Cancer cells often have a high demand for antiapoptotic proteins in order to resist programmed cell death. CDK9 inhibition selectively targets survival proteins and reinstates apoptosis in cancer cells. We designed a series of 4-thiazol-2-anilinopyrimidine derivatives with functional groups attached to the C5-position of the pyrimidine or to the C4-thiazol moiety and investigated their effects on CDK9 potency and selectivity. One of the most selective compounds, 12u inhibits CDK9 with IC(50) = 7 nM and shows over 80-fold selectivity for CDK9 versus CDK2. X-ray crystal structures of 12u bound to CDK9 and CDK2 provide insights into the binding modes. This work, together with crystal structures of selected inhibitors in complex with both enzymes described in a companion paper, (34) provides a rationale for the observed SAR. 12u demonstrates potent anticancer activity against primary chronic lymphocytic leukemia cells with a therapeutic window 31- and 107-fold over those of normal B- and T-cells.
Cancer cells often have a high demand for antiapoptotic proteins in order to resist programmed cell death. CDK9 inhibition selectively targets survival proteins and reinstates apoptosis in cancer cells. We designed a series of 4-thiazol-2-anilinopyrimidine derivatives with functional groups attached to the C5-position of the pyrimidineor to the C4-thiazol moiety and investigated their effects on CDK9 potency and selectivity. One of the most selective compounds, 12u inhibits CDK9with IC(50) = 7 nM and shows over 80-fold selectivity for CDK9 versus CDK2. X-ray crystal structures of 12u bound to CDK9 and CDK2 provide insights into the binding modes. This work, together with crystal structures of selected inhibitors in complex with both enzymes described in a companion paper, (34) provides a rationale for the observed SAR. 12u demonstrates potent anticancer activity against primary chronic lymphocytic leukemia cells with a therapeutic window 31- and 107-fold over those of normal B- and T-cells.
Cyclin-dependent kinases (CDKs) can generally
be classified into
two main groups based on whether their primary role is in the control
of cell cycle progression or regulation of transcription. Multiple
CDKs control the cell cycle and are considered essential for normal
proliferation, development, and homeostasis. CDK4/cyclin D, CDK6/cyclin
D, and CDK2/cyclin E facilitate the G1-S phase transition by sequentially
phosphorylating the retinoblastoma protein (Rb), while CDK1/, CDK2/cyclin
A, and CDK1/cyclin B are essential for S-phase progression and G2-M
transition, respectively.[1]Most CDK
inhibitors have been developed as potential cancer therapeutics
based on the premise that they might counteract the uncontrolled proliferation
of cancer cells by targeting the cell-cycle regulatory functions of
CDKs. However in recent years, this understanding of the cellular
functions and regulatory roles of CDKs has been challenged.[2,3] The observations that cancer cell lines and some embryonic fibroblasts
lacking CDK2 proliferate normally and that CDK2 knockout mice are
viable[4,5] suggest that this CDK performs a nonessential
role in cell-cycle control. Furthermore, redundancy of CDK4 and CDK6
was also suggested in cells that enter the cell cycle normally.[6] It has been demonstrated that mouse embryos deficient
in CDKs 2, 3, 4, and 6 develop to mid-gestation, as CDK1 can form
complexes with their cognate cyclins and subsequently phosphorylate
Rb protein. Inactivation of Rb in turn activates E2F-mediated transcription
of proliferation factors.[7] In cells depleted
of CDK1/cyclin B, CDK2/cyclin B is readily detectable and can facilitate
G2/M progression.[3] These studies suggest
that specifically targeting individual cell-cycle CDKs may not be
an optimal therapeutic approach because of a high level of functional
redundancy and compensatory mechanisms.By contrast, the hypothesis
that inhibition of transcriptional
CDKs might be an effective anticancer strategy has gained considerable
support following the observation that many cells rely on the production
of short-lived mitotic regulatory kinases and apoptosis regulators
such as Mcl-1 for their survival.[2,8] The transcriptional
CDKs, particularly CDK9/cyclin T and CDK7/cyclin H, are involved in
the regulation of RNA transcription. CDK7/cyclin H is a component
of transcription factor IIH (TFIIH) that phosphorylates the serine-5
residues within the heptad repeats of RNA polymerase II (RNAPII) C-terminal
domain (CTD) to initiate transcription.[9,10] CDK9/cyclin
T, the catalytic subunit of positive transcription elongation factor
P-TEFb,[11,12] phosphorylates two elongation repressors,
i.e., the DRB-sensitive-inducing factor (DSIF) and the negative elongation
factor (NELF), and position serine-2 of the CTD heptad repeat to facilitate
productive transcription elongation.[2,13] While CDK7
is also recognized as a CDK-activating kinase (CAK),[10] CDK9 appears to have a minimal effect on cell-cycle regulation.[14]During the past decade an intensive search
for pharmacological
CDK inhibitors has led to the development of several clinical candidates
and to the realization that inhibition of the transcriptional CDKs
underlies their antitumor activity.[2,15] Flavopiridol
(alvocidib), the first CDK inhibitor to enter clinical trials, is
the most potent CDK9 inhibitor identified to date and has demonstrated
marked antitumor activity in chronic lymphocytic leukemia (CLL).[16,17] Flavopiridol has been shown to inhibit multiple CDKs[18] and other kinases,[19] but the
primary mechanism responsible for its observed antitumor
activity in CLL appears to be the CDK9-mediated down-regulation of
transcription of antiapoptotic proteins.[20,21]R-Roscovitine (seliciclib) is the first orally
bioavailable CDK inhibitor that targets CDK2, CDK7, and CDK9 (IC50 ≈ 0.1, 0.5, and 0.8 μm, respectively).[22−24] During evaluation in phase I oncology monotherapy and combination
chemotherapy clinical trials it was shown to be well tolerated and
some evidence of disease stabilization was reported.[15] Phase II clinical trials are underway in non-small-cell
lung cancer (NSCLC) patients. R-Roscovitine has demonstrated
selective induction of apoptosis in cancer cells by down-regulation
of antiapoptotic proteins through transcriptional CDK inhibition.[25,26] Other CDK inhibitors including AZD5438,[27] R547,[28,29] and AT519[30] have
also been evaluated in clinical trials.While there are several
pan-CDK inhibitors in clinical studies,[27,29−31] CDK9 inhibitors with good potency and selectivity
have only recently emerged.[32,33] To further exploit
the sensitivity of the 4-hetertoarylpyrimidine pharmacophore (type
I, Figure 1) that specifically targets the
CDK9-ATPgatekeeper residue Phe103 and the ribose-binding pocket,
we prepared a series of 5-substituted-2-anilino-4-(thiazol-5-yl)pyrimidines
(type II, Figure 1) and 4-(4-substituted-thiazol-5-yl)-N-phenylpyrimidin-2-amines (type III, Figure 1).
4-(Thiazol-5-yl)-2-(phenylamino)pyrimidine derivatives.Here, we report the synthesis, SAR, crystal structural
analysis
and biological evaluation of a novel class of 5-substituted-2-anilino-4-(thiazol-5-yl)pyrimidines
and 4-(4-substituted-thiazol-5-yl)-N-phenylpyrimidin-2-amines.
Structure–activity relationship (SAR) analysis reveals the
importance of the C5-group of the pyrimidine core, in the context
of a bulky substituted aniline moiety, for CDK9 potency and selectivity.
A nanomolar Ki inhibitor of CDK9, 12u demonstrates excellent selectivity with over 80-fold selectivity
for CDK9 versus CDK2. This compound inhibits cellular CDK9-mediated
RNA polymerase II transcription, reduces the expression level of Mcl-1
antiapoptotic protein, and subsequently triggers apoptosis in humancancer cell lines and primary CLL cells. In a companion paper[34] we also give a detailed structural analysis
of several 5-substituted-2-anilino-4-(thiazol-5-yl)pyrimidines bound
to CDK2 and CDK9.
Chemistry
Synthesis of 2-anilino-4-thiazolpyrimidine
type I compounds shown
in Table 1 was carried out according to the
methods described previously.[32,35] The chemistry for synthesis
of 5-substituted-2-anilino-4-(thiazol-5-yl)pyrimidines (type II) is
outlined in Scheme 1. Treatment of 1-methylthioureawithethyl 2-chloro-3-oxobutanoate in pyridine resulted in ethyl 4-methyl-2-(methylamino)thiazole-5-carboxylate
(1). The 2-methylamino group in 1 was found
to interfere in subsequent reactions and was therefore masked as the tert-butoxycarbonate (2). Alkylation of 2 with cyanomethanide afforded tert-butyl
5-(2-cyanoacetyl)-4-methylthiazol-2-yl(methyl)carbamate (3, R′ = CN) in a 72% yield. Converting 3 to enaminone
(4, R′ = CN) was achieved conveniently by refluxing
in N,N-dimethylformamide–dimethylacetal
(DMF–DMA).[35]4 (R′
= CN) can also be obtained by bromination of 1-(4-methyl-2-(methylamino)thiazol-5-yl)ethanone
(5), followed by treatment withsodium cyanide and then
DMF–DMA. However, because of the requirement for highly toxic
sodium cyanide, this procedure is not recommended for routine synthesis.
Enaminone (6) was treated with SelectFluor[36,37] in methanol at 0 °C, producing 3-(dimethylamino)-2-fluoro-1-(4-methyl-2-(methylamino)thiazol-5-yl)prop-2-en-1-one
(4, R′ = F). The analogue 2-chloro-3-(dimethylamino)-1-(4-methyl-2-(methylamino)thiazol-5-yl)prop-2-en-1-one
(4, R′ = Cl) was obtained by treating 6 withN-chlorosuccinimide.[38] Preparation of alkyl-substituted enaminones (4, R′
= Me, Et, or Pr) started from tert-butyl methyl(4-methylthiazol-2-yl)carbamate
(8), followed by treatment with alkylaldehyde (R′′CH2CHO) in the presence of LDA to yield 9, which
was then oxidized withmanganese dioxide.[39,40] Compound 8 was obtained by condensation reaction between
1-chloropropan-2-one (7) and 1-methylthiourea.
Table 1
Structure and Biological Activity
Summary
kinase inhibition Ki (nM)a
compd
R′
R
CDK9T1
CDK1B
CDK2A
CDK7H
cytotoxicity GI50 (μM),bHCT-116
Ia
H
m-NO2
2
5
3
417
0.09
12a
CN
m-NO2
6
6
1
260
0.04
12b
OH
m-NO2
932
1424
>5000
>5000
3.20
Ib
H
m-SO2NH2
2
6
4
1960
0.05
12c
CN
m-SO2NH2
6
12
4
114
0.55
12d
OH
m-SO2NH2
>5000
>5000
>5000
>5000
2.03
12e
F
m-SO2NH2
4
4
3
91
<0.01
12f
Cl
m-SO2NH2
11
19
10
685
0.03
12g
Me
m-SO2NH2
5
62
34
1176
0.30
12h
Et
m-SO2NH2
98
788
845
1285
3.81
12i
Pr
m-SO2NH2
>5000
ND
>5000
ND
4.50
12j
CN
m-4-acetylpiperazin-1-yl
7
43
43
92
0.22
12k
CN
m-piperazin1-yl
5
42
56
68
0.23
12l
CN
p-4-acetylpiperazin-1-yl
22
45
26
316
0.82
12m
CN
p-piperazin-1-yl
6
79
39
71
0.03
12n
CN
m-piperidin-1-yl
9
35
42
286
0.93
12o
F
m-4-acetylpiperazin-1-yl
7
26
42
302
0.35
12p
F
m-piperazin-1-yl
4
24
20
193
0.31
12q
F
m-morpholin-4-yl
3
18
20
473
0.16
12r
Cl
m-piperazin-1-yl
4
88
45
155
0.26
12s
F
m-1,4-diazepan-1-yl
5
47
85
111
0.64
12t
CN
m-4-acetyl-1,4-diazepan-1-yl
7
91
131
210
0.33
12u
CN
m-1,4-diazepan-1-yl
7
94
568
46
0.42
Ic
H
m-1,4-diazepane-1-yl
19
195
320
433
0.66
The ATP concentrations used in these
assays were within 15 μM of Km,
i.e., 45, 45, 90, and 45 μM for CDK1/cyclin B, CDK2/cyclin A,
CDK7/cyclin H/MAT1, and CDK9/cyclin T1, respectively. The data given
are mean values derived from two replicates. Apparent inhibition constants
(Ki) were calculated from IC50 values and the appropriate Km (ATP)
values for each kinase.[35]
Antiproliferative activity by MTT
48 h assay. The data given are mean values derived from at least three
replicates.
Scheme 1
Synthesis
of 3-(Dimethylamino)-2-(4-methyl-2-(methylamino)thiazole-5-carbonyl)acrylonitrile
and Derivatives
Synthesis
of 3-(Dimethylamino)-2-(4-methyl-2-(methylamino)thiazole-5-carbonyl)acrylonitrile
and Derivatives
Reagents and conditions:
(a)
di-tert-butyl dicarbonate, 4-dimethylaminopyridine
(DMAP), DCM, rt, 1 h, 93%; (b) LDA, MeCN, THF, −78 °C,
1.5 h, 72%; (c) N,N-dimethylformamide–dimethylacetal
(DMF–DMA), reflux, overnight or microwave, Δ, 20–45
min, 30–80%; (d) SelectFluor, MeOH, 0 °C, 1 h; or NCS,
MeOH, rt 0.5 h, 30%; (e) 1-methylthiourea, MeOH, pyridine, rt 4 h,
88%; (f) LDA, R′CH2CHO, MeCN, THF, −78 °C,
1–1.5 h, 53–77%; (g) MnO2, CHCl3, Δ, 3 h, 65–85%.The ATP concentrations used in these
assays were within 15 μM of Km,
i.e., 45, 45, 90, and 45 μM for CDK1/cyclin B, CDK2/cyclin A,
CDK7/cyclin H/MAT1, and CDK9/cyclin T1, respectively. The data given
are mean values derived from two replicates. Apparent inhibition constants
(Ki) were calculated from IC50 values and the appropriate Km (ATP)
values for each kinase.[35]Antiproliferative activity by MTT
48 h assay. The data given are mean values derived from at least three
replicates.Reaction conditions required for preparing 4 greatly
varied depending on the R′ group of the intermediates 3 and 10. With R′ as a carbonitrile, i.e.,
an electron withdrawing group, 4 (R′ = CN) was
conveniently obtained under mild reaction conditions. Conversely,
an electron donating R′ group slowed the rate of reaction,
requiring heat for an extended period of time. The enaminones (4, R′ = Me, Et, or n-propyl), for
instance, were obtained under high temperature microwave conditions
with low yields. The reaction became more difficult with a bulkier
alkyl group; thus, attempting the synthesis of 4, where
R′ = isopropyl, failed even under harsher reaction conditions
because of unfavorable electronic and steric effects of the bulky
isopropyl group. Pyrimidine ring formation reaction was performed
under conditions similar to those we have developed for the synthesis
of 2-anilinoamino-4-(heteroaryl)pyrimidines.[35] The limiting factor in the preparation of type II analogues was
the efficiency of the condensation reactions between the substituted
phenylguanidines (11) and 4 (Scheme 2). In general, microwave-aided protocols were more
effective in terms of reducing reaction times and improving yields
in the pyrimidine condensation reactions compared to conventional
methodology. Analogues 12b and 12d, where
R′ = OH, were obtained from condensation of 4 (R′
= Cl) with corresponding phenylguanidines 11, followed
by in situ hydrolysis.
Scheme 2
Synthesis of 4-(4-Methylthiazol-5-yl)-2-(phenylamino)pyrimidine-5-carbonitrile
and Derivatives
Reagents and conditions:
(a)
2-methoxyethanol, microwave, 200–300 W, Δ, 20–45
min.
Synthesis of 4-(4-Methylthiazol-5-yl)-2-(phenylamino)pyrimidine-5-carbonitrile
and Derivatives
Reagents and conditions:
(a)
2-methoxyethanol, microwave, 200–300 W, Δ, 20–45
min.In order to extend the SARs, we prepared
another series of compounds
with functional group R′ at the C4-position of the thiazol
ring system (type III, Figure 1); the chemistry
is outlined in Scheme 3. 2-Methylaminothiazol-5-ylethanone
derivative 14 was obtained by bromination of 5 in the presence of PTSA to afford 13, which was then
treated withdi-tert-butyl dicarbonate, followed
by reaction withmethyl 2,2-difluoro-2-(fluorosulfonyl)acetate in
the presence of a catalytic amount of CuI.[41] Converting 14 to the corresponding enaminone 15 (R1 = Boc, R′′ = CH2CF3) was achieved by reacting 14 withDMF–DMA
as described above. To prepare analogue 17, 1,1,1-trifluoropentane-2,4-dione
was treated withhydroxy(tosyloxy)iodobenzene,[42] followed by reaction with1-methylthiourea and then di-tert-butyl dicarbonate. tert-Butyl (5-acetylthiazol-2-yl)(methyl)carbamate
(20) was obtained by cyclization reaction between chloroacetone
chloride and N′-carbamothioyl-N,N-dimethylformimidamide (18),[43] giving 19, followed by treating
the latter withdi-tert-butyl dicarbonate. To prepare tert-butyl(5-acetyl-4-cyclopropylthiazol-2-yl)(methyl)carbamate
(24), bromination of 1-cyclopropylethanone (21) yielded 2-bromo-1-cyclopropylethanone (22), which was
subsequently reacted with1-methylthiourea and then di-tert-butyl dicarbonate to afford 4-cyclopropyl-N-methylthiazol-2-amine
(23). Acetylation was achieved by LDA-mediated alkylation
reaction between 23 and acetaldehyde,[39] followed by oxidation of the resulting thiazol-5-ylethanol
intermediate withMnO2. 4-Phenylthiazol derivative (26) was prepared by condensation reaction between 2-chloro-1-phenylethanone
and 1-methylthiourea to afford N-methyl-4-phenylthiazol-2-amine,
followed by the Friedel–Crafts acylation reaction. Finally,
the enaminones (15) were converted to the corresponding
pyrimidines (27a–l) upon treatment
with the appropriate phenylguanidines[32,35] under microwave
irradiation conditions.
Scheme 3
Synthesis of N-Methyl-5-(2-(phenylamino)pyrimidin-4-yl)-4-thiazol-2-amine
Derivatives
We previously identified
a series of 2-herteroaryl-4-anilinopyrimidine CDK inhibitors.[32,35,44] Many of these compounds showed
potent CDK9 inhibitory activity. The lead compounds demonstrated excellent
pharmaceutical properties and in vivo antitumor efficacy.[32] However, CDK9 specificity was not achieved,
as they cross-reacted with other cell cycle CDKs, in particular CDK2.Previously established SARs of 2-anilino-4-(thiazol-5-yl)pyrimidines
(type I, Figure 1) with respect to CDK2 suggested
the importance of substituents at the C2-position in the thiazole
ring.[32,35] It was found that introduction of amino
functions, in the context of either meta- or para-substituted anilines
at the C2-pyrimidine ring, resulted in a significant increase of inhibitory
activity not only against CDK2 but also against CDK9. Cocrystal structures
of some of these inhibitors bound to CDK2 revealed that the thiazole
C2-amino group interacted strongly with the Asp145 residue of CDK2,
enhancing the hydrophobic interactions of the thiazol C4-methyl group
with the Phe80gatekeeper residue of CDK2 (Phe103 in CDK9). Additional
hydrogen bonding interactions between the thiazole C2-amino groups
and Gln131 and Asp86 of CDK2 were also observed. Substitution of the
thiazole C2-amino group with a C2-methylaminoorC2-ethylamino appeared
to have a detrimental effect on CDK2 and CDK4 activity while having
only a minimal effect on CDK9 potency.[32,35] A bulkier
group, such as phenyl, pyridyl, or other heterocycles at this position,
however, led to significantly reduced activity against all CDKs. We
therefore designed a series of 5-substituted-4-thiazolpyrimidines
in the context of the C2-methylamino to improve potency and selectivity
against CDK9.The SAR analysis of the pyrimidinyl C2-aniline
moiety was previously
described.[32,35] It was shown that many meta-
or para-substitutions of the aniline, in the context of the 4-thiazolpyrimidine,
were well tolerated and manipulation of these substituents led to
a number of inhibitors possessing varying CDK selectivity profiles.
In many cases, meta-substituted anilines gave rise to selectivity
for CDK9 over CDK2 compared with their para-substituted aniline analogues.
However, substituents in the ortho position abolished CDK-inhibitory
activity in all cases.It is recognized that the ATP-binding
sites are highly conserved
among kinases,[45] but the nonconserved hydrophobic
region, which is not occupied by ATP, and the so-called “gatekeeper”
region can be exploited for inhibitor design.[46][47−50]A cocrystal structure of flavopiridol bound to CDK9 showed that the
hydrophobic region that accommodates the chlorophenyl ring of flavopiridol
is more open in CDK9 than CDK2. CDK9Gly112 takes the place of CDK2Lys89 and creates a less crowded and a different electrostatic environment.[51] Analysis of the previously published 2-anilino-4-(thiazol-5-yl)pyrimidineCDK2-bound crystal structures and their corresponding models of CDK9
binding[32,35] suggested that an appropriate functional
group at either the C5-pyrimidineor the C4-thiazol moiety might enhance
interactions with the CDK9gatekeeper region. We thus investigated
the potency and selectivity of a series of 5-substituted 2-anilino-4-(thiazol-5-yl)-pyrimidines
against CDKs and characterized their cellular antitumor activity.
The results are summarized in Table 1.Compound Ia (R′ = H, R = m-NO2) is a highly potent pan-CDK inhibitor. Substitution
of hydrogen at C5-pyrimidine in Ia with a carbonitrile
group results in compound 12a (R′ = CN, R = m-NO2) that exhibits a similar potency and selectivity
profile. Both compounds inhibit CDK9, CDK1, and CDK2 potently with Ki values ranging from 1 to 6 nM but are significantly
less active toward CDK7. Both compounds are highly effective antiproliferative
agents with respective GI50 values of 90 and 40 nM in the
HCT-116humancolon cancer cell line. Replacement of the C5-carbonitrilewith a C5-hydroxyl group in 12b (R′ = OH, R = m-NO2) results in over 155-fold and 230-fold
loss in CDK9 and CDK1 inhibition, respectively. This replacement also
abolishes CDK2 and CDK7 inhibitory activity and significantly reduces
cellular antiproliferative activity. A compound containing the m-sulfonamideaniline ring, 12c (R′ =
CN, R = m-SO2NH2), shows similar
potencies against CDK1, CDK2, and CDK9 but a 17-fold or a 10-fold
loss in CDK7 inhibition and cellular toxicity, respectively, compared
to Ib (R′ = H, R = m-SO2NH2). Again, introducing a hydroxyl group at C5-pyrimidine,
in the context of m-sulfonamideaniline, is not tolerated; 12d (R′ = OH, R = m-SO2NH2) shows little biological activity in the enzymatic
and cellular assays. These results demonstrate the importance of C5-substitution
of the pyrimidine and that a protic orhydrogen bond donating function
at this position has a detrimental effect on biological activity.
Compound 12e (R′ = F, R = m-SO2NH2), a potent pan-CDK inhibitor (Ki = 3–7 nM), is the most potent antiproliferative
agent of this series, with GI50 < 10 nM against HCT-116
cells. Analogue 12f, where R′ = Cl, R = m-SO2NH2, however, displays a >3-fold
reduced CDK inhibitory activity and cellular potency compared to 12e. A more interesting trend toward CDK9 selectivity is observed
withC5-alkylpyrimidines; 12g (R′ = Me, R = m-SO2NH2) exhibits a CDK9 inhibitory
potency similar to that of 12ewith Ki = 5 nM but enhances selectivity for CDK9with >7-fold
lower effectiveness against other CDKs. However, this selectivity
results in 12g showing over 30-fold reduced cytotoxicity
in HCT-116 cells compared to 12e. With further introduction
of a bulkier alkyl group, CDK9 inhibitory activity dramatically decreases;
thus, 12h (R′ = Et, R = m-SO2NH2) is 20-fold less potent against CDK9 than 12g, while 12i (R′ = Pr, R = m-SO2NH2) is not active against CDKs at concentrations
up to 5 μM. As expected, 12h and 12i are also less cytotoxic in cancer cells with respective GI50 values of 3.81 and 4.50 μM.Retaining the C5-carbonitrilepyrimidine core but replacing the m-sulfonamide with
a bulkier 1-(piperazin-1-yl)ethanoneorpiperazine leads to the corresponding 12j (R′
= CN, R = m-1-(piperazin-1-yl)ethanone) or 12k (R′ = CN, R = m-piperazine). This
not only maintains CDK9 potency but also increases selectivity (∼10-fold)
over CDK2 compared to 12c, indicating the tolerance of
a large ring system in the corresponding CDK9 binding region. Compounds 12m–r, bearing heterocyclic piperidine,
1-(piperazin-1-yl)ethanone, piperazine, ormorpholine at the meta-
or para-position of the aniline in the context of a C5-carbonitrileor C5-halogen pyrimidine moiety, display favorable CDK9 inhibitory
activity withlow nanomolar potencies and possesses over 4-fold selectivity
for CDK9. The exception is compound 12l (R′ =
CN, R = p-1-(piperazin-1-yl)ethanone), which shows
a >3-fold loss of potency against CDK9. All these analogues demonstrate
excellent antiproliferative activity with GI50 values ranging
from 0.03 to 0.93 μM.Introduction of a bulkier heterocylic
(1,4-diazepan-1-yl)ethanoneor1,4-diazepane at the meta-position of the aniline affords 12s–u, displaying appreciable selectivity
for CDK9 versus CDK2. Compound 12u, in particular, shows
a >80-fold enhanced CDK9 selectivity over CDK2. Compounds 12s–u effectively inhibit tumor cell growth
with
GI50 values of 0.64, 0.33, and 0.42 μM, respectively.
Replacement of the C5-carbonitrileor C5-fluoridewith a C5-hydrogen
affords Ic (R′ = H, R = m -1,4-diazepane).
However, this replacement results in a >2-fold loss in CDK9 inhibitory
activity but a more significant drop in CDK2 selectivity when compared
with12s and 12u. These further support
the role of the carbonitrileorfluoride substitution at the C5-pyrimidine
in favoring potency and selectivity against CDK9 over CDK2.In general, all C5-substituted pyrimidine analogues are also potent
CDK1 inhibitors, with activity comparable to that of CDK2 as shown
in Table 1. An exception is compound 12u which targets CDK1 and CDK2with Ki values of 94 and 568 nM, respectively, being 6-fold more
potent for CDK1 than for CDK2. It is apparent that this combined inhibition
of CDK9, CDK1, and CDK2 results in significant cytotoxicity in cancer
cells. 12e, a nanomolar CDK9, CDK1, and CDK2 inhibitor,
for example, is the most potent cytotoxic agent of this chemical class,
with GI50 < 10 nM against HCT-116 cells. This is consistent
with the finding that cancer cells expressing shRNA targeting a combination
of CDK2, CDK1, and CDK9 were most effective in induction of apoptosis
of cancer cells, and targeting CDK9, CDK1, and CDK2 has been proposed
as an anticancer strategy.[3]Most
of the analogues described here are significantly less effective
as CDK7 inhibitors when compared to their activity against other CDKs,
suggesting that CDK7 inhibition is not a requirement for the observed
cellular cytotoxicity: many compounds demonstrate excellent antiproliferative
activity irrespective of modest CDK7 inhibition (Table 1). Compounds Ib and 12g, for example,
inhibit CDK7with Ki > 1 μM,
but
both exhibit excellent antiproliferative activity with GI50 = 0.05 and 0.30 μM, respectively.In order to assess
whether modification of the C4-methyl of the
thiazole is tolerated, we prepared a series of substituted C4-thiazolpyrimidine
derivatives; the SAR is summarized in Table 2. Replacement of the C4-methyl with phenyl (27a, R′
= Ph, R = m-NO2) is not tolerated, and
no inhibitory activity against CDKs is detected up to 5 μM.
However, this compound exhibits potent antiproliferative activity
with a GI50 of 60 nM, indicating potential non-CDK kinase
targets. Substitution of C4-trifluoroethyl, in the context of m-nitroanilinopyrimidine, yields 27b (R′
= CH2CF3, R = m-NO2). This compound exhibits excellent selectivity for CDK9 (Ki = 0.134 μM), being inactive against
other CDKs at concentrations up to 5 μM. Despite its high selectivity, 27b still displays good antiproliferative activity with GI50 < 1 μM in HCT-116cancer cells. However, analogues 27c (R′ = CH2CF3, R = m-OH), 27d (R′ = CH2CF3, R = p-OH), and 27e (R′
= CH2CF3, R = m-SO2NH2) reduce CDK9 inhibition and selectivity when compared
to 27b. Keeping the benzenesulfonamide moiety but replacing
the 4C-trifluoroethyl with a C4-trifluoromethyl, C4-hydrogen, or C4-cyclopropyl
affords compounds 27f–k. All of these
compounds possess significantly enhanced inhibitory activity not only
against CDK9 but also against other CDKs. As expected, these compounds
are extremely cytotoxic to cancer cells with GI50 values
in the range of 0.01–0.41 μM. However, substitution of
the benzenesulfonamidewith1,4-diazepan-1-ylaniline yields 27l (R′ = cyclopropyl, R = m-1,4-diazepane).
This analogue shows a significant loss of activity which suggests
that the benzenesulfonamide moiety is a key contributor to optimal
CDK inhibition and cellular potency of this series.
Table 2
Structure and Biological Activity
Summary
kinase inhibition Ki (nM)a
compd
R′
R
CDK9T1
CDK1B
CDK2A
CDK7H
cytotoxicity GI50 (μM),bHCT-116
27a
Ph
m-NO2
>5000
>5000
>5000
>5000
0.06
27b
CH2CF3
m-NO2
134
>5000
>5000
>5000
0.90
27c
CH2CF3
m-OH
245
556
282
3474
2.60
27d
CH2CF3
p-OH
278
334
347
1543
3.70
27e
CH2CF3
m-SO2NH2
156
154
226
1984
0.91
27f
CF3
m-SO2NH2
2
1.5
2
47
0.05
27g
CF3
p-SO2NH2
3
0.5
1.5
64
0.01
27h
H
p-SO2NH2
5
1
1
54
0.13
27i
H
m-SO2NH2
3
4
5
40
0.02
27j
cyclopropyl
m-SO2NH2
16
18
2
273
0.41
27k
cyclopropyl
p-SO2NH2
8
2
13
255
0.08
27l
cyclopropyl
m-1,4-diazepan-1-yl
66
176
326
464
0.77
The ATP concentrations used in these
assays were within 15 μM of Km.
The data given are mean values derived from two replicates. Apparent
inhibition constants (Ki) were calculated
from IC50 values and the appropriate Km (ATP) values for each kinase.[35]
Antiproliferative activity
by MTT
48 h assay. The data given are mean values derived from at least three
replicates.
The ATP concentrations used in these
assays were within 15 μM of Km.
The data given are mean values derived from two replicates. Apparent
inhibition constants (Ki) were calculated
from IC50 values and the appropriate Km (ATP) values for each kinase.[35]Antiproliferative activity
by MTT
48 h assay. The data given are mean values derived from at least three
replicates.
Cocrystal Structures of 12u Bound to CDK9/Cyclin
T and CDK2/Cyclin A
As one of the most selective CDK9 inhibitors
in the chemical series, 12u was cocrystallized withCDK9/cyclin
T and CDK2/cyclin A in order to explain the observed SAR. The crystal
structure and refinement data are summarized in Table 3. A more thorough rationalization of the SARs provided by
the determination of five additional inhibitor cocrystal structures
bound to CDK9/cyclin T and CDK2/cyclin A is provided in the companion
paper.[34] As shown in Figure 2, 12u adopts a similar binding mode within the
CDK9 and CDK2ATP binding sites located between the N- and C-terminal
lobe, and the thiazole, pyrimidine, and aniline moieties occupy similar
positions. In both CDK9 and CDK2, 12uhydrogen-bonds
with the kinase hinge regions. The N1-pyrimidine accepts a hydrogen
bond from the peptide nitrogen of Cys106 (Leu83 in CDK2), while the
C2-NH of the pyrimidine ring donates a hydrogen bond to the peptide
carbonyl of Cys106. At the back of the ATP binding site the C5-carbonitrile
group exploits the hydrophobic region close to the gatekeeper residue
Phe103 (Phe80 in CDK2) to form a favorable lone pair−π
interaction. The CDK2/cyclin A/12u structure was determined
at a higher resolution and shows a water molecule trapped in a pocket
behind the C5-carbonitrile (Figure 2B and Figure 2D). This water molecule forms a hydrogen-bond network
with the backbone of residue Asp145 and with the side chain of Glu51.
In the adenine site the pyrimidine ring is sandwiched between the
hydrophobic side chains of Ala46 (Ala31 in CDK2) and Leu156 (Leu134
in CDK2), with which it forms extensive van der Waals interactions.
The hydrogen of the C2-methylaminothiazole binds to Asp167 in CDK9
and to the corresponding residue Asp145 in CDK2. At the front of the
ATP binding pocket, the aniline ring is contacted from above by Ile25
(Ile10 in CDK2) to make favorable van der Waals interactions with
both enzymes.
Table 3
Crystallographic Parameters of CDK2/Cyclin
A/12u and CDK9/Cyclin T/12u
CDK2/cyclin A/12u
CDK9/cyclin T/12u
Data Collection
beamline
Diamond I-03
Diamond I-03
space group
P212121
H3
unit cell (Å)
a = 73.81; b = 134.55; c = 149.17
a = b = 172.80; c = 98.88
unit cell (deg)
α = β = γ = 90
α = β = 90; γ = 120
resolution (highest resolution
shell) (Å)
29.83–2.26(2.38–2.26)
86.40–3.08(3.25–3.08)
total observations
308028 (43211)
70331 (10036)
unique observations
69656 (9807)
20079 (2956)
Rmerge
0.071 (0.496)
0.058 (0.511)
multiplicity
4.4 (4.4)
3.5 (3.4)
mean I/σI
13.4 (2.8)
15.4 (2.4)
completeness (%)
99.2 (96.6)
99.0 (99.6)
Refinement Statistics
highest
resolution shell
(Å)
2.29–2.26
3.09–3.17
total number of atoms
9233
4649
total number of waters
363
16
R
18.10 (23.09)
16.11 (28.6)
Rfree
21.71 (28.66)
19.72 (34.2)
rms bonds (Å)
0.003
0.011
rms angles
(deg)
0.702
1.433
Figure 2
Cocrystal structures of 12u bound to CDK9/cyclin
T1
(PDB code 4BCG) and CDK2/cylin A (PDB code 4BCP). The structures of CDK9/T1/12u (A, C)
are shown. Compound 12u bound to CDK2/cyclin
A (B, D) and showing two binding orientations. Electron density around 12u is shown as a wire mesh (A, B). Selected CDK9 and CDK2
residues are drawn in ball-and-stick representation. Hydrogen bonds
in all panels are depicted by dotted lines.
Cocrystal structures of 12u bound to CDK9/cyclin
T1
(PDB code 4BCG) and CDK2/cylin A (PDB code 4BCP). The structures of CDK9/T1/12u (A, C)
are shown. Compound 12u bound to CDK2/cyclin
A (B, D) and showing two binding orientations. Electron density around 12u is shown as a wire mesh (A, B). Selected CDK9 and CDK2
residues are drawn in ball-and-stick representation. Hydrogen bonds
in all panels are depicted by dotted lines.The very weak electron density of the 1,4-diazepan-1-ylaniline
moiety of 12u suggests that it is not bound tightly to
CDK2. Two conformations of 12u were consistent with the
observed electron density and represent possible alternative binding
modes. Neither of these modes suggests either favorable or unfavorable
interactions, correlating to the relative absence of electron density
for the 1,4-diazepane ring. In the CDK9 complex, however, the 1,4-diazepane
ring clearly adopts an “inward” conformation orientated
toward the thiazole ring. This may be more favorable for 12u because of a reduced solvent exposure of the hydrophobic 1,4-diazepane
ring in this orientation.Although the inhibitor interactions
withCDK2 and CDK9 are mostly
conserved, there are, however, two significant differences. First,
as seen for other CDK9/cyclin T inhibitor structures described in
the companion paper,[34] the binding of the
inhibitor induces a lowering of the glycine-rich loop into the ATP
binding site. The loop adopts several conformations, and this inherent
flexibility is reflected in the higher b-factors.
By contrast, the conformation of the glycine rich loop in CDK2/cyclin
A appears to be relatively unaltered on inhibitor binding (PDB 2JGZ). Second, in comparison with the apo structure of
CDK9/cyclin T (PDB 3BLH)[51] the backbone of the hinge region adapts
to inhibitor binding by shifting away from 12u. This
shift enables a hydrogen bond to form between the C2-NH of the pyrimidinewith the peptide carbonyl of Cys106 at an optimal length of 2.8 Å.
These two observations support the hypothesis that CDK9 has a more
flexible ATP binding pocket than CDK2.We propose that the greater
flexibility of the ATP-binding site
of CDK9 enables the large flexible anilino-1,4-diazepine of 12u, in the context of the C5-carbonitrilepyrimidine moiety,
to be well accommodated by CDK9. In contrast, the crystal structure
of 12u bound to CDK2 shows that this ring adopts an orientation
either “inward” or “outward”, suggesting
that the CDK2 binding pocket is too crowded for 12u.
This variation in the ability of the kinases to adapt and readily
accommodate inhibitors offers an explanation for the high potency
and selectivity of 12u toward CDK9.
Compound 12u Is a Potent Antiproliferative Agent
Compound 12u was screened against a panel of kinases
using biochemical assays and showed no inhibitory activity at concentrations
up to 5 μM against a panel of kinases, including BCR-Abl, CaMK1,
IKK, Lck, MARK2, PKA, PKB, PKC, and cSRC (Table 4). The antiproliferative effects of 12u against a panel
of nine tumor cell lines and three nontranformed cell lines were examined
using a 48 h MTT assay as summarized in Table 5A. To investigate cell-type sensitivity, we included HCT-116 colon
carcinoma (wild-type and mutant p53, respectively), MCF-7 breast carcinoma
(wild-type p53, pRb positive, ER positive and containing CDK4/cyclin
D and CDK6/cyclin D), and MDA-MB-468breast carcinoma cells (mutant
p53, pRb negative, ER negative and lacking CDK4/cyclin D and CDK6/cyclin
D),[52] and other cell lines. Similar sensitivity
is observed for cells with different p53, Rb, and CDK4/6 status. Compound 12u suppresses tumor cell proliferation with GI50 values ranging from 0.38 to 0.78 μM, irrespective of the tumor
cell type. However, all nontransformed cell lines, i.e., microvascular
endothelial cell line HMEC-1 and embryonic lung fibroblasts MRC-5
and WI-38, are significantly less sensitive to 12u treatment
(GI50 = 3.12–5.96 μM). The time-course assays
were performed using A2780ovarian cancer and MRC-5 and HMEC-1 nontransformed
cell lines. As shown in Table 5B, 24 h treatment
with12u, as well as withflavopiridol, is sufficient
to achieve maximal growth inhibition of 12u in A2780cancer cells. Again, 12u is significantly less toxic
in the HMEC-1 and MRC-5 nontransformed cells. In contrast, flavopiridol
fails to demonstrate any significant differential effects between
the cancer cells and noncancerous cell lines.
Table 4
Inhibitory Activity of 12u against Protein Kinases
protein kinases
inhibition
IC50 (nM)a
CDK1/cyclin B
188
CDK2/cyclin A
1126
CDK6/cyclin
D3
>5000
CDK7/cyclin H
92
CDK9/cyclin T1
14
BCR-Abl
>5000
CaMK1
>5000
IKKβ
>5000
Lck
>5000
MAPK2
>5000
PKA
>5000
PKBα
>5000
PKCα
>5000
cSRC
>5000
The ATP concentrations used in these
assays were within 15 μM of Km.
The data given are mean values derived from two replicates.
Table 5
Antiproliferative Activity of 12u and Flavopiridol against a Panel of Human Humour and Nontranformed
Cell Lines (A) and Activity by MTT Time-Course Experiments (B)a
A
human cell
line
origin
designation
48 h MTT,GI50 ± SD (μM)
colon carcinoma
HCT-116 (p53wt, pRb+)
0.420 ± 0.020
HCT-116
(p53null)
0.780 ± 0.060
HCC
2998 (p53wt, pRb+)
0.385 ± 0.091
breast carcinoma
MCF-7 (p53wt, pRb+, ER+)
0.690 ± 0.020
MDA-MB468 (p53mut, pRb–, ER–)
0.402 ± 0.026
ovarian carcinoma
A2780
0.320 ± 0.010
cervical carcinoma
HeLa
0.630 ± 0.140
renal carcinoma
TK 10
0.747 ± 0.076
pancreatic carcinoma
PANC-1
0.590 ± 0.078
microvascular endothelial
HMEC-1
3.120 ± 0.320
embryonic lung fibroblast
MRC-5
5.960 ± 0.720
WI-38
5.490 ± 0.630
The data given are mean values
derived from at least three replicates ± SD.
The ATP concentrations used in these
assays were within 15 μM of Km.
The data given are mean values derived from two replicates.The data given are mean values
derived from at least three replicates ± SD.
Compound 12u Effectively Induces Cancer Cell Apoptosis
Cell death induced by therapeutic agents can occur through caspase-dependent
or -independent apoptosis or by necrosis. To assess whether apoptosis
is contributing to the cytotoxic effect of 12u, we used
annexin V/PI (propidium iodide) surface staining in A2780cancer cells
following treatment with12u for 48 h (Figure 3A). Compound 12u induced cell apoptosis
at the GI50 (the concentration of 12u required
to inhibit 50% of cell proliferation by MTT assay) in a dose-dependent
manner. At the GI50 concentration 12u causes
38% annexin V-positive cells and the percentage increases to 50% at
5GI50. Withflavopiridol the same treatments results in
39% and 54% apoptotic cells at GI50 and 5GI50, respectively. Concurrent treatment with 5GI50 of either 12uorflavopiridol together with 50 μM of the pan-caspase
inhibitor Z-VAD-fmk suppresses apoptosis, suggesting a caspase-dependent
mechanism of apoptosis induction.[18]
Figure 3
Cellular mode
of action of 12u. (A) A2780 cells were
exposed to 12u, flavopiridol, or 12u (or
flavopiridol) + 25 μM Z-VAD-fmk for 48 h and analyzed by annexin
V/PI staining. The percentage of cells undergoing apoptosis was defined
as the sum of early apoptosis (annexin V-positive cells) and late
apoptosis (annexin V-positive and PI-positive cells). (B) 12u induces caspase-3 activity in A2780 after treatment for a period
of 24 h, but 12u is less effective in nontransformed
endothelial cell line HMEC-1 upon the same treatment. Vertical bars
represent the mean ± SD of two independent experiments. Values
significantly different from DMSO vehicle control are marked with
asterisks: (∗) P < 0.05; (∗∗) P < 0.0001. (C) Cell-cycle analysis of A2780 cells treated
with 12u or flavopiridol for 24 h. (D) Western blot analysis
of A2780 cells treated with 12u or flavopiridol for 24
h. DMSO diluent was used as control in each experiment, and β-actin
antibody was used as an internal control.
Cellular mode
of action of 12u. (A) A2780 cells were
exposed to 12u, flavopiridol, or12u (orflavopiridol) + 25 μM Z-VAD-fmk for 48 h and analyzed by annexin
V/PI staining. The percentage of cells undergoing apoptosis was defined
as the sum of early apoptosis (annexin V-positive cells) and late
apoptosis (annexin V-positive and PI-positive cells). (B) 12u induces caspase-3 activity in A2780 after treatment for a period
of 24 h, but 12u is less effective in nontransformed
endothelial cell line HMEC-1 upon the same treatment. Vertical bars
represent the mean ± SD of two independent experiments. Values
significantly different from DMSO vehicle control are marked with
asterisks: (∗) P < 0.05; (∗∗) P < 0.0001. (C) Cell-cycle analysis of A2780 cells treated
with12uorflavopiridol for 24 h. (D) Western blot analysis
of A2780 cells treated with12uorflavopiridol for 24
h. DMSO diluent was used as control in each experiment, and β-actin
antibody was used as an internal control.Activation of caspase-3 activity by 12u was confirmed
in A2780cancer cell following exposure to drug for 24 h and was used
to compare that in HMEC-1 untransformed endothelial cells (Figure 3B). Compound 12u significantly activates
caspase-3 activity in the tumor cells starting at GI50 concentration
(p < 0.001), and the effect is further enhanced
at higher concentrations. In contrast, no such activity is detected
in the HMEC-1 cells up to 10GI50 concentrations of 12u. These results confirm that the cytotoxicity induced by 12u is mediated through the preferential induction of apoptosis
in cancer cell lines and corroborates the MTT cytotoxic potency.As 12u showed potent CDK1 inhibition in biochemical
kinase assays, we next investigated its effects on cell cycle progression.
A2780 cells were treated with12u (orflavopiridol) for
a period of 24 h at GI50 and 5GI50 concentrations,
respectively (Figure 3C). The cells showed
no alteration in cell cycle distribution at concentrations less than
5GI5012u, at which concentration accumulation
of cells in G2/M phase of the cell cycle was detected. This confirms
that 12u has a lower cellular CDK1 inhibitory activity
compared to that of CDK9. A similar cell cycle profile is observed
withflavopiridol (Figure 3C).
Compound 12u Inhibits CDK9 Activity and Down-Regulates
Mcl-1 in Cancer Cells
We next investigated the cellular mode
of action of 12u by Western blot analysis (Figure 3D). Treatment of A2780 cells with12u for a period of 24 h showed that phosphorylation at Ser-2 CTD of
RNAPII was reduced at the GI50 and abrogated at 5GI50, confirming cellular CDK9 inhibition. The same treatment
caused down-regulation of Mcl-1 and HDM2 but had little effect on
the levels of Bcl-2 expression. Induction of apoptosis was indicated
by PARP cleavage. Analogous results were obtained withflavopiridol,
with inhibition of the phosphorylation of Ser-2 of RNAPII CTD, reduction
of Mcl-1 and HDM2, and induction of cleaved PARP being observed.
Ex Vivo Antitumor Activity in Primary Chronic Lymphocytic Leukemia
Cells
The potency and selectivity of 12u were
further evaluated in patient-derived CLL cells (Table 6), as well as age-matched normal B-cells and T-cells, using
an annexin V-FITC apoptosis assay. As shown in Figure 4A, the compound exhibits excellent activity with a mean LD50 of 2.60 μM ± 1.1 μM against CLL cells (the
concentration of 12u required to kill 50% of the CLL
cells following exposure for 48 h). Figure 4B shows that 12u induces a dose-dependent increase in
apoptosis in CLL cells as denoted by an increased annexin V positivity.
In contrast, little toxicity is observed in the normal B- and T-cells
with LD50 of >80 and >280 μM, respectively
(Figure 4A). To determine whether the cytotoxicity
induced
by 12u is caspase-dependent, primary CLL cells were incubated
with various concentrations of 12u for 24 h, followed
by flow cytometric assessment of active caspase-3. As shown in Figure 4C, the caspase-3 activity is significantly induced
at 1.0 μM 12u (P < 0.05) and
is further enhanced in a dose-dependent manner at 5 μM (P < 0.0001) and 10 μM (P <
0.0001) when compared with untreated controls. These data support
the conclusion that 12u-induced cytotoxicity is mediated
via the activation of the effector caspase-3.
Table 6
Clinical Characteristics of the CLL
Patients (n = 10) in This Study
patient characteristics
number
mean age (years)
68
sex (male/female)
7/3
previously untreated/treated
10/0
Binet stage (A/B/C)
8/2/0
IGHV gene
mutation (mutated/unmutated)
8/2
CD38 expression (<20%/≥20%)
7/3
ZAP-70 expression (<20%/≥20%)
7/3
Figure 4
Compound 12u shows selective toxicity against CLL
cells and induces its effects through the induction of apoptosis.
Primary CLL cells and normal B-cells were cultured in the presence
of increasing concentrations of 12u for 48 h. (A) Sigmoidal
dose–response curves for 12u in CLL cells and
normal B-cells and T-cells. (B) Compound 12u caused a
dose-dependent increase in annexin V-positive cells, and this was
preceded by (C) a dose-dependent increase in caspase-3 activation
after 24 h in culture ((∗) P < 0.05, (∗∗) P < 0.0001). (D) Mcl-1 protein expression was significantly
inhibited by 12u at 8 h in all the primary CLL samples
tested (P < 0.0001).
Compound 12u shows selective toxicity against CLL
cells and induces its effects through the induction of apoptosis.
Primary CLL cells and normal B-cells were cultured in the presence
of increasing concentrations of 12u for 48 h. (A) Sigmoidal
dose–response curves for 12u in CLL cells and
normal B-cells and T-cells. (B) Compound 12u caused a
dose-dependent increase in annexin V-positive cells, and this was
preceded by (C) a dose-dependent increase in caspase-3 activation
after 24 h in culture ((∗) P < 0.05, (∗∗) P < 0.0001). (D) Mcl-1 protein expression was significantly
inhibited by 12u at 8 h in all the primary CLL samples
tested (P < 0.0001).CLL cells are characterized by resistance to apoptosis
mediated
by up-regulation of Bcl-2 family proteins. Mcl-1 is the most important
antiapoptotic member of the Bcl-2 protein family and is overexpressed
in the majority of patientswith CLL at baseline. Increased levels
of Mcl-1 are associated with both drug resistance and inferior survival.[53,54] Down-regulation of Mcl-1 is sufficient to induce apoptosis in CLL
cells.[55] A correlation between lower Mcl-1
protein and mRNA levels with known biologic prognostic markers and
improved outcomes in patientswith CLL has been reported.[53,56] In the present study, primary CLL cells derived from 10 patients
were cultured with 1 μM 12u for 8 h and examined
for the effect on Mcl-1 protein. Figure 4D
shows that the levels of Mcl-1 protein expression are consistently
inhibited by 12u in all CLL patient samples (P < 0.0001) irrespective of stages of the disease. The
change in Mcl-1 protein expression precedes evidence of apoptosis
induction suggesting that the inhibition is a trigger for apoptosis
rather than a consequence of apoptosis induction.
Conclusion
In this communication we describe the synthesis
and SAR of a series
of 5-substituted-4-(thiazol-5-yl)-2-(phenylamino)pyrimidines and 4-(4-substituted-thiazol-5-yl)-N-phenylpyrimidin-2-amines. Many compounds inhibit CDK9
activity at low nanomolar concentrations and exhibit very potent antiproliferative
activity in tumor cells. Optimization led to the discovery of 12u, one of the most selective CDK9 inhibitors in the series,
being >80-fold more potent for CDK9 versus CDK2. The cocrystal
structures
of 12u bound in CDK9/cyclin T and CDK2/cyclin A provide
a rationale for the observed potency and selectivity. Compound 12u was examined in more detail regarding its cellular mode
of action. The study demonstrates that by inhibiting cellular RNAPII
transcriptional activity, 12u mediates down-regulation
of the antiapoptotic protein Mcl-1, thereby rendering cells sensitive
to apoptosis. Significantly, 12u exhibits excellent antitumor
activity in primary CLL cells but shows little toxicity in healthy
normal B- and T-cells. In keeping with this finding, Mcl-1 is not
detectable in normal B- and T-cells (data not shown), indicating that
Mcl-1 may not be a critical regulator of survival in normal lymphocytes.
In contrast, CLL cells appear to have a requirement for this protein
in order to maintain viability.[53] This
study provides a rationale for further development of CDK9 inhibitors
for the treatment of CLL and other humanmalignancies.
Experimental Section
Chemistry
Chemical reagents and solvents were obtained
from commercial sources. When necessary, solvents were dried and/or
purified by standard methods. 1H NMR and 13C
NMR spectra were obtained using a Bruker 400 Ultrashield spectrometer
at 400 and 100 MHz, respectively. These were analyzed using the Bruker
TOPSPIN 2.1 program. Chemical shifts are reported in parts per million
relative to internal tetramethylsilane standard. Coupling constants
(J) are quoted to the nearest 0.1 Hz. The following
abbreviations are used: s, singlet; d, doublet; t, triplet; q, quartet;
m, multiplet; br, broad. High resolution mass spectra were obtained
using a Waters 2795 single quadrupole mass spectrometer/micromass
LCT platform. Purity for final compounds was greater than 95% and
was measured using Waters high performance liquid chromatography (Waters
2487 dual λ absorbance detector) with Phenomenex Gemini-NX 5u
C18 110A 250 mm × 4.60 mm column, UV detector at 254 nm, using
system A (10% MeOH containing 0.1% TFA for 4 min, followed by linear
gradient 10–100% MeOH over 6 min at a flow rate of 1 mL/min),
system B (10% MeCN containing 0.1% TFA for 2 min, followed by linear
gradient 10–100% over 10 min at a flow rate of 1 mL/min), and
system C (10% MeCN containing 0.1% TFA for 4 min, followed by linear
gradient 10–100% over 10 min at a flow rate of 1 mL/min). Flash
chromatography was performed using either a glass column packed withsilica gel (200–400 mesh, Aldrich Chemical) or prepacked silica
gel cartridges (FlashMaster systems, Biotage). Melting points were
determined with an Electrothermal melting point apparatus.
To a solution of ethyl 2-chloroacetoacetate (13.8 mL, 100
mmol) in 100 mL of MeOH were added 1-methylthiourea and 3 mL of pyridine,
and the mixture was stirred at room temperature for 4 h. The mixture
was concentrated and the precipitate was washed with saturated NaHCO3 solution, filtered, and dried to offer the title compound
as a white solid (17.46 g, 85% yield), mp 88–90 °C. 1H NMR (CDCl3): δ 1.35 (t, 3H, J = 7.2 Hz, CH3), 2.53 (s, 3H, CH3), 2.99 (s,
3H, CH3), 4.28 (q, 2H, J = 7.2 Hz, CH2). 13C NMR (CDCl3): δ 14.78, 17.67,
31.27, 60.28, 107.47, 159.79, 162.37, 171.08. HR-MS (ESI+): m/z [M + H]+ calcd
for C8H13 N2O2S, 201.0698,
found 201.0463.
To a solution of ethyl 4-methyl-2-(methylamino)thiazole-5-carboxylate
(10.0 g, 50.0 mmol) in DCM were added 4-dimethylaminopyridine (DMAP)
(1.0 g) and di-tert-butyl dicarbonate (12.0 g, 55.0
mmol), and the reaction was continued for 8 h at room temperature.
After completion of the reaction, the mixture was washed with 5% aqueous
HCl, followed by saturated NaHCO3 solution, brine, dried
over MgSO4, and filtered. The organic solution was concentrated
to dryness, and the title compound was obtained via recrystallization
from hexane as a white solid (14 g, 93%), mp 148–150 °C. 1H NMR (CDCl3): δ 1.34 (t, 3H, J = 7.2 Hz, CH3), 1.60 (s, 9H, 3 × CH3),
2.64 (s, 3H, CH3), 3.55 (s, 3H, CH3), 4.29 (q,
2H, J = 7.2 Hz, CH2). 13C NMR
(CDCl3): δ 14.35, 17.37, 28.14, 33.94, 60.54, 83.85,
116.19, 153.13, 156.56, 162.68, 163.10. HR-MS (ESI+): m/z [M + H]+ C13H21N2O4S, 301.1222, found 301.1312.
To a solution of 2 (6.0 g, 20.0 mmol)
in 6 mL of anhydrous THF was added 1.50 mL of acetonitrile (1.3 mmol).
The mixture was cooled at −78 °C, and LDA was added dropwise
over 10 min. The reaction was continued for 2 h. After completion
of the reaction, 10 mL of H2O was added and the mixture
was acidified with dilute HCl solution and extracted withCHCl3 (3 × 50 mL). The combined organic phase was washed withbrine, dried over MgSO4, and concentrated to dryness. The
mixture was purified by using PE/EtOAc as elutant to afford the title
compound as a white solid (4.25 g, 72%), mp 119–121 °C. 1H NMR (CDCl3): δ 1.61 (s, 9H, 3 × CH3), 2.68 (s, 3H, CH3), 3.59 (s, 3H, CH3), 3.86 (s, 2H, CH2). 13C NMR (CDCl3): δ 18.71, 28.09, 32.23, 34.26, 84.76, 113.76, 122.37, 153.11,
159.36, 163.14, 179.25. HR-MS (ESI+): m/z [M + H]+ calcd for C13H18 N3O3S, 296.1069, found 296.1130.
To a well-stirred solution
of 3-(dimethylamino)-1-(4-methyl-2-(methylamino)thiazol-5-yl)prop-2-en-1-one
(6)[35] (5.0 mmol) in MeOH under
ice bath was added SelecFluor (7.5 mmol), and the mixture was stirred
for 1 h. After completion of the reaction, the mixture was concentrated
and purified by column chromatography using EtoAc/MeOH to yield the
title compound. Yellow solid (30%). 1H NMR (DMSO-d6): δ 2.40 (s, 3H, CH3), 2.83
(d, 3H, J = 4.8 Hz, CH3), 3.04 (apparent
s, 6H, 2 × CH3), 6.88 (d, 1H, J =
30.4 Hz, CH), 8.04 (d, 1H, J = 4.4 Hz, NH). HR-MS
(ESI+): m/z [M + H]+ calcd for C10H15FN3OS, 244.0920,
found 244.0849.
To an ice-cooled solution
of 6 (2.0 mmol) in 50 mL of methanol was added N-chlorosuccinmide (20 mmol) over 10 min dropwise. After
being stirred at room temperature for 30 min, the mixture was concentrated
and purified by column chromatography using EtoAc to yield the title
compound as a light yellow solid (43%), mp 141–143 °C. 1H NMR (CDCl3): δ 2.38 (s, 3H, CH3), 2.99 (s, 3H, CH3), 3.26 (s, 6H, 2 × CH3), 6.44 (s, 1H, NH), 7.42 (s, 1H, CH). 13C NMR (CDCl3): δ 18.28, 31.98, 43.33, 101.88, 117.86, 149.13, 153.84,
171.91, 180.77. HR-MS (ESI+): m/z [M + H]+ calcd for C10H15ClN3OS, 260.0624, found 260.0541.
N,4-Dimethylthiazol-2-amine
was obtained from 1-methylthiourea and 1-chloropropan-2-one using
the method for preparing 1 as off-white solid (88%),
mp 68–70 °C. 1H NMR (CDCl3): δ
2.96 (d, 3H, J = 1.2 Hz, CH3), 2.96 (s,
3H, CH3), 6.06 (q, 1H, J = 1.2 Hz, CH),
6.15 (brs, 1H, NH). 13C NMR (CDCl3): δ
17.36, 31.95, 100.06, 148.66, 171.61. HR-MS (ESI+): m/z [M + H]+ calcd for C5H9N2OS, 129.0486, found 129.0372. The
title compound was obtained from N,4-dimethylthiazol-2-amine
and di-tert-butyl dicarbonate by the method for preparing 2 as light yellow liquid (55%). 1H NMR (CDCl3): δ 1.59 (s, 9H, 2 × CH3), 2.35 (d,
3H, J = 1.2 Hz, CH3), 3.55 (s, 3H, CH3), 6.48 (q, 1H, J = 0.8 Hz, CH). 13C NMR (CDCl3): δ 17.41, 28.20, 34.32, 82.91, 108.39,
147.11, 153.13, 161.20. HR-MS (ESI+): m/z [M – tert-butyl + H]+ calcd for C6H9N2O2S, 173.0385, found 173.0224.
General Procedure for Preparation of 9 (R′
= Me, Et, or Pr)
To a solution of 8 (10.0 mmol)
in 15 mL of anhydrous THF cooling at −78 °C was added
25.0 mmol of LDA dropwise. After the mixture was stirred for 30 min
the corresponding aldehyde (12.0 mmol) was added and the reaction
was continued for a further 1 h. After completion of the reaction,
10 mL of water was added and the mixture was washed with 2 M aqueous
HCl solution. After removal of THF, the mixture was extracted withCHCl3 (3 × 30 mL) and the combined organic layers
were washed withbrine, dried over MgSO4, filtered, and
concentrated to dryness. The title compound was obtained by column
chromatography using EtOAc/PE as eluant.
General Procedure for Preparation of 10 (R′
= Me, Et, or Pr)
A solution of corresponding 9 in CHCl3 (1.5 mmol/mL) was treated withMnO2 (5.0 equiv), and the mixture was refluxed for 3 h. Upon completion
of the reaction, the mixture was filtered through Celite, and the
filtrate was concentrated to dryness.
Yellow solid (65%), mp
82–84 °C. 1H NMR (CDCl3): δ
0.92 (t, 3H, J = 7.2 Hz, CH3), 1.30–1.43
(m, 2H, CH2), 1.59 (s, 9H, 3 × CH3), 1.62–1.72
(m, 2H, CH2), 2.64 (s, 3H, CH3), 2.75 (t, 2H, J = 7.2 Hz, CH2), 3.55 (s, 3H, CH3). 13C NMR (CDCl3): δ 13.88, 18.38, 22.37,
26.54, 28.14, 34.07, 42.73, 84.01, 125.16, 153.16, 155.50, 161.67,
193.78. HR-MS (ESI+): m/z [M + H]+ C15H25N2O3S, 313.1586, found 313.1620.Preparations of 4 (R1 = Boc, R′ = Me, Et or Pr) were done by heating 10 in DMF–DMA using the method described previously[35] or by heating in a Discovery microwave at 140
°C for 45 min. The mixture was concentrated and used for the
pyrimidine formation reaction without further purification.
General Procedure for Preparation of Ic and 12a–u
A mixture of the appropriate
3-(dimethylamino)-1-(4-methyl-2-(methylamino)thiazol-5-yl)prop-2-en-1-oneor 4 and 1-phenylguanidine (11)[32,35] (2 equiv mmol) in 2-methoxyethanol (0.2 mL/mmol) was heated in a
microwave at 100–140 °C for 20–45 min. When the
mixture was cooled, the residue was purified by flash chromatography
using appropriate mixtures of EtoAc/PEorEtoAc/MeOH as the eluant.
The products were further purified by crystallization from EtOAc–MeOH
mixtures.
To an ice-cooled solution of 1-(4-methyl-2-(methylamino)thiazol-5-yl)ethanone
(3.4 g, 20.0 mmol) in 50 mL of CHCl3 was added N-bromosuccinimide (3.5 g, 20.0 mmol). After being stirred
at room temperature for 3 h, the mixture was washed with saturated
aqueous NaHCO3, and the organic layer was dried over MgSO4, filtered, and concentrated. The precipitates were collected
and washed withMeOH to afford the title compound as a white solid
(2.76 g, 56%), mp 149 °C (dec). 1H NMR (DMSO-d6): δ 2.41 (s, 3H, CH3), 2.87
(d, 3H, J = 4.8 Hz, CH3), 4.74 (s, 2H,
CH2), 8.57 (d, H, J = 4.4 Hz, NH). HR-MS
(ESI+): m/z [M + H]+ calcd for C7H10BrN2OS, 250.9677,
found 250.9607. tert-Butyl 5-acetyl-4-(bromomethyl)thiazol-2-yl(methyl)carbamate
was obtained by reacting 13 and di-tert-butyl dicarbonate as white solid (98%), mp 84–86 °C. 1H NMR (CDCl3): δ 1.61 (s, 9H, 3 × CH3), 2.54 (s, 3H, CH3), 3.60 (s, 3H, CH3), 5.34 (s, 2H, CH2). HR-MS (ESI+): m/z [M + H]+ calcd for C12H18Br N2O3S, 351.0210, found
351.0338.
A solution of tert-butyl
5-acetyl-4-(bromomethyl)thiazol-2-yl(methyl)carbamate (8.45 g, 24.0
mmol) in DMF was treated withmethyl 2,2-difluoro-2-(fluorosulfonyl)acetate
(5.7 mL, 45.0 mmol) and CuI (2.9 g, 15.0 mmol). The mixture was heated
at 80 °C for 12 h. After completion of the reaction, the mixture
was purified by chromatography using EtoAc/PE to afford the title
compound as a white solid (4.13g, 51%), mp 93–95 °C. 1H NMR (CDCl3): δ 1.61 (s, 9H, 3 × CH3), 2.53 (s, 3H, CH3), 3.59 (s, 3H, CH3), 4.03 (q, 2H, J = 10.4 Hz, CH2). 13C NMR (CDCl3): δ 28.10, 30.84, 34.12, 35.00
(q, J = 30 Hz), 84.36, 124.15 (q, J = 276 Hz), 127.96, 146.63 (q, J = 3 Hz), 153.26,
161.96. 190.21. HRMS (ESI+): m/z [M + H]+ calcd for C13H18F3N2O3S, 339.0990, found 339.1058.
To a solution of 1,1,1-trifluoropentane-2,4-dione
(1.0 g, 6.4 mmol) in 15 mL of acetonitrile was added hydroxy(tosyloxy)iodobenzene
(3.0 g, 7.7 mmol), and the mixture was refluxed for 1 h. When the
mixture was cooled, 1-methylthiourea (0.69 g, 7.7 mmol) was added
and the mixture was heated under reflux for 4 h. The mixture was concentrated
and purified using PE/EtOAc as eluant to yield the title compound
as a white solid (2.66 g, 53%), mp 133–135 °C. 1H NMR (DMSO-d6) δ: 2.42 (d, 3H, J = 0.8 Hz, CH3), 2.88 (d, 3H, J = 4.8 Hz, CH3), 8.86 (bs, 1H, NH). 13C NMR
(DMSO-d6): δ 29.87 (d, J = 3 Hz), 31.59, 120.44 (q, J = 271 Hz), 125.57,
141.93 (q, J = 36 Hz), 171.05, 187.48. HR-MS (ESI+): m/z [M + H]+ calcd for C7H8 F3N2OS,
225.0309, found 225.0309.
A mixture of N-methylthiourea (9.0 g, 100 mmol)
and DMF–DMA (12 mL, 120 mmol) in 50 mL of CHCl3 was
refluxed overnight. The mixture was concentrated and the resulting
precipitate was collected by filtration to afford 18 as
white solid (14.3 g, 98%), mp 109–110 °C. 1H NMR (CDCl3): δ 3.02 (d, 0.9H, J = 5.2 Hz, CH3), 3.04 (s, 2.1H, CH3), 3.13
(s, 0.9H, CH3), 3.15 (s, 2.1H, CH3), 3.19 (s,
0.9H, CH3), 3.21 (d, 2.1H, J = 5.2 Hz,
CH3), 6.88 (brs, 1H, NH), 8.85 (s, 0.3H, CH), 8.88 (s,
0.7H, CH). HR-MS (ESI+): m/z [M + H]+ calcd for C5H12N3S, 146.0752; found 146.0638.
1-(2-(Methylamino)thiazol-5-yl)ethanone (19)
A mixture of N,N-dimethyl-N′-(methylcarbamothioyl)formimidamide (3.62 g, 25.0
mmol) and chloroacetone chloride (2 mL, 25.0 mmol) in 50 mL of acetonitrile
was refluxed for 4 h. After completion of the reaction, the mixture
was concentrated, neutralized by saturated NaHCO3 solution,
and dried over air to yield the title compound as a white solid (3.10
g, 79%), mp 163–164 °C. 1H NMR (DMSO-d6): δ 2.35 (s, 3H, CH3), 2.88
(s, 3H, CH3), 8.00 (s, H, thiazol-H),
8.54 (br s, 1H, NH). 13C NMR (DMSO-d6): δ 26.03, 31.55, 127.66, 150.04, 175.29, 188.75. HR-MS
(ESI+): m/z [M + H]+ calcd for C6H9N2OS, 157.0436;
found 157.0269.
A solution of 1-cyclopropylethanone (8.41
g, 100 mmol) in 30 mL of methanol was cooled in an ice bath, and bromine
(5.15 mL, 100 mmol) was added dropwise. The mixture was stirred for
1 h before being warmed to room temperature. After the mixture was
stirred for a further 3 h, 50 mL of water was added. The mixture was
extracted by diethyl ether (3 × 100 mL) and the combined organic
phase was washed withbrine, dried over MgSO4, and concentrated
to yield 22 as a colorless oil (12.49 g, 77%). 1H NMR (CDCl3): δ 0.98–1.04 (m, 2H, CH2), 1.08–1.14 (m, 2H, CH2), 2.14–2.23
(m, 1H, CH), 4.02 (s, 2H, CH2). A solution of 22 (12.49 g, 76.0 mmol) in 30 mL of methanol was treated with1-methylthiourea
(6.84 g, 76.0 mmol), and an amount of 2 mL of pyridine was added.
After being stirred at room temperature overnight, the mixture was
concentrated and basified with saturated NaHCO3 solution.
The precipitate was filtered and dried over air. 4-Cyclopropyl-N-methylthiazol-2-amine was obtained as a white solid after
recrystallization from PE/EtOAc (5.70 g, 49%). 1H NMR (CDCl3): δ 0.76–0.81 (m, 2H, CH2), 0.81–0.87
(m, 2H, CH2), 1.81–1.90 (m, 1H, CH), 2.95 (s, 3H,
CH3), 5.46 (br s, 1H, NH), 7.07 (s, 1H, thiazol-H). HR-MS
(ESI+): m/z [M + H]+ calcd for C7H11N2S, 155.0643;
found 155.0452. The latter was treated withdi-tert-butyl dicarbonate to afford the title compound 23 as
a brown liquid (55%). 1H NMR (CDCl3) δ:
0.81–0.90 (m, 4H, 2 × CH2), 1.58 (s, 9H, 3
× CH3), 1.89–1.99 (m, 1H, CH), 3.52 (s, 3H,
CH3), 6.45 (s, 1H, thiazol-H). HR-MS (ESI+): m/z [M + H]+ calcd for C12H19N2O2S, 255.1167; found
255.1184.
A solution of 23 (3.81g, 15.0
mmol) in 20 mL of anhydrous THF at −78 °C was treated
with 25.0 mmol of LDA. After the mixture was stirred for 30 min, acetaldehyde
(20.0 mmol, 1.12 mL) was added and the reaction was continued for
2 h. After completion of the reaction, 20 mL of water was added and
the mixture was treated with 2 M aqueous HCl solution. After concentration,
the mixture was extracted withCHCl3 (3 × 50 mL).
The combined organic phase was washed withbrine, dried over MgSO4, filtered, and concentrated to dryness. The resulting mixture
was dissolved in 30 mL of CHCl3 and then treated withMnO2 (10 equiv). The mixture was heated under reflux for 4 h.
The reaction mixture was purified using PE/EtOAc as elutant to afford
the title compound as a yellow solid (3.51 g, 79%), mp 89–90
°C. 1H NMR (CDCl3): δ 0.98–1.08
(m, 2H, CH2), 1.09–1.18 (m, 2H, CH2),
1.60 (s, 9H, 3 × CH3), 2.51 (s, 3H, CH3), 2.98–3.10 (m, 1H, CH), 3.49 (s, 3H, CH3). 13C NMR (DMSO-d6): δ 10.12,
12.10, 28.14, 30.92, 33.86, 83.95, 124.81, 153.15, 161.43, 162.18,
190.96. HR-MS (ESI+): m/z [M + H]+ calcd for C14H21N2O3S, 297.1273, found 297.1296.
To a solution of N-methyl-4-phenylthiazol-2-amine
(1.90 g, 10 mmol) and anhydrous AlCl3 (6.7 g, 50 mmol)
in 100 mL of CH2Cl2 under nitrogen gas was added
acetyl chloride (1.42 mL, 20 mmol) over 15 min. The reaction was continued
for 2 h. The mixture was cooled in an ice bath and quenched using
MeOH. After being concentrated, the mixture was purified by column
chromatography using EtOAc/PE as eluant to affort the title compound
as a white solid (1.55g, 67%), mp 229–231 °C. 1H NMR (DMSO-d6): δ 1.90 (s, 1H,
CH3), 2.88 (d, 1H, J = 4.8 Hz, CH3), 7.42–7.50
(m, 3H, 3 × Ph-H), 7.50–7.56 (m, 2H, 2 × Ph-H), 8.52
(br q, 1H, J = 4.8 Hz, NH). 13C NMR (DMSO-d6): δ 28.64, 31.40, 124.06, 128.53, 129.49,
129.80, 136.19, 159.09, 171.46, 189.10. HR-MS: m/z [M + H]+ calcd for C12H13N2OS, 233.0749, found, 233.0658. N-Methyl-4-phenylthiazol-2-amine
was obtained from benzoyl chloride and 1-methylthiourea as a white
solid (100%), mp 143–144 °C. 1H NMR (DMSO-d6): δ 2.99 (s, 3H, CH3), 7.18
(s, 1H, thiazol-H), 7.36–7.43 (m, 1H, Ph-H), 7.43–7.50
(m, 2H, 2 × Ph-H), 7.75–7.81 (m, 2H, 2 × Ph-H), 8.73
(br s, 1H, NH). HRMS: m/z [M + H]+ calcd for C10H11N2S, 191.0643;
found 191.0512.
CDK9330 (residues 1–330)/cyclin
T1 (residues 1–259, Q77R, E96G, F241L) compounds were expressed,
purified, and crystallized as described previously.[51] Crystals were grown by vapor diffusion against a reservoir
containing 14% PEG1000, 100 mM sodium potassium phosphate, pH 6.2,
500 mM NaCl, 4 mM TCEP. A crystal was soaked in mother liquor containing
also 1 mM 12u and 15% glycerol for 45 min before cryocooling
in liquid nitrogen.CDK2/cyclin A was expressed and purified
as described previously.[57] Purified protein
was incubated with12u, filtered, and cocrystallized
in 1.25 M ammonium sulfate, 0.5 M potassium chloride, 100 mM Hepes,
pH 7.0, 5 mM DTT at 4 °C. Crystals were cryoprotected and frozen
in 7 M sodium formate in the presence of 1 mM 12u.Diffraction data for the CDK9/cyclin T1/12u and CDK2/cyclin
A/12u were collected from single crystals at Diamond
Light Source beamline I03. Diffraction data for CDK9/cyclin T/12u were processed with XDS[58] and
SCALA (CCP4).[59] PHENIX.refine[60] was used for rigid body refinement with a model
derived from 3BLH as the initial model. REFMAC[61] was used
for subsequent TLS and restrained refinement. Jelly body restraints
to an external model (3BLH) were used during refinement. CDK2/cyclin A data were
processed using XDS[58] and SCALA.[59] Molecular replacement was performed by the program
PHASER[62] using a search model derived from
PDB entry 3DDQ. Ligand restraints were defined using PHENIX, and structures were
refined and rebuilt using PHENIX.refine and COOT.[63]
Kinase Assay
Inhibition of CDKs and other kinases was
measured by radiometric assay Millipore’s KinaseProfiler according
to the protocols detailed at http://www.millipore.com/drugdiscovery/dd3/, where ATP concentration for each specific kinase assay was set
within 15 μM of the apparent Km for
ATP where determined. Half-maximal inhibition (IC50) values
were calculated from 10-point dose–response curves, and apparent
inhibition constants (Ki) were calculated
from the IC50 values and Km (ATP) values for the kinases in question as described.[35] The assay details can also be found in the Supporting Information.
Cell Culture
All cancer cell lines were obtained from
the cell bank at the Centre for Biomolecular Sciences, University
of Nottingham, U.K. The HMEC-1 cell line was purchased (ECACC), and
cells were cultured in essential mediumwith 10% fetal bovine serum
(FBS), 7.5% sodium bicarbonate, 1% 0.1 mM nonessential amino acids,
1% 1 M HEPES, 1% 200 mM l-glutamine, and 1% penicillin. Other
cell lines were maintained in RPMI-1640with 10% FBS.
Proliferation Assays
MTT (3-(4,5-dimethylthiazol-2-yl)-2,5-diphenyltetrazolium
bromide, Sigma) assays were performed as reported previously.[35] Compound concentrations required to inhibit
50% of cell growth (GI50) were calculated using nonlinear
regression analysis.
Caspase-3/7 Assay
Activity of caspase 3/7 was measured
using the Apo-ONE homogeneous caspase-3/7 kit (Promega G7790).[18]
Cell Cycle Analysis and Detection of Apoptosis
Cells
(4× 105) were cultured for 48 h in medium alone orwith varying concentrations of inhibitor. Cell cycle status was analyzed
using a Beckman Coulter EPICS-XL MCL flow cytometer, and data were
analyzed using EXPO32 software. Apoptosis was also confirmed using
FITCannexin V/PI (propidium idodide) staining after cells were cultured
in medium only orwith varying concentrations of inhibitors according
to the protocols (BD Bioscience). The annexin V/PI-positive apoptotic
cells were enumerated using flow cytometry. The percentage of cells
undergoing apoptosis was defined as the sum of early apoptosis (annexin
V-positive cells) and late apoptosis (annexin V-positive and PI-positive
cells). The pan-caspase inhibitor Z-Val-Ala-Asp-(OMe)-CH2F (Z-VAD-fmk, Sigma) was dissolved in DMSO and used at 25 μM.
Detection of Apoptosis in Primary CLL Cells
Freshly
isolated primary CLL cells and normal B- and T-cells were cultured
in RPMIwith 10% fetal calf serum and l-glutamine, penicillin,
and streptomycin. Cells were maintained at 37 °C in an atmosphere
containing 95% air and 5% CO2 (v/v). CLL cells (106/mL) were treated with inhibitor for 48 h. Subsequently, cells
were labeled withCD19-APC (Caltag) and then resuspended in 200 μL
of binding buffer containing 4 μL of annexin V-FITC (Bender
Medsystems, Vienna, Austria). Apoptosis was quantified in the CD19+ CLL cells, CD19+ normal B-cells, and CD3+ normal T-cells using an Accuri C6 flow cytometer and FlowJo software
(TreeStar). LD50 values were calculated from line-of-best-fit
analysis of the sigmoidal dose–response curves.
Western Blots
Western blotting was performed as described.[20] Antibodies used were as follows: total RNAP-II
(8WG16), phosphorylated RNAP-II Ser-2 (Covance), Bcl-2 (Dako, Denmark
A/S), MDM2 and β-actin (Sigma-Aldrich), Mcl-1, PARP (Cell Signaling
Technology). Both anti-mouse and anti-rabbit immunoglobulin G (IgG)
horseradish peroxidase-conjugated antibodies were obtained from Dako.
Statistical Analysis
All experiments were performed
in triplicate and repeated at least twice, representative experiments
being selected for figures. Statistical significance of differences
for experiments was determined using one-way analysis of variance
(ANOVA), with a minimal level of significance at p < 0.01.
Authors: Paul T Nyffeler; Sergio Gonzalez Durón; Michael D Burkart; Stéphane P Vincent; Chi-Huey Wong Journal: Angew Chem Int Ed Engl Date: 2004-12-27 Impact factor: 15.336
Authors: David E MacCallum; Jean Melville; Sheelagh Frame; Kathryn Watt; Sian Anderson; Athos Gianella-Borradori; David P Lane; Simon R Green Journal: Cancer Res Date: 2005-06-15 Impact factor: 12.701
Authors: Shudong Wang; Gary Griffiths; Carol A Midgley; Anna L Barnett; Michael Cooper; Joanna Grabarek; Laura Ingram; Wayne Jackson; George Kontopidis; Steven J McClue; Campbell McInnes; Janice McLachlan; Christopher Meades; Mokdad Mezna; Iain Stuart; Mark P Thomas; Daniella I Zheleva; David P Lane; Robert C Jackson; David M Glover; David G Blake; Peter M Fischer Journal: Chem Biol Date: 2010-10-29
Authors: Airlie J McCoy; Ralf W Grosse-Kunstleve; Paul D Adams; Martyn D Winn; Laurent C Storoni; Randy J Read Journal: J Appl Crystallogr Date: 2007-07-13 Impact factor: 3.304
Authors: Muhammed H Rahaman; Yingyi Yu; Longjin Zhong; Julian Adams; Frankie Lam; Peng Li; Ben Noll; Robert Milne; Jun Peng; Shudong Wang Journal: Invest New Drugs Date: 2018-09-08 Impact factor: 3.850
Authors: Alison J Hole; Sonja Baumli; Hao Shao; Shenhua Shi; Shiliang Huang; Chris Pepper; Peter M Fischer; Shudong Wang; Jane A Endicott; Martin E Noble Journal: J Med Chem Date: 2013-01-29 Impact factor: 7.446