Down-regulation of p21-activated serine/threonine kinase 1 is involved in loss of mesencephalic dopamine neurons

The p21-activated serine/threonine kinase 1 (PAK1) has a role in a variety of cellular functions like cell motility, cell cycle, survival, and even death. In cancer cells, PAK has functional roles in growth, invasion, and metastasis [1‐5]. Accordingly, the activity of PAK1 is up-regulated through protein overexpression or gene amplification in cancer tissues [4, 5]. Although the roles of PAK1 have been widely studied in a variety of cancers and cancer cells, other roles of PAK1 still remain to be elucidated [6].

Tumor formation and progression requires changes in cellular events such as the expression of oncogenes, activation of anti-apoptotic signaling pathways, and simultaneous inactivation of pro-apoptotic signals as well as of tumor suppressors. Apoptosis (type I cell death), also known as programmed cell death, is a finely regulated mechanism initiated by signaling in response to a variety of stresses cells. Thus, it has important roles in development and tissue homeostasis in multicellular organisms. The susceptibility to apoptotic cell death is dependent on the relative level of anti-apoptotic proteins of Bcl-2 family members (Bcl-2, Bcl-x_L, Mcl-1) to that of pro-apoptotic proteins of the family (Bad, Bax, Bak, Bok) or the interaction of each protein. Anti-apoptotic protein Bcl-2, which acts as a proto-oncogene, is a member of protective complexes that help to maintain the integrity of mitochondrial membrane. Pro-apoptotic proteins, such as Bad and Bax, can bind to Bcl-2. This binding inhibits the function of Bcl-2 and induces the release of pro-apoptotic factors by neutralizing the effects of pro-apoptotic proteins, and cell death eventually [7‐9]. PAK1 affects the activity of pro-apoptotic signals in a Raf-1 dependent or independent manner. Raf-1 activated by PAK1 is translocalized to the mitochondria, and then it phosphorylates Bad, which is a pro-apoptotic protein [10]. Alternatively, PAK1 directly phosphorylates Bad on Ser112 and Ser136, inducing inhibition of the interaction with Bcl-2 or Bcl-x_L and subsequently resulting in cell survival [11].

In the nervous system, dysregulation of PAK signaling is implicated in neurodegenerative diseases, such as Alzheimer’s disease (AD) and Huntington’s disease (HD), and mental retardation [12]. Down-regulation of PAK1 and PAK3, brain specific forms, has been reported in AD brain [13]. Also, PAK1 overexpression enhanced the toxicity in a HD cellular model [14]. Rac-PAK signaling was found to be defective in the mouse model of Fragile X syndrome, Fmr1-KO mouse, suggesting that PAK1 might also be involved in the pathogenesis of mental retardation [15]. In addition, knockout of PAK1 induced defects in brain development [16]. However, up to now, molecular details involved in Parkinson’s disease (PD) are limited.

It has been widely thought that cancer or neurodegenerative disease may originate from opposite fate of cells, to death or persistent growth, respectively. As a matter of fact, some neurodegenerative diseases, such as PD, have an evident inverse co-morbidity with many cancers [17]. A certain protein, known as a genetic determinant, may decide directions of the disease to either cancers or neurodegenerative diseases according to the cellular condition in which the determinant is activated or inactivated, or the cell type, whether or not the cells have the ability to proliferate in response to the same external stimuli that can induce DNA damage or cell cycle activation [18]. Even importance of genetic determinants or molecular determinants in the pathogenesis of such diseases, not much research has been performed.

The contradictory roles of PAK1 in cancers and AD and in cell survival and cell death suggest the possibility of PAK as a new molecular determinant whose activity could determine cell fate. Hence, we examined the role of PAK1 in neuronal death. Here, we report novel molecular mechanisms involved in the death of mesencephalic DA neurons mediated by PAK1 and suggest a new molecular target for a therapeutic agent in PD.

Since a very long time, there have been many discussions on the inverse comorbidity between some cancers and neurodegenerative diseases [17]. A genetic determinant can decide the fate of cells, depending on the situations or the characteristic of cells [18]. Previous studies have described that overactivation or overexpression of PAK1 is strongly correlated with cancers [4, 5]. In our study, down-regulation of PAK1 activity, and not amount, was correlated with cell death and that occurs at the initial stage of cell death induced by oxidative stress, such as 6-OHDA treatment (Fig. 1, 2). These results suggest that PAK1 might function as a molecular determinant that may decide the direction for development of neurodegenerative diseases, such as Parkinson’s disease (Fig. 7).

It has been widely accepted that apoptosis (type I cell death) is the major cause of neuronal loss in PD [31]. Some studies, using postmortem brain sections of PD patients, have reported that loss of the nigral neurons is due to apoptosis induced by an increase in Bax (a proapoptotic member of the Bcl-2 family)-dependent mitochondrial membrane permeability [31‐33]. If cell fate is determined by the ratio of pro-apoptotic to anti-apoptotic proteins, a reduction in Bcl-2 could increase Bax-triggered apoptosis. Of course, some part of this apoptotic cell death might originate from dephosphorylation of Bad induced by inactivation of PAK signaling [10, 11]. However, if we consider that the expression of Bax in the melanized nigral neurons is more related with PD brains [32, 33], Bax-triggered apoptosis induced by the reduction in Bcl-2 could be a major pathway. Even in our conclusion, the possibility that two pathways act synergistically or simultaneously, cannot be excluded. Thus, in order to clarify this, more studies are needed.

Unwanted proteins should be cleared by cellular clearance systems. The failure of clearance system could increase the number of misfolded and aggregated proteins. Thus, the ubiquitin-proteasome system is closely linked with the development of PD. Indeed, mutations of genes involved in the ubiquitin-proteasome system are found in several types of familial PD and proteasomal enzymatic activity is reduced in sporadic PD [34]. In our study (Fig. 2, 3) and previous reports [22, 23], oxidative stress regulated PAK1 activity. Subsequently, PAK1 could regulate the level of Bcl-2 protein through the ubiquitin-proteasome system by modulating its phosphorylation status through the extracellular signal-regulated kinase (ERK) pathway [22, 23]. However, these mechanisms only work when ubiquitin-proteasome system is intact or partially working. Protein aggregates, known as Lewy bodies, in nigral neurons have toxic effects. Production of protein aggregates might be due to an imbalance between the amount of unwanted proteins and the capacity to remove them. Indeed, the reduction of removal capacity is correlated with aging [35, 36].

Extracellular signals that lead to the activation or inactivation of PAK1 have not been clearly elucidated. In particular, inactivation mechanisms including dephosphorylation mechanisms still remain to be elucidated. Although PP2A [37, 38], POPX1/2 [39] and PP2Cα [40] have been reported as PAK1 regulators in different cellular contexts, there has been no report on the phosphatase for PAK1 in response to oxidative stress. Our data suggest calcineurin, PP2B, as a phosphatase for PAK1 in the signaling involved in oxidative stress (Fig. 3a). Indeed, the inhibitors of PP2B decreased the death rate of neuroblastoma cells induced by 6-OHDA treatment (Fig. 3b), suggesting a new target for PD therapy. For treatment of PD, levodopa (L-dopa), a dopamine precursor has been widely used. However, long-term treatment causes some side effects such as psychiatric symptoms, wearing-off phenomenon, or on/off phenomenon. The focus of new therapy has been on the development of agents that suppress or delay neuronal death [21]. Cyclosporin A, an immunosuppressant, has been used to suppress the immune response after transplantation. Interestingly, treatment of Cyclosporin A protects cortical neurons from traumatic brain injury [41, 42] and attenuates the decrease of tyrosine hydroxylase in the nigrostriatal DA neurons and DA neurons itself in a 6-OHDA animal model of PD [21]. Our data also showed that treatment with Cyclosporin A increased the viability of neuroblastoma cells treated with the neurotoxin, 6-OHDA (Fig. 3). Thus we suggest a novel molecular mechanism for neuroprotective effects of Cyclosporin A. Inactivation of PAK1 induced by oxidative stress might be blocked by the inhibition of PP2B by Cyclosporin A, thereby increasing cell viability (Fig. 7). However, more studies are required for the clinical application of Cyclosporin A in PD patients.

These results indicate the roles of PAK1 in neuronal death and neurodegeneration induced by oxidative stress. Down-regulation of PAK1 activity was strongly correlated with cell death and loss of mesencephalic DA neurons. Conversely, up-regulation of PAK1 activity recovered loss of mesencephalic DA neurons and behavioral defects in a hemiparkinsonian animal model. Taken together, our data provided novel mechanisms in the pathogensis of PD and a useful therapeutic target.