PI3K/Akt is arguably the most important cell-survival signaling pathway for neurons. As the key negative regulator of the PI3K/Akt pathway, PTEN is an important target of study for neuroprotection during neurodegeneration. In this work, we are the first to demonstrate NO-mediated redox regulation as the mechanism of PTEN protein degradation. We also demonstrate that NO rapidly induces S-nitrosylation of PTEN, thereby inactivating it. Moreover, NO, but not H2
induces PTEN protein degradation. Loss of PTEN protein is reportedly associated with myocardial and brain ischemia [27
], presumably as a cellular adaptive stress response to activate the pro-survival PI3K/Akt signaling. Herein, we show that PTEN loss also occurs in neurons in response to a variety of neurotoxins (e.g., glutamate and Aβ peptides) as well as in chronic neurodegenerative conditions such as AD and PD brains. Hence, our findings on NO-mediated PTEN protein degradation may represent a common mechanism underlying PTEN loss in these acute and chronic degenerative conditions, in which NO plays a critical pathophysiological role.
It is widely accepted that oxidative stress is one of the earliest changes that occurs in the pathogenesis of AD, arising from the imbalance between increased production of reactive oxygen and nitrogen species and impaired antioxidant defenses, as reflected in the accumulation of oxidative damage to macromolecules detected in MCI, the clinical precursor of AD, and AD brains [30
-induced modification and S-nitrosylation represent the two dominant oxidative events through targeted modifications of critical Cys residues in proteins. H2
-induced PTEN oxidation was reported to cause the formation of an intra-chain disulfide bond between C71-124 [22
], and is reflected in faster mobility species by band-shift assay on non-reducing gels (Figure ). SNOC did not induce a band shift of PTEN on non-denaturing gels and thus unlikely induced a major conformational change due to intra-chain disulfide bond formation. However, our mutagenesis data suggest the involvement of overlapping residues on C71 and C124. It is, therefore, reasonable to predict that these two oxidative modifications can compete with each other when both oxygen and NO species are present. Moreover, our data indicate that NO-mediated oxidation is the predominant form of PTEN in aging brains and in MCI/AD brains (Figure ), which may generalize to other neurodegenerative diseases such as PD.
Although our mutagenesis studies (Figure ) cannot determine the nitrosylated sites unambiguously, the results strongly suggest that C83 is likely the most significant physiological site of S-nitrosylation on PTEN. It is possible that multiple Cys residues are involved depending on the spatial and temporal concentrations of NO. It is well known that the C124 residue is critical for the enzymatic activity of PTEN [32
]; mutation of this residue results in inactivation of both the protein and phospholipid lipase activities of PTEN. Our data [see additional file 3
] also indicate its importance in PTEN protein stability; mutation on this Cys residue renders PTEN resistance to SNOC-induced protein degradation. It warrants further investigation of the underlying mechanisms. Structural analysis based on the solved crystal structure of PTEN [33
] indicates that Cys83 is not in close vicinity to C124, which is located face-to-face to C71 (Figure ). It is therefore not yet clear mechanistically how C83, in conjunction with C71 and C124, affects the lipid phosphatase activity of PTEN.
Our results demonstrate NO-mediated PTEN protein degradation via UPS, as evidenced by enhanced ubiquitination. There are several precedents showing that protein S-nitrosylation can be functionally coupled to its ubiquitination and modulate protein degradation [34
]. It is possible that S-nitrosylation of PTEN plays a direct causative role in its degradation, as evidenced by enhanced ubiquitination upon SNOC treatment (Figure ). Given the role of phosphorylation in modulating PTEN protein stability and activity previously revealed by cancer cell models [14
], it is also possible that NO signal induces alteration on PTEN phosphorylation status which is the cause of enhanced ubiquitination and protein degradation. Interestingly, the two putative kinases identified as responsible for phosphorylation of the two major clusters on PTEN (Ser/Thr cluster 380/382/383 and Thr366/Ser370), namely glycogen synthase kinase GSK3β and casein kinase CK2, are both implicated in neurodegeneration [39
]. Moreover, our unpublished data show that OA can prevent PTEN dephosphorylation to a significant extent following exposure to SNOC, suggesting that PP2A, PP1 and perhaps PP2B, all major protein phosphatases implicated in AD [41
], may play a role in modulating PTEN stability. Thus, the detailed interplay between kinases and phosphatases warrants further investigation.
Although we made our initial observation in MCI/AD brain samples, the loss of PTEN is also reported to occur in both myocardial [27
] and cerebral ischemia/reperfusion [28
]. Therefore, we believe that PTEN inhibition in these acute conditions mediates subsequent activation of PI3K/Akt signaling, which is known to be key to the endogenous protective effect, similar to the neuronal adaptive response. Paradoxically, we found that loss of PTEN occurs in human brains with AD, accompanied by elevated P-Akt in AD-affected regions. Moreover, the elevated P-Akt has often been detected in the same neuron bearing neurofibrillary tangles, which is a major pathological hallmark of AD consisting of hyperphosphorylated tau protein [13
]. Although a reduction of P-Akt was also reported by previous studies [42
]; we repeatedly found elevated P-Akt levels in degenerating neurons in AD brains by immunohistochemistry (unpublished data), which is consistent with several other reports [11
]. Taken with our earlier finding that downregulated PTEN resulted in tau hyperphosphorylation and aggregation, we speculate that the loss of PTEN may be a contributing factor to neurodegeneration over the course of the disease due to chronically or excessively activated Akt signaling, which we speculate to be detrimental, as has been reported in several other chronic diseases [44
]. Akt is activated in samples from patients with chronic heart failure; biochemical analyses demonstrated that chronic Akt activation induces feedback inhibition [46
]. This nascent theory appears to be supported by a finding that conditional PTEN ablation in the forebrain region which caused impaired synaptic structure and function with concomitant constitutive activation of Akt and mTOR signaling [47
]. An alternative view would be that the increase in Akt signaling occurs in response to damage, as an adaptive response, but fails to reach significantly protective levels under these chronic conditions. Only future experiments will be able to distinguish between these alternatives.