A novel DYRK1A - Wiley Online Library

3 downloads 0 Views 1MB Size Report
Feb 28, 2013 - (methyl 9-(2,4-dichlorophenylamino) thiazolo[5,4-f]quinazoline- ... f]quinazoline-2-carbimidate); FBS, fetal bovine serum; HEK cell, human.
JOURNAL OF NEUROCHEMISTRY

| 2015 | 133 | 440–451

doi: 10.1111/jnc.13018

,1

,1

*From Diaxonhit, Paris, France †Normandie Univ, COBRA, UMR 6014 & FR 3038; Univ Rouen; INSA Rouen; CNRS, IRCOF, 1 rue Tesniere, Mont St Aignan Cedex, Paris, France

Abstract The dual-specificity tyrosine phosphorylation-regulated kinase 1A (DYRK1A) gene is located within the Down Syndrome (DS) critical region on chromosome 21 and is implicated in the generation of Tau and amyloid pathologies that are associated with the early onset Alzheimer’s Disease (AD) observed in DS. DYRK1A is also found associated with neurofibrillary tangles in sporadic AD and phosphorylates key AD players (Tau, amyloid precursor, protein, etc). Thus, DYRK1A may be an important therapeutic target to modify the course of Tau and amyloid beta (Ab) pathologies. Here, we describe EHT 5372 (methyl 9-(2,4-dichlorophenylamino) thiazolo[5,4-f]quinazoline2-carbimidate), a novel, highly potent (IC50 = 0.22 nM) DYRK1A inhibitor with a high degree of selectivity over 339

kinases. Models in which inhibition of DYRK1A by siRNA reduced and DYRK1A over-expression induced Tau phosphorylation or Ab production were used. EHT 5372 inhibits DYRK1Ainduced Tau phosphorylation at multiple AD-relevant sites in biochemical and cellular assays. EHT 5372 also normalizes both Ab-induced Tau phosphorylation and DYRK1A-stimulated Ab production. DYRK1A is thus as a key element of Abmediated Tau hyperphosphorylation, which links Tau and amyloid pathologies. EHT 5372 and other compounds in its class warrant in vivo investigation as a novel, high-potential therapy for AD and other Tau opathies. Keywords: Alzheimer’s disease, amyloid precursor, protein, DYRK1A, EHT 5372, kinase, Tau. J. Neurochem. (2015) 133, 440–451.

Alzheimer disease (AD) is the major cause of dementia in the elderly affecting more than 30 million people worldwide and is pathologically characterized by brain senile plaques with beta-amyloid peptide (Ab) deposition and neurofibrillary tangles (NFT) harboring abnormally hyperphosphorylated Tau. Hyperphosphorylated Tau accumulates in the somatodendritic compartment of neurons and aggregates in insoluble filaments that form the NFT lesions. Abnormal Tau hyperphosphorylation changes its conformation and reduces its affinity to microtubules (Lindwall and Cole 1999), resulting in microtubule instability, NFT formation, neuronal cell dysfunction, and ultimately cell death, as neurons are no longer able to maintain axonal transport (LaPointe et al.

Received August 22, 2014; revised manuscript received November 26, 2014; accepted December 07, 2014. Address correspondence and reprint requests to Laurent Desire, Diaxonhit, 63-65 Boulevard Massena, 75013 Paris, France. E-mail: [email protected] 1 These authors contributed equally to this work. Abbreviations used: AD, Alzheimer disease; APP, amyloid precursor, protein; Ab, b-amyloid; DS, Down Syndrome; DYRK1A, Dualspecificity tYrosine-Regulated Kinase-1A; EGCG, epigallocatechingallate; EHT 5372, (methyl 9-(2,4-dichlorophenylamino) thiazolo[5,4f]quinazoline-2-carbimidate); FBS, fetal bovine serum; HEK cell, human embryonic kidney cell; hERG, (human Ether-a-go-go-Related Gene); HRP, horseradish peroxidase; MTT, 3-(4,5-dimethylthiazol-2-yl)-2,5diphenyltetrazolium bromide; NFT, neurofibrillary tangles; SD, standard deviation; siRNA, small interference RNA.

440

© 2014 International Society for Neurochemistry, J. Neurochem. (2015) 133, 440--451

New DYRK1A inhibitor as anti-Tau and amyloid therapy

2009). In AD, sequential hyperphosphorylation of Tau on multiple amino acids correlates with the severity of NFT pathology (Augustinack et al. 2002) and the appearance of these lesions in a predictable manner correlates with neuronal loss and the degree of cognitive deficit (Grundke-Iqbal et al. 1986; Braak and Braak 1991). Recent studies have provided evidence of Tau secretion at the synapse, a role for extracellular soluble Tau monomers in the initiation of Tau pathology spreading, and of trans-synaptic propagation of Tau pathology via a prion-like mechanism (Liu et al. 2012; Michel et al. 2014). In addition, Tau also appears to play a direct role in Ab-mediated toxicity at the synapse, which is supported by observations that reducing levels of Tau in mouse transgenic models protects neurons from Ab insults (Roberson et al. 2007; Ittner et al. 2010). Pharmacological inhibition of Tau phosphorylation at certain key sites, that regulate its functional activity or promote its aggregation into NFTs, may provide a promising therapeutic alternative to BACE (Beta-site APP-cleaving enzyme), c-secretase or Ab inhibitors, which to date have been unsuccessful in human trials. Past studies have identified more than 20 Serine/Threonine kinases able to phosphorylate Tau in vitro, these include GSK3 (Tau kinase I, GSK3b, and GSK3a), CDK5 (Tau kinase II), MARK (MAP/microtubule affinity-regulating kinase), PKA (Protein kinase A), and ERK1/2 (Sergeant et al. 2008; Martin et al. 2013). However, whether all of these kinases actually play a role in human tauopathies is currently not known. The human Dual-specificity tyrosine-regulated kinase-1A (DYRK1A) kinase is associated with NFTs in Down Syndrome (DS) and sporadic Alzheimer’s disease (AD) (Wegiel et al. 2008, 2011). The DYRK1A gene is located on the Down Syndrome Critical Region of human chromosome 21, resulting in the over-expression of DYRK1A, along with the amyloid precursor protein (APP). This pathogenic gene dosage imbalance results in consistent alterations in cognition and memory because of defective neurogenesis and formation of neuronal structures. Furthermore, over-expression studies in cell cultures and transgenic models of DS have implicated the DYRK1A kinase in the generation of both Tau and amyloid pathologies associated with early onset AD that is uniformly observed in DS (Kimura et al. 2007; Ryoo et al. 2007; Park et al. 2009). Increasing evidences also implicate DYRK1A in neurodegenerative processes. For instance, DYRK1A protein levels are elevated in AD brains (Ferrer et al. 2005; Kimura et al. 2007) and some studies, but not others, have identified DYRK1A as a genetic factor associated with AD (Kimura et al. 2007; Vazquez-Higuera et al. 2009). DYRK1A kinase may favor the appearance of some pathological hallmarks of neurodegeneration by phosphorylating key disease components. In AD, these include Tau, as DYRK1A phosphorylates Tau at T212, a residue that is hyperphosphorylated in AD brains and primes Tau for

441

phosphorylation by GSK3 at other sites (Woods et al. 2001). A high-throughput siRNA screening of 572 kinases (Azorsa et al. 2010), has confirmed DYRK1A role in Tau hyperphosphorylation on various amino acids (S202, T212, T231, S396, S404, AT8 (pS202 + T205), 12E8 (pS262 + pS356)). In addition, other key DYRK1A substrates involved in AD include alternative splicing factors that control Tau splicing, but also APP and presenilin (PS1), which are responsible for Ab production (reviewed in Frost et al. 2011; Smith et al. 2012). DYRK1A kinase is thus a novel, high-potential therapeutic target for pharmacological interventions seeking to modify the course of AD. In our screening efforts to discover new scaffolds for the inhibition of DYRK1A, we identified a series of new DYRK1A inhibitors. Here, we report on the biochemical and biological characterization of (methyl 9-(2,4-dichlorophenylamino) thiazolo[5,4-f]quinazoline-2-carbimidate) (EHT 5372), a novel, highly potent (IC50 = 0.22 nM) DYRK1A inhibitor with a high degree of selectivity over 339 kinases. Models in which inhibition of DYRK1A by siRNA reduced and DYRK1A over-expression induced Tau phosphorylation or Ab production were used to further evaluate EHT 5372. EHT 5372 inhibits DYRK1A-induced Tau phosphorylation at multiple AD-relevant sites in biochemical assay, cell lines, and primary cortical neurons without affecting cell viability. In addition, it normalizes Ab-induced Tau phosphorylation in neuronal cells and also normalizes DYRK1A-induced Ab production in APP over-expressing cells. These results indicate that EHT 5372 and other compound in its class warrant further investigation as candidates Tau and amyloiddirected therapeutics to alter the onset or progression of Alzheimer’s disease and other tauopathies.

Methods Materials and compounds EHT 5372 (methyl 9-(2,4-dichlorophenylamino) thiazolo[5,4-f] quinazoline-2-carbimidate) was synthetized as described in Leblond et al. 2013). Harmine, epigallocatechin-gallate (EGCG), and TG003 were provided by COBRA laboratory based in Rouen, France. L41 was synthesized at Diaxonhit following the experimental procedures described by Debdab et al. (2011). The DYRK1A expression plasmid encodes a highly active DYRK1A-GFP fusion protein and was a kind gift of Dr. Walter Becker (Institut f€ ur Pharmakologie und Toxikologie, Aachen, Germany). Antibodies raised against the following epitopes were used: DYRK1A (Invitrogen, Carlsbad, CA, USA; 1 : 1000 dilution), phospho-Tau Thr212 (pT212, Invitrogen; 1 : 2500 dilution), phospho-Tau Ser396 (pS396, Invitrogen; 1 : 5000 dilution), phospho-Tau Thr231 (pT231, Invitrogen; 1 : 500 dilution), phospho-Tau Ser262 (pS262, Abcam, Cambridge, GB; 1 : 1000 dilution), phospho-Tau Ser356 (pS356, Abcam; 1 : 5000 dilution), phospho-Tau Ser202/ Thr205 (AT8, Thermo Scientific Bioblock, Illkirch, France; 1 : 500 dilution). Total Tau antibodies were Tau-5 (Tebu; 1 : 200 dilution) and Tau-46 (Sigma, St. Louis MO, USA; 1 : 1000 dilution).

© 2014 International Society for Neurochemistry, J. Neurochem. (2015) 133, 440--451

442

S. Coutadeur et al.

Secondary goat anti-mouse and anti-rabbit antibodies conjugated to horseradish peroxidase (HRP) were from Bio-Rad Laboratories, Marnes la Coquette, France and used at 1 : 5000 dilution. Dharmacon’s SMARTpool ON-TARGETplus siRNA for DYRK1A and DYRK1B, the control scrambled siRNA and the DharmaFECT formulation reagents were purchased from Thermo Scientific. Cell culture and treatments HEK293 cells (ATCC #CRC-1573) were maintained in Modified Eagle’s medium + Earle’s salt supplemented with 10% fetal bovine serum (FBS), 2 mM L-glutamine (Sigma, France), 1X NonEssential Amino Acids and antibiotics. SH-SY5Y cells (ECACC#94030304) were maintained in Modified Eagle’s medium/F12K (1 : 1, v/v) supplemented with 10% FBS, 2 mM Lglutamine, 1X Non-Essential Amino Acids, 1X sodium pyruvate and antibiotics. HEK293 cells were transiently transfected with a human 4R2N Tau construct under CMV (cytomegalovirus) promoter in the presence or absence of the DYRK1A expression vector using FugeneHD X-tremeGENE HP DNA transfection reagent (Roche Diagnostics, Meylan, France), according to the manufacturer’s protocol. The SH-SY5Y-Tau clone was generated by transfecting the Tau441 expression vector using FugeneHD, followed by 3 weeks of G418 (200 lg/mL, Sigma) selection. Single clones were isolated and cultured in 96-well plates, passaged into 24-well plates and screened for Tau over-expression by western blot and immunostaining using the Tau-5 antibody. Primary cortical neurons cultures were generated from 16-day-old Wistar rat embryos (Janvier, Le Genest-Saint-Isle, France) as described in Marcade et al. (2008). Dissociated cells were plated in poly-L-lysine, laminine-coated plates in medium (Neurobasal, 1X B27, 2 mM L-glutamine, and 1X penicillin/streptomycin, Sigma) containing 5% horse serum and 5% FBS at 37°C in a humidified 5% CO2 incubator. Four days after plating, cells were treated with 2.5 lM of Cytosine Arabinofuranoside. After 4 days, one half of the medium was changed with medium containing 2% horse serum. The culture was allowed to mature in this medium for 7–10 days. Amyloid beta peptide preparations were realized by solubilizing Ab42 in 1,1,1,3,3,3-hexafluoro-2-Propanol (HFIP, Sigma) to 1 mM. After evaporation of the solvent with Speedvac, the dried peptide film was resuspended in dimethylsulfoxide (DMSO), diluted in phosphate-buffered saline, and maturated at 4°C to form Ab oligomers. The oligomer solution was centrifuged, and the resulting supernatant was incubated on cells as described. Two independent preparations from two different Ab vials were used. For all phosphorylation-related experiments, cells were changed with new culture medium and treated with 0.1–10 lM DYRK1A inhibitor EHT 5372 or 0.1% DMSO as vehicle for 24 h. The kinase inhibitor was tested in at least three independent experiments performed in duplicate. Cell viability was determined using a standard 3-(4,5-dimethylthiazol-2-yl)-2,5-diphenyltetrazolium bromide assay (Promega, Charbonnieres, France). After treatment, cells were washed twice with cold phosphate-buffered saline and harvested for Western blotting or ELISA in the presence of protease (Roche Diagnostics) and phosphatases inhibitor cocktails (Sigma). Sarkosyl fractions were prepared as in Michel et al. (2014). After a trypsin wash, cells were lysed in 50 lL of Triton lysis buffer (1% Triton X-100, 50 mM Tris, 150 mM NaCl, protease inhibitors). The samples were centrifuged at 14 000 g for 30 min, and the

supernatant was collected (Triton fraction). The pellet was washed with Triton lysis buffer and then solubilized in sarkosyl lysis buffer (1% sarkosyl, 50 mM Tris, 150 mM NaCl, protease inhibitors). The fractions were immunoblotted using Tau-46. Standard Western blotting procedures were used. All shown Western blots images are representative of three to five experiments that gave similar results. Quantification was performed using ImageJ software (http:// imagej.nih.gov/ij/index.html). The human Tau [pS396] ELISA kit was from Invitrogen and was used according to the manufacturer’s protocol. IC50 determinations were performed using Prizm software (GraphPad Software, Inc. La Jolla, CA). Mann–Whitney U-test was used to determine the significance between the data means. Significance values are as follows: *p < 0.05; **p < 0.01; ***p < 0.001 versus corresponding control.

In vitro kinase assay DYRK1A kinase activity was determined by incubating 20 ng of recombinant human DYRK1A protein (Invitrogen) with 40 ng of 4R2N recombinant human Tau 441 (SignalChem, Interchim, Montlucßon, France) in kinase buffer (25 mM Tris-HCl (pH 7.5), 5 mM beta-glycerophosphate, 2 mM dithiothreitol, 0.1 mM Na3VO4, 10 mM MgCl2) and 1 mM ATP in a final volume of 10 lL for 30 min at 30°C. Test compounds were added 10 min prior to the addition of kinase buffer, ATP, and Tau. Final DMSO concentration was 0.2%. The reaction was inactivated upon addition of western blot sample buffer and immediately heated for 10 min at 95°C. Phosphorylated Tau was analyzed by western blot using phospho-Tau antibodies and a secondary goat anti-rabbit HRP antibody in 5% bovine serum albumin. Membranes were stripped and reprobed with mouse monoclonal anti-human total Tau-5 and a secondary goat anti-mouse HRP in 5% milk (EHT 5372, Harmine, EGCG). For TG003 and L41 total Tau blots shown in Fig. 2b, equal fractions of the reactions mixtures were run on a separate gel and analyzed with Tau-5 antibody. Small molecule profiling on 339 kinases panel In vitro profiling of the 339 kinases panel was performed at Reaction Biology Corporation (www.reactionbiology.com, Malvern, PA, USA) using the ‘HotSpot’ assay platform. The DYRKtide sequence [RRRFRPASPLRGPPK] was used for DYRK1-4 kinases. Compounds were delivered into the reaction, followed 20 min later by addition of a mixture of ATP (Sigma) and 33P ATP (Perkin Elmer, Waltham, MA, USA). ATP final concentration was 10 lM. Reactions were carried out at 20°C for 120 min. After subtraction of background derived from control reactions containing inactive enzyme, kinase activity data were expressed as the percent of remaining kinase activity in test samples compared to vehicle (DMSO) reactions.

Results Anti-kinase activity of EHT 5372 EHT 5372 is the current lead compound of a novel chemical class of thiazoloquinazoline compounds developed as a novel class of DYRK inhibitors (Leblond et al. 2013). Activity against the four members of the DYRK family, DYRK1A, DYRK1B, DYRK2, DYRK3, and DYRK4 was determined

© 2014 International Society for Neurochemistry, J. Neurochem. (2015) 133, 440--451

New DYRK1A inhibitor as anti-Tau and amyloid therapy

EHT 5372 compound demonstrated good selectivity over 339 tested kinases, with inhibitory activities displayed toward the CMGC group only. Partial inhibition, below 50%, were noticed on discoidin domain receptor tyrosine kinase 1 (DDR1) and Haspin. We then compared the potency of EHT 5372 on DYRK1A and DYRK1B with that of the few tool compounds currently described (harmine, EGCG, L41 and TG003) on the ‘HotSpot’ platform and found that EHT 5372 was much more potent on both kinases, in particular being more than 30-fold more potent than the most active one, leucettine L41 (Table 1).

DDR1 IC50 < 1 nM 7 nM < IC50 < 200 nM Partial activity

DYRK1A DYRK1B GSK3β GSK3α

443

DYRK2 DYRK3 CLK4 CLK1 CLK2 Haspin

EHT 5372 inhibits the direct phosphorylation of Tau by DYRK1A Using an in vitro phosphorylation assay with recombinant human DYRK1A and Tau proteins (Fig. 2a), we confirmed that DYRK1A directly phosphorylated Tau protein at Serine 396 (pS396). Phosphorylation occurred only in the presence of Tau protein, DYRK1A protein, and ATP. When added to the reaction, EHT 5372 potently and dose-dependently inhibited Tau phosphorylation at pS396 (Fig. 2a). We then compared EHT 5372 effect with that of harmine, epigallocatechin-gallate EGCG, TG003, and L41. The observed pS396 Tau phosphorylation was inhibited by harmine at the concentration of 1 lM and above, consistent with its reported IC50 (0.7 lM) by Frost et al. (2011). L41 and TG003 also inhibited pS396 Tau phosphorylation at 1 lM and above, whereas EGCG showed only activity at 10 lM, and surprisingly followed a bell shaped curve in this assay. Because all these reference compounds were less active than EHT 5372 on the two above kinase assays, they were not further investigated.

Fig. 1 The kinome activity map for EHT 5372 (methyl 9-(2,4-dichlorophenylamino) thiazolo[5,4-f]quinazoline-2-carbimidate) with identified hits highlighted.

at Reaction Biology Corporation using their ‘HotSpot’ assay platform. EHT 5372 inhibited DYRK1A activity with an IC50 of 0.22 nM and DYRK1B with an IC50 of 0.28 nM. IC50 for DYRK2 and DYRK3 were higher (10.8 and 93.2 nM, respectively) and there was no inhibition on DYRK4. The compound was subsequently profiled at 1000x its IC50 (250 nM) on Reaction Biology Corporation’s full (339) kinase panel and IC50 were determined on hits with more than 50% inhibition. Figure 1 shows the kinome activity map for EHT 5372 with the hits highlighted. As shown in Table 1, EHT 5372 displayed minimal impact on the CDC2-like kinase (CLK) family, with more than 100x selectivity over CLK1 (CLK1: IC50 = 22.8 nM; CLK2: IC50 = 88.8 nM; CLK3:IC50 > 10 lM; CLK4: no inhibition) and on the glycogen synthase kinase 3 (GSK3) family (GSK3a: IC50 = 7.44 nM; GSK3b: IC50 = 221 nM). Thus

Inhibition of DYRK1A by EHT 5372 reduces Tau phosphorylation in HEK293 cells To determine the effect of EHT 5372 on cellular DYRK1A-induced Tau phosphorylation, we initially co-

Table 1 IC50 of EHT 5372 on the hits of a selectivity profile performed on a total of 339 kinases IC50 (nM)

DYRK1A

DYRK1B

DYRK2

DYRK3

DYRK4

GSK3a

CLK1

CLK2

CLK3

CLK4

GSK3b

EHT 5372 Selectivity ratio

0.22 1

0.28 1.28

10.8 49.1

93.2 423.6

n.i. n.d.

7.44 33.8

22.8 103.6

88.8 403.6

> 10000 n.d.

59 268.1

221 1004.5

IC50 (nM)

DYRK1A

DYRK1B

Harmine TG003 L41 EGCG

21.8 24.01 7.60 11130

27.8 34.39 37 1244

Selectivity ratios are evaluated by dividing potencies from the off-target and DYRK1A. Values for other DYRK1A inhibitors (Harmine, EGCG, L41, and TG003) were determined for DYRK1A and DYRK1B only. n.i.: no inhibition; n.d.: not determined.

© 2014 International Society for Neurochemistry, J. Neurochem. (2015) 133, 440--451

444

S. Coutadeur et al.

EHT 5372 (µM)

(a)

Compound (µM)

(b)

0 0.01 0.03 0.1 0.3 1

Tau DYRK1A ATP 64 kDa-

64 kDa-

– + +

+ + –

+ – +

+ + +

+ + +

+ + +

+ + +

+ + +

0 0.01 0.1 1

Tau – DYRK1A + ATP +

+ + + pS396

64 kDa(pS396)

Tau-5

64 kDa(Tau-5)

64 kDa(pS396)

+ + –

+ – +

+ + +

+ + +

+ + +

+ + +

3

+ + +

10

+ + +

Harmine

EGCG

64 kDa(Tau-5) 64 kDa(pS396) TG003 64 kDa(Tau-5) 64 kDa(pS396)

L-41

64 kDa(Tau-5) Fig. 2 Dose-dependent inhibition of Tau phosphorylation by EHT 5372 (methyl 9-(2,4-dichlorophenylamino) thiazolo[5,4-f]quinazoline2-carbimidate) and other DYRK1A inhibitors in an in vitro kinase assay. Western-blot analysis of Tau phosphorylation induced by incubation of Tau, DYRK1A and ATP in the presence of various concentrations of the indicated compound. Representative Western blots demonstrating

the inhibitory effect of EHT 5372 are shown (a) Phosphorylated Tau was detected using phospho-specific S396 (pS396) antibody. (b) Effect of previously described DYRK1A inhibitors harmine, epigallocatechin-gallate (EGCG), TG003 and L41on pS396 Tau levels. Global Tau (phosphorylated and non-phosphorylated) was detected by Tau-5 antibody (a and b).

transfected HEK293 cells, that express DYRK1A but are devoid of endogenous Tau, with DYRK1A and Tau expression vectors and monitored Tau phosphorylation at S396 position by western blot and ELISA. Figure 3a shows that Tau transfection alone in HEK293 cells results in the phosphorylation of Tau by endogenous kinases at S396. When DYRK1A was co-transfected to maximize DYRK1A effect on Tau and the dynamic range of the assay in the context of pathologically increased DYRK1A expression, Tau phosphorylation was increased by more than twofold. Cells were then either non-transfected or co-transfected with DYRK1A and Tau expression vectors and treated after 24 h with various concentrations of EHT 5372. After 24 h incubation, DYRK1A, pS396 Tau, and total Tau were detected by ELISA and western blot. Cell viability was determined in parallel using 3-(4,5-dimethylthiazol-2-yl)2,5-diphenyltetrazolium bromide. EHT 5372 dose-dependently reduced pS396-Tau levels with an IC50 of 1.7 lM whereas cell viability remained over 87% in all conditions (Fig 3b). An almost parallel result was obtained in western

blot (Fig. 3c). EHT 5372 dose dependently reduced pS396Tau levels, whereas total Tau remained globally unchanged. In fact, inhibition of DYRK1A by EHT 5372 reduced cellular Tau phosphorylation at multiple phosphoepitopes (Fig. 3c). Inhibition of DYRK1A by EHT 5372 reduces cellular Tau phosphorylation at multiple phosphoepitopes in neuronal cells Rat cortical neurons were first incubated with various doses of EHT 5372 for 24 h and cell lysates were assayed for Tau phosphorylation using the antibodies indicated in Fig. 4a. On most phosphoepitopes, EHT 5372 was active at the concentration of 0.1 lM (Fig. 4a). We subsequently confirmed these results using neuroblastoma SH-SY5Y-Tau441 cells, a stable clone over-expressing hTau441. Cells were treated with EHT 5372 for 24 h and cell lysates were assayed for pS396 levels by ELISA. Inhibition of DYRK1A by EHT 5372 dose-dependently reduced cellular Tau phosphorylation. At the dose of 10 lM, EHT 5372 reduced pS396-Tau

© 2014 International Society for Neurochemistry, J. Neurochem. (2015) 133, 440--451

New DYRK1A inhibitor as anti-Tau and amyloid therapy

445

(a)

***

(c)

***

(b)

**

** ***

***

µM EHT5372 IC50 = 1.73 µM Fig. 3 DYRK1A increases Tau phosphorylation in HEK293 cells and its inhibition by EHT 5372 (methyl 9-(2,4-dichlorophenylamino) thiazolo[5,4-f]quinazoline-2-carbimidate) reduces cellular Tau phosphorylation on multiple amino acids. (a) Transient transfection of DYRK1A increases Tau phosphorylation. HEK293 cells were transiently co-transfected by DYRK1A, Tau, or both expression vectors and Tau phosphorylation was assessed using a phospho-S396specific (pS396) ELISA. Control: untransfected cells. B, C, Inhibition of DYRK1A by EHT 5372 reduces cellular Tau phosphorylation in

HEK293 cells transiently co-transfected by DYRK1A and Tau expression vectors. Cell cultures were treated for 24 h with various doses of EHT 5372 and cell lysates were assayed for Tau phosphorylation using the pS396-specific ELISA (b) or various phospho-specific (T231, AT8 (S202/T205), S396, T212), DYRK1A, total Tau (Tau-46) and actin antibodies (c). Control in (b): untreated, DYRK1A and Tau transfected, cells. Control in (c): untransfected and untreated cells. Mean values  SD from five independent experiments performed in duplicate are shown. **p < 0.01; ***p < 0.001.

levels by 69% (Fig. 4c). In both cellular models, EHT 5372 had no significant effect on total Tau level, or cell viability (Fig. 4a–c). SH-SY5Y cells (and HEK293 cells) express both DYRK1A and DYRK1B in the form of several isoforms, whereas DYRK1B is almost undetectable in mature rat cortical neurons (Fig. 4d). Given that EHT 5372 also inhibits DYRK1B, we investigated whether DYRK1B may impact Tau phosphorylation in SH-SY5Y-Tau441 cells. Cells were transfected for 96 h with scrambled control (25 nM), DYRK1A or DYRK1B siRNA (10–50 nM) and analyzed for DYRK1A, DYRK1B, or pS396 Tau levels. From 10 nM, DYRK1A and DYRK1B siRNA very strongly reduced the cellular level of DYRK1A and DYRK1B, respectively (Fig. 4e). However, only DYRK1A siRNA reduced pS396 Tau levels, with maximal inhibition already achieved at 10 nM. DYRK1B siRNA or the control siRNA did not affect pS396 Tau levels. Total Tau remained unaffected and these western blot results were confirmed using the phospho-S396specific ELISA (Fig. 4f), indicating that DYRK1A, and not DYRK1B, phosphorylates Tau in these cells.

Amyloid beta peptide stimulates Tau phosphorylation, which can be reduced by EHT 5372 Soluble Ab42 oligomers induce neuronal Tau hyperphosphorylation on AT8 epitope (Brouillette et al. 2012). Ab42 also induces DYRK1A expression (Kimura et al. 2007). Thus, SH-SY5Y-Tau441 cells and cortical neurons were treated with Ab42 oligomers preparations in the presence or absence of various concentrations of EHT 5372 for 48 h to assess EHT 5372 effect on Tau phosphorylation induced by Ab42. Cell lysates were assayed using the anti-pS396 (SH-SY5Y cells) or the AT8 antibody (cortical neurons). Ab42 stimulated Tau phosphorylation at both pS396 and AT8 epitopes and this effect was dose-dependently inhibited by EHT 5372 treatment in both cell types, reverting phosphorylated Tau levels back to the control situation (Fig. 5). Neither Ab42 nor EHT 5372 exhibited a significant effect on total Tau or actin levels in this assay. Insoluble Tau is operationally defined as the fraction of Tau that pellets in the presence of the detergent sarkosyl. Because Tau phosphorylation is closely related to its aggregation, we investigated sarkosyl-insoluble fractions of

© 2014 International Society for Neurochemistry, J. Neurochem. (2015) 133, 440--451

446

S. Coutadeur et al.

(a)

(b)

pS396 tau (% of control)

(c)

(d)

120 100

pS396 tau

**

80 60

**

40

***

20 0 Cont

(e)

Cell viability

0.1

1 5 μM EHT 5372

10

(f)

**

**

**

Fig. 4 Inhibition of DYRK1A by EHT 5372 (methyl 9-(2,4-dichlorophenylamino) thiazolo[5,4-f]quinazoline-2-carbimidate) reduces cellular Tau phosphorylation on multiple amino acids in neuronal cells. (a) Effect of EHT 5372 on Tau phosphorylation in rat cortical neurons. Primary neuronal cell cultures were treated with various doses of EHT 5372 for 24 h and cell lysates were assayed for Tau phosphorylation (T231, AT8 (S202/T205), S356, S396, T212), total Tau (Tau-46) or actin. (b) Cortical neurons viability by 3-(4,5-dimethylthiazol-2-yl)-2,5-diphenyltetrazolium bromide (MTT) in the samples is shown. (c) Dose-dependent inhibition of Tau phosphorylation at S396 by EHT 5372 in SH-SY5Y-Tau441 cells. Cells were treated with EHT 5372 for 24 h and SH-SY5Y cell lysates were analyzed for Tau (Tau-46), pS396 by ELISA and cell viability using MTT. Mean values  SD from five independent experiments performed

in duplicate are shown. (d) Endogenous DYRK1A and DYRK1B protein expression in wild-type SH-SY5Y cells, cortical neurons and HEK293 cells. (e) Effect of DYRK1A or DYRK1B down-regulation by siRNA on Tau phosphorylation. SH-SY5Y-Tau441 cells were transfected for 96 h with scrambled (Sc) DYRK1A or DYRK1B siRNA and analyzed for DYRK1A, DYRK1B or pS396 Tau levels. A representative western blot demonstrating the effect of DYRK1A or DYRK1B siRNA on DYRK1A or DYRK1B level is shown. (f) Determination of pS396 Tau levels by pS396-specific ELISA after DYRK1A or DYRK1B siRNA-mediated down-regulation in SH-SY5Y-Tau441 cells. Cont: siRNA vehicle-treated control cells. Mean values  SD from three independent experiments performed in duplicate are shown. **p < 0.01; ***p < 0.001.

Tau from the primary cortical neurons cultures. Ab42 increased the sarkosyl-insoluble Tau fraction and co-treatment with EHT 5732 prevented such increase. These data

indicated that Ab42 treatment of neuronal cells increased Tau insolubility and that DYRK1A is central to Ab-mediated Tau phosphorylation and aggregation.

© 2014 International Society for Neurochemistry, J. Neurochem. (2015) 133, 440--451

New DYRK1A inhibitor as anti-Tau and amyloid therapy

+ 25 μg/mL Aβ42

(a) Cont

0

0.1

1

5

pS396

-70 kDa

Tau-46

-70 kDa

AT8

-55 kDa

Actin

-42 kDa

Tau-46 (sarkosyl fraction)

Fold induction (% of control)

*

3.5

*

3

Cortical neurons

-55 kDa

4

2.5

SH-SY5YTau441 cells

-55 kDa

Tau-46 (Triton fraction)

(b)

μM EHT 5372

* *

*

pS396 (SH-S5Y5-Tau441 cells) AT8 Cortical Tau-46 (sark.) neurons Actin

2 1.5 1 0.5 0 Cont

0

0.1

1

5

μM EHT 5372 + 25 μg/mL Aβ42

Fig. 5 DYRK1A inhibition by EHT 5372 (methyl 9-(2,4-dichlorophenylamino) thiazolo[5,4-f]quinazoline-2-carbimidate) prevents Abinduced Tau phosphorylation in neuronal cells. Ab42 preparation was added to either primary cortical neurons or SH-SY5Y-Tau441 cells (overnight incubation) in the presence or absence of various concentrations of EHT 5372 and cell lysates were tested for Tau phosphorylation using pS396 antibody (for SH-SY5Y) or AT8 antibody (for cortical neurons). Sarkosyl fractions (sark.) from cortical neurons were immunoblotted using Tau-46. (a) Representative western blot from three independent experiments. (b) Quantification (ImageJ software). Mean values  SD from three independent experiments performed in duplicate are shown. *p < 0.05.

Inhibition of DYRK1A by EHT 5372 reduces Ab production More Ab is produced in DYRK1A over-expressing cell lines and transgenic mice (Kimura et al. 2007; Ryoo et al. 2007). Here, we used HEK293 cells over-expressing APP to test whether DYRK1A inhibition by EHT 5372 effectively modulates Ab levels. Ab40 was used as the readout, being more than 10 times more produced than Ab42 by these cells (data not shown). First, HEK293-APP cells were treated with different concentrations of EHT 5372 for 24 h. This treatment resulted in a dose-dependent reduction in Ab40 levels in the culture supernatant. The calculated IC50 was 1.06 lM (Fig. 6a). To confirm the role of DYRK1A as a modulator of Ab production, we then either down-regulated its expression using siRNA, or over-expressed DYRK1A and quantified

447

Ab40 that accumulated in the supernatant for 24 h. Figure 6b shows DYRK1A levels visualized by western blot after incubation with vehicle only, the scrambled control siRNA, and various doses of DYRK1A siRNA for 96 h. Extinction of DYRK1A expression was maximal from 10 nM DYRK1A siRNA. Therefore, this concentration was used to determine the impact of DYRK1A down-regulation on Ab40 production during the last 24 h of the experiment. We also used the DYRK1B-specific siRNA, which was also very active in down-regulating DYRK1B protein levels in HEK293-APP cells at 10 nM (data not shown). Figure 6c shows that DYRK1A siRNA reduced Ab production by 50%, whereas DYRK1B siRNA was ineffective. Vehicle and scrambled siRNA (10 nM) controls also did not impact Ab levels. Therefore, DYRK1A, but not DYRK1B downregulation, altered Ab production in these cells. In parallel, we transiently transfected the DYRK1A expression vector in HEK293-APP cells in the presence or absence of various doses of EHT 5372. The western blot in Fig. 6d shows the different DYRK1A proteins separated by their respective size: the lower bands correspond to endogenous DYRK1A isoforms, whereas the upper band corresponds to the over-expressed DYRK1A-GFP fusion protein. DYRK1A over-expression modestly but reproducibly induced Ab40 levels by 1.5-fold when Ab40 accumulated for 24 h was quantified in the cell supernatant (Fig. 6e). At the concentration of 1 lM, EHT 5372 fully blocked the induction of Ab by DYRK1A, normalizing Ab levels that resulted from elevated DYRK1A. At higher concentration (10 lM), EHT 5372 further decreased Ab40 levels, without affecting cell viability.

Discussion The intracellular accumulation of hyperphosphorylated Tau and neurofibrillary degeneration are hallmark pathological events in tauopathies. The Tau lesions correlate well with clinical symptoms in AD patients (Braak and Braak 1991). Ab is also a major aggregating protein in AD, but the Ab deposit pattern follows a less defined path and the relationship between amyloid accumulation and neurofibrillary degeneration is a major unanswered question. For example, the toxic effects of Ab are upstream to changes in Tau but conversely, Tau appears to be necessary for the progression of neurodegeneration in cell culture and animal transgenic models (Rapoport et al. 2002; Roberson et al. 2007). Tau function is regulated by phosphorylation by coordinate actions of Tau kinases and phosphatases (reviewed in Martin et al. 2013; Wang et al. 2013), and alternative splicing, which controls the Tau 3R/ 4R variants ratio (Shi et al. 2008; Qian et al. 2011). There are up to 85 identified phosphorylation sites on the longest Tau isoform and numerous sites are associated with Tau dysfunction and neurodegeneration (Schraen-Maschke et al. 2008; Wang

© 2014 International Society for Neurochemistry, J. Neurochem. (2015) 133, 440--451

448

S. Coutadeur et al.

(a)

(d) DYRK1A-GFP Cont 0

100 50 0 —50 —6

—4

—2

—0

Cont Veh Scr 5 DYRK1A

-130 kDa -90 kDa

(e)

siRNA DYRK1A 10 25 50 nM

-90 kDa

(c) Aβ levels (% of control)

10 µM EHT 5372

2

EHT 5372 log [µM]

(b)

1

DYRK1A -GFP DYRK1A endo

IC50 = 1.06

150 125 100

Aβ level (% of control)

% Aβ (pg/mL)

150

*

200

*

150 100

50 0

Cont

DYRK1A DYRK1A + DYRK1A + EHT 5372 EHT 5372 1 μM 10 μM

75

***

50 25 0

Cont

Veh Scr. siRNA siRNA siRNA DYRK1B DYRK1A

Fig. 6 Inhibition of DYRK1A reduces and over-expression of DYRK1A induces Ab production in HEK293-amyloid precursor, protein (APP) cells. (a) Inhibition of DYRK1A by EHT 5372 (methyl 9-(2,4dichlorophenylamino) thiazolo[5,4-f]quinazoline-2-carbimidate) dosedependently reduces Ab40 levels. (b–e), Effect of DYRK1A expression down-regulation by siRNA (b and c) or over-expression (d and e) on Ab40 levels in HEK293-APP cells. (b) DYRK1A levels visualized by western blot after incubation with siRNA vehicle, control siRNA

(25 nM), and various doses of DYRK1A siRNA (5–5 nM). (c) Effect of the siRNA treatments (10 nM each) on Ab40 levels, as determined by ELISA. (d) DYRK1A was over-expressed in HEK293-APP cells and both endogenous (endo) and over-expressed DYRK1A (DYRK1A-GFP) were detected by western blot. (e) Effect of DYRK1A over-expression on Ab levels in HEK293-APP cells and its inhibition by EHT 5372. Mean values  SD from three independent experiments performed in duplicate are shown. *p < 0.05; ***p < 0.001.

et al. 2013). It is thought that pharmacological inhibition of Tau phosphorylation at certain key sites that regulate the functional activity of Tau or that promote the aggregation of Tau into NFTs may provide a promising approach for the treatment of AD and other tauopathies. Here, we implicate DYRK1A as a crucial player in Ab-mediated Tau phosphorylation and Tau-associated pathology. DYRK1A belongs to the CMGC group of the eukaryotic kinome (CMGC: cyclin-dependent kinases (CDKs), mitogen-activated protein kinases (MAPKs), glycogen synthase kinases, and CLKs) (Becker et al. 1998). Increased activity of DYRK1A has been reported in various brain regions in subjects that suffer from DS and neurodegenerative diseases including familial and sporadic AD. (Ferrer et al. 2005; Kimura et al. 2007). The over-expression of DYRK1A probably contributes to neurofibrillary degeneration and dementia through promoting abnormal hyperphosphorylation of Tau at several sites seen in Tau isolated from AD brain (Wegiel et al. 2008, 2011; Wang et al. 2013). A limited number of tool compounds are available to probe the activity of DYRK1A (Smith et al. 2012; Medda et al. 2013; Becker et al. 2014). The most widely used are two plant compounds, epigallocatechin-gallate (EGCG) and

harmine. EGCG was shown to rescue brain defects of DYRK1A-over-expressing mice (Guedj et al. 2009) despite modest efficacy in DYRK1A assays, complicated pharmacokinetic properties and poor bioavailability, and has been clinically tested with some positive outcome on cognitive deficits in DS individuals (De la Torre et al. 2014). Harmine, a b-carboline alkaloid compound, showed potent inhibition ockler et al. 2009) and Tau of DYRK1A (IC50 = 33 nM, G€ phosphorylation. EGCG also interacts with MAP kinases, phosphatases, DNA, RNA, DNA methyltransferase, topoisomerases (Patra et al. 2008), whereas harmine presents activities on MAO-A, 5-HT2A, imidazoline receptors, CDK1, 2, and 5 (Li et al. 2011). This spectrum of nonkinase off-target activities is responsible for various side effects, such as harmine’s associated CNS effects, and compromises their clinical usefulness. Recently, the leucettine L41 compound has also been described on DYRK1A (IC50 DYRK1A = 40 nM) and Cdc2-like kinases (Clk) (IC50 CLK1 = 15 nM; Debdab et al. 2011). L41 inhibits the phosphorylation of serine/arginine-rich proteins (SRp), a family of proteins regulating pre-RNA splicing. L41 also provided neuroprotection against glutamate-induced cell death in HT22 hippocampal cells and against APP-induced

© 2014 International Society for Neurochemistry, J. Neurochem. (2015) 133, 440--451

New DYRK1A inhibitor as anti-Tau and amyloid therapy

cell death in mouse brain slices (Tahtouh et al. 2012a). Finally L41 may prevent in vivo cognitive impairments because of intracerebroventricular injection of amyloid-b 25-35 (Tahtouh et al. 2012b). We describe here EHT 5372, a novel, potent (IC50 = 0.22 nM) DYRK1A inhibitor. EHT 5372 was synthesized (Leblond et al. 2013) in an effort to develop novel DYRK1 inhibitors with good potency, selectivity, and ‘drug-like’ profile, and was found more potent than all the available reference compounds in two kinase assays. Harmine, L41, TG003, and EGCG were run in parallel to EHT 5372 to provide head to head comparison, as only partial information from different in vitro kinase assays are available for these molecules. Differences in the potency of EHT 5372 and the reference compounds were noted between our two kinase assays, which can be explained by the very different experimental conditions (such as ATP, kinase and substrate concentrations, or test format) as well as type of substrate (DYRKtide: an optimized peptide substrate for DYRK1A which is unrelated to Tau-Fig. 1 vs. full-length 4R2N hTau441, Fig. 2). We also noticed that EHT 5372 exhibited lower potency in cell based assays, which is frequently observed for compounds and may be attributed to cell penetration, solubility or stability parameters. EHT 5372 exhibited a high degree of selectivity over 339 tested kinases, demonstrating inhibitory activities toward the CMGC group only, with at least 100-fold selectivity over CLKs (which are ubiquitous key regulators of pre-mRNA splicing). Almost equipotent activities toward DYRK1A and DYRK1B were observed, owing to high identity of the two amino acid sequences which are 84% identical in the N-terminus and the catalytic domain (Leder et al. 1999). DYRK1B (also called Mirk) is an attractive target for oncology that is expressed at very low levels in normal cells (Friedman 2007). Accordingly, DYRK1B protein is almost undetectable in rat cortical neurons and our siRNA knockdown experiments in SH-SY5Y and HEK293 cells have demonstrated that DYRK1B is unlikely to play some role in the effects of EHT 5372 on Tau phosphorylation and Ab production. EHT 5372 is more than 30-fold more potent on DYRK1A than on GSK3a and more than 1000-fold more potent on GSK3b. Despite this, it cannot be excluded that some effects reported here on Tau are mediated at least in part via some level of GSK3a inhibition in cells, especially at the highest concentrations of EHT 5372. However, inhibition of Tau phosphorylation occurs through DYRK1A inhibition upon EHT 5372 treatment as 1/EHT 5372 prevents DYRK1A-induced Tau phosphorylation at multiple sites and 2/ EHT 5372 inhibits Tau phosphorylation at comparable magnitude than that achieved with DYRK1A siRNA knock-down. Previous studies have shown that DYRK1A inhibitors including harmine reduce Tau phosphorylation independently of cell toxicity, whereas significant reduction in total

449

Tau levels also occurs prior to toxicity (Frost et al. 2011). Then, cell viability decreases rapidly, likely owing to too important reduction in Tau levels. Here, EHT 5372 concentrations as high as 10 lM do not impact total Tau levels, and do not reduce cell viability by more than 20%. The rapid decrease in viability observed with harmine between 8 lM (20% toxicity) and 16 lM (20% viability) by Frost et al. (2011) in H4 neuroglioma cells may suggest that EHT 5372 could become rapidly toxic when used above 10 lM, if EHT 5372 follows this pattern. Alternatively, it is possible that EHT 5372 does not follow the same pattern than harmine, or that this mechanism is shifted to higher compound concentrations. Higher concentrations were not tested in cell lines because we focused on analyzing the effects of EHT 5372 at non-toxic concentrations. Thus, one limitation of our study is that cellular activity versus toxicity profiles of EHT 5372 and harmine cannot be formally compared and ranked. Interestingly, EHT 5372 shows high cell viability in cortical neurons, at a concentration as high as 25 lM. This concentration was not tested in cell lines. While it remains to be demonstrated that dividing and non-dividing cells are differentially impacted by EHT 5372, an hypothesis not tested here, it likely that viability of both dividing and nondividing cells will be important characteristics of EHT 53720 s in vivo profile. Increased Tau phosphorylation was observed following the application of Ab42 oligomers to SH-SY5Y cells and primary cortical neurons cultures. In these cultures, EHT 5372 prevented Ab-induced Tau hyperphosphorylation. Tau phosphorylation is closely related to its aggregation (Ballatore et al. 2007). Treatment of cortical neurons with Ab also increased the sarkosyl-insoluble Tau fraction and cotreatment with EHT 5732 prevented such increase. When the levels of this insoluble pool of Tau are increased, neurons are predisposed to neurofibrillary tangle (Ryoo et al., 2007). The type of amyloid species that were generated was not characterized and we acknowledge the fact that the experimental systems used here minimize the complexity found in vivo in terms of types of Ab species, oligomers, and Tau phosphorylation pattern. Nevertheless, our data fully support DYRK1A as a key element of Ab-mediated Tau hyperphosphorylation and aggregation, and indicate that EHT 5732 is able to normalize this pathogenic event. Elevated Ab levels are detected in the hippocampus of DYRK1A transgenic mice and in the brain of DS patients, suggesting that DYRK1A over-expression promotes Ab production. Indeed, other DYRK1A substrates include APP and presenilin (PS1) (reviewed in Wegiel et al. 2011; Smith et al. 2012; Becker et al. 2014). Ryoo et al. (2008) revealed that DYRK1A phosphorylates APP at T668 in vitro and in mammalian cells. This affects APP cellular trafficking and is thought to facilitate the cleavage of APP by secretases, thereby enhancing the production of Ab (Vingtdeux et al. 2005). Here, we show that DYRK1A over-expression increased, whereas DYRK1A siRNA knock-down reduced

© 2014 International Society for Neurochemistry, J. Neurochem. (2015) 133, 440--451

450

S. Coutadeur et al.

Ab levels in HEK293-APP cells. Accordingly, EHT 5372 inhibited Ab production both in the context of endogenous DYRK1A expression or following DYRK1A over-expression. Thus, inhibiting DYRK1A activity with EHT 5372 may have the potential to reduce amyloid plaques formation. Interestingly, the observed stimulation of DYRK1A expression by Ab in cells and in APP/PS1 mice cells highlights a positive feedback between DYRK1A and Ab that has been suggested to accelerate disease progression (Kimura et al. 2007). Therefore, our data confirm that DYRK1A induction could be a causal link between Tau phosphorylation and Ab production. DYRK1A is thus a novel, high-potential target for AD and other tauopathies for the following reasons: 1/ DYRK1A phosphorylates key AD proteins, 2/ DYRK1A directly links amyloid and Tau pathologies in AD and DS, 3/ DYRK1A acts as a priming kinase for GSK3, a major player in AD, 4/ DYRK1A is involved in regulating pre-mRNA splicing of Tau that is associated with AD, 5/ DYRK1A over-expression is associated with the early AD phenotype that occurs in Down syndrome patients, and 6/ DYRK1A appears to be very dose-dependent, since increases as small as 1.5-fold produces major DS-like pathologies in the mouse. Concerning this last point, in DS, three copies of both the APP and DYRK1A genes result in increased levels of DYRK1A and APP by 50 and 55%, respectively (Wegiel et al. 2008). This small change in DYRK1A gene dosage results in consistent alterations of cognition and memory. This is compatible with the objective to normalize, but not fully block, DYRK1A activity by treating at low drug concentrations, thereby lowering the potential for side effects. In the adult organism, preliminary experiments have indicated that EHT 5372 is well tolerated upon chronic administration in rodents and does not impact the human Ether-a-go-go-Related Gene channel. Therefore, the selective DYRK1A inhibitor EHT 5372 and this class of compounds warrant further investigation as a novel, high-potential therapy for AD and other tauopathies.

Acknowledgments and conflict of interest disclosure This work was supported by MESR (Ministere Enseignement Superieur & Recherche, France) PhD grant to AF. TB thanks the LABEX SynOrg (ANR-11-LABX-0029) for financial support. HB, CD, A-SC, AG, MP, and LD are current employees of Diaxonhit

References Augustinack J. C., Schneider A., Mandelkow E. M. and Hyman B. T. (2002) Specific tau phosphorylation sites correlate with severity of neuronal cytopathology in Alzheimer’s disease. Acta Neuropathol. 103, 26–35. Azorsa D. O., Robeson R. H., Frost D. et al. (2010) High-content siRNA screening of the kinome identifies kinases involved in Alzheimer’s disease-related tau hyperphosphorylation. BMC Genom. 12, 11–25.

Ballatore C., Lee V. M. and Trojanowski J. Q. (2007) Tau-mediated neurodegeneration in Alzheimer’s disease and related disorders. Nat. Rev. Neurosci. 8, 663–672. Becker W., Weber Y., Wetzel K., Eirmbter K., Tejedor F. J. and Joost H. G. (1998) Sequence characteristics, subcellular localization, and substrate specificity of DYRK-related kinases, a novel family of dual specificity protein kinases. J. Biol. Chem. 273, 25893– 25902. Becker W., Soppa U. and Tejedor F. J. (2014) DYRK1A: a potential drug target for multiple down syndrome neuropathologies. CNS Neurol. Disord. Drug Targets. 13, 26–33. Braak H. and Braak E. (1991) Neuropathological stageing of Alzheimerrelated changes. Acta Neuropathol. 82, 239–259. Brouillette J., Caillierez R., Zommer N., Alves-Pires C., Benilova I., Blum D., De Strooper B. and Buee L. (2012) Neurotoxicity and memory deficits induced by soluble low-molecular-weight amyloid-b1-42 oligomers are revealed in vivo by using a novel animal model. J. Neurosci. 32, 7852–7861. De la Torre R., De Sola S., Pons M. et al. (2014) Epigallocatechin-3gallate, a DYRK1A inhibitor, rescues cognitive deficits in Down syndrome mouse models and in humans. Mol. Nutr. Food Res. 58, 278–288. Debdab M., Carreaux F., Renault S. et al. (2011) Leucettines, a class of potent inhibitors of cdc2-like kinases and dual specificity, tyrosine phosphorylation regulated kinases derived from the marine sponge leucettamine B. Modulation of alternative pre-RNA splicing. J. Med. Chem. 54, 4172–4186. Ferrer I., Barrachina M., Puig B., Martınez de Lagran M., Martı E., Avila J. and Dierssen M. (2005) Constitutive Dyrk1A is abnormally expressed in Alzheimer disease, Down syndrome, Pick disease, and related transgenic models. Neurobiol. Dis. 20, 392–400. Friedman E. (2007) Mirk/Dyrk1B in cancer. J. Cell. Biochem. 102, 274– 279. Frost D., Meechoovet B., Wang T., Gately S., Giorgetti M., Shcherbakova I. and Dunckley T. (2011) b-carboline compounds, including harmine, inhibit DYRK1A and tau phosphorylation at multiple Alzheimer’s disease-related sites. PLoS ONE 6, e19264. G€ockler N., Jofre G., Papadopoulos C., Soppa U., Tejedor F. J. and Becker W. (2009) Harmine specifically inhibits protein kinase DYRK1A and interferes with neurite formation. FEBS J. 276, 6324–6337. Grundke-Iqbal I., Iqbal K., Tung Y.C., Quinlan M., Wisniewski H.M. and Binder L.I. (1986) Abnormal phosphorylation of the microtubule-associated protein tau (tau) in Alzheimer cytoskeletal pathology. Proc. Natl Acad. Sci. USA 83, 4913–4917. Guedj F., Sebrie C., Rivals I. et al. (2009) Green tea polyphenols rescue of brain defects induced by overexpression of DYRK1A. PLoS ONE 4, e4606. Ittner L. M., Ke Y. D., Delerue F. et al. (2010) Dendritic function of tau mediates amyloid-beta toxicity in Alzheimer’s disease mouse models. Cell 142, 387–397. Kimura R., Kamino K., Yamamoto M. et al. (2007) The DYRK1A gene, encoded in chromosome 21 Down syndrome critical region, bridges between beta-amyloid production and tau phosphorylation in Alzheimer disease. Hum. Mol. Genet. 16, 15–23. LaPointe N. E., Morfin G., Pigino G., Gaisina I. N., Kozikowski A. P., Binder L. I. and Brady S. T. (2009) The amino terminus of tau inhibits kinesin-dependent axonal transport: implications for filament toxicity. J. Neurosci. Res. 87, 440–451. Leblond B., Casagrande A.-S., Desire L., Foucourt A. and Besson T. Tricyclic Pyrimidines As Inhibitors of DYRK1A/DYRK1B As Potential Treatment for Down’s Syndrome or Alzheimer’s Disease. Patent Application Number: WO 2013/026806A1. Publication date: February 28, 2013. Priority date: August 17, 2012.

© 2014 International Society for Neurochemistry, J. Neurochem. (2015) 133, 440--451

New DYRK1A inhibitor as anti-Tau and amyloid therapy

Leder S., Weber Y., Altafaj X., Estivill X., Joost H. G. and Becker W. (1999) Cloning and characterization of DYRK1B, a novel member of the DYRK family of protein kinases. Biochem. Biophys. Res. Commun. 254, 474–479. Li Y., Sattler R., Yang E. J., Nunes A., Ayukawa Y., Akhtar S., Ji G., Zhang P. W. and Rothstein J. D. (2011) Harmine, a natural betacarboline alkaloid, upregulates astroglial glutamate transporter expression. Neuropharmacol. 60, 1168–1175. Lindwall G. and Cole R. D. (1999) Phosphorylation affects the ability of tau protein to promote microtubule assembly. J. Biol. Chem. 259, 5301–5305. Liu L., Drouet V., Wu J.W., Witter M.P., Small S.A., Clelland C. and Duff K. (2012) Trans-synaptic spread of tau pathology in vivo. PLoS ONE 7, e31302. Marcade M., Bourdin J., Loiseau N., Peillon H., Rayer A., Drouin D., Schweighoffer F. and Desire L. (2008) Etazolate, a neuroprotective drug linking GABA(A) receptor pharmacology to amyloid precursor protein processing. J. Neurochem. 106, 392–404. Martin L., Latypova X., Wilson C. M., Magnaudeix A., Perrin M. L., Yardin C. and Terro F. (2013) Tau protein kinases: involvement in Alzheimer’s disease. Ageing Res. Rev. 12, 289–309. Medda F., Smith B., Gokhale V., Shaw A., Dunckley T. and Hulme C. (2013) Beyond Secretases: kinase inhibitors for the treatment of Alzheimer’s Disease. Annu. Rev. Med. Chem. 48, 57–71. Michel C. H., Kumar S., Pinotsi D., Tunnacliffe A., St George-Hyslop P., Mandelkow E., Mandelkow E. M., Kaminski C. F. and Kaminski Schierle G. S. (2014) Extracellular monomeric tau protein is sufficient to initiate the spread of tau protein pathology. J. Biol. Chem. 289, 956–967. Park J., Song W. J. and Chung K. C.(2009) Function and regulation of Dyrk1A: towards understanding Down syndrome. Cell. Mol. Life Sci. 66, 3235–3240. Patra S. K., Rizzi F., Silva A., Rugina D. O. and Bettuzzi S. (2008) Molecular targets of (-)-epigallocatechin-3-gallate (EGCG): specificity and interaction with membrane lipid rafts. J. Physiol. Pharmacol. 59, 217–235. Qian W., Liang H., Shi J., Jin N., Grundke-Iqbal I., Iqbal K., Gong C. X. and Liu F. (2011) Regulation of the alternative splicing of tau exon 10 by SC35 and Dyrk1A. Nucleic Acids Res. 39, 6161– 6171. Rapoport M., Dawson H. N., Binder L. I., Vitek M. P. and Ferreira A. (2002) Tau is essential to beta-amyloid-induced neurotoxicity. Proc. Natl Acad. Sci. USA 99, 6364–6369. Roberson E. D., Scearce-Levie K., Palop J. J., Yan F., Cheng I. H., Wu T., Gerstein H., Yu G. Q. and Mucke L. (2007) Reducing endogenous tau ameliorates amyloid beta-induced deficits in an Alzheimer’s disease mouse model. Science 316, 750–754. Ryoo S. R., Jeong H. K., Radnaabazar C. et al. (2007) DYRK1Amediated hyperphosphorylation of Tau. A functional link between Down syndrome and Alzheimer disease. J. Biol. Chem. 282, 34850–34857.

451

Ryoo S. R., Cho H. J., Lee H. W. et al. (2008) Dual-specificity tyrosine (Y)-phosphorylation regulated kinase 1A-mediated phosphorylation of amyloid precursor protein: evidence for a functional link between Down syndrome and Alzheimer’s disease. J. Neurochem. 104, 1333–1344. Schraen-Maschke S., Sergeant N., Dhaenens C. M. et al. (2008) Tau as a biomarker of neurodegenerative diseases. Biomark. Med. 2, 363– 384. Sergeant N., Bretteville A., Hamdane M. et al. (2008) Biochemistry of Tau in Alzheimer’s disease and related neurological disorders. Expert Rev. Proteomics 5, 207–224. Shi J., Zhang T., Zhou C. et al. (2008) Increased dosage of Dyrk1A alters alternative splicing factor (ASF)-regulated alternative splicing of tau in Down syndrome. J. Biol. Chem. 283, 28660–28669. Smith B., Medda F., Gokhale V., Dunckley T. and Hulme C. (2012) Recent advances in the design, synthesis, and biological evaluation of selective DYRK1A inhibitors: a new avenue for a disease modifying treatment of Alzheimer’s? ACS Chem. Neurosci. 3, 857–872. Tahtouh T., Elkins J. M., Filippakopoulos P. et al. (2012a) Selectivity, cocrystal structures, and neuroprotective properties of leucettines, a family of protein kinase inhibitors derived from the marine sponge alkaloid leucettamine B. J. Med. Chem. 55, 9312–9330. Tahtouh T., Fedorov O., Lozach O., Carreaux F., Bazureau J. P., Meunier J., Maurice T., Knapp S. and Meijer L.(2012b) Leucettines, a family of pharmacological inhibitors of DYRKs & CLKs kinases derived from the marine sponge Leucettamine B. Planta Med. 78, PD1. Vazquez-Higuera J. L., Sanchez-Juan P., Rodrıguez-Rodrıguez E. et al. (2009) DYRK1A genetic variants are not linked to Alzheimer’s disease in a Spanish case-control cohort. BMC Med. Genet. 10, 129. Vingtdeux V., Hamdane M., Gompel M. et al. (2005) Phosphorylation of amyloid precursor carboxy-terminal fragments enhances their processing by a gamma-secretase-dependent mechanism. Neurobiol. Dis. 20, 625–637. Wang J. Z., Xia Y. Y., Grundke-Iqbal I. and Iqbal K. (2013) Abnormal hyperphosphorylation of tau: sites, regulation, and molecular mechanism of neurofibrillary degeneration. J. Alzheimers Dis. S1, S123–S139. Wegiel J., Dowjat K., Kaczmarski W. et al. (2008) The role of overexpressed DYRK1A protein in the early onset of neurofibrillary degeneration in Down syndrome. Acta Neuropathol. 116, 391–394. Wegiel J., Gong C. X. and Hwang Y. W. (2011) The role of DYRK1A in neurodegenerative diseases. FEBS J. 278, 236–245. Woods Y. L., Cohen P., Becker W., Jakes R., Goedert M., Wang X. and Proud C. G. (2001) The kinase DYRK phosphorylates proteinsynthesis initiation factor eIF2Bepsilon at Ser539 and the microtubule-associated protein tau at Thr212: potential role for DYRK as a glycogen synthase kinase 3-priming kinase. Biochem. J. 355, 609–615.

© 2014 International Society for Neurochemistry, J. Neurochem. (2015) 133, 440--451