The tau kinase signaling cascade represents a critical pathogenic mechanism in Alzheimer's disease (AD) and related tauopathies. Hyperphosphorylation of the microtubule-associated protein tau leads to its aggregation into neurofibrillary tangles (NFTs), a hallmark neuropathological feature strongly correlated with cognitive decline. Understanding the kinases that regulate tau phosphorylation is essential for developing disease-modifying therapeutics.
Tau is a natively unfolded protein primarily expressed in neurons, where it promotes microtubule assembly and stability. In the adult human brain, six tau isoforms are produced through alternative splicing of the MAPT gene, ranging from 352 to 441 amino acids[1]. Tau contains over 80 potential phosphorylation sites, primarily serine and threonine residues, with lesser tyrosine phosphorylation[2].
Physiological tau phosphorylation regulates its microtubule-binding capacity, synaptic functions, and neuronal viability. However, pathological hyperphosphorylation disrupts tau's ability to bind microtubules, leading to microtubule instability and promoting tau aggregation into insoluble paired helical filaments (PHFs) and NFTs[3].
The balance between tau kinases and phosphatases determines phosphorylation state. In AD, this balance shifts dramatically toward increased kinase activity and/or decreased phosphatase activity, particularly in the temporal lobe and hippocampus[4].
GSK-3β is a serine/threonine kinase encoded by the GSK3B gene, constitutively active in neurons under normal conditions[5]. It exists as two isoforms: GSK-3α (51 kDa) and GSK-3β (47 kDa), with GSK-3β being the predominant isoform in the brain and the primary kinase implicated in tau phosphorylation[6].
GSK-3β activity is regulated through multiple mechanisms:
GSK-3β phosphorylates tau at over 40 sites, making it the principal kinase responsible for pathological tau hyperphosphorylation[11]. Key sites include:
| Site | Effect on Tau |
|---|---|
| Thr181 | Early phosphorylation marker, CSF biomarker |
| Ser199 | Major GSK-3β site |
| Ser202 | Phosphorylated in early NFT formation |
| Thr205 | Important for tau aggregation |
| Ser212 | Co-localizes with early pathological changes |
| Thr217 | Emerging biomarker, correlates with early AD |
| Ser235 | Priming site for further phosphorylation |
| Ser396 | Major site in PHFs, affects aggregation |
| Ser404 | Modulates tau filament formation |
The sequential phosphorylation model suggests GSK-3β initiates tau hyperphosphorylation at priming sites, then propagates to additional sites in a "spread" pattern that mirrors the anatomical progression of NFT pathology in AD[12].
Multiple lines of evidence implicate GSK-3β in AD pathogenesis:
Several signaling pathways converge on GSK-3β regulation:
PI3K/Akt pathway: Akt phosphorylates GSK-3β at Ser9, inhibiting its activity. Aβ disrupts this pathway, removing a critical brake on GSK-3β[17].
Wnt/β-catenin pathway: GSK-3β is a component of the β-catenin destruction complex. Wnt signaling inhibits GSK-3β, but this pathway is dysregulated in AD[18].
MAPK pathways: ERK and p38 MAPK can regulate GSK-3β activity through phosphorylation events[19].
NMDA receptor signaling: Excitatory synaptic activity can modulate GSK-3β through calcium-dependent mechanisms[20].
CDK5 is a serine/threonine kinase with sequence similarity to cyclin-dependent kinases, but its activity is not cell-cycle dependent. Instead, CDK5 is primarily active in post-mitotic neurons due to its requirement for neuronal activators p35 and p39[21].
CDK5 phosphorylates tau at multiple sites, some overlapping with GSK-3β and some unique:
| Site | Significance |
|---|---|
| Ser202 | Overlaps with GSK-3β, early pathological marker |
| Thr205 | Important for tau conformation |
| Ser235 | Priming site |
| Ser404 | Modulates aggregation propensity |
CDK5-mediated phosphorylation at Ser202 and Thr205 produces conformationally distinct tau species that may be especially prone to aggregation[23].
PKA phosphorylates tau at multiple sites, particularly Ser214 and Ser262, with the latter being a microtubule-binding domain site[28]. PKA activity is regulated by cAMP and is responsive to neurotransmitter signaling, particularly through β-adrenergic and dopamine receptors[29].
CaMKII phosphorylates tau at Ser262 and Thr205, sites important for microtubule binding and aggregation[30]. Given CaMKII's central role in synaptic plasticity and calcium signaling, its dysregulation may link synaptic dysfunction to tau pathology.
CK1 isoforms (CK1δ, CK1ε) phosphorylate tau at multiple sites including Ser202, Thr205, and Ser409[31]. CK1 activity is increased in AD brains, and it may initiate tau phosphorylation cascades[32].
tyrosine phosphorylation of tau (particularly Tyr18, Tyr29, and Tyr394) is increasingly recognized as pathological[35]. Src family kinases including Fyn, Src, and Lck can phosphorylate these sites, and tau tyrosine phosphorylation may facilitate subsequent serine/threonine phosphorylation[36].
Multiple pharmaceutical companies have developed GSK-3β inhibitors:
Lithium: The oldest GSK-3 inhibitor, but has limited brain penetration and significant side effects at therapeutic doses[37].
Tideglusib (NP-031112): A selective GSK-3 inhibitor that reached Phase II clinical trials for AD and PSP. Results showed good safety but inconclusive efficacy[38].
AZD1080: A potent GSK-3 inhibitor that reversed memory deficits in transgenic AD mice, but was not advanced to clinical trials[39].
Given the complexity of tau kinase networks, strategies targeting multiple kinases simultaneously may be more effective:
The phosphorylation state of tau reflects the balanced activity of kinases and phosphatases. The primary tau phosphatase is protein phosphatase 2A (PP2A), which accounts for approximately 70% of tau dephosphorylation activity in the brain[43].
In AD, PP2A activity is reduced through multiple mechanisms:
The combination of increased kinase activity and decreased phosphatase activity creates a "double hit" promoting tau hyperphosphorylation[44].
PP2A is the major tau phosphatase, but protein phosphatase 1 (PP1), PP2B (calcineurin), and PP5 also contribute to tau dephosphorylation. Each of these phosphatases is affected in AD:
PP2A: Reduced activity in AD brain correlates with cognitive decline. The PP2A inhibitor SET (a phosphoprotein that accumulates in AD) contributes to reduced PP2A activity.
PP1: Involved in synaptic plasticity and memory formation. PP1 activity is modulated by dopamine and other neurotransmitters that are affected early in AD.
PP2B (Calcineurin): Calcium-activated phosphatase that dephosphorylates tau. Its activity is dysregulated by calcium homeostasis disruption in AD.
The phosphatases themselves can be regulated by kinases—PKA can phosphorylate and inhibit PP2A, creating another layer of cross-talk in the kinase-phosphatase network.
Tau phosphorylated at specific kinase-specific sites has diagnostic and prognostic value:
These biomarkers enable:
Multiple transgenic models have been developed to study tau kinase involvement:
GSK-3β transgenic models: Express mutant GSK-3β (e.g., GSK-3βS9A, a constitutively active form) under neuronal promoters. These mice develop tau hyperphosphorylation and memory deficits.
p25 inducible models: Conditional p25 overexpression leads to hyperactive CDK5, producing robust tau pathology and neurodegeneration.
APP/PSEN1 models: Express human mutant APP and PS1, producing Aβ that indirectly activates tau kinases. These models show kinase activation preceding tangle formation.
MAPT P301L models: Express mutant tau (P301L) that aggregates readily. Combining with kinase overexpression accelerates pathology.
| Compound | Company | Phase | Notes |
|---|---|---|---|
| Tideglusib | Noscendo | II | AD, PSP; safe but inconclusive efficacy[38:1] |
| AZD1080 | AstraZeneca | Preclinical | Reversed memory deficits in mice[39:1] |
| AR-014418 | Roche | I | AD; development discontinued |
| LY-2090314 | Eli Lilly | I/II | Cancer; limited CNS penetration |
| Compound | Stage | Notes |
|---|---|---|
| Roscovitine | Research | Poor brain penetration, toxic at high doses |
| Dinaciclib | Research | Multi-CDK inhibitor, limited CNS penetration |
| Peptide inhibitors | Preclinical | p5-tat, cell-penetrating peptides |
Toxicity: GSK-3β has many cellular roles; broad inhibition causes on-target/off-tumor effects including pancreatic β-cell dysfunction, stem cell differentiation effects, cardiac hypertrophy, and increased tumorigenesis risk in periphery[48].
Brain penetration: Many inhibitors fail to achieve adequate brain concentrations due to P-glycoprotein efflux.
Complexity: Single kinase inhibition may be insufficient given redundant pathways and compensatory mechanisms.
Timing: Intervention may need to occur before substantial pathology accumulates, requiring pre-symptomatic identification.
Instead of directly inhibiting GSK-3β, targeting upstream activators may provide more selective modulation:
Tau kinases are activated by neuroinflammatory processes:
Metabolic alterations affect tau kinase activity:
Synaptic activity modulates tau kinases:
Recent evidence suggests tau pathology spreads through interconnected neural networks:
The concept of tau as a prion-like protein has gained traction:
Emerging technologies reveal cell-type-specific kinase expression:
Phospho-tau species provide molecular readouts of disease stage:
| Stage | Phospho-tau Pattern | Clinical Correlation |
|---|---|---|
| Preclinical | pSer202, pThr181 | Asymptomatic, biomarker positive |
| MCI | pThr217, pSer235 | Mild cognitive impairment |
| Moderate AD | pSer396, pSer404 | Clear cognitive deficits |
| Severe AD | Multiple phosphorylated sites | Severe dementia, high NFT burden |
Phospho-tau measurements can track therapeutic efficacy:
The tau kinase signaling cascade represents a central therapeutic target in Alzheimer's disease. GSK-3β and CDK5 remain the primary targets, but the complex kinase network and compensatory mechanisms present significant challenges. Emerging strategies focusing on:
The interplay between kinases, phosphatases, aggregation mechanisms, and spread pathways creates multiple therapeutic opportunities. Successful translation will require careful patient selection, adequate brain penetration, and appropriate dosing to balance efficacy with toxicity.
Goedert et al. Tau isoforms (1989). 1989. ↩︎
Hanger et al. Tau phosphorylation sites (2009). 2009. ↩︎
Liu et al. Kinase/phosphatase balance (2008). 2008. ↩︎
Woodgett, GSK-3 (1990). 1990. ↩︎
Hooper et al. GSK-3 in brain (2008). 2008. ↩︎
Bhat et al. GSK-3 Tyr216 (2003). 2003. ↩︎
Plattner et al. Priming phosphorylation (2006). 2006. ↩︎
Avila et al. GSK-3 tau sites (2010). 2010. ↩︎
Arendt et al. Sequential phosphorylation (2016). 2016. ↩︎
Leroy et al. GSK-3 activity in AD (2002). 2002. ↩︎
Wellington et al. GSK3B polymorphisms (2000). 2000. ↩︎
Lucas et al. GSK-3 transgenic mice (2001). 2001. ↩︎
Moloney et al. Aβ activates GSK-3 (2010). 2010. ↩︎
Kitagishi et al. PI3K/Akt/GSK-3 (2012). 2012. ↩︎
Caricasole et al. Wnt signaling in AD (2003). 2003. ↩︎
Zhang et al. MAPK-GSK-3 cross-talk (2011). 2011. ↩︎
Wei et al. NMDA/GSK-3 (2010). 2010. ↩︎
Tsai et al. CDK5 p35 (1994). 1994. ↩︎
Patrick et al. p25 generation (1999). 1999. ↩︎
Ahmad et al. CDK5 tau sites (2006). 2006. ↩︎
Cruz et al. Calpain/p25 (2003). 2003. ↩︎
Shiroma et al. CDK5/GSK-3 cooperation (2005). 2005. ↩︎
Zheng et al. CDK5 inhibitors (2007). 2007. ↩︎
Liu et al. PKA tau (2006). 2006. ↩︎
Zheng et al. cAMP/PKA signaling (2010). 2010. ↩︎
Bhide et al. CaMKII tau (1996). 1996. ↩︎
Greenberg et al. CK1 tau (1994). 1994. ↩︎
Flajolet et al. CK1 in AD (2007). 2007. ↩︎
Arendt et al. MAPK activation (2000). 2000. ↩︎
Munoz & Ammit, p38 MAPK (2010). 2010. ↩︎
Lebouvier et al. Tau tyrosine phosphorylation (2009). 2009. ↩︎
Bhaskar et al. Fyn tau (2005). 2005. ↩︎
Chiu et al. Lithium GSK-3 (2000). 2000. ↩︎
del Ser et al. Tideglusib Phase II (2013). 2013. ↩︎ ↩︎
Bennett et al. AZD1080 (2017). 2017. ↩︎ ↩︎
Mairet-Coello et al. Roscovitine (2012). 2012. ↩︎
Zheng et al. CDK5/p25 inhibitor (2013). 2013. ↩︎
Vasilje et al. Combination therapy (2019). 2019. ↩︎
Liu et al. PP2A tau (2005). 2005. ↩︎
Sontag & Sontag, Kinase/phosphatase dysregulation (2006). 2006. ↩︎
Blennow et al. CSF tau biomarkers (2020). 2020. ↩︎
Janelidze et al. pThr217 blood biomarker (2020). 2020. ↩︎
Chiotis et al. pSer396 PET (2020). 2020. ↩︎
Beurel et al. GSK-3 toxicity (2015). 2015. ↩︎
Heneka et al. NLRP3 inflammasome (2015). 2015. ↩︎
Jiang et al. Diabetes and tau (2019). 2019. ↩︎
Wu et al. Tau spreading mechanisms (2016). 2016. ↩︎
Velmez et al. Spatial transcriptomics tau (2021). 2021. ↩︎