3-Aryl-4-(arylhydrazono)-1H-pyrazol-5-ones: Highly ligand efficient and potent inhibitors of GSK3β

Article abstract of DOI:10.1016/j.bmcl.2010.01.072

A series of 3-aryl-4-(arylhydrazono)-1H-pyrazol-5-one inhibitors of GSK3β was developed from a low molecular weight, highly ligand efficient screening hit 1. Hit-to-lead optimization led to a number of highly potent inhibitors, while maintaining the high ligand efficiency of the screening hit.

Full text of DOI:10.1016/j.bmcl.2010.01.072

Bioorganic & Medicinal Chemistry Letters 20 (2010) 1661–1664
Contents lists available at ScienceDirect
Bioorganic & Medicinal Chemistry Letters
journal homepage: www.elsevier.com/locate/bmcl
3-Aryl-4-(arylhydrazono)-1H-pyrazol-5-ones: Highly ligand efficient
and potent inhibitors of GSK3b
*
Michael Arnost, Al Pierce, Ernst ter Haar, David Lauffer, Jaren Madden, Kirk Tanner, Jeremy Green
Vertex Pharmaceuticals Incorporated, 130 Waverly Street, Cambridge, MA 02139, USA
a r t i c l e i n f o
a b s t r a c t
Article history:
A series of 3-aryl-4-(arylhydrazono)-1H-pyrazol-5-one inhibitors of GSK3b was developed from a low
molecular weight, highly ligand efficient screening hit 1. Hit-to-lead optimization led to a number of
highly potent inhibitors, while maintaining the high ligand efficiency of the screening hit.
Ó 2010 Elsevier Ltd. All rights reserved.
Received 16 December 2009
Revised 7 January 2010
Accepted 11 January 2010
Available online 21 January 2010
Keywords:
GSK3b
Kinase inhibitors
Pyrazolones
Ligand efficiency
Binding efficiency
Hit-to-lead
Lead-like
Glycogen synthase kinase (GSK) is a Ser/Thr kinase, originally
identified as the enzyme responsible for the inactivating phosphor-
posed the concepts of ligand,15 or binding16 efficiency as measures
of the average value of the contributions of each atom within the
binding molecule to the overall energy of binding to the target.
We have chosen to use the binding efficiency index (BEI), which
is simply calculated from activity and molecular weight,16 to pro-
vide guidance in optimizing potency. Thus, BEI = (pKi/MW), where
molecular weight is in kiloDaltons. For compound 1, BEI = 26.2,
which was the second highest value that we recorded in this screen
and amongst the highest we have observed in any kinase screen.
With this starting point we believed that highly potent GSK3b
inhibitors with properties suitable for lead optimization could be
identified.17
In our initial binding model of compound 1 we hypothesized
that the 4-hydrazino group was most likely forming an internal
hydrogen bond with the pyrazolone carbonyl forming a pseudo-
bicyclic 6,5-ring system, with a single substituent, the 2-chloro-
phenyl group. In this model, the pyrazolone heteroatoms would
likely form two hydrogen bonds with the kinase hinge region, sim-
ilar to those formed between GSK3 and a number of inhibitors,
such as indirubin and derivatives (e.g., BIO),18 maleimides (e.g.,
SB-415286),19 and staurosporine20 ( Fig. 1). In this model, the
2-chlorophenyl ring of 1 occupies space that is also, in part, occu-
pied by these inhibitors, while the pyrazolone 3-position is open
and presents a suitable vector for additional interaction with the
enzyme active site.
ylation of glycogen synthase.1 Two isoforms, GSK3
a (51 kDa) and
GSK3b (47 kDa) are known and there is high homology in their
kinase domains. The functional differences between the isoforms
are currently unclear, but they are differently distributed2 and
not entirely functionally redundant, as GSK3b knockout in mice
was shown to be embryonically lethal.3 Subsequent research has
identified GSK3 as a key kinase in a variety of signalling pathways
and networks, and a potential therapeutic target for a number of
diseases,4,5 including bipolar disorder,6,7 schizophrenia,7 Alzhei-
mer’s disease,8–10 cardiac disease,11 and diabetes.12,13 Much of
the complexity of the GSK3 signalling network remains to be
unravelled and small molecule intervention has begun to shed
light on the overriding pharmacological action of GSK3 modula-
tion. A number of GSK3 inhibitors have now entered clinical tri-
als14 the results of which will help define the ultimate practical
utility of GSK3 inhibition.
In the course of screening our compound collection for inhibi-
tors of GSK3b, we identified compound 1, (Z)-4-(2-(2-chloro-
phenyl)hydrazono)-1H-pyrazol-5(4H)-one, as
a low molecular
weight inhibitor (MW = 222.6; Ki = 1.49 M: Figure 1). This com-
l
pound was notable amongst the screening hits in displaying an
unusually high ligand efficiency. Various researchers have pro-
To explore and develop structure–activity relationships of the
pyrazolones, a limited series of analogs was prepared by the gen-
eral method shown in Scheme 1.
* Corresponding author. Tel.: +1 6174446315.
E-mail address: jeremy_green@vrtx.com (J. Green).
0960-894X/$ - see front matter Ó 2010 Elsevier Ltd. All rights reserved.
doi:10.1016/j.bmcl.2010.01.072

1662
M. Arnost et al. / Bioorg. Med. Chem. Lett. 20 (2010) 1661–1664
Br
O
N
H
3
N
HN
HN
HN
N
O
O
O
NO2
N
N
N
OH
O
NH
Cl
O
H
O
N
HN
N
H
Cl
H
HO
Staurosporine
1
BIO
SB-415286
Figure 1. Compound 1 and structurally related inhibitors of GSK3b.
R1
O
(a)
R2
O
O
N
NH
R1
R2
OMe
NH2
(c)
NH
N
R1
O
(b)
N2+Cl-
R2
N
NH
Scheme 1. Synthesis of pyrazolones. Reagents and conditions: (a) NH2NH2, EtOH; (b) NaNO2, HCl, H2O, 0 °C; (c) NaOAc, EtOH, H2O.
In Scheme 1, ketoesters (prepared as needed from methylke-
GSK3b complexed with staurosporine (Pdb code 1Q3D) gives two
possible binding modes shown in Figure 2. The ligand conforma-
tion shown in Figure 2A places the chlorine unfavourably close to
the carbonyl of Val135. In the alternate binding mode shown in
Figure 2B, the o-chlorophenyl ring is rotated approximately 180°,
so that there is no steric clash with the protein, but the intramolec-
ular contact between the chlorine atom and the hydrazone is
disfavoured. To address this liability and further explore the SAR
of the series, we prepared a more diverse set of pyrazolone deriv-
atives, with variation at R1 and R2. Data is shown in Table 2.
It is apparent from the data contained in Table 2 that this
scaffold is particularly suited to very potent inhibition of the
target GSK3b. Numerous analogs possess low nanomolar and
sub-nanomolar inhibition constants. With R1 fixed as phenyl,
3- and 4-substitution at R2 afforded compounds that were both
potent and efficient inhibitors (e.g., 12, 14, and 15). Fixing R2
as 4-methoxyphenyl (16–23) identified R1 as 3-pyridyl (17), as
well as 3-methoxyphenyl (19), 4-methoxyphenyl (20) and 3,4-
dimethoxyphenyl (21) as potent and efficient inhibitors. The
additional methoxy group of the 3,4,5-trimethoxyphenyl-substi-
tuted compound 22, though quite potent, was notably less ligand
efficient. The R1 combination of 3-pyridyl with 4-methoxy in
compound 23 was potent (2 nM) and more efficient than 1. Fur-
ther optimization of this class of GSK3b inhibitors is exemplified
by compounds 24 to 62. As noted above, it does not necessarily
follow that the most potent inhibitors are the most efficient
inhibitors. Compounds 33, 43, 52, 53, and 62, are all sub-
10 nM GSK3b inhibitors, yet are significantly less efficient (BEI
<22) than the screening hit, 1. These compounds all bear 4-ami-
no derivatives that would be expected to extend into solvated
space. While the morpholino group might improve the solubility
of these molecules, there is a significant efficiency cost of intro-
ducing a 100 Da substituent that contributes little, or anything,
to the binding energy. Meanwhile, compounds 37, 39, and 54
are not only potent inhibitors (GSK3b Ki <5 nM), but are signifi-
cantly improved in ligand efficiency (BEI >28) relative to 1.
tones by the method of Jung et al.21) were converted to pyrazo-
lones in good yield by treatment with hydrazine in ethanol.
Subsequent treatment with freshly prepared aryl diazonium salts
afforded the desired products, which were purified by HPLC.22 In
an initial evaluation of the screening hit we prepared a number
of analogs from available ketoesters and the 2-chlorophenyl diazo-
nium salt. Results of this series are summarized in Table 1.
Results from this survey showed that, in general, the pyrazolone
3-substituent had little effect on the inhibitory potency of the mol-
ecules, with the exception of compound 9, which was 30-fold more
potent than the screening hit 1. Interestingly, this marked
improvement in potency for compound 9 does not translate into
a binding efficiency comparable to compound 1. In fact, compound
9 is comparable in binding efficiency to the parent phenyl deriva-
tive, 4. A careful analysis of the binding modes of these compounds
suggested that their binding affinity might be limited by the pres-
ence of the ortho-chloro substituent. Glide docking24 and subse-
quent minimization of compound 1 into the crystal structure of
Table 1
3-Substituted 4-(2-(2-chlorophenyl)hydrazono)-1H-pyrazol-5(4H)-one inhibitors of
GSK3b
Cl
R1
N
N
N
H
N
O
H
Compd
R1
MW
GSK3b, Ki, nM23
BEI
1
2
3
4
5
6
7
8
9
H
Me
iPr
Ph
222.6
236.7
264.7
298.7
299.7
299.7
299.7
328.8
358.8
1490
460
>3700
720
650
200
250
850
44
26.2
26.8
<20.5
20.6
20.6
22.4
22.0
18.5
20.5
2-Pyridyl
3-Pyridyl
4-Pyridyl
3-MeO-Ph
3,4-(MeO)2-Ph
Selectivity for these compounds against
a panel of kinases
(including kinases, such as CDK2, known to bind GSK3 inhibi-
tors25 and a variety of other Ser/Thr, Tyr and lipid kinases) is

M. Arnost et al. / Bioorg. Med. Chem. Lett. 20 (2010) 1661–1664
1663
Figure 2. Binding model of compound 1 docked in GSK3b (PDB ID: 1q3d).20
Table 2 (continued)
Table 2
N-, 3-Substituted 4-(2-hydrazono)-1H-pyrazol-5(4H)-one inhibitors of GSK3b
Compd R1
R2
MW
GSK3b, Ki,
BEI
nM12
R1
R2
37
38
39
40
41
42
43
4-MeO-Ph
3-MeO-Ph
4-CN-Ph
324.3 0.8
319.3 <2
295.3
295.3 14
337.4 110
365.3 16
28.0
>27.2
29.5
26.6
20.6
21.3
21.3
N
N
4-MeO-Ph
4-MeO-Ph
4-MeO-Ph
4-MeO-Ph
4-MeO-Ph
4-MeO-Ph
N
H
3-Pyridyl
4-Pyridyl
4-NMe2-Ph
4-NEt2-Ph
4-Morpholino-
Ph
2
N
O
H
Compd R1
R2
MW
GSK3b, Ki,
BEI
nM12
379.4
8
10
11
12
13
14
15
Ph
Ph
Ph
Ph
Ph
Ph
Ph
3-Cl-Ph
4-Cl-Ph
2-MeO-Ph
3-MeO-Ph
4-MeO-Ph
264.3 99
298.7 120
298.7 23
294.3 >4000
294.3 35
294.3 20
26.5
23.2
25.6
<18.3
25.3
26.2
44
45
46
47
48
49
50
51
52
53
3,4-(MeO)2-Ph
3,4-(MeO)2-Ph
3,4-(MeO)2-Ph
3,4-(MeO)2-Ph
3,4-(MeO)2-Ph
3,4-(MeO)2-Ph
3,4-(MeO)2-Ph
3,4-(MeO)2-Ph
3,4-(MeO)2-Ph
3,4-(MeO)2-Ph
Ph
3-Cl-Ph
4-Cl-Ph
324.3 <3.5
358.8 0.4
358.8 <2
354.4 0.4
349.3 <2
325.3 <3.5
325.3 <3.5
367.4 87
>26.1
26.2
>24.2
26.5
>24.9
>26.0
>26.0
19.2
3-MeO-Ph
4-CN-Ph
3-Pyridyl
4-Pyridyl
4-NMe2-Ph
4-NEt2-Ph
4-Morpholino-
Ph
16
17
18
19
20
21
22
2-Pyridyl
3-Pyridyl
4-Pyridyl
3-MeO-Ph
4-MeO-Ph
3,4-(MeO)2-Ph
3,4,5-(MeO)3-
Ph
4-MeO-Ph
4-MeO-Ph
4-MeO-Ph
4-MeO-Ph
4-MeO-Ph
4-MeO-Ph
4-MeO-Ph
295.3 640
295.3 20
295.3 30
324.3 18
21.0
26.1
25.5
23.9
26.8
>24.5
20.9
395.5
409.4
2
3
22.0
20.8
324.3
2
354.4 <2
384.4
325.3
9
2
54
55
56
57
58
59
60
61
62
4-MeO-3-
pyridyl
4-MeO-3-
pyridyl
4-MeO-3-
pyridyl
4-MeO-3-
pyridyl
4-MeO-3-
pyridyl
4-MeO-3-
pyridyl
4-MeO-3-
pyridyl
4-MeO-3-
pyridyl
Ph
295.3 4.5
329.7
28.3
25.8
20.2
25.9
25.2
25.8
26.3
23.9
19.1
23
4-MeO-3-
pyridyl
4-MeO-Ph
26.7
4-Cl-Ph
3
2-MeO-Ph
3-MeO-Ph
4-CN-Ph
3-pyridyl
4-Pyridyl
4-CO2H-Ph
325.3 270
325.3 3.7
320.3 8.3
296.3 23
296.3 16
339.3 7.7
380.4 50
24
25
26
27
28
29
30
31
32
33
34
3-MeO-Ph
3-MeO-Ph
3-MeO-Ph
3-MeO-Ph
3-MeO-Ph
3-MeO-Ph
3-MeO-Ph
3-MeO-Ph
3-MeO-Ph
3-MeO-Ph
3-MeO-Ph
Ph
294.3 39
25.2
24.6
24.5
25.6
26.3
27.4
25.5
25.9
24.2
23.8
19.1
3-Cl-Ph
4-Cl-Ph
3-MeO-Ph
4-CN-Ph
2-Pyridyl
3-Pyridyl
4-Pyridyl
4-CO2H-Ph
4-NMe2-Ph
4-Morpholino-
Ph
328.8
328.8
324.3
319.3
295.3
8
9
5
4
8
295.3 30
295.3 23
338.3 6.5
337.8
9
379.4 59
4-MeO-3-
pyridyl
4-Morpholino-
Ph
35
36
4-MeO-Ph
4-MeO-Ph
3-Cl-Ph
4-Cl-Ph
328.8 1.9
26.5
26.5
328.8
2

1664
M. Arnost et al. / Bioorg. Med. Chem. Lett. 20 (2010) 1661–1664
Table 3
By employing the simple binding efficiency index calculation to
a screening hit, it has been possible to optimize a new class of
GSK3b inhibitors from 1.5 M to <1 nM, without compromising
binding efficiency. In doing so, molecular weights of the most po-
tent and efficient molecules have been controlled and lie well
within the range of lead-like molecules,17 allowing for further opti-
mization of properties necessary for an orally available, efficacious
therapeutic molecule.
Kinase selectivity profile of compounds 37, 39 and 54
l
MeO
MeO
MeO
N
N
N
N
N
N
H
OMe
N
N
N
N
N
H
H
N
N
N
O
O
O
H
H
H
Acknowledgements
54
37
39
Kinase
37 Ki (nM)
39 Ki (nM)
54 Ki (nM)
The authors thank Ann Grippo, Yu-Ping Luong, Christine Mem-
mott, Cameron Stuver Moody and Paul Taslimi, for enzymology
support.
GSK3b
CDK2
Flt3
0.8
95
66
2
62
130
4.5
76
290
IRAK4
JAK2
JNK3
PI3Kg
KDR
MET
PKA
Plk1
690
1500
>4000
>4000
1000
3400
>4000
>4000
>4000
>4000
>4000
>4000
1100
3500
2500
540
1000
>4000
>4000
>4000
>4000
>4000
1800
>1600
>4000
1300
1800
>4000
>4000
>4000
>4000
>4000
690
Supplementary data
Supplementary data associated with this article can be found, in
the online version, at doi:10.1016/j.bmcl.2010.01.072.
References and notes
ROCK1
Src
Syk
1. Cohen, P.; Yellowlees, D.; Aitken, A.; Donella-Deana, A.; Hemmings, B. A.;
Parker, P. J. Eur. J. Biochem. 1982, 124, 21.
2. Lau, K. F.; Miller, C. C.; Anderton, B. H.; Shaw, P. C. J. Pept. Res. 1999, 54, 85.
3. Hoeflich, K. P.; Luo, J.; Rubie, E. A.; Tsao, M.-S.; Jin, O.; Woodgett, J. R. Nature
2000, 406, 86.
4. Rayasam, G. V.; Tulasi, V. K.; Sodhi, R.; Davis, J. A.; Ray, A. Br. J. Pharmacol. 2009,
156, 885.
shown in Table 3 and shows good overall selectivity (minimally
>20Â) for an early lead series.25
During the course of this work we were able to obtain the X-ray
crystallographic structure26 of compound 46 (GSK3b Ki <2 nM; BEI
>24.2) bound to GSK3b, which confirms the expected binding
mode ( Fig. 3). In this structure, hydrogen-bonding contacts are
made between the pyrazolone NH and CO and the backbone of
hinge residues Asp133 and Val135, respectively. In addition, the
methoxy substituents (R1 = 3,4-dimethoxyphenyl) make hydro-
gen-bonding contacts with both the backbone NH of Asp200 and
the sidechain amine of Lys85. It appears that this set of hydrogen
bonds, plus the hydrophobic contacts made by the two phenyl
rings of the scaffold, form a maximally efficient binding core in
GSK3. Ligands lacking any of these elements are generally less po-
tent, though they may be equally efficient, while molecules con-
taining additional functionality fall off in binding efficiency,
though they may be equally potent.
5. Cohen, P.; Goedert, M. Nat. Rev. Drug Disc. 2004, 3, 479.
6. Rowe, M. K.; Wiest, C.; Chuang, D. Neurosci. Biobehav. Rev. 2007, 31, 920.
7. Jope, R. S.; Roh, M. S. Curr. Drug Targets 2006, 7, 1421.
8. Hernandez, F.; Avila, J. FEBS Lett. 2008, 582, 3848.
9. Hooper, C.; Killick, R.; Lovestone, S. J. Neurochem. 2008, 104, 1433.
10. Savage, M. J.; Gingrich, D. E. Drug Dev. Res. 2009, 70, 125.
11. Sugden, P. H.; Fuller, S. J.; Weiss, S. C.; Clerk, A. Br. J. Pharmacol. 2008, 153, S137.
12. Lee, J.; Kim, M. S. Diabetes Res. Clin. Pract. 2007, 77, S49.
13. Wagman, A. S.; Johnson, K. W.; Bussiere, D. E. Curr. Pharm. Des. 2004, 10, 1105.
14. Medina, M.; Castro, A. Curr. Opin. Drug Discov. Devel. 2008, 11, 533.
15. Hopkins, A. L.; Groom, C. R.; Alex, A. Drug Discovery Today 2004, 9, 430.
16. Abad-Zapatero, C.; Metz, J. T. Drug Discovery Today 2005, 10, 464.
17. Teague, S. J.; Davis, A. M.; Leeson, P. D.; Oprea, T. Angew. Chem., Int. Ed. 1999, 38,
3743.
18. Meijer, L.; Skaltsounis, A.-L.; Magiatis, P.; Polychronopoulos, P.; Knockaert, M.;
Leost, M.; Ryan, X. P.; Vonica, C. A.; Brivanlou, A.; Dajani, R.; Crovace, C.;
Tarricone, C.; Musacchio, A.; Roe, M. S.; Pearl, L.; Greengard, P. Chem. Biol. 2003,
10, 1255.
19. Coghlan, M. P.; Culbert, A. A.; Cross, D. A. E.; Corcoran, S. L.; Yates, J. W.; Pearce,
N. J.; Rausch, O. L.; Murphy, G. J.; Carter, P. S.; Cox, L. R.; Mills, D.; Brown, M. J.;
Haigh, D.; Ward, R. W.; Smith, D. G.; Murray, K. J.; Reith, A. D.; Holder, J. C.
Chem. Biol. 2000, 7, 793.
20. Bertrand, J. A.; Thieffine, S.; Vulpetti, A.; Cristiani, C.; Valsasina, B.; Knapp, S.;
Kalisz, H. M.; Flocco, M. J. Mol. Biol. 2003, 333, 393.
21. Jung, M. E.; Lau, P. Y.; Mansuri, M. M.; Speltz, L. M. J. Org. Chem. 1985, 50, 1087.
22. Compounds described were prepared according to the following general
procedure: To a solution of aniline (1.1 equiv) in 6 M HCl (2 mL/mmol) at 0 °C
is added NaNO2 (1.1 equiv) and the mixture stirred at 0 °C for 30 min. The
solution is added to the pyrazolone (1 equiv in 50% aqueous EtOH, 3 mL/mmol)
and potassium acetate (6 equiv) and the mixture stirred at 0 °C for a further
30 min. The precipitate is filtered, washed with water and dried. Products are
purified by HPLC.
23. Compounds were assayed by modification the coupled assay method described
in: Fox, T.; Coll, J. T.; Xie, X.; Ford, P. J.; Germann, U. A.; Porter, M. D.;
Pazhanisamy, S.; Fleming, M. A.; Galullo, V.; Su, M. S.; Wilson, K. P. Protein Sci.
1998, 7, 2249. Assay details are provided in the Supplementary data.
24. Friesner, R. A.; Banks, J. L.; Murphy, R. B.; Halgren, T. A.; Klicic, J. J.; Mainz, D. T.;
Repasky, M. P.; Knoll, E. H.; Shelley, M.; Perry, J. K.; Shaw, D. E.; Francis, P.;
Shenkin, P. S. J. Med. Chem. 2004, 47, 1739.
25. GSK3b–CDK2 inhibitor selectivity and models for designing improved
selectivity are reported by Vulpetti, A.; Crivori, P.; Cameron, A.; Bertrand, J.;
Brasca, M. G.; D’Alessio, R.; Pevarello, P. J. Chem. Inf. Model. 2005, 45, 1282.
26. Crystallographic data for the compound 46 bound to GSK3b has been deposited
with the RCSB Protein Data Bank. PDB ID: 3L1S.
Figure 3. X-ray crystal structure of compound 46 bound to GSK3b (PDB ID: 3L1S).