Mechanism of Inhibition of PP2A Activity and Abnormal Hyperphosphorylation of Tau by I₂^PP2A/SET

Abstract

Protein phosphatase-2A (PP2A) activity, which is compromised in Alzheimer disease brain, is regulated by two endogenous inhibitors, one of them being I₂^PP2A, a 277 amino acid long protein also known as SET. Here we report that both the amino terminal fragment (I_2NTF; aa 1–175) and the carboxy terminal fragment (I_2CTF; aa 176–277) of I₂^PP2A inhibit PP2A by binding to its catalytic subunit PP2Ac and cause hyperphosphorylation of tau. The C-terminal acidic region in I_2CTF and Val 92 in I_2NTF are essential for their association with PP2Ac and inhibition of the phosphatase activity.

1. Introduction

Abnormal hyperphosphorylation and aggregation of microtubule associated protein tau into paired helical filaments/neurofibrillary tangles is a hallmark of Alzheimer disease (AD) and related neurodegenerative disorders, called tauopathies [1–3]. The hyperphosphorylated tau sequesters normal tau, MAP1 and MAP2, which results in the breakdown of the microtubule network and, probably in a progressive retrograde degeneration of the affected neurons [4].

Protein phosphatase 2A (PP2A) is a heterotrimeric enzyme, consisting of one catalytic C subunit, one scaffolding A subunit, and one of several structurally distinct, regulatory B subunits [5]. The phosphorylation of tau that suppresses its microtubule binding and assembly activities in adult mammalian brain is primarily regulated by PP2A [6]. PP2A regulates tau phosphorylation, both directly as tau phosphatase and indirectly by regulating the activities of several tau kinases which include CaM Kinase II, PKA, MAP kinase kinase (MEK 1/2), extracellular regulated kinase (ERK 1/2), GSK-3β, and P70S6 kinase [7]. PP2A activity in AD brain is decreased [8].

The activity of PP2A is regulated by an inhibitor protein, I₂^PP2A [9–11]; This protein, also known as SETα, TAF-1β, and PHAPII protein [12–14] was discovered as a translocated gene fused to the CAN gene in a patient with acute undifferentiated leukemia [14]. I₂^PP2A is a 277 amino acid long polypeptide with an apparent molecular weight of 39 kDa in SDS-PAGE. This protein is widely expressed in various tissues and localizes primarily in the nucleus [12,15,16] where it has been reported to block DNase activity and acetylation of histones [17]. Previously, we showed that I₂^PP2A is selectively cleaved at N175 into I_2NTF (N-terminal fragment) and I_2CTF (C-terminal fragment) and translocated from the neuronal nucleus to the cytoplasm, co-localized with neurofibrillary tangles in the affected areas of AD brain [18]. However, the exact molecular mechanism of the inhibition of PP2A activity and generation of abnormal hyperphosphorylation of tau by I₂^PP2A in the cell was not understood. The present study shows that I₂^PP2A inhibits PP2A activity dose-dependently and cleavage of this inhibitor into I_2NTF and I_2CTF not only results in its translocation from the cell nucleus to the cytoplasm but also potentiates its inhibitory activity, for each of these two fragments can interact with the PP2A catalytic subunit PP2Ac. I_2CTF interacts with PP2Ac through its highly acidic C-terminal domain. Interaction of I_2NTF with PP2Ac and its ability to inhibit the phosphatase activity is completely inhibited when valine 92 in the inhibitor is mutated to alanine.

2. Materials and Methods

For generation of I₂^PP2A and its deletion mutant plasmids, GST pull-down assays, coimmunoprecipitation, PP2A activity assays, and immunocytochemical staining, see Supplementary Data.

3. Results

3.1 I₂^PP2A interaction with PP2Ac subunit

I₂^PP2A was previously shown to inhibit PP2A activity in vitro [9,11]. To date, it is not clear whether I₂^PP2A can decrease PP2A activity in cells and whether this reduction involves inhibition of the PP2A expression level. To address to this question, we carried out transient transfection of NIH3T3 cells with different amounts of pCMV2B-I₂^PP2A. After 48 hours transfection the cell lysates were collected and one half of the sample was employed for Western blots (Fig. 1A, B), whereas the other half was subjected to immunoprecipitation with anti-PP2Ac, followed by PP2A activity assay (Fig. 1C). The FLAG-tagged I₂^PP2A expression in transfected cells increased with the amount of DNA used for transfection (Fig. 1A). The expression of PP2Ac, however, was not altered (Fig. 1B), suggesting that the decrease in PP2A activity in transfected cells (Fig. 1C) was due to PP2A inhibition and not to a reduction of the PP2A expression level.

Fig. 1. Inhibition of PP2A activity by I₂^PP2A and its interaction with PP2Ac.

(A) NIH3T3 cells were transfected with different amounts of pCMV2B-I₂^PP2A to express FLAG-tagged I₂^PP2A or, as a control, with vector alone. After 48 hours transfection, cells were lysed and expression of PP2Ac, I₂^PP2A, and, as a loading control, β-actin were determined by Western blots. (B) The relative PP2Ac expression level was normalized by the expression of β-actin. PP2Ac expression level did not show any significant difference among different amounts of expression of I₂^PP2A in NIH3T3 cells. One way ANOVA, p = 0.1369; t-test, vector vs. 1 μg, p = 0.2189; vector vs. 2 μg, p = 0.5031; vector vs. 3 μg cDNA, p = 0.1992. (C) PP2A was immunoprecipitated from cell lysates with anti-PP2Ac and the phosphatase activity assayed colorimetrically using pNPP as a substrate. I₂^PP2A inhibited the PP2A activity in a pCMV2B-I₂^PP2A dose-dependent fashion. *p<0.05 compared with control. (D) GST-pull down assay. Rat brain extract (Ext.), used as a source of PP2A holoenzyme, was incubated with Sepharose 4B beads bearing GST, GST-I₂^PP2A, or GST-PP2A-A. After washing, bound PP2Ac, PP2A-B, and PP2A-A were detected by Western blots. The GST, GST-I₂^PP2A and GST-PP2A-A used in pull down assay are shown in the lowest panel. GST-I₂^PP2A pulled down PP2Ac (PP2A catalytic subunit) but not PR55α (PP2A Bα regulatory subunit) or PR65α (PP2A A regulatory subunit). From left to right, Lane 2 (Ext.) is input, and Lane 3 (GST/Ext.) is a negative control. Positions of protein size markers are indicated on the left of the panel. Error bars in panels B and C indicate means ± SE; n=3.

I₂^PP2A was previously shown to inhibit PP2A1 (the ABC complex), PP2A2 (the AC dimer), and PP2Ac (the C subunit) in a non-competitive manner [9,19]. However, no direct evidence was reported to confirm this association. To study the interaction between PP2A and I₂^PP2A, in vitro pull-down assays were carried out using bacterially expressed GST alone (negative control), GST-PP2A-A (positive control), or GST-I₂^PP2A. Rat brain extract as the source of PP2A holoenzyme was added to purified GST-PP2A-A, GST-I₂^PP2A, or GST alone, on glutathione Sepharose 4B beads and the interactions were detected by Western blots developed with antibodies to PP2A subunits A, B, or C. We found that I₂^PP2A could pull down PP2Ac but not PP2A-A or PP2A-B subunits (Fig. 1D). These findings suggest that I₂^PP2A interacts with the PP2A catalytic subunit rather than the PP2A regulatory subunits PP2A-A or PP2A-B.

I_2FL, I_2NTF and I_2CTF localize differently in HEK293FT cells

Previously we reported that in AD brain I₂^PP2A is selectively cleaved at N175 and is translocated from the neuronal nucleus to the cytoplasm [18]. To test whether the N-terminal fragment (I_2NTF, aa1–175) or C-terminal fragment (I_2CTF, aa176–277) or both can translocate from the cell nucleus to the cytoplasm, we generated three constructs corresponding to I₂^PP2A full length (I_2FL), I_2NTF, and I_2CTF, each tagged with myc, and inserted them in pcDNA3.1 vector (Fig. 2A) for expression in eukaryotic cells. To confirm the integrity of the constructs, the plasmids were multiplied in dH5α E. coli, purified and digested by BamHI and Xhol restriction enzymes. The resulting DNA fragments were separated and extracted from 1% agarose gel (Fig. 2B). The accuracy of the DNA fragments was attested with subsequent sequencing. Finally, the plasmids were transfected either in HEK293FT cells or NIH3T3 cells and by Western blot analysis we confirmed the expression of myc-tagged I_2FL, I_2NTF, and I_2CTF, at 39 kDA, 22 kDa, and 20 kDa apparent molecular weight, respectively (Fig. 2C).

Fig. 2. Intracellular localization of I_2FL, I_2NTF, and I_2CTF with reference to tau.

(A) I_2FL-myc, I_2NTF-myc, and I_2CTF-myc cDNA were generated using pEGFP-N3/I₂^PP2A (wt) as a template. The PCR inserts as well as pcDNA3.1 were digested with BamHl and Xhol. The products were subsequently ligated into pcDNA3.1 vector and transformed into E. Coli DH4-α. (B) After amplification, the purified plasmids were digested with BamHl and Xhol, separated by electrophoresis, and sequenced. (C) Left panel: Western blots specific for the C-terminal myc tag of each protein confirmed the expression of I_2FL and I_2NTF. Right panel: Western blots specific for the C-terminal myc tag of each protein confirmed the expression of I_2FL and I_2CTF. (D) Immunocytochemical staining of HEK293FT cells transfected with GFP-Tau and pcDNA3.1 or pcDNA3.1-I_2FL, -I_2NTF, or -I_2CTF. I_2FL tagged with c-myc was localized exclusively in the nucleus of HEK293FT cells. I_2NTF tagged with c-myc was diffused all over the HEK293FT cells. I_2CTF tagged with c-myc, although concentrated in the nucleus, was also found distributed within the cytoplasm. Magnification bar = 10 μm.

To assess the intracellular localization, we co-transfected HEK293FT cells with pcDNA3.1-I_2FL and its fragments with an exclusively cytoplasmic GFP fused tau protein. Then, 72 hours post-transfection, the HEK293FT cells were fixed with 4% paraformaldehyde solution and analyzed by immunocytochemistry. The antibody to myc recognized the exogenously expressed I₂^PP2A. While as expected I_2FL was found within the nucleus, I_2NTF and I_2CTF were found diffused both in the nucleus and the cytoplasm (Fig 2D).

3.2 PP2Ac co-immunoprecipitates with I_2FL, I_2NTF, and I_2CTF

Employing GST-pull down assays we showed above (Fig. 1D) that in vitro I_2FL could interact with PP2Ac but not with PP2A-A or PP2A-B. To confirm this interaction in cells, HEK293FT cells were transfected with myc-tagged I_2FL, I_2NTF, or I_2CTF for 72 hours, followed by immunoprecipitation with anti-myc from these cell lysates and analysis by Western blots. PP2Ac co-immunoprecipitated with I_2FL as well as I_2NTF and I_2CTF (Figs. 3A, B). These findings suggested that cleavage of I₂^PP2A into I_2NTF and I_2CTF could result in an increased inhibition of PP2A activity.

Fig. 3. Co-immunoprecipitation of PP2Ac with I_2FL, I_2NTF, or I_2CTF.

HEK293FT cells transfected to overexpress c-myc-tagged I_2FL, I_2NTF or I_2CTF were lysed and c-myc tag antibody was used for immunoprecipitation. The pulled down proteins were separated on SDS-polyacrylamide gels, then transferred onto two stacked PVDF membranes of decreasing pore size and analyzed by Western blots. (A) Upper row: Western blots on the 0.45 μm PVDF membrane developed with c-myc tag antibody. Lower row: Western blots on the 0.22 μm PVDF membrane developed with c-myc tag antibody. I_2FL, I_2NTF, and I_2CTF were all immunoprecipitated with the c-myc antibody (see framed bands). (B) Western blots developed with anti-PP2Ac. NC=negative control, i.e. non-treated with the vector. Positions of protein molecular markers are shown on the left of the blots.

3.3 Inhibition of PP2A activity and generation of hyperphosphorylated tau by I_2FL, I_2NTF, and I_2CTF

Next we proceeded to assess the effect of interactions of I_2FL, I_2NTF and I_2CTF with PP2A on the phosphatase activity and on phosphorylation of tau. A previous study had reported that I₂^PP2A inhibits PP2A through its amino terminal region [12]. We, therefore, first examined whether I_2NTF could inhibit PP2A activity. We transfected HEK293FT cells with pcDNA3.1-I_2FL, -I_2NTF, or vector alone for 72 hours, and then assayed okadaic acid sensitive PP2A activity towards pSer199 tau [20] in the cell lysates. We found that PP2A activity was reduced to ~70% and ~50% in I_2FL and I_2NTF transfected cells, respectively (Fig. 4A). In order to study the effect of this PP2A inhibition on abnormal hyperphosphorylation of tau, we expressed in tau₄₄₁-stably transfected PC12 cells [21] I_2FL or I_2NTF for 72 hours. We then examined the phosphorylation of tau at multiple known pathological sites by Western blots of the cell lysates developed with various phospho-specific antibodies. We found both in I_2FL and in I_2NTF transfected cells abnormal hyperphosphorylation of tau at M4 (Thr231/Ser235), pS404 (Ser404), PHF-1 (Ser396/404), and 12E8 (Ser262/356) sites. The tau hyperphosphorylation at M4, PHF-1, and 12E8 sites, which are believed to be major sites in Alzheimer-type neurofibrillary degeneration, was quite robust in cells transfected with I_2NTF (Fig. 4B,C).

Fig. 4. Inhibition of PP2A activity by I_2FL and I_2NTF.

(A) Relative PP2A activity in transiently transfected HEK293FT cells with N-terminal myc tagged I_2FL and I_2NTF. I_2FL and I_2NTF reduced PP2A activity to ~70% and ~50%, respectively, in HEK293FT cells. *p<0.05 compared with the vector transfection control. (B) Tau₄₄₁ stably transfected PC12 cells were transiently transfected to express myc tagged I_2FL or I_2NTF and tau phosphorylation at M4 (pT231/pS235), pS404, PHF-1 (pS396/404) and 12E8 (pS262/356), total tau (R134d), as well as transfected I₂^PP2A were detected by Western blots from same amounts of lysates. The myc-tagged proteins were detectedwith anti-myc monoclonal 9E10 (1:100 in 5% defatted milk and 0.1% Tween-20) from Millipore/Chemicon. (C) The level of phosphorylated tau was quantified by densitometry and normalized to the amount of total tau. *p<0.05 compared with the vector transfection control. Positions of protein size markers are indicated on the left of the blots. Positions of I_2FL and I_2NTF bands are labeled with arrows. NSB = non-specific band.

Since both in the GST-pull down assays as well as the co-immunoprecipitation assays we had found that not only I_2NTF but also I_2CTF interacted with PP2Ac (see above), we studied the effect of I_2CTF in comparison to I_2FL on PP2A activity and hyperphosphorylation of tau in cells. In lysates of NIH3T3 cells which were transfected with I_2FL, I_2CTF, or vector alone, we found a ~31% and ~53% reduction in PP2A activity, respectively (Fig. 5A). Tau₄₄₁ stably transfected PC12 cells transfected with pcDNA3.1-I_2FL or -I_2CTF showed a marked increase in abnormal hyperphosphorylation of tau at M4 (Thr231/Ser235), pS404 (Ser404), pS262 (Ser262), and 12E8 (Ser262/346) sites (Fig. 5B,C).

Fig. 5. Inhibition of PP2Ac activity by I_2FL and I_2CTF.

(A) Relative PP2A activity in transiently transfected cells with C-terminal myc tagged constructs was detected as described in Fig. 1C. I_2FL reduced PP2A activity to ~69% whereas I_2CTF inhibited PP2A activity to ~47%. *p<0.05 compared with the vector transfection control. (B) Tau phosphorylation at M4 (pT231/pS235), pS404, pS262, 12E8 (pS262/356), total tau (R1334D) and I₂^PP2A were detected by Western blots from same amounts of lysates from I_2FL and I_2CTF transfected cells. The myc-tagged proteins were detected with anti-myc monoclonal 9B11 (1:1,000 in 5% defatted milk and 0.1% Tween-20). (C) The level of phosphorylated tau was quantified by densitometry and normalized using the amount of total tau. *p<0.05 compared with the vector transfection control. The rest of the details are the same as in Figure 4.

3.4 Domains of I₂^PP2Ainvolved in its binding to PP2A catalytic subunit

Finally, based on the structural features of I₂^PP2A (TAF-Iβ) characterized by a subtype specific N-terminal region, a coiled-coil region involved in dimerization of TAF-I, a putative nuclear localization signal, and a long stretch of acidic amino acids at C-terminal ends [12,22], we attempted to define the PP2Ac binding domain in I₂^PP2A. For this purpose, based on the structure, a series of the deletion mutants of I₂^PP2A, I_2NTF, I_2CTF and F4 corresponding to the C-terminal acidic region were generated (Fig. 6A). During these studies Val 92 in I_2NTF was accidently mutated to Ala. While we corrected this mutation by mutagenesis, we also decided to study the effect of this point mutation on interaction of I_2NTF with PP2A. In vitro pull-down assays were carried out with bacterially expressed GST fusion I₂^PP2A full length protein or its mutants. Rat brain extract was added to purified GST fusion proteins or GST alone on glutathione Sepharose 4B beads and PP2A catalytic subunit that was pulled down by these beads was detected by Western blots. The results showed that I₂^PP2A and its mutants containing C-terminal acidic region (I₂^PP2AF4 and I_2CTF) had the ability of binding with PP2Ac, whereas, the mutation of valine 92 to alanine in the leucine rich region of I_2NTF lost its PP2Ac binding ability (Fig. 6B and C). Thus, the minimal region of I_2CTF required for the binding of PP2Ac is localized at the C-terminal acidic region (amino acid 225–277), and Val 92 in I_2NTF is not only necessary for the binding but also for its inhibitory activity (Fig 6D).

Fig. 6. Identification of minimal and critical regions of I₂^PP2A required for interaction with PP2Ac.

(A) Schematic diagram of the constructs of functional domains of I₂^PP2A employed for the interaction studies. (B, C) GST-pull down assays with rat brain extract as a source of PP2A holoenzyme. The pull down assays were carried out using 0.5 μg of GST-fusion protein per mg brain extract, except in cases of GST-I_2CTF (0.1 μg was employed), and GST alone (double the amount, i.e. 1 μg was employed). I_2CTF, F4, and I_2NTF but not I_2NTFV92A interacted with PP2Ac. (D) Relative PP2A activity in transiently transfected NIH3T3 cells with C-terminal myc tagged constructs I_2NTF-myc can reduce PP2A activity and mutation of V92 to A in I_2NTF results in a complete loss of this inhibitory activity. *p<0.05.

4. Discussion

I₂^PP2A/SET is a multifunctional protein. It is believed to have a nuclear localization signal between amino acid residues 168–181 and is imported in the cell nucleus mainly by importin alpha-3 [23]. In its primary location in the cell nucleus it protects cells both by inhibiting DNAse [24] and acetylation of histones [17]. Translocation of I₂^PP2A from the nucleus to the cytoplasm has been reported to be cytotoxic and exacerbates DNA damage [23]. Generation of intracytoplasmic juxtamembrane domains of APP and APLP2 by caspase has also been shown to lead to the translocation of I₂^PP2A from the nucleus to the cytoplasm and cell death [25]. In vitro studies had shown that I₂^PP2A was also a potent inhibitor of PP2A [19] and that this inhibitory activity was localized in its amino terminal region [26]. Previously, we discovered that in AD brain I_2FL is selectively cleaved at N175 into I_2NTF and I_2CTF, and translocated from the neuronal nucleus to the cytoplasm where it colocalizes with abnormally hyperphosphorylated tau/neurofibrillary tangles [18]. A subsequent study showed that the cleavage of I₂^PP2A at N175 can be catalyzed by a lysosomal/endosomal enzyme, asparaginyl endopeptidase (also known as legumain) which occurs during cerebral acidosis caused by ischemic stroke or seizures [27]. However, neither the exact molecular mechanism by which I₂^PP2A inhibits PP2A activity nor how its cleavage into I_2NTF and I_2CTF contributes to abnormal hyperphosphorylation of tau were understood. The present study for the first time shows that I₂^PP2A decreases PP2A activity in the cell by directly interacting with and inhibiting, and not by affecting the expression of PP2Ac. Furthermore, while I_2FL localizes in the cell nucleus, both I_2NTF and I_2CTF are diffused into the cytoplasm where they interact with PP2A through PP2Ac and inhibit the phosphatase activity. I_2NTF and I_2CTF appear to be quite robust in inhibiting PP2A activity and causing abnormal hyperphosphorylation of tau. I_2CTF interacts with PP2Ac through its carboxy terminal acidic region. In I_2NTF valine 92 is required for its interaction with PP2Ac and inhibition of the phosphatase activity. The inhibition of PP2A activity by I_2FL, I_NTF, and I_2CTF was observed in all three cell lines, i.e. NIH3T3, HEK293FT, and PC12 cells, studied.

Because of the small sizes of I_2NTF and I_2CTF as compared with I_2FL, these proteins transferred to the PVDF membrane at quite variable rates and, therefore, it was not possible to normalize the inhibition of PP2A activity or hyperphosphorylation of tau from the intensity of the protein bands seen in Western blots. However, hyperphosphorylation of tau at various sites observed in I_2NTF or I_2CTF as compared with I_2FL-transfected cells indicated that the two fragments of I₂^PP2A probably inhibited PP2A activity towards phosphotau in a substrate-specific manner. Relative to I_2FL, the expression of I_2NTF caused a high level of hyperphosphorylation of tau at Thr231/Ser235 (M4 site), Ser262/356 (12E8 site), and Ser396/404 (PHF-1 site) but not at Ser404, and I_2CTF at Thr231/Ser235 (M4 site). Previously, in an in vivo study in rat, we showed that I_2CTF caused hyperphosphorylation of tau at Ser396 [28].

In AD brain the activity of PP2A is compromised [8] and tau is abnormally hyperphosphorylated [1,3]. The AD abnormally hyperphosphorylated tau probably disrupts the microtubule network and leads to neurodegeneration by sequestration of normal tau as well as microtubule associated proteins MAP1 and MAP2 [4]. In the present study, the PP2A inhibition by I₂^PP2A, especially I_2NTF and I_2CTF, resulted in abnormal hyperphosphorylation of tau at Thr231/Ser235, Ser262/356, and Ser396/404 in cultured cells. These phosphorylation sites are known to inhibit binding of tau to microtubules [7] and cause neurodegeneration, and cognitive impairment [29]. Since the inhibition of PP2A by the carboxy terminal region of I₂^PP2A was not previously known, based on the present study, we expressed I_2CTF in the brain and found inhibition of PP2A activity, neurodegeneration associated with abnormal hyperphosphorylation of tau, and intraneuronal accumulation of Aβ and impairment in spatial learning and memory in rats [28]. These in vivo findings on the PP2A inhibitory activity of I_2CTF are in agreement with the present study.

PP2A, which accounts for ~70% of the total phosphoseryl/phosphothreonyl protein phosphatase activity in human brain, is the major tau phosphatase [30]. PP2A regulates phosphorylation of tau both directly and by regulating the activities of several tau protein kinases which include PKA, CaMKII, MEK1/2, ERK1/2, and p70S6 kinase [7]. I₂^PP2A binds the neuronal cdk5 activator p35 and enhances the ckd5/p35 activity [31]. Knockdown and overexpression of I₂^PP2A in the brain have been found to decrease and increase GSK-3β activity, respectively [28,32]. Thus, a dysregulation of I₂^PP2A can have a profound effect on affected neurons, both through activation of several protein kinases and abnormal hyperphosphorylation of tau.

Besides I₂^PP2A, PP2A activity is also regulated by methylation and phosphotyrosinylation of PP2Ac and by I₁^PP2A, also known as PHAPI (putative histocompatibility leukocyte antigen class II associated protein-1), mapmodulin, pp32, and LANP [9]. Both mRNA and protein expressions of I₁^PP2A are elevated and the inhibitor is co-localized with neurofibrillary tangles in AD brain [18]. Demethylation of PP2Ac, which decreases its activity, was reported in AD brain [33,34]. Phosphorylation of PP2Ac at tyrosine 307, which inhibits its activity, was found to be elevated in AD brain [35]. Microarray analyses showed an upregulation of the expression of SET genes in AD hippocampus [36]. While the mRNA and protein expressions of I₂^PP2A are elevated in AD brain, reported by us previously [18], the dual action of I_2NTF and I_2CTF in inhibiting PP2A activity shown in the present study shows how the cleavage and translocation of this inhibitor could work as the last straw which broke the camel’s back in producing neurofibrillary degeneration in AD.

Highlights.

• Molecular mechanism of protein phosphatase-2A activity by I₂^PP2A/SET

• Role of cleavage of I₂^PP2A into I_2NTF and I_2CTF in inhibition of PP2A activity

Abbreviations

AD

Alzheimer disease

PP2A

protein phosphatase 2A

MEK1/2

MAP kinase kinase

ERK 1/2

extracellular regulated kinase

PHAP1

putative histocompatibility leukocyte antigen class II associated protein-1