Synthesis and activity of tetrapeptidic HTLV-I protease inhibitors possessing different P3-cap moieties

Article abstract of DOI:10.1016/j.bmc.2008.03.055

The causative agent behind adult T-cell leukemia and tropical spastic paraparesis/HTLV-I-associated myelopathy is the human T-cell leukemia virus type 1 (HTLV-I). Tetrapeptidic HTLV-I protease inhibitors were designed on a previously reported potent inhibitor KNI-10516, with modifications at the P₃-cap moieties. All the inhibitors showed high HIV-1 protease inhibitory activity (over 98% inhibition at 50 nM) and most exhibited highly potent inhibition against HTLV-I protease (IC₅₀ values were less than 100 nM).

Full text of DOI:10.1016/j.bmc.2008.03.055

Available online at www.sciencedirect.com
Bioorganic & Medicinal Chemistry 16 (2008) 5795–5802
Synthesis and activity of tetrapeptidic HTLV-I protease
inhibitors possessing different P3-cap moieties
Meihui Zhang,a,b Jeffrey-Tri Nguyen,a Henri-Obadja Kumada,a Tooru Kimura,a
Maosheng Cheng,b Yoshio Hayashia,c and Yoshiaki Kisoa,*
aDepartment of Medicinal Chemistry, Center for Frontier Research in Medicinal Science and 21st Century COE Program,
Kyoto Pharmaceutical University, Yamashina-ku, Kyoto 607-8412, Japan
bKey Laboratory of New Drugs Design and Discovery of Liaoning Province, School of Pharmaceutical Engineering,
Shenyang Pharmaceutical University, Shenyang 110016, PR China
cDepartment of Medicinal Chemistry, Tokyo University of Pharmacy and Life Sciences, Hachioji, Tokyo 192-0392, Japan
Received 18 February 2008; revised 21 March 2008; accepted 22 March 2008
Available online 27 March 2008
Abstract—The causative agent behind adult T-cell leukemia and tropical spastic paraparesis/HTLV-I-associated myelopathy is the
human T-cell leukemia virus type 1 (HTLV-I). Tetrapeptidic HTLV-I protease inhibitors were designed on a previously reported
potent inhibitor KNI-10516, with modifications at the P3-cap moieties. All the inhibitors showed high HIV-1 protease inhibitory
activity (over 98% inhibition at 50 nM) and most exhibited highly potent inhibition against HTLV-I protease (IC50 values were less
than 100 nM).
ꢀ 2008 Elsevier Ltd. All rights reserved.
1. Introduction
with a median survival time of 13 months in aggressive
cases.7 The poor prognosis for ATL patients is partly
due to the innate resistance of HTLV-I-infected T-cells
to apoptosis and thus to conventional chemotherapy
regimens.8 At the present time, there is no effective treat-
ment for ATL and HTLV-I infection. Therefore, cura-
tive therapies against ATL are urgently needed.
Human T-cell leukemia virus type 1 (HTLV-I) is the eti-
ologic agent for adult T-cell leukemia (ATL) and other
chronic inflammatory diseases, such as tropical spastic
paraparesis/HTLV-I-associated
myelopathy
(TSP/
HAM) and HTLV-I-associated uveitis.1 HTLV-I causes
ATL, an invariably fatal expansion of virus-infected
CD4+ T cells, in 2–6% of infected individuals after a
long period of 20–50 years.2 HTLV-I infections are en-
demic in high-risk groups in South Japan, the Caribbean
Basin, South America, Central and West Africa, the
Middle East, and the Pacific Region.3 Currently, it is
estimated that 20–30 million people worldwide are
infected with HTLV-I.4 All routes of HTLV-I virus
transmission require close contact with infected T-lym-
phocytes. Three main routes of HTLV-I transmission
have been recognized, including transmission from
mother to child through breast-feeding, sexual, and
blood–borne transmissions. In Japan, where 15–25%
individuals are carriers, the infection is spread mainly
via breast milk.5,6 The prognosis of ATL patients is poor
HTLV-I was the first human retrovirus to be identified
and isolated in the early 1980s from patients with
ATL.9,10 In common with other retroviruses, HTLV-I
encodes a virus-specific aspartic protease responsible
for the proliferation of the retrovirus. HTLV-I protease
is a small homodimer composed of two identical sub-
units containing 125 amino acids each.11–13 As in most
human retroviruses, there are three large open reading
frames in the genome of HTLV-I that encode for the
Gag (48 kDa), Pol (99 kDa), and Env (54 kDa) polypro-
teins, and a number of spliced, open reading frames that
encode for short regulatory proteins (Rex, Tax, etc.).
HTLV-I proteolytic cleavage of the HTLV-I Gag pre-
cursor polyprotein yields matrix (MA), capsid (CA)
and nucleocapsid (NC) proteins of the mature virion,
while the cleavage of the Pol precursor polyprotein af-
fords reverse transcriptase/ribonuclease H (RT–RH)
and integrase (IN). Because of its crucial role in viral
Keywords: HTLV-I; ATL; HIV-1; Aspartic protease inhibitor.
*
Corresponding author. Tel.: +81 75 595 4635; fax: +81 75 591
9900; e-mail: kiso@mb.kyoto-phu.ac.jp
0968-0896/$ - see front matter ꢀ 2008 Elsevier Ltd. All rights reserved.
doi:10.1016/j.bmc.2008.03.055

5796
M. Zhang et al. / Bioorg. Med. Chem. 16 (2008) 5795–5802
maturation, HTLV-I protease is an attractive target for
the development of inhibitors to treat HTLV-I infection.
of HTLV-I protease inhibitors to improve the chance
for cellular penetration, and reduce the peptidic nature
of the inhibitors so as to avoid premature digestion by
other proteases. Thus far, the P3-cap moiety has not
been investigated to optimize the efficiency of inhibition.
Our past study showed that the replacement of a proline
P4 moiety by an acetyl P3-cap moiety significantly de-
creased HTLV-I protease inhibitory activity.20 In the
work described herein, using compound 5a as a starting
reference compound, we explored the significance of the
capping moiety of the P3 residue on HTLV-I protease
inhibition.
Although HTLV-I protease is in the same family of
aspartic proteases as HIV-1 protease, HTLV-I protease
has received little attention compared to its more fa-
mous cousin HIV-1 protease that has been extensively
studied.14 Due to 45% sequence similarity between
HTLV-I and HIV-1 proteases at the active site, some
HIV-1 protease inhibitors had been tested on HTLV-I
protease.15–17 However, differences in the amino acid se-
quences of HIV-1 and HTLV-I proteases produce differ-
ences in susceptibilities to these inhibitors.
Consequently, we decided to design HTLV-I protease
inhibitors on the sequence of a Gag substrate (1) that
is processed at the matrix-capsid cleavage site,18 and
synthesized potent octapeptidic HTLV-I protease inhib-
itor KNI-10161 (2, IC50 = 159 nM) that possessed
(2S,3S)-3-amino-2-hydroxy-4-phenylbutyric acid (all-
ophenylnorstatine, Apns) with a hydroxymethylcarbon-
2. Chemistry
The synthesized HTLV-I protease inhibitors 9 and 5a–p
were prepared by standard solution phase peptide cou-
pling chemistry (Scheme 1). Isobutylamine was coupled
with Na- tert-butyloxycarbonyl (R)-5,5-dimethyl-1,3-thi-
azolidine-4-carboxylic acid (Boc-Dmt-OH) using ben-
zotriazol-1-yloxy-tris(dimethylamino)phosphonium hexa-
fluorophosphate (BOP) in the presence of triethylamine
(TEA), followed by Boc-deprotection with 4 N HCl in
dioxane to obtain compound 6. Compound 7 was pre-
pared by coupling compound 6 and Boc-Apns-OH with
BOP. The same deprotection/condensation procedure
was repeated for the successive introduction of Boc-Tle-
OH and Boc-Phg-OH. In the last step, inhibitors 5a–p
were obtained by coupling compound 9 with their respec-
tive carboxylic acids. Compounds 5k–l are racemates.
yl isostere as a P1–P1 transition-state mimic (Fig. 1).19
0
0
The replacement of the P1 Pro residue by (R)-5,5-di-
methyl-1,3-thiazolidine-4-carboxylic acid (Dmt) fol-
lowed by truncation and natural amino acid
substitution studies performed on compound 2 led to a
more potent hexapeptidic HTLV-I protease inhibitor
KNI-10166 (3, IC50 = 88 nM).20 Compound 3 was fur-
ther refined by non-natural amino acid isostere substitu-
tion with L-(+)-a-phenylglycine (Phg) and L-tert-leucine
(Tle), and by a truncation study to uncover pentapepti-
dic inhibitor KNI-10247 (4, IC50 = 144 nM).21 The
0
0
replacement of the P2 Ile residue by a P1 -cap moiety in-
creased HTLV-I protease inhibitory activity in tetrapep-
tidic inhibitor KNI-10516 (5a, IC50 = 107 nM). In our
previous publications, we attempted to minimize the size
3. Results and discussion
HTLV-I protease inhibition assay was performed using
L40I HTLV-I protease mutant as previously reported.21
Enzymatic assay data and computer assisted docking
experiments0 from our previous studies suggested that
the P3-to-P1 residues were strong determinants of inhib-
itory activity against HTLV-I protease.20,21 In the cur-
rent study, we described our efforts to further optimize
KNI-10516-related compounds by exploring modifica-
tions at the P3-cap position. We were particularly inter-
ested in studying the effect of attaching various moieties
to the P3 amide nitrogen atom. An X-ray crystal struc-
ture of a truncated, L40I mutant HTLV-I protease in
complex with an inhibitor revealed that the S4 subsite,
where the P3-cap moiety would presumably interact,
would accommodate for hydrophobic residues.22 Inter-
estingly, similar associations had been made for peptidic
substrates and the S4 subsite.12 In consideration of these
facts, we opted to synthesize HTLV-I protease inhibitors
that possess hydrophobic P3-cap moieties. We believed
that van der Waals interactions play a key role in the
S4 region of the active site, and that finding the perfect
‘fit’ would increase inhibitory activity against the
enzyme.
Several interesting structure–activity relationships were
observed with hydrogen bond interactions, chain length,
branching, and bulkiness (Table 1). Most interestingly,
the absence of a P3-cap moiety in compound 9 greatly
Figure 1. Processive development of potent HTLV-I protease inhib-
itors, KNI-10161 (2), KNI-10166 (3), KNI-10247 (4) and KNI-10516
(5a), from a MA/CA substrate (1).

M. Zhang et al. / Bioorg. Med. Chem. 16 (2008) 5795–5802
5797
Scheme 1. Synthesis of compounds 9 and 5a–p. Reagents and conditions: (a) Boc-Dmt-OH, BOP, TEA, DMF, rt, 15 h; (b) 4 N HCl/dioxane, rt, 1 h;
(c) Boc-Apns-OH, BOP, TEA, DMF, rt, 15 h; (d) Boc-Tle-OH, BOP, TEA, DMF, rt, 15 h; (e) Boc-Phg-OH, BOP, TEA, DMF, rt, 15 h; (f) RCOOH,
BOP, TEA, DMF, rt, 15 h.
Table 1. Structures and activity of P3-substituted HTLV-I Protease
Inhibitors 9 and 5a–p
compounds 5d (chain length) and 5f (a-carbon branch-
ing), whereas compound 5l exhibited higher inhibitory
activity than compounds 5d (chain length) and 5h (b-
Compound
R
HTLV-I protease
inhibition,
IC50 SEM (nM)
carbon branching). Comparing compounds 5f (a-carbon
branching) to 5h (b-carbon branching) suggested that a-
carbon branching was a greater determinant of inhibi-
tory activity than b-carbon branching, which was also
confirmed by compounds 5j (a-carbon branching) and
5i (b-carbon branching), as well as compounds 5k
(a-carbon branching) and 5l (b-carbon branching).
However, increasing bulkiness at the c-carbon reduced
inhibitory activity against HTLV-I protease, as exempli-
fied by compounds 5d, 5m and 5n. Consequently, a-car-
bon branching effect was a greater determinant of
HTLV-I protease inhibitory activity than b-branching,
which in its turn, was greater than c-branching. Hence,
in agreement with the general trend observed with pep-
tidic substrates,12 we observed, by comparing a-, b- and
c-carbon branching effects, that the distance to the
9
KNI-10673
Not
applicable
331.6 0.5
5a
5b
5c
5d
5e
5f
KNI-10516
KNI-10571
KNI-10570
KNI-10616
KNI-10575
KNI-10574
KNI-10572
KNI-10573
KNI-10633
KNI-10614
KNI-10615
KNI-10618
KNI-10617
KNI-10652
KNI-10651
KNI-10634
Me
Et
107.1 0.6
104.5 0.5
104.3 0.5
92.0 0.5
96.9 0.6
93.5 0.5
82.8 0.6
101.7 0.5
83.3 0.6
80.4 1.3
84.2 1.3
88.2 0.5
96.5 0.5
104.8 0.6
98.2 0.5
79.2 0.6
nPr
nBu
cPr
iPr
5g
5h
5i
tBu
iBu
tPe
5j
C(Me2)Et
aCH(Me)nPr
CH2aCH(Me)Et
iPe
5k
5l
5m
5n
5o
5p
0
CH2CH2Ph
cHx
Ph
P1–P1 catalytic site was inversely proportional to the ex-
tent of change in HTLV-I protease inhibitory activity.
Lastly, we examined the effect of introducing a six-mem-
ber ring system at the a-carbon and observed that a
compound with a P3-cap cyclohexyl ring (5o) exhibited
equipotency to that of a cyclopropyl ring (5e), but lower
potency than compound 5k. On the other hand, com-
pound 5p with a P3-cap phenyl ring seemed to provide
a better inhibitory profile than compounds 5e and 5k.
This trend suggested that a P3-cap phenyl ring moiety
(5p) provided a more appropriate bulkiness than a
cyclohexyl ring moiety (5o). Looking at the inhibitory
profiles, although we hypothesized various structure–
activity relationships from this group of compounds,
the reliability of the trends is still in question considering
that all of the amide P3-cap moieties (5a–p) were extre-
mely potent HTLV-I protease inhibitors with little vari-
ations in IC50 values, from 79 to 107 nM, although the
standard-error-of-the-mean was only 1 nM. Nonethe-
less, this observation suggested that the nature of the
amide P3-cap moieties was a lower determinant of inhib-
itory activity. Computer-assisted docking experiments
revealed that the pocket, where the P3-cap moiety re-
sides, was near the exposed surface of the enzyme
(Fig. 2A), and thereby, other than a hydrogen bond
interaction with the amide oxygen (cf. 9), would have
a Racemate.
reduced the inhibitory activity against HTLV-I protease
(cf. 5a–p), because, according to computer-assisted
docking experiments, the amide oxygen from the P3-
cap moiety might be involved in hydrogen bond interac-
tion with Leu57A from the flap of the HTLV-I protease
(Fig. 2B). We also examined the effect of P3-cap chain
elongation and observed that the inhibitory profiles of
compounds 5a–d suggested that HTLV-I protease inhib-
itory potency might improve proportionally with chain
length. Intuitively, we explored branching on the P3-
cap moiety, and according to compounds 5b,e–g, methyl
branching at the a-carbon next to the amide carbonyl
seemed to increase HTLV-I protease inhibitory activity,
whereas methyl branching at the b-carbon also showed a
trend with increased inhibitory potency as exemplified
by compounds 5c,h,i. To confirm the hypothesized
structure–activity relationships on the P3-cap moiety,
compound 5j showed a shared benefit of chain length
(cf. 5c) and a-carbon branching (cf. 5g). From the same
line of reasoning, compound 5k was more potent than

5798
M. Zhang et al. / Bioorg. Med. Chem. 16 (2008) 5795–5802
Figure 2. Computer model of KNI-10634 (5p) in the active site of HTLV-I protease. Dotted lines represent possible hydrogen bond interactions. (A)
Portions of the P3 moiety residue near the exposed surface of the HTLV-I protease. (B) The surface of some HTLV-I protease heavy atoms is hidden
to clearly depict various hydrogen bond interactions.
a lesser influence on inhibitory activity. To confirm the
speculative structure–activity relationships, various
groups of compounds, where the P1 -cap and P3-cap
moieties, respectively, differ from each other, are cur-
rently being evaluated so that structure–activity rela-
tionship trends can be statistically validated.
in the amino acid sequences of HIV-1 and HTLV-I pro-
teases do indeed produce differences in susceptibilities
to inhibitors, in that HIV-1 protease is more susceptible
to inhibitors than HTLV-I protease. On a side note, our
past studies23–27 have shown that HIV-1 protease is also
less specific than Plasmepsin II, an aspartic protease
associated with malaria. These observations also imply
that should the need arise, HTLV-I protease inhibitors
and Plasmepsin inhibitors could become lead com-
pounds in the design of novel HIV-1 protease
inhibitors.
0
A computer-assisted docking model was generated for
one of the more potent HTLV-I protease inhibitor, com-
pound 5p, described in the current study (Fig. 2). In the
model, multiple possible hydrogen bond interactions ex-
isted throughout the backbone of compound 5p and
HTLV-I protease’s Asp32A, Asp32B, Gly34A, Asp36A,
Leu57A, Ala59A, and Ala59B of the dimer, and thereby
anchoring the inhibitor in its docked position within the
HTLV-I protease (Fig. 2B). Similar to HIV-1 and
Plasmepsin II protease inhibitors, a water molecule
mediated the hydrogen bond interactions between
Ala59A and Ala59B in the hairpin regions of HTLV-I
protease’s flaps and the inhibitor. Moreover, compound
5p’s transition-state mimic hydroxymethylcarbonyl isos-
teric moiety interacted with the protease’s catalytic
Asp32A and Asp32B. As a result, a hydrogen bond
network was formed in the vicinity of the critical
4. Conclusions
In the present study, small peptide-based P3-cap substi-
tuted HTLV-I protease inhibitors containing
a
hydroxymethylcarbonyl isostere were designed and syn-
thesized on the basis of the substrate transition-state
mimicry. The HTLV-I protease profiles of these new
and extremely potent HTLV-I protease inhibitors
showed chemical structure dependence at the P3-cap
moiety, especially the hydrogen bond interaction involv-
ing Leu57A. These inhibitors also exhibited high
potency against HIV-1 protease. The overall most
potent HTLV-I protease inhibitor within this study
was KNI-10634 (5p, IC50 = 79 nM).
0
P1–P1 residues. As shown in Figure 2A, the P3 residue’s
phenyl side-chain fitted snugly in the S3 pocket, while
the P3-cap moiety resided near the edge of the active site
in a region where various hydrophobic groups could be
accommodated.
5. Experimental
5.1. Chemistry
All synthesized inhibitors exhibited nearly complete
anti-HIV-1 protease activity at 50 nM (P99% inhibi-
tion), and interestingly, our results reflected the obser-
´
vations made by To¨zser and co-workers that HTLV-I
protease expresses higher specificity than its HIV-1 pro-
5.1.1. General methods. Reagents and solvents were pur-
chased from commercial suppliers and used without fur-
1
ther purification. H NMR spectra were measured on a
tease cousin.15 Whereas the To¨zser research group re-
´
ported narrower activation profiles from peptidic
substrates, we observed narrower inhibition profile
from our HTLV-I protease inhibitors. In simpler words,
in agreement with our past reports,20,21 in general, po-
tent HTLV-I protease inhibitors exhibit extremely po-
tent HIV-1 protease inhibition, whereas potent HIV-1
protease inhibitors do not necessarily exhibit potent
HTLV-I protease inhibitory activity. Hence, differences
JEOL JNM-AL300 FT (300 MHz) spectrometer with
tetramethylsilane (TMS) as an internal standard at
ambient temperature. Data are reported as follows:
chemical shift (ppm), multiplicity (s = singlet, d = dou-
blet, t = triplet, q = quartet, dd = double doublet,
bs = broad singlet, m = multiplet) and number of pro-
tons. LC MS was recorded on a Finnigan SSQ-7000
spectrometer. MALDI-TOF MS was performed on a

M. Zhang et al. / Bioorg. Med. Chem. 16 (2008) 5795–5802
5799
Voyager-DE RP spectrometer and all target compounds
were less than 1 m/z from calculated values. Preparative
HPLC was obtained from a C18 reverse phase column
(20 · 250 mm; YMC Pack ODS SH-343-5AM). Assay
HPLC was performed on a C18 reverse phase column
(3.0 · 75 mm; YMC Pack ODS AS-3E7). Analytical
HPLC was performed using a C18 reverse phase column
(4.6 · 150 mm; YMC Pack ODS AM-302) with a binary
solvent system: linear gradient of CH3CN in 0.1% aque-
ous trifluoroacetic acid (TFA) at a flow rate of 0.9 mL/
min, detected at UV 230 nm. The purity of the desired
compounds and intermediates was confirmed by analyt-
ical HPLC and was greater than 99% in two different
systems (gradient: system A, 0–100%/ 40 min; system
B, 30–90%/ 40 min).
5.1.6. (R)-N-Isobutyl-3-{(2S,3S)-3-[(2S)-2-amino-3,3-di-
methyl]butanoylamino-2-hydroxy-4-phenyl}butanoyl-5,5-
dimethyl-1,3-thiazolidine-4-carboxamide (H-Tle-Apns-
Dmt-NHiBu, 8). Compound 8 was prepared from Boc-
Tle-OH and compound 7 by general method A and B.
Yield: 81%; 1H NMR (DMSO-d6): d = 0.67–0.99 (m,
5 · 3H), 1.37 (s, 3H), 1.51 (s, 3H), 1.70–1.77 (m, 1H),
2.65–2.80 (m, 2H), 2.82–3.10 (m, 2H), 3.46 (bs, 1H),
3.5 (bs, 1H), 4.11 (bs, 1H), 4.40–4.57 (m, 2 · 1H),
4.90–5.04 (m, 2H), 5.32 (bs, 2H), 6.90–7.40 (m, 5H),
8.17 (t, 1H), 8.52 (d, 1H); MS (TOF) m/z = 507.305
[M+H]+; Analytical HPLC: Rt = 22.27 min (system A).
5.1.7. (R)-N-Isobutyl-3-{{(2S,3S)-3-{(2S)-2-[(2S)-2-amino-
2-phenyl]acetylamino-3,3-dimethyl}butanoylamino-2-hydro-
xy-4-phenylbutanoyl}}-5,5-dimethyl-1,3-thiazolidine-4-
carboxamide (H-Phg-Tle-Apns-Dmt-NHiBu, 9). Com-
pound 9 was prepared from Boc-Phg-OH and com-
5.1.2. General method A: amide bond formation. To a
solution of substituted amine (1.0 mmol) in N,N-di-
methyl formamide (DMF, 5 mL) were added Boc-Xaa-
OH or carboxylic acid (1.1 mmol), BOP (1.1 mmol),
and Et3N (adjusted to pH 8), and stirred for 15 h at
rt. After removal of the solution in vacuo, the residue
was dissolved in EtOAc (20 mL), washed with 10% citric
acid aq (3 · 15 mL), 10% NaHCO3 aq (3 · 15 mL) and
brine (1 · 15 mL), dried over anhydrous Na2SO4, and
concentrated in vacuo to give the crude product which
was used directly to the next step without purification,
or purified by preparative HPLC in the case of the target
compounds. The purified target compounds were imme-
diately lyophilized to afford their respective white amor-
phous powder.
1
pound 8 by general method A and B. Yield: 77%; H
NMR (DMSO-d6): d = 0.51–0.98 (m, 5 · 3H), 1.31 (s,
3H), 1.46 (s, 3H), 1.65–1.76 (m, 1H), 2.50–2.78 (m,
2H), 2.80–2.95 (m, 2H), 3.5 (bs, 1H), 4.11 (bs, 1H),
4.20–4.33 (m, 2 · 1H), 4.38 (s, 1H), 4.86 (d, 1H), 4.98
(d, 1H), 5.07 (bs, 1H), 6.90–7.12 (m, 5H), 7.30–7.40
(m, 5H), 7.91 (t, 1H), 8.03 (d, 1H), 8.45 (d, 1H), NH2
not observed; MS (TOF) m/z = 640.012 [M+H]+,
663.239 [M+Na]+; Analytical HPLC: Rt = 24.36 min
(system A), 12.11 min (system B).
5.1.8. (R)-N-Isobutyl-3-{{(2S,3S)-3-{(2S)-2-[(2S)-2-acetyl-
amino-2-phenyl]acetylamino-3,3-dimethyl}butanoylamino-
2-hydroxy-4-phenyl}}butanoyl-5,5-dimethyl-1,3-thiazoli-
dine-4-carboxamide (Ac-Phg-Tle-Apns-Dmt-NHiBu, 5a).
Compound 5a was prepared from acetic acid and com-
5.1.3. General method B: Boc-deprotection. 4 N HCl in
1,4-dioxane (5 mL) was added to the afforded crude
Boc-protected product (ca. 1.0 mmol) and stirred for
1 h at rt. The solution was evaporated under reduced
pressure and used directly to the next step without puri-
fication. A small portion of the product was purified by
preparative HPLC. The desired fractions were collected
and immediately lyophilized to afford white amorphous
powder, compound 6, 7, 8, or 9.
1
pound 9 by general method A. Yield: 60%; H NMR
(CDCl3)d = 0.70–0.92 (m, 5 · 3H), 1.40–1.58 (m,
2 · 3H), 1.63–1.80 (m, 1H), 2.06 (s, 3H), 2.2–2.6 (bs,
1H), 2.58–2.85 (m, 2H), 2.88–3.22 (m, 2H), 4.04–4.96
(m, 4 · 1H + 2H), 5.58 (t, 1H), 6.32 (d, 1H), 6.35–6.40
(bs, 1H), 6.66 (t, 1H), 6.90–7.24 (m, 5H), 6.98 (d, 1H),
7.26–7.35 (m, 5H); MS (TOF) m/z = 704.762 [M+Na]+,
720.763 [M+K]+; Analytical HPLC: Rt = 28.35 min
(system A), 17.94 min (system B).
5.1.4. (R)-N-Isobutyl-5,5-dimethyl-1,3-thiazolidine-4-car-
boxamide (H-Dmt-NHiBu, 6). Compound 6 was pre-
pared from Boc-Dmt-OH and isobutylamine by
general methods A and B. Yield: 98%; 1H NMR
(CDCl3): d = 0.90 (d, 2 · 3H), 1.42 (s, 3H), 1.64 (s,
3H), 1.77–1.86 (m, 1H), 3.12 (t, 2H), 4.35 (s, 1H), 4.53
(s, 2H), 6.4 (bs, 1H), 7.46 (t, 1H); MS (LC) m/z =
216.89 [M+H]+; Analytical HPLC: Rt = 16.34 min
(system A).
5.1.9. (R)-N-Isobutyl-3-{{(2S,3S)-3-{(2S)-2-[(2S)-2-propio-
nylamino-2-phenyl]acetylamino-3,3-dimethyl}butanoyl-
amino-2-hydroxy-4-phenyl}}butanoyl-5,5-dimethyl-1,3-
thiazolidine-4-carboxamide (EtCO-Phg-Tle-Apns-Dmt-
NHiBu, 5b). Compound 5b was prepared from propionic
acid and compound 9 by general method A. Yield: 65%;
1H NMR (CDCl3)d = 0.78–0.92 (m, 5 · 3H), 1.16 (t,
3H), 1.40–1.57 (m, 2 · 3H), 1.62–1.80 (m, 1H), 2.2–2.7
(bs, 1H), 2.30 (q, 2H), 2.62–2.85 (m, 2H), 2.89–3.24 (m,
2H), 4.05–5.02 (m, 4 · 1H + 2H), 5.53–5.60 (m, 1H),
6.32 (d, 1H), 6.45–6.55 (bs, 1H), 6.70 (t, 1H), 6.90 (d,
1H), 6.92–7.21 (m, 5H), 7.26–7.35 (m, 5H); MS (TOF)
m/z = 718.987 [M+Na]+; Analytical HPLC: Rt =
29.64 min (system A), 20.26 min (system B).
5.1.5.
(R)-N-Isobutyl-3-[(2S,3S)-3-amino-2-hydroxy-4-
phenyl]butanoyl-5,5-dimethyl-1,3-thiazolidine-4-carboxam-
ide (H-Apns-Dmt-NHiBu, 7). Compound 7 was prepared
from Boc-Apns-OH and compound 6 by general method
1
A and B. Yield: 59%; H NMR (CDCl3): d = 0.80–0.91
(m, 2 · 3H), 1.41–1.51 (m, 2 · 3H), 1.70–1.78 (m, 1H),
2.1 (bs, 1H), 2.94–3.02 (m, 2H), 3.03–3.19 (m, 2H),
3.75–3.88 (bs, 1H), 4.22–4.35 (m, 2 · 1H), 4.46 (d, 1H),
4.86 (d, 1H), 7.00–7.14 (bs, 2H), 7.24–7.35 (m, 5H),
7.9–8.2 (bs, 1H); MS (TOF) m/z = 394.615 [M+H]+; Ana-
lytical HPLC: Rt = 21.66 min (system A).
5.1.10. (R)-N-Isobutyl-3-{{(2S,3S)-3-{(2S)-2-[(2S)-2-buty-
rylamino-2-phenyl]acetylamino-3,3-dimethyl}butanoylami-
no-2-hydroxy-4-phenyl}}butanoyl-5,5-dimethyl-1,3-thiazol-
idine-4-carboxamide (nPrCO-Phg-Tle-Apns-Dmt-NHiBu,

5800
M. Zhang et al. / Bioorg. Med. Chem. 16 (2008) 5795–5802
5c). Compound 5c was prepared from n-butyric acid and
compound 9 by general method A. Yield: 62%; 1H NMR
(CDCl3)d = 0.78–0.97 (m, 6 · 3H), 1.40–1.58 (m,
2 · 3H), 1.60–1.76 (m, 1H + 2H), 1.8–2.2 (bs, 1H), 2.24
(t, 2H), 2.66–2.84 (m, 2H), 2.86–3.23 (m, 2H), 4.04–4.99
(m, 4 · 1H + 2H), 5.51 (d, 1H), 6.15–6.25 (bs, 1H), 6.26–
6.50 (bs, 1H), 6.66 (t, 1H), 6.73–6.82 (m, 1H), 6.91–7.25
(m, 5H), 7.26–7.36 (m, 5H); MS (TOF) m/z = 710.984
[M+H]+, 732.931 [M+Na]+, 748.858 [M+K]+; Analytical
HPLC: Rt = 30.56 min (system A), 22.19 min (system B).
NHiBu, 5g). Compound 5g was prepared from pivalic
acid and compound 9 by general method A. Yield:
59%; H NMR (CDCl3)d = 0.75–0.92 (m, 5 · 3H), 1.22
(s, 3 · 3H), 1.48–1.58 (m, 2 · 3H), 1.69–1.80 (m, 1H),
2.0–2.4 (bs, 1H), 2.62–2.81 (m, 2H), 2.85–3.26 (m,
2H), 3.99–5.06 (m, 4 · 1H + 2H), 5.42–5.46 (m, 1H),
6.21 (d, 1H), 6.29 (d, 1H), 6.72 (t, 1H), 6.86 (d, 1H),
6.92–7.25 (m, 5H), 7.26–7.37 (m, 5H); MS (TOF)
m/z = 725.045 [M+H]+, 747.044 [M+Na]+, 763.027
[M+K]+; Analytical HPLC: Rt = 32.27 min (system A),
25.20 min (system B).
1
5.1.11. (R)-N-Isobutyl-3-{{(2S,3S)-3-{(2S)-2-[(2S)-2-penta-
noylamino-2-phenyl]acetylamino-3,3-dimethyl}butanoylami-
no-2-hydroxy-4-phenyl}}butanoyl-5,5-dimethyl-1,3-thiazoli-
dine-4-carboxamide (nBuCO-Phg-Tle-Apns-Dmt-NHiBu,
5d). Compound 5d was prepared from valeric acid and
compound 9 by general method A. Yield: 59%; 1H
NMR (CDCl3)d = 0.71–0.94 (m, 6 · 3H), 1.30–1.37 (m,
2H), 1.47–1.64 (m, 2H + 2 · 3H), 1.6–2.0 (bs, 1H), 1.70–
1.84 (m, 1H), 2.26 (t, 2H), 2.60–2.83 (m, 2H), 2.90–3.23
(m, 2H), 4.00–4.99 (m, 4 · 1H + 2H), 5.47–5.51 (m,
1H), 6.10–6.22 (m, 1H), 6.22–6.35 (bs, 1H), 6.65–6.71
(m, 1H), 6.72–6.80 (m, 1H), 6.91–7.25 (m, 5H), 7.26–
7.36 (m, 5H); MS (TOF) m/z = 724.977 [M+H]+,
746.951 [M+Na]+, 762.969 [M+K]+; Analytical HPLC:
Rt = 32.30 min (system A), 24.36 min (system B).
5.1.15.
(R)-N-Isobutyl-3-{{(2S,3S)-3-{(2S)-2-[(2S)-2-(3-
methyl)butanoylamino-2-phenyl]acetylamino-3,3-dimethyl}-
butanoylamino-2-hydroxy-4-phenyl}}butanoyl-5,5-dimethyl-
1,3-thiazolidine-4-carboxamide (iBuCO-Phg-Tle-Apns-Dmt-
NHiBu, 5h). Compound 5h was prepared from isovaleric
acid and compound 9 by general method A. Yield: 56%;
1H NMR (CDCl3)d = 0.77–0.96 (m, 7 · 3H), 1.47–1.58
(m, 2 · 3H), 1.61–1.78 (m, 1H), 2.0–2.4 (bs, 1H), 2.08–
2.15 (m, 1H + 2H), 2.59–2.78 (m, 2H), 2.82–3.22 (m,
2H), 4.04–5.02 (m, 4 · 1H + 2H), 5.54–5.57 (m, 1H),
6.27 (d, 1H), 6.40 (d, 1H), 6.69 (t, 1H), 6.84 (d, 1H),
6.90–7.25 (m, 5H), 7.26–7.35 (m, 5H); MS (TOF) m/
z = 724.933 [M+H]+, 746.940 [M+Na]+, 762.912
[M+K]+; Analytical HPLC: Rt = 31.82 min (system A),
23.86 min (system B).
5.1.12. (R)-N-Isobutyl-3-{{(2S,3S)-3-{(2S)-2-[(2S)-2-cyclo-
propanecarbonylamino-2-phenyl]acetylamino-3,3-dimethyl}-
butanoylamino-2-hydroxy-4-phenyl}}butanoyl-5,5-dimethyl-
1,3-thiazolidine-4-carboxamide (cPrCO-Phg-Tle-Apns-
Dmt-NHiBu, 5e). Compound 5e was prepared from
cyclopropanecarboxylic acid and compound 9 by
5.1.16. (R)-N-Isobutyl-3-{{(2S,3S)-3-{(2S)-2-[(2S)-2-(3,3-
dimethyl)butanoylamino-2-phenyl]acetylamino-3,3-dimethyl}-
butanoylamino-2-hydroxy-4-phenyl}}butanoyl-5,5-dimethyl-1,
3-thiazolidine-4-carboxamide (tPeCO-Phg-Tle-Apns-
Dmt-NHiBu, 5i). Compound 5i was prepared from
tert-butylacetic acid and compound 9 by general method
A. Yield: 58%; 1H NMR (CDCl3)d = 0.70–0.93 (m,
5 · 3H), 1.01 (s, 3 · 3H), 1.47–1.58 (m, 2 · 3H), 1.67–
1.80 (m, 1H), 2.0–2.4 (bs, 1H), 2.13 (d, 2H), 2.60–2.82
(m, 2H), 2.85–3.24 (m, 2H), 4.00–5.00 (m, 4 · 1H +
2H), 5.51 (d, 1H), 6.13 (d, 1H), 6.16–6.25 (bs, 1H),
6.67 (t, 1H), 6.73 (d, 1H), 6.90–7.26 (m, 5H), 7.27–7.36
(m, 5H); MS (TOF) m/z = 738.909 [M+H]+, 760.904
[M+Na]+, 776.911 [M+K]+; Analytical HPLC:
Rt = 33.15 min (system A), 25.55 min (system B).
1
general method A. Yield: 57%; H NMR (CDCl3)d =
0.75–0.98 (m, 2H + 5 · 3H), 1.46–1.57 (m, 1H +
2 · 3H), 1.71–1.77 (m, 1H), 2.0–2.5 (bs, 1H), 2.68–
2.82 (m, 2H), 2.83–3.25 (m, 2H), 4.07–5.04 (m,
4 · 1H + 2H), 5.53–5.58 (m, 1H), 6.35 (d, 1H), 6.54–
6.68 (m, 2 · 1H), 6.72 (t, 1H), 6.92–7.25 (m, 5H),
7.26–7.35 (m, 5H); MS (TOF) m/z = 708.921 [M+H]+,
730.972 [M+Na]+, 746.905 [M+K]+; Analytical HPLC:
Rt = 30.46 min (system A), 21.40 min (system B).
5.1.13. (R)-N-Isobutyl-3-{{(2S,3S)-3-{(2S)-2-[(2S)-2-isobu-
tyrylamino-2-phenyl]acetylamino-3,3-dimethyl}butanoyl-
amino-2-hydroxy-4-phenyl}}butanoyl-5,5-dimethyl-1,3-thi-
azolidine-4-carboxamide (iPrCO-Phg-Tle-Apns-Dmt-NHiBu,
5f). Compound 5f was prepared from isobutyric acid
and compound 9 by general method A. Yield: 65%;
1H NMR (CDCl3)d = 0.74–0.95 (m, 5 · 3H), 1.10–1.18
(m, 2 · 3H), 1.42–1.58 (m, 2 · 3H), 1.71–1.81 (m, 1H),
2.42–2.50 (m, 1H), 2.2–2.5 (bs, 1H), 2.67–2.84 (m,
2H), 2.87–3.25 (m, 2H), 4.05–5.04 (m, 4 · 1H + 2H),
5.48–5.53 (m, 1H), 6.29 (d, 1H), 6.45 (d, 1H), 6.68 (t,
1H), 6.82 (d, 1H), 6.92–7.25 (m, 5H), 7.26–7.35 (m,
5H); MS (TOF) m/z = 710.990 [M+H]+, 732.939
[M+Na]+, 748.938 [M+K]+; Analytical HPLC: Rt =
30.99 min (system A), 22.34 min (system B).
5.1.17. (R)-N-Isobutyl-3-{{(2S,3S)-3-{(2S)-2-[(2S)-2-
(2,2-dimethyl)butanoylamino-2-phenyl]acetylamino-3,3-
dimethyl}butanoylamino-2-hydroxy-4-phenyl}}butanoyl-
5,5-dimethyl-1,3-thiazolidine-4-carboxamide [EtC(Me2)-
CO-Phg-Tle-Apns-Dmt-NHiBu, 5j]. Compound 5j was
prepared from 2,2-dimethylbutyric acid and compound
9 by general method A. Yield: 61%; 1H NMR
(CDCl3)d = 0.78–0.92 (m, 6 · 3H), 1.18 (s, 2 · 3H),
1.48–1.59 (m, 2H + 2 · 3H), 1.7–2.2 (bs, 1H), 1.72–1.78
(m, 1H), 2.61–2.81 (m, 2H), 2.84–3.27 (m, 2H), 3.98–
5.03 (m, 4 · 1H + 2H), 5.44 (d, 1H), 6.12 (d, 1H), 6.19
(d, 1H), 6.70 (t, 1H), 6.85 (d, 1H), 6.90–7.21 (m, 5H),
7.26–7.37 (m, 5H); MS (TOF) m/z = 739.093 [M+H]+,
761.035 [M+Na]+, 777.069 [M+K]+; Analytical HPLC:
Rt = 33.36 min (system A), 26.68 min (system B).
5.1.14. (R)-N-Isobutyl-3-{{(2S,3S)-3-{(2S)-2-[(2S)-2-piv-
aloylamino-2-phenyl]acetylamino-3,3-dimethyl}butanoyla-
mino-2-hydroxy-4-phenyl}}butanoyl-5,5-dimethyl-1,3-thi-
azolidine-4-carboxamide (tBuCO-Phg-Tle-Apns-Dmt-
5.1.18. (R)-N-Isobutyl-3-{{(2S,3S)-3-{(2S)-2-[(2S)-2-
(2-methyl)pentanoylamino-2-phenyl]acetylamino-3,3-di-
methyl}butanoylamino-2-hydroxy-4-phenyl}}butanoyl-5,5-

M. Zhang et al. / Bioorg. Med. Chem. 16 (2008) 5795–5802
5801
dimethyl-1,3-thiazolidine-4-carboxamide [nPrCH(Me)CO-
Phg-Tle-Apns-Dmt-NHiBu, racemate, 5k]. Compound 5k
was prepared from 2-methyl-n-valeric acid and com-
pound 9 by general method A. The compound was iso-
811.126 [M+K]+; Analytical HPLC: Rt = 32.64 min (sys-
tem A), 26.00 min (system B).
5.1.22. (R)-N-Isobutyl-3-{{(2S,3S)-3-{(2S)-2-[(2S)-2-cyclo-
hexanecarbonylamino-2-phenyl]acetylamino-3,3-dimethyl}-
butanoylamino-2-hydroxy-4-phenyl}}butanoyl-5,5-dimethyl-
1,3-thiazolidine-4-carboxamide (cHxCO-Phg-Tle-Apns-Dmt-
NHiBu, 5o). Compound 5o was prepared from cyclohex-
anecarboxylic acid and compound 9 by general method
A. Yield: 55%; 1H NMR (CDCl3)d = 0.78–0.91 (m,
5 · 3H), 1.22–1.58 (m, 3 · 2H + 2 · 3H), 1.67–1.90 (m,
1H + 2 · 2H), 2.15–2.24 (m, 1H), 2.6–2.9 (bs, 1H),
2.67–2.86 (m, 2H), 2.90–3.24 (m, 2H), 4.00–5.02 (m,
4 · 1H + 2H), 5.46–5.50 (m, 1H), 6.33 (d, 1H), 6.40
(d, 1H), 6.60–6.80 (m, 2 · 1H), 6.90–7.25 (m, 5H),
7.26–7.36 (m, 5H); MS (TOF) m/z = 750.277 [M+H]+,
772.803 [M+Na]+, 788.769 [M+K]+; Analytical HPLC:
Rt = 33.28 min (system A), 27.11 min (system B).
1
lated by HPLC as a racemate. Yield: 55%; H NMR
(CDCl3)d = 0.80–0.91 (m, 6 · 3H), 1.13 (d, 3H), 1.20–
1.78 (m, 1H + 2 · 2H + 2 · 3H), 2.30–2.35 (m, 1H),
2.5–2.9 (bs, 1H), 2.65–2.83 (m, 2H), 2.87–3.23 (m, 2H),
4.08–5.05 (m, 4 · 1H + 2H), 5.50–5.59 (m, 1H), 6.30–
6.40 (bs, 1H), 6.50–6.56 (bs, 1H), 6.68 (t, 1H), 6.85 (d,
1H), 6.91–7.25 (m, 5H), 7.26–7.34 (m, 5H); MS (TOF)
m/z = 760.894 [M+Na]+, 776.931 [M+K]+; Analytical
HPLC: Rt = 32.94 min (system A), 26.63 min
(system B).
5.1.19. (R)-N-Isobutyl-3-{{(2S,3S)-3-{(2S)-2-[(2S)-2-
(3-methyl)pentanoylamino-2-phenyl]acetylamino-3,3-
dimethyl}butanoylamino-2-hydroxy-4-phenyl}}butanoyl-
5,5-dimethyl-1,3-thiazolidine-4-carboxamide [EtCH(Me)
CH2CO-Phg-Tle-Apns-Dmt-NHiBu, racemate, 5l]. Com-
pound 5l was prepared from 3-methyl-n-valeric acid
and compound 9 by general method A. The compound
5.1.23. (R)-N-Isobutyl-3-{{(2S,3S)-3-{(2S)-2-[(2S)-2-ben-
zoylamino-2-phenyl]acetylamino-3,3-dimethyl}butanoyla-
mino-2-hydroxy-4-phenyl}}butanoyl-5,5-dimethyl-1,3-thi-
1
was isolated by HPLC as a racemate. Yield: 54%; H
azolidine-4-carboxamide
(PhCO-Phg-Tle-Apns-Dmt-
NMR (CDCl3)d = 0.78–0.92 (m, 7 · 3H), 1.11–1.29 (m,
1H), 1.30–1.43 (m, 1H), 1.45–1.58 (m, 2 · 3H), 1.60–
1.82 (m, 1H), 1.83–1.95 (m, 1H), 1.97–2.08 (m, 1H),
2.20–2.30 (m, 1H), 2.60–2.83 (m, 2H), 2.88–3.26 (m,
2H), 4.02–5.00 (m, 4 · 1H + 2H), 5.53 (d, 1H), 6.18 (d,
1H), 6.22–6.36 (m, 1H), 6.60–6.70 (m, 1H), 6.75–6.84
(m, 1H), 6.90–7.25 (m, 5H), 7.26–7.36 (m, 5H), OH not
observed; MS (TOF) m/z = 738.994 [M+H]+, 761.034
[M+Na]+, 777.036 [M+K]+; Analytical HPLC:
Rt = 33.31 min (system A), 25.71 min (system B).
NHiBu, 5p). Compound 5p was prepared from benzoic
acid and compound 9 by general method A. Yield:
61%; 1H NMR (CDCl3)d = 0.71–0.93 (m, 5 · 3H),
1.48–1.55 (m, 2 · 3H), 1.6–2.0 (bs, 1H), 1.65–1.80 (m,
1H), 2.62–2.85 (m, 2H), 2.88–3.20 (m, 2H), 4.03–4.91
(m, 4 · 1H + 2H), 5.68 (d, 1H), 6.19 (d, 1H), 6.22–
6.35 (m, 2 · 1H), 6.67 (t, 1H), 6.90–7.26 (m, 5H),
7.30–7.86 (m, 2 · 5H); MS (TOF) m/z = 744.873
[M+H]+, 766.838 [M+Na]+, 782.867 [M+K]+; Analyti-
cal HPLC: Rt = 32.19 min (system A), 25.16 min (system
B).
5.1.20. (R)-N-Isobutyl-3-{{(2S,3S)-3-{(2S)-2-[(2S)-2-(4-
methyl)pentanoylamino-2-phenyl]acetylamino-3,3-dimeth-
yl}butanoylamino-2-hydroxy-4-phenyl}}butanoyl-5,5-dimethyl-
1,3-thiazolidine-4-carboxamide (iPeCO-Phg-Tle-Apns-Dmt-
NHiBu, 5m). Compound 5m was prepared from 4-methyl-
n-valeric acid and compound 9 by general method A.
Yield: 59%; 1H NMR (CDCl3)d = 0.76–0.92 (m,
7 · 3H), 1.47–1.58 (m, 1H + 2H + 2 · 3H), 1.6–2.0
(bs, 1H), 1.74–1.80 (m, 1H), 2.26 (t, 2H), 2.64–2.85
(m, 2H), 2.89–3.24 (m, 2H), 3.99–4.95 (m, 4 · 1H +
2H), 5.47–5.51 (m, 1H), 6.15 (d, 1H), 6.24 (d, 1H),
6.68 (t, 1H), 6.75 (d, 1H), 6.90–7.19 (m, 5H), 7.25–
7.36 (m, 5H); MS (TOF) m/z = 738.962 [M+H]+,
760.862 [M+Na]+, 776.865 [M+K]+; Analytical HPLC:
Rt = 33.52 min (system A), 26.46 min (system B).
5.2. HIV-1 and HTLV-I protease inhibition assays
Compounds 9 and 5a-p were evaluated for HIV-1 prote-
ase inhibitory activity at 50 nM of the test compound,
and HTLV-I protease inhibitory activity as IC50 values,
using previously reported procedures.21 The HTLV-I
protease used in the assay was an L40I mutant of the
wild-type protease to prevent autolysis and improve
stability.22
5.3. Computer-assisted docking experiments
Attempts at ab initio computer-assisted docking of com-
pound 5p, which exhibited the more potent HTLV-I
protease inhibitory activity, in a truncated mutant
HTLV-I protease (AB dimer of PDB 2B7F; truncated
from 125 to 116 residues; L40I mutation)22 using all
the different methods provided by the 2006.0804 release
of Chemical Computing Group’s Molecular Operating
Environment software under MMFFX94 force-field,
were unsuccessful in that a rational pose, that depicted
the HMC transition-state mimic interactions usually ob-
served under X-ray crystallography studies, could not be
found. Consequently, we built a complex of compound
5p and HTLV-I protease, based on the X-ray crystallog-
raphy data of similar inhibitor-protease complexes,
namely PDB 1MRX and 2B7F in the 0absence of water.
The relative coordinates of the P1 ꢀ P1 (Apns-Dmt) res-
5.1.21. (R)-N-Isobutyl-3-{{(2S,3S)-3-{(2S)-2-[(2S)-2-(3-phenyl)-
propanoylamino-2-phenyl]acetylamino-3,3-dimethyl}but-
anoylamino-2-hydroxy-4-phenyl}}butanoyl-5,5-dimethyl-
1,3-thiazolidine-4-carboxamide
(PhCH2CH2CO-Phg-
Tle-Apns-Dmt-NHiBu, 5n). Compound 5n was prepared
from 3-phenylpropanoic acid and compound 9 by gen-
eral method A. Yield: 55%; 1H NMR (CDCl3)d =
0.70–0.95 (m, 5 · 3H), 1.44–1.82 (m, 1H + 2 · 3H,
partly covered by H2O), 2.52–2.84 (m, 2 · 2H), 2.85–
3.26 (m, 2 · 2H), 3.91–4.96 (m, 4 · 1H + 2H), 5.41 (t,
1H), 5.85–6.09 (m, 2 · 1H), 6.58–6.68 (m, 1H), 6.69–
6.79 (m, 1H), 6.90–7.39 (m, 3 · 5H), OH not observed;
MS (TOF) m/z = 772.855 [M+H]+, 795.059 [M+Na]+,

5802
M. Zhang et al. / Bioorg. Med. Chem. 16 (2008) 5795–5802
ki, K.; Kozuru, M.; Araki, K.; Kawano, F.; Niimi, M.;
Tobinai, K.; Hotta, T.; Shimoyama, M. Br. J. Haematol.
2001, 113, 375.
idues of potent HIV-1 protease inhibitor KNI-577 (PDB
1MRX)28 were merged into an HTLV-I protease inhibitor,
Ac-Ala-Pro-Gln-Val-Statine-Val-Met-His-Pro
(PDB
8. Ravandi, F.; Kantarjian, H.; Jones, D.; Dearden, C.;
Keating, M.; O’Brien, S. Cancer 2005, 104, 1808.
9. Yoshida, M.; Seiki, M.; Yamaguchi, K.; Takatsuki, K.
Proc. Natl. Acad. Sci. USA. 1984, 81, 2534.
10. Clark, S. C.; Arya, S. K.; Wong-Staal, F.; Matsumoto-
Kobayashi, M.; Kay, R. M.; Kaufman, R. J.; Brown, E.
L.; Shoemaker, C.; Copeland, T.; Oroszlan, S.; Smith, K.;
Sarngadharan, M. G.; Lindner, S. G.; Gallo, R. C. Proc.
Natl. Acad. Sci. USA. 1984, 81, 2543.
2B7F),22 by triangulation using the carboxylic acid’s
carbons of the two proteases’ two catalytic aspartic
acids (AB dimers), and the two inhibitors’ catalytic P1
hydroxy group as points of alignment. After the
HIV-1 protease was removed, all heavy atoms were fixed
except for the side-chains of HTLV-I protease’s Asp32A
and Asp32B, the polarization state of the Asps was
modified accordingly, and the model was energy mini-
mized under MMFFX94 force-field. The merged inhib-
itor’s positions of all heavy atoms that had hydrogen
bond interactions in the respective original model were
fixed and the inhibitor was energy minimized in the rigid
HTLV-I protease. The model was meticulously modified
accordingly to resemble compound 5p while being en-
ergy minimized after each major modification. The gen-
erated model was ‘‘water-soaked’’ to identify the water
molecule that mediated the interactions between the
flaps of HTLV-I protease and the inhibitor. In the final
step, the free inhibitor and water molecule were energy
minimized inside the rigid HTLV-I protease inhibitor
(Fig. 2).
11. Kobayashi, M.; Ohi, Y.; Asano, T.; Hayakawa, T.; Kato,
K.; Kakinuma, A.; Hatanaka, M. FEBS Lett. 1991, 293,
106.
´
12. To¨zser, J.; Zahuczky, G.; Bagossi, P.; Louis, J. M.;
Copeland, T. D.; Oroszlan, S.; Harrison, R. W.; Weber, I.
T. Eur. J. Biochem. 2000, 267, 6287.
13. Heidecker, G.; Hill, S.; Lloyd, P. A.; Derse, D. J. Virol.
2002, 76, 13101.
´
14. Louis, J. M.; Oroszlan, S.; To¨zser, J. J. Biol. Chem. 1999,
274, 6660.
´
15. Kadas, J.; Weber, I. T.; Bagossi, P.; Miklossy, G.; Boross,
P.; Oroszlan, S.; To¨zser, J. J. Biol. Chem. 2004, 279,
´
27148.
16. Ding, Y. S.; Rich, D. H.; Ikeda, R. A. Biochemistry 1998,
37, 17514.
17. Daenke, S.; Schramm, H. J.; Bangham, C. R. M. J. Gen.
Virol. 1994, 75, 2233.
Acknowledgements
18. Ha, J. J.; Gaul, D. A.; Mariani, V. L.; Ding, Y. S.; Ikeda,
R. A.; Shuker, S. B. Bioorganic Chem. 2002, 30, 138.
19. Maegawa, H.; Kimura, T.; Arii, Y.; Matsui, Y.; Kasai, S.;
Hayashi, Y.; Kiso, Y. Bioorg. Med. Chem. Lett. 2004, 14,
5925.
20. Kimura, T.; Nguyen, J.-T.; Maegawa, H.; Nishiyama, K.;
Arii, Y.; Matsui, Y.; Hayashi, Y.; Kiso, Y. Bioorg. Med.
Chem. Lett. 2007, 17, 3276.
21. Nguyen, J.-T.; Zhang, M.; Kumada, H.-O.; Itami,
A.; Nishiyama, K.; Kimura, T.; Cheng, M.; Hay-
ashi, Y.; Kiso, Y. Bioorg. Med. Chem. Lett. 2008,
18, 366.
This research was supported in parts by The Frontier
Research Program, The 21st Century COE Program
from The Ministry of Education, Culture, Sports, Sci-
ence and Technology, Japan; and Japan Society for
the Promotion of Science’s Post-Doctoral Fellowship
for Foreign Researchers. We are grateful to Mr. T.
Hamada for mass spectrometry and HIV-1 enzymatic
assay determinations.
22. Li, M.; Laco, G. S.; Jaskolski, M.; Rozycki, J.; Alexan-
dratos, J.; Wlodawer, A.; Gustchina, A. Proc. Natl. Acad.
Sci. USA. 2005, 102, 18332.
23. Nezami, A.; Luque, I.; Kimura, T.; Kiso, Y.; Freire, E.
Biochemistry 2002, 41, 2273.
24. Nezami, A.; Kimura, T.; Hidaka, K.; Kiso, A.; Liu, J.;
Kiso, Y.; Goldberg, D. E.; Freire, E. Biochemistry 2003,
42, 8459.
25. Kiso, A.; Hidaka, K.; Kimura, T.; Hayashi, Y.; Nezami,
A.; Freire, E.; Kiso, Y. J. Peptide Sci. 2004, 10, 641.
26. Abdel-Rahman, H. M.; Kimura, T.; Hidaka, K.; Kiso, A.;
Nezami, A.; Freire, E.; Hayashi, Y.; Kiso, Y. Biol. Chem.
2004, 385, 1035.
References and notes
1. Poiesz, B. J.; Ruscetti, F. W.; Gazdar, A. F.; Bunn, P. A.;
Minna, J. D.; Gallo, R. C. Proc. Natl. Acad. Sci. USA.
1980, 77, 7415.
2. Nitta, T.; Kanai, M.; Sugihara, E.; Tanaka, M.; Sun, B.;
Nagasawa, T.; Sonoda, S.; Saya, H.; Miwa, M. Cancer
Soc. 2006, 97, 836.
3. Proietti, F. A.; Carneiro-Proietti, A. B. F.; Catalan-
Soares, B. C.; Murphy, E. L. Oncogene 2005, 24, 6058.
4. Nicot, C. Am. J. Hematol. 2005, 78, 232.
5. Arisawa, K.; Soda, M.; Akahoshi, M.; Fujiwara, S.;
Uemura, H.; Hiyoshi, M.; Takeda, H.; Kashino, W.;
Suyama, A. Cancer Sci. 2006, 97, 535.
27. Hidaka, K.; Kimura, T.; Tsuchiya, Y.; Kamiya, M.;
Ruben, A. J.; Freire, E.; Hayashi, Y.; Kiso, Y. . Bioorg.
Med. Chem. Lett. 2007, 17, 3048.
28. Vega, S.; Kang, L.-W.; Velazquez-Campoy, A.; Kiso, Y.;
Amzel, L. M.; Freire, E. Proteins 2004, 55, 594.
6. Bittencourt, A. L. Rev. Inst. Med. Trop.. Sao Paulo 1998,
40, 245.
7. Yamada, Y.; Tomonaga, M.; Fukuda, H.; Hanada, S.;
Utsunomiya, A.; Tara, M.; Sano, M.; Ikeda, S.; Takatsu-