Thio- and oxoflavopiridols, cyclin-dependent kinase 1-selective inhibitors: Synthesis and biological effects

Article abstract of DOI:10.1021/jm000231g

Flavopiridol analogues, thio- and oxoflavopiridols which contain a sulfur (16) or oxygen (18) atom tinker between a chromone ring and the hydrophobic side chain, are selective cyclin-dependent kinase 1 (CDK1) inhibitors with an IC₅₀ of 110 and

Full text of DOI:10.1021/jm000231g

4126
J . Med. Chem. 2000, 43, 4126-4134
Th io- a n d Oxofla vop ir id ols, Cyclin -Dep en d en t Kin a se 1-Selective In h ibitor s:
Syn th esis a n d Biologica l Effects
Kyoung Soon Kim,*,§ J ohn S. Sack,† J ohn S. Tokarski,† Ligang Qian,§ Sam T. Chao,§ Leslie Leith,§
Yolanda F. Kelly,§ Raj N. Misra,§ J ohn T. Hunt,§ S. David Kimball,§ William G. Humphreys,# Barris S. Wautlet,‡
J anet G. Mulheron,‡ and Kevin R. Webster‡
Departments of Oncology Chemistry, Oncology Drug Discovery, Structural Biology and Modelling, and Metabolism and
Pharmacokinetics, Bristol-Myers Squibb Pharmaceutical Research Institute, Princeton, New J ersey 08543-4000
Received J une 5, 2000
Flavopiridol analogues, thio- and oxoflavopiridols which contain a sulfur (16) or oxygen (18)
atom linker between a chromone ring and the hydrophobic side chain, are selective cyclin-
dependent kinase 1 (CDK1) inhibitors with an IC50 of 110 and 130 nM. These analogues were
prepared from key intermediate 7 by substituting the ethyl sulfoxide. Enantio pure intermediate
piperidone 10 was obtained from the racemic piperidone 8 via a very efficient “dynamic kinetic
resolution” in 76% yield. Hydrophobic side chains such as chlorophenyl or tert-butyl produced
potent CDK1 inhibitory activity, while hydrophilic side chains such as pyrimidine or aniline
caused a severe reduction in CDK inhibitory activity. These analogues are competitive inhibitors
with respect to ATP, and therefore activity was dependent upon the CDK subunit without
being affected by the cyclin subunit or protein substrate. Thio- and oxoflavopiridols 16 and 18
are not only selective within the CDK family but also discriminated between unrelated serine/
threonine and tyrosine protein kinases. CDK1 selective thio- and oxoflavopiridol analogues
inhibit the colony-forming ability of multiple human tumor cell lines and possess a unique
antiproliferative profile in comparison to flavopiridol.
In tr od u ction
Toward this end, several ATP-competitive small
organic molecules have been reported in the literature
as CDK inhibitors. Flavopiridol (1),7 butyrolactone (2),8
olomoucine (3),9 and kenpaullone (4)10 and subsequently
several analogues11 of olomoucine and flavopiridol have
been reported. While butyrolactone (2) and olomoucine
(3) inhibit both CDK1 and CDK2 without having ap-
preciable CDK4 inhibitory activity, flavopiridol (1) is
known to have a broad spectrum of CDK inhibitory
activity against CDK1, -2, and -4 and is significantly
less active against unrelated kinases (Table 3).
The cyclin-dependent kinases (CDKs) are serine/
threonine protein kinases which are the driving force
behind the cell cycle and cell proliferation.1,2 CDKs are
multisubunit enzymes composed of at least a catalytic
subunit (CDK) and a regulatory subunit (cyclin). To date
at least 10 CDKs and 15 cyclins have been identified.
Once activated, CDKs phosphorylate and modulate the
activity of a variety of cellular proteins. These proteins
can be categorized as tumor suppressors (e.g., RB, p53),
transcription factors (e.g., E2F-DP, RNA pol II), replica-
tion factors (replication protein A), and organizational
factors that influence cellular and chromatin structures
(e.g., histone H1, lamin A, MAP4).3 Individual CDKs
perform distinct roles in cell cycle progression and can
be classified as either G1, S, or G2/M phase enzymes.1
The kinase activity of CDK complexes is tightly regu-
lated and cell cycle-dependent. This regulation occurs
at the level of cyclin gene transcription, cyclin protein
degradation, and posttranslational modification of the
CDK subunit. In addition, two families of CDK inhibi-
tory proteins have been identified: the KIP proteins
p21, p27, and p57 and the INK4 proteins p15, p16, p18,
and p19.4 Aberrations in each of these regulatory
pathways have been found in a large percentage of
human tumors including melanoma, lymphoma, and
carcinomas of the breast, prostate, colon, non-small-cell
lung, ovarian, pancreatic, and squamous cell.5 As a
result, there is considerable interest in restoring normal
cell cycle control in tumor cells by targeting CDKs or
CDK-associated molecules.6
* To whom correspondence should be addressed. Tel: (609) 252-
5181. Fax: (609) 252-6601. E-mail: kyoung.kim@bms.com.
§ Department of Oncology Chemistry.
‡ Department of Oncology Drug Discovery.
These attributes plus the fact that flavopiridol is a
highly potent inhibitor of tumor cell proliferation make
† Department of Structural Biology and Modelling.
#
Department of Metabolism and Pharmacokinetics.
10.1021/jm000231g CCC: $19.00 © 2000 American Chemical Society
Published on Web 10/14/2000

Thio- and Oxoflavopiridols
J ournal of Medicinal Chemistry, 2000, Vol. 43, No. 22 4127
Sch em e 1
it an interesting target for further analogue synthesis.
We set out to prepare analogues which would be more
selective within the CDK family and possess a unique
spectrum of antitumor activity. Herein, we report a
series of thio- and oxoflavopiridols that are CDK1-
selective inhibitors of the prototype 5 where the pip-
eridylchromone ring is coupled to the hydrophobic R
group via a sulfur or oxygen atom linker.
chiral piperidone 10 which was obtained from racemic
piperidone 813 via a very efficient “dynamic kinetic
resolution”14 (Scheme 1). Heating racemic piperidone 8
in the presence of dibenzoyl-D-tartaric acid in methanol
provides optically pure piperidone 10 in 76% yield after
removing dibenzoyl-D-tartaric acid. Desired enantio-
meric salt 9 was quite insoluble in methanol, while the
opposite enantiomeric salt was soluble. There is a
difference not only in solubility between these two
diastereomeric salts but also in thermodynamic stabil-
ity. The relative ratio of the two diastereomeric salts
existing in hot methanol solution (not in the precipi-
tates) was ca. 4:1 in favor of the desired enantiomeric
salt 9 as determined by chiral HPLC column. We believe
that a facile in situ epimerization of the C-4 chiral center
under the reaction conditions, the thermodynamic sta-
bility difference between the two diastereomeric salts
in favor of the desired 4R-enantiomer of piperidone 9,
and a large solubility difference between the two dia-
stereomeric salts in methanol enable this “dynamic
kinetic resolution” to be very efficient. Enantio pure
piperidone 10 was prepared on 100-g scale by this
process. Reduction of piperidone 10 using Dibal-H
provided the mixture of cis-alcohol 11 (56% yield) and
trans-alcohol 12 (13% yield) after purification.
Acylation, which produced concomitant demethyla-
tion, followed by treatment of 13 with carbon disulfide
produced thiol 14 (Scheme 2). Depending on the nu-
cleophiles, either thioether 6, obtained by the ethylation
of thiol 14, or its diastereomeric mixture of sulfoxide 7
was used. Sulfoxide 7 generally led to the higher yields
of the product under milder reaction conditions. Dem-
ethylation of thioether 15 produced the desired thiofla-
vopiridol 16. Analogues shown in Table 1 (22-26) were
prepared in a similar manner using racemic intermedi-
ate thioethyl 19 or sulfoxide 20 (Scheme 3).
We originally envisioned that thioether 6 or alkyl
sufoxide intermediate 7 would serve as a key intermedi-
ate. Replacement of the thioalkyl or alkyl sulfoxide
group with the appropriate nucleophiles would expedite
analogue synthesis.12 Interestingly, the thio and oxo
analogues (Nu ) SR or OR) demonstrated CDK1-
selective inhibitory activities and displayed unique
biological properties, presumably due to an altered
selectivity profile. The synthesis, structure-activity
relationships (SARs), and biological properties of these
compounds are described below.
Resu lts a n d Discu ssion
The biological activities of flavopiridol and thio-, oxo-,
and other flavopiridol analogues were compared in a
series of protein kinase and clonogenic (colony-forma-
tion) assays. As illustrated in Table 1, flavopiridol (1)
is a potent, broad-spectrum inhibitor of CDKs. Ana-
Ch em istr y
The key chiral intermediates for the analogue syn-
thesis, thioether 6 and sulfoxide 7, were prepared from

4128 J ournal of Medicinal Chemistry, 2000, Vol. 43, No. 22
Kim et al.
Sch em e 2
logues containing a sulfur (16) or oxygen (18) atom
linker retain significant potency against CDK1 and gain
selectivity within the family of CDKs (Table 1). Thio-
and oxoflavopiridols 16 and 18 exhibit selectivity over
CDK2 by ∼20-fold and 50-150-fold over CDK4, respec-
tively. These compounds showed competitive CDK1
inhibition with respect to ATP, and their selectivities
were dependent upon the CDK subunit without being
affected by the cyclin subunit (Table 2) or protein
substrate (either histone H1 or Rb, data not shown).
Thus far the tert-butyl analogue 23 is the most potent
Table 3, thio- and oxoflavopiridols 16 and 18 are
significantly more selective for CDK1 than other CDK
inhibitors described to date. In addition, each appears
to retain the excellent selectivity profile of flavopiridol
(1) when tested against unrelated serine/threonine and
tyrosine protein kinases. In fact, the oxygen-linked
analogue 18 is less active against PKC in comparison
to flavopiridol (1).
The structures of flavopiridol and thioflavopiridol 16
in complex with CDK2 were determined by X-ray cry-
stallography, revealing the molecular basis of inhibition
by these molecules.15 Figure 1a,b shows the crystal
structures of flavopiridol and thioflavopiridol 16, respec-
tively, in the active site of CDK2 with a number of re-
sidues comprising the site. In an effort to understand
the relative CDK1 selectivity of thioflavopiridol 16 over
flavopiridol, an analysis of the X-ray ligand-protein
contacts, a comparison of the sequences of CDK1, CDK2,
and CDK4, and molecular modeling were performed.
The contacts of the benzopyran ring and the piperidinyl
ring of both flavopiridol and thioflavopiridol 16 with
CDK2 in the X-ray structures are found to be similar
to those observed by Kim et al. for deschloroflavopiri-
CDK1 inhibitor of the sulfur-linked series with IC50
)
80 nM as a racemic mixture. This compound also retains
significant selectivity within the CDK family. In gen-
eral, hydrophobic substituents with the sulfur or oxygen
atom linker exhibit more potent CDK1/cyclin B inhibi-
tory activity compared to the less hydrophobic ana-
logues, and an aromatic ring is not required for good
potency as seen with the tert-butyl analogue 23. On the
other hand, more polar substituents, 24 and 25, signifi-
cantly reduce activity against CDK1. Insertion of a
nitrogen atom linker (25) results in a severe reduction
in potency against all CDKs tested. As illustrated in

Thio- and Oxoflavopiridols
J ournal of Medicinal Chemistry, 2000, Vol. 43, No. 22 4129
Sch em e 3
Ta ble 1. CDK Activity Data of the Analogue
Ta ble 3. Activity Data of CDK Inhibitors against Other
Different Kinases
IC50 (µM)
CDK1/ CDK2/ CDK4/
Cyc B1 Cyc E Cyc D1 MAP PKA PKC EGFR
compd
1, flavopiridol
0.03
0.11
0.13
7.0
0.17 0.10
2.10 16.2
2.11 6.15
19.0 >50
>25 >50
>25
30 2000 1000
94 260 160
14.0 22.0
16
16.1 >50
18
>50 >50
>50
440
590
olomoucine
7.0
1000
butyrolactone I 0.60
1.50 1000
IC50 (µM)
numbering from SWISS-PROT):17 CDK1(84-89), SM-
DLKK;CDK2(84-89),HQDLKK;CDK4(97-102),DQDL-
RT.
CDK1/ CDK2/ CDK4/
Cyc B1 Cyc E Cyc D1
compd
X
R
1, flavopiridol none 2-chlorophenyl
0.03
0.46
0.11
0.13
6.10
0.44
0.08
6.40
0.17
3.93
2.10
2.11
0.10
2.06
19
16
18
21
22
23
24
S
S
O
S
S
S
S
ethyl (()
The aforementioned differences occur in the region
where the chlorophenyl ring of thioflavopiridol 16
resides and may help explain the observed selectivity.
The CDK1 selectivity of thioflavopiridol 16 compared
to flavopiridol is due to a relative decrease in CDK2
binding. The selectivity may result in part from a less-
than-optimal interaction of thioflavopiridol 16 with
Ile10, a residue conserved in CDK1, CDK2, and CDK4,
coupled with a greater number of compensating interac-
tions with CDK1 than with CDK2 or CDK4. The X-ray
structure reveals that the chlorine atom of flavopiridol
favorably interacts with Ile10 as a van der Waals
distance of 2.99 Å to the Cγ of Ile10 would indicate
suggesting that this may be one of the reasons that
flavopiridol increases kinase inhibition by a factor of 6
over the deschloro analogue.18 In the CDK2 complex
with thioflavopiridol 16 the side chain of Ile10 is
observed to be shifted only slightly when the complex
is superposed by backbone atoms with that of the
flavopiridol-CDK2 complex. However, the sulfur linker
of thioflavopiridol 16 places the chlorophenyl ring
further away from the side chain of Ile10, relative to
flavopiridol, and closer to Lys89 which is observed to
be within contact distance of thioflavopiridol. The
hydrocarbon portion of the side chain of Lys89 is within
van der Waals contact of the chlorophenyl ring. In
addition, the basic nitrogen of Lys89 adopts a geometry
with respect to the chlorophenyl ring suitable for an
2-chlorophenyl (-)
2-chlorophenyl (-)
2-chlorophenyl (+)
phenyl (()
16.2
6.15
4.40 >25
6.59
1.07
4.10
2.07
82.5
tert-butyl (()
4,6-dimethyl-
pyrimidine (()
phenyl (()
40.4
25
26
NH
16.3
2.50
>25
9.69
>25
3.70
none N-piperidyl(()
Ta ble 2. Activity Data of CDK Inhibitors Depending on Cyclin
Subunit
IC50 (µM)
CDK1/
Cyc B1
CDK1/
Cyc A
CDK2/
Cyc E
CDK2/
Cyc A
compd
1, flavopiridol
16
0.03
0.11
0.05
0.16
0.17
5.9
0.07
1.5
dol.16 An analysis of the sequence alignment of CDK1,
CDK2, and CDK4 shows only subtle differences in the
residues that comprise the buried ATP-binding site;
thus selectivity is most likely due to interactions other
than those deep in the ATP-binding site. An amino acid
comparison of the surface-exposed hinge region and an
analysis of flavopiridol and thioflavopiridol 16 docked
in homology models of CDK1 and CDK4 (built from the
X-ray structure of CDK2, results not published) reveal
that several differences in this particular area may be
playing a role in specificity (aligned sequences with

4130 J ournal of Medicinal Chemistry, 2000, Vol. 43, No. 22
Kim et al.
F igu r e 1. (a) Binding site of the flavopiridol-CDK2 X-ray crystal structure. Flavopiridol is shown in ball-and-stick with a number
of CDK2 residues comprising the site. Nitrogen atoms are colored blue, oxygen atoms are colored red, chlorine is colored orange,
sulfur is colored yellow, carbon atoms in thioflavopiridol are colored dark green, and carbon atoms in CDK2 are colored light
green. (b) Binding site of thioflavopiridol 16-CDK2 X-ray crystal structure. Thioflavopiridol is shown in ball-and-stick with a
number of CDK2 residues comprising the site. Nitrogen atoms are colored blue, oxygen atoms are colored red, chlorine is colored
orange, sulfur is colored yellow, carbon atoms in thioflavopiridol are colored dark green, and carbon atoms in CDK2 are colored
light green.
Ta ble 4. Clonogenic Assay Data
Con clu sion s
IC50 (nM)
Enantio pure thio- and oxoflavopiridols were prepared
compd
HCT116
A2780
PC3
Mia PaCa-2
via efficient “dynamic kinetic resolution” of intermediate
piperidone 8, and representative thio- and oxoflavopiri-
dols 16 and 18 display CDK1 selective inhibitory
activity over CDK2 by 20-fold and 50-150-fold over
CDK4, respectively. These compounds do not have any
appreciable inhibitory activity against other kinds of
kinases such as MAP, PKA, PKC, and EGFR. It is
postulated that the CDK1 selectivity of the hydrophobic
group-substituted analogues containing a sulfur and
oxygen linker presented in this paper is due to sequence
differences in the solvent-exposed binding pocket. In a
clonogenic assay using human tumor cell lines, nonse-
lective CDK inhibitor flavopiridol and CDK1-selective
thioflavopiridol 16 displayed a different spectrum of
activity depending on the cell lines. We believe that
these small CDK1-selective inhibitor molecules provide
us with a unique opportunity to further study the
biological role of CDK1/cyclin B.
1, flavopiridol
13
210
15
870
10
20
36
30
16
amino-aromatic interaction.19 This type of interaction
involving Lys89 was also observed in the crystal struc-
ture of a CDK2-olomoucine complex.20 In the flavopiri-
dol-CDK2 complex it is observed that Lys89 is in a
different position and closer to Gln85 with a distance
from the Lys side chain nitrogen to the Gln side chain
oxygen of 3.59 Å. Although Lys89 is common to CDK1
and CDK2, sequence differences near this residue, e.g.
Met85 (CDK1) vs Gln85 (CDK2), would not only in-
crease the hydrophobic nature of this area but also affect
packing preferences, including the conformational ori-
entation of Lys89, such that it is steered toward the
chlorophenyl ring of thioflavopiridol. CDK2/CDK4 se-
lectivity of thioflavopiridol may be rationalized in part
by the fact that the corresponding residue of Lys89
(CDK2) in CDK4 is Thr, and thus a potential favorable
interaction is lost in binding to CDK4.
Exp er im en ta l Section
All new compounds were homogeneous by thin-layer chro-
matography and reversed-phase HPLC (>99%). Flash chro-
matography was carried out on E. Merck Kieselgel 60 silica
gel (230-400 mesh). Preparative HPLC was run on YMC OD
S-10 50- × 500-mm column eluting with a mixture of solvents
A and B (starting from 10% solvent B to 100% solvent B over
30 min gradient time; solvent A: 10% MeOH-90% H2O-0.1%
TFA; solvent B: 90% MeOH-10% H2O-0.1% TFA; flow rate:
Thioflavopiridol 16, the most CDK1-selective com-
pound, was further compared to flavopiridol (1) in a
clonogenic assay using four human tumor cell lines
(Table 4). This assay demonstrated that changing the
CDK selectivity profile impacted the antiproliferative
activity of these small molecules. Thioflavopiridol 16
was found to be 16- and 58-fold less active than
flavopiridol (1) when tested against HCT116 colorectal
carcinoma or A2780 ovarian carcinoma cells. However,
thioflavopiridol 16 is equally effective at inhibiting
colony formation of PC-3 prostate carcinoma or Mia
PaCa-2 pancreatic carcinoma cells. These data suggest
that a CDK1-selective inhibitor may have a unique
spectrum of antitumor activity and perhaps with altered
or lesser side effects. Thus, thioflavopiridol 16, a potent
and more selective analogue of flavopiridol, may war-
rant further preclinical evaluation.
1
84 mL/min; UV: 254 nm). H and 13C NMR spectra were ob-
tained on a J EOL CPF-270 spectrometer operating at 270 or
67.5 MHz, respectively, and are reported as ppm downfield
from an internal tetramethylsilane standard. The abbrevia-
tions of qn and sx in 1H NMR refer to quintet and sextet, res-
pectively. Melting points are uncorrected. Tetrahydrofuran
(THF) and xylenes were dried by distillation from sodium. N,N-
Dimethylformamide (DMF) was dried over 4A molecular
sieves.
(R)-1-Meth yl-4-(2,4,6-tr im eth oxyp h en yl)-3-p ip er id in o-
n e (10). A mixture of (()-1-methyl-4-(2,4,6-trimethoxyphenyl)-
3-piperidinone13 (1.60 g, 5.73 mmol) and dibenzoyl-D-tartaric
acid (2.28 g, 315 mmol) in 10 mL of methanol was heated at
reflux temperature until it became a homogeneous solution
under argon atmosphere and it was cooled to room tempera-

Thio- and Oxoflavopiridols
J ournal of Medicinal Chemistry, 2000, Vol. 43, No. 22 4131
ture. After stirring overnight at room temperature the pre-
cipitated solid was filtered, washed with a small amount of
methanol to obtain the first crop of (R)-1-methyl-4-(2,4,6-tri-
methoxyphenyl)-3-piperidinone dibenzoyl-D-tartaric acid salt,
7 (2.65 g). The filtrate solution was concentrated to a volume
of ca. 6 mL and was stirred at ambient temperature overnight.
The second crop of the chiral salt (0.24 g) was obtained by
filtering the solid and washing with a small amount of
methanol. The combined solid salt (2.89 g) was dissolved in a
mixture of CH2Cl2 (40 mL) and aqueous NaOH solution (12
mL of 0.5 NaOH), the organic solution was separated, washed
with brine, dried over MgSO4 and concentrated under reduced
pressure to obtain the title compound (1.60 g, 76% yield) as a
white solid: mp 131-133 °C; [R]25D +31° (MeOH, c 1.0). Optical
purity of 99.9% of this enantiomer was checked by HPLC on
OD Chiracel column (250 × 4.6 mm) eluting with 30% 2-pro-
panol in hexane containing 0.1% triethylamine (flow rate: 1.0
mL/min) at 254 nm: 1H NMR (CDCl3) δ 6.28 (s, 2H), 4.01 (m,
1H), 3.92 (s, 3H), 3.88 (s, 6H), 3.58 (m, 1H), 3.13 (m, 1H), 2.97
(m, 1H), 2.61 (m, 1H), 2.51 (s, 3H), 2,48 (m, 1H), 2.12 (m, 1H);
13C NMR (CDCl3) δ 207.0, 151.5, 159.7, 109.7, 92.3, 67.7, 56.9,
56.4, 55.7, 47.2, 44.6, 30.5; MS (ESI) 446 (M + H)+.
(3S ,4R )-2-(E t h ylt h io)-8-(3-h yd r oxy-1-m e t h yl-4-p ip -
er id yl)-5,7-d im eth oxy-4H-1-ben zop yr a n -4-on e (6). To a
suspension of 13 (6.5 g, 21.01 mmol) in anhydrous THF (85
mL) under argon was added carbon disulfide (7 mL). The
slurry was cooled to 0 °C (ice bath) and lithium bis(trimeth-
ylsilyl)amide (1.0 M solution in THF, 126 mL) was slowly
added at such a rate that the temperature of the slurry was
maintained below 18 °C. After the addition was complete, the
solution was gradually warmed to room temperature and
stirred for 2 h. Water (50 mL) was added to the mixture,
stirred and concentrated in vacuo. The residue was cooled to
0 °C, methanol (80 mL) was added followed by 50% aqueous
trifluoroacetic acid (30 mL). After stirring at room temperature
for 1.0 h. the solution was concentrated in vacuo. The residue
was triturated with a small amount of ethyl acetate and the
solid was filtered to obtain crude thiol 14 (6.5 g). This crude
thiol compound was used directly for the next step without
further purification.
To a stirred solution of crude thiol 14 (6.5 g, 21.01 mmol)
at 0 °C in anhydrous DMF (75 mL) under argon was added
cesium carbonate (35 g). The resulting suspension was stirred
at 0 °C for 30 min, then iodoethane (1.75 mL) was added. The
mixture was stirred at 0 °C for 1 h and then at room
temperature for 18 h. The reaction mixture was diluted with
water (300 mL) and the product was extracted with dichlo-
romethane (3 × 300 mL). The combined dichloromethane
extracts was washed with 10% aqueous lithium chloride, water
and brine, dried over anhydrous Na2SO4 and concentrated in
vacuo to give crude 6 as a gum (5.0 g). The crude material
was purified by flash chromatography (ethyl acetate-metha-
nol-triethylamine/10:0.5:0.2 and 10:1:0.2) to give 3.81 g of
pure 6 (50%) as a solid: [R]25D -74.0° (MeOH, c 0.42); 1H NMR
(CDCl3) δ 1.43 (t, J ) 7.6 Hz, 3H), 1.58 (m, 1H), 2.05 (m, 1H),
2.23 (m, 1H), 2.35 (s, 3H), 3.02 (m, 5H), 3.26 (s, broad, 1H),
3.35 (m, 1H), 3.86 (m, 1H), 3.94 (s, 3H), 3.97 (s, 3H), 6.43 (s,
1H), 6.46 (s, 1H); 13C NMR (CDCl3) δ 157.8, 165.1, 162.1, 159.7,
157.6, 110.3, 109.9, 108.8, 92.5, 69.9, 62.8, 57.1, 56.4, 56.0, 46.4,
(3S,4R)-1-Meth yl-4-(2,4,6-tr im eth oxyp h en yl)-3-p ip er i-
d in ol (11). To a stirred solution of 10 (25.0 g, 89.5 mmol) in
CH2Cl2 (225 mL) at -78 °C was added dropwise a solution of
diisobutylaluminum hydride (180 mmol, 180 mL of 1 M
solution in CH2Cl2) under argon atmosphere with maintaining
the reaction temperature below -65 °C. The reaction mixture
was stirred for 3.5 h at -78 °C after the completion of addition.
Trifluoroacetic acid (50 mL) was added dropwise to the reaction
mixture at -78 °C. After stirring the mixture for 15 min,
MeOH (250 mL) was added. The mixture was warmed to room
temperature and concentrated to obtain a solid residue, which
was stirred with aqueous NaOH solution (2 N, 750 mL) for 15
min. The product was extracted with ethyl acetate (3 × 750
mL). The combined ethyl acetate solution was washed with
brine, dried over Na2SO4 and concentrated to give a gum (25.0
g, 99%). This material was dissolved in CH2Cl2 and purified
by flash chromatography (EtOAc:MeOH:Et3N/100:20:0.2) to
afford a mixture of cis- and trans-alcohols as a foam (22.0 g).
This isomeric mixture was dissolved in 30% 2-propanol in
hexanes and passed through a Chiracel AD column (50 × 50
mm; Daicel Chem. Ind. Ltd.) eluting with 30% 2-propanol in
hexanes containing 0.2% Et3N to obtain the title compound
38.4. 25.5, 24.5, 14.2; MS (ESI) 380 (M + H)+. Anal. (C19H25
NO5S‚0.25H2O) C, H, N, S.
-
(3S,4R)-2-(Eth ylsu lfin yl)-8-(3-h yd r oxy-1-m eth yl-4-p ip -
er id in yl)-5,7-d im eth oxy-4H-1-ben zop yr a n -4-on e, Tr iflu o-
r oa cetic Acid Sa lt (7). To a solution of 6 (1.67 g, 4.41 mmol)
in CH2Cl2 (18 mL) containing TFA (1.7 mL) at 0 °C was added
m-chloroperbenzoic acid (1.42 g, 50-60% contents) as a solid
with stirring. The mixture was stirred at 0 °C for 1 h and then
quenched by adding dimethyl sulfide (1 mL). The mixture was
stirred for 5 min and concentrated in vacuo. To the residue
was added ethyl ether (150 mL) and the mixture was stirred
at room temperature for 30 min. The precipitated solid was
filtered and dried to obtain the sulfoxide 7 (2.15 g) as a
diastereomeric mixture, which was contaminated by a very
small amount of overoxidized sulfone product. This crude
diastereomeric mixture was used for the next reaction without
any further purification: 1H NMR (CD3OD) δ 6.66 (s, 1H),
6.62(s, 1H), 6.34 (s, 1H), 6.32 (s, 1H), 4.21 (s, 1H), 4.15 (s, 1H),
3.66-3.04 (m, 8H), 2.91 (s, 3H), 1.88 (m, 1H), 1.32 (t, J ) 7.3
Hz, 3H), 1.26 (t, J ) 7.3 Hz, 3H); 13C NMR (CD3OD) δ 182.3,
182.2, 171.0, 169.8, 165.0, 164.7, 162.8, 158.4 158.1, 110.4,
109.6, 107.6, 107.5, 106.4, 106.2, 101.4, 100.9, 67.8, 67.7, 61.9,
61.7, 56.8, 56.7, 47.4, 46.6, 44.4, 44.3, 37.5, 37.3, 23.5, 6.1, 5.2;
MS (ESI) 396 (M + H)+.
(3S,4R)-2-[(2-Ch lor op h en yl)th io]-8-(3-h yd r oxy-1-m eth -
yl-4-p ip e r id in yl)-5,7-d im e t h oxy-4H -1-b e n zop yr a n -4-
on e (15). A mixture of potassium tert-butoxide (1.12 g, 10
mmol) and 2-chlorothiophenol (1.16 g, 8 mmol) in THF (10 mL)
was stirred at room temperature under argon atmosphere for
10 min. To this mixture at 0 °C was added a solution of 7 (1.0
g, 1.96 mmol) in THF (10 mL). The resulting mixture was
stirred at 0 °C for 1 h and it was directly loaded onto a
chromatography column (SiO2) and eluted with ethyl acetate
(ca. 1 L) followed by EtOAc:MeOH:Et3N (100:15:2) to afford
the desired product. This material was dissolved in CH2Cl2
(100 mL), washed with aqueous NaHCO3 solution, dried over
Na2SO4 and concentrated in vacuo to afford 15 (830 mg, 92%)
as a pure cis-alcohol 11 (14.1 g, 56%): mp 111-112 °C (lit.13
1
109-111 °C); [R]25 -53.8° (MeOH, c 1.0) (lit.13 -54.13°); H
D
NMR (CDCl3) δ 6.13 (s, 2H), 3.82 (m, 1H), 3.82 (s, 9H), 3.32
(m, 1H), 3.00 (m, 2H), 2.87 (m, 1H), 2.30 (s, 3 H), 2.10 (m,
3H), 1.40 (m, 1H); 13C NMR (CDCl3) δ159.6, 159.2, 111.5, 91.5,
70.4, 62.6, 57.1, 55.8, 55.3, 46.4, 36.8, 24.3.
(3S,4R)-4-(3-Acetyl-2-h yd r oxy-4,6-d im eth oxyp h en yl)-1-
m eth yl-3-p ip er id in ol (13). A solution of 11 (13.8 g, 49 mmol)
in CH2Cl2 (250 mL) at 0 °C was added BF3 etherate (50 mL)
followed by acetic anhydride (40 mL). The resulting mixture
was stirred at room temperature overnight and concentrated.
The residue was cooled in an ice bath and quenched with
MeOH (300 mL) and the mixture was stirred at room tem-
perature for 15 min and concentrated in vacuo. The residue
was dissolved in MeOH (300 mL) and stirred with 20%
aqueous KOH solution (200 mL) for 48 h. It was mostly
concentrated and pH of the residue was adjusted to 9.5 with
4 N hydrochloric acid. The product was extracted with CH2-
Cl2 (5 × 100 mL) and the combined CH2Cl2 extracts was dried
over Na2SO4 and concentrated to afford a solid. This solid was
purified by flash column chromatography (EtOAc:MeOH:Et3N/
100:15:0.2) to obtain a pure product which was triturated with
diisopropyl ether (150 mL) to give 13 (12.5 g, 82.6%) as a light
yellow solid: mp 182 °C (lit.13 184-186 °C); [R]25 -34.5°
D
(MeOH, c 1.0) (lit.13 -32.65°); 1H NMR (CDCl3) δ 14.40 (s, 1H),
6.11 (s, 1H), 4.51 (s, broad, 1H), 4.03 (s, 3H), 4.01 (s, 3H), 4.0
(s, 1H), 3.50 (m, 1H), 3.16 (m, 3H), 2.73 (s, 3H), 2.46 (s, 3H).
2.36 (m, 1H), 2.22 (m, 1H), 1.56 (m, 1H); 13C NMR (CDCl3) δ
203.3, 264.0, 163.7, 161.7, 109.9, 105.6, 86.1, 69.8, 62.2, 56.6,
55.2, 55.0, 46.0, 36.4, 32.8, 23.7; MS (ESI) 310 (M + H)+.

4132 J ournal of Medicinal Chemistry, 2000, Vol. 43, No. 22
Kim et al.
1
as a solid: mp 185-187 °C; [R]25 -59.8° (MeOH, c 0.41); H
MS (ESI) 418 (M + H)+. Anal. (C21H20ClNO6‚0.5H2O‚1.15CF3-
D
NMR (CD3OD) δ 1.54 (m, 1H), 2.33 (m, 2H), 2.48 (s, 3H), 3.08
(m, 4H), 3.70 (m, 1H), 3.96 (s, 3H), 3.98 (s, 3H), 4.60 (s, 1H),
6.00 (s, 1H), 6.63 (s, 1H),7.47 (m, 1H), 7.58 (m, 1H), 7.66 (d, J
) 8.2 Hz, 1H), 7.80 (d, J ) 6.6 Hz, 1H); 13C NMR (CD3OD) δ
179.3, 166.2, 165.9, 162.4, 160.3, 140.8, 139.1, 134.6, 133.3,
130.9, 129.1, 113.0, 111.6, 109.8, 95.5, 70.5, 63.8, 58.6, 57.9,
57.8, 46.8, 40.3, 25.7; MS (ESI) 462 (M + H)+.
COOH) C, H, N, F, Cl.
(()-(3SR,4RS)-2-(Eth ylth io)-5,7-dih ydr oxy-8-(3-h ydr oxy-
1-m eth yl-4-p ip er id in yl)-4H-1-ben zop yr a n -4-on e (19). To
a solution of racemic 6 (144 mg, 0.4 mmol) in 1,2-dichloroet-
hane (2 mL) at room temperature was added a solution of
boron tribromide (800 mg, 3.2 mmol). The mixture was stirred
at 80 °C for 5 h, cooled to 0 °C, quenched by methanol (5 mL),
and it was neutralized by aqueous NaHCO3 solution. After
concentrating in vacuo the residue was purified by preparative
HPLC to obtain the trifluoroacetic acid salt of 19 (121 mg, 67%)
(3S,4R)-2-[(2-Ch lor op h en yl)t h io]-5,7-d ih yd r oxy-8-(3-
h yd r oxy-1-m e t h yl-4-p ip e r id in yl)-4H -1-b e n zop yr a n -4-
on e (16). To a stirred solution of 15 (2.6 g, 5.63 mmol) in 1,2-
dichloroethane (50 mL) under argon was slowly added boron
tribromide (2.0 M solution in 1,2-dichloroethane, 28 mL). After
the addition was complete, the suspension was heated at 85-
90 °C for 4 h. The mixture was cooled and then concentrated
in vacuo. To the residue at -30 °C (dry ice-acetonitrile bath)
methanol (50 mL) was slowly added and the solution was
gradually warmed to room temperature. Sodium bicarbonate
(8 g) was added as a solid, stirred for 30 min and the mixture
was acidified with trifluoroacetic acid to pH 2.0. The mixture
was concentrated in vacuo to give a semisolid product (1.69
g). This material was purified by preparative HPLC to obtain
the product as a trifluoroacetic acid salt. The trifluoroacetate
salt was dissolved in methanol (70 mL) containing aqueous 1
N hydrochloric acid (15 mL), concentrated in vacuo to a small
volume and then lyophilized to give 1.94 g (69%) of the HCl
1
as a solid: mp 105-107 °C; H NMR (CD3OD) δ 6.27 (s, 1H),
6.15 (s, 1H), 4.24 (s, 1H), 3.63-3.12 (m, 8H), 2.90 (s, 3H), 1.85
(m, 1H), 1.42 (t, J ) 7.6 Hz, 3H); 13C NMR (CD3OD) δ 182.0,
171.7, 163.9, 162.2, 158.1, 106.9, 106.7, 105.2, 101.0, 68.0, 61.8,
56.8, 44.2, 37.4, 26.6, 23.3, 14.3; MS (ESI) 352 (M + H)+. Anal.
(C17H21NO5S‚1.0H2O‚1.1CF3COOH) C, H, N.
(()-(3SR,4RS)-2-(Eth ylsu lfin yl)-8-(3-h yd r oxy-1-m eth yl-
4-piper idin yl)-5,7-dih ydr oxy-4H-1-ben zopyr an -4-on e, Tr i-
flu or oa cetic Acid Sa lt (20). To a solution of 19 (0.99 g, 2.0
mmol) in CH2Cl2 (20 mL) containing TFA (0.8 mL) at 0 °C
was added m-chloroperbenzoic acid (0.576 g, 50-60% contents)
as a solid with stirring. The mixture was stirred at 0 °C for 2
h and then quenched by adding dimethyl sulfide (0.15 mL).
The mixture was stirred for 5 min and concentrated in vacuo.
To the residue was added ethyl ether (150 mL) and the mixture
was stirred at room temperature for 30 min. The precipitated
solid was filtered and dried to obtain the trifluoroacetic acid
salt of sulfoxide 20 (0.955 g, 96%) as a diastereomeric mixture,
which was contaminated by a very small amount of overoxi-
dized sulfone product. This crude diastereomeric mixture was
used for the next reaction without any further purification:
1H NMR (CD3OD) δ 6.66 and 6.62 (s, ea, 1H), 6.34 and 6.32
(s, ea, 1H), 4.21 and 4.15 (m, 1H), 3.66-3.04 (m, 8H), 2.91 (s,
3H), 1.88 (m, 1H), 1.32 and 1.26 (t, ea, J ) 7.3 Hz, 3H); 13C
NMR (CD3OD) δ 182.3 and 182.2, 171.0, 169.8, 165.0, 164.7,
162.8, 158.4 and 158.1, 110.4, 109.6, 107.6 and 107.5, 106.4
and 106.2, 101.4, 100.9, 67.8 and 67.7, 61.9 and 61.7, 56.8 and
56.7, 47.4, 46.6, 44.4 and 44.3, 37.5 and 37.3, 23.5, 6.1, 5.2.
salt of 16 as a light yellow powder: mp 110 °C; [R]25 -7.1°
D
(MeOH, c 0.32); 1H NMR (CD3OD) δ 1.66 (m, 1H), 2.90 (s, 3H),
3.50-2.91 (m, 7H), 4.05 (m, 1H), 5.93 (s, 1H), 6.29 (s, 1H),
7.87-7.50 (m, 4H); 13C NMR (CD3OD) δ 182.2, 168.7, 164.3,
162.2, 157.9, 140.3, 138.8, 133.9, 132.3, 129.9, 127.4, 108.3,
106.8, 105.2, 101.4, 67.8, 61.7, 56.8, 44.4, 37.1, 23.1; MS (ESI)
434 (M + H)+. Anal. (C21H20NO5SCl‚HCl‚1.1H2O) C, H, N, S,
Cl.
(3S,4R)-2-(2-Ch lor op h en oxy)-8-(3-h yd r oxy-1-m eth yl-4-
p ip er id in yl)-5,7-d im eth oxy-4H-1-ben zop yr a n -4-on e (17).
A mixture of 2-chlorophenol (1.0 mL, 9.65 mmol) and potas-
sium tert-butoxide (1.20 g, 10.69 mmol) in anhydrous THF (12
mL) was stirred for 10 min at room temperature, then cooled
in an ice bath and sulfoxide 7 (1.15 g, 2.26 mmol) was added.
The reaction mixture was stirred at 0-5 °C for 1.5 h. After
adding acetic acid (0.25 mL) to the mixture it was directly
loaded onto a silica gel column and eluted with ethyl acetate
followed by EtOAc:MeOH:Et3N (100:20:2) to obtain after
concentration a product containing trifluoroacetic acid and
triethylamine. This material was mixed with chloroform (60
mL) and aqueous NaHCO3 (10 mL) solution, the organic layer
was separated, dried over magnesium sulfate and concentrated
(()-(3SR,4RS)-2-(P h en ylt h io)-5,7-d ih yd r oxy-8-(3-h y-
d r oxy-1-m et h yl-4-p ip er id in yl)-4H -1-b en zop yr a n -4-on e
(22). A mixture of 19 (50 mg, 0.1 mmol), thiophenol and
NaHCO3 (100 mg) in DMF (0.5 mL) was stirred at 100 °C for
2 h, cooled to room temperature and the solid was filtered off.
The filtrate solution was purified by preparative HPLC to
obtain the trifluoroacetic acid salt of 22 (21 mg, 35%) as a
1
solid: mp 82-84 °C; H NMR (CD3OD) δ 7.73-7.59 (m, 5H),
6.25 (s, 1H), 5.86 (s, 1H), 4.08 (s, 1H), 3.52-2.93 (m, 6H), 2.88
(s, 3H), 1.68 (m, 1H); 13C NMR (CD3OD) δ 182.1, 171.0, 164.1,
162.1, 157.9, 136.6, 132.0, 131.5, 128.0, 107.7, 106.7, 105.1,
101.3, 67.8, 61.7, 56.7, 44.3, 37.1, 23.1; MS (ESI) 400 (M +
H)+. Anal. (C21H21NO5S‚0.5H2O‚1.6CF3COOH) C, H, N.
in vacuo to obtain 17 (760 mg, 76%) as a floppy glassy material:
[R]25 -55.6° (MeOH, c 0.7); H NMR (CDCl3) δ 7.54 (dd, J )
1
D
7.0, 2.3 Hz, 1H), 7.40-7.34 (m, 3H), 6.44 (s, 1H), 5.39 (s, 1H),
3.96 (s, 3H), 3.93 (s, 3H), 3.74 (s, broad, 1H), 3.19-2.85 (m,
6H), 2.28 (s, 3H), 1.85-2.05 (m, 1H), 1.48-1.60 (m, 1H); 13C
NMR (CDCl3) δ 179.4, 164.1, 162.2, 159.9, 155.1, 147.8, 131.8,
128.7, 128.1, 127.3, 123.4, 111.0, 108.2, 93.2, 91.1, 70.1, 62.8,
57.1, 56.7, 56.3, 46.7, 46.4, 38.4, 24.8; MS (ESI) 446 (M + H)+.
(()-(3SR,4RS)-2-(ter t-Bu tylth io)-5,7-d ih yd r oxy-8-(3-h y-
d r oxy-1-m et h yl-4-p ip er id in yl)-4H -1-b en zop yr a n -4-on e
(23). A mixture of sulfoxide 20 (50 mg, 0.1 mmol), tert-
butylthiol (45 mg, 0.5 mmol) and KOtBu (56 mg, 0.5 mmol) in
THF (1.5 mL) was stirred at room temperature for 0.5 h.
Trifluoroacetic acid (20 µL) was added to the reaction mixture,
and it was purified by preparative HPLC to obtain the
trifluoroacetic acid salt of 23 (32 mg, 84%) as a solid: mp 70-
(3S,4R)-2-(2-Ch lor op h en oxy)-5,7-d ih yd r oxy-8-(3-h y-
d r oxy-1-m et h yl-4-p ip er id in yl)-4H -1-b en zop yr a n -4-on e
(18). To a solution of 17 (760 mg, 1.70 mmol) in 1,2-
dichloroethane (20 mL) at room temperature was added a
solution of boron tribromide (8.5 mL of 2 M solution in 1,2-
dichloroethane, 17.0 mmol). The mixture was stirred at 80 °C
for 5 h, cooled and concentrated in vacuo. To the residue at
-30 °C methanol (10 mL) was added followed by solid NaHCO3
(1.5 g) and it was stirred at room temperature for 10 min. After
acidifying the mixture with TFA (0.5 mL) it was concentrated
to a small volume and the residue was purified by preparative
HPLC to obtain the trifluoroacetic acid salt of 18 (505 mg, 53%)
1
72 °C; H NMR (CD3OD) δ 6.48 (s, 1H), 6.30 (s, 1H), 4.25 (s,
1H), 3.70-3.47 (m, 3H), 3.34-3.15 (m, 3H), 2.91 (s, 3H), 1.82
(m, 1H), 1.53 (s, 9H); MS (ESI) 380 (M + H)+; HRMS FAB (M
+ H)+ calcd for C19H25NO5S 380.1532, found 380.1541.
(()-(3SR,4RS)-2-[(4,6-Dim eth ylp yr im idin -2-yl)th io]-5,7-
d ih yd r oxy-8-(3-h yd r oxy-1-m et h yl-4-p ip er id in yl)-4H -1-
ben zop yr a n -4-on e (24). A mixture of sulfoxide 20 (50 mg,
0.1 mmol), 4,6-dimethyl-2-mercaptopyrimidine (56 mg, 0.4
mmol) and KOtBu (45 mg, 0.4 mmol) in THF (1.0 mL) was
stirred at room temperature for 2 h. Trifluoroacetic acid (20
µL) was added to the reaction mixture, and it was purified by
preparative HPLC to obtain the trifluoroacetic acid salt of 24
as a solid: mp 105 °C (softened); [R]25 -22.4° (MeOH, c 0.7);
D
1H NMR (CD3OD) δ 7.65 (d, J ) 8.2 Hz, 1H), 7.48 (m, 3H),
6.30 (s, 1H), 5.20 (s, 1H), 4.20 (m, 1H), 3.51 (m, 3H), 3.18 (m,
3H), 2.87 (s, 3H), 1.80 (m, 1H); 13C NMR (CD3OD) δ 186.5,
168.5, 164.8, 162.8, 155.8, 149.2, 133.3, 131.2, 130.8, 128.4,
125.3, 107.8, 104.8, 102.1, 89.1, 68.6, 62.5, 57.6, 45.0, 38.2, 24.0;
1
(33 mg, 76%) as a solid: mp 93-95 °C; H NMR (CD3OD) δ
7.08 (s, 1H), 6.76 (s, 1H), 6.33 (s, 1H), 4.16 (s, 1H), 3.56-3.04

Thio- and Oxoflavopiridols
J ournal of Medicinal Chemistry, 2000, Vol. 43, No. 22 4133
Ta ble 5. Summary of the Crystallographic Data Collection,
Reduction, and Refinement
CT) using a Filtermate universal harvester (Packard Instru-
ment Co., Meriden, CT) and the filters were quantitated using
a
TopCount 96-well liquid scintillation counter (Packard
flavopiridol
thioflavopiridol 16
P212121
Instrument Co., Meriden, CT). Dose-response curves were
generated to determine the concentration required to inhibit
50% of kinase activity (IC50). Compounds were dissolved at
10 mM in dimethylformamide (DMF) and evaluated at six
concentrations, each in triplicate. The final concentration of
DMF in the assay equaled 2%. IC50 values were derived by
nonlinear regression analysis and have a coefficient of variance
(SD/mean, n ) 6) ) 16%.
CDK2/Cyclin E Kin a se Assa y. Kinase reactions consisted
of 5 ng of baculovirus expressed GST-CDK2/cyclin E (human)
complex, 0.5 µg GST-RB fusion protein (amino acids 776-928
of retinoblastoma protein), 0.2 µCi [γ-33P]ATP, 25 µM ATP in
50 µL kinase buffer (50 mM Hepes, pH 8.0, 10 mM MgCl2, 1
mM EGTA, 2 mM DTT). Reactions were incubated for 45 min
at 30 °C and stopped by the addition of cold trichloroacetic
acid (TCA) to a final concentration 15%. TCA precipitates were
collected onto GF/C unifilter plates (Packard Instrument Co.,
Meriden, CT) using a Filtermate universal harvester (Packard
Instrument Co., Meriden, CT) and the filters were quantitated
using a TopCount 96-well liquid scintillation counter (Packard
Instrument Co., Meriden, CT). Dose-response curves were
generated to determine the concentration required inhibiting
50% of kinase activity (IC50). Compounds were dissolved at
10 mM in DMF and evaluated at six concentrations, each in
triplicate. The final concentration of DMF in the assay equaled
2%. IC50 values were derived by nonlinear regression analysis
and have a coefficient of variance (SD/mean, n ) 6) ) 14%.
CDK4/Cyclin D1 Kin a se Assa y. Kinase reactions con-
sisted of 150 ng of baculovirus expressed GST-CDK4/cyclin D1
(human), 280 ng of Stag-cyclin D1, 0.5 µg GST-RB fusion
protein (amino acids 776-928 of retinoblastoma protein), 0.2
µCi [γ-33P]ATP, 25 µM ATP in 50 µL kinase buffer (50 mM
Hepes, pH 8.0, 10 mM MgCl2, 1 mM EGTA, 2 mM DTT).
Reactions were incubated for 1 h at 30 °C and stopped by the
addition of cold trichloroacetic acid (TCA) to a final concentra-
tion 15%. TCA precipitates were collected onto GF/C unifilter
plates (Packard Instrument Co., Meriden, CT) using a Filter-
mate universal harvester (Packard Instrument Co., Meriden,
CT) and the filters were quantitated using a TopCount 96-
well liquid scintillation counter (Packard Instrument Co.,
Meriden, CT). Dose-response curves were generated to de-
termine the concentration required inhibiting 50% of kinase
activity (IC50). Compounds were dissolved at 10 mM in DMF
and evaluated at six concentrations, each in triplicate. The
final concentration of DMF in the assay equaled 2%. IC50
values were derived by nonlinear regression analysis and have
a coeficient of variance (SD/mean, n ) 6) )18%.
Kinase assays for PKA and EGFR were carried out using a
TCA precipitation procedure identical to those described
earlier. PKA (2 units) from bovine heart (Sigma Chemical Co.)
was reacted with 500 ng of GST-TK fusion protein substrate
(generated using pGEX-2TK expression construct, Pharmacia)
for 1 h at 30 °C in 50 mM Tris pH 7.5, 10 mM MgCl2 and 25
µM [γ-32P]ATP. EGFR (50 ng catalytic domain produced in bac-
ulovirus infected Sf9 cells as a GST-fusion protein) was reacted
with 1 µg poly(Glu-Tyr) for 1 h at 30 °C in 50 mM Tris pH 7.5,
10 mM MgCl2, 2 mM MnCl2 and 25 µM [γ-32P]ATP.
space group
P212121
real space cell parameters:
a (Å2)
53.27
71.67
72.38
42445
13272
7.76
52.99
70.76
71.87
74340
17687
9.55
b (Å2)
c (Å2)
number of observations
number of unique reflections
Rsymm (on R) (%)
max observed resolution (Å2)
completeness (%)
2.2
1.9
85.3
74.1
crystallographic R-factor (%)
average B-factor (protein) (Å2)
average B-factor (ligand) (Å2)
residues in final model
total number of atoms
rms (bonds) (Å2)
23.7
28.0
22.46
35.37
298
20.98
52.46
298
2426
0.014
1.936
1.810
2427
0.014
1.948
1.696
rms (angles) (deg)
rms (impropers) (deg)
(m, 6H), 2.80 (s, 3H), 2.40 (s, 6H), 1.70 (m, 1H); MS (ESI) 430
(M + H)+. Anal. (C21H23N3O5S‚1.0H2O‚1.3CF3COOH) C, H, N,
F, S.
(()-(3SR,4RS)-2-(P h en yla m in o)-5,7-d ih yd r oxy-8-(3-h y-
d r oxy-1-m et h yl-4-p ip er id in yl)-4H -1-b en zop yr a n -4-on e
(25). A mixture of 19 (50 mg, 0.1 mmol) and aniline (0.1 mL)
was stirred at 120 °C for 16 h, cooled to room temperature
and it was purified by preparative HPLC to obtain the
trifluoroacetic acid salt of 25 (26 mg, 46%) as a solid: mp 243-
1
245 °C dec; H NMR (CD3OD) δ 7.46-7.23 (m, 5H), 6.35 (s,
1H), 5.44 (s, 1H), 4.26 (s, 1H), 3.90-3.06 (m, 6H), 2.90 (s, 3H),
1.76 (m, 1H); 13C NMR (CD3OD) δ 183.4, 164.5, 163.2, 161.3,
155.9, 138.5, 131.0, 127.1, 124.3, 112.0, 103.7, 95.8, 68.9, 62.1,
57.1, 44.5, 36.5, 23.3; MS (ESI) 383 (M + H)+. Anal.
(C21H22N2O5‚0.5H2O‚1.5CF3COOH) C, H, N.
(()-(3S R ,4R S )-2-N -P ip e r id yl-5,7-d ih yd r oxy-8-(3-h y-
d r oxy-1-m et h yl-4-p ip er id in yl)-4H -1-b en zop yr a n -4-on e
(26). A mixture of 19 (50 mg, 0.1 mmol) and piperidine (0.1
mL) was stirred at 120 °C for 2 h, cooled to room temperature
and it was purified by preparative HPLC to obtain the
trifluoroacetic acid salt of 26 (42 mg, 68%) as a solid: mp 81-
1
83 °C; H NMR (CD3OD) δ 6.20 (s, 1H), 4.90 (s, 1H), 4.25 (s,
1H), 3.61-3.18 (m, 10H), 2.92 (s, 3H), 1.85-1.74 (m, 7H); 13
C
NMR (CD3OD) δ 183.0, 164.4, 162.9, 162.0, 155.1, 106.6, 103.4,
101.0, 68.5, 62.2, 57.0, 48.1, 44.6, 37.9, 26.7, 25.4, 23.6; MS
(ESI) 375 (M + H)+. Anal. (C20H26N2O5‚0.75H2O‚2.0CF3COOH)
C, H, N.
Cr ysta l Str u ctu r e of Hu m a n CDK2. Crystals of CDK2
were grown at room temperature by the hanging drop vapor
diffusion method using 35% poly(ethylene glycol) 5000 monom-
ethyl ester, 0.2 M ammonium acetate, 0.1 M HEPES, pH 7.0,
as a reservoir solution. The hanging drop consisted of 10λ of
the concentrated protein mixed with an equal volume of
reservoir solution. Prior to data collection, crystals were
transferred to a vial containing the mother liquor with 2.4-
3.0 nM ligand added and allowed to soak for 3 days. For X-ray
crystallographic data collection, the crystal was transferred
to a vial containing the mother liquor with 15% glycerol added.
The crystal was then flash cooled by transferring the crystal
on a nylon loop to a stream of liquid nitrogen at 100 K for
data collection. Data were collected on a Siemens Hi-Star area
detector system using mirror focused Cu KR radiation from a
Rigaku RU-2000 rotating anode. The X-ray intensity data was
reduced using program XENGEN21 and refined by the method
of simulated annealing using program XPLOR.22 Details of the
data collection and reduction are listed in Table 5.
CDK1/Cyclin B1 Kin a se Assa y. Kinase reactions con-
sisted of 100 ng of baculovirus expressed GST-CDK1/cyclin B1
(human) complex, 1 µg histone HI (Boehringer Mannheim,
Indianapolis, IN), 0.2 µCi [γ-33P]ATP, 25 µM ATP in 50 µL
kinase buffer (50 mM Tris, pH 8.0, 10 mM MgCl2, 1 mM EGTA,
0.5 mM DTT). Reactions were incubated for 45 min at 30 °C
and stopped by the addition of cold trichloroacetic acid (TCA)
to a final concentration 15%. TCA precipitates were collected
onto GF/C unifilter plates (Packard Instrument Co., Meriden,
The MAP kinase and PKC reactions were quantitated by
spotting onto p81 phosphocellulose as described for the PKC
assay system from Gibco BRL. MAP kinase p42, of xenopus
origin (Santa Cruz Biotechnology Inc.), was reacted with 1 mM
substrate peptide (APRTPGGRR) for 1 h at 30 °C in 50 mM
Tris pH 8.0, 10 mM MgCl2, 0.5 mM EGTA, 1mM DTT and 25
µM [γ-32P]ATP. PKC was assayed using the PKC assay system
from Gibco BRL at an ATP concentration of 20 µM. PKC
(mixture of R, â and γ isoforms from rat brain) was purchased
from Upstate Biotechnology.
Clon ogen ic Assa y. Colony growth inhibition was measured
for four human tumor cell lines (HCT116 colorectal carcinoma,
A2780 ovarian carcinoma, A549 lung carcinoma and Mia
PaCa-2 pancreatic carcinoma) using a standard clonogenic

4134 J ournal of Medicinal Chemistry, 2000, Vol. 43, No. 22
Kim et al.
(8) (a) Kiriyama, N.; Nitta, K.; Sakaguchi, Y.; Taguchi, Y.; Yama-
moto, Y. Studies on the Metabolic Products of Aspergillus terreus.
III. Metabolite of the Strain IFO 8835. Chem. Pharm. Bull. 1977,
25, 2593-2601. (b) Kitagawa, M.; Okabe, T.; Ogino, H.; Matsu-
moto, H.; Suzuki-Takahashi, I.; Kokubo, T.; Higashi, H.; Saitoh,
S.; Taya, Y.; Yasuda, H.; Ohba, Y.; Nishimura, S.; Tanaka, N.;
Okuyama, A. Butyrolactone I, a selective inhibitor of cdk2 and
cdc2 kinase. Oncogene 1993, 8, 2425-2432.
(9) (a) Vesely, J .; Havlicek, L.; Strnad, M.; Blow, J .; Donella-Deana,
A.; Pinna, L.; Letham, D. S.; Kato, J .; Detivaud, L.; Leclerc, S.;
Meijer, L. Inhibition of cyclin-dependent kinases by purine
analogues. Eur. J . Biochem. 1994, 224, 771-786. (b) Abraham,
R.; Acquarrrrone, M.; Andersen, A.; Asensi, A.; Belle, R.; Berger,
F.; Bergounioux, C.; Brun, G.; Buquet-Fagot, C.; Fagot, D.; Glab,
N.; Goudeau, H.; Goudeau, M.; Guerrier, P.; Houghton, P.;
Hendriks, H.; Kloareg, B.; Lippai, M.; Marie, D.; Maro, B.;
Meijer, L.; Mester, J .; Mulner-Lorillon, O.; Poulet, S.; Schier-
enberg, E.; Schutte, B.; Vaulot, D.; Verlhac, M. Cellular effects
of olomoucine, an inhibitor of cyclin-dependent kinases. Biol.
Cell. 1995, 83, 105-120 (c) Havlicek, L.; Hanus, J .; Vesely, J .;
Leclerc, S.; Meijer, L.; Shaw, G.; Strnad, M. Cytokinin-Derived
Cyclin-Dependent Kinase Inhibitors: Synthesis and cdc2 Inhibi-
tory Activity of Olomoucine and Related Compounds. J . Med.
Chem. 1997, 40, 408-412.
(10) Schultz, C.; Linj, A.; Leost, M.; Zaharevitz, D. W.; Gussio, R.;
Sausville, E. A.; Meijer, L.; Kunick, C. Paullones, a Series of
Cyclin-Dependent Kinase Inhibitors: Synthesis, Evaluation of
CDK1/Cyclin B Inhibition, and in Vitro Antitumor Activity. J .
Med. Chem. 1999, 42, 2909-2919.
(11) (a) Norman, T. C.; Gray, N. S.; Koh, J . T.; Schultz, P. G. A
Structure-Based Library Approach to Kinase Inhibitors. J . Am.
Chem. Soc. 1996, 118, 7430-7431. (b) Imbach, P.; Capraro, H.;
Furet, P.; Mett, H.; Meyer, T.; Zimmermann, J . 2,6,9-Trisubsti-
tuted Purines: Optimization Towards Highly Potent and Selec-
tive CDK1 Inhibitors. Bioorg. Med. Chem. Lett. 1999, 9, 91-96.
(c) Murthi, K. K.; Dubay, M.; McClure, C.; Brizuela, L.; Boisclair,
M. D.; Worland, P. J .; Mansuri, M. M.; Pal, K. Structure-
Activity Relationship Studies of Flavopiridol Analogues. Bioorg.
Med. Chem. Lett. 2000, 10, 1037-1041.
(12) Vlahos, C. J .; Matter, W. F.; Kwan, Y. H.; Brown, R. F. A specific
Inhibitor of Phosphatidylinositol 3-Kinase, 2-(4-Morpholinyl)-8-
phenyl-4H-1-benzopyran-4-one (LY294002). J . Biol. Chem. 1994,
269, 5241-5248.
(13) (a) Kattige, S. L.; Naik, R. G.; Lakdawalla, A. D.; Alihussein, N.
D.; Rupp, R. H.; Souza, N. J . 4H-Benzopyran-4-one Compounds
Which Have Antiinflammatory Or Immunomodulating Action.
U.S. Patent 4,900,727, Feb 13, 1990. (b) Naik, R. G.; Kattige, S.
L.; Bhat, B.; Alreja, B.; Souza, N. J .; Rupp, R. H. An Antiin-
flammatory Cum Immunomodulatory Piperidinylbenzopyranone
from Dysoxylum Binectariferum: Isolation, Structure and Total
Synthesis. Tetrahedron 1988, 44, 2081-2086.
(14) Noyori, R.; Tokunaga, M.; Kitamura, M. Stereoselective Organic
Synthesis via Dynamic Kinetic Resolution. Bull. Chem. Soc. J pn.
1995, 68, 36-56.
(15) Structure determination protocols used for crystallography:
Lawrie, A. M.; Noble, M. E. M.; Tunnah, P.; Brown, N. R.;
J ohnson, L. N.; Endicott, J . A. Protein kinase inhibition by
staurosporine revealed in details of the molecular interaction
with CDK2. Nat. Struct. Biol. 1997, 4, 796-801.
assay. Briefly, 200 cells/well were seeded into 6-well tissue
culture plates (Falcon, Franklin Lakes, NJ ) and allowed to
attach for 18 h. Assay medium consisted of either McCoy’s 5a,
RPMI-1640, Ham’s F12K or DMEM plus 10% fetal bovine
serum for the HCT116, A2780, A549 and Mia PaCa-2 cell lines,
respectively. Cells were then treated in duplicate with a six
concentration dose-response curve. The maximum concentra-
tion of DMF never exceeded 0.25%. Compound was replaced
and the colonies were fed with fresh media every third day.
Colony number was scored on day 10 using a BioTran III
colony counter (New Brunswick Scientific Co., Edison, NJ ).
The compound concentration required to inhibit 50% of colony
formation (IC50) was determined by nonlinear regression
analysis. The coefficient of variance (SD/mean, n ) 3) ) 30%.
Ack n ow led gm en t. Microanalysis and mass spectra
were kindly provided by the Bristol-Myers Squibb
Department of Analysis Research and Development.
CDK2-expressing baculovirus and purification proce-
dures were provided by Nikola P. Pavletich, Memorial
Sloan-Kettering Cancer Center, New York.
Refer en ces
(1) (a) Hunter, T.; Pines, J . Cyclins and Cancer II. Cyclin and CDK
inhibitors come of age. Cell 1994, 79, 573-582. (b) Sherr, C.
Cancer cell cycles. Science (Washington, DC) 1996, 274, 1672-
1677.
(2) A review article: Pines, J . The cell cycle kinases. Semin. Cancer
Biol. 1994, 5, 305-313.
(3) (a) Nigg, E. A. Cellular substrates of p34cdc2 and its companion
cyclin-dependent kinases. Trends Cell Biol. 1993, 3, 296-301.
(b) Rickert P.; Seghezzi, W.; Shanahan, F.; Cho, H.; Lees, E.
Cyclin C/CDK8 is a novel CTD kinase associated with RNA
polymerase II. Oncogene 1996, 12, 2631-2640. (c) Dynlacht, B.
D.; Moberg, K.; Lees, J . A.; Harlow, E.; Zhu, L. Specific
regulation of E2F family members by cyclin-dependent kinases.
Mol. Cell. Biol. 1997, 17, 3867-3875. (d) Okata, K.; Hisanaga,
S.; Sugita, M.; Okuyama, A.; Murofushi, H.; Kitazawa, H.; Chari,
S.; Bulinski, J . C.; Kishimoto, T. MAP4 is the in vivo substrate
for cdc2 kinase in HeLa cells: identification of an M-phase
specific and a cell cycle-independent phosphorylation site in
MAP4. Biochemistry 1997, 36, 15873-15883.
(4) (a) Xiong, Y.; Hannong, G. J .; Zhang, H.; Casso, D.; Kobayashi,
R.; Beach, D. p21 is a universal inhibitor of cyclin kinases.
Nature 1993, 366, 701-704. (b) Lukas, J .; Parry, D.; Aagaard,
L.; Mann, D. J .; Bartkova, J .; Strauss, M.; Peters, G.; Bartek, J .
Retinoblastoma protein-dependent cell-cycle inhibition by the
tumor suppressor p16. Nature 1995, 375, 503-506. (c) Polyak,
K.; Kato, J . Y.; Solomon, M. J .; Sherr, C. J .; Massague, J .;
Roberts, J . M.; Koff, A. p27Kip 1, a cyclin-dependent Cdk
inhibitor, links transforming growth factor-beta and contact
inhibition to cell cycle arrest. Genes. Dev. 1994, 8, 9-22.
(5) (a) Kamb, A.; Gruis, N. A.; Weaver-Feldhaus, J .; Liu, Q.;
Harshman, K.; Tavtigian, S. V.; Stockert, E.; Day III, R. S.;
J ohnson, B. E.; Skolnik, M. H. A Cell Cycle Regulators, Poten-
tially Involved in Genesis of Many Tumor Types. Science 1994,
264, 436-440. (b) Nobori, T.; Miura, K.; Wu, D. J .; Lois, A.;
Takabayashi, K.; Carson, D. A. Deletions of the cyclin-dependent
kinase-4 inhibitor gene in multiple human cancers. Nature
(London) 1994, 368, 753-756. (c) Pines, J . Cyclins, CDKs and
cancer. Semin. Cancer Biol. 1995, 6, 63-72. (d) Hartwell, L. H.;
Kastan, M. B. Cell cycle control and cancer. Science 1994, 266,
1821-1828.
(16) Filgueira de Azevedo, W., Jr.; Mueller-Dieckmann, H.-J.; Schulze-
Gahmen, U.; Worland, P. J .; Sausville, E. A.; Kim, S.-H.
Structural basis for specificity and potency of a flavonoid
inhibitor of human CDK2, a cell cycle kinase. Proc. Natl. Acad.
Sci. U.S.A. 1996, 93, 2735-2740.
(17) Bairoch, A.; Apweiler, R. The SWISS-PROT protein sequence
database and its supplement TrEMBL in 2000. Nucleic Acids
Res. 2000, 28, 45-48.
(6) (a) Webster, K. R. The therapeutic potential of targeting the cell
cycle. Exp. Opin. Invest. Drugs 1998, 7 (6), 1-23. (b) Rosania,
G. R.; Chang, Y.-T. Targeting hyperproliferative disorders with
cyclin dependent kinase inhibitors. Exp. Opin. Ther. Patents
2000, 10, 215-230. (c) Sielecki, T. M.; Boylan, J . F.; Benfield,
P. A.; Trainor, G. L. Cyclin-Dependent Kinase Inhibitors: Useful
Targets in Cell Cycle Regulation. J . Med. Chem. 2000, 43, 1-18.
(7) (a) Kaur, G.; Stetler-Stevenson, M.; Sebers, S.; Worland, P.;
Sedlacek, H.; Myers, C.; Czech, J .; Naik, R.; Sausville, E. Growth
Inhibition with Reversible Cell Cycle Arrest of Carcinoma Cells
by Flavone L86-8275. J . Natl. Cancer Inst. 1992, 84, 1736-1740.
(b) Carlson, B.; Dubay, M. M.; Sausville, E. A.; Brizuela, L.;
Worland, P. J . Flavopiridol Induces G1 Arrest with Inhibition
of Cyclin-dependent Kinase (CDK) 2 and CDK4 in Human
Breast Carcinoma Cells. Cancer Res. 1996, 56, 2973-2978. (c)
Sedlacek, H. H.; Czech, J .; Naik, R.; Kaur, G.; Worland, P.;
Losiewicz, M.; Parker, B.; Carlson, B.; Smith, A.; Senderowicz,
A.; Sausville, E. Flavopiridol (L868275; NSC649890), a new
kinase inhibitor for tumor therapy. Int. J . Oncol. 1996, 9, 1143-
1168
(18) Losiewicz, M. D.; Carlson, B. A.; Sausville, E. A.; Nauk, R. G.;
Narayanan, V. L.; Worland, P. J . Proceedings of the 88th Annual
Meeting of American Association for Cancer Research, Toronto,
Canada, 1995; p 35 (abstract).
(19) Burley, S. K.; Petsko, G. A. Amino-aromatic interactions in
proteins. FEBS Lett. 1986, 203, 139-143.
(20) Schulze-Gahmen, U.; Brandsen, J .; J ones, H. D.; Morgan, D. O.;
Meijer, L.; Vesely, J .; Kim, S.-H. Multiple modes of ligand
recognition: Crystal structures of cyclin-dependent protein
kinase 2 in complex with ATP and two inhibitors, olomoucine
and isopentenyladenine. Proteins Struct. Funct. Genet. 1995, 22,
378-391.
(21) Howard, A. J .; Nielsen, C.; Xuong, N. H. Software for
a
Diffractometer with multiwire Area Detector. Methods Enzymol.
1985, 114, 452-472.
(22) Bru¨nger, A. T.; Kuriyan, J .; Karplus, J . Crystallographic R-factor
refinement by molecular dynamics. Science 1987, 235, 458-460.
J M000231G