Monitoring protein phosphatase 1 isoform levels as a marker for cellular stress

Neurotoxicology and Teratology 26 (2004) 387 – 395 www.elsevier.com/locate/neutera

Monitoring protein phosphatase 1 isoform levels as a marker for cellular stress Fa´tima Camo˜es Amador, Ana Gabriela Henriques, Odete A.B. da Cruz e Silva, Edgar F. da Cruz e Silva* Centro de Biologia Celular, Universidade de Aveiro, 3810-193 Aveiro, Portugal Received 9 October 2003; received in revised form 19 December 2003; accepted 22 December 2003

Abstract Reversible protein phosphorylation is a central mechanism regulating many biological functions, and abnormal protein phosphorylation can have a devastating impact on cellular control mechanisms, including a contributing role in neurodegenerative processes. Hence, many promising novel drug development strategies involve targeting protein phosphorylation systems. In this study, we demonstrate that various cellular stresses relevant to neurodegeneration can specifically affect the protein expression levels of protein phosphatase 1 (PP1). PP1 levels were altered upon exposure of PC12 and COS-1 cells to aluminium, Abeta peptides, sodium azide, and even heat shock. Particularly interesting, given PP1’s involvement in aging and neurodegeneration, was the consistent decrease in PP1g1 levels in response to stress agents. In fact, alterations in the expression levels of PP1 appear to correspond to an early response of stress induction, that is, before alterations in heat shock proteins can be detected. Our data suggest that monitoring PP1 isoform expression could constitute a useful diagnostic tool for cellular stress, possibly even neurodegeneration. Additionally, our results strengthen the rationale for signal transduction therapeutics and indicate that altering the specific activity of PP1 either directly or by targeting its regulatory proteins may be a useful therapeutic development strategy for the future. D 2004 Elsevier Inc. All rights reserved. Keywords: Alzheimer’s disease; Aluminium; Abeta; Oxidative stress; Phosphorylation

1. Introduction Synaptic transmission is a complex molecular process involving second messengers, protein phosphatases and protein kinases, with their various substrates, activators, and inhibitors [22]. Numerous signaling pathways have been described, where protein phosphorylation/dephosphorylation plays a central regulatory role, effectively modulating the function of several key proteins involved in synaptic transmission, including voltage- and ligandgated channels, neurotransmitter release, and neurotransmitter transporters [41]. Inasmuch as most protein phosphorylation processes occur on serine (Ser) and threonine (Thr) residues, the function and control of the Ser/Thrspecific protein phosphatases are particularly relevant.

* Corresponding author. Tel.: +351-234-370-983; fax: +351-234-426408. E-mail address: [email protected] (E.F. da Cruz e Silva). 0892-0362/$ – see front matter D 2004 Elsevier Inc. All rights reserved. doi:10.1016/j.ntt.2003.12.007

Several types of Ser/Thr phosphatases have been identified in eukaryotic cells [38], but one of the most widely expressed and highly regulated members of this family is protein phosphatase 1 (PP1) [10,12]. The three known mammalian PP1 catalytic subunit genes (PP1a, PP1h, and PP1g) are particularly abundantly expressed in mammalian brain [15]. Indeed, numerous studies have highlighted the importance of PP1 in the control of brain function. For example, Morishita et al. [32] demonstrated that adequate activation of NMDA receptors allows PP1 to gain access to synaptic substrates and be recruited to synapses where its activity is necessary to sustain long-term depression (LTD). Moreover, recent reports have linked PP1 to the efficacy of learning and memory by limiting the capacity of acquisition of new knowledge and favouring the observed memory decline with age [20]. The vital role played by PP1 in the control of LTD and synaptic plasticity [28,29,35] has been further validated by our previous observations on its subcellular localization. Both PP1a and PP1g1 were shown to be highly and specifically

388

F.C. Amador et al. / Neurotoxicology and Teratology 26 (2004) 387–395

concentrated in dendritic spines [36]. These results suggested that spines may have developed as specialized signal transduction organelles, enabling neurons to integrate the multitude of signals received and elicit an appropriate response. More recently, the highly specialized localization of PP1 was shown to result from the interaction of the relevant catalytic subunit with a regulatory protein responsible for tethering the catalytic subunit to the spine cytoarchitecture [1]. Abnormal protein phosphorylation has been associated with several neurodegenerative disorders, such as Alzheimer’s disease (AD), Parkinson’s disease, and Huntington’s disease. For example, in AD, tau protein (the major component of neurofibrillary tangles, a hallmark lesion observed in AD) is highly hyperphosphorylated at several Ser/Thr sites [6]. Several lines of evidence point to abnormal phosphatase activity being associated with tau hyperphosphorylation. Another neuropathological hallmark of AD is the extracellular deposits known as senile plaques. The major constituent of senile plaques is a toxic fragment termed Abeta, which is produced from the cleavage of the Alzheimer’s amyloid precursor protein (APP). Again, protein phosphorylation is not only known to stimulate APP processing [16]; more specifically, PP1 was shown to be involved in the control of this process [13,14]. The abovementioned neuropathological hallmarks of AD patients also show a characteristically high level of aluminium [26], present in the core of senile plaques [42] and in neurofibrillary tangles [37]. Hence, aluminium represents one of many factors contributing to neuropathologies, such as AD. Another commonly addressed factor is oxidative stress. The latter appears to be important on several levels, including the induction of abnormal APP processing and potentially increasing Abeta production [18,19]. Among the processes associated with AD, apoptosis [17,40,43], increased levels of heat shock proteins [44,45], and altered protein phosphorylation may be particularly relevant. Several studies have demonstrated that abnormal phosphorylation due to protein phosphatase inhibition may be involved in neurodegenerative diseases and associated processes. Thus, microinjection of okadaic acid (a potent and specific inhibitor of the Ser/Thr-specific phosphatases of the PP1/PP2A/PP2B family) into rat hippocampus induced neuronal stress and neurodegeneration [41]. The exquisite enrichment of PP1 in dendritic spines, associated with its postulated involvement in aging-related memory defects, coupled with many observations linking abnormal phosphorylation of key proteins with neurodegeneration, led us to investigate the effect of several stress conditions associated with AD on the expression of PP1. Given its localization at the brain’s sites of action (the dendritic spines), any factors impacting negatively on synaptic transmission and signaling, such as neurotoxic or stressinducing compounds, might be expected to affect PP1 levels.

2. Materials and methods 2.1. Cell culture A rat pheochromocytoma cell line (PC12) with neuronal-like properties and a nonneuronal monkey kidney cell line (COS-1) were used in the experiments described. PC12 cells were cultured in RPMI 1640 (Invitrogen) supplemented with 10% (v/v) heat-inactivated horse serum (Invitrogen), 5% (v/v) heat-inactivated foetal bovine serum (Invitrogen), 1% (v/v) antimycotic –antibiotic solution (Invitrogen), and 0.85 g/l NaHCO3. COS-1 cells were maintained in Dulbecco’s modified Eagle’s medium (Sigma) supplemented with 10% (v/v) heat-inactivated foetal bovine serum, 1% (v/v) antibiotic – antimycotic solution, and 3.7 g/l NaHCO3. Both cell lines were grown routinely in 100-mm tissue culture dishes and maintained at 37 jC in an atmosphere of 5% CO2. For experimental procedures, cells were collected, counted, and seeded onto 6-well plates. When using PC12 cells, the plates were previously treated with 100 Ag/ml poly-Lornithine (Sigma). 2.2. Experimental cell treatments To ascertain the effect of several types of stress related to AD on the levels of expressed PP1, both PC12 and COS-1 cells were exposed to increasing concentrations of aluminium or sodium azide, or to heat shock or Abeta, as described below. 2.2.1. Exposure to aluminium Several studies have implicated aluminium and other metals in the aetiology of AD, including the finding of high aluminium concentrations in senile plaques [42]. Thus, aluminium-induced stress was investigated for its effects on PP1 expression. PC12 and COS-1 cells were grown in 6-well plates and washed with serum-free and phosphatefree medium. Cells were then incubated for various times up to 48 h in medium supplemented with the indicated concentrations of aluminium (AlCl3.6H2O). Cell viability assays were carried out on both cell lines by the MTT method [33] at 24 and 48 h. This method is based on the reduction of MTT, a water soluble tetrazolium salt, by mitochondrial dehydrogenase, to an insoluble intracellular purple formazan. The extent of reduction of MTT was measured spectrophotometrically at 570 nm. Briefly, after treatment of the cells, the medium was removed and 0.5 mg/ml MTT (Sigma) solution (in serum-free DMEM or serum-free RPMI) was added and incubated for 3 h at 37 jC. The resulting insoluble formazan precipitates were solubilized with 0.04 M HCl/ Isopropanol. The absorbance of the converted dye was measured at 570 nm in a Cary 50 spectrophotometer (Varian). The cellular viability was expressed as a percentage of control cells.

F.C. Amador et al. / Neurotoxicology and Teratology 26 (2004) 387–395

389

induction of HSPs under stress conditions have been associated with AD. Thus, oxidative stress was induced by treating cells with sodium azide (a cytochrome c oxidase inhibitor) and the effect on PP1 expression determined. Briefly, cells were cultured as described above, grown for 2 h in medium containing 0 – 10 mM sodium azide and allowed to recover for up to 24 h. Samples were collected as described below, immediately after the 2 h exposure period (short-term effect) or after a long recovery period (long-term effect). The effect of increasing sodium azide concentration on cell viability was also assessed by the MTT reduction method described above.

2.2.2. Exposure to Abeta Given the importance of Abeta in AD and other neurodegenerative disorders, cells were also exposed to this peptide. Abeta 1– 40 and Abeta 25 –35 were obtained from Sigma, prepared as 1 mM stock solutions in water, and cells were incubated with 10 –20 AM of each peptide in serum-free medium for 24 h. PP1 expression was compared to control cells incubated in the absence of the peptides. 2.2.3. Exposure to heat shock and sodium azide Other stress factors may be relevant to neurodegenerative conditions. Particularly, oxidative stress and the

A – Expression in PC12 cells

(hr) 0

12

24

36

1 mM AlCl3

0.5 mM AlCl3

0.1 mM AlCl3

-

48

12

24

36

48

12

24

36

48

β-tubulin PP1α PP1γ1

B – PP1α α expression 120

***

% of control

100

**

***

***

** ***

80

**

***

12 hr 24 hr

60

36 hr 40 48 hr 20 0 0.1 mM

1 mM

0.5 mM

C –PP1γ1 expression 120

% of control

100

*

**

80

***

*** *** ***

60

12 hr

***

24 hr

*** ***

36 hr

40 48 hr 20 0 0.1 mM

0.5 mM

1 mM

Fig. 1. Effect of aluminium on PP1 isoform expression levels in PC12 cells. PC12 cells were incubated with increasing concentrations of aluminium for the indicated times. The expression levels of PP1a and PP1g1 were analyzed by immunoblotting with isoform-specific antibodies using duplicate membranes. hTubulin levels were also assessed as a control (A). Results obtained from at least three independent experiments were quantified and are represented graphically for PP1a (B) and PP1g1 (C). Values shown represent percentage of control and are expressed as mean F S.E.M. Values statistically different from control are indicated: P < 0.05 (*), P < 0.01 (**) and P < 0.001 (***), respectively.

390

F.C. Amador et al. / Neurotoxicology and Teratology 26 (2004) 387–395

Substantial induction of heat shock proteins (like HSP70) was achieved by transiently incubating cultures for 90 min at 44 jC and allowing them to recover for up to 24 h. 2.3. Sample collection and immunodetection After each treatment, the conditioned medium was removed and the cells were washed with phosphate-buffered saline. Samples were collected in boiling 1% (v/v) SDS and sonicated for 30 s. Protein determination was performed using the BCA protein assay [39], and normalized protein samples were electrophoretically separated by 12% SDS – PAGE [27]. Separated proteins were electrotransferred onto a nitrocellulose membrane [8], and specific proteins were identified by immunodetection.

For immunodetection, the membrane was saturated in 5% (v/v) nonfat dry milk in TBS-T (50 mM Tris, 150 mM NaCl, 0.05% Tween 20, pH 8.0) for 1 h [24]. PP1a and PP1g1 isoforms were detected using isoform-specific antipeptide rabbit polyclonal antibodies prepared exactly as described in da Cruz e Silva et al. [15]. Incubation of the membranes with the primary antibody was carried out for 2 h at room temperature. Detection was achieved using a horseradish peroxidase-conjugated antirabbit IgG secondary antibody (Pierce). The blots were then washed extensively, and PP1 isoforms were visualised on X-ray film using the ECL chemiluminescent method (Amersham). HSP70 and h-tubulin were detected in a similar manner using specific antibodies (obtained from Stressgen and Zymed, respectively) and detected using an alkaline phos-

Fig. 2. Effect of aluminium on PP1 protein expression levels in COS-1 cells. COS-1 cells were treated and analysed as described in the legend to Fig. 1. Duplicate immunoblots were prepared for PP1a and PP1g1 analysis. The same blots were subsequently used to measure h-tubulin as a control (A). Panels (B) and (C) represent the quantification of at least three independent experiments for PP1a (B) and PP1g1 (C). Values statistically different from control are indicated: P < 0.05 (*), P < 0.01 (**) and P < 0.001 (***), respectively.

F.C. Amador et al. / Neurotoxicology and Teratology 26 (2004) 387–395

391

phatase-labelled secondary antibody in conjunction with a corresponding colorimetric detection method (NBT/BCIP).

decreased to 57 F 8.1% for 0.1 mM AlCl3, 43.9 F 3.9% for 0.5 mM AlCl3, and 41.3 F 4.2% for 1 mM AlCl3.

2.4. Quantification and data analysis

3.1.1. PP1a expression The effect of AlCl3 (0.1, 0.5, or 1 mM) on PP1a expression was evaluated by immunoblot analysis in both cell lines as a function of time of exposure (12, 24, 36, and 48 h). Data is shown as percentage of control (the level of expression obtained for cells incubated without aluminium, at each time, were taken as 100%). In general, for both cell lines, a decrease in PP1a expression was observed with increasing aluminium concentrations and with increasing time of exposure at each concentration (Figs. 1A and 2A). Statistically significant decreases in PP1a were obtained at shorter incubation times as the aluminium concentration increased (Figs. 1B and 2B).

Quantitative analysis of the autoradiographic film exposures was carried out using a GS-710 calibrated imaging densitometer (BioRad). Data are expressed as mean F S.E.M. of triplicate determinations, from at least three independent experiments. Statistical significance analysis was conducted by one-way analysis of variance (ANOVA) with Student’s t test. Unless otherwise noted, a level of statistical significance is considered P < 0.05 versus control.

3. Results 3.1. Effect of aluminium on PP1 expression Prior to monitoring the effects on protein expression levels, cell viability was evaluated using the MTT assay. No significant effect was observed for all aluminium concentrations following a 12 h exposure, whereas statistically significant decreases in cellular viability were detected after 24 and 48 h of exposure (data not shown). In PC12 cells, a 20 –25% decrease in viability was measured at 24 h for all aluminium concentrations. After 48 h, cellular viability

3.1.2. PP1g1 expression The effect of AlCl3 (0.1, 0.5, or 1 mM) on PP1g1 protein expression levels was similarly evaluated with time (12, 24, 36, and 48 h), in both PC12 and COS-1 cell lines. Again, a general trend was observed regarding decreasing PP1g1 expression with time and with increasing aluminium concentration (Figs. 1A and 2A). Similar to the results obtained for PP1a, statistically significant decreases in PP1g1 expression were obtained in PC12 cells only after 36 and 48 h of incubation with 0.1 mM AlCl3 ( P < 0.05 and P < 0.01, respectively). In contrast, 0.5 mM AlCl3 yielded significant

A – Abeta 1 - 40 PC12

Abeta (µM)

0

10

COS-1 20

0

10

20

β-tubulin PP1α PP1γ1 HSP70

B – Abeta 25 - 35 PC12

Abeta (µ µM)

0

10

COS-1 20

0

10

20

β-tubulin PP1α PP1γ1 HSP70

Fig. 3. Effect of Abeta peptides on PP1 protein expression levels. PC12 and COS-1 cells were incubated with 10 and 20 AM Abeta 1 – 40 (A) or Abeta 25 – 35 (B) during 24 h in serum-free medium. Duplicate immunoblots were prepared and used to assess PP1a and PP1g1 levels. The same membranes were reused to monitor HSP70 and h-tubulin levels.

392

F.C. Amador et al. / Neurotoxicology and Teratology 26 (2004) 387–395

decreases earlier, from 24 h onwards, and 1 mM AlCl3 produced significant decreases at all time points sampled (Fig. 1C). Noteworthy, in COS-1 cells, 0.1 mM AlCl3 produced a slight but statistically significant increase in PP1g1 expression, whereas higher aluminium concentrations decreased its levels. This biphasic pattern of PP1g1 expression may reflect a transient stress response in this cell line. 3.1.3. HSP70 and h-tubulin expression HSP70 expression levels were also monitored under the same experimental conditions in both cell lines. Whereas a small decrease in HSP70 was detected in PC12 cells with increasing aluminium concentration, no changes could be measured in COS-1 cells (data not shown). h-tubulin expression was used as a control marker utilizing the immunoblots previously used for assessing PP1 expression. No differences were detected in h-tubulin expression under the conditions tested, in both cell lines (Figs. 1A and 2A). 3.2. Effect of Abeta on PP1 expression Two different potentially toxic Abeta peptides were tested on both PC12 and COS-1 cells. Their effect on PP1 expression levels was dependent on the peptide and the cell line used (Fig. 3). Abeta 1 –40, presumably the less toxic peptide, yielded small but opposing effects on the levels of PP1a and PP1g1, particularly with PC12 cells (Fig. 3A). Exposure to Abeta 25 – 35, the more toxic peptide, produced similar results to those obtained with the larger peptide. In both cell lines, expression levels of PP1g1 decreased (Fig. 3B). However, PP1a expression increased in PC12 cells but decreased slightly in COS-1 cells. h-Tubulin and HSP70 expression levels remained unaltered under our experimental conditions (Fig. 3). Interestingly, in contrast to PP1, although HSP70 has been used as a molecular marker for stress, it did not change in response to Abeta.

3.3. Effect of heat shock on PP1 expression We went on to study the effect of heat shock and the consequent induction of HSPs on PP1 levels. Cells were subjected to a brief period of growth under increased temperature conditions and then allowed to recover for up to 24 h. As expected, this treatment produced a robust increase in HSP70 expression levels for both cell lines (Fig. 4). Notably, in contrast, heat shock produced a highly significant decrease in the expression levels of both PP1 isoforms, in both PC12 and COS-1 cells (Fig. 4). For example, a decrease of approximately 50% was measured in PC12 cells for the levels of PP1g1. 3.4. Effect of oxidative stress on PP1 expression We also included sodium azide in our analysis, which is typically used as an inducer of oxidative stress, because oxidative stress has been associated with AD and other neurodegenerative disorders. Hence, PC12 and COS-1 cells were incubated with 10 mM NaN3 and allowed to recover for up to 24 h (Fig. 5). Overall, PC12 cells appeared to be more sensitive to sodium azide exposure than COS-1 cells. Thus, the levels of PP1a, PP1g1, and HSP70 all decreased markedly (up to 40% of control) in PC12 cells following sodium azide exposure (Fig. 5A). In contrast, sodium azide produced no effect on the levels of HSP70 in COS-1 cells, and only moderate decreases for PP1a and PP1g1 (Fig. 5A). h-tubulin expression levels did not alter with sodium azide (data not shown). Experiments with short term exposure to sodium azide reinforced the potential biphasic response previously observed for PP1 expression levels upon exposure to cellular stress. Thus, in contrast to the long recovery periods described above, the expression of PP1g1 increased with increasing sodium azide concentration during the 2-h exposure period (Fig. 5B), reaching a maximum of approxi-

Fig. 4. Effect of heat shock on PP1 protein expression levels. PC12 and COS-1 cells were incubated at 44 jC for 1.5 h and allowed to recover for 18 and 24 h in serum-free medium. Cell lysates were then collected and used to prepare duplicate immunoblots to analyse PP1a and PP1g1 expression. The same immunoblots were used to assess HSP70 and h-tubulin expression.

F.C. Amador et al. / Neurotoxicology and Teratology 26 (2004) 387–395

393

Fig. 5. Effect of oxidative stress on PP1 protein expression levels. PC12 and COS-1 cells were incubated in the presence of 10 mM sodium azide for 2 h, and the effect on PP1a and PP1g1 expression was analyzed following recovery periods of 18 and 24 h. The same membranes were used to monitor HSP70 and htubulin expression (A). Short-term effects of increasing sodium azide concentration (0 – 10 mM) on PP1g1 were assessed immediately after the 2 h incubation period with 50 mM 2-deoxyglucose (B).

mately 130% and 140% for PC12 and COS-1 cells, respectively.

4. Discussion The neurotoxicology of aluminium has been shown to mimic many of the pathophysiological features of AD. Indeed, the accumulation of aluminium in the brain may contribute to the cholinergic deficiency observed in AD patients [7]. By potentiating lipid peroxidation, aluminium affects the uptake of choline in nerve terminals [3], potentially contributing to the cholinergic dysfunction and neuronal cell degeneration known to occur in AD. In our experimental model, aluminium caused a decrease in PP1 expression levels. Although the precise molecular link between aluminium toxicity and decreased PP1 expression remains to be elucidated, several lines of evidence provide interesting clues. Chronic aluminium exposure impairs long-term potentiation and depression in the rat dentate gyrus in vivo, potentially suggesting that aluminium affects both presynaptic and postsynaptic mechanisms of synaptic transmission [11]. Interestingly, PP1 is not only highly enriched in dendritic spines [36] and necessary for maintaining LTD [32], it was also linked to age-related memory and learning deficits [20]. Thus, although more experimentation is required, it is tempting to hypothesize that the toxic effect of aluminium may be mediated via its effect on PP1. Therefore, at least in our experimental models, PP1 levels can be used to monitor aluminium toxicity, thus indicating

the potential use of PP1 as a diagnostic marker for cellular stress and neurodegeneration. Aluminium is known to potentiate oxidative stress induced by iron and to promote inflammatory events, although this process is not dependent on reactive oxygen species production induced by transition metals [5]. Inasmuch as aluminium accumulation was significantly increased in the nerve endings, in the presence of an oxidizing system, a possible role was proposed for aluminium in the promotion and enhancement of oxidant-induced damage believed to occur in neuronal degeneration [2]. Experimentally, we simulated oxidative stress by adding sodium azide (a cytochrome c oxidase inhibitor) and determining the effect of short (2 h) and long (18 – 24 h) exposures. Although sodium azide is likely to have other nonspecific effects, it provoked a decrease in PP1 expression, as was observed with the other stress-inducing agents. Taken together, these observations suggest that PP1 may be a sensitive ‘sensor’ of cellular stress, and its expression levels may be used as a monitoring and diagnostic tool. Abeta peptides are commonly used to induce neurotoxicity and to delineate similarities to AD in several in vitro models [9,25,30]. Results from various laboratories indicate that Abeta 25– 35 appears to be more toxic than Abeta 1– 40, although the latter may be pathologically more relevant. Our results showed that exposure to Abeta 1 – 40 (Fig. 3A) and Abeta 25– 35 (Fig. 3B) only affected PP1a and PP1g1 expression slightly. However, some interesting trends were observed. For example, PP1g1 expression decreased in both cell lines and with both peptides, similar to the aluminium-

394

F.C. Amador et al. / Neurotoxicology and Teratology 26 (2004) 387–395

induced effects. Under the conditions used, the PC12 cell line (with neuronal characteristics) proved to be more sensitive to Abeta exposure, as measured by the effect on PP1g1 levels (Fig. 3). Interestingly, the Abeta 25 – 35 peptide appeared to produce more robust effects than the full Abeta 1 – 40 peptide, consistent with the postulated higher toxic potential of the former. However, PP1a levels did not parallel the PP1g1 response. In general, PP1a levels tended to increase slightly upon exposure to the toxic peptides (Fig. 3). Hence, just as with aluminium, exposure to Abeta induced decreased PP1g1 expression. Previous experiments [21] indicated that Abeta 1 –42 and aluminium induce stress in the endoplasmic reticulum in rabbit hippocampus. It would appear that both agents induce nuclear translocation of gadd153 and of the inducible transcription factor NF-nB, which in turn cause a decrease in the levels of Bcl-2 in the endoplasmic reticulum and nucleus. NF-nB is also known to be induced under oxidative stress conditions. It is possible then that aluminium, Abeta, and even oxidative stress share aspects of cellular neurotoxicity which can lead to apoptosis. This is particularly significant because not only is Abeta the main component of senile plaques, but they have also been reported to accumulate high levels of aluminium, and furthermore, apoptosis has been associated with AD. Interestingly, although altered HSP expression has been associated with AD, Abeta-induced stress did not result in changes in HSP70 expression levels under the experimental conditions used (Fig. 3). HSP expression may be induced by chronic exposure to increased Abeta levels, as would be expected to occur in AD patients. As expected, HSP70 was robustly induced by briefly exposing cells to elevated temperatures (Fig. 4). Under these conditions both PP1a and PP1g1 expression levels were clearly reduced (Fig. 4). Thus, considering the data obtained with Abeta and heat-shock-induced stress, it appears that whereas HSP70 expression is only induced under conditions of extreme stress (as exemplified by heat shock), effects on PP1 expression levels can be detected even in response to mild/moderate stress. Exposure to sodium azide (another robust stress agent) also resulted in reduced PP1 expression (Fig. 5A). Again, PC12 cells appeared to be the more sensitive of the two cell lines. However, contrary to what might be expected, brief exposure to sodium azide (2 h) resulted in a dose-dependent increase in PP1g1 expression (Fig. 5B). Thus, it appears that PP1 levels may respond in a biphasic manner to cellular stress. In our model, brief and mild stress result in increased PP1 levels (as exemplified by the 2 h sodium azide treatment), whereas chronic and more aggressive stress agents (as exemplified by aluminium, Abeta 25 – 35, and chronic sodium azide treatment) decrease PP1 expression. Hence, PP1 expression levels may provide a highly sensitive molecular marker for neurodegenerative stress conditions. The mechanism by which different stressors decrease PP1 levels was not addressed but

warrants future experimentation. Two possibilities are the stimulation of PP1 degradation and the inhibition of its synthesis. Checking the mRNA levels of PP1a and PP1g1 should yield interesting insights into this question. In conclusion, altered PP1 expression could be detected under all cellular stress conditions tested. Curiously, reduced PP1 activity has been implicated in tau hyperphosphorylation [4,23,31], which can promote AD pathology and neurodegeneration. Although further experimentation is needed to validate our results in clinically relevant samples, it appears that PP1 may exhibit a typical biphasic response, increasing in expression under mild stress conditions and decreasing under prolonged stress conditions. In fact, decreased PP1a and PP1g1 mRNA expression was recently reported in samples from AD patients [34]. Our data supports these findings and demonstrates that alterations should also be detected at the protein level. The response is similar in different cell lines and upon exposure to different stresses: aluminium, Abeta, oxidative stress, and heat shock. Clearly, our observations need to be validated further in animal models. Taken together, the data presented here suggest that PP1g1 may prove to be particularly useful as an early diagnostic molecular marker for conditions that result in mild neurodegenerative stress. Furthermore, PP1 and its regulatory proteins, which can regulate or mediate PP1 functions, represent interesting potential targets for signal transduction therapeutics.

Acknowledgements This work was supported by grants from the Fundacß a˜o para a Cieˆncia e Tecnologia of the Portuguese Ministry for Science and Higher Education to EFCS (POCTI/CBO/ 39799/2001) and OABCS (POCTI/NSE/33520/99 and BCI/ 34349/99), and from the V Framework Program of the European Union to OABCS (QLK3-CT-2001-02362), and by the Centro de Biologia Celular of the Universidade de Aveiro.

References [1] P.B. Allen, C.C. Ouimet, P. Greengard, Spinophilin, a novel protein phosphatase 1 binding protein localized to dendritic spines, Proc. Natl. Acad. Sci. U. S. A. 94 (1997) 9956 – 9961. [2] F.C. Amador, M.S. Santos, C.R. Oliveira, Lipid peroxidation facilitates aluminium accumulation in rat brain synaptosomes, J. Toxicol. Environ. Health 58 (1999) 427 – 435. [3] F.C. Amador, M.S. Santos, C.R. Oliveira, Lipid peroxidation and aluminium effects on the cholinergic system in nerve terminals, Neurotox. Res. 3 (2001) 223 – 233. [4] L. Baum, R. Seger, J.R. Woodgett, S. Kawabata, K. Maruyama, M. Koyama, J. Silver, T. Saitoh, Overexpressed tau protein in cultured cells is phosphorylated without formation of PHF: implication of phosphoprotein phosphatase involvement, Brain Res. Mol. Brain Res. 34 (1995) 1 – 17. [5] A. Becaria, S.C. Bondy, A. Campbell, Aluminum and copper inter-

F.C. Amador et al. / Neurotoxicology and Teratology 26 (2004) 387–395

[6]

[7]

[8]

[9] [10]

[11]

[12] [13] [14]

[15]

[16]

[17]

[18]

[19]

[20]

[21]

[22] [23]

[24]

[25]

act in the promotion of oxidative but not inflammatory events: implications for Alzheimer’s disease, J. Alzheimer’s Dis. 5 (2003) 31 – 38. M. Bennecib, C.X. Gong, I. Grundke-Iqbal, K. Iqbal, Role of protein phosphatase-2A and 1- in the regulation of GSK-3, cdk5 and cdc2 and the phosphorylation of tau in rat forebrain, FEBS Lett. 485 (2000) 87 – 93. H. Bielarczyk, M. Tomaszewicz, A. Szutowicz, Effect of aluminium on acetyl-CoA and acetylcholine metabolism in nerve terminals, J. Neurochem. 70 (1998) 1175 – 1181. W.N. Burnette, ‘‘Western blotting’’: electrophoretic transfer of proteins from sodium dodecyl sulfate – polyacrylamide gels to unmodified nitrocellulose and radiographic detection with antibody and radioiodinated protein A, Anal. Biochem. 112 (1981) 195 – 203. J. Carter, C.F. Lippa, Beta-amyloid, neuronal death and Alzheimer’s disease, Curr. Mol. Med. 1 (2001) 733 – 777. H. Ceulemans, W. Stalmans, M. Bollen, Regulator-driven functional diversification of protein phosphatase-1 in eukaryotic evolution, Bioessays 24 (2002) 371 – 381. J. Chen, M. Wang, D. Ruan, J. She, Early chronic aluminium exposure impairs long-term potentiation and depression to the rat dentate gyrus in vivo, Neuroscience 112 (2002) 879 – 887. P.T.W. Cohen, Protein phosphatase 1—targeted in many directions, J. Cell Sci. 115 (2002) 241 – 256. E.F. da Cruz e Silva, O.A.B. da Cruz e Silva, Protein phosphorylation and APP metabolism, Neurochem. Res. 28 (2003) 1553 – 1561. E.F. da Cruz e Silva, O.A.B. da Cruz e Silva, C.T.B.V. Zaia, P. Greengard, Inhibition of protein phosphatase 1 stimulates secretion of Alzheimer amyloid precursor protein, Mol. Med. 1 (1995) 535 – 541. E.F. da Cruz e Silva, C.A. Fox, C.C. Ouimet, E. Gustafson, S.J. Watson, P. Greengard, Differential expression of protein phosphatase 1 isoforms in mammalian brain, J. Neurosci. 15 (1995) 3375 – 3389. O.A.B. da Cruz e Silva, K. Iverfeldt, T. Oltersdorf, S. Sinha, I. Lieberburg, T.V. Ramabhadran, T. Suzuki, S.S. Sisodia, S. Gandy, P. Greengard, Regulated cleavage of Alzheimer beta-amyloid precursor protein in the absence of the cytoplasmic tail, Neuroscience 57 (1993) 873 – 874. M. Dragunow, R.L. Faull, P. Lawlor, E.J. Beilharz, K. Singleton, E.B. Walker, E. Mee, In situ evidence for DNA fragmentation in Huntington’s disease striatum and Alzheimer’s disease temporal lobes, NeuroReport 6 (1995) 1053 – 1057. D. Gabuzda, J. Busciglio, L.B. Chen, P. Matsudaira, B.A. Yankner, Inhibition of energy metabolism alters the processing of amyloid precursor protein and induces a potentially amyloidogenic derivative, J. Biol. Chem. 269 (1994) 13623 – 13628. L. Gasparini, M. Racchi, L. Benussi, D. Curti, G. Binetti, A. Bianchetti, M. Trabucchi, S. Govoni, Effect of energy shortage and oxidative stress on amyloid precursor protein metabolism in COS cells, Neurosci. Lett. 231 (1997) 113 – 117. D. Genoux, U. Haditsch, M. Knobloch, A. Michalon, D. Storm, I.M. Mansuy, Protein phosphatase 1 is a molecular constraint on learning and memory, Nature 418 (2002) 970 – 975. O. Ghribi, M.M. Herman, D.A. DeWitt, M.S. Forbes, J. Savory, Abeta(1 – 42) and aluminum induce stress in the endoplasmic reticulum in rabbit hippocampus, involving nuclear translocation of gadd 153 and NF-kappaB, Brain Res. Mol. Brain Res. 30 (2001) 30 – 38. P. Greengard, The neurobiology of slow synaptic transmission, Science 294 (2001) 1024 – 1030. K. Iqbal, A.D. Alonso, J.A. Gondal, C.X. Gong, N. Haque, S. Khatoon, A. Sengupta, J.Z. Wang, I. Grundke-Iqbal, Mechanism of neurofibrillary degeneration and pharmacologic therapeutic approach, J. Neural Transm., Suppl. 59 (2000) 213 – 222. D.A. Johnson, J.W. Gautsch, J.R. Sportsman, J.H. Elder, Improved technique utilizing nonfat dry milk for analysis of proteins and nucleic acids transferred to nitrocellulose, Gene Anal. Tech. 1 (1984) 3. P. Kienlen-Campard, S. Miolet, B. Tasiaux, J.N. Octave, Intracellular

[26]

[27] [28]

[29] [30]

[31]

[32]

[33]

[34]

[35]

[36]

[37]

[38] [39]

[40]

[41]

[42]

[43]

[44]

[45]

395

amyloid-beta 1 – 42, but not extracellular soluble amyloid-beta peptides, induces neuronal apoptosis, J. Biol. Chem. 277 (2002) 15666 – 15670. T.P.A. Kruck, D.R. McLachlan, Mechanisms of aluminium neurotoxicity—relevance to human disease, in: H. Sigel, A. Siegel (Eds.), Metal ions in Biological Systems, Marcel Dekker, New York, 1988, pp. 285 – 314. U.K. Laemmli, Cleavage of structural proteins during the assembly of the head of bacteriophage T4, Nature 227 (1970) 680 – 685. J. Lisman, A mechanism for the Hebb and the anti-Hebb processes underlying learning and memory, Proc. Natl. Acad. Sci. U. S. A. 86 (1989) 9574 – 9578. R.C. Malenka, Synaptic plasticity in the hippocampus, Cell 78 (1994) 535 – 538. M.P. Mattson, J. Partin, J.G. Begley, Amyloid beta-peptide induces apoptosis-related events in synapses and dendrites, Brain Res. 807 (1998) 167 – 176. S.E. Merrick, J.Q. Trojanowski, V.M. Lee, Selective destruction of stable microtubules and axons by inhibitors of protein serine/threonine phosphatases in cultured human neurons, J. Neurosci. 17 (1997) 5726 – 5737. W. Morishita, J.H. Connor, H. Xia, E.M. Quilan, S. Shenolikar, R.C. Malenka, Regulation of synaptic strength by protein phosphatase 1, Neuron 32 (2001) 1133 – 1148. T. Mosmann, Rapid colorimetric assay for cellular growth and survival: application to proliferation and cytotoxicity assays, J. Immunol. Methods 65 (1983) 55 – 63. E.J. Mufson, S.E. Counts, S.D. Ginsberg, Gene expression profiles of cholinergic nucleus basalis neurons in Alzheimer’s disease, Neurochem. Res. 27 (2002) 1035 – 1048. R.M. Mulkey, S. Endo, S. Shenolikar, R. Malenka, Involvement of a calcineurin/inhibitor-1 phosphatase cascade in hippocampal long-term depression, Nature 369 (1994) 486 – 488. C.C. Ouimet, E.F. da Cruz e Silva, P. Greengard, The a and g1 isoforms of protein phosphatase 1 are highly and specifically concentrated in dendritic spines, Proc. Natl. Acad. Sci. U. S. A. 92 (1995) 3396 – 3400. D.P. Perl, Aluminum and Alzheimer’s disease: methodologic approaches, in: H. Sigel, A. Siegel (Eds.), Metal Ions in Biological Systems, Marcel Dekker, New York, 1988, pp. 259 – 283. S. Shenolikar, Protein serine/threonine phosphatases—new avenues for cell regulation, Annu. Rev. Cell Biol. 10 (1994) 55 – 86. P.K. Smith, R.I. Krohn, G.T. Hermanson, A.K. Mallia, F.H. Gartner, M.D. Provenzano, E.K. Fujimoto, N.M. Goeke, B.J. Olson, D.C. Klenk, Measurement of protein using bicinchoninic acid, Anal. Biochem. 150 (1985) 76 – 85. J.H. Su, A.J. Anderson, B.J. Cummings, C.W. Cotman, Immunohistochemical evidence for apoptosis in Alzheimer’s disease, NeuroReport 5 (1994) 2529 – 2533. ´ rias, Neurotoxic and synaptic effects of okadaic R. Tapia, F. Pena, C. A acid, an inhibitor of protein phosphatases, Neurochem. Res. 24 (1999) 1423 – 1430. S. Tokutake, H. Nagase, S. Morisaki, S. Oyanagi, Aluminium detected in senile plaques and neurofibrillary tangles is contained in lipofuscin granules with silicon, probably as aluminosilicate, Neurosci. Lett. 185 (1995) 99 – 102. J.C. Troncoso, R.R. Sukhov, C.H. Kawas, V.E. Koliatsos, In situ labeling of dying cortical neurons in normal aging and in Alzheimer’s disease: correlations with senile plaques and disease progression, J. Neuropathol. Exp. Neurol. 55 (1996) 1134 – 1142. B.C. Yoo, S.H. Kim, N. Cairns, M. Fountoulakis, G. Lubec, Deranged expression of molecular chaperones in brains of patients with Alzheimer’s disease, Biochem. Biophys. Res. Commun. 280 (2001) 249 – 258. B.C. Yoo, R. Seidl, N. Cairns, G. Lubec, Heat – shock protein 70 levels in brain of patients with Down syndrome and Alzheimer’s disease, J. Neural Transm., Suppl. 57 (1999) 315 – 322.