Examining the correlations between GSK-3 inhibitory properties and anti-convulsant efficacy of valproate and valproate-related compounds

Article abstract of DOI:10.1016/j.bmcl.2004.09.013

A family of compounds based upon the chemical structure of valproate were synthesized and assayed for their ability to inhibit glycogen synthase kinase (GSK)-3 α and β activity in vitro. A family of compounds based upon the chemical structure of valproate were synthesized and assayed for their ability to inhibit glycogen synthase kinase (GSK)-3 α and β activity in vitro. This data is correlated to the known anti-convulsant properties of these compounds in order to determine the potential role of GSK-3 inhibition in the therapeutic efficacy of these drugs.

Full text of DOI:10.1016/j.bmcl.2004.09.013

Bioorganic & MedicinalChemistry Letters 14 (2004) 5465–5467
Examining the correlations between GSK-3 inhibitory
properties and anti-convulsant efficacy of valproate and
valproate-related compounds
Geoff H. Werstuck,a,* Anna J. Kim,a Timothy Brenstrum,b Stephan A. Ohnmacht,b
Ella Pannab and Alfredo Caprettab,*
aDepartment of Biochemistry and Biomedical Sciences, Henderson Research Centre, 711 Concession Street, Hamilton,
Ontario, Canada L8V 1C3
bDepartment of Chemistry, McMaster University, Hamilton, Ontario, Canada L8S 4M1
Received 18 August 2004; revised 6 September 2004; accepted 7 September 2004
Available online 1 October 2004
Abstract—A family of compounds based upon the chemical structure of valproate were synthesized and assayed for their ability to
inhibit glycogen synthase kinase (GSK)-3 a and b activity in vitro. This data is correlated to the known anti-convulsant properties of
these compounds in order to determine the potentialroel of GSK-3 inhibition in the therapeutic efficacy of these drugs.
ꢀ 2004 Elsevier Ltd. All rights reserved.
Valproic acid (VPA, 2-propylpentanoic acid) is a potent
and widely prescribed drug that acts both as an anti-
convulsant in the treatment of epilepsy and as a mood
stabilizer to control bipolar disorder.1,2 The anti-convul-
sant effects of valproate occur at plasma concentrations
over 0.35mM and toxicity is observed at concentrations
over 1.4mM.1,3 Exposure to valproate is known to
induce a variety of cellular responses, however the mech-
anism(s) by which valproate alleviates convulsions and
modifies behavior is still unclear.4–6 One potential, phy-
siologically relevant target of valproate is the enzyme
glycogen synthase kinase (GSK)-3. The two isoforms,
GSK-3 a and b, were initially identified as homologous
kinases that phosphorylate the rate-limiting enzyme in
glycogen synthesis.7 Over the last 20 years the list of
GSK-3 substrates has grown to encompass a broad
array of proteins that are involved in several important
regulatory and developmental pathways including
translation initiation factor eIF2be, transcription factors
CREB, c-Jun, c-Myc, c-Myb, HSF-1, ATP-citrate lysase
and tau.8 As a result, a great deal of interest has been
focused upon GSK-3 as a potentialtherapeutic target
in the treatment of cancer, diabetes, and neurodegener-
ative diseases.
The ability of valproate to inhibit GSK-3 activity
remains controversial.5,9 A number of compounds that
are chemically related to valproate have been shown to
exhibit similar anti-convulsant properties in animal
models. Very little is known, however, regarding the
effects of these compounds on potential cellular targets.
The present report provides insight into the effect of
VPA on GSK-3 activity in an attempt to determine
the relationship between GSK-3 inhibition and the thera-
peutic efficacy of VPA as an anti-convulsant. We have
prepared a small library of compounds including some
with previously determined anti-convulsant activity
and tested for their ability to inhibit GSK-3 a and b.
As shown in Scheme 1, commercially available valproic
acid (1) was converted to its corresponding acid chloride
and then treated with either methanolic ammonia to
give valpromide (2) or hydroxylamine to give the val-
propylhydroxamic acid (3). Esterification of 1 under
Fischer conditions allowed for the preparation of methyl
valproate (4) which could, in turn, be treated with LDA
at À78ꢁC to give the enolate and converted to either the
a-iodo (5) or the a-methyl( 6) systems shown. Exposure
of 5 to DBU allowed for elimination to the a,b-unsatu-
rated ester (7), which could be readily hydrolyzed to its
Keywords: GSK-3; Valproate.
*
Corresponding authors. Tel.: +1 9055259140x27318; fax: +1
9055222509 (G.H.W.); e-mailaddresses: gwerstuck@thrombosis.
hhscr.org; capretta@mcmaster.ca
0960-894X/$ - see front matter ꢀ 2004 Elsevier Ltd. All rights reserved.
doi:10.1016/j.bmcl.2004.09.013

5466
G. H. Werstuck et al. / Bioorg. Med. Chem. Lett. 14 (2004) 5465–5467
Table 1. GSK-3 Inhibitory Activity
O
R
O
OMe
O
R
GSK-3 activity (% control) Anti-convulsant
activity
a
b
R
Control100 6.6
1
100 5.4
80.4 8.6a
96.6 14.5
54.2 8.7a
104.5 0.3
35.5 7.4a ++
1: R = OH
2: R = NH2
3: R = NH-OH
4: R = H
5: R = I
7: R = OMe
8: R = OH
9: R = NH2
2
83.9 10.4 +++
49.1 18.5a +++
81.8 2.2a
6: R = CH3
3
4
6
99.7 13.2 122.5 18.6
O
R
O
R
61.3 12.9a
100.0 7.7
67.4 12.9a
99.9 0.4
87.9 8.9
99.2 4.4
88.3 15.4
0.5 0.3a
33.3 10.5a
91.3 9.4
O
R
7
8
10
++
24.8 11.9a +++
122.7 23.1
15
16
110.1 23.9
12: R = OH
13: R = OMe
14: R = OiPr
15: R = OMe
16: R = OiPr
17: R = OH
18: R = NH2
10: R = OH
11: R = NH2
17
57.7 12.4a
56.1 5.2a
3.0 0.6a
19.5 1.6a
4.6 1.5a
+
+
19 (0.5mM)
19 (4mM)
LiCl(20mM)
SB415286 (20lM)
H
N
a
O
OH
22.1 6.4
1.5 0.1a
O
O
NO2
4mM of each compound was tested unless stated otherwise; n P 3.
aP < 0.05; quantification of anti-convulsant activity is relative to val-
proate (++) and is based on the findings of Levi et al.,12 Elmazar
et al.,13 and Keane et al.14 in animalmodelsystems. Significantly more
effective (+++) and less effective (+) anti-convulsants are indicated.
No data regarding anti-convulsant activity is known for all other
entries.
HO
N
H
Cl
19
SB 415286
Scheme 1. Synthesis of inhibitor library.
corresponding acid (8) then converted to amide (9). In
the same way, the a-methylsystem ( 6) could be trans-
formed to its corresponding acid (10) or amide (11).
Elaboration of commercially available valeric acid (12)
allowed for access to the so-called 4-ene-VPA system.
Esterification of 12 to either its methyl( 13) or isopropyl
ester (14) set the stage for conversion to the appropriate
enolate prior to treatment with allyl bromide to provide
systems 15 and 16. Hydroylsis of either ester affords 4-
ene-VPA (17), which could be converted to its acid chlo-
ride then derivatized to amide 18. 2-Ethylbutyric acid
(19) was obtained from a commercialsource. Alcom-
pounds were fully characterized using NMR and MS.
It should be noted that compounds containing chiral
centers were prepared as racemates.
ity at 0.5mM. At higher concentrations this compound
completely inhibits both isoforms. Interestingly, we have
identified two compounds, 4 and 17, that appear to
show significant specificity toward GSK-3 b and have
no effect on the a isoform. The trend toward strong
inhibitory activity on b was observed in valproate and
several other compounds as well. Because the active sites
of GSK-3 a and b are highly homologous,11 the isoform
specific nature of this family of inhibitors suggests that
they act distalto the substrate-binding site. Further
studies are required to determine the specific mode of
inhibition for the valproate-related compounds pre-
sented in Table 1. Previously identified inhibitors have
been shown to compete for GSK-3 binding to ATP
(i.e., maleimides, including SB415286) or magnesium
(i.e., lithium).8,11
With compounds in hand, GSK-3 a and b activities
were determined in vitro by assaying the ability of
the recombinant kinase to phosphorylate a synthetic
substrate.10 The inhibitory activity of each of the com-
pounds with respect to GSK-3 is presented in Table 1.
The numbers shown represent the percentage activity
of GSK-3 in the presence of the compound in relation
to the controlexperiment. Lithium, a seelctive inhibitor
of GSK-3 and a maleimide-based inhibitor, SB415286
were included in the assay as controls.
Several of the compounds in our library have previously
been tested for their ability to act as anti-convulsants in
experimentalanimalmodesl. Thus, by comparing these
data sets, we can look for correlations that may indicate
a role for GSK-3 inhibitory action in anti-convulsant
efficacy. We can conclude from this data that GSK-3
inhibitory function does not correlate with anti-convul-
sant activity. This is evident from compound 8 that has
no GSK-3 inhibitory activity but has been shown to be
an effective anti-convulsant.12,13 Conversely ethyl buty-
ric acid, which is the most effective GSK-3 inhibitor that
we have tested, has only limited anti-convulsant proper-
ties.14 To the best of our knowledge, established GSK-3
inhibitors like lithium or SB415286 have not been shown
to have any anti-convulsant activities.
Our results show that VPA is a direct mmol inhibitor of
both GSK-3 a and b. This result is similar to what was
observed by Chen et al.9 and in contrast to the findings
5
of Phielet a.l
The reason for this discrepancy is not
obvious as similar experimental conditions and sub-
strates were used in all cases. Several of the compounds
showed no inhibitory effect on either isoform including
2, 6, 8, 15 and 16. The most effective GSK-3 inhibitor
identified was ethylbutyric acid (compound 19). Ethyl
butyric acid shows significant GSK-3 b inhibitory activ-
In conclusion we have identified novel GSK-3 inhibitors
based on the structure of valproate. Our data suggests
that GSK-3 inhibition does not play a role in the anti-

G. H. Werstuck et al. / Bioorg. Med. Chem. Lett. 14 (2004) 5465–5467
5467
7. Plyte, S. E.; Hughes, K.; Nikolakaki, E.; Pulverer, B. J.;
Woodgett, J. R. Biochim. Biophys. Acta 1992, 1114, 147–
162.
8. Doble, B. W.; Woodgett, J. R. J. Cell Sci. 2003, 116, 1175–
1186.
convulsant efficacy of this family of compounds. Further
analysis is required to determine the mechanism and
physiological relevance of GSK-3 inhibition by
valproate.
9. Chen, G.; Huang, L.-D.; Jiang, Y.-M.; Manji, H. K. J.
Neurochem. 1999, 72, 1327–1330.
Acknowledgements
10. GSK-3 activity was measured by adding 15lM substrate
(phospho-glycogen synthase peptide-2 (pGS-2) (Upstate
Biotech.)) and 0.5Ci/lL [c32P]-ATP to recombinant GSK-
3a/b (New England Biolabs and Upstate Biotech) in a
reaction mixture containing 20mM MOPS, 50lM EDTA,
0.25mM Mg acetate, 5mM MgCl2, 5mM b-glycerol
phosphate, 1mM EGTA, 0.25mM Na3VO4, 0.25mM
The authors wish to thank F. Hoffmann-La Roche AG,
McMaster University, the NaturalSciences and Engi-
neering Research Councilof Canada, Heart and Stroke
Foundation (New Investigator Award) and a Canadian
Institutes for Health Research Grant (#MOP-62910) for
their financialsupport.
DTT and 35lM ATP in a totalvoulme of 40
lL.
Valproate (0-4mM) or test compound (0-4mM) was
added to GSK-3 before the substrate and ATP were
added. Lithium (20mM) was used as a controlfor GSK-3
inhibition. After 30min at 30ꢁC samples were placed on
ice, then spotted onto Whatman P81 phosphocellulose
paper and washed 3· with 0.75% o-phosphoric acid and
once with acetone. 32P incorporation onto the substrate
was determined by scintillation counting and total counts
minus background were normalized to GSK-3 activity in
the absence of inhibitor. Results are presented as the
mean SD. The StudentÕs t-test was performed to deter-
mine significance. For all analyses, P < 0.05 was consid-
ered statistically significant.
References and notes
1. Bowden, C. L.; Brugger, A. M.; Swann, A. C.; Calabrese,
J. R.; Janicak, P. G.; Petty, F.; Dilsaver, S. C.; Davis, J.
M.; Rush, A. J.; Small, J. G.; Garza-Trevino, E. S.; Risch,
S. C.; Goodnick, P. J.; Morris, D. D. J. Am. Med. Assoc.
1994, 271, 918–924.
2. Penry, J. K.; Dean, J. C. J. Clin. Psych. 1989, 50, 17–22.
3. McElroy, S. L.; Keck, P. E., Jr. In Textbook of Psycho-
pharmacology; Schatzberg, A. F., Nemeroff, C. B., Eds.;
American Psychiatric: Washington, DC, 1995; pp 351–375.
4. MacDonald, R. L.; Bergey, G. K. Brain Res. 1979, 170,
558–562.
11. Ali, A.; Hoeflich, K. P.; Woodgett, J. R. Chem. Rev. 2001,
101, 2527–2540.
12. Levi, M.; Yagen, B.; Bialer, M. Pharm Res. 1997, 14, 213–
217.
13. Elmazar, M. M. A.; Hauck, R. S.; Nau, H. J. Pharm. Sci.
1993, 82, 1255–1258.
5. Phiel, C. J.; Zhang, F.; Huang, E. Y.; Guenther, M. G.;
Lazar, M. A.; Klein, P. S. J. Biol. Chem. 2001, 276, 36734–
36741.
6. Williams, R. S. B.; Cheng, L.; Mudge, A. W.; Harwood,
A. J. Nature 2002, 417, 292–295.
14. Keane, P. E.; Simiand, J.; Mendes, E.; Santucci, V.;
Morre, M. Neuropharmacology 1983, 22, 875–879.