Tau and GSK-3β are Critical Contributors to α-Synuclein-Mediated Post-Stroke Brain Damage

Post-stroke secondary brain damage is significantly influenced by the induction and accumulation of α-Synuclein (α-Syn). α-Syn-positive inclusions are often present in tauopathies and elevated tau levels and phosphorylation promotes neurodegeneration. Glycogen synthase kinase 3β (GSK-3β) is a known promoter of tau phosphorylation. We currently evaluated the interaction of α-Syn with GSK-3β and tau in post-ischemic mouse brain. Transient focal ischemia led to increased cerebral protein–protein interaction of α-Syn with both GSK-3β and tau and elevated tau phosphorylation. Treatment with a GSK-3β inhibitor prevented post-ischemic tau phosphorylation. Furthermore, α-Syn interaction was observed to be crucial for post-ischemic GSK-3β-dependent tau hyperphosphorylation as it was not seen in α-Syn knockout mice. Moreover, tau knockout mice show significantly smaller brain damage after transient focal ischemia. Overall, the present study indicates that GSK-3β catalyzes the α-Syn-dependent tau phosphorylation and preventing this interaction is crucial to limit post-ischemic secondary brain damage.


Introduction
α-Synuclein (α-Syn) is one of the abundant proteins in the brain that is known to cause synucleinopathies and chronic neurodegeneration (Clinton et al., 2010;Lashuel et al., 2013;Savica et al., 2013). It is highly induced in both human and rodent brains following stroke (Kim et al., 2016(Kim et al., , 2018. The accumulation and aggregation of α-Syn promote ischemic brain damage, whereas its deletion, silencing, or inhibition decreases infarction, mitochondrial fragmentation, oxidative stress, apoptosis, and autophagy and better sensorimotor and cognitive recovery after transient focal ischemia in rodents (Kim et al., 2016(Kim et al., , 2018Unal-Cevik et al., 2011). α-Synpositive inclusions are frequently observed in tauopathies and increased tau levels and phosphorylation are known to occur in Parkinson's disease (Moussaud et al., 2014). This indicates that α-Syn and tau promote each other's fibrillization and solubility indicating a synergistic mechanism of secondary brain damage (Clinton et al., 2010;Postina, 2008). In CNS, both α-Syn and tau are predominately induced in neurons (Kim et al., 2016;Pluta et al., 2018).
Tau plays a crucial role in protecting microtubule integrity that is essential for vesicular transport, polarity, and signal transduction in axons and dendrites (Chen & Jiang, 2019). However, tau phosphorylation leads to its aggregation and formation of neurofibrillary tangles that promote apoptosis after cerebral ischemia (Gordon-Krajcer et al., 2007;Morioka et al., 2006;Wen et al., 2007). Tau phosphorylation exacerbates and its deficiency protects the post-ischemic rodent brain (Bi et al., 2017). Glycogen synthase kinase 3β (GSK-3β) promotes tau hyperphosphorylation and its inhibition reduces brain damage and improves neurologic function after focal ischemia (Toral-Rios et al., 2020;Wang et al., 2016Wang et al., , 2017. We presently evaluated if α-Syn and GSK-3β interactively modulate tau phosphorylation and the downstream brain damage after transient focal ischemia in adult mice.

GSK-3β Inhibitor Administration
A cell-permeable GSK-3β Inhibitor VIII, CAS 487021-52-3 (Cat # 361,549; Calbiochem USA) that regulates the biological activity of GSK-3β by phosphorylation and dephosphorylation at different residues and prevents tau phosphorylation at a GSK-3β-specific site was injected (4 mg/kg dissolved in 5 mg/ml of DMSO and supplemented with 0.9% sterile NaCl; IV) at 30 min of reperfusion following transient MCAO. Mice in the control group received an equal volume of vehicle (Venna et al., 2015).

Co-Immunoprecipitation (Co-IP)
The protein-protein interaction between α-Syn, tau, and GSK-3β was analyzed using Co-IP Assay (Cat. # ab206996; Abcam USA) as described previously (Kim et al., 2021). Briefly, the ipsilateral peri-infarct cortex was homogenized in a non-denaturing lysis buffer containing a cocktail of protease and phosphatase inhibitors. The homogenate was centrifuged (10,000 g; 5 min at 4 °C) and 300 µg protein equivalent of the supernatant was incubated overnight at 4 °C with antibodies against α-Syn or GSK-3β (each at 1:50; Cell Signaling Technology USA) and with 25 µl of protein A/G Sepharose beads slurry (1 h at 4 °C). The precipitated protein complex was eluted, washed in buffer, incubated (5 min at 80 °C) with NuPAGE LDS Sample Buffer (Cat. # NP0008; Life Technologies USA) and immunoblotted to evaluate the protein-protein interaction. Rabbit IgG was used as a control.

Western Blotting
Ipsilateral peri-infarct cortex was homogenized in T-PER Tissue Protein Extraction Reagent (Cat. # 78,510; Life Technologies USA) supplemented with a cocktail of protease and phosphatase inhibitors. Homogenates were centrifuged (10,000 g for 10 min at 4 °C) and 40 µg protein equivalents were electrophoresed, transferred to nitrocellulose membranes, and probed with antibodies against α-Syn, tau, p-tau, GSK-3β, p-GSK-3β (S9), and GAPDH (each at 1:1,000; Cell Signaling Technology USA), followed by HRP-conjugated (1:3,000) or IRDye Infrared Fluorescent Dye-conjugated (1:20,000; LI-COR) anti-rabbit or anti-mouse antibodies. Protein bands were detected using enhanced chemiluminescence or scanned with a NIR spectrum (between 680 and 800 nm) using a LI-COR Odyssey analyzer and the band intensity was quantitated with Image Studio software (LI-COR Biotechnology USA).

Infarct Volume Assessment
As infarct size after transient MCAO in mice is known to reach ~ 90% to 95% by 24 h of reperfusion, it was assessed with T2-MRI (4.7 T small animal system scanner; Agilent Technologies USA) at that time. Briefly, mice were anesthetized with isoflurane and 8-10 coronal brain slices at 1.0 mm thickness and 20 × 20 mm 2 fields were scanned under isoflurane anesthesia as described previously (Chelluboina et al., 2021). The respiration rate was monitored during the imaging. The Infarct size was computed blindly, and the total infarct volume was configured by numeric integration of data from serial coronal sections factoring-in sectional intervals and corrected for edema using NIH ImageJ software with an FDF plugin as described before (Chelluboina et al., 2021;Kim et al., 2018;Mehta et al., 2015Mehta et al., , 2022.

Transient Focal Ischemia-Induced Phosphorylation of GSK-3β and Tau
The p-tau levels increased (by 2-fold to 2.5-fold; p < 0.05) and p-GSK-3β levels decreased (by 3-fold to 4-fold; p < 0.05) in the ipsilateral peri-infarct cortex at 6 h, 12 h, and 24 h of reperfusion following transient MCAO in adult mice compared with sham ( Fig. 1A and B). However, protein levels of total tau and total GSK-3β were unaltered at any reperfusion period tested compared with sham ( Fig. 1A and B).

α-Syn Forms a Complex with p-tau and p-GSK-3β After Focal Ischemia
We previously demonstrated that induction of α-Syn after focal ischemia in rodents increases DRP1 phosphorylation that promotes mitochondrial fragmentation (Kim et al., 2016). To understand if α-Syn promotes GSK-3β phosphorylation and the further downstream tau hyperphosphorylation that modulates DRP1 and mitochondrial damage after focal ischemia, we immunoprecipitated protein complex with α-Syn antibodies and evaluated the interaction of α-Syn with p-tau, and p-GSK-3β at 12 h of reperfusion following transient MCAO in the ipsilateral cerebral cortex. As expected, α-Syn protein levels increased significantly in the post-ischemic brain ( Fig. 2A). α-Syn binding to p-tau increased and to p-GSK-3β decreased significantly after transient MCAO compared to sham ( Fig. 2A), whereas α-Syn binding to total tau and total GSK-3β were not altered after transient MCAO ( Fig. 2A).
Using α-Syn −/− mice, we investigated if post-ischemic GSK-3β activation and tau hyperphosphorylation are α-Syn dependent. In α-Syn −/− mouse brain, total GSK-3β and total tau protein levels did not change following transient MCAO and 12 h of reperfusion in the ipsilateral cortex compared to sham (Fig. 2B). This is similar to what we observed in wild-type mice. However, α-Syn deletion prevented postischemic tau hyperphosphorylation seen in wild-type mice (Fig. 2B). We further evaluated the interaction of α-Syn, tau, and GSK-3β following immunoprecipitation with GSK-3β antibody. In α-Syn −/− mice subjected to transient MCAO and 12 h of reperfusion, both GSK-3β phosphorylation and tau phosphorylation were similar to sham (Fig. 2B), unlike in wild-type control. This indicates that GSK-3β bound to tau is S9 phosphorylated (inactive), which cannot prevent tau phosphorylation in the wild-type mice when it was decreased following transient MCAO. However, in α-Syn −/− mice, the reversal of GSK-3β phosphorylation prevents tau hyperphosphorylation. This suggests that α-Syn availability is essential for GSK-3β suppression and the resulting tau hyperphosphorylation in the post-ischemic brain.

Inhibition of GSK-3β Prevented Tau Hyperphosphorylation in the Ischemic Brain
We injected mice with a potent GSK-3 inhibitor VIII that prevents tau phosphorylation at a GSK-3-specific site to confirm the role of GSK-3β in controlling tau hyperphosphorylation (Venna et al., 2015). GSK-3β inhibitor given IV at 30 min of reperfusion significantly reversed the decrease in GSK-3β S9 phosphorylation and subsequently decreased tau phosphorylation at 12 h of reperfusion following transient MCAO compared to the vehicle control group (Fig. 3). This confirms that GSK-3β mediates tau phosphorylation after focal ischemia. Fig. 1 Transient focal ischemiainduced tau phosphorylation but prevented GSK3β S9 phosphorylation. Protein abundance of p-tau increased (A) and p-GSK-3β decreased (B) in the peri-infarct cortex of mice at 6 h, 12 h, and 24 h of reperfusion following transient MCAO. Total tau and GSK3β protein levels were unaltered after focal ischemia. Each band in the blot is the biological representation. Values are mean ± SD (n = 4/ group); *p < 0.05 compared with sham by Mann-Whitney U test

Tau Deletion Decreased Brain Damage After Transient Focal Ischemia
We previously demonstrated that mice lacking α-Syn suffered less brain damage and recovered motor functions after focal ischemia better than wild-type mice. As shown above, α-Syn seems to influence GSK-3β and its downstream tau signaling. As tau hyperphosphorylation and its aggregation are known to promote neurodegeneration (Ballatore et al., 2007), we tested the effect of focal ischemia in Tau −/− mice. Tau −/− mice showed no total tau or p-tau protein after transient MCAO or sham surgery (Fig. 4A). Tau −/− mice showed significantly smaller infarct volume (by 38%; p < 0.05) at 1 day of reperfusion following 1 h of transient MCAO compared with Tau +/+ mice (Fig. 4B).
The current studies show that the basal levels of GSK-3β and tau are unaltered after focal ischemia, but their phosphorylation status changes significantly. Importantly, GSK-3β S9 phosphorylation is reduced, leading to its activation and downstream tau phosphorylation. The key component Fig. 2 α-Syn interacted with p-tau and p-GSK-3β after transient focal ischemia. The protein lysates from peri-infarct cortical tissue of mice subjected to transient MCAO and 12 h of reperfusion were immunoprecipitated with α-Syn antibody. Immunoprecipitated proteins were subjected to western blotting and probed with antibodies against tau, p-tau, GSK-3β, and p-GSK-3β. Following transient MCAO, α-Syn/p-tau interaction increased while α-Syn/p-GSK-3β interaction decreased compared with sham (A). We then repeated the experiment with protein lysates from α-Syn −/− mice subjected to transient MCAO and 12 h of reperfusion. As α-Syn protein is absent in α-Syn −/− mice, we used GSK-3β antibodies for immunoprecipitation. In α-Syn −/− mice, there was no change in interaction of p-GSK-3β and p-tau in sham or ischemic mice, indicating the essential nature of α-Syn in this process (B). Each band in the blot is the biological representation. Values are mean ± SD (n = 4/group); *p < 0.05 compared with sham by Mann-Whitney U test that enables this seems to be the increased post-ischemic expression of α-Syn, which binds to both GSK-3β and tau. Previous studies showed that α-Syn is directly involved in tau hyperphosphorylation (Duka et al., 2006). Our studies confirm this by showing that post-ischemic tau phosphorylation does not change in α-Syn knockout mice (known to be protected after focal ischemia) compared to sham, indicating that α-Syn binding is critical. Following transient MCAO, tau phosphorylation peaked at 12 h of reperfusion and remained elevated up to 24 h of reperfusion. GSK-3β is the major kinase that phosphorylates tau, and we observed reduced GSK-3β S9 phosphorylation between 6 h and 24 h of reperfusion. GSK-3β activity is inhibited when the S9 residue is phosphorylated (Duka et al., 2009), leading to GSK-3β activation and subsequent downstream tau phosphorylation. Thus, further interaction and inhibitor studies were conducted at 12 h of reperfusion following transient MCAO. Our studies also show that the GSK-3β inhibitor promoted S9 phosphorylation and prevented downstream tau phosphorylation. This indicates the essentiality of GSK-3β in Tau phosphorylation and an association between α-Syn, GSK-3β, and tau in mediating post-ischemic brain damage.
Tau is a neuron-specific cytoskeletal protein involved in the assembly and stabilization of microtubules regulating axonal transport and maintaining neuronal morphology (Guo et al., 2017). Progressive tau hyperphosphorylation leads to its aggregation/deposition, impairing normal cellular functions and leading to excitotoxicity, whereas lowering tau is neuroprotective (Roberson et al., 2007). Tau can be phosphorylated at various serine, threonine, and tyrosine residues, including S199-202, Thr205, and S396-404, which are linked to abnormal tau processing in tauopathies. Studies have revealed that tau phosphorylation at S396-404 is one of the earliest events in neurodegenerative diseases, where nearly half of all structures harboring p-S396-404 form early aggregates with a well-preserved neuronal soma (Mondragón-Rodríguez et al., 2014). GSK-3β can directly phosphorylate tau at S396, S400, and S404 without the involvement of other kinases (Morfini et al., 2002), and this favors the interaction of tau with α-Syn and GSK3β. Consistent with these reports, we observed a significant increase in p-tau (S404) levels and interaction with α-Syn and GSK-3β following transient MCAO. This suggests that tau S404 phosphorylation is an indicator of neuronal damage after ischemic stroke. We previously reported that post-ischemic mitochondrial dysfunction (increase in Drp1 and p-Drp1 protein levels) and oxidative stress could be curtailed by silencing or repressing α-Syn (Kim et al., 2016(Kim et al., , 2018. The oxidative damage and mitochondrial dysfunction accompany an increase in tau phosphorylation (S396/404) in the hippocampus of aged mice (Torres et al., 2021), suggesting that tau and α-Syn are linked in aggravating post-ischemic secondary brain damage. Consistent with this, the present study shows that tau knockout mice that lack total tau as well as p-tau show significantly lower infarction following transient focal ischemia. Since the tau protein regulates synaptic function, tau deletion or knockdown in mice could lower brainderived neurotrophic factor, leading to deficiencies in longterm depression and long-term hippocampal potentiation in aging (Gonçalves et al., 2020;Velazquez et al., 2018). Our studies also indicate a progression from α-Syn to GSK-3β to tau in post-ischemic brain damage. Fig. 3 GSK-3β inhibitor VIII prevented GSK-3β S9 dephosphorylation, leading to reduced tau phosphorylation after transient focal ischemia. Treatment with GSK-3β Inhibitor VIII at 30 min of reperfusion following transient MCAO reversed post-ischemic changes in both GSK-3β phosphorylation and tau phosphorylation compared to vehicle control. GSK-3β Inhibitor VIII did not change the protein levels of total tau and total GSK-3β. Each band in the blot is the biological representation. Values are mean ± SD (n = 4/group); *p < 0.05 compared with vehicle control by Mann-Whitney U test Overall, the current study demonstrates that a complex formed by α-Syn, tau, and GSK-3β is necessary for GSK-3β-catalyzed tau hyperphosphorylation in the post-ischemic brain and that α-Syn serves as a key initiator of tau-mediated ischemic brain damage. Disrupting these interactions might be a novel strategy to prevent post-ischemic brain damage. Fig. 4 Tau deletion decreased post-ischemic secondary brain damage. Tau −/− mice showed no tau or p-tau protein expression in the peri-infarct cortex of mice subjected to 1 h of transient MCAO and 1 day of reperfusion compared with Tau +/+ mice (A). T2-MRI showed reduced infarct volume in Tau −/− mice compared to Tau +/+ mice on day 1 of reperfusion (B). Each band in the blot is the biological representation. Values are mean ± SD (n = 4/group in A and 8/group in B); *p < 0.05 by the Mann-Whitney U test