Elevation of casein kinase 1ε associated with TDP‐43 and tau pathologies in Alzheimer's disease

Abstract

Alzheimer's disease (AD) is characterized by the presence of extracellular amyloid β plaques and intraneuronal neurofibrillary tangles of hyperphosphorylated microtubule‐associated protein tau in the brain. Aggregation of transactive response DNA‐binding protein of 43 kDa (TDP‐43) in the neuronal cytoplasm is another feature of AD. However, how TDP‐43 is associated with AD pathogenesis is unknown. Here, we found that casein kinase 1ε (CK1ε) phosphorylated TDP‐43 at Ser403/404 and Ser409/410. In AD brains, the level of CK1ε was dramatically increased and positively correlated with the phosphorylation of TDP‐43 at Ser403/404 and Ser409/410. Overexpression of CK1ε promoted its cytoplasmic aggregation and suppressed TDP‐43‐promoted tau mRNA instability and tau exon 10 inclusion, leading to an increase of tau and 3R‐tau expressions. Levels of CK1ε and TDP‐43 phosphorylation were positively correlated with the levels of total tau and 3R‐tau in human brains. Furthermore, we observed, in pilot immunohistochemical studies, that the severe tau pathology was accompanied by robust TDP‐43 pathology and a high level of CK1ε. Taken together, our findings suggest that the elevation of CK1ε in AD brain may phosphorylate TDP‐43, promote its cytoplasmic aggregation and suppress its function in tau mRNA processing, leading to acceleration/exacerbation of tau pathology. Thus, the elevation of CK1ε may link TDP‐43 to tau pathogenesis in AD brain.

Abbreviations

AD

Alzheimer's disease

ALS

amyotrophic lateral sclerosis

CDK5

cyclin‐dependent kinase 5

CK1ε

casein kinase 1ε

Dyrk1A

dual‐specificity tyrosine/Y‐phosphorylation regulated kinase 1A

EC

entorhinal cortex

FTLD‐U

frontal temporal lobar degeneration with ubiquitin inclusion

GSK‐3β

glycogen synthase kinase‐3β

NFT

neurofibrillary tangle

PDPK

proline‐directed protein kinase

TDP‐43

transactive response DNA‐binding protein of 43 kDa

UTR

untranslated region

Introduction

Neuronal microtubule‐associated protein tau is hyperphosphorylated and aggregated to form neurofibrillary tangles in the brains of individuals with Alzheimer's disease (AD) 21, 22 and related neurodegenerative disease, termed tauopathies 19, 38, 55. The numbers of NFTs are correlated with the clinical symptoms of AD 1, 6, suggesting pathological changes of tau may be pivotal for neurodegeneration of AD 29.

In AD brain, tau lesion is initiated in the entorhinal cortex (EC) and follows a stereotypical pattern to progressively propagate to the limbic system and eventually to widespread isocortex regions 7, 8. Cerebellum expresses 1/4 of tau seen in the cortices 27 and is lack of tau pathology 37. Furthermore, tau pathology has been developed only in the mice with overexpression of wild‐type or mutated human tau up to many folds of the endogenous murine tau 3, 15, 46, 51, 52, 61, but has not been observed in wild‐type rodents 47. These data suggest that the level of tau expression may be critical for the development of tau pathology. Inhibition of tau expression in some mouse models protects them against cognitive impairment 50, 52, 57.

Adult normal human brain expresses six tau isoforms by alternative splicing of exons 2, 3 and 10 of its pre‐mRNA 20. Exon 10 encodes the second microtubule‐binding repeat and its alternative splicing generates tau isoforms with three or four microtubule‐binding repeats, named 3R‐tau or 4R‐tau, which are under developmental and cell type‐specific regulation 20, 36. In normal adult human brain, approximately equal levels of 3R‐ and 4R‐tau are expressed 18. It is well known that alteration in the 3R‐tau/4R‐tau ratio, as a result of dysregulation of alternative splicing of tau exon 10, is sufficient to trigger neurodegeneration in frontotemporal dementia, Pick's disease, corticobasal degeneration and progressive nuclear palsy 17, 29.

Transactive response DNA‐binding protein of 43 kDa (TDP‐43) encoded by the TARDBP gene is originally found to associate with frontal temporal lobar degeneration with ubiquitin inclusion (FTLD‐U) and amyotrophic lateral sclerosis (ALS) 4, 44. TDP‐43 is aggregated to form cytoplasmic inclusion in affected neurons in these diseases 25. Consequently, TDP‐43 pathology has been reported in a number of other neurodegenerative diseases, many associated with tau pathology, including AD 4, 26, 34, 56. TDP‐43 proteinopathy has been observed in up to 2/3 of sporadic cases of AD 33, 34. TDP‐43 is a DNA‐ and RNA‐binding protein and involves in mRNA processing 9. We recently reported that TDP‐43 promotes tau mRNA instability and tau exon 10 inclusion 23, 24. Overexpression of TDP‐43 decreases Aβ plaque deposition and increases abnormal tau aggregation in vivo 14. Thus, pathological changes of TDP‐43 may accelerate/exacerbate tau pathology via dysregulation of tau mRNA processing.

Pathological TDP‐43 in inclusion is hyperphosphorylated 5, 25, 28, 43. Several kinases, including casein kinase 1 (CK1) 25, 35, CK2 11, CDC7 40 and TTBK1/2 41, are able to phosphorylate TDP‐43 in vitro and in vivo and promote its pathological accumulation and neurotoxicity 13, 39, 41, 45. The level of CK1ε was increased dramatically in AD brain 12, 16. However, the impact of CK1ε‐TDP‐43 in tau mRNA processing and tau pathogenesis remains unclear. In the present study, we show that CK1ε phosphorylates TDP‐43, promotes its cytoplasmic aggregation and suppresses its function in promoting tau mRNA instability and tau exon 10 inclusion. Increased level of CK1ε is associated with TDP‐43 phosphorylation, levels of total tau and 3R‐tau and TDP‐43 and tau proteinopathies in AD brains. These results suggest that the elevation of CK1ε is associated with TDP‐43 and tau proteinopathies.

Materials and Methods

Plasmids, reagents and antibodies

pCI/TDP‐43 tagged with haemagglutinin (HA), pCI/CK1ε tagged with HA and Flag and pEGFP/tau 3′‐UTR were constructed as described previously 12, 24. pCI/SI9‐LI10 containing a tau minigene, SI9‐LI10, comprising tau exons 9, 10, 11 and part of intron 9 and the full‐length of intron 10 was a gift from Dr Jianhua Zhou of the University of Massachusetts Medical School 62. pGEX‐6P1/TDP‐43 was constructed by PCR from pCI/TDP‐43. All the constructs were confirmed by DNA sequence analysis. CK1ε‐specific inhibitor PF4800567 (3‐(3‐chloro‐phenoxymethyl)‐1‐(tetrahydro‐pyran‐4‐yl)‐1H‐pyrazolo[3,4‐d]pyrimidin‐4‐ylamine) was purchased from Tocris (Minneapolis, MN, USA). Anti‐CK1ε (H60), anti‐CK1ε (4D7), anti‐CDK5 and anti‐GAPDH were purchased from Santa Cruz Biotechnology Inc (Santa Cruz, CA, USA). Polyclonal anti‐TDP‐43 (A260), anti‐GSK3β and anti‐GFP were purchased from cell signaling technology (Danvers, MA, USA). Monoclonal antibody 8D9 was raised against a histidine‐tagged protein containing the first 160 residues of rat Dyrk1A 59. Polyclonal anti‐pS379‐TDP‐43, anti‐pS403/404‐TDP‐43 and anti‐pS409/410‐TDP‐43 were produced by Abmart company (Shanghai, China). For immunohistochemistry, polyclonal anti‐TDP‐43 was from the Proteintech Group Inc (Chicago, IL, USA), polyclonal anti‐pS409/410‐TDP‐43 was from Cosmo Bio Co. LTD (Koto‐Ku, Tokyo, Japan) and mouse monoclonal anti‐pS202/pT205‐tau (AT8) was from ThermoFisher Scientific (Rockford, IL, USA). Anti‐HA, cocktail for bacteria and GST beads were purchased from Sigma‐Aldrich Corp (St Loius, MO, USA). Peroxidase‐conjugated anti‐mouse and anti‐rabbit IgG were purchased from Jackson Immuno Research Laboratories (West Grove, PA, USA). Oregon Green 488‐conjugated goat anti‐rabbit IgG, Alexa Fluor 555‐conjugated goat anti‐mouse IgG and TO‐PRO‐3 iodide (642/661) were purchased from Invitrogen (Invitrogen, Carlsbad, CA, USA).

Human brain tissue

Frontal cortices of ten AD and nine age‐matched normal human brains used in this study (Table 1) were obtained without the identification of donors from the Sun Health Research Institute Donation Program (Sun City, AZ, USA). All brain samples were histopathologically confirmed and stored at −80°C until used. Neurofibrillary pathology was staged according to Braak and Braak 7. Tangle and plaque scores were a density estimate and were designated none, sparse, moderate or frequent (0, 1, 2 or 3 for statistics), as defined according to CERAD AD criteria. Five areas (frontal, temporal, parietal, hippocampal and entorhinal) were examined and the scores were added up for a maximum of 15. The use of frozen human brain tissue was in accordance with the National Institutes of Health guidelines and approved by our institutional review committee. The tissue was homogenized in cold buffer consisting of 50 mM Tris‐HCl, pH 7.4, 8.5% sucrose, 2.0 mM EDTA, 10 mM β‐mercaptoethanol, 1.0 mM orthovanadate, 50 mM NaF, 1.0 mM 4‐(2‐aminoethyl)benzenesulfonyl fluoride hydrochloride (AEBSF) and 10 μg/mL each of aprotinin, leupeptin, and pepstatin and stored at −80°C for Western blots analysis.

Table 1.

Human brain tissue of Alzheimer's disease (AD) and control (Con) cases used in this study.

Case	Age at death (years)	Gender	PMI (h)	Braak stage†	Tangle scores‡	Plaque scores‡
AD1	83	F	3.0	VI	12.4	13.4
AD2	74	M	2.75	VI	14.66	10.25
AD3	79	F	1.5	VI	14.66	14.0
AD4	73	F	2.0	V	15.00	14.0
AD5	81	M	3.0	V	11.00	12.5
AD6	76	M	2.33	VI	15.00	12.5
AD7	72	M	2.5	VI	15.00	14
AD8	74	F	2.83	VI	15.00	14.4
AD9	76	M	4.0	V	15.00	11.75
AD10§	78	M	1.83	VI	15.00	13.25
Mean ± S.D.	76.6 ± 3.60		2.57 ± 0.71		14.27 ± 1.40	13.00 ± 1.28
Con1	85	F	2.75	II	5.0	2
Con2	82	F	2.0	II	4.25	0.25
Con3	70	F	2.0	I	0	8.25
Con4	73	M	2.0	III	2.75	0
Con5	78	M	1.66	I	0	0
Con6	80	M	3.25	II	2.75	2.5
Con7	80	M	2.16	I	1.0	12.75
Con8	83	F	3.25	II	0.75	1
Con9	82	F	2.25	II	3.50	0
Mean ± S.D.	79.22 ± 4.87		2.37 ± 0.58		2.22 ± 1.86	2.97 ± 4.51
AD11¶	80	F	2.25	VI	14.5	12.5
AD12¶	85	F	1.66	V	12.0	12.25
Con10¶	85	M	2.5	II	4.25	0

^†Neurofibrillary pathology was staged according to Braak and Braak 7.

^‡Tangle and plaque scores were a density estimate and were designated none, sparse, moderate or frequent (0, 1, 2 or 3 for statistics), as defined according to CERAD AD criteria. Five areas (frontal, temporal, parietal, hippocampal and entorhinal) were examined and the scores were added up for a maximum of 15.

^§Case was not used in the tau measurement by immuno‐dot blot.

^¶Cases were used for the immunohistochemical study.

Abbreviation: PMI = postmortem interval.

Cell culture and transfection

Human embryonic kidney cell line (HEK‐293FT) and human cervix epithelia cell line (HeLa) were cultured in Dulbecco's modified Eagle's medium (DMEM) supplemented with 10% fetal bovine serum (FBS) (Invitrogen), 100 U/mL penicillin and 100 μg/mL streptomycin and incubated in a humidified atmosphere containing 5% CO₂ at 37°C. Cells were seeded to culture plates and all transfections were performed in triplicate with FuGENE HD (Promega, Madison, WI, USA) or Lipofectamine^TM 2000 (Invitrogen) according to the manufacturer's instructions.

For the PF4800567 treatment, HEK‐293FT cells were transfected with corresponding plasmids for 40 h and then treated with various concentrations of PF4800567 for 8 h. The cells were harvested for biochemical studies.

Expression and purification of recombinant GST‐TDP‐43 fusion protein from Escherichia coli

pGEX6P1/TDP‐43 cultured in LB medium was induced with 0.4 mM IPTG after OD₆₀₀ was 1.0 for 4 h at RT, then cells were centrifuged at 5000 × g for 10 minutes at 4°C and the pellet was resuspended and sonicated in phosphate‐buffered saline (PBS) with 1 mM dithiothreitol (DTT) and cocktail for bacteria (1:1000). Then, the lysate was added to detergents 1% Triton X‐100, 0.1% NP40, 0.1% sodium deoxycholate and 0.1% SDS, agitated at 4°C for 30 minutes, then centrifuged at 3000 × g for 15 minutes. The supernatant was incubated with glutathione‐Sepharose beads for 1 h at 4°C. The beads were washed with Tris‐buffered saline (TBS, 50 mM Tris‐HCl, pH 7.4, 150 mM NaCl) and eluted with elution buffer (10 mM reduced glutathione in 50 mM Tris‐HCl, pH 8.5). The purified protein was dialyzed against 5 mM MES, pH 6.8, 0.5 mM EGTA and lyophilized.

GST pulldown

GST and GST‐TDP‐43 were purified by affinity purification with glutathione‐Sepharose but without elution from the beads. The GST or GST‐TDP‐43 bead was incubated with crude extract from rat brain homogenate in buffer (50 mM Tris‐HCl, pH 7.4, 8.5% sucrose, 50 mM NaF, 1 mM Na₃VO₄, 50 nM okadaic acid, 0.1% Triton X‐100, 2 mM EDTA, 1 mM PMSF, 10 μg/mL aprotinin, leupeptin and pepstatin). After 4 h incubation at 4°C, the beads were washed with washing buffer (50 mM Tris‐HCl, pH 7.4, 150 mM NaCl, 0.1% Triton X‐100, 1 mM PMSF, 2 μg/mL aprotinin, leupeptin, pepstatin and 1 mM DTT) six times, and then the bound proteins were eluted by boiling in 2×Laemmli sample buffer and the samples were subjected to Western blot analysis.

Phosphorylation of TDP‐43 by CK1ε in vitro

GST‐TDP‐43 was phosphorylated by immunopurified CK1ε in the reaction buffer (50 mM Tris‐HCl, pH 7.4, 2 mM MgCl₂, 10 mM β‐ME, 1 mM EGTA and 0.2 mM ATP) at 30°C for 0, 0.5, 1 or 2 h. The reaction was stopped by boiling with an equal volume of 2×Laemmli sample buffer. The reaction products were analyzed by Western blot developed with site‐specific and phosphorylation‐dependent anti‐TDP‐43 antibodies.

Immunoprecipitation

TDP‐43 or CK1ε tagged with HA was overexpressed separately in HEK‐293FT cells. The cells were washed with PBS and lysed in IP lysis buffer (50 mM Tris‐HCl, pH 7.4, 150 mM NaCl, 0.2% sodium deoxycholate, 0.1% NP‐40, 0.1% Triton X‐100, 1 mM Na₃VO₄, 50 mM NaF, 2 mM EDTA, 1 mM AEBSF, and 10 μg/mL of aprotinin, leupeptin and pepstatin) on ice for 30 minutes, then centrifuged at 15 000 × g, 4°C for 5 minutes. The supernatant was incubated with protein G beads precoupled with monoclonal anti‐HA at 4°C for 4 h. The beads were washed with lysis buffer twice and with 50 mM Tris‐HCl (pH 7.4) twice and the bound CK1ε was used for above in vitro phosphorylation assay and bound TDP‐43 was analyzed by Western blot developed with site‐specific and phosphorylation‐dependent TDP‐43 antibodies.

Immunofluorescent staining of cultured cells

HeLa cells were plated in poly‐d‐lysine precoated 24‐well plate onto coverslips one day before transfection. The cells were transfected with pCI/TDP‐43·HA and pCI/CK1ε‐HA·Flag for 48 h with Lipofectamine^TM 2000 (Invitrogen), then the cells were fixed with 4% paraformaldehyde in PBS for 30 minutes at room temperature. After washing with PBS, the cells were blocked with 10% goat serum in 0.2% Triton X‐100/PBS for 2 h at 37°C after permeabilization, and then incubated with polyclonal anti‐TDP‐43 (Proteintech, 1:1000) and monoclonal anti‐CK1ε (Santa Cruz, 4D7, 1:1000) overnight at 4°C. The cells were washed with PBS and incubated for 2 h with secondary antibody (Oregon Green 488‐conjugated goat anti‐rabbit IgG and Alexa Fluor 555‐conjugated goat anti‐mouse IgG, 1:500) plus TO‐PRO‐3 iodide (1:1000, ThermoFisher Scientific) at room temperature. After washing with PBS, the cells were mounted with antifade mounting media (Roche, Diagnostic, USA) and observed using a Nikon EZ‐C1 laser scanning confocal microscope.

Immunohistochemistry

Free‐floating sections, 40 μm in thickness, of the frontal cortex of AD and control cases were washed with three changes of PBS, subsequently subjected to permeabilization and removal of endogenous peroxidase activity, blocked with normal goat serum and then incubated with primary antibodies overnight at 4°C. The primary antibodies used were rabbit polyclonal anti‐TDP‐43 (0.3 μg/mL, Proteintech), rabbit polyclonal anti‐pS409/410‐TDP‐43 (1:1000, Cosmo Bio Co. LTD) and mouse monoclonal anti‐pS202/pT205‐tau (AT8, 0.2 μg/mL, ThermoFisher Scientific). After washing in PBS, the sections were incubated with species‐matched horseradish peroxidase‐conjugated secondary antibodies (1:2000, Jackson ImmunoResearch) at room temperature for 2 h, washed and developed in 0.05% diaminobenzidine (Sigma) plus 0.015% hydrogen peroxide (Fisher Scientific). Sections were mounted on microscopic slides, air‐dried overnight, rehydrated and counterstained with Mayer's hematoxylin (Sigma), dehydrated in ascending concentrations of ethanol, cleared in Histoclear (National Diagnostics), and coverslipped. Photomicrographs were taken using a Nikon 90i digital microscope.

Double‐labeling immunofluorescence was employed to assess colocalization of tau pathology with TDP‐43 pathology. Immunofluorescence staining was performed using a protocol similar to that described above. The primary antibodies were mixed AT8 and rabbit anti‐TDP‐43, and secondary antibodies were mixed Alexa Fluor 555‐conjugated goat anti‐mouse IgG (1:1200) and Oregon Green 488‐conjugated goat‐anti‐rabbit IgG (1:1200). After incubation with secondary antibodies, sections were washed and incubated with TO‐PRO‐3 Iodide (1:1000, ThermoFisher Scientific) at room temperature for 15 minutes. After washing twice in PBS, sections were mounted on microscopic slides and coverslipped with anti‐fade mounting medium. Photomicrographs were taken using a Nikon EZ‐C1 laser scanning confocal microscope.

In both chromogen‐based immunohistochemistry and immunofluorescence staining, nonimmunized normal goat serum in replace of the primary antibody was used as a negative control.

Western blot and immuno‐dot blot

For Western blot, cultured cells were lysed using 1×Laemmli SDS sample buffer directly and brain homogenates were diluted using 2×Laemmli SDS sample buffer at 1:1 ratio, followed by heating at 100°C for 5 minutes. The brain homogenates were a transparent solution and the proteins were well dissolved. Samples were subjected to SDS‐PAGE and electrically blotted onto polyvinylidene fluoride membrane (Millipore). The membrane was subsequently blocked in 5% fat‐free milk‐TBS for 30 minutes, incubated with primary antibodies in TBS overnight, washed with TBST (TBS with 0.1% Tween20), incubated with HRP‐conjugated secondary antibody for 2 h at RT, washed with TBST, and incubated with the ECL Western Blotting Substrate (Thermo Scientific) and exposed to the HyBlot CL^® autoradiography film (Denville Scientific, Inc., Holliston, MA, USA). Specific immunostaining was quantified using the Multi Gauge software V3.0 from Fuji Film.

For immuno‐dot blot, the brain homogenates in the same protein concentration (5 mg/mL) in the 1×Laemmli sample buffer were diluted 40 times with dilution buffer (0.2% BSA in TBS containing 50 mM NaF, 1 mM Na₃VO₄, and 2 µg/mL each of aprotinin, leupeptin and pepstatin) after denaturing, which was identified as 1. Then the samples were diluted serially using dilution buffer and applied onto a nitrocellulose membrane (Schleicher and Schuell, Keene, NH, USA) at 5 µL/grid (7 × 7 mm). The blot was placed in a 37°C oven for 1 h to allow the protein to bind to the membrane and was processed as described above for Western blot.

Reverse transcription‐PCR (RT‐PCR) and real‐time quantitative PCR (qPCR)

Total cellular RNA was extracted from cultured cells using the RNeasy mini kit (Qiagen) according to the manufacturer's instruction. One microgram of total RNA was used for first‐strand cDNA synthesis with oligo‐(dT)₁₈ using the Omniscript reverse transcription kit (Invitrogen). PCR was performed using PrimeSTAR^TM HS DNA Polymerase (Takara Bio Inc.) with forward primer 5′‐GGTGTCCACTCCCAGTTCAA‐3′ and reverse primer 5′‐CCCTGGTTTATGATGGATGTTGCCTAATGAG‐3′ to measure the alternative splicing of tau exon 10. The conditions of PCR were: 98°C for 5 minutes, 98°C for 10 s, 68°C for 40 s for 30 cycles and then 68°C 10 minutes for extension. The PCR products were resolved on 1.5% agarose gels, visualized by ethidium bromide staining and quantitated using the Multi Gauge software V3.0 from Fuji Film.

The qPCR assay was performed in a final volume of 25 µL containing EvaGreen qPCR master mixture 12.5 µL (2×, Agilent Technologies, Santa Clara, CA, USA), 1 µL DNA template, 0.5 µL (10 µM) of forward and reverse primers and 0.375 µL of dye (1:500). The qPCR was performed using these primers: GFP (forward, 5′‐TGAACCGCATCGAGCTGAAGGG‐3′; reverse, 5′‐ACCTTGATGCCGTTCTTCTGCTTG‐3′) and human GAPDH (forward, 5′‐CATGAGAAGTATGACAACAGCCT‐3′; reverse, 5′‐AGTCCTTCCACGATACCAAAGT‐3′). Amplification was conducted in an MX3000P real‐time PCR system (Stratagene) and the PCR conditions were 95°C for 10 minutes, 95°C for 30 s, 55°C for 1 minute, 72°C for 1 minute for 40 cycles and then 95°C for 1 minute, 55°C for 30 s, and 95°C for 30 s. The fluorescence signals were collected between 72°C and 95°C for the melting curve analysis. The cyclic threshold (Ct) value was analyzed by the 2^−△△CT method.

Statistical analyses

The data were presented as the mean ± S.D where appropriate. Data points were compared with the unpaired two‐tailed Student's t‐test (for data with normal distribution) or Mann‐Whitney test (for data with nonnormal distribution). For the analysis of the correlation between TDP‐43 and CK1ε or total tau or 3R‐tau/total‐tau, Pearson (for data with normal distribution) or Spearman (for data with nonnormal distribution) correlation coefficient was calculated.

Results

Morphological varieties of TDP‐43 inclusion in AD brain

The TDP‐43 pathology was reported in up to 67% cases of AD 34. To study the morphology of TDP‐43 pathology in AD brain, we immunostained frontal cortical sections from AD brain with anti‐TDP‐43 and anti‐pS409/410‐TDP‐43. We observed a robust TDP‐43 pathology with these two antibodies (Figure 1A,B), in addition to diffuse nuclear staining by TDP‐43 (Figure 1A, insert), but not pSer409/410‐TDP‐43 (Figure 1B). TDP‐43 inclusions exhibited a variety of morphologies (Figure 1), which included amorphous (Figure 1A‐i,B‐i), restricted (Figure 1C‐ii), skein‐like (Figure 1A‐iii), dystrophic neurites (Figure 1A‐iv, 1B‐iv), compact round inclusions (Figure 1A‐v,vi) and extra‐condensed cobblestone‐like inclusions (Figure 1A‐vii) or granular somatodendritic inclusions (Figure 1B‐viii). We also observed a few clusters of compact round inclusions (Figure 1B‐ix) with anti‐pSer409/410‐TDP‐43. Interestingly, the cobblestone‐like highly condensed inclusions did not seem to be associated with any neuronal or other cellular profiles (Figure 1A‐vii,B‐vii), a phenomenon verified by the immunofluorescence staining described below. We did not detect in the brain sections of the AD cases but we examined apparent neuronal intranuclear TDP‐43 inclusions, such as the cat eye‐like aggregate‐bearing nuclear profiles seen in ALS/FTLD‐U brain 10. We found granular cytoplasmic inclusions (Figure 1B‐viii) in AD brain sections immunostained for pSer409/410‐TDP‐43, but not in TDP‐43‐immunostained sections from the same tissues.

Figure 1.

Morphological varieties of TDP‐43 pathology in AD brain. Free‐floating sections of frontal cortex of AD cases were immunostained with anti‐TDP‐43 (A) or anti‐pS409/410‐TDP‐43 (B). The high magnified insert in (A) shows a presentative nuclear localization of TDP‐43. The morphology of TDP‐43 pathology in AD brain can be divided into four categories: I) somatodendritic inclusion, which could be amorphous (i), restricted (ii), skein‐like (iii) or granular (viii); II) dystrophic neurites (iv); III) compact round inclusion in the absence (v) or presence (vi) of apparent cellular profiles in proximity; and IV) highly condensed cobblestone‐like spots (vii). Clusters of compact round inclusions (ix) were also observed. Scale bars: 20 μm for low magnification, 10 μm for high magnification.

CK1ε interacts with and phosphorylates TDP‐43

To investigate whether proline‐directed protein kinases (PDPKs) interact with and phosphorylate TDP‐43, we performed a GST pulldown assay. We expressed recombinant GST‐TDP‐43 fusion protein and bound it onto GSH‐beads. After incubation of the GST‐TDP‐43 coupled GSH‐beads with rat brain extract overnight at 4°C, we detected the pulldown proteins by Western blot. The results revealed that glycogen synthase kinase‐3β (GSK‐3β), dual‐specificity tyrosine/Y‐phosphorylation regulated kinase 1A (Dyrk1A) and cyclin‐dependent kinase 5 (CDK5) were not pulled down by TDP‐43, CK1ε was pulled down by GST‐TDP‐43, but not by GST itself. (Figure 2A). These results suggest that TDP‐43 may interact with CK1ε, not with PDPKs.

Figure 2.

CK1ε interacts with and phosphorylates TDP‐43. A. GST‐TDP‐43 or GST coupled with glutathione–sepharose was incubated with rat brain extract. After washing, bound proteins were subjected to Western blots with indicated antibodies. B,C. Recombinant GST‐TDP‐43 was phosphorylated by CK1ε for various times as indicated. The phosphorylation products of TDP‐43 were analyzed by Western blots developed with site‐specific and phosphorylation‐dependent TDP‐43 antibodies (B). Levels of phosphorylated TDP‐43 at individual sites were plotted against the phosphorylation time (C). D. TDP‐43 or together with CK1ε was expressed in HEK‐293FT cells. The phosphorylation products of TDP‐43 were analyzed by Western blots developed with site‐specific and phosphorylation‐dependent TDP‐43 antibodies. E,F. TDP‐43 was overexpressed in HEK‐293FT cells for 40 h, then the cells were treated with 20 and 50 μM PF4800567, a CK1ε‐specific inhibitor, for 8 h. TDP‐43 was immunoprecipitated with anti‐HA and analyzed by Western blots with indicated antibodies (E). Average levels of phosphorylated TDP‐43 at each site were plotted against the concentration of PF4800567 (F).

It is reported that aggregated TDP‐43 in FTLD‐U/ALS is hyperphosphorylated at Ser379, Ser403, Ser404, Ser409 and Ser410 25, 28, 43. CK1 and CK2 are presumably the responsible kinases 35. To study the phosphorylation of TDP‐43 by CK1ε, we incubated recombinant GST‐TDP‐43 with CK1ε immunopurified by monoclonal anti‐HA from HEK‐293FT cells in the presence of ATP for various times as indicated and analyzed the phosphorylation products by Western blots developed with site‐specific and phosphorylation‐dependent anti‐TDP‐43. We found that TDP‐43 was phosphorylated by CK1ε at Ser379, Ser403/404 and Ser409/410 in a time‐dependent manner (Figure 2B,C).

To determine the phosphorylation of TDP‐43 by CK1ε in cultured cells, we overexpressed TDP‐43 and CK1ε in HEK‐293FT cells and analyzed the phosphorylation by Western blots. We found that co‐expression of CK1ε clearly increased TDP‐43 at Ser379, Ser403/404 and Ser409/410 (Figure 2D), suggesting that CK1ε may phosphorylate TDP‐43 at these sites.

PF4800567 is a novel and specific inhibitor of CK1ε 58. To confirm the role of CK1ε in TDP‐43 phosphorylation, we overexpressed TDP‐43 tagged with HA in HEK‐293FT cells for 40 h, and then treated the cells with 20 and 50 μM PF4800567 for 8 h. TDP‐43 was immunoprecipitated by monoclonal anti‐HA and then analyzed by Western blots. We found that the levels of TDP‐43 phosphorylation were decreased at Ser403/404 and Ser409/410 and slightly decreased at Ser379 in the cells treated with PF4800567 in a dose‐dependent manner (Figure 2E,F), supporting that CK1ε phosphorylates TDP‐43 at Ser403/404 and Ser409/410 in cultured cells.

Elevated level of CK1ε associated with an increase in TDP‐43 phosphorylation in AD brain

Expression of CK1ε is increased in AD brain 12. To study the impact of the increased CK1ε level on TDP‐43 phosphorylation in AD brains, we analyzed the levels of CK1ε and TDP‐43 phosphorylation in the homogenates of frontal cortices from controls and AD cases by Western blots (Figure 3A). We found that consistent with previous findings, the level of CK1ε was increased significantly in AD brains (Figure 3A,B). The phosphorylation levels of TDP‐43 at Ser403/404 (Figure 3A,D) and at Ser409/410 (Figure 3A,E), but not at Ser379 (Figure 3A,C), were increased in AD brains. Furthermore, the level of CK1ε was positively correlated with the phosphorylation level of TDP‐43 at Ser403/404 (Figure 3G) and at Ser409/410 (Figure 3H), but not at Ser379 (Figure 3F). These results suggest that the increased CK1ε level could contribute to the upregulation of phosphorylation of TDP‐43 at Ser403/404 and Ser409/410 in AD brain.

Figure 3.

CK1ε level is elevated in AD brain and is associated with an increase in TDP‐43 phosphorylation at Ser403/404 and Ser409/410. A. Brain frontal cortex homogenates from nine controls and ten AD cases were analyzed by Western blots. B–E. The levels of CK1ε (B), TDP‐43 phosphorylated at Ser379 (C), Ser403/404 (D) and Ser409/410 (E) are represented as scattered dots with mean ± SD. F–H. The level of CK1ε was plotted against the level of TDP‐43 phosphorylated at Ser379 (F), Ser403/404 (G) or Ser409/410 (H). The correlation between CK1ε and TDP‐43 was analyzed by Pearson (G and H) and Spearman (F) correlation. *P < 0.05; **P < 0.01; ****P < 0.0001.

CK1ε suppresses TDP‐43's function on tau mRNA processing

We recently reported that TDP‐43 suppresses tau expression by promoting the instability of its mRNA via its 3′‐untranslated region (3′‐UTR) 24. To study whether CK1ε affects the function of TDP‐43 on tau expression, we used the GFP‐tailed with 3′‐UTR of tau to study the GFP expression as described previously 24. We co‐transfected pEGFP/tau 3′‐UTR with pCI/TDP‐43 or together with pCI/CK1ε into HEK‐293FT cells or treated the cells with 50 μM PF4800567 for 8 h and then measured GFP expression by real‐time PCR and Western blots. We found that the overexpression of TDP‐43 suppressed GFP expression at both mRNA and protein levels (Figure 4A,B). This suppression by TDP‐43 was reduced clearly by co‐expression of CK1ε and enhanced slightly by PF4800567 (Figure 4A,B), suggesting that CK1ε may suppress the TDP‐43's function in promoting tau mRNA instability via its 3′‐UTR.

Figure 4.

CK1ε suppresses the function of TDP‐43 in tau mRNA processing and promotes TDP‐43 aggregation. A,B. GFP‐tailed with tau 3′‐UTR was co‐expressed with TDP‐43 and/or CK1ε in HEK‐293FT cells or treated the cells with 50 μM PF4800567 for 8 h. The levels of GFP mRNA and protein were analyzed by qRT‐PCR and Western blots developed with indicated antibodies, respectively. C,D. Overexpression of CK1ε together with TDP‐43 or treated with 50 μM PF4800567 for 8 h in HEK‐293FT cells transfected with mini‐tau gene, pCI/SI9‐LI10. The splicing products were analyzed by RT‐PCR and expression of TDP‐43 and CK1ε were analyzed by Western blots. The ratio of tau exon 10 inclusion/exclusion were calculated. E. TDP‐43 was co‐expressed with CK1ε in HeLa cells and immunostained with anti‐TDP‐43 (Green) and anti‐CK1ε (Red). Scale bar, 20 µm. F. Cells with cytoplasmic TDP‐43 aggregates were quantitated. The data are presented as mean ± SD. *P < 0.05; **p < 0.01.

TDP‐43 promotes tau exon 10 inclusion 23. To determine whether CK1ε regulates TDP‐43's function in tau exon 10 inclusion, we co‐transfected tau mini‐gene SI9‐LI10, consisting tau exons 9,10 and 11 and partial intron 9 and full of intron 10 as described previously, with TDP‐43 and CK1ε into HEK‐293FT cells or treated the cells with 50 μM PF4800567 for 8 h, and analyzed the splicing products of tau exon 10 by RT‐PCR. We found that enhanced tau exon 10 inclusion by TDP‐43 was ameliorated by co‐expression of CK1ε and increased by the PF4800567 treatment (Figure 4C,D), suggesting that CK1ε suppresses the TDP‐43's activity in the enhancement of tau exon 10 inclusion.

To learn the effect of CK1ε on TDP‐43 subcellular localization and aggregation, we co‐overexpressed TDP‐43 and CK1ε in HeLa cells which have easily visible cytoplasm and nucleus, immunostained the cells with anti‐TDP‐43 and anti‐CK1ε. We found that TDP‐43 was expressed mainly in the nucleus and CK1ε was in both the nucleus and the cytoplasm (Figure 4E). A fraction of CK1ε in the nucleus was colocalized with TDP‐43 (Figure 4E). We observed stress granule‐like TDP‐43 aggregates in the cells with overexpression of CK1ε (Figure 4E). The cytoplasmic TDP‐43 aggregates were seen in ~10% cells with TDP‐43 overexpression alone, but in ~50% of cells with co‐expression of both TDP‐43 and CK1ε (Figure 4F), suggesting that CK1ε promotes TDP‐43 aggregation.

Levels of CK1ε and phosphorylated TDP‐43 are correlated positively with total tau and 3R‐tau levels

We previously reported that the level of 3R‐tau, resulting from the exclusion of tau exon 10, is increased in AD brain 53. We thus investigated the association between the levels of CK1ε and phosphorylation of TDP‐43 with 3R‐tau level. We determined the levels of 3R‐tau and total tau in control and AD brains using immuno‐dot blots. We found that, as expected, the ratio of 3R‐tau/total tau was increased significantly in AD brains (Figure 5A,B). We performed linear regression analyses and found that the CK1ε level was positively correlated with total tau and 3R‐tau levels in AD and control brains (Figure 5C,D). In addition, the phosphorylation levels of TDP‐43 at Ser403/404 or Ser409/410, but not at Ser379, was also positively correlated with total tau and 3R‐tau levels (Figure 5C,D). These results suggest that the phosphorylation of TDP‐43 by CK1ε may affect tau expression and contribute to the increase in total tau and 3R‐tau levels in AD brain.

Figure 5.

Phosphorylation of TDP‐43 is associated with the increase in the levels of total tau and 3R‐tau/total tau in AD brain. A,B. Levels of 3R‐tau (RD3) and total‐tau (R134d) in frontal cortices from nine AD and nine age‐matched control cases were determined by immuno‐dot blots (A) and quantified by densitometry (B). C,D. The levels of total tau (C) and the ratio of 3R‐tau/total tau (D) in frontal brain homogenates determined by immuno‐dot blots were plotted against the levels of CK1ε, pS379‐TDP‐43, pS403/404‐TDP‐43, and pS409/410‐TDP‐43, respectively. The Pearson (CK1ε, pS403/404‐TDP‐43 and pS409/410‐TDP‐43 to total tau or 3R‐tau/total tau) and Spearman (pS379‐TDP‐43 to total tau or 3R‐tau/total tau) correlation coefficient r were calculated. The data are presented as mean ± SD. ***P < 0.001.

Elevation of CK1ε expression associated with TDP‐43 and tau pathologies

The level of CK1ε is increased in AD brain and the upregulation of CK1ε is involved in tau pathogenesis in AD 12, 16. To study the impact of the level of CK1ε on TDP‐43 and tau pathologies, we performed immunohistochemical staining for phosphorylated tau (AT8), TDP‐43 and pSer409/410‐TDP‐43 in adjacent frozen sections of the frontal cortices from two AD and one control cases, in which the cerebrocortical tissue showed profound differences in expression level of CK1ε (Figure 6C). We found that the AD case with a high level of CK1ε showed both extra‐dense AT8‐positive inclusions and dense TDP‐43 inclusions; in contrast, the AD case with a low level of CK1ε expression showed sparse AT8‐positive inclusions and correspondingly very limited TDP‐43 pathology, as evidenced by immunostaining with antibodies against pan‐TDP‐43 and against pSer409/410‐TDP‐43 (Figure 6A,C). To further examine the association between TDP‐43 pathology and tau pathology, we colabeled TDP‐43 and AT8 with double immunofluorescence staining and the results showed that TDP‐43 and AT8 were colocalized in some but not all somatodendritic compartments of neurons with inclusions (Figure 6B). In addition, we found TDP‐43 pathology in the absence of AT8 immunoreactivity in somatodendritic compartments of the same neurons and dystrophic neurites and as highly condensed cobblestone‐like structures (Figure 6B).

Figure 6.

High level of CK1ε is associated with profound tau and TDP‐43 pathologies in AD. A. Immunohistochemical staining of phosphorylated tau (pSer202/pThr205, AT8), TDP‐43 and phosphorylated TDP‐43 (pSer409/410). For tau staining, magnified views of the boxed areas in low magnification are shown below the low‐magnification images. High magnification as inserts shows the representative morphology of staining. AD case 1 showed much higher density than AD case 2 of AT8‐positive tangles (hollow arrows) and dystrophic neurites (hollow arrowheads) and TDP‐43‐positive and pSer409/410‐TDP‐43‐positive inclusions (filled dark arrows). Filled dark arrowheads indicate seemingly normal nuclear TDP‐43 staining. B. Double immunofluorescence staining showed the association of tau pathology with TDP‐43 pathology. Sections were counterstained with TOPRO 3 iodide for nuclei. Within numerous AT8‐positive neurofibrillary tangles (arrows a) and dystrophic neurites (arrows b), TDP‐43 inclusions were seen to colocalize with AT8 staining in neurofibrillary tangles (arrows c) or to separately locate in the somatodendritic compartment (arrows d) and dystrophic neurites (arrows e) or appear as cobblestone‐like condensed spots of varying sizes (arrows f). Defuse and weak staining of nuclei by TDP‐43 (arrows g) were also seen. C. Western blots showed the expression levels of CK1ε, TDP‐43, pSer409/410‐TDP‐43, total tau (R134d) and pSer202/pThr205‐tau (AT8). AD case 1 exhibited much higher levels of the proteins examined, except TDP‐43, than AD case 2 and the control case. Scale bars: (A) 500 μm for low magnification, 10 μm for inserts and 50 μm for the rest; (B) 20 μm.

Discussion

TDP‐43 proteinopathy has been observed in >57% of AD cases 34, 35. TDP‐43 is considered as the third protein linked to AD 31. We recently reported that TDP‐43 regulates the tau mRNA processing, including promoting tau mRNA instability via acting on its 3′‐UTR and tau exon 10 inclusion 23, 24. In the present study, we found that CK1ε phosphorylated TDP‐43 at Ser403/404 and Ser409/410, enhanced its cytoplasmic aggregation and suppressed its functions on promoting tau mRNA instability and tau exon 10 inclusion. The level of CK1ε was increased and associated with the levels of TDP‐43 phosphorylation at Ser403/404 or Ser409/410, total tau and 3R‐tau in AD and control brains. This pilot immunohistochemical study showed robust TDP‐43 and tau proteinopathies in AD brain with high level of CK1ε in the pilot. These findings suggest that increased CK1ε in AD brain may phosphorylate TDP‐43 and promote its cytoplasmic aggregation, leading to less TDP‐43 available to regulate the tau mRNA processing, resulting in an increase of total tau and 3R‐tau and tau pathology. Thus, CK1ε may link TDP‐43 to tau pathogenesis in AD via dysregulation of TDP‐43's function on tau mRNA processing. TDP‐43 proteinopathy as a result of increased CK1ε may accelerate/exacerbate tau pathology and neurofibrillary degeneration in AD brain.

TDP‐43 pathology is originally reported to be associated with ALS and FTLD 44. Subsequent studies suggested TDP‐43 pathology presented in 2/3 of the brains of patients with AD 2, 34, and is associated with a higher likelihood of the clinical expression of clinical Alzheimer's‐type dementia, including with greater memory loss, cognitive impairment and dementia, naming and functional decline, and hippocampal atrophy 30, 31, 33, 34, 42. In the present study, we observed robust intraneuronal cytoplasmic inclusion and dystrophic neurites stained with anti‐TDP‐43 in AD brains. The morphology of TDP‐43 proteinopathy is similar to that seen in FTLD/ALS brains, amorphous, restricted, skein‐like, or granular somatodendritic inclusion, dystrophic neurites, compact round inclusion in the absence or presence of apparent cellular profiles in proximity and highly condensed cobblestone‐like spots. Especially, these condensed spots did not seem to be associated with any neuronal or other cellular profiles, which were not reported previously in FTLD/ALS brains. Unlike in FTLD/ALS brains, we did not observe nuclear inclusion, such as cat eye‐like aggregation of TDP‐43 in AD brain sections.

TDP‐43 in inclusions is hyperphosphorylated at multiple sites, including Ser379, Ser403/404 and Ser409/410 25, 28, 43. Phosphorylation of Ser409/410 of TDP‐43 is a consistent feature in all sporadic and familiar forms of TDP‐43 proteinopathies 43, which has been validated in AD 5. In the present study, we found robust pSer409/410‐TDP‐43 positive inclusions in AD brains. The morphologies of pSer409/410‐TDP‐43 inclusion were variform as reported in the brains of FTLD/ALS. In addition, we observed small and multiple granules in the cytoplasm immunostained with anti‐pSer409/410‐TDP‐43 in AD brain sections, which was not seen in TDP‐43‐immunostained brain sections from the same patient. Intensive TDP‐43 proteinopathy was presented in AD brain with higher level of CK1ε, but very rare in AD brain with lower level of CK1ε and no TDP‐43 pathology in control brain.

The level of CK1ε was increased in AD brains 12, 16, which may play important roles in neurodegenerative disease 48. It was reported that CK1ε is able to phosphorylate TDP‐43 in vitro at multiple sites, including Ser379, Ser403/404 and Ser409/410 25, 35. Phosphorylation of TDP‐43 by CK1ε promotes oligomerization and enhances toxicity 13. Here we found that TDP‐43 was phosphorylated by CK1ε from HEK‐293FT cells at Ser379, Ser403/404 and Ser409/410 in vitro. However, selective inhibition of CK1ε by PF4800567 significantly reduced the phosphorylation at Ser403/404 and Ser409/410, but not at Ser379 in cultured cells, suggesting that Ser379 may not be sufficiently phosphorylated by CK1ε ex vivo. Overexpression of CK1ε promoted cytoplasmic TDP‐43 aggregation. However, here we could not confirm that these aggregates are not stress granules. In AD brains, the phosphorylation level of TDP‐43 at Ser403/404 or Ser409/410, but not at Ser379, was significantly increased and was positively correlated with CK1ε level. Of note, phosphorylation of Ser409/410 is a consistent feature in all sporadic and familial forms of TDP‐43 proteinopathies 43. In C. elegans models with ALS‐related mutants overexpression, phosphorylation of TDP‐43 at Ser409/410 drives ALS‐related TDP‐43 mutants' toxicity 39. Thus, hyperphosphorylation of TDP‐43, especially at Ser409/410, by CK1ε is critical in TDP‐43 pathogenesis in AD brain.

Overexpression of TDP‐43 increases abnormal tau aggregation but decreases Aβ plaque in mice 14. We recently reported that TDP‐43 takes part in the regulation of tau mRNA processing. Overexpression of TDP‐43 suppressed tau expression by acting on the 3′‐UTR and promoting tau mRNA instability 24. Highest level of TDP‐43 is accompanied by the lowest tau expression in the cerebellum of rat and human 24, 27, where no tau pathology has been observed in AD brains 37. Thus, pathological changes in TDP‐43 may affect the tau mRNA processing. In the present study, we determined the role of CK1ε in TDP‐43's function in tau mRNA processing. We found that suppressed GFP expression by TDP‐43 was attenuated by co‐expression of CK1ε, suggesting that CK1ε suppressed TDP‐43's function in promoting tau mRNA instability. The levels of CK1ε and TDP‐43 phosphorylation at either Ser403/404 or Ser409/410, but not at Ser379, were positively correlated with tau levels in AD and control brains. In the case of tau exon 10 splicing, we also found that enhanced tau exon 10 inclusion was suppressed by CK1ε co‐expression. The levels of CK1ε and Ser403/404 and Ser409/410 phosphorylation, but not Ser379, were positively correlated with 3R‐tau level resulting from tau exon 10 exclusion in AD and control brains. More severe tau pathology was observed in the AD brain with higher CK1ε level. These results suggest that the increased CK1ε level may promote TDP‐43 phosphorylation and cytoplasmic aggregation, leading to an increase of 3R‐tau and total tau expression, which contribute or accelerate tau pathology of AD.

In the present study, we showed the pathological relationship between TDP‐43 and tau in AD brains. However, TDP‐43 proteinopathy was first reported in the brains of patients with FTLD/ALS, in which tau pathology was negligible 44. People who develop ALS are with an average age of 55 at the time of diagnosis. Thus, the onset of ALS is much earlier than that of sporadic AD, in which disease starts after the age of 65. It appears that the changes in TDP‐43 are insufficient to induce tau pathology under these conditions. However, several studies have shown that AD patients with TDP‐43 pathology showed greater memory loss and cognitive impairment 30, 33, leading to TDP‐43 that is considered as the third protein of AD 31. Here, we found that TDP‐43 phosphorylation by CK1ε promoted its cytoplasmic aggregation, made it unable to promote tau mRNA instability and tau exon 10 inclusion. Thus, pathological changes of TDP‐43 may accelerate/exacerbate tau pathology and neurodegeneration in AD.

We previously reported that Dyrk1A enhances 3R‐tau expression via phosphorylating splicing factors, ASF 54, SC35 49 and SRp55 60. In AD brain, truncation and activation of Dyrk1A by activated calpain I is associated with the 3R‐tau/4R‐tau ratio. TDP‐43 does not have proline‐directed Ser/Thr sites, thus, Dyrk1A could not phosphorylate it. In addition, it was reported that GSK‐3β phosphorylates SC35 and regulates tau exon 10 splicing. In the present study, we also did not observe the interaction between TDP‐43 and GSK‐3β.

Conclusion

In the present study, we found that CK1ε phosphorylated TDP‐43 at Ser403/404 and Ser409/410, promoted its cytoplasmic aggregation, resulting in the enhancement of tau mRNA stability and tau exon 10 exclusion, consequence increasing 3R‐tau and total tau expression. An increased level of CK1ε in AD brain may contribute to tau pathology by enhancing tau and 3R‐tau expression resulted from stabilized tau mRNA and tau exon 10 exclusion as results of nuclear depletion of TDP‐43 by its phosphorylation and cytoplasmic aggregation. Thus, CK1ε may link TDP‐43 to tau pathology and TDP‐43 proteinopathy may accelerate tau pathology by promoting the expression of total tau and changing the ratio of 3R‐tau/4R‐tau.

Author's Contributions

J. G. performed experiments, analyzed results and wrote the first draft of the manuscript. W.H., X.T., S.Q. and D.C. performed experiments. C. G. and K. I. discussed results and edited manuscript. F. L. designed and performed experiments, analyzed and interpreted results, and wrote the manuscript.

Consent for Publication

All authors gave consent for publication.

Competing Interests

The authors declare that they have no conflicts of interest with the contents of this article.

Data Availability Statement

The data that support the findings of this study are available from the corresponding author upon reasonable request.