Background
Frontotemporal lobar degeneration (FTLD) is a progressive neurodegenerative disease clinically diagnosed by evidence of personality and behavioral changes and language dysfunction [
1]. Following Alzheimer’s disease (AD), FTLD is the second most prevalent form of presenile dementia affecting 10-30 per 100,000 individuals between the ages of 45 and 65 years. In general, FTLD features atrophy of the frontal and temporal lobes resulting from neuron loss [
2]. FTLD is sub-classified into three major pathological subtypes based on the presence of aggregated protein deposits of either TDP-43 or tau inclusions, with a small subset of cases exhibiting FUS-related pathology [
2]. FTLD-TDP accounts for roughly 50% of cases whereas FTLD-tau accounts for approximately 45% of cases [
2]. FTLD-TDP presents with aberrantly processed, ubiquitinated, and phosphorylated TDP-43 in neuronal inclusions and dystrophic neurites. FTLD-tau, on the other hand, is characterized by hyperphosphorylated aggregates of tau, which form tangles and pick bodies in neurons, glia, and neurites. Several mutations in
MAPT reduce tau’s affinity for microtubules and increase its aggregation rate and can cause FTLD-tau [
3].
Phosphorylation of TDP-43 at serine residues 409 and 410 remains a consistent pathological feature in ALS and FTLD-TDP [
4]. Phosphorylation of TDP-43 reduces TDP-43 protein turnover, increases cellular mislocalization of TDP-43, drives protein aggregation, and promotes neurodegeneration [
5‐
9]. Likewise, hyperphosphorylated tau protein is a hallmark of several neurodegenerative disorders including AD, progressive supranuclear palsy (PSP), corticobasal degeneration (CBD), and FTLD. Phosphorylated tau has been implicated in the formation of toxic tau aggregates that promote neurodegeneration [
10‐
15].
The kinases TTBK1 and TTBK2 have been implicated in a number of neurodegenerative diseases. TTBK1 protein is increased in AD and influences the aggregation of tau [
16,
17]. Furthermore, transgenic mouse lines expressing full-length human TTBK1 exhibit age-dependent detriments consistent with neurodegeneration, including learning impairment, neurofilament aggregation, microgliosis, altered CDK5/p35 activity, and decreased expression of NMDA receptors [
18]. A double-transgenic mouse model expressing FTLD mutant tau and human TTBK1 shows increased accumulation of oligomeric tau and enhanced motor neuron loss, suggesting a direct role of TTBK1 in accelerating tau-related neurodegeneration [
19]. Mutations in TTBK2 cause spinocerebellar ataxia type 11, a disorder exhibiting both loss of Purkinje cells and widespread deposition of tau [
20]. TTBK2 plays an essential role in the initiation of ciliogenesis during embryonic development through the regulation of microtubule dynamics [
21‐
23]. TTBK1 is solely expressed in the CNS and reproductive tissues, whereas TTBK2 is ubiquitously expressed throughout many tissues including liver, skeletal muscle, pancreas, heart, and brain [
24].
Purified recombinant human TTBK1 and TTBK2 can directly phosphorylate both TDP-43 at S409/410 [
9] and tau at S198, S199, S202, and S422 [
17] and S208 and 210 [
25]. Additionally, both TTBK1 and TTBK2 co-localize with phosphorylated TDP-43 (pTDP) in human postmortem tissues from both FTLD and ALS cases [
9] and phosphorylated tau (ptau) in AD cases [
16]. Since TTBK1/2 phosphorylate both TDP-43 and tau, a common pathway could be involved in the initiation of FTLD. While TTBK1/2 directly phosphorylate TDP-43 and tau in vitro
, it remains unclear how TTBK1/2 activity in vivo influences disease onset and progression.
To examine how TTBK1/2 contribute to both TDP-43 and tau phosphorylation, we analyzed their effects on lifespan, proteostatic function, and neurodegeneration in the context of tau or TDP-43 transgenic animal models. We also examined the expression of TTBK1 and TTBK2 in post-mortem human brain. We demonstrate that TTBK1/2 kinase expression leads to significant neurodegenerative phenotypes in our transgenic tau and TDP-43 models and is also a consistent feature of FTLD-tau and FTLD-TDP.
Methods
C. elegans Strains
The N2 (Bristol) strain of
C. elegans was used for all experimental controls and maintained as described [
26]. Strains were maintained at 20 °C on OP50 seeded nematode growth media (NGM). All experiments were performed at room temperature unless otherwise designated. Construction and characterization of TDP-43 transgenic (tg) (CK410) and tau (high-expression) (CK144) lines used were described previously [
8]. Tau (low-expression) (CK1044) tg strains were generated by introducing wild type human full length 1N4R splice isoform tau cDNAs driven by the pan neuronal
aex-3 promoter (
Paex-3::Tau) into the
C. elegans genome.
TTBK1 and TTBK2 kinase domain (hTTBK1-cat and hTTBK2-cat) strains were constructed by introducing human TTBK1 or TTBK2 cDNA encoding the kinase domain, driven by the pan neuronal
rgef-1 promoter (
Prgef-1::hTTBK1cat,
Prgef-1::hTTBK2cat) (TTBK1:CK1051; TTBK2:CK646, CK645) into the
C. elegans genome.
Prgef-1::hTTBK1cat was microinjected into N2 at a concentration of 30 ng/μl with an
elt-2::mCherry coinjection marker at a concentration of 25 ng/μl.
Prgef-1::hTTBK2cat was microinjected into N2 at a concentration of 50 ng/μl with an
elt-2::mCherry coinjection marker at a concentration of 25 ng/μl. For all transgenic strains, extrachromosomal arrays were then integrated by exposing animals to a dose of 4000 Rad Gamma rays and subsequently outcrossed back into the N2 background at least twice. TTBK1 and TTBK2 strains used were CK1051, CK645, CK646. Strain CZ1200 [
27], which carries an integrated
Punc25::GFP transgene marker in GABAergic motor neurons, was a generous gift from Dr. Y. Jin. CK1051 and CK646 were crossed with CK1044, CK144, and CK410 to generate homozygous double transgenic
C. elegans. CK646/CK1044 and CK1051/CK144 double transgenic were subsequently crossed with CZ1200 to produce triple transgenic lines with GFP marked GABAergic neurons.
Radial locomotion assay
Behavior was assessed by placing 15-20 age-matched (L4) C. elegans at the center of a 150 mm NGM plate supplemented with 5× peptone (5xPEP), with a uniform OP50 bacterial lawn. After 1 h of free movement at room temperature, the radial distance traveled from the origin by each animal was measured. Distance from the origin traveled per unit of time was expressed in micrometers per second to give a radial velocity. The assay was performed in triplicate by an observer blinded to genotype and statistical analyses were performed using GraphPad Prism software.
Lethality assay
CK144 was crossed with CK646 or CK645 and made homozygous for either the tau or hTTBK2-cat transgene to create the F1 population. From a single F1 parent, each F2 progeny was isolated onto an individual plate and the subsequent F3 offspring were scored for transgene expression. Animals that did not survive past L2 were characterized as larval lethal. Animals that survived into adulthood but did not produce progeny or died prior to egg laying were classified as adult sterile. A chi-squared analysis was performed to assess significance.
Neurodegeneration assays
Strains with Punc25::GFP-tagged (GABA)-ergic motor neurons were generated by crossing to the reporter strain CZ1200. Strains were staged to day 1 of adulthood and immobilized on a 2% agarose pad with 0.01% sodium azide. Live VD and DD GABAergic neurons were assessed under fluorescent microscopy on DeltaVision Elite (Applied Precision, Issaquah, WA) imaging system using an Olympus 60× oil objective. The number of live neurons, number of dorsal cord gaps, and percentage of neurons with aberrantly branched neuronal commissures were scored. Statistical significance was analyzed by performing a One-way ANOVA with a Tukey’s post-hoc test using GraphPad Prism statistical software.
Lifespan analysis
C. elegans were synchronized to L4 stage from a timed egg lay on NGM plates at 20 °C. Lifespan plates were prepared from 30 mm NGM plates that were seeded with 10X concentrated OP50 and treated with 10 mg/ml 5-fluorodeoxyuridine (FUDR). One hundred animals per strain were assayed at 25 °C. Animals were checked daily for signs of movement by observing locomotion and pharyngeal pumping. Towards the end of life, animals were tapped lightly with a platinum pick to look for a response. Animals that failed to respond were counted as dead and removed from the plate. Animals that died of bursting, bagging, or mishandling were censored from the data. Statistical significance was analyzed by performing a Chi-squared analysis with a Mantel-Cox test using GraphPad Prism software.
Immunoblotting
Mixed-stage populations of
C. elegans were grown on 150 mm 5XPEP plates, washed with M9 buffer, and frozen with liquid nitrogen. Protein lysates were prepared by sonication of frozen
C. elegans pellets in lysis buffer (10 mM Tris-HCL pH 7.5, 5 mM EDTA, 10% sucrose) at 70% amplitude for 10 s, repeated three times. The lysate was loaded and resolved on precast 4-15% gradient SDS-PAGE gels (BioRad) and transferred to PVDF membrane (Bio-Rad Immun-blot PVDF membrane) at 100 V for 32 m. Human TDP-43 was detected with the commercially available monoclonal antibody ab57105 (Abcam, 1:2500) directed at human TDP-43 amino acids 1-261. TDP-43 phosphorylated at S409/S410 was detected by a commercially available monoclonal antibody (Cosmobio, Catalog # TIP-PTD-M01, 1:1000). Total Tau was detected with a pan-tau polyclonal antibody rb17025 (V. Lee lab, 1:3000) [
28]. Tau phosphorylated at T181 was detected by AT270 (Thermo Scientific; 1:15,000). Tau phosphorylated at S202 was detected by CP13 (1:500), a generous gift from Dr. Peter Davies (Albert Einstein College of Medicine, Bronx, NY). Tau phosphorylated at T231 was detected by AT180 (Thermo Scientific, 1:2000). Tau phosphorylated at S396/404 was detected by PHF-1 (P. Davies lab, 1:2000) [
29]. Load controls were detected by probing for β-Tubulin as previously described [
14].
Human post-mortem brain lysate was prepared by homogenization of tissue in lysis buffer (50 mM HEPES pH 7.5, 1 mM EDTA, 150 mM NaCl, 10% Glycerol, 0.1% Triton X-100, 1 mM PMSF, 1 protease inhibitor pellet (Roche cOmplete Mini)) followed with a 10 s sonication at 50% amplitude using a Branson Sonifier with micro tip. Total protein lysate was loaded in 5XSDS buffer (5% SDS, 200 mM DTT, 50 mM Tris pH 6.8, 5 mM EDTA, 50% sucrose, 0.05% Bromophenol Blue) and resolved on a precast 4-15% gradient SDS-PAGE gel (BioRad) at 200 V and transferred to PVDF membrane (Bio-Rad Immun-blot PVDF membrane) at 100 V for 30 m. TTBK1 was detected with the commercially available polyclonal antibody directed at N-terminal amino acids 240-270 of human TTBK1 (Abcam, ab103944, 1:1000). TTBK2 was detected with the commercially available antibody directed at the N-terminus (Abcam, ab67839).
Post-mortem human tissue
We obtained de-identified samples of postmortem tissue from the University of Washington Alzheimer’s Disease Research Center (ADRC) Neuropathology Core (PI, Dr. C. Dirk Keene) after receiving human subjects approval (University of Washington human subjects division approval: HSD# 06-0492-E/A 01). FTLD cases were selected on the basis of having an autopsy-confirmed diagnosis of FTLD-tau or FTLD-TDP. Control samples were from neurologically healthy control participants, who were of a similar age and were confirmed to be negative for neuropathologic changes of FTLD-tau or FTLD-TDP using routine and immunohistochemical assays. Frontal cortex (prefrontal middle frontal gyrus) and hippocampus (at the level of the lateral geniculate nucleus) samples were dissected at the time of autopsy in coronally sliced brains fixed approximately 3 weeks in 10% neutral buffered formalin according to routine protocols. Samples were processed and embedded in paraffin according to standard protocols.
Immunohistochemistry and Immunofluorescence
Formalin-fixed, paraffin-embedded human brain tissue samples were sectioned by the University of Washington Alzheimer’s Disease Research Center neuropathology core (Seattle, WA) onto standard charged glass microscope slides. Primary antibodies used for immunohistochemistry were anti-TTBK1 (Abcam, 1:100) and anti-TTBK2 (Abgent, 1:200). In order to minimize variability, sections from all cases (normal and affected subjects) were stained simultaneously for each antibody. Briefly, 5 μm sections from the frontal cortex and hippocampus were deparaffinized in xylene, rehydrated through graded alcohols, and an antigen retrieval step consisting of autoclaving sections in citrate buffer (1.8 mM citric acid/ 8.2 mM sodium citrate) was performed. Sections were treated for endogenous peroxidases with 3% hydrogen peroxide, blocked in 5% milk, incubated with primary antibody overnight at 4 °C, followed by biotinylated secondary antibody for 45 min at room temperature. Finally, sections were incubated in an avidin-biotin complex (Vector’s Vectastain Elite ABC kit, Burlingame, CA) and the reaction product was visualized with 0.05% diaminobenzidine (DAB)/0.01% hydrogen peroxide in PBS. Specificity of these antibodies has been previously shown [
9]. Immunohistochemistry photomicrographs were taken with a digital camera and imported into Adobe Photoshop for mounting. To optimize visualization of staining, photomicrographs were modified when necessary by adjusting brightness and contrast.
For double label immunofluorescence experiments, sections were co-immunostained with TTBK1 or TTBK2 and pathological tau as detected by pT231 specific monoclonal antibody AT180 (ThermoScientific). AlexaFluor 647 goat anti-mouse and 568 goat anti-rabbit secondary antibodies (Molecular Probes) were used and autofluorescence was quenched with 0.1% Sudan Black. Microscopy was performed on a Delta Vision microscope (GE, Inc) using a 60× or 100× oil immersion objective, a sCMOS camera, and 2 × 2 binning. Image analysis was performed using softWoRx 6.0 Beta software (GE, Inc). Human brain samples stained for AT180 and TTBK1 or TTBK2 were imaged on a Leica TCS SP5 II confocal microscope using a 63× oil immersion objective. Colocalization analysis of confocal images were conducted in ImageJ 1.51n using Coloc 2.
Discussion
Previous work has shown that TTBK1 and TTBK2 phosphorylate both tau [
17,
25] and TDP-43 [
30]. However, the full extent to which tau and TDP-43 are phosphorylated by TTBK1 and TTBK2 in vivo have yet to be described. It remains unknown whether changes in TTBK1 and TTBK2 activity, abundance, and proteolytic processing influence tau- and TDP-43-proteinopathies. To address these gaps in knowledge, we generated transgenic
C. elegans expressing active human TTBK1 and TTBK2 kinase domains, and evaluated their effects on tau and TDP-43 transgenic models of FTLD. We also characterized TTBK1 and TTBK2 expression patterns and levels in both FTLD-tau and FTLD-TDP subtypes.
We found that hTTBK1-cat expression dramatically increased total and phosphorylated protein levels of human tau and TDP-43 in transgenic C. elegans resulting in exacerbated behavioral phenotypes and neurodegeneration. Interestingly, while hTTBK2-cat expression drove accumulation of total and phosphorylated tau and behavioral defects in tau transgenic animals, hTTBK2-cat was relatively neutral in TDP-43 transgenic animals. This suggests an in vivo selectivity of TTBK1 and TTBK2 towards their phosphorylation targets that differs from their in vitro ability to phosphorylate purified TDP-43. Furthermore, these data could also reflect regional or neuronal subtype selectivity by TTBK2.
Although increases in TTBK1 expression have been previously shown in human AD cases [
16], there have been no studies examining changes in TTBK1 or TTBK2 abundance in FTLD. In this study we show that both TTBK1 and TTBK2 protein levels increase as compared to age-matched controls in FTLD-TDP and FTLD-tau cases. These findings were seen in both immunoblot analyses and immunohistochemistry studies. These results suggest that changes in TTBK1 and TTBK2 abundance or processing may influence their kinase activities towards tau and TDP-43 in FTLD-tau and FTLD-TDP.
Importantly, this study suggests distinct target selectivity between TTBK1 and TTBK2 kinase activity despite high homology between the kinase domains (88% identity and 96% similarity) [
24]. The differential regulation of TTBK1 vs TTBK2 activity is poorly understood. However, here we show that TTBK2 kinase domain has a greater influence on promoting tau-induced neurodegeneration than TTBK1, but it does not appear to affect TDP-43 in vivo. Likewise, TTBK1 is able to phosphorylate both TDP-43 and tau in vivo, but appears to have a relatively modest effect on tau compared to TTBK2. Investigations into the regulation of TTBK1 and TTBK2 kinase activity and substrate specificity are important next steps in determining the roles of TTBK1 and TTBK2 in FTLD.
Conclusions
The identification of TTBK1 and TTBK2 as both tau and TDP-43 kinases indicates a possible shared mechanism for the initiation of TDP-43 proteinopathy and tauopathy in FTLD. This is supported by the pathological presence of either phosphorylated TDP-43 or tau in the majority of FTLD cases, and the elevated protein expression of TTBK1 and TTBK2 in both FTLD-tau and FTLD-TDP. Therefore, the development of drugs to selectively inhibit TTBK1 and TTBK2 may be a common therapeutic strategy for both FTLD-tau and FTLD-TDP. In general, kinases have become among one of the most important classes of drug targets [
31]. TTBK1 in particular makes an attractive drug target due to its restricted expression in neurons [
24], unlike other tau kinases such as GSK3 and CDK5 that are ubiquitously expressed [
32,
33]. Furthermore, TTBK1 is one of the first kinases to be identified as both a tau and TDP-43 kinase and our results suggest that TTBK1 kinase activity is able to induce neurodegenerative phenotypes in both tau and TDP-43 backgrounds. TTBK1 selective kinase inhibitors therefore represent a potential means to treat both FTLD-tau and FTLD-TDP.
Acknowledgements
We thank the reviewers for useful suggestions and commentary. We thank Elaine Loomis, Aleen Saxton, Samantha Rice, and Kaili Chickering for exceptional technical assistance. We thank Allison Beller for outstanding administrative assistance. We thank Virginia Lee for the generous gift of the pan-neuronal tau antibody. We thank Peter Davies for the generous gift of the CP13 and PHF1 ptau antibodies. We thank Dr. Yishi Jin for the CZ1200 C. elegans strain. We thank the Developmental Studies Hybridoma Bank (NICHD) for the β-Tubulin antibody E7. We thank WormBase (WS251) for C. elegans genetic information. We thank the CGC for provision of some C. elegans strains, which is funded by the NIH Office of Research Infrastructure Programs (P40 OD010440).