Protein phosphatase 2A and Alzheimer’s disease
Protein phosphatase 2A (PP2A) is one of the most important serine/threonine phosphatase in mammalian brains [
1]. It accounts for as much as 1% of total cellular proteins in most tissue and cells. As a highly conserved enzyme [
2], it has ubiquitous distribution and plays important roles in development, cell growth, and transformation [
3,
4]. The PP2A holoenzyme is composed of scaffolding/structural A, regulatory/targeting B and catalytic C subunits. Together with another protein phosphatase—protein phosphatase 1 (PP1), PP2A accounts for more than 80% of the total serine/threonine phosphatase activity in mammalian cells. Correspondingly, the PP2A activity was tightly and precisely controlled. Until now, we have known the holoenzyme composition, PP2A inhibitors and post-translational modifications of the catalytic subunit dominate the regulation of PP2A activity [
1].
Alzheimer’s disease (AD) is one of the most common neurodegenerative diseases in the developed countries. It is characterized by the presence of a large number of extracellular deposits of β-amyloid (Aβ) and intracellular neurofibrillary tangles (NFTs) in the brain. The latter is well correlated with dementia severity [
5]. NFTs are composed of bundles of paired helical filaments (PHF), the main component of which is abnormally hyperphosphorylated microtubule-associated protein tau [
6]. In normal adult brains, tau proteins were found mostly in axons and have the function of promoting the polymerization of tubulins and stabilizing the microtubulin. In AD brains, tau proteins were hyperphosphorylated, translocated to cell body and aggregated into PHF, correspondingly, the axonal degeneration happened in neurons. Until now the reason and role of tau hyperphosphorylation in AD pathogenesis have not been fully understood. It is believed that tau hyperphosphorylation promotes NFT formation, synapse degeneration and neuron loss [
7,
8]. Hence, investigating the upstream reasons and underlying mechanism of tau hyperphosphorylation may contribute a lot to disclosing the pathogenesis of AD, and then, promote the clinical treatments.
AD-like abnormal hyperphosphorylation of tau mainly occurs on serine and threonine sites [
9]. Tau hyperphosphorylation in AD is thought to be caused by an imbalance of the corresponding tau candidate protein kinases (PKs) and protein phosphatases (PPs). Numerous researches have suggested that quite a lot of PKs and PPs were involved in tau hyperphosphorylation, such as glycogen synthase kinase-3 (GSK-3), mitogen-activated protein kinases (MAPKs), extracellular signal-regulated protein kinase (ERK), C-jun amino terminal kinase (JNK), and protein phosphatase (PP)2A. Among various kinases and phosphatases involved, PP2A is the most implicated [
9,
10,
11,
12,
13]. Also, PP2A is suggested to be involved in Aβ production. When PP2A was inhibited in N2a cells with okadaic acid (OA), the phosphorylation of APP (precursor of Aβ) at Thr-668 was increased and the secretion of both sAPPα and sAPPβ were enhanced [
14].
Decreased PP2A activity in AD brains and its relationship with tauopathy and Aβ over-production
The activities of several protein phosphatases, including PP1, PP2A and PP5 were decreased in AD brains [
9,
15,
16], while down-regulation of PP2A is thought to play the major role in tau hyperphsophorylation in AD. An
in vitro testing shows that PP2A, PP1, PP5 and PP2B accounted for approximately 71%, 11%, 10% and 7%, respectively, of the total tau phosphatase activity of human brain [
9]. Other evidences indicating that PP2A is the major tau phosphatase are as follows: (1) PP2A has a much stronger ability than PP-1 and PP-2B in terms of dephosphorylation of PHF-tau and restoration of the ability of tau protein to promote microtubule assembly [
17,
18,
19]; (2) when PP2A was inhibited in cultured cells, competent brain slices or in rat brain, tau was hyperphosphorylated at several PHF-tau sites [
20,
21,
22,
23]; (3) hyperphosphorylation and altered compartmentalization of tau were observed in transgenic mice with a dominant negative mutant form of the catalytic subunit of PP2A [
24].
The direct evidence that PP2A is involved in Aβover-production is lacking, but studies investigating the effects of OA (PP1 and PP2A inhibitor) on the regulation of APP phosphorylation and metabolism were explored by several research groups. The results seemed to be conflicting or inconsistent. Although incubation of various cell types with OA leads to stimulation of APP secretion [
25-
28], it is not clear whether these effects are preferentially mediated by PP1 [
26] and/or PP2A [
27]. Moreover, OA has been reported to either decrease [
27,
29] or increase[
30,
31] Aβproduction. But notably, phosphorylation at Thr-668 has been linked to increased amyloidogenic processing of APP [
32,
33,
34]. Accordingly, enhanced APP phosphorylation at the Thr-668 site is correlated with increased Aβ production [
14].
All these reports supported the idea that PP2A was the key phosphatase involved in AD pathogenesis, hence, indicating the upstream reasons and underlying mechanism of PP2A inactivation might contribute a lot to disclosing the pathogenesis of AD, and then, promote the clinical treatments.
Mechanisms of PP2A inactivation in AD brains
The regulation of PP2A is very complicated. Until now, the mechanism for PP2A inactivation in AD brains has not been fully understood. Decreased mRNA levels of PP2A catalytic subunit alpha (the major PP2Ac isoform in brain) and PP2A regulatory subunits PR55γ and PR61ϵ were found in the CA3 hippocampus of AD [
35]. Consistent with this report, Sontag et al found that the protein level of PP2A catalytic subunit was decreased in the frontal and temporal lobes [
36]. In our research, we also got a significant decrease of the PP2A catalytic subunit level in AD temporal lobes by using two different antibodies [
37]. At the same time, a significant decrease of Bα subunit of PP2A was observed [
36]. Due to the lacking of specific antibodies to PP2A PR55γ and PR61ϵ, the protein levels of these two regulatory subunits could not be tested [
36]. Since the holoenzyme PP2ABαC is the major neuronal PP2A isoform that binds to tau [
38], and the PP2A heterotrimer is the major functional form
in vivo [
1], the decreased level of Bα may partly account for the down-regulated PP2A activity in AD.
Another regulation pathway on PP2A activity is the regulation of PP2A inhibitors. Two specific, non-competitive and heat-stable inhibitors of PP2A were purified from bovine kidney and termed I
1PP2A and I
2PP2A [
39]. Both proteins inhibit all holoenzyme forms of PP2A, probably by binding directly to the catalytic subunit [
40]. Tanimukai et al found: (1) a significant increase in the neocortical levels of I
1PP2A and I
2PP2A mRNA expressions in AD as compared with those in control cases; (2) a shift in the intracellular distribution of I
2PP2A from its primarily nuclear location to the cytoplasm in neurons; (3) an increase in the I
2PP2A cleavage activity in AD brain, and (4) a co-localization of the two inhibitors with PP2A in neuronal cytoplasm and with the abnormally hyperphosphorylated tau in neurons with early- to middle-stage neurofibrillary degeneration [
41]. These studies suggest the possible involvement of I
1PP2A and I
2PP2A in the down-regulation of PP2A and abnormal hyperphosphorylation of tau in AD.
The third way of PP2A regulation is holoenzyme composition and post-translational modification [
1,
42]. The PP2A holoenzyme is composed of scaffolding/structural A, regulatory/targeting B and catalytic C subunits. The C subunit has phosphatase activity. C and A subunits constitute the core dimer enzyme of PP2A. It is suggested that the PP2A
D represents at least one third of the total cellular PP2A [
43]. PP2A
D and B subunits compose the PP2A holoenzyme. Until now, people have found two A, two C and at least 18 B isoforms, which provides the possibility that about 75 different dimeric and trimeric PP2A holoenzymes can be generated. This specific holoenzyme composition provides many possibilities for regulation. The regulatory subunits determine the cellular and subcellular localizations, and the substrate specificity of PP2A holoenzymes. And the presence or absence of additional subunits can affect the catalytic activity of PP2A towards the same substrate [
1]. Phosphorylation at Tyr307/Thr304 and methylation at Leu309 of the N-terminal are two major post-translational modifications of PP2Ac [
44].
One of the possibilities of reduced PP2A activity in AD was thought to be caused by the reduced level of PP2A heterotrimeric enzyme PP2ABαC, which is the major isoform of neuronal PP2A that binds to tau [
36]. In a further investigation, the down-regulation of PP2A carboxyl Leu309 methylation was suggested to be involved in decreased PP2ABαC level in AD brain, since the methylation of PP2Ac at Leu309 seemed important for the holoenzyme composition [
45,
46]. In several experimental systems, the decreased methylation of PP2A were induced by some suggested upstream factors of AD such as Aβ, estrogen deficiency and impaired homocysteine metabolism, and the methylation of PP2A correlated with the increased tau phosphorylation and accumulation [
47,
48]. These results indicate that the increased demethylation of PP2A (L309) may contribute to the reduced PP2A activity observed in the AD brain.
PP2A phosphorylation on both Thr304 and Tyr307 residues is associated with its inactivation [
49-
51]. Thr304 is phosphorylated by “autophosphorylation-activated protein kinase” [
49]. Tyr307 is a target for p60v-src, p56lck, and epidermal growth factor and insulin receptors [
50-
52]. PP2A reactivation occurs through its unique ability to catalyze intramolecular autodephosphorylation [
49,
50]. In a recent research [
37], we investigated whether or not increased Tyr307 phosphorylation of PP2A happened in AD and its relationship with tau hyperphosphorylation. The results showed an increase of PP2Ac-Yp307 in 100 000 g pellets of AD brain tissue homogenate, and an aberrant accumulation of PP2Ac-Yp307 in neurons that bear pre-tangles or tangles in the susceptible brain regions such as the entorhinal cortical cortex and the hippocampus. In three different experimental systems, the over-production of Aβ was correlated with PP2Ac-Y307 phosphorylation, PP2A inactivation and tau hyperphosphorylation. In addition, estrogen deficiency could also induce PP2Ac-Y307 phosphorylation and tau hyperphosphorylation. These findings further explained the mechanism of PP2A inactivation in AD.
Remaining questions
Though a lot of work has been done to disclose the role of PP2A in AD pathogenesis, large areas still remain unclear. For example, why is the PP2A catalytic subunit level decreased in AD? Which types of PP2A holoenzymes are involved in AD neurons? Why is PP2A specifically inactivated in some neurons? What are the precise roles of PP2A in β-amyloid metabolism? And what are the up-stream reasons that induce the abnormal demethylation and phosphorylation of PP2A? We hope the solution to these questions may eventually lead to the discovery of therapeutic agents that can counteract PP2A de-regulation in AD.
Summary
As the major tau protein phosphatase, the activity of PP2A is decreased in AD brain, thus promoting the abnormal hyperphosphorylation of tau, which is supposed to be involved in NFT formation and neurodegeneration. PP2A is also involved in the APP secreting pathway, and probably participates in the Aβ production. Decreased PP2Ac level, decreased PP2A holoenzyme composition, increased level of PP2A inhibitors, increased PP2Ac Leu309 demethylation and Tyr307 phosphorylation could partly explain the mechanisms of PP2A inactivation in AD. Aβ over-production, estrogen deficiency and impaired homocysteine metabolism are the possible up-stream factors that inactivate PP2A in AD neurons (Fig. 1). Further studies are required to disclose the role of PP2A in AD.
Higher Education Press and Springer-Verlag Berlin Heidelberg