinatorial chemistry; hydroxyethylamine isosteres; aspartyl
protease inhibitors.
*
Corresponding author. Tel.: þ1-510-642-4488; fax: þ1-510-642-8369;
e-mail: jellman@uclink.berkeley.edu
Scheme 1.
0040–4020/02/$ - see front matter q 2002 Published by Elsevier Science Ltd.
PII: S0040-4020(02)00629-4

6306
M. Chino et al. / Tetrahedron 58 (2002) 6305–6310
Scheme 2. Reagents and conditions: (a) TfN3, K2CO3, cat CuSO4, aqueous MeOH, CH2Cl2, rt, 92% yield for 2a, 68% yield for 2b; (b) isobutyl chloroformate,
N-methylmorpholine, THF, 2408C, then CH2N2, 08C; (c) aqueous HBr, CH2Cl2, 08C; (d) NaBH4, THF, H2O, 65% yield over 3 steps for 4a, 60% yield over 3
steps for 4b.
is well documented that for aspartyl protease inhibitors
either alcohol stereoisomer can provide the greatest
inhibitory activity depending upon both the protease that
is targeted and the functionality displayed on the inhibitor.5
The efficiency of the sequence was demonstrated by the
preparation of 4a and 4b, which incorporate the phenyl-
alanine and protected lysine side chains, in 58 and 41%
overall yields, respectively. In addition, minimal epimeriza-
tion (,2%) occurs during the halomethylation and
reduction steps as determined by chiral HPLC analysis of
4a and 4b. The enantiomers of 4a and 4b were used as
HPLC standards and were prepared from the corresponding
D-amino acids.
Scheme 3.
protected amino acids (Scheme 2). Because a very large
number of natural and unnatural amino acids are com-
mercially available, this approach provides access to bromo
alcohols 4 incorporating a large number of diverse P1 side
chains. First, the amino acid is stereospecifically converted
to a-azido carboxylic acid 2 by using trifluoromethansul-
fonyl azide according to the procedures of Wong13 and
Pelletier.14 The a-bromo ketone 3 is then prepared in a one-
pot sequence by activation of carboxylic acid 2 with
isobutyl chloroformate and N-methylmorpholine followed
by addition of diazomethane and then aqueous HBr.
Reduction with NaBH4 provides the desired alcohol 4 as a
1:1 ratio of stereoisomers. The 1:1 mixture of alcohol
stereoisomers is ideal for the discovery of inhibitors since it
For the solid-phase synthesis of aspartyl protease inhibitors,
the secondary alcohol intermediate 4 is coupled to DHP
resin (Scheme 3).15 While neither camphorsulfonic acid nor
toluenesulfonic acid were effective catalysts for loading
alcohol 4 to support, pyridinium toluenesulfonate (PPTs)
resulted in good loading efficiencies providing 5a and 5b in
0.57 and 0.58 mmol/g loading levels, respectively. Because
the conditions to load alcohol 4 are acidic and may result in
cleavage of the Boc group, prior to further synthetic
transformations 5b was treated with tert-butyl carbonic
anhydride to ensure that the amine side chain was
completely protected.
Scheme 4. Reagents and conditions: (a) iBuNH2, NMP, 808C; (b) N-Fmoc-4-aminobenzenesulfonyl choride, iPr2NEt, THF, rt; (c) SnCl2, PhSH, Et3N, THF,
rt; (d) N-succinimidyl carbonate of 3-(S)-hydroxytetrahydrofuran, i Pr2NEt, THF, rt; (e) 20% piperidine in NMP; (f) 95:5 THF/H2O, 15 min.

M. Chino et al. / Tetrahedron 58 (2002) 6305–6310
6307
To demonstrate that support-bound intermediates 5 can
4.2. Trifluoromethanesulfonyl azide (triflyl azide)13
effectively be converted to aspartyl protease inhibitors, the
FDA approved HIV protease inhibitor amprenavir,16 9a,
was synthesized (Scheme 4). A derivative of amprenavir,
which incorporates the lysine side chain at the P1 position,
9b, was also prepared to demonstrate the compatibility of
the solid-phase sequence with reactive side chain function-
ality. The syntheses were initiated by adding isobutylamine
to 5a and 5b followed by reaction of the resulting secondary
amines with N-Fmoc-4-aminobenzenesulfonyl chloride to
provide the tertiary sulfonamide intermediates 6a and 6b.
Reduction of the azide with thiophenol/Et3N/SnCl2 (4:5:1)
as described by Bartra and co-workers provided the amines
7a and 7b,17 which were acylated with the N-succinimidyl
carbonate of 3-(S )-hydroxytetrahydrofuran to provide
carbamates 8a and 8b. Removal of the Fmoc protecting
groups with 20% piperidine in NMP followed by treatment
with 95:5 TFA/water for 15 min provided inhibitors 9a
(amprenavir) and 9b in 66 and 70% overall yields after
reverse-phase HPLC purification, respectively.
A solution of NaN3 (595 mg, 9.15 mmol) in 1.5 mL of H2O
was cooled in an ice bath and treated with 2.5 mL of
CH2Cl2. The resulting biphasic mixture was stirred
vigorously and treated with Tf2O (523 mg, 1.85 mmol)
over a period of 5 min. The reaction mixture was stirred at
08C for 2 h. The organic phase was then separated and the
aqueous phase was extracted twice with CH2Cl2. The total
volume of the reagent solution was 5 mL. The combined
organic layers were washed once with saturated Na2CO3
solution and the resulting triflyl azide solution used without
further purification. WARNING: According to the literature
report the triflyl azide solution should not be heated or
concentrated.
4.2.1. Synthesis of a-azido acid 2a (R5Bn). The diazo
transfer reaction was performed according to the procedure
of Pelletier for converting a-amino acids to a-azido acids.14
L-Phe (460 mg, 2.79 mmol) was combined with K2CO3
(577.5 mg, 4.19 mmol) and CuSO4 pentahydrate (6.98 mg,
27.9 mmol), distilled H2O (9 mL), and CH3OH (18 mL).
Triflyl azide in CH2Cl2 (15 mL, ,5.55 mmol, 2 equiv.)
prepared according to the procedure above was added with
stirring, and the resulting mixture was stirred at ambient
temperature overnight. Subsequently, the organic solvents
were removed under reduced pressure and the aqueous
slurry was diluted with H2O (50 mL). This slurry was then
acidified to pH 6 with 2N HCl aqueous solution and diluted
with 0.25 M, pH 6.2 phosphate buffer (50 mL) and extracted
with EtOAc (4£) to remove the sulfonamide byproduct.
The aqueous phase was then acidified to pH 2 with 2N HCl
aqueous solution. The product was obtained from another
round of EtOAc extractions (3£). The EtOAc extracts were
combined, dried (Na2SO4) and evaporated to dryness giving
474 mg (89% yield) of the a-azido acid 2a (R¼Bn) as a
pale yellow oil. The spectral data correlated exactly with the
literature values.14
3. Conclusion
In summary, we report an efficient procedure for the solid-
phase synthesis of hydroxyethylamine-based aspartyl pro-
tease inhibitors with 1,2-bromoalcohol 4 serving as the key
intermediate. Because 4 is prepared in four straightforward
steps from readily available amino acids in good overall
yields and without racemization, this method should provide
rapid access to inhibitors incorporating a large number of
different side chains at the P1 position.
4. Experimental
4.1. General methods
Materials were obtained from commercial suppliers and
employed without further purification unless otherwise
stated. DHP resin was purchased from NovaBiochem (San
Diego, CA). The following solvents were distilled under N2
from the specified drying agents: tetrahydrofuran (THF) and
diethyl ether (Et2O) from sodium/benzophenone ketyl, and
methylene chloride (CH2Cl2), dichloroethane and pyridine
from calcium hydride. Diazomethane was generated in situ
using the following procedure.18 To a 0.7 M suspension of
p-toluenesulfonylmethylnitrosamide (Diazold) in absolute
ethanol was added small aliquots of 40 wt% KOH aqueous
solution until the Diazold suspension became white.
Diazomethane gas was produced and transferred by cannula
(two fine-polished pipets connected by rubber tubing) under
positive N2 flow into the stirring THF solution of the second
reagent. Infrared spectra (IR) were recorded with a Perkin–
Elmer 1600 Series Fourier transform spectrometer using
4.2.2. Synthesis of a-azido acid 2b (R5(CH2)4NHBoc).
Compound 2b was prepared according to the procedure
used to prepare 2a. From 686 mg (2.79 mmol) of N-1-Boc-
Lys was obtained 516 mg (68% yield) of 2b as a pale yellow
oil. The spectral data correlated exactly with the literature
values.14
4.2.3. Bromomethyl alcohol 4a (R5Bn). Isobutyl chloro-
formate (287 mL, 2.2 mmol) was added to a solution of
azido acid 2a (382 mg, 2.0 mmol) and N-methylmorpholine
(242 mL, 2.2 mmol) in THF (20 mL) at 2408C, and the
reaction mixture was stirred for 15 min. The reaction
mixture was then filtered and diazomethane (from 1.33 g,
6.2 mmol of 1-methyl-3-nitro1-nitrosoguanidine and
1.8 mL of 40 wt% aqueous KOH in 10 mL of EtOH) was
added slowly. The reaction flask was stoppered and was
maintained at 08C in a refrigerator overnight. The reaction
mixture was then treated with 48% HBr aq (410 mL) and
was stirred for 15 min. The reaction mixture was diluted
with EtOAc (20 mL) and then was extracted with 15 wt%
aqueous citric acid, then washed with saturated sodium
bicarbonate (with CO2 evolution), and then washed with
brine. The organic layer was dried over anhydrous sodium
sulfate, filtered, and concentrated under reduced pressure.
1
NaCl plates, and only partial spectral data is listed. All H
and 13C NMR spectra were obtained in CDCl3 unless
1
otherwise stated, and chemical shifts for H and 13C NMR
are recorded in parts per million relative to the internal
solvent peak at 7.26 and 77.0 ppm, respectively. Coupling
constants are reported in hertz. LC–MS analyses were
conducted using a Hewlett Packard 1100 MSD with an
Agilent Zorbax SB-C18 column.

6308
M. Chino et al. / Tetrahedron 58 (2002) 6305–6310
The residue was dissolved in 20 mL of 95:5 THF/H2O. To
the resulting solution was added sodium borohydride (0.1 g,
2.6 mmol) gradually at rt. The reaction mixture was stirred
for 1 h and then was neutralized with aqueous 1N HCl. After
extraction with EtOAc (3£20 mL), the organic extracts
were washed with brine, dried (Na2SO4), and concentrated
under reduced pressure. Purification by silica-gel chroma-
tography with 85:15 hexanes/EtOAc afforded 350 mg (65%
yield over 3 steps) of 4a (R¼Bn) as a white solid: IR
residue was purified by silica-gel chromatography with
80:20 hexane/EtOAc to afford 80.3 mg of benzoylated
compound 4b (97%). In the same manner, the 4b prepared
from D-N-1-Boc-Lys was converted to the corresponding
benzoylated compound. 1H NMR (400 MHz, CDCl3, for 1:1
mixture): d ppm 1.42 (s, 9H£2), 1.40–1.74 (m, 6H£2),
3.03–3.20 (br, 2H£2), 3.56–3.88 (m, 3H£2), 4.55 (brs,
1H£2), 5.23 (m, 1H), 5.31 (m, 1H), 7.45 (m, 2H£2), 7.60
(m, 1H£2), 8.06 (m, 2H£2). HRMS calcd for C19H27-
BrN4O4þH 455.12939, found 455.12966. The HPLC
analysis of a 1:1 mixture of these compounds (5 mL,
concentration¼2.5 mg/mL) using a DAICEL Chiralcel
column with EtOH/hexane (1.5:98.5) as an eluent at a
flow rate of 1.0 mL/min at wavelength of 230 nm afforded
four peaks of equal area corresponding to the four possible
diastereomers. The retention time (RT) for L- and D-form
was 23.3 and 26.3 min and 19.7 and 21.9 min, respectively.
HPLC analysis of 4b derived from L-N-1-Boc-Lys estab-
lished that no racemization had occurred during the
synthesis sequence [rt (peak area), 23.3 min (50%),
26.3 min (50%)].
(NaCl): 3418, 2107, 1262, 1050 cm21
.
1H NMR
(500 MHz): d 2.33 (brs, 1H), 2.40 (brs, 1H), 2.84 (dd,
J¼14.0, 8.6 Hz, 1H), 3.03 (dd, J¼13.7, 7.9 Hz, 1H), 3.08
(dd, J¼13.7, 6.7 Hz, 1H), 3.15 (dd, J¼14.0, 3.6 Hz, 1H),
3.48 (dd, J¼10.3, 5.4 Hz, 1H), 3.53 (dd, J¼10.4, 6.7 Hz,
1H), 3.61–3.79 (m, 3H£2), 7.26–7.30 (m, 3H£2), 7.33–
7.39 (m, 2H£2). 13C NMR (125 MHz): d 35.2, 36.9, 37.0,
37.4, 64.8, 66.0, 71.9, 72.2, 127.0, 127.1, 128.7, 128.8,
129.2, 129.4, 136.6, 136.9. Anal. Calcd for C10H12BrN3O:
C, 44.46; H, 4.48; N, 15.56. Found: C, 44.73; H, 4.64; N,
15.28.
Determination of enantiopurity of bromomethyl alcohol 4a.
Both enantiomers of 4a were synthesized from L- and
D-Phe. The HPLC analysis of a 1:1 mixture of these
compounds (5 mL, concentration¼2.5 mg/mL) using a
DAICEL Chiralcel column with i PrOH/hexane (2:98) as
an eluent at a flow rate of 1.0 mL/min at wavelength of
210 nm afforded four peaks of equal area corresponding to
the four possible diastereomers. The retention time for 4a
prepared from L- and D-Phe was 20.8 and 22.2 min and 16.5
and 31.0 min, respectively. HPLC analysis of 4a derived
from L-Phe established that racemization was less than 2.5%
[rt (peak area), 16.5 min (,1%), 20.8 min (46%), 22.2 min
(52%), 31.0 min (,1%)].
4.2.5. N-Fmoc-4-aminobenzenesulfonic acid.19 Sulphani-
lic acid (6.80 g, 35.6 mmol) was dissolved in a solution of
saturated aqueous NaHCO3 (17 mL) and reacted with
FmocCl (0.92 g, 4.82 mmol) at pH 8 with mechanical
stirring. The reaction mixture was stirred for 12 h, and then
the precipitated solid was filtered off and washed with
Et2O. The filtrate was dried under reduced pressure. The dry
residue was dissolved in a mixture of dry toluene (25 mL)
and anhydrous DMF (2.5 mL) and thionyl chloride
(1.00 mL, 13.7 mmol) was then added at 08C. The reaction
mixture was stirred for 12 h at room temperature, then
poured into an ice-water mixture and then neutralized with
NaHCO3. The aqueous phase was extracted with EtOAc
(2£25 mL). The organic layer was washed with brine, dried
(NaSO4), and concentrated. Recrystallization of the crude
product from benzene and hexane afforded 811 mg (54%
yield) of desired product as a white crystalline solid: IR
(NaCl): 2360, 1716, 1589, 1524, 1373, 1320, 1220,
4.2.4. Bromomethyl alcohol 4b (R5(CH2)4NHBoc).
Compound 4b was prepared according to the procedure
used to prepare 4a starting with 545 mg (2.0 mmol) of 2b.
Purification by silica-gel chromatography with 70:30
hexanes/EtOAc afforded 420 mg (60% yield for 3 steps)
of 4b (R¼(CH2)4NHBoc) as a white solid: IR (NaCl): 3433,
2925, 2359, 2099, 1688, 1524 cm21. 1H NMR (500 MHz, in
acetone-d6): d 1.38 (s, 9H£2), 1.38–1.79 (m, 6H£2),
3.03–3.13 (m, 2H£2), 3.37–3.44 (m, 1H), 3.48–3.66 (m,
5H), 3.82–3.86 (m, 1H), 3.87–3.91 (m, 1H), 5.98 (brs,
1H£2). 13C NMR (125 MHz): d 22.9, 23.1, 28.4, 29.6, 29.9,
35.1, 37.1, 44.8, 54.3, 63.5, 64.7, 72.3, 72.6, 156.1, 156.2.
HRMS (FABMS) calcd for [M]þ (C12H23BrLiN4O3)
357.13135, found 357.111960.
1170 cm21
.
1H NMR (500 MHz): d 4.28 (t, J¼6.0 Hz,
1H), 4.66 (d, J¼6.0 Hz, 2H), 7.35 (td, J¼7.5, 1.0 Hz, 2H),
7.44 (t, J¼7.5 Hz, 2H), 7.53–7.60 (m, 2H), 7.61 (d,
J¼7.5 Hz, 2H), 7.80 (d, J¼7.5 Hz, 2H), 7.96 (d,
J¼7.5 Hz). 13C NMR (125 MHz): d 46.9, 67.3, 118.1,
120.1, 124.7, 127.2, 128.0, 128.7, 138.1, 141.4, 143.3,
144.1, 152.5. Anal. Calcd for C21H16 ClNO4S: C, 60.94; H,
3.90; N, 3.38. Found: C, 61.14; H, 4.05; N, 3.30.
4.3. General solid-phase synthesis methods
Determination of enantiopurity of bromomethyl alcohol 4b.
Both enantiomers of 4b were synthesized from L- and D-N-
1-Boc-Lys and were then benzoylated to introduce a
chromophore for HPLC analysis. To a solution of 4b
(64.0 mg, 0.182 mmol) prepared from L-N-1-Boc-Lys in
2 mL of CH2Cl2 were added pyridine (64.4 mL,
0.797 mmol), 4-(dimethylamino)pyridine (DMAP, 2.4 mg,
0.020 mmol), and benzoyl chloride (46.3 mL, 0.399 mmol).
The mixture was stirred at ambient temperature overnight.
The reaction was quenched by addition of sat. NaHCO3 aq.
(3 mL). The aqueous layer was extracted with CH2Cl2
(3£5 mL). The organic layers were combined, dried over
Na2SO4, and concentrated under reduced pressure. The
All solid-phase reaction mixtures were stirred at the slowest
rate. For the general solid-phase workup procedure, the
reaction solution was filtered away from the support-bound
material using polypropylene cartridges with 70 mm PE frits
(Speed Accessories) attached to Teflon stopcocks.
Cartridges and stopcocks were purchased from Applied
Separations (Allentown, PA). The support-bound material
was thoroughly washed with various solvents as described
in the specific experimental sections.
4.3.1. Support-bound bromide 5a (R5Bn). To a mixture
of 144 mg (0.141 mmol) of DHP resin in 0.5 mL of

M. Chino et al. / Tetrahedron 58 (2002) 6305–6310
6309
dichloroethane at room temperature was added a solution of
93.5 mg (0.212 mmol) of alcohol 4a (R¼Bn) in 0.5 mL of
dichloroethane. The reaction mixture was stirred for 10–
12 h at 608C followed by the addition of 88.5 mg
(0.353 mmol) of PPTs. The resin was then washed with
NMP (3£), THF (3£), CH2Cl2 (3£), and Et2O (3£),
flushed with N2, and then dried under reduced pressure, and
stored at 2208C.
0.570 mequiv./g, 58% loading efficiency based on the
0.98 mequiv./g loading level reported for the purchased
DHP resin.
4.3.3. Synthesis of aspartyl protease inhibitor 9. Support-
bound bromide 5 (,10 mmol) was added to a vial. A 1.0 M
solution of n-butylamine in NMP (1 mL) was then added to
the vial. The vial was sealed and then the reaction mixture
was heated at 808C for 36 h. The reaction mixture was then
transferred to a polypropylene cartridge with a 70 mm PE
frit attached to a Teflon stopcock for the work up process.
The resin was washed with NMP (3£), THF (3£), CH2Cl2
(3£), and ether (3£), and then dried under vacuum.
Acylation with 0.3 M sulfonyl chloride7 and 0.6 M i Pr2EtN
in THF (1.0 mL) was carried out overnight. The resin was
washed with NMP (3£), THF (3£), CH2Cl2 (3£), and ether
(3£), and then dried under vacuum. Reduction of the azide
was accomplished using 0.2 M SnCl2, 0.8 M PhSH, and
1.0 M Et3N in THF (1 mL) for 4 h8. The resin was washed
with 50 vol% aqueous THF solution (3£), THF (3£),
CH2Cl2 (3£) and ether (3£), and then dried under vacuum.
The resulting support-bound amine was then acylated using
1 mL of a THF stock solution that was 0.3 M in the
N-succinimidyl carbonate of 3-(S)-hydroxytetrahydro-
furan9 and 0.6 M in i Pr2EtN. After allowing the acylation
reaction to proceed overnight, the resin was washed with
NMP (4£), THF (2£), CH2Cl2 (3£) and ether (3£). Resin
5 was then treated with 20 vol% piperidine in NMP (1 mL)
for 30 min. The resin was washed with NMP (3£), THF
(3£), CH2Cl2 (3£), and ether (3£) and dried under
vacuum. The resulting resin was treated with 95:5
TFA/H2O for 15 min followed by filtration. The resin was
then rinsed with 95:5 TFA/H2O (1£) and CH2Cl2 (2£). The
combined filtrates were concentrated to dryness. Toluene
was added to azeotrope the TFA during the concentration
step. The mixture was then purified by reverse-phase HPLC
(Varian Microsorb C18 (Si) column R 0089100C5) using a
linear gradient starting with 5:95, 0.1% TFA/acetonitrile:
0.1% TFA/H2O and ending with 0.1% TFA/acetonitrile
over 120 min. Purified product 9 was stored at 2208C.
The loading level of the resin was determined according to
the following procedure. Resin 5a (30.4 mg) was treated
with 95:5 TFA/H2O (1.0 mL). Stirring was carried out by
rotating at room temperature on a wrist action shaker for
15 min. After filtration, the resin was washed with 95:5
TFA/H2O (1£) and CH2Cl2 (2£), and the combined
filtrates were concentrated under reduced pressure to
provide alcohol 4a as colorless oil. The loading level was
then determined by integration of the NMR signals of 4a
against a known amount of p-xylene as an internal standard
(500 MHz, MeOH-d4). The loading level was
0.576 mequiv./g, which corresponds to a 59% loading
efficiency based on the 0.98 mequiv./g loading level
reported for the purchased DHP resin.
4.3.2. Support-bound bromide 5b (R5(CH2)4NHBoc).
To a mixture of 144 mg (0.141 mmol) of DHP resin in
0.5 mL of dichloroethane at room temperature was added
a solution of 74.4 mg (0.212 mmol) of alcohol 4b
(R¼(CH2)4NHBoc) in 0.5 mL of dichloroethane. The
reaction mixture was stirred for 10–12 h at 608C
followed by the addition of 88.5 mg (0.353 mmol) of
PPTs. The reaction mixture was transferred to
a
polypropylene cartridge with a 70 mm PE frit attached
to a Teflon stopcock for the work up process. The resin
was immediately washed with NMP (3£), THF (3£),
CH2Cl2 (3£), and Et2O (3£), flushed with N2, and then
dried under reduced pressure.
Because some of the N-Boc group might be cleaved under
the acidic conditions of the resin-loading step, the resin was
subjected to Boc2O to ensure complete protection. The
reaction vessel was charged with 0.353 M Boc2O (77.0 mg,
0.353 mmol) in THF (1.0 mL) at room temperature
followed by the addition of 61 mL (0.35 mmol) of i Pr2NEt.
Mixing was accomplished by rotating on a wrist-action
shaker at room temperature for 10–12 h. The reaction
solution was then drained and the resin was immediately
washed with NMP (3£), THF (3£), CH2Cl2 (3£), and Et2O
(3£), flushed with N2, dried under reduced pressure, and
stored at 2208C.
4.3.4. Inhibitor 9a (R5Bn, amprenavir (vx 478)). The
solid-phase synthesis procedure was followed starting with
19.1 mg (11.0 mmol) of support-bound bromide 5a.
Reverse-phase HPLC purification gave 3.7 mg (66%
yield) of inhibitor 13a as a white powder: IR (NaCl):
2921, 1708, 1689, 1595, 1314, 1148 cm21 1H NMR
.
(500 MHz, in MeOH-d4): d 0.86–0.91 (m, 6H), 1.28–2.19
(m, 9H), 2.75–3.17 (m, 6H), 3.51–3.91 (m, 7H), 6.70 (d,
J¼8.7 Hz, 2H), 7.47 (d, J¼8.7 Hz, 2H). HRMS (FABMS)
calcd for [MþH]þ (C25H36N3O6S) 506.2325, found
506.2330.
The loading level of the resin was determined according to
the following procedure. Resin 5b (23.9 mg) was treated
with 95:5 TFA/H2O (1.0 mL). Stirring was carried out by
rotating at room temperature on a wrist action shaker for
15 min. After filtration, the resin was washed with 95:5
TFA/H2O (1£) and CH2Cl2 (2£), and the combined
filtrates were concentrated under reduced pressure to
provide alcohol 4b (lacking the Boc group) as colorless
oil. The loading level was then determined by integration of
the NMR signals of 4b (lacking the Boc group) against a
known amount of p-xylene as an internal standard
(500 MHz, MeOH-d4. The loading level was
4.3.5. Inhibitor 9b (R5(CH2)4NHBoc). The solid-phase
synthesis procedure was followed starting with 15.5 mg
(8.8 mmol) of support-bound bromide 5b. Reverse-phase
HPLC purification provided 3.0 mg (70% yield) of inhibitor
13b as a colorless oil: IR (NaCl): 3399, 1678, 1203,
1139 cm21. 1H NMR (500 MHz, in MeOH-d4): d 0.72–0.98
(m, 6H), 1.53–2.33 (m, 5H), 2.53–3.20 (m, 5H), 3.42–3.87
(m, 5H), 4.93–5.09 (m, 1H), 6.67–6.70 (m, 1H), 7.16–7.79
(m, 8H). HRMS (FABMS) calcd for [MþNa]þ (C22H38-
N4ONaO6S) 509.2410, found 509.2397.

6310
M. Chino et al. / Tetrahedron 58 (2002) 6305–6310
A.; Sun, Y.; Kuntz, I. D.; Ellman, J. A. Chem. Biol. 1997, 4,
297–309.
Acknowledgments
9. Bi, X.; Lin, B.; Haque, T.; Lee, C. E.; Skillman, A. G.; Kuntz,
I. D.; Ellman, J. A.; Lynch, G. J. Neurochem. 2000, 74,
1469–1477.
This research was supported by NIH grant GM50451.
Support for M. C. from Mitsubishi Pharma Corporation is
also gratefully acknowledged. M. W. thanks the Japan
Society for the Promotion of Science (JSPS Fellows No.
01798).
10. Garcia, M.; Platet, N.; Liaudet, E.; Laurent, V.; Derocq, D.;
Brouillet, J. P.; Rochefort, H. Stem Cells 1996, 14, 642–650.
11. Haque, T. S.; Skillman, A. G.; Lee, C. E.; Habashita, H.;
Gluzman, I. Y.; Ewing, T. J. A.; Goldberg, D. E.; Kuntz, I. D.;
Ellman, J. A. J. Med. Chem. 1999, 42, 1428–1440.
12. Olsen, V.; Cawley, N. X.; Brandt, J.; Egel-Mitani, M.; Loh,
Y. P. Biochem. J. 1999, 339, 407–411.
References
1. Macdonald, S. J. F.; Smith, P. W. Drug Discov. Today 2001, 6,
947–953.
13. Alper, P. B.; Hung, S.-C.; Wong, C.-H. Tetrahedron Lett.
1996, 34, 6029–6032.
2. Maly, D. J.; Huang, L.; Ellman, J. A. ChemBiolChem. 2002, 3,
16–37.
14. Lundquist, IV J. T.; Pelletier, J. C. Org. Lett. 2001, 3,
781–783.
3. Lee, C. E.; Kick, E. K.; Ellman, J. A. J. Am. Chem. Soc. 1998,
20, 9735–9748.
15. Thompson, L. A.; Ellman, J. A. Tetrahedron Lett. 1994, 35,
9333–9336.
4. Kick, E. K.; Ellman, J. A. J. Med. Chem. 1995, 38,
1427–1430.
16. Kim, E. E.; Baker, C. T.; Dwyer, M. D.; Murcko, M. A.; Rao,
B. G.; Tung, R. D.; Navia, M. A. J. Am. Chem. Soc. 1995, 117,
1181–1182.
5. For an excellent summary of aspartyl protease targets and
inhibitors see: Leung, D.; Abbenante, G.; Fairlie, D. P. J. Med.
Chem. 2000, 43, 305–341.
17. Bartra, M.; Romea, P.; Urpi, G.; Vilarrasa, J. Tetrahedron
1990, 46, 587–594.
18. Lombardi, P. Chem. Ind. 1990, 708.
6. Thompson, L. A.; Tebben, A. J. Annu. Rep. Med. Chem. 2001,
36, 247–256.
19. N-Fmoc-4-aminobenzenesulfonyl chloride was prepared
according to the following literature procedure: Jameson,
G. W.; Elmore, D. T. Biochem. J. 1974, 141, 555–565.
7. Wolfe, M. S. J. Med. Chem. 2001, 44, 2039–2060.
8. Kick, E. K.; Roe, D. C.; Skillman, A. G.; Liu, G.; Ewing, T. J.