IMPLICATION OF CYCLIN-DEPENDENT KINASE 5 IN THE
NEUROPROTECTIVE PROPERTIES OF LITHIUM
E. G. JORDÀ,a1E. VERDAGUER,b1A. M. CANUDAS,a
A. JIMÉNEZ,aS. GARCIA DE ARRIBA,bC. ALLGAIER,b
M. PALLÀSa2AND A. CAMINSa2*
aUnitat de Farmacologia i Farmacognòsia, Facultat de Farmàcia, Uni-
versitat de Barcelona, Nucli Universitari de Pedralbes, E-08028
bRudolf-Boehm-Institut für Pharmakologie und Toxikologie, Universität
Leipzig, Härtelstra?e, 16-18, D-04107 Leipzig, Germany
Abstract—Although numerous studies have demonstrated a
neuroprotective and anti-apoptotic role of lithium in neuronal
cell cultures, the precise mechanism by which this occurs,
remains to be elucidated. In this study, we evaluated the
lithium-mediated neuroprotection against colchicine-induced
apoptosis in cultured cerebellar granule neurons. Previously,
it has been demonstrated that colchicine mediates apoptosis
in cerebellar granule neurons through cytoskeletal alteration
and activation of an intrinsic pro-apoptotic pathway. Recently
we also demonstrated a potential role of cyclin-dependent
kinase 5 (cdk5) in this pathway. Here we report that colchi-
cine induces dephosphorylation in Ser-9 and phosphoryla-
tion in Tyr-216, and thus activation, of glycogen synthase
kinase-3? in cerebellar granule neurons, and that this modi-
fication is inhibited by the presence of 5 mM lithium. How-
ever, the selective glycogen synthase kinase-3? inhibitors
SB-415286 and SB-216763 were unable to prevent colchicine-
induced apoptosis in these cells, suggesting that the anti-
apoptotic activity of lithium is not mediated by glycogen
synthase kinase-3? under these conditions. On the other
hand, 5 mM lithium prevented the colchicine-induced in-
crease in cdk5 expression and breakdown of cdk5/p35 to
cdk5/p25. In addition, we show that up-regulation of cdk5/p25
is unrelated to inhibition of the activity of myocyte enhancer
factor 2, a pro-survival transcription factor. These data sug-
gest a previously undescribed neuroprotective mechanism of
lithium associated with the modulation of cdk5/p35 or cdk5/
p25 expression. © 2005 IBRO. Published by Elsevier Ltd. All
Key words: apoptosis, cdk5/p25, lithium, cerebellar granule
Lithium salts are widely used prophylactically in the treat-
ment of bipolar depression. Moreover, in the last decade a
new potential application of lithium as a neuroprotective
drug has been suggested (Li et al., 2002). This is sup-
ported by the observation that lithium protects against
several pro-apoptotic stimuli in neuronal cell preparations,
such as ?-amyloid-induced neuronal cell death (Alvarez et
al., 1999), potassium deprivation-induced apoptosis in cer-
ebellar granule neurons (Mora et al., 2001, 2002), gluta-
mate neurotoxicity (Nonaka et al., 1998a; Hashimoto et al.,
2003), apoptosis mediated by ouabain in SH-SY5Y neu-
roblastoma cells (Hennion et al., 2002), ceramide-induced
apoptosis in neurons (Centeno et al., 1998) and neurotox-
icity caused by ?-bungarotoxin, a neurotoxin derived from
Bungarus multicinctus venom (Tseng and Lin-Shiau,
2002). In vivo studies also indicate a neuroprotective role
of lithium. Thus, it has recently been demonstrated that low
doses of lithium protect against both ischemic injury and
quinolinate-induced neurotoxicity in a rat model of Hunting-
ton’s disease (Wei et al., 2001). Furthermore, it has also
been shown, using transgenic mice expressing tau protein,
that chronic lithium treatment decreases tau phosphoryla-
tion through inhibition of GSK-3? (Hong et al., 1997; Love-
stone et al., 1999; Pérez et al., 2003).
Several mechanisms have been proposed to explain
the neuroprotective properties of lithium: Chen and
Chuang (1999) demonstrated that long-term lithium treat-
ment caused an increase in the expression of the anti-
apoptotic protein bcl-2 and a decrease in the levels of the
pro-apoptotic factors p53 and Bax, while Mora et al. (2002)
demonstrated that lithium inhibits the dephosphorylation of
PKB. Additional neuroprotective mechanisms have been
described, for example, lithium increases the activity of
Akt, modulates NMDA receptor activity and reduces the
intracellular calcium increase mediated by NMDA receptor
stimulation (Hashimoto et al., 2003). However, the best-
studied pathway involved in the neuroprotective properties
of lithium is the inhibition of glycogen synthase kinase-3?
(GSK-3?) activity, an enzyme with pro-apoptotic properties
that is involved in tau phosphorylation and plays an impor-
tant role in Alzheimer’s disease (Hongisto et al., 2003).
Recently it has been demonstrated that lithium protects
cerebellar granule neurons (CGNs) from apoptosis in-
duced by serum-potassium deprivation through the inhibi-
tion of MEF2 hyperphosphorylation. The mechanism as-
sociated with this neuroprotective effect is unknown, al-
though it has been demonstrated that is not mediated
through inhibition of GSK-3? (Linseman et al., 2003). Con-
sequently, Linseman et al. (2003) concluded that addi-
tional neuroprotective pathways must exist to explain the
mechanism of lithium-mediated neuroprotection.
Colchicine-induced apoptosis in CGNs is a model of
programmed cell death mediated by cytoskeletal alter-
ation. In this model, colchicine activates the intrinsic ap-
1The first two authors contributed equally to this work.
E-mail address: firstname.lastname@example.org (A. Camins).
Abbreviations: cdk5, cyclin-dependent kinase 5; CGN, cerebellar gran-
ule neuron; FCS, fetal calf serum; GSK-3?, glycogen synthase kinase-
3?; MEF2, myocyte enhancer factor 2; PBS, phosphate-buffered
saline; PI, propidium iodide; TBS-T, 50 mM Tris, NaCl 1.5%, 0.05%
Tween 20, pH 7.5.
Neuroscience 134 (2005) 1001–1011
0306-4522/05$30.00?0.00 © 2005 IBRO. Published by Elsevier Ltd. All rights reserved.
optotic pathway (Bonfoco et al., 1995; Gorman et al., 1999;
Volbracht et al., 2001; Fonfría et al., 2002). Previous stud-
ies have demonstrated a release of pro-apoptotic mito-
chondrial proteins, such as cytochrome C and AIF (ap-
optosis inducing factor), after treatment with colchicine
(Gorman et al., 1999; Fonfría et al., 2002). Thus, colchicine
can induce apoptosis through the activation of caspase-
dependent and caspase-independent pathways. Studies
performed in our laboratory also suggest a role of cdk5, a
kinase implicated in microtubular disruption in Alzheimer’s
disease, in the pro-apoptotic pathway initiated by colchi-
cine in CGNs (Jordà et al., 2003).
In this study, we show that colchicine increases the
expression of cdk5 and pro-apoptotic protein p25. In con-
trast, lithium downregulates expression of these proteins
suggesting the existence of a previously undescribed pro-
apoptotic pathway modulated by lithium that may be im-
portant in the understanding of the neuroprotective and
anti-apoptotic effects of this drug.
Eagle’s basal medium (BME) and fetal calf serum (FCS) were
from GIBCO (Life Technologies, Paisley, UK). Lithium, propidium
iodide, colchicine, SB-415286, SB-216763 and all other chemicals
were from Sigma (St. Louis, MO, USA).
Primary cultures of cerebellar granule cells were prepared from 7
day-old Sprague–Dawley rat pups according to the method of
Verdaguer et al. (2002). Cerebella, freed of meninges, were
trypsinized and treated with DNase. Cell suspensions were ad-
justed to 8.0?105cells/ml and then cells were plated on poly-L-
lysine-coated 96 or 24 well plates at a density 3.2?105cells/cm2.
Cultures were grown in Eagle’s basal medium containing 10%
FCS, 2 mM L-glutamine, 0.1 mg/ml gentamycin and 25 mM KCl.
Cytosine arabinoside (10 ?M) was added 16–18 h after plating to
inhibit the growth of non-neuronal cells. Cultures prepared by this
method were enriched to contain more than 95% granule neurons.
The care and use of these animals were performed according to
the policy on the use of animals in neuroscience research pub-
lished by the Society for Neuroscience. The protocols were ap-
proved by a review committee of the University of Barcelona under
the supervision of the “Generalitat de Catalunya.”
Treatment of CGNs and survival assay
CGNs were used after 7–10 days in vitro. Colchicine was diluted
in culture medium and added to the cell culture up to 1 ?M. Lithium
(1–5 mM) was added 24 h prior to colchicine addition. To inves-
tigate the effect of lithium on serum-potassium deprivation condi-
tions it was added to the medium, at the precise concentrations
(0.1–5 mM) 24 h before complete medium was replaced by de-
prived medium, without serum and containing 5 mM KCl (this
condition is referred to as S/K deprivation) and lithium at the exact
concentrations. Cell death was determined 24 h after colchicine
addition or 12 h after S/K deprivation using the MTT assay as
follows: MTT was added to the cells at a final concentration of
250 ?M and incubated for 1 h to allow the reduction of MTT to
produce a dark blue formazan product (Hansen et al., 1989). The
media was removed, and cells were dissolved in dimethylsulfox-
ide. Formazan formation was then assayed by measuring the
change in absorbance at 595 nm on a microplate reader (BioRad
Laboratories, CA, USA). Viability results were expressed as per-
centages of the absorbance measured in untreated cells. CGNs
were used after 7–10 days in vitro.
Analysis of apoptosis by flow cytometry
Apoptosis was measured after 24 h of colchicine treatment or 12 h
after S/K deprivation. Briefly, the culture medium was removed
and cells were collected by pipetting and washed with PBS. Pro-
pidium iodide (PI; 10 ?g/ml) was added 30 min before flow cy-
tometry experiments on an Epics XL flow cytometer. The instru-
ment was set up with the standard configuration: excitation of the
sample was performed using an air-cooled argon-ion laser at
488 nm and 15 mW power. Forward scatter (FSC), side scatter
(SSC) and red (620 nm) fluorescence for PI were acquired. Op-
tical alignment was based on optimized signal from 10 nm fluo-
rescent beads (Immunocheck, Epics Division). Time was used as
a control of the stability of the instrument. Red fluorescence was
projected on a 1024 monoparametrical histogram. Aggregates
were excluded, gating single cells by their area vs. peak fluores-
Detection of apoptotic nuclei by PI staining
PI staining was used to evaluate morphological evidence of
apoptosis, e.g. condensed nuclei. After treatment, cells were
fixed in 4% paraformaldehyde/phosphate-buffered saline solu-
tion (PBS), pH 7.4, for 1 h at room temperature. After washing
with PBS, they were incubated for 3 min with a solution of PI in
PBS (10 ?g/ml). Coverslips were mounted in Mowiol® 4-88.
Stained cells were visualized under UV illumination using the
20? objective of a Nikon Eclipse fluo microscope and digitized
images were captured.
Apoptotic cells contained shrunken, brightly fluorescent ap-
optotic nuclei with condensed chromatin, compared with non-
apoptotic cells. Apoptotic cells were scored by counting at least
500 cells for each sample in three different experiments.
Transmission electron microscopy
Neurons were fixed by addition of 2.5% glutaraldehyde in 0.1 M
PBS (pH 7.4) at 4 °C for 2 h and then post-fixed in 1% osmium
tetroxide and 0.8% potassium ferricyanide in the same buffer at
4 °C overnight. Samples were dehydrated in graded concentra-
tions of acetone and embedded in Spurr resin. Ultrathin sections
were obtained with an Ultracut E (Reichert-Jung) ultramicrotome,
stained with uranyl acetate and lead citrate, and then analyzed
under a transmission electron microscope (H-800, Hitachi) oper-
ated at 100 kV.
Caspase activity determination
nitroaniline was used for the determination of caspase-3 activity,
Ac-VEID-pNA for caspase-6 activity and Ac-LEHD-pNA for
caspase-9 activity, according to the following method: after treat-
ment with colchicine in presence or absence of lithium 5 mM,
CGNs were collected in a lysis buffer (50 mM HEPES, 100 mM
NaCl, 0.1% CHAPS, 0.1 mM EDTA, pH 7.4) and 0.5 ?g/?l of
protein was then incubated at 37 °C in 96-well plates containing
200 ?M colorimetric substrate in assay buffer (50 mM HEPES,
100 mM NaCl, 0.1% CHAPS, 10 mM dithiothreitol, 0.1 mM EDTA,
pH 7.4) for 24 h. Absorbance of the cleaved product was mea-
sured at 405 nm in a microplate reader (BioRad). Results were
expressed as arbitrary units of absorbance.
Colorimetric substrate Ac-DEVD-p-
Immunoblot assay of ?-spectrin digestion.
caspase-3 activity, ?-spectrin degradation, was analyzed by
Western blot analysis using aliquots containing 5 ?g of protein
per sample. Briefly, samples were placed in sample buffer
As measure of
E. G. Jordà et al. / Neuroscience 134 (2005) 1001–10111002
(0.5 M Tris–HCl pH 6.8, 10% glycerol, 2% w/v SDS, 5% v/v
2-?-mercaptoethanol, 0.05% Bromophenol Blue) and dena-
tured by boiling at 95–100 °C for 5 min. Samples were sepa-
rated by electrophoresis on 5% acrylamide gels. Thereafter,
proteins were transferred to polyvinylidene fluoride (PVDF)
sheets (ImmobilonTM-P, Millipore Corporation, Bedford, MA,
USA) using a transblot apparatus (BioRad). Membranes were
blocked overnight with 5% non-fat milk dissolved in TBS-T
buffer (50 mM Tris; NaCl 1.5%; 0.05% Tween 20, pH 7.5). They
were then incubated with a primary antibody against ?-spectrin
(Chemicon International, Temecula, CA, USA). After 90 min,
blots were washed thoroughly in TBS-T buffer and incubated
for 1 h with a peroxidase-conjugated IgG secondary antibody
(Amersham Corporation, Arlington Heights, IL, USA). Immuno-
reactive protein was visualized using a chemiluminescence-
based detection kit according to the manufacturer’s protocol
(ECL kit; Amersham Corporation). Protein loading was routinely
monitored by Phenol Red staining of the membrane.
For immunoblotting, proteins (30 ?g/lane) were size-separated by
SDS-PAGE (10% acrylamide) and transferred to polyvinylidene
fluoride membranes as described above. Membranes were incu-
bated in the presence of 5% non-fat milk and incubated overnight
at 4 °C with primary antibodies recognizing either cdk5 (Ab-4,
1:1000; NeoMarkers Inc., Fremont, CA, USA), p-GSK-3? (Ser 9)
(1:500; Affinity BioReagents polyclonal antibodies), p-GSK-3?
(Tyr 216, sc-11758), GSK-3? (Leu-17, sc-8257) and MEF-2 (Ser
59, sc-13919-R) (1:500; Santa Cruz Biotechnology, Santa Cruz,
CA, USA). Afterward, immunoreactive protein was visualized as
For immunocytochemistry experiments, CGNs were grown on
sterile coverslips. After stimuli, cells were washed twice in PBS
and fixed in 4% paraformaldehyde/PBS, pH 7.4 for 1 h at room
temperature. They were pre-incubated for 30 min in PBS contain-
ing 0.3% Triton X-100 and 30% normal horse serum at room
temperature. The cultures were immunostained with antibodies
specific for cdk5 (Ab-4, 1:400 dilution; NeoMarkers, Inc.),
p-GSK-3? (Ser 9), p-GSK3? (Tyr 216) or p-MEF2 (Ser 59) (1:400;
Santa Cruz Biotechnology) followed by rhodamine-conjugated anti-
rabbit IgG or anti-mouse IgG (1:200). Subsequently, coverslips
were thoroughly washed and mounted in Mowiol® 4-88 and cells
were then imaged using fluorescence microscopy at 100? oil
immersion objective (Nikon Eclipse).
Values presented are the mean?S.E.M. from at least four exper-
iments performed in four to six independent cultures. Data were
analyzed by ANOVA followed by the Tukey-Kramer multiple com-
Lithium inhibits colchicine-induced apoptosis
Cytotoxicity was evaluated using the MTT assay after in-
cubation of CGNs for 24 h with 1 ?M colchicine. Exposure
of CGNs to colchicine resulted in a significant decrease in
formazan product. Pre-treatment with lithium (1–5 mM) for
24 h prior to addition of colchicine (1 ?M) significantly
inhibited neuronal cell death, with an increase in cell via-
bility from 60% in colchicine-treated CGNs to 95% follow-
ing 5 mM lithium pre-treatment (Fig. 1). The analysis of
apoptotic nuclei by flow cytometry revealed that colchicine
treatment significantly increased the number of aneuploid
cells and that lithium reversed DNA fragmentation (Fig. 2).
Furthermore, we used PI labeling to analyze the percent-
age of nuclei containing condensed chromatin, as an ap-
optotic index to confirm the neuroprotective effects of lith-
ium. Fifty-five percent of colchicine-treated CGNs had con-
densed nuclei, compared with only 20% of those treated
with 5 mM lithium (Fig. 2). The percentage of cells with
condensed nuclei in control samples was approximately
10%. Analysis of nuclear morphology by electron micros-
copy confirmed the results obtained by PI staining, thus
revealing an apoptosis-like chromatin condensation upon
colchicine treatment that was significantly reduced in the
presence of 5 mM lithium (Fig. 2).
We measured the enzymatic activity of caspases 3, 6
and 9 to further demonstrate that colchicine-induced neu-
rotoxicity involves activation of the caspase pathway. Ex-
posure of CGNs to colchicine (1 ?M) for 24 h induced a
significant increase in caspase activity that was prevented
1 10 15 30
1 2 34 5
Lithium [mM] SB415286 [µM] SB-216763 [µM]
110 15 30
Fig. 1. Assessment of cell viability by MTT assay on colchicine-induced cell death in CGNs. Effects of lithium and GSK-3? inhibitors. Means?S.E.M.
of n (three to five) determinations from four different experiments are given. Significant differences vs. the colchicine group: * P?0.05, ** P?0.01.
E. G. Jordà et al. / Neuroscience 134 (2005) 1001–10111003
by a 24 h pre-treatment with lithium (Fig. 3). Moreover, we
measured the enzymatic activity of caspase-3 by Western
blot analysis of ?-spectrin. Upon caspase-3 activation
?-spectrin is degraded to a 120 kDa fragment, SBDP120.
Protein samples from CGNs treated with 1 ?M colchicine
contained a specific 120 kDa band corresponding to
SBDP120, indicative of caspase-3 activation. In the pres-
ence of 5 mM lithium there is a reduction in this band (Fig. 3).
Neuroprotective effects of lithium on serum/potassium
deprivation-induced apoptosis in CGNs
To determine whether the neuroprotective and antiap-
optotic properties of lithium, we have evaluated the role of
lithium in a well-established model of apoptosis in CGNs,
S/K deprivation. As it has been described previously, the
treatment with increasing concentrations of lithium (1 mM–
5 mM) jointly S/K deprivation for 12 h showed a significant
increase in neuronal cell viability (Fig. 4). In addition, in this
paradigm, lithium 5 mM attenuates S/K-induced apoptosis,
measured by flow cytometry (DNA fragmentation) and nu-
clear condensed nuclei (Fig. 4).
Evaluation of the role of GSK-3? in
colchicine-induced apoptosis in CGNs
GSK-3? is a key regulator of neuronal cell death and it has
been demonstrated that lithium selectively inhibits the ac-
A B C D E
Colchicine, 1 µM
% PI positive nuclei
% FACS analysis
Control Li, 5 mMSB2,15 µ
Fig. 2. Evidences of apoptotic cell death induced by colchicine. Upper panel. Quantitative evaluation of colchicine-induced apoptosis and the effect
of lithium, SB-216763 and SB-452861 by fluorescence microscopy (open bars, left axis) and flow cytometry (black bars, right axis) using PI staining.
Results are expressed as the percentage of PI positive nuclei and apoptotic cells and represent the mean?S.E.M. Values derive from counts of at
least 500 nuclei in four or more random microscope fields per condition and error bars represent the S.E.M. of four experiments. Lower panel.
Representative microscopy fields shown fluorescence labeling of DNA in the nuclei of control (A), 1 ?M colchicine (B), colchicine pre-treated with 5 mM
lithium (C), colchicine pre-treated with 15 ?M SB-216763 (D), and colchicine pre-treated with 15 ?M SB-415286 (E). Scale bar?10 ?m. Represen-
tative electronic microscopy fields of neurons treated with 1 ?M colchicine in the absence (G) or presence of 5 mM lithium (H), control cells (F). Cultures
were fixed with 2.5% glutaraldehyde for transmission electron microscopy analysis. In colchicine-treated neurons, chromatin condensed to form
crescent-shaped and spherical structures.
E. G. Jordà et al. / Neuroscience 134 (2005) 1001–10111004
tivity of this enzyme (Kaytor and Orr, 2002; Chuang et al.,
2002). Accordingly, we determined whether colchicine in-
duced GSK-3? activation using an antibody against p-GSK-
3? (Ser-9) and GSK-3? (Tyr-216). Analysis of lysates from
colchicine-treated CGNs revealed that colchicine treatment
induced the dephosphorylation of GSK-3? (Ser-9) and the
phosphorylation of GSK-3? (Tyr-216), therefore, activates
GSK-3?. Both effects were prevented by 5 mM lithium. In
contrast, levels of unphosphorylated GSK-3?, analyzed
using a GSK-3? (Leu-17) antibody, remained unchanged
under the various experimental conditions (Fig. 5).
Next we determined whether inhibition of GSK-3?
mimics the anti-apoptotic effect of lithium on colchicine
treatment in CGNs using two well-known selective inhibi-
tors of GSK-3?, namely SB-415286 and SB-216763
(Cross et al., 2001; Facci et al., 2003). When cultures were
treated with colchicine in the presence of varying concen-
trations of either SB-415286 or SB-216763, we observed
no increase in neuronal survival using the MTT method
(Fig. 1). Flow cytometric analysis revealed an increase in
PI fluorescence in response to 1 ?M colchicine, indicating
neuronal DNA fragmentation, that was not prevented by
the addition of SB-415286 or SB-216763 at any of the
concentration tested (Fig. 2). Analysis of the condensed
nuclei after treatment with 1 ?M colchicine revealed no
difference in nuclear morphology in response to addition of
SB-415286 or SB-216763 (Fig. 2). These data suggest
that selective inhibition of GSK-3? is not sufficient to pre-
vent colchicine-induced apoptosis in CGNs.
Lithium prevents colchicine- and serum/potassium
deprivation-induced cdk5/p25 expression in CGNs
We have previously demonstrated that inhibition of cdk5
prevents colchicine-induced apoptosis (Jordà et al., 2003).
Since our results suggest that specific inhibition of GSK-3?
is insufficient to prevent the neurotoxic effects of colchicine
on CGNs, we evaluated the expression of cdk5 and p35/
p25 in response to lithium pre-treatment. Western-blots
revealed an increase in the expression of cdk5 protein
Fig. 3. Determination of caspase pathway involvement in colchicine induced apoptosis. Upper panel: Colchicine (1 ?M)-induced increase in
caspase-3, -6, -9 activity in CGNs and the effect of pretreatment with lithium (5 mM for 24 h) Results are shown as the mean?S.E.M. of three
determinations from three different cultures. Significant differences vs. respective controls: * P?0.05, ** P?0.01 vs. respective control values (Ct);
#P?0.05 vs. respective colchicine values. Lower panel: Expression of the ?-spectrin degradation product SBDP120 as an indicator of caspase-3
activity in CGNs treated with colchicine or colchicine plus lithium.
E. G. Jordà et al. / Neuroscience 134 (2005) 1001–1011 1005
after colchicine and S/K deprivation treatment, which was
clearly inhibited by 24 h of pre-treatment with 5 mM lithium.
p25 Levels were also increased after colchicine and S/K
deprivation treatment, but not when cells were pre-treated
with lithium (Fig. 6). Furthermore, immunocytochemistry
with an antibody against cdk5 showed that colchicine treat-
ment induced nuclear localization of the protein (Fig. 6).
Colchicine does not induce MEF-2 activation
Recently, it has been demonstrated that glutamate through
cdk5/p25 activation, induces MEF-2 phosphorylation and in-
activation (Gong et al., 2003). Furthermore, MEF-2 is neces-
sary for neuronal survival (Mao et al., 1999). Thus, we tested
whether the activation of cdk5/p25 by colchicine also inacti-
vates MEF-2. Western-blot and immunocytochemical analy-
sis revealed no effect of colchicine treatment on MEF-2 when
compared with serum and potassium withdrawal-induced of
MEF-2 activation. As described (Linseman et al., 2003), S/K
deprivation induced changes in MEF-2 phosphorylation, that
are prevented by lithium. These data confirm a role of lithium
in this apoptotic pathway (Fig. 7).
Lithium is a drug used widely in the treatment of bipolar
disorders, but since the demonstration of its ability to inhibit
1mM 2mM3mM 4mM5mM
S/K deprivation, 12h
CTS/K S/K+Li, 5mM
% FACS Analysis
Fig. 4. Neuroprotective effect of lithium in S/K deprivation paradigm. Upper panel: Effects of lithium on serum/potassium deprivation-induced cell
death in CGNs. Lower panel. Representative microscopy fields shown fluorescence labeling of DNA in the nuclei of control (A), 12 h serum potassium
deprivation (B), serum potassium deprivation treated with 5 mM lithium and quantitative evaluation of serum-potassium deprivation-induced apoptosis
and the effect of lithium (5 mM) by fluorescence microscopy (open bars, left axis) and flow cytometry (black bars, right axis) using PI staining. Scale
bar?10 ?m. Means?S.E.M. of n (three to five) determinations from four different experiments are given. * P?0.05, *** P?0.001, Significant
differences vs. control group.###P?0.001 significant differences vs. the S/K deprivation group.
E. G. Jordà et al. / Neuroscience 134 (2005) 1001–10111006
GSK-3?, many lines of evidence from in vitro and in vivo
studies have highlighted its potential application as a neuro-
protective drug (Nonaka et al., 1998b; Li and El-Mallahk,
2000; Mora et al., 2002; Kaytor and Orr, 2002; Kang et al.,
2003). Although several mechanisms have been sug-
gested to explain the neuroprotective properties of lithium,
the complete pathway remains to be elucidated. The evi-
dence that GSK-3? is able to phosphorylate tau, suggests
that lithium may be a potentially effective drug in the treat-
ment of Alzheimer’s disease (Chuang et al., 2002). Fur-
thermore, lithium may have potential applications in the
treatment of other neurodegenerative diseases involving
cytoskeletal alteration and GSK-3? activation (Brandt
2001; Chuang et al., 2002).
Colchicine is a neurotoxin that induces cytoskeletal
alterations and is a useful in vitro tool for the evaluation of
drugs with potential applications in the treatment of neuro-
degenerative diseases involving cytoskeletal alteration
(Brandt, 2001; Canu and Callissano, 2003). In this study
we have shown that acute lithium treatment effectively
inhibits colchicine-induced apoptosis via an inhibition of
The apoptotic pathway can be divided into three se-
quential phases: the induction, effector and degradation
phases (Kroemer et al., 1997; Chang et al., 2002). Al-
though lithium inhibits the degradation phase of colchicine-
induced apoptosis, these results are not sufficient to ex-
plain its anti-apoptotic properties.
Analysis of GSK-3? activation by immunodetection
techniques indicated that the presence of lithium leads to
an inhibition of this enzyme. However, our results argue
against a prominent and exclusive role of GSK-3? in the
Fig. 5. Activation of GSK-3? after colchicine treatment. Effect of lithium. Upper panel: Representative Western blot for phospho-GSK-3? (Tyr-216),
phospho-GSK-3? (Ser9) and GSK-3? (Leu-17) under several experimental conditions. Bar chart show semiquantitative bands optical densities (O.D.)
in arbitrary units) in the same experimental conditions. Significant differences vs. respective controls: * P?0.05, respective control values;#P?0.05
vs. respective colchicine values. Lower panel: Immunodetection of phospho-GSK-3? (Tyr-216) and phospho-GSK-3? (Ser-9) in colchicine-treated
CGNs. Both experiments indicate an activation of GSK-3? without modification of expression levels. (Scale bar?5 ?m).
E. G. Jordà et al. / Neuroscience 134 (2005) 1001–10111007
apoptotic pathway utilized by colchicine, because the se-
lective GSK-3? inhibitors, SB-415286 and SB-216763, did
not inhibit colchicine-induced apoptosis. Previous studies
demonstrated a neuroprotective role of these GSK-3? in-
hibitors against excitotoxicity and potassium deprivation
(Cross et al., 2001; Facci et al., 2003). The observation
that roscovitine, a cdk1/2/5 inhibitor, that unlike other cdk
inhibitors, does not inhibit GSK-3?, exhibits neuroprotec-
tive effects, confirms our hypothesis that additional targets
must be involved in colchicine-induced apoptosis in CGNs
(Jordà et al., 2003).
Previously, we have suggested that the anti-apoptotic
and neuroprotective effects of cdk inhibitors on colchicine-
induced neurotoxicity may be mediated by inhibition of
cdk5 (Jordà et al., 2003), and it has been proposed that
both cyclin-dependent kinases and GSK-3? are involved in
the pathogenesis of Alzheimer’s disease through their ca-
aert et al., 2002; Lim and Qi, 2003; Lee and Tsai, 2003).
Cdk5 is a cyclin-dependent kinase not involved in cell cycle
proliferation that is widely present in the brain and plays an
important role in neuronal development (Nguyen et al.,
2002). In the brain the kinase activity cdk5 is regulated by
two neuron-specific co-activators, p35 and p39. After a
neurotoxic stimulus, p35 and p39 are converted into p25
and 29, respectively. Recent studies have demonstrated
that cdk5/p25 acts as pro-apoptotic signal involved in the
process of neuronal cell death in several neurodegenera-
tive diseases (Sengupta et al., 1997; Lee and Tsai, 2003).
The cytoplasmatic localization and overexpression of
cdk5/p25 is associated with tau phosphorylation and thus
contributes to Alzheimer’s disease progression. Cdk5/p25
can also localize to the nucleus and modulate gene tran-
scription (Gong et al., 2003).
Our results suggest that colchicine and S/K withdrawal
induce the up-regulation of cdk5 expression and the break-
down of p35 to form the more stable p25. The increase in
the expression of these proteins contributes to apoptotic
Fig. 6. Role of Cdk5/p25 activation in colchicine-induced apoptosis. Effect of lithium. Upper panel: Representative Western blot of cdk5 and p35/p25.
Cdk5 is overexpressed and translocated into the nucleus in CGNs after colchicine treatment, effects being blocked in the presence of 5 mM lithium.
Western blot analysis of p35/p25 immunoreactivity in CGNs under the same conditions, showing the p25/p35 ratio change. Similar results were
obtained in response to serum/potassium withdrawal in absence or presence of 5 mM lithium. Bar chart show semiquantitative p25/p35 ratio (O.D.,
arbitrary units) in the same experimental conditions. Significant differences vs. respective controls: ** P?0.01, control value;##P?0.01 vs. respective
neurotoxic stimuli. Lower panel: Immunostaining of CGNs 24 h after colchicine treatment in the absence (B) or presence (C) of lithium demonstrated
an increase on cdk5 expression after treatment with neurotoxin (A). Control cells. (Scale bar?25 ?m.)
E. G. Jordà et al. / Neuroscience 134 (2005) 1001–10111008
cell death. In contrast, acute treatment of CGNs with lith-
ium prevents these colchicine-induced changes.
In addition, our results provide further evidence that
colchicine mimics some of the abnormalities in enzymatic
activation associated with Alzheimer’s disease. Immuno-
histochemistry with an anti-cdk5 antibody showed that af-
ter a treatment with colchicine cdk5 exhibits a nuclear
localization. These results are in agreement with earlier
studies suggesting that the nuclear presence of cdk5/p25
could have an important role in the apoptotic process.
It has been reported previously that lithium increases
the expression of the pro-survival transcription factor
MEF-2 (Linseman et al., 2003; Gong et al., 2003) which
is a target for direct phosphorylation by cdk5, thereby
inhibiting its pro-survival activity. As expected, S/K de-
privation modifies the expression of MEF-2 levels that
were increased in presence of lithium 5 mM. However
our results suggest that MEF-2 is not involved in the
apoptotic pathway initiated by colchicine, since this neu-
rotoxin did not modify the expression levels of MEF-2.
Changes in MEF-2 expression were observed after S/K
deprivation. These changes were inhibited by lithium, in
contrast colchicine treatment did not modifies this path-
way suggesting that the apoptotic route mediated by
CGNs in both paradigms is different. Thus, meanwhile
both stimuli induced changes in cdk5 expression, in S/K
deprivation cdk5 mediated apoptosis probably through a
nuclear alteration of MEF-2, fact that did not occur with
colchicine-induced apoptosis in CGNs. Recently, sev-
eral authors suggest that cdk5 is involved in tau dephos-
phorylation by a mechanism dependent of phosphatase
phosphorylation and a cross-talk between cdk5, GSK-3?
and phosphatase PP-1 could also exists (Zambrano et
al., 2004; Morfini et al., 2004). In this way a possible
pathway involved in colchicine-induced apoptosis could
be mediated by cdk5 through phosphatase PP-1 phos-
phorylation and tau dephosphorylation.
Recently, using a transgenic mouse model that exhib-
its progressive deposition of ?-amyloid in cortical brain
regions, it has been demonstrated that cdk5/p25 are spe-
cifically involved in tau hyperphosphorylation, with GSK-3?
and other MAP kinases playing a lesser role (Otth et al.,
2002). These in vivo data are in agreement with our results
in vitro and suggest an important role for cdk5/p25 in the
pathogenesis of Alzheimer’s disease, where lithium could
offer a potential treatment.
Fig. 7. Role of MEF-2 in colchicine-induced apoptosis. Effect of lithium phospho-MEF-2 (Ser 59) immunoreactivity in CGNs. Immunocytochemistry studies
showed that level of phospho-MEF-2 was maintained in both control (A) and colchicine-treated cells in the absence (B) or presence of 5 mM lithium (C). In
contrast, in S/K deprived CGNs, the immunoreactivity of phospho-MEF-2 (Ser 59) was reduced (D) and this was prevented by 5 mM lithium (E). (F)
Representative Western blots showing that colchicine does not cause modification of phospho-MEF-2 (Ser 59) in CGNs, while protein levels were
substantially decreased after 24 h of serum and potassium withdrawal, compared with control samples. Pre-treatment with lithium did not induce modification
of phospho-MEF-2 (Ser 59) levels in colchicine-treated cultures but does it in 24 h of serum and potassium withdrawal (scale bar?5 ?m).
E. G. Jordà et al. / Neuroscience 134 (2005) 1001–10111009
Taken together, the results of this study show that cdk5
regulation is involved in the neuroprotective properties of
lithium. Thus, the neuroprotective and anti-apoptotic prop-
erties of lithium, through the modulation of GSK-3? and
cdk5/p25, give support to the suggestion that drugs such
as flavopiridol and induribins, dual inhibitors of these two
enzymes, have potential applications in the treatment of
Acknowledgments—The authors would like to thank Maria
Teresa Iglesias for excellent technical assistance, and the mi-
croscopy service of the University of Barcelona for help with
electron microscopy experiments. The excellent secretarial
support of Ms Mar Morales is greatly appreciated. We thank the
Language Assessment Service of the University of Barcelona
for revising the manuscript. This study was supported by grants
SAF2002-00790 from Ministerio de Educación y Ciencia
(FEDER founds), FISS G03/137, FISS G03/167 PI04/1300
from Instituto de Salud Carlos III and Interdisziplinäre Zentrum
für Klinische Forschung (IZKF) at the Faculty of Medicine of the
University of Leipzig (Project C20). E.V. is the recipient of a
Alvarez G, Muñoz-Montaño JR, Satrústegui J, Avila J, Bogónez E,
Díaz-Nido J (1999) Lithium protects cultured neurons against
?-amyloid-induced neurodegeneration. FEBS Lett 453:260–264.
Bonfoco E, Ceccatelli S, Manzo L, Nicotera P (1995) Colchicine in-
duces apoptosis in cerebellar granule cells. Exp Cell Res 218:
Brandt R (2001) Cytoskeletal mechanisms of neuronal degeneration.
Cell Tissue Res 305:255–265.
Canu N, Callissano P (2003) In vitro cultured neurons for molecular
studies correlating apoptosis with events related to Alzheimer dis-
ease. Cerebellum 2:270–278.
Centeno F, Mora A, Fuentes JM, Soler G, Claro E (1998) Partial
lithium-associated protection against apoptosis induced by C2-
ceramide in cerebellar granule neurons. Neuroreport 9:199–
Chang LK, Putcha GV, Deshmukh M, Johnson EM (2002) Mitochon-
drial involvement in the point of no return in neuronal apoptosis.
Chen RW, Chuang DW (1999) Long term lithium treatment sup-
presses p53 and bax expression but increases Bcl-2 expression.
J Biol Chem 274:6039–6042.
Chuang DM, Chen RW, Chalecka-Franaszek E, Ren M, Hashimoto R,
Senatorov V, Kanai H, Hough C, Hiroi T, Leeds P (2002) Neuro-
protective effects of lithium in cultured cells and animal models of
diseases. Bipolar Disord 4:129–136.
Cross DA, Culbert AA, Chalmers KA, Facci L, Skaper SD, Reith AD
(2001) Selective small-molecule inhibitors of glycogen synthase
kinase-3 activity protect primary neurons from death. J Neurochem
Facci L, Stevens DA, Skaper SD (2003) Glycogen synthase kinase-3
inhibitors protect central neurons against excitotoxicity. Neurore-
Fonfría E, Daré E, Benelli M, Suñol C, Ceccatelli S (2002) Transloca-
tion of apoptosis-inducing factor in cerebellar granule cells ex-
posed to neurotoxic agents inducing oxidative stress. Eur J Neu-
Gong X, Tang X, Wiedmann M, Wang X, Peng J, Zheng D, Blair LA,
Marshall J, Mao Z (2003) Cdk5-mediated inhibition of the protec-
tive effects of transcription factor MEF2 in neurotoxicity-induced
apoptosis. Neuron 38:33–46.
Gorman AM, Bonfoco E, Zhivotovsky B, Orrenius S, Ceccatelli S
(1999) Cytochrome c release and caspase-3 activation during
colchicine-induced apoptosis of cerebellar granule cells. Eur
J Neurosci 11:1067–1072.
Hansen MB, Nielsen SE, Berg K (1989) Re-examination and further
development of a precise and rapid dye method for measuring cell
growth/cell kill. J Immunol Methods 119:203–210.
Hashimoto R, Fujimaki K, Jeong MR, Christ L, Chuang DM (2003) Lithi-
um-induced inhibition of Src tyrosine kinase in rat cerebral cortical
neurons: a role in neuroprotection against N-methyl-D-aspartate re-
ceptor-mediated excitotoxicity. FEBS Lett 538:145–148.
Hennion JP, Adnan M, Mo UHF, Mallakh RS (2002) Evaluation of
neuroprotection by lithium and valproic acid against ouabain-
induced cell damage. Bipolar Disord 4:201–206.
Hong M, Chen DCR, Klein PS, Lee VMY (1997) Lithium reduces tau
phosphorylation by inhibition of glycogen synthase kinase-3. J Biol
Hongisto V, Smeds N, Brecht S, Herdegen T, Courtney MJ, Coffey ET
(2003) Lithium blocks the c-Jun stress response and protects
neurons via its action on glycogen synthase kinase 3. Mol Cell Biol
Jordà EG, Verdaguer E, Canudas AM, Jimenez A, Bruna A, Caelles C,
Bravo R, Escubedo E, Pubill D, Camarasa J, Pallas M, Camins A
(2003) Neuroprotective action of flavopiridol, a cyclin-dependent
kinase inhibitor, in colchicine-induced apoptosis. Neuropharmacol-
Kang HJ, Noh JS, Bae YS, Gwag BJ (2003) Calcium dependent
prevention of neuronal apoptosis by lithium ion: essential role of
phosphoinositide 3-kinase and phospholipase C?. Mol Pharmacol
Kaytor MD, Orr HT (2002) The GSK3 beta signaling cascade and
neurodegenerative disease. Curr Opin Neurobiol 12:275–278.
Knockaert M, Grengard P, Meijer L (2002) Pharmacological inhibitors
of cyclin-dependent kinases. Trends Pharmacol Sci 23:417–425.
Kroemer G, Zamzami N, Susin SA (1997) Mitochondrial control of
apoptosis. Immunol Today 18:44–51.
Lee MS, Tsai LH (2003) Cdk5: One of the links between senile plaques
and neurofibrillary tangles? J Alzheimers Dis 5:127–137.
Li R, El-Mallahk RS (2000) A novel evidence of different mechanisms
of lithium and valproate neuroprotective action on human SY5Y
neuroblastoma cells: caspase-3 dependency. Neurosci Lett 294:
Li X, Bijur GN, Jope RS (2002) Glycogen synthase kinase-3?, mood
stabilizers, and neuroprotection. Bipolar Disord 4:137–144.
Lim CB, Qi R (2003) Cyclin-dependent kinases in neuronal develop-
ment and degeneration. J Alzheimers Dis 5:329–335.
Linseman DA, Cornejo BJ, Le SS, Meintzer MK, Laessig TA, Bouchard
RJ, Heidenreich KA (2003) A myocyte enhancer factor 2D
(MEF2D) kinase activated during neuronal apoptosis is a novel
target inhibited by lithium. J Neurochem 85:1488–1499.
Lovestone S, Davis DR, Webster MT, Kaech S, Brion JP, Matus A,
Anderton BH (1999) Lithium reduces tau phosphorylation: effects
in living cells and in neurons at therapeutic concentrations. Biol
Mao Z, Bonni A, Xia F, Nadal-Vicens M, Greenberg ME (1999) Neu-
ronal activity-dependent cell survival mediated by transcription
factor MEF2. Science 286:785–790.
Mora A, Sabio G, González-Polo RA, Cuenda A, Alessi DR, Alonso
JC, Fuentes JM, Soler G, Centeno F (2001) Lithium inhibits
caspase 3 activation and dephosphorylation of PKB and GSK3
induced by K?deprivation in cerebellar granule cells. J Neurochem
Mora A, Sabio G, Soler G, Centeno F (2002) Different dependence of
lithium and valproate on PI3K/PKB pathway. Bipolar Disord 4:
Morfini G, Szebenyi G, Brown H, Pant HC, Pigino G, DeBoer S, Beffert
U, Brady ST (2004) A novel CDK5-dependent pathway for regu-
E. G. Jordà et al. / Neuroscience 134 (2005) 1001–10111010
lating GSK3 activity and kinesin-driven motility in neurons. EMBO Download full-text
Nguyen MD, Mushynski WE, Julien JP (2002) Cycling at the interface
between neurodevelopment and neurodegeneration. Cell Death
Nonaka S, Hough CJ, Chuang DM (1998a) Chronic lithium treatment
robustly protects in the central nervous system against excitotox-
icity by inhibiting N-methyl-D-aspartate receptor-mediated calcium
influx. Proc Natl Acad Sci U S A 95:2642–2647.
and carbamazepine. J Pharmacol Exp Ther 286:539–547.
Otth C, Concha II, Arendt T, Stieler J, Schliebs R, Gonzalez-Billault C,
Maccioni R (2002) A?PP induces cdk5-dependent tau hyperphos-
phorylation in transgenic mice Tg2576. J Alzheimers Dis
Pérez M, Hernández F, Lim F, Díaz-Nido J, Avila J (2003) Chronic
lithium treatment decreases mutant tau protein aggregation in a
transgenic mouse model. J Alzheimers Dis 5:301–308.
Sengupta A, Wu Q, Grundke-Iqbal I, Iqbal K, Singh TJ (1997) Poten-
tiation of GSK-3-catalyzed Alzheimer-like phosphorylation of hu-
man tau by cdk5. Mol Cell Biochem 167:99–105.
Tseng WP, Lin-Shiau SY (2002) Long-term lithium treatment prevents
neurotoxic effects of ?-bungarotoxin in primary cultured neurons.
J Neurosci Res 69:633–641.
Verdaguer E, Garcia-Jorda E, Canudas AM, Dominguez E, Jimenez A,
Pubill D, Escubedo E, Pallas JC, Camins A (2002) Kainic acid-
induced apoptosis in cerebellar granule neurons: an attempt at cell
cycle re-entry. Neuroreport 13:413–416.
Volbracht C, Leist M, Kolb SA, Nicotera P (2001) Apoptosis in
caspase-inhibited neurons. Mol Med 7:36–48.
Wei H, Qin ZH, Senatorov VV, Wei W, Wang Y, Qian Y, Chuang DM
(2001) Lithium suppresses excitotoxicity-induced striatal lesions in
a rat model of Huntington’s disease. Neuroscience 106:603–612.
Zambrano CA, Egana JT, Nunez MT, Maccioni RB, Gonzalez-Billault
C (2004) Oxidative stress promotes tau dephosphorylation in neu-
ronal cells: the roles of cdk5 and PP1. Free Radic Biol Med
(Accepted 23 April 2005)
(Available online 27 June 2005)
E. G. Jordà et al. / Neuroscience 134 (2005) 1001–10111011