Login| Sign Up| Help| Contact|

Patent Searching and Data


Title:
USE OF PHOSPHORYLATED TAU AND P38GAMMA TO TREAT A NEUROLOGICAL CONDITION
Document Type and Number:
WIPO Patent Application WO/2017/147654
Kind Code:
A1
Abstract:
The present invention relates to a method of treating or preventing a neurological condition mediated by a tau-dependent signalling complex in neurons of a subject, comprising treating the subject to: (a) promote phosphorylation of one or more amino acid residues of tau, wherein the phosphorylation of the amino acid residues causes disruption of the tau-dependent signalling complex in neurons of the subject; or (b)introduce a variant of tau that causes disruption of the tau-dependent signalling complex in neurons of the subject. The invention also relates to vectors, compositions and kits for treating or preventing a neurological condition mediated by a tau-dependent signalling complex in neurons of a subject.

Inventors:
ITTNER LARS MATTHIAS (AU)
ITTNER ARNE ANSELM (AU)
Application Number:
PCT/AU2017/050180
Publication Date:
September 08, 2017
Filing Date:
March 01, 2017
Export Citation:
Click for automatic bibliography generation   Help
Assignee:
NEWSOUTH INNOVATIONS PTY LTD (AU)
International Classes:
A61K38/17; A61P9/10; A61P25/08; A61P25/28
Domestic Patent References:
WO2007094867A22007-08-23
WO2007062167A22007-05-31
WO2012009442A22012-01-19
WO2008124066A12008-10-16
WO2013086583A12013-06-20
Foreign References:
US20160038613A12016-02-11
US5264221A1993-11-23
Other References:
BUEE-SCHERRE, V. ET AL.: "Phosphorylation of microtubule-associated protein tau by stress-activated protein kinases in intact cells", FEBS LETTERS, vol. 515, 2002, pages 151 - 154, XP004347759
GOEDERT, M. ET AL.: "Phosphorylation of microtubule-associated protein tau by stress- activated protein kinases", FEBS LETTERS, vol. 409, 1997, pages 57 - 62, XP000906954
SABIO, G. ET AL.: "Stress- and mitogen-induced phosphorylation of the synapse- associated protein SAP90/PSD-95 by activation of SAPK3/p38y and ERKl/ERK2", BIOCHEM. J., vol. 380, 2004, pages 19 - 30, XP003008876
LONG, D. ET AL.: "p38g mitogen-activated protein kinase suppresses chondrocyte production of MMP-13 in response to catabolic stimulation", OSTEOARTHRITIS AND CARTILAGE, vol. 18, 2010, pages 1203 - 1210, XP027255886
RANKIN, C. ET AL.: "Pseudo-phosphorylation of tau at Ser202 and Thr205 affects tau filament formation", MOLECULAR BRAIN RESEARCH, vol. 138, 2005, pages 84 - 93, XP027784716
JEGANATHAN, S. ET AL.: "Proline-directed Pseudo-phosphorylation at AT 8 and PHF1 Epitopes Induces a Compaction of the Paperclip Folding of Tau and Generates a Pathological ( MC -1) Conformation", THE JOURNAL OF BIOLOGICAL CHEMISTRY, vol. 283, no. 46, 14 November 2008 (2008-11-14), pages 32066 - 32076, XP055575942, DOI: 10.1074/jbc.M805300200
POOLER, A. ET AL.: "Dynamic association of tau with neuronal membranes is regulated by phosphorylation", NEUROBIOLOGY OF AGING, vol. 33, 2012, pages 431.e27 - 431.e38, XP028397295
SUN, Q. ET AL.: "Pseudo-hyperphosphorylation causing AD-like changes in tau has significant effects on its polymerization", BIOCHEMISTRY, vol. 48, no. 25, 2009, pages 6002 - 6011, XP055412283
TENREIRO, S. ET AL.: "Protein phosphorylation in neurodegeneration: friend or foe?", FRONTIERS IN MOLECULAR NEUROSCIENCE, vol. 7, 2014, pages 1 - 30, XP055412284
ITTNER, L. ET AL.: "Dendritic function of tau mediates amyloid-beta toxicity in Alzheimer's disease mouse models", CELL, vol. 142, no. 3, 2010, pages 387 - 97, XP028931189
"Genbank", Database accession no. NP_001290181
PEARSON ET AL.: "CLUSTAL", 1994
GENETICS COMPUTER GROUP (GCG: "the ALIGN program (Version 2.0) and GAP, BESTFIT, BLAST, FASTA, and TFASTA", WISCONSIN GENETICS SOFTWARE PACKAGE
A ITTNER ET AL., J EXP MED, vol. 209, 2012, pages 2229 - 46
GRIEGERSAMULSKI: "Adeno-associated virus as a gene therapy vector: vector development, production and clinical applications", ADVANCES IN BIOCHEMICAL ENGINEERING/BIOTECHNOLOGY, vol. 99, 2005, pages 119 - 145, XP009125595
AE HARASTA ET AL., NEUROPSYCHOPHARMACOLOGY, vol. 40, 2015, pages 1969 - 1978
NALDINI ET AL.: "In vivo gene delivery and stable transduction of nondividing cells by a lentiviral vector", SCIENCE, vol. 272, 1996, pages 263 - 267, XP000583652, DOI: 10.1126/science.272.5259.263
LOIS: "Germline transmission and tissue-specific expression of transgenes delivered by lentiviral vectors", SCIENCE, vol. 295, 2002, pages 868 - 872, XP002288582, DOI: 10.1126/science.1067081
VOGEL ET AL.: "A single lentivirus vector mediates doxycycline-regulated expression of transgenes in the brain", HUM GENE THER., vol. 15, no. 2, 2004, pages 157 - 165
KOZARSKYWILSON, CURRENT OPINION IN GENETICS AND DEVELOPMENT, vol. 3, 1993, pages 499 - 503
SOUTHGATE ET AL.: "Gene transfer into neural cells in vitro using adenoviral vectors", CURRENT PROTOCOLS IN NEUROSCIENCE, vol. 4, no. 23, 2008
AKLI ET AL.: "Transfer of a foreign gene into the brain using adenovirus vectors", NATURE GENETICS, vol. 3, no. 3, 1993, pages 224 - 228, XP000672891, DOI: 10.1038/ng0393-224
"Remington's Pharmaceutical Science", 1985, MACK PUBLISHING COMPANY
FB ENGEL ET AL., GENES DEV, vol. 19, 2005, pages 1175 - 87
C STURCHLER-PIERRAT ET AL., PROC NATL ACAD SCI U S A, vol. 94, 1997, pages 13287 - 92
A PROBST ET AL., ACTA NEUROPATHOL, vol. 99, 2000, pages 469 - 81
I DEWACHTER ET AL., J NEUROSCI, vol. 22, 2002, pages 3445 - 53
KL TUCKERM MEYERYA BARDE, NAT NEUROSCI, vol. 4, 2001, pages 29 - 37
AR POGOZELSKI ET AL., PLOS ONE, vol. 4, 2009, pages e7934
G SUMARA ET AL., CELL, vol. 136, 2009, pages 235 - 48
ITTNER ET AL., PROC. NATL. ACAD. SCI. U.S.A., vol. 105, 2008, pages 15997 - 16002
GIBSON ET AL., NAT. METHODS, vol. 6, 2009, pages 343 - 345
ITTNER ET AL., NAT. PROTOC., vol. 2, 2007, pages 1206 - 1215
LM ITTNER ET AL., CELL, vol. 142, 2010, pages 387 - 397
CV VORHEESMT WILLIAMS, NAT PROTOC, vol. 1, 2006, pages 848 - 58
HORNER ET AL., NAT. PROTOC., vol. 8, 2013, pages 1961 - 1984
KE ET AL., ACTA NEUROPATHOL, vol. 130, 2015, pages 661 - 678
VAN EERSEL ET AL., NEUROPATHOL. APPL. NEUROBIOL, vol. 41, 2015, pages 906 - 925
AA ITTNERA GLADBACHJ BERTZLS SUHLM ITTNER, ACTA NEUROPATHOL COMMUN, vol. 2, 2014, pages 149
M WEIERGRABERM HENRYJ HESCHELERN SMYTHT SCHNEIDER, BRAIN RES BRAIN RES PROTOC, vol. 14, 2005, pages 154 - 64
AB TORTR KOMOROWSKIH EICHENBAUMN KOPELL, J NEUROPHYSIOL, vol. 104, 2010, pages 1195 - 210
BRYKSINET, BIOTECHNIQUES, vol. 48, 2010, pages 463 - 465
M AVITZOUR ET AL., FEBS J, vol. 274, 2007, pages 963 - 75
VON JONQUIERES ET AL., PLOS ONE, vol. 8, 2013, pages e65646
G VON JONQUIERES ET AL., PLOS ONE, vol. 8, 2013, pages e65646
T FATHYD KEP GUNNINGJ GOTZLM ITTNER, NAT PROTOC, vol. 4, 2009, pages 78 - 85
WIENKEN ET AL., NAT. COMMUN, vol. 1, 2010, pages 100
DOLAI ET AL., CANCER RES., vol. 76, 2016, pages 2766 - 2777
THINGHOLM ET AL., NAT. PROTOC, vol. 1, 2006, pages 1929 - 1935
MP LAMBERT ET AL., PROC NATL ACAD SCI U S A, vol. 95, 1998, pages 6448 - 53
ITTNER ET AL., ACTA NEUROPATHOL. COMMUN, vol. 2, 2014, pages 149
STURCHLER-PIERRAT ET AL., PROC. NATL. ACAD. SCI. U.S.A., vol. 94, 1997, pages 13287 - 13292
C BALLATOREVM LEEJQ TROJANOWSKI, NATURE REVIEWS. NEUROSCIENCE, vol. 8, 2007, pages 663 - 72
C HAASSDJ SELKOE, NATURE REVIEWS. MOLECULAR CELL BIOLOGY, vol. 8, 2007, pages 101 - 12
K IQBAL, F LIU, CX GONG, C ALONSO ADEL, I GRUNDKE-IQBAL, ACTA NEUROPATHOL, vol. 118, 2009, pages 53 - 69
EM MANDELKOWE MANDELKOW, COLD SPRING HARB PERSPECT MED, vol. 2, 2012, pages a006338
ES MUSIEKDM HOLTZMAN, NAT NEUROSCI, vol. 18, 2015, pages 800 - 6
M RAPOPORT, HN DAWSON, LI BINDER, MP VITEK, A FERREIRA,M RAPOPORT, HN DAWSON, LI BINDER, MP VITEK, A FERREIRA, PROC NATL ACAD SCI USA, vol. 99, 2002, pages 6364 - 9
ED ROBERSON ET AL., SCIENCE, vol. 316, 2007, pages 750 - 4
JJ PALOPL MUCKE, NAT NEUROSCI, vol. 13, 2010, pages 812 - 8
GE HARDINGHAMH BADING: "Nature reviews", NEUROSCIENCE, vol. 11, 2010, pages 682 - 96
Q WANGDM WALSHMJ ROWANDJ SELKOER ANWYL, J NEUROSCI, vol. 24, 2004, pages 3370 - 8
S LI ET AL., J NEUROSCI, vol. 31, 2011, pages 6627 - 38
AA ITTNER, A GLADBACH, J BERTZ, LS SUH, LM ITTNER, NEUROPATHOL COMMUN, vol. 2, 2014, pages 149
MA FABIAN ET AL., NAT BIOTECHNOL, vol. 23, 2005, pages 329 - 36
MB MENONS DHAMIJAA KOTLYAROVM GAESTEL, AUTOPHAGY, 2015
G BUZSAKIEI MOSER, NAT NEUROSCI, vol. 16, 2013, pages 130 - 8
R GOUTAGNYJ JACKSONS WILLIAMS, NAT NEUROSCI, vol. 12, 2019, pages 1491 - 3
RT CANOLTY ET AL., SCIENCE, vol. 313, 2006, pages 1626 - 8
AB TORTRW KOMOROWSKIJR MANNSNJ KOPELLH EICHENBAUM, PROC NATL ACAD SCI USA, vol. 106, 2009, pages 20942 - 7
Y RONG, X LU, A BERNARD, M KHRESTCHATISKY, M BAUDRY, J, NEUROCHEM, vol. 79, 2001, pages 382 - 90
M AARTS ET AL., SCIENCE, vol. 298, 2002, pages 846 - 50
M GOEDERT ET AL., FEES LETT, vol. 409, 1997, pages 57 - 62
S MONDRAGON-RODRIGUEZ ET AL., J BIOL CHEM, vol. 287, 2012, pages 32040 - 53
LM ITTNERJ GOTZ: "Nature reviews", NEUROSCIENCE, vol. 12, 2011, pages 65 - 72
JZ WANGQ WUA SMITHI GRUNDKE-IQBALK IQBAL, FEES LETT, vol. 436, 1998, pages 28 - 34
V BUEE-SCHERRERM GOEDERT, FEES LETT, vol. 515, 2002, pages 151 - 4
A CAVALLINI ET AL., J BIOL CHEM, vol. 288, 2013, pages 23331 - 47
See also references of EP 3423080A4
Attorney, Agent or Firm:
GRIFFITH HACK (AU)
Download PDF:
Claims:
CLAIMS :

A method of treating or preventing a neurological condition mediated by a tau-dependent signalling complex in neurons of a subject, comprising treating the subject to:

a. promote phosphorylation of one or more amino acid residues of tau, wherein the phosphorylation of the amino acid residues causes disruption of the tau-dependent signalling complex in neurons of the subject; or

b. introduce a variant of tau that causes disruption of the tau-dependent signalling complex in neurons of the subject.

The method of claim 1, wherein treating the subject comprises administering an agent which:

a. promotes phosphorylation of one or more amino acid residues of tau, wherein the

phosphorylation of the amino acid residues causes disruption of the tau-dependent signalling complex; or

b. introduces a variant of tau that causes

disruption of the tau-dependent signalling complex in neurons of the subject.

The method of claim 1, wherein the one or more amino acid residues of tau is threonine at position 205 of tau (T205 ) .

The method of any one of claims 1 to 3, wherein the subject is treated by administering an agent that elevates ρ38γ activity, or the activity of a variant of ρ38γ, in neurons of the subject. The method of claim 4, wherein the agent comprises ρ38γ or a variant thereof, or a nucleic acid that is capable of expressing ρ38γ or a variant thereof in neurons of the subject.

The method of claim 4 or 5 , wherein the agent comprises a nucleic acid that is capable of

expressing ρ38γ or a variant thereof in neurons of the sub ect .

The method of any one of claims 4 to 6, wherein ρ38γ comprises the amino acid sequence of SEQ ID NO: 2.

The method of any one of claims 4 to 7, wherein the variant of ρ38γ comprises an amino acid sequence that is at least 60%, 65%, 70%, 75%, 80%, 85%, 90%, 95 or 99% identical to the amino acid sequence of ρ38γ (SEQ ID NO: 2) .

The method of any one of claims 4 to 8, wherein the variant of ρ38γ comprises a PDZ interaction motif.

The method of any one of claims 4 to 9, wherein the variant of ρ38γ is a constitutively active variant of ρ38γ.

The method of claim 10, wherein the constitutively active variant of ρ38γ (p38yCA) comprises SEQ ID NO: 3 The method of any one of claims 1 to 11, wherein the tau-dependent signalling complex comprises PSD-95 and tau .

The method of any one of claims 1 to 12, wherein the tau-dependent signalling complex comprises PSD-95, tau and FYN.

The method of any one of claims 1 to 13, wherein the neurological condition is selected from the group consisting of Alzheimer' s disease, stroke and epilepsy.

A method of treating or preventing a neurological condition mediated by a tau-dependent signalling complex in neurons of a subject, comprising

administering an agent which elevates ρ38γ activity, or the activity of a variant of ρ38γ, in the neurons of the subject.

The method of claim 15, wherein the agent comprises ρ38γ or a variant thereof, or a nucleic acid that is capable of expressing ρ38γ or a variant thereof in neurons of the subject.

The method of claim 15 or 16, wherein the agent comprises a nucleic acid that is capable of

expressing ρ38γ or a variant thereof in neurons of the sub ect .

The method of any one of claims 15 to 17, wherein the variant of ρ38γ comprises an amino acid sequence that is at least 60%, 70%, 75%, 80%, 85%, 90%, 95 or 99% identical to the amino acid sequence of ρ38γ (SEQ ID NO: 2) .

The method of any one of claims 15 to 18, wherein the variant of ρ38γ comprises a PDZ interaction motif.

The method of any one of claims 15 to 19, wherein the variant of ρ38γ is a constitutively active variant of ρ38γ (p38yCA) .

The method of claim 20, wherein the constitutively active variant comprises SEQ ID NO: 3. 22. The method of any one of claims 15 to 21, wherein the tau-dependent signalling complex comprises PSD-95 and tau .

The method of any one of claims 15 to 22, wherein the tau-dependent signalling complex comprises PSD-95, tau and FYN.

The method of any one of claims 15 to 23, wherein neurological condition is selected from the group consisting of Alzheimer' s disease, stroke and epilepsy.

25. A vector for treating or preventing a neurological condition mediated by a tau-dependent signalling complex in neurons of a subject, comprising: a. a nucleic acid sequence encoding ρ38γ or a variant thereof; or

b. a nucleic acid sequence encoding a variant of tau that causes disruption of the tau-depende signalling complex in neurons of the subject.

The vector of claim 25, wherein the nucleic acid sequence is operably linked to a regulatory sequence for expressing the ρ38γ or a variant thereof, or the variant of tau, in neurons of the subject.

The vector of claim 25 or 26, wherein the vector is a viral vector.

The vector of claim 27, wherein the viral vector is an adeno-associated viral (AAV) vector.

The vector of claim 28, wherein the AAV vector is AAV1, AAV9, AAVrhlO or AAVcy5.

An adeno-associated viral vector for treating or preventing a neurological condition mediated by a tau-dependent signalling complex in neurons of a subject, comprising:

a. a nucleic acid sequence encoding ρ38γ or a variant thereof; or

b. a nucleic acid sequence encoding a variant of tau that causes disruption of the tau-dependent signalling complex in neurons of the subject.

31. The vector of claim 30, wherein the nucleic acid

encoding ρ38γ or a variant thereof, or variant of tau, is operably linked to regulatory sequence for expressing the ρ38γ or a variant thereof, of the variant of tau, in neurons of the subject.

The vector of any one of claims 25 to 31, wherein the variant of ρ38γ comprises an amino acid sequence which is at least 60%, 65%, 70%, 75%, 80%, 85%, 90%, 95 or 99% identical to the amino acid sequence of ρ38γ (SEQ ID NO: 1) .

33. The vector of any one of claims 25 to 32, wherein the variant of ρ38γ comprises a PDZ interaction motif.

34. The vector of any one of claims 25 to 33, wherein the variant of ρ38γ is a constitutively active mutant of ρ38γ.

35. The vector of claim 34, wherein the constitutively active mutant of ρ38γ comprises SEQ ID NO: 3.

36. The vector of any one of claims 25 to 35, wherein the tau-dependent signalling complex comprises PSD-95 and tau . 37. The vector of any one of claims 25 to 36, wherein the tau-dependent signalling complex comprises PSD-95, tau and FYN.

38. The vector of any one of claims 25 to 37, wherein the neurological condition is selected from the group consisting of Alzheimer' s disease, stroke and epilepsy. A composition comprising the vector of any one of claims 25 to 38.

A method of disrupting a signalling complex

comprising PSD-95, tau and FYN in a neuron,

comprising contacting the neuron with an agent which a. promotes phosphorylation of one or more amino acid residues of the tau, wherein the phosphorylation of the amino acid residues causes disruption of the signalling complex; or b. introduces a variant of tau that causes disruption of the tau-dependent signalling complex .

The method of claim 40, wherein the amino acid residue of tau that causes disruption of, or reduces formation of, the tau-dependent signalling complex i threonine at position 205 of tau (T205) .

The method of claim 40, wherein the neuron is contacted with an agent that elevates ρ38γ activity, or the activity of a variant of ρ38γ, in the neuron.

The method of claim 41, wherein the agent comprises ρ38γ or a variant thereof, or a nucleic acid that is capable of expressing ρ38γ or a variant thereof, in the neuron. 44. The method of claim 39 or 42, wherein the agent

comprises a nucleic acid that is capable of

expressing ρ38γ, or a variant thereof, in the neuron. The method of any one of claims 41 to 43, wherein ρ38γ comprises the amino acid sequence of SEQ ID NO: 2.

The method of any one of claims 41 to 43, wherein the variant of p38γ comprises an amino acid sequence that is at least 60%, 65%, 70%, 75%, 80%, 85%, 90%, 95 or 99% identical to the amino acid sequence of ρ38γ (SEQ ID NO: 1) .

The method of any one of claims 41 to 43, wherein the variant of ρ38γ comprises a PDZ interaction motif.

48. The method of any one of claims 41 to 43, wherein the variant of ρ38γ is a constitutively active mutant of ρ38γ.

49. The method of claim 47, wherein the constitutively active mutant of ρ38γ (p38yCA) comprises an amino acid substitution of aspartic acid to alanine at position

179 of ρ38γ.

50. The method of any one of claims 39 to 48, wherein the neuron is in a subject.

51. A method of treating Alzheimer's disease in a subject comprising introducing into neurons of the subject: a. a nucleic acid capable of expressing ρ38γ, or a variant thereof; or b. a nucleic acid capable of expressing a variant of tau that causes disruption of the tau- dependent signalling complex.

A method of treating stroke in a subject, comprising introducing into neurons of the subject:

a. a nucleic acid capable of expressing ρ38γ, or a variant thereof; or

b. a nucleic acid capable of expressing a variant of tau that causes disruption of the tau- dependent signalling complex.

A method of treating epilepsy in a subject,

comprising introducing into neurons of the subject: a. a nucleic acid capable of expressing ρ38γ, or a variant thereof; or

b. a nucleic acid capable of expressing a variant of tau that causes disruption of the tau- dependent signalling complex.

An agent which: (a) promotes phosphorylation of one or more amino acid residues of tau, wherein the phosphorylation of the amino acid residues causes disruption of the tau-dependent signalling complex in neurons of the subject; or (b) introduces a variant of tau that causes disruption of the tau-dependent signalling complex in neurons of the subject, for use in the treating or preventing a neurological

condition mediated by a tau-dependent signalling complex in neurons of a subject.

An agent which elevates ρ38γ activity, or the activity of a variant of ρ38γ, in neurons of a subject, for use in the treating or preventing a neurological condition mediated by a tau-dependent signalling complex in neurons of a subject.

An agent which: (a) promotes phosphorylation of one or more amino acid residues of tau, wherein the phosphorylation of the amino acid residues causes disruption of the tau-dependent signalling complex in neurons of the subject; or (b) introduces a variant of tau that causes disruption of the tau-dependent signalling complex in neurons of the subject, for use in disrupting, or reducing formation of, a signalling complex comprising PSD-95, tau and FYN in a neuron.

An agent which: (a) promotes phosphorylation of one or more amino acid residues of tau, wherein the phosphorylation of the amino acid residues causes disruption of the tau-dependent signalling complex in neurons of the subject; or (b) introduces a variant of tau that causes disruption of the tau-dependent signalling complex in neurons of the subject, for use in treating Alzheimer's disease in a subject.

An agent which: (a) promotes phosphorylation of one or more amino acid residues of tau, wherein the phosphorylation of the amino acid residues causes disruption of the tau-dependent signalling complex in neurons of the subject; or (b) introduces a variant of tau that causes disruption of the tau-dependent signalling complex in neurons of the subject, for use in treating stroke in a subject.

An agent which: (a) promotes phosphorylation of one or more amino acid residues of tau, wherein the phosphorylation of the amino acid residues causes disruption of the tau-dependent signalling complex in neurons of the subject; or (b) introduces a variant of tau that causes disruption of the tau-dependent signalling complex in neurons of the subject, for use in treating epilepsy in a subject.

Description:
Use of phosphorylated tau and p38gamma to treat a neurological condition

Field

The present invention relates to a method of treating or preventing a neurological condition mediated by a tau- dependent signalling complex in neurons of a subject, to a vector for treating or preventing such neurological conditions, and to compositions comprising a vector for treating such conditions. Background

Excitotoxicity of neurons is a pathological process by which neurons are damaged or killed by excessive stimulation. Such stimulation occurs when glutamatergic receptors, such as, for example, NMDA-type receptors (NR) , are overact ivated by neurotransmitters such as, for example, glutamic acid. Excitotoxicity can also be induced by excitotoxins such as amyloid-β (Αβ) .

Excitotoxicity is believed to play a prominent role in neurological conditions such as various forms of neurodegenerative disease including Alzheimer's disease (AD), frontotemporal dementia, Huntington's disease,

Parkinson's disease. Excitotoxicity is also associated with epilepsy, and neuronal damage which occurs following stroke.

Alzheimer's disease (AD) is the most prevalent form of dementia and is the most common neurodegenerative disease. AD is estimated to affect as many as 1% of adults 60 years of age and over.

AD is characterised by brain atrophy, neural loss, extracellular Αβ plaques, and intracellular

neurofibrillary tangle (NFTs) containing aberrantly phosphorylated tau.

Tau is an axonal protein that, under non-pathological conditions, regulates microtubule stability and

microtubule dependent processes. Tau has also been found to reside in a post-synapt ic signalling complex that mediates Αβ-induced excitotoxicity, and potentially other excitotoxicity . In AD, tau becomes aberrantly

phosphorylated, and accumulates in the somatodendritic compartments of neurons, aggregates and eventually forms neurofibrilar tangles (NFT) . Progression of NFT pathology throughout the brain correlates with disease progression in Alzheimer's disease.

The prevailing theory in AD is that Αβ triggers toxic events including tau phosphorylation causing neuronal dysfunction and death. In support, depleting tau prevents Αβ toxicity in AD mouse and cell culture models. Αβ- toxicity in AD is therefore considered in the art to be mediated by phosphorylated tau in the pathogenesis of AD.

It would be advantageous to provide alternative methods of treating AD and other neurological conditions.

Summary

The inventors have found that, contrary to the teaching in the art, phosphorylation of tau at particular amino acid residues causes disruption of tau-dependent signalling complexes, and prevents or reduces

excitotoxicity and Αβ-induced toxicity.

A first aspect provides a method of treating or preventing a neurological condition mediated by a tau- dependent signalling complex in neurons of a subject, comprising treating the subject to:

(a) promote phosphorylation of one or more amino acid residues of tau, wherein the phosphorylation of the amino acid residues causes disruption of the tau-dependent signalling complex in neurons of the subject; or

(b) introduce a variant of tau that causes disruption of the tau-dependent signalling complex in neurons of the subject.

A second aspect provides a method of treating or preventing a neurological condition mediated by a tau- dependent signalling complex in neurons of a subject, comprising administering an agent which:

(a) promotes phosphorylation of one or more amino acid residues of tau, wherein the

phosphorylation of the amino acid residues causes disruption of the tau-dependent signalling complex in neurons of the subject; or

(b) introduces a variant of tau that causes

disruption of the tau-dependent signalling complex in neurons of the subject.

An alternative second aspect provides an agent which: (a) promotes phosphorylation of one or more amino acid residues of tau, wherein the phosphorylation of the amino acid residues causes disruption of the tau-dependent signalling complex in neurons of the subject; or (b) introduces a variant of tau that causes disruption of the tau-dependent signalling complex in neurons of the

subject, for use in the treating or preventing of a neurological condition mediated by a tau-dependent signalling complex in neurons of a subject, or use of an agent which: (a) promotes phosphorylation of one or more amino acid residues of tau, wherein the phosphorylation of the amino acid residues causes disruption of the tau-dependent signalling complex in neurons of the subject; or (b) introduces a variant of tau that causes disruption of the tau-dependent signalling complex in neurons of the

subject, in the manufacture of a medicament for the treating or preventing of a neurological condition

mediated by a tau-dependent signalling complex in neurons of a subject.

A third aspect provides a method of treating or preventing a neurological condition mediated by a tau- dependent signalling complex in neurons of a subject, comprising administering an agent which elevates ρ38γ activity, or the activity of a variant of ρ38γ, in the neurons of the subject. An alternative third aspect provides an agent which elevates ρ38γ activity, or the activity of a variant of ρ38γ, in neurons of a subject, for use in the treating or preventing of a neurological condition mediated by a tau- dependent signalling complex in neurons of a subject, or use of an agent which elevates ρ38γ activity, or the activity of a variant of ρ38γ, in neurons of a subject, in the manufacture of a medicament for the treating or preventing of a neurological condition mediated by a tau- dependent signalling complex in neurons of a subject.

A fourth aspect provides a vector for treating or preventing a neurological condition mediated by a tau- dependent signalling complex in neurons of a subject, comprising :

(a) a nucleic acid sequence encoding ρ38γ or a

variant thereof; or

(b) a nucleic acid sequence encoding a variant of tau that causes disruption of the tau-dependent signalling complex in neurons of the subject.

A fifth aspect provides an adeno-associated viral vector for treating or preventing a neurological condition mediated by a tau-dependent signalling complex in neurons of a subject, comprising:

(a) a nucleic acid sequence encoding ρ38γ or a variant thereof; or

(b) a nucleic acid sequence encoding a variant of tau that causes disruption of the tau-dependent signalling complex in neurons of the subject. A sixth aspect provides a method of disrupting, or reducing formation of, a signalling complex comprising PSD-95, tau and FYN in a neuron, comprising contacting the neuron with an agent which:

(a) promotes phosphorylation of one or more amino acid residues of the tau, wherein the phosphorylation of the amino acid residues causes disruption of the signalling complex; or

(b) introduces a variant of tau that causes disruption of the tau-dependent signalling complex.

An alternative sixth aspect provides an agent which: (a) promotes phosphorylation of one or more amino acid residues of tau, wherein the phosphorylation of the amino acid residues causes disruption of the tau-dependent signalling complex in neurons of the subject; or (b) introduces a variant of tau that causes disruption of the tau-dependent signalling complex in neurons of the

subject, for use in disrupting, or reducing formation of, a signalling complex comprising PSD-95, tau and FYN in a neuron, or use of an agent which (a) promotes

phosphorylation of one or more amino acid residues of tau, wherein the phosphorylation of the amino acid residues causes disruption of the tau-dependent signalling complex in neurons of the subject; or (b) introduces a variant of tau that causes disruption of the tau-dependent signalling complex in neurons of the subject, in the manufacture of a medicament for disrupting, or reducing formation of, a signalling complex comprising PSD-95, tau and FYN in a neuron.

A seventh aspect provides a method of treating

Alzheimer's disease in a subject comprising administering an agent which:

(a) promotes phosphorylation of one or more amino acid residues of tau, wherein the phosphorylation of the amino acid residues causes disruption of the tau- dependent signalling complex in neurons of the subject; or

(b) introduces a variant of tau that causes disruption of the tau-dependent signalling complex in neurons of the subject.

An alternative seventh aspect provides an agent which: (a) promotes phosphorylation of one or more amino acid residues of tau, wherein the phosphorylation of the amino acid residues causes disruption of the tau-dependent signalling complex in neurons of the subject; or (b) introduces a variant of tau that causes disruption of the tau-dependent signalling complex in neurons of the

subject, for use in treating Alzheimer's disease in a subject, or use of an agent which (a) promotes

phosphorylation of one or more amino acid residues of tau, wherein the phosphorylation of the amino acid residues causes disruption of the tau-dependent signalling complex in neurons of the subject; or (b) introduces a variant of tau that causes disruption of the tau-dependent signalling complex in neurons of the subject, in the manufacture of a medicament for treating Alzheimer's disease in a subject. An eighth aspect provides a method of treating

Alzheimer's disease in a subject comprising introducing into neurons of the subject:

(a) a nucleic acid capable of expressing ρ38γ, or

variant thereof; or

(b) a nucleic acid capable of expressing a variant of tau that causes disruption of the tau-dependent signalling complex. An alternative eighth aspect provides: (a) a nucleic acid capable of expressing ρ38γ, or variant thereof; or (b) a nucleic acid capable of expressing a variant of tau that causes disruption of the tau-dependent signalling complex, for use in treating Alzheimer's disease in a subject, or use of (a) a nucleic acid capable of expressing ρ38γ, or variant thereof; or (b) a nucleic acid capable of

expressing a variant of tau that causes disruption of the tau-dependent signalling complex, in the manufacture of a medicament for treating Alzheimer's disease in a subject.

A ninth aspect provides a method of treating stroke in a subject comprising administering an agent which:

(c) promotes phosphorylation of one or more amino acid residues of tau, wherein the phosphorylation of the amino acid residues causes disruption of the tau- dependent signalling complex in neurons of the subject; or

(d) introduces a variant of tau that causes disruption of the tau-dependent signalling complex in neurons of the subject.

An alternative ninth aspect provides an agent which: (a) promotes phosphorylation of one or more amino acid

residues of tau, wherein the phosphorylation of the amino acid residues causes disruption of the tau-dependent signalling complex in neurons of the subject; or (b) introduces a variant of tau that causes disruption of the tau-dependent signalling complex in neurons of the

subject, for use in treating stroke in a subject, or use of an agent which (a) promotes phosphorylation of one or more amino acid residues of tau, wherein the

phosphorylation of the amino acid residues causes disruption of the tau-dependent signalling complex in neurons of the subject; or (b) introduces a variant of tau that causes disruption of the tau-dependent signalling complex in neurons of the subject, in the manufacture of a medicament for treating stroke in a subject.

A tenth aspect provides a method of treating stroke in a subject, comprising introducing into neurons of the subject :

(a) a nucleic acid capable of expressing ρ38γ, or

variant thereof; or

(b) a nucleic acid capable of expressing a variant of tau that causes disruption of the tau-dependent signalling complex.

An alternative tenth aspect provides: (a) a nucleic acid capable of expressing ρ38γ, or variant thereof; or (b) a nucleic acid capable of expressing a variant of tau that causes disruption of the tau-dependent signalling complex, for use in treating stroke in a subject, or use of (a) a nucleic acid capable of expressing ρ38γ, or variant thereof; or (b) a nucleic acid capable of expressing a variant of tau that causes disruption of the tau-dependent signalling complex, in the manufacture of a medicament for treating stroke in a subject.

An eleventh aspect provides a method of treating

epilepsy in a subject comprising introducing into neurons of the subject an agent which:

(a) promotes phosphorylation of one or more amino acid residues of tau, wherein the phosphorylation of the amino acid residues causes disruption of the tau- dependent signalling complex in neurons of the subject; or

(b) introduces a variant of tau that causes disruption of the tau-dependent signalling complex in neurons of the subject.

An alternative eleventh aspect provides an agent which: (a) promotes phosphorylation of one or more amino acid residues of tau, wherein the phosphorylation of the amino acid residues causes disruption of the tau-dependent signalling complex in neurons of the subject; or (b) introduces a variant of tau that causes disruption of the tau-dependent signalling complex in neurons of the

subject, for use in treating epilepsy in a subject, or use of an agent which (a) promotes phosphorylation of one or more amino acid residues of tau, wherein the

phosphorylation of the amino acid residues causes

disruption of the tau-dependent signalling complex in neurons of the subject; or (b) introduces a variant of tau that causes disruption of the tau-dependent signalling complex in neurons of the subject, in the manufacture of a medicament for treating epilepsy in a subject.

A twelfth aspect provides a method of treating epilepsy in a subject, comprising introducing into neurons of the subject :

(a) a nucleic acid capable of expressing ρ38γ, or

variant thereof; or

(b) a nucleic acid capable of expressing a variant of tau that causes disruption of the tau-dependent signalling complex. An alternative twelfth aspect provides: (a) a nucleic acid capable of expressing ρ38γ, or variant thereof; or (b) a nucleic acid capable of expressing a variant of tau that causes disruption of the tau-dependent signalling complex, for use in treating epilepsy in a subject, or use of (a) a nucleic acid capable of expressing ρ38γ, or variant thereof; or (b) a nucleic acid capable of expressing a variant of tau that causes disruption of the tau-dependent signalling complex, in the manufacture of a medicament for treating epilepsy in a subject.

A thirteenth aspect provides a composition for treating or preventing a neurological condition mediated by a tau- dependent signalling complex in neurons of a subject,

comprising an agent which:

(a) promotes phosphorylation of one or more amino

acid residues of tau, wherein the

phosphorylation of the amino acid residues causes disruption of the tau-dependent signalling complex in neurons of the subject; or

(b) introduces a variant of tau that causes

disruption of the tau-dependent signalling complex in neurons of the subject. A fourteenth aspect provides a composition comprising a vector described herein.

A fifteenth aspect provides a kit for treating or

preventing a neurological condition mediated by a tau-dependent signalling complex in neurons of a subject, comprising an agent which:

(a) promotes phosphorylation of one or more amino

acid residues of tau, wherein the

phosphorylation of the amino acid residues causes disruption of the tau-dependent

signalling complex in neurons of the subject; or (b) introduces a variant of tau that causes

disruption of the tau-dependent signalling complex in neurons of the subject.

A sixteenth aspect provides a kit comprising a vector described herein. A seventeenth aspect provides a transgenic non-human animal comprising a transgenic nucleic acid sequence which is capable of expressing in neurons of the transgenic animal ρ38γ or a variant thereof, or a variant of tau that causes disruption of the tau-dependent signalling complex.

An eighteenth aspect provides a method of assessing whether a neurological condition can be treated or

prevented by a method described herein, comprising the steps of:

(a) providing a test animal suffering from the

neurological condition or exhibiting a phenotype which is a model for the neurological condition; (b) crossing the test animal with a transgenic animal to obtain progeny, the transgenic animal comprising a transgenic nucleic acid sequence which is capable of expressing in neurons of the animal ρ38γ or a variant thereof, or a variant of tau that causes disruption of the tau-dependent signalling complex; and

(c) assessing the severity of the neurological

condition or the phenotype which is a model for the neurological condition in progeny expressing the transgenic nucleic acid sequence. Description of the Drawings

Figure 1A is a schematic diagram showing the domain structure of p38 MAP kinases including a dendrogram showing degree of similarity. As can be seen, ρ38γ has a unique C-terminal PDZ interaction motif.

Figure IB shows the results of polymerase chain reaction (PCR) on genomic DNA from mice with targeted alleles for p38a, ρ38β, ρ38γ and ρ38δ. f, floxed allele, -, knockout allele, +, wild-type allele.

Figure 1C shows the results of western blots of cortical extracts of control mice (f/f or +/+) confirmed expression of ρ38 , ρ38β and ρ38γ, but not ρ38δ in brains. Antibody specificity was shown by probing extracts of mice with individual knockout or p38 MAPKs . Aneu, neuron-specific knockout of p38 . GAPDH showed equal loading. BM, bone marrow .

Figure 2A are graphs showing reduced seizure latency

(left) and linear regression slopes (right) of p38y + + and p38y mice injected with 30mg/kg PTZ. Mean seizure severity was markedly increased in ρ38γ ~/~ compared to ρ38γ +/+ mice injected with 30mg/kg PTZ (**P < 0.01; ****p < 0.0001; n=10-12) .

Figure 2B are photographs showing co-localization of ρ38γ and post-synaptic PSD-95 (arrows), but not pre-synaptic synaptophysin (Syp) in neurons. Scale bar, Ιμπι.

Figure 2C is a graph showing early mortality in

APP23.ρ38γ +/+ (n=62) was further augmented in APP23.ρ38γ ~/~ (n=43) mice, while ρ38γ +/+ (n=49) and 38y ~ (n=48) mice presented with normal survival (****p < 0.0001, ***p < 0.001) . Figures 2D—F show the spatial working memory deficits in APP23.ρ38γ +/+ (n=10), and more so APP23.ρ38γ ~/~ (n=8) compared to p38y + + (n=10) and ρ38γ ~/~ (n=10) mice using Morris-water-maze (MWM) (**P < 0.01; *P < 0.05) .

Figure 2D is representative MWM path traces. Dashed squares, location of hidden platform.

Figure 2E is a graph showing escape latency was increased in APP23.p38y + + , and more so in APP23 ,ρ38γ ~/~ mice, but comparable to p38y + + and ρ38γ ~/~ mice.

Figure 2F is a graph showing the time in quadrant

(seconds) in a MWM test of p38y- +/+ , p38y ~7~ , APP23p38y- +/+ and APP23p38y _ _ mice. APP23 ,ρ38γ ~/~ mice spent less time in the targeted (Ql) and more time in the opposite quadrant (Q4) during probe trials, compared to APP23.p38y + + , p38y + + and ρ38γ ~/~ mice.

Figure 2G shows representative EEG traces of APP23.p38y + + , APP23.ρ38γ ~/~ , ρ38γ +/+ and p38y ~ mice, with bouts of hypersynchronicity (green) in APP23.p38y + + and APP23. ρ38γ ~ _ , but not p38y + + and ρ38γ ~/~ mice.

Figure 2H is a graph showing markedly increased numbers of spike trains in APP23 ,ρ38γ ~/~ compared to APP23.p38y + + mice (n=6-8; **p < 0.01) . No spike trains were detected in p38y + + and ρ38γ ~/~ recordings .

Figure 21 is a graph showing the number of spikes per minute was increased in APP23 ,ρ38γ ~/~ compared to

APP23.p38y + + mice, but rare in p38y + + and ρ38γ ~/~ mice (n=6-8; ***P < 0.001 **P < 0.01, *P < 0.05) .

Figure 2J shows a representative phase-amplitude

comodulograms computed for interictal hippocampal LFPs recordings showing reduced cross-frequency coupling (CFC) around 8 Hz in APP23.p38y + + compared to p38y + + and ρ38γ ~/~ mice. CFC at ~8Hz was virtually lost in APP23 ,ρ38γ ~/~ mice. Figure 2K is a graph showing the modulation index computed for phase-amplitude distributions was reduced in

APP23.ρ38γ +/+ (n=8) and more so in APP23.ρ38γ ~/~ (n=8) compared to ρ38γ +/+ (n=6) and 38y ~ (n=6) mice (***p < 0.001, **P < 0.01) .

Figure 3A is a graph showing seizure latencies in

p38y _ ~ , Alzl7.p38y + + and Alzl7.p38y _ _ mice following i.p. administration of 30 mg/kg PTZ. Further reduction in seizure latencies following 30 mg/kg PTZ i.p. was observed in Alz 17 ,ρ38γ ~/~ mice compared to those already reduced in ρ38γ ~ ~ compared to ρ38γ + + and Alz 17.ρ38γ + + mice (n=10-12; **P < 0.01; *P < 0.05; ns, not significant) .

Figure 3B is a graph showing linear regression analysis of seizure latency curves in (3A) (n=10-12; ***p < 0.001; **p < 0.01) .

Figure 3C is a graph showing further enhanced mean seizure severity after 30 mg/kg PTZ in Alz 17 ,ρ38γ ~/~ mice compared to those already increased in ρ38γ ~/~ compared to ρ38γ +/+ and Alz 17.ρ38γ +/+ mice (n=10-12; ***P < 0.001, **P < 0.01, *P < 0.05) .

Figure 3D is a graph showing seizure latencies after

30mg/kg PTZ were profoundly reduced in tau + + . ρ38γ ~/~ compared to tau + + ,ρ38γ +/+ mice, and were markedly increased in both tau _/" .ρ38γ +/+ and tau ' ^ .ρ38γ ' ^ mice (n=10-12; **P < 0.01; *P < 0.05) .

Figure 3E is a graph showing a linear regression analysis of seizure latency curves in (D) (n=10-12; ***p < 0.001; **p < 0.01; ns, not significant)

Figure 3F is a graph showing mean seizure severity was increased in tau + + ,ρ38γ ~/~ compared to tau + + .ρ38γ +/+ mice, but was similarly reduced in tau / .p38y^ / ^ and tau / .p38y ^/^ mice after 30 mg/kg PTZ injection (n=10-l; ***P < 0.001; **P < 0.01; *P < 0.05) . Figure 3G is a graph showing percent survival of

APP23.ρ38γ ~7~ . tau ~7~ mice compared with APP23.p38 y +/+ tau ~7~ , APP23.p38 γ +/+ and APP23. p38 mice over 300 days.

Figure 3H is a graph showing escape latency of p38 γ + + , ρ38γ ~7~ , APP23.ρ38γ +/+ , APP23. ρ38γ ~/~ , ΑΡΡ23. ρ38γ +/+ . tau ~/~ and ΑΡΡ23.ρ38γ ~/~ . ta _/~ mice following Morris Water Maze (MWM) test .

Figure 31 is a graph showing time in quadrant during MWM test for ρ38γ +/+ , ρ38γ ~7~ , APP23.ρ38γ +/+ , APP23.ρ38γ ~ ~ ,

ΑΡΡ23. ρ38γ +/+ . tau ~/~ and ΑΡΡ23. ρ38γ ~/~ . tau ~/~ mice.

Figure 4A is a photograph showing that more PSD-95/tau/Fyn complexes were immunoprecipitated from Alz 17 ,ρ38γ ~/~ than Alz 17 ,ρ38γ +/+ brains, despite comparable total levels of PSD-95, tau and Fyn. GAPDH confirmed equal loading.

Figure 4B is a graph showing quantification of tau and Fyn bound to PSD-95 detected in (4A) (n=6; ***p < 0.001; *P < 0.05) .

Figure 4C is a photograph showing the results of

immunoprecipitat ion (IP) of PSD-95/tau/Fyn complexes from cells transfected with FLAG-PSD-95, tau and Fyn. Co- transfection of wild-type ρ38γ (WT) mitigated, and of constitutive active ρ38γ (CA) abolished, PSD-95/tau/Fyn interaction .

Figure 4D is a graph showing quantification of tau and Fyn bound to PSD-95 detected in (C) (n=6; ***p < 0.001; **P < 0.01; *P < 0.05) .

Figure 4E shows that ρ38γ WT and CA ρ38γ failed to disrupt PSD-95/tau/Fyn complexes immunoprecipitated from cells in the presence of p38 inhibitor.

Figure 4F is a graph showing quantification of tau and Fyn bound to PSD-95 detected in (E) (n=6; ***p < 0.001; **P < 0.01; *P < 0.05) . Figure 4G shows that consistently more PSD-95/tau/Fyn complexes were immunoprecipitated from cortical lysates of ρ38γ ~/~ than ρ38γ +/+ mice 0, 5 and 15 minutes after

injection with 30 mg/kg PTZ.

Figure 4H is a graph showing quantification of tau and Fyn bound to PSD-95 detected in (4G) (n=6; ***p < 0.001; **P < 0.01; *P < 0.05) .

Figure 41 shows more tau, Fyn, NMDA receptor subunits 1 (NR1) and 2B (NR2B) we immunoprecipitated in complexes with PSD-95 from brains of 38y ~ than ρ38γ +/+ mice. This was further enhanced in APP23.ρ38γ / compared to

APP23.ρ38γ +/+ mice. Total levels of APP (22C11), PSD-95, tau, Fyn, NR1, NR2B and ρ38γ were, however, comparable in 38y ~ , ρ38γ +/+ , APP23.ρ38γ ~/~ and APP23.ρ38γ +/+ mice.

Figure 4J is a graph showing quantification of tau, Fyn, NR1 and NR2B bound to PSD-95 detected in (41) (n = 6-8; ***P < 0.001; **P < 0.01; *P < 0.05) .

Figure 5A shows cells transfected with tau and wild-type (WT) or constitutive active (CA) ρ38γ are predominantly being phosphorylated at T205 and less at S199, but

virtually not at S396 and S404. GAPDH showed equal

loading .

Figure 5B shows the results of immunoprecipitation of PSD95/tau/Fyn complexes from cells co-transfected with PSD95, Fyn and wild-type or mutant human tau (S199A, S199D, T205A, T205E) . The results show that mimicking phosphorylation at T205 (T205E) quantitatively disrupted the interaction of PSD95, Fyn and tau, while the tau variant T205A increased it. Mutation of S199 had no effect on PSD-95/tau/Fyn complexes. Figure 5C is a graph showing quantification of tau and Fyn bound to PSD-95 detected in (Fig. 5B) (n=6; ***P < 0.001; *P < 0.05; ns, not significant) .

Figure 5D shows the results of imnmunoprecipitat ion of PSD95/tau/Fyn complexes from cells co-transfected with

PSD95, Fyn, wild-type or mutant human tau, with or without ρ38γ εΑ . Co-expression of PSD-95, Fyn and WT tau with ρ38γ εΑ abolished PSD-95/tau/Fyn complex formation, while transfection of T205A tau completely prevented the effects of p38Y CA on PSD-95/T205A tau/Fyn interaction.

Figure 5E is a graph showing quantification of tau and Fyn bound to PSD-95 detected in Figure 5D (n=4; ***p < 0.001; **P < 0.01) .

Figure 5F is a graph showing the effect of tau variants on Αβ-induced toxicity as determined by LDH release in hippocampal neurons. Αβ (0.05 or 0.5μΜ) -induced toxicity (measured by LDH release) was reduced in T205E compared to WT and T205A tau-expressing neurons. Cytotoxicity induced by H2O2 (3uM) was similar for all tau variants. (n=6 independent experiments; **p < 0.01; *P < 0.05) .

Figure 5G is an image showing localization of ρ38γ and ρ38γ εΑ in cultured hippocampal neurons. Both, AAV-expressed WT and constitutive active (CA) ρ38γ localized to

dendritic spines in cultured hippocampal neurons ^3Tub, β3^ιΛυ1ίη) , similar to endogenous ρ38γ (see Fig. 2B) .

Control neurons expressed AAV18 GFP . Scale bar, 1 μπι.

Figure 5H is a graph showing expression of ρ38γ WT and more so of ρ38γ εΑ reduced toxicity induced by Αβ (0.05 or 0.5μΜ) but not H2O2 (3uM) in hippocampal neurons,

determined by LDH release. (n=6 independent experiments; ***P < 0.001; **P < 0.01) .

Figure 51 is a graph showing expression of ρ38γ and more so of ρ38γ εΑ in C57B1/6 brains increased seizure latencies after administration of PTZ (50mg/kg i.p.) compared to mice that received AAV-GFP (n=8-10; **P < 0.01) .

Figure 5J is a graph showing the results of linear

regression analysis of seizure latency curves in Figure 51 (n=8-10; . ***P < 0.001; **P < 0.01) .

Figure 5K is a graph showing the degree of improvement in seizure latencies (linear regression slopes) mediated by expression of both WT and CA ρ38γ positively correlated with level of ρ38γ expression in individual mice

challenged with 50mg/kg PTZ (n=8-10; *P < 0.05) .

Figure 6A is a graph showing the susceptibility of ρ38γ knockout mice and ρ38γ + + control mice to excitotoxic seizures induced following i.p injection of 50mg/kg body weight pentylenetetrazole (PTZ) . Seizure latency was reduced in ρ38γ knockout (ρ38γ ~/~ ) as compared to control (ρ38γ+/+) mice following 50mg/kg PTZ (*P < 0.05; n=9-10) . Figure 6B is a graph showing the results of linear

regression of analysis of seizure latency curves in Figure 6A ( ****p < 0.0001; n=9-10) .

Figure 6C is a graph showing mean seizure severity in p38y + + and p38y mice following i.p. administration of 50 mg/kg PTZ. Mean seizure severity was reduced in p38y + + mice compared to p38y mice (*P<0.05) .

Figure 7A is a graph showing the susceptibility of ρ38γ _/~ knockout APP23 mice and ρ38γ + + APP23 mice to excitotoxic seizures induced following i.p injection of 30mg/kg body weight PTZ. ρ38γ _ ~ and APP23.ρ38γ + + presented similar reduced seizure latencies compared to non-transgenic ρ38γ + + mice after PTZ injection. The seizure latency was even further reduced in APP23 ρ38γ 1 mice (n=10-12; **P < 0.01; *P < 0.05) . Figure 7B is a graph showing the results of linear

regression analysis of seizure latency curves in Figure 7A (n=10-12; ***P < 0.001; *P < 0.05) .

Figure 7C is a graph showing mean seizure severity

following PTZ administration (30mg/kg BW i.p.) . Seizure severity was significantly increased in ρ38γ ~/~ and

APP23.p38y + + compared to non-transgenic ρ38γ +/+ mice (n=10- 12; ***P < 0.001; *P < 0.05) . APP23.ρ38γ ~/~ mice showed a trend to even further enhanced seizure.

Figure 8A is a graph showing the length of swim paths of ρ38γ +/+ , ρ38γ ~7~ , APP23.ρ38γ +/+ and APP23.ρ38γ ~7~ mice in a Morris-water-maze (MWM) to assess memory impairment.

Longer swim paths indicated memory acquisition deficits in APP23.ρ38γ +/+ , that were worse in APP23.ρ38γ ~/~ mice, compared to normal learning in ρ38γ +/+ and ρ38γ ~/~ mice (**p < 0.01; *P < 0.05; ns, not significant) .

Figure 8B is a graph showing escape latencies, and Figure 8C is a graph showing average speeds, of ρ38γ + + , ρ38γ _ ~ , APP23. p38γ + + and ΑΡΡ23.ρ38γ mice in the Morris-water-maze (MWM) . Escape latencies and average speeds were similar using visual cued platform, confirming visual and motor competency .

Figure 9A is a diagram of a representative raw interictal EEG (LFP), band pass filtered signals for theta (4-12 Hz) and gamma (25-100 Hz) oscillations, gamma amplitude envelope and theta phase in APP23.ρ38γ +/+ and APP23.ρ38γ / and non-transgenic control ρ38γ +/+ and ρ38γ ~/~ mice.

Figure 9B is a graph showing spectral power analysis of interictal EEGs showed a shift to lower theta frequencies in APP23.ρ38γ +/+ and APP23.ρ38γ ~/~ compared to ρ38γ +/+ and ρ38γ ~/~ recordings (n=6-8) . Dashed boxes mark low and high theta bands . Figure 9C is a graph showing quantification (area-under- curve, AUC) of spectral power of low frequency theta (4-8 Hz) in APP23.p38y + + and more so in APP23 ,ρ38γ ~/~ compared to ρ38γ +/+ and 38y ~ recordings (***p < 0.001) .

Figure 9D is a graph showing that spectral power of high frequency theta power (8-12 Hz) in Figure 9B was decreased in APP23.ρ38γ +/+ and APP23.ρ38γ ~/~ compared to ρ38γ +/+ and ρ38γ ~/~ recordings (***p < 0.001; ns, not significant) .

Note that aberrant power of high frequency theta (8-12 Hz) in APP23 mice was not affected by deletion of ρ38γ.

Figure 9E is a graph showing gamma spectral power analysis of interictal EEG in APP23.ρ38γ +/+ , APP23.ρ38γ ~/~ , ρ38γ +/+ and ρ38γ ~/~ recordings (n=6-8) . Dashed boxes mark gamma band .

Figure 9F is a graph showing quantification (AUC) of the graph shown in Figure 9E . The results showed increased spectral power of gamma (25-100 Hz) in APP23.ρ38γ+/+ and APP23 ,ρ38γ ~/~ compared to ρ38γ +/+ and ρ38γ ~/~ recordings (***p < 0.001) .

Figure 9G is a graph showing phase-amplitude plot computed for interictal hippocampal LFPs recordings showing a reduction in APP23.ρ38γ +/+ (n=8) and loss in APP23.ρ38γ ~/~ (n=8) of phase-amplitude coupling (CFC) compared to ρ38γ +/+ (n=6) and ρ38γ-/- (n=6) mice.

Figure 10A shows immunoblots in which both full-length

(FL) WT and CA ρ38γ precipitated together with PSD-95 from cells transfected with PSD-95 and ρ38γ variants. Notable, deletion of the C-terminal PDZ-binding motif (ΔΡπι) in both WT and CA ρ38γ abolished the interaction with PSD-95.

Figure 10B shows immunoblots in which both WT and CA ρ38γ precipitated together with tau from cells transfected with V5-tagged tau and ρ38γ variants. GAPDH confirmed equal loading . Figure IOC shows immunoprecipitation (IP) of PSD-95/tau complexes from cells transfected with PSD-95 and tau

(hTau40) . Co-transfection of wild-type ρ38γ (WT) mitigated and of constitutive active ρ38γ (CA) abolished PSD-95/tau interaction. GAPDH confirmed equal loading. Figure IOC also shows a graph showing quantification of tau bound to PSD-95 as detected in IPs (n=5; ***P < 0.001; **P < 0.01) . Figure 10D is an immunoblot showing Fyn and both, WT and CA ρ38γ precipitated together with tau from cells

transfected with V5-tagged tau, Fyn and ρ38γ variants. GAPDH confirmed equal loading.

Figure 11A is schematic diagram of tau domains and major phosphorylation sites, including non-SP/TP and SP/TP sites. N1/N2: N-terminal inserts encoded by exons 2/3;

Pro: proline-rich domain; Rl-4: microtubule-binding repeats .

Figure 11B is the results of an in vitro kinase assay using recombinant tau and ρ38γ in absence (-) or presence (+) of adenosinetriphosphate (ATP) and followed by

immunoblotting for ρ38γ, tau (Taul3) and phosphorylation site specific antibodies showed phosphorylation of tau at S199, T205, S396 and S404, but not other sites tested by ρ38γ.

Figure 11C is the results of an in vitro kinase assay using recombinant wild-type tau or variants with indicated serines/threonines mutated to Alanine and ρ38γ in absence (-) or presence (+) of ATP which confirmed site-specific phosphorylation of S199, T205, S396 and S404 by ρ38γ.

Figure 12A shows hippocampal neurons with adeno-associated virus (AAV) -mediated expression of human wildtype (WT) ,

T205A or T205E mutant tau which were exposed to 0.05 μΜ

Αβ42 or vehicle. Cytotoxicity was detected 24 later by EthDl uptake in WT and T205A, but not T205E tau expressing neurons. Scale bar, 10 μπι.

Figure 12B shows immunoblots in which similar expression of WT, T205A and T205E tau was observed in hippocampal neurons. GAPDH confirmed equal loading.

Figure 13 shows lower magnification of cells shown in Figure 5G: Both, AAV-expressed WT and constitutive active (CA) ρ38γ localized to dendritic spines in cultured hippocampal neurons ^3Tub, β3-tubulin) , similar to endogenous ρ38γ (see Figure 1) . Control neurons expressed AAV-GFP. Scale bar, ΙΟμπι. Broken lines indicated optical fields shown at higher magnification in Figure 5G.

Figure 14A shows forbrains of mice infected with AAV constructs. Brains show widespread AAV-mediated

expression of GFP or ΗΑ-ρ38γ. NC, negative control. Scale bar, 250 μπι. Broken lines indicate insets.

Figure 14B is an immunoblot of cortical lysates of mice intracranially injected with AAV carrying GFP, HA-tagged ρ38γ or HA-tagged p38yCA which shows higher expression of ρ38γ than p38yCA. GAPDH confirmed equal loading. Ctrl, lysate from cells transfected with ΗΑ-ρ38γ.

Figure 15 is a graph showing mean seizure severity was significantly reduced in C57B1/6 mice with AAV-mediated expression of p38yCA challenged with PTZ (50mg/kg i.p.) compared to GFP-expressing controls (n=8-10; *P < 0.05) . Figures 16—18 show the spatial working memory deficits in

APP23.AAV GFP , AAV GFP , AAV p38YCA , and APP23. AAV p38YCA mice using Morris-water-maze (MWM) .

Figure 16A is representative MWM path traces for

APP23.AAV GFP , AAV GFP , AAV p38YCA , and APP23. AAV p38YCA mice . Dashed squares is the location of hidden platform. Also shown is a graph showing that escape latency was decreased in AAV GFP , AAV p38YCA , and APP23. AAV p38YCA as compared to APP23.AAV GFP mice.

Figure 17 is a graph showing AAV p38YCA mice spent more time in the targeted (Ql) and less time in the opposite

quadrant (Q4) during probe trials, compared to APP23.AAV GFP mice .

Figure 18 is a graph showing escape latency over 3 days was decreased in AAV GFP , AAV p38YCA , and APP23. AAV p38YCA as compared to APP23.AAV GFP mice.

Figure 19 is graphs showing effect of AAV mediated

expression of tau wild type (tau _/~ .AAV tau WT ) , GFP (tau ^ .AAV GFP), tauT205A (tau ~7~ .AAV tau T205A ) , or tau T205E (tau- _ T205E —/—

.AAV tau ) m tau mice on (A) seizure latency and seizure grade ( (B) is a linear regression of the slopes of

(A) ) ; and (C) mean seizure severity, following PTZ-induced seizures by administration of 50mg/kg of PTZ.

Figure 20 is (A) an image of cross-frequency coupling (CFC) ; and (B) is a graph showing the modulation index, in APP23. AAVp38y CA mice compared with APP23. AAVGFP , AAV. GFP and AAV.p38y CA mice (n=5 to 6) (left) . *P<0.05; ns : not significant. Error bars indicate SEM.

Figure 21 (A) shows stimulus image-location pairing possibilities in differential paired associate learning (dPAL) task in Bussey-Saksida touchscreen operant chamber used in pPAL trial in (B) . + indicates the correct image location pairing and - indicates the incorect pairing. The 6 image location pairings were randomised across trials;

(B) is a graph of the number of correct pPAL trials over time in APP23.ρ38γ +/+ , APP23.ρ38γ ~7~ , ρ38γ +/+ and ρ38γ ~7~ mice during touchscreen operant chamber testing; and (C) is a graph showing area under the curve analysis of correct trials per minute curves in (B) , (n=8-10); ***P<0.001, **P<0.01, *P<0.05, ns: not significant; (left) two-way ANOVA: F (3, 941) =60.90; a=0.05; SAidak post-hoc; (right) one-way ANOVA: F ( 3 , 41 ) =11.43 ; a=0.05; Sidak post-hoc) .

Figure 22 A-C shows the results of a pairwise

discrimination task in Bussey-Saksida touchscreen operant chamber which shows minor impairment of discrimination memory in APP23.p38g-/- mice. (A) shows the stimulus used for the analysis of the pairwise discrimination task; (B) shows a graph of the number of correct trials per minute for consecutive testing days for APP23.ρ38γ + + , APP23.ρ38γ ~ ~ , ρ38γ +/+ and ρ38γ ~7~ mice (η=8; *Ρ<0.05 for ΑΡΡ23.ρ38γ ~7~ vs ρ38γ ~7~ ; α=0.05; F ( 3 , 100 ) =3.561 ; 2-way ANOVA with Sidak's multiple comparisons post-hoc test) ; and (C) is a graph showing area under the curve (AUC) analysis of correct trials per minute curves in (B) (n=8; *P<0.05 for

APP23.ρ38γ ~7~ vs ρ38γ ~7~ α=0.05; F ( 3 , 28 ) =2.984 ; ANOVA with Sidak's multiple comparisons post-hoc test) .

Figure 23 is (A) an image of a representative Western blot of brain extracts from human controls (Braak 0) and humans suffering from Alzheimer's Disease at different

neuropathological disease stages (Braak I-VI) set out in Table 3, (B) is a graph showing the levels of ρ38γ in the western blot in (A) normalised to GAPDH, both (A) and (B) showing markedly reduced levels of ρ38γ as AD advances, and a trend towards reduction in early disease stages, (n=4- 5/group; *P<0.05; ns, not significant; a=0.05;

F ( 3 , 13 ) =5.435 ; ANOVA with Sidak's multiple comparisons post-hoc test) .

Figure 24 is an image of representative EEG (LFP) traces in 4 month-old non-transgenic control p38y ++ and ρ38 ^, and APP23.ρ38γ / *, APP23.ρ38γ /~ , APP23.ρ38γ / * . tau ~ ' ~ and APP23.ρ38γ .tau^ mice. Note that deletion of tau results in absent hypersynchronous activity (grey boxes) .

Figure 25 is a graph showing numbers of hypersynchronous epileptiform activity (spikes per minute) in ρ38γ +/+ ,ρ38γ ~/~ , APP23.ρ38γ +/+ , ΑΡΡ23.ρ38γ ' ' , ΑΡΡ23.ρ38γ +/+ . tau ~/~ and

ΑΡΡ23 ,ρ38γ '/' . tau _/~ mice. Hypersynchronous epileptiform activity in APP23.ρ38γ +/+ . tau'' ' and APP23.ρ38γ '/' . tau'' ' mice were similar to levels seen in nontransgenic control ρ38γ +/+ and ρ38γ ' ' mice (n=6-8; ***P < 0.001; ns, not significant; =0.05; F(5, 223)=45.12; ANOVA with Sidak' s multiple comparisons post-hoc test) .

Figure 26 is graphs showing (A) spectral power analysis of theta frequencies (4 - 12 Hz) in interictal sections of APP23.ρ38γ +/+ , APP23.ρ38γ ' ' , APP23.ρ38γ +/+ . tau ~/~ , APP23.ρ38γ ' /' . tau '/' , ρ38γ +/+ and ρ38γ '/' recordings (η=6-8) . Note that theta shift to lower theta frequencies (4 - 8 Hz) in APP23 recordings was not reversed upon deletion of tau in

APP23.ρ38γ +/+ . tau ' ' ' and APP23.ρ38γ '/' . tau-/-. Dashed boxes mark low and high theta bands; and (B) gamma spectral power (25 - 100 Hz) of interictal sections of APP23.ρ38γ +/+ and more so APP23.ρ38γ ' was reverted in APP23.ρ38γ +/+ . tau ' and APP23.ρ38γ '/' . tau ~/~ to levels of ρ38γ +/+ and ρ38γ '/' recordings (n=6-8) . Dashed boxes mark gamma band.

Figure 27 is an image of a representative phase-amplitude comodulograms of interictal hippocampal LFPs recordings showed reduced and virtually lost cross-frequency coupling (~8Hz) in APP23.ρ38γ +/+ and APP23.ρ38γ '/' respectively compared to ρ38γ +/+ and ρ38γ '/' mice. Deletion of tau resulted in restored cross-frequency coupling in both

APP23.ρ38γ +/+ . tau '' ' and APP23.ρ38γ '/' . tau ~ ~ recordings (n=6- 8) .

Figure 28 are graphs showing (A) the averaged modulation index for coupling of theta phase and gamma amplitude in recordings from APP23.ρ38γ +/+ , APP23.ρ38γ ~/~ ,

APP23.ρ38γ +/+ . tau ~/~ , APP23.ρ38γ ~/~ . tau ~/~ , ρ38γ +/+ and 38y ~ mice (n=6-8; ***P < 0.001; **P < 0.01; ns, not

significant; n=6-8; =0.05; F(5, 111)=17.31; ANOVA with Sidak' s multiple comparisons post-hoc test) . Deletion of tau resulted in restored and similar levels of cross- frequency coupling in both APP23.p38y + + . tau _/~ and

APP23.ρ38γ ~/~ . tau^ mice; and (B) a Phase-amplitude plot showing the relation of gamma amplitude across the theta phase computed for interictal hippocampal LFPs shows reduction in APP23.ρ38γ +/+ (n=8) and loss in APP23.ρ38γ ~/~ (n=8) of phase-amplitude coupling (CFC) compared to ρ38γ +/+ (n=6) and ρ38γ-/- (n=6) mice. However, deletion of tau results in restored coupling across the theta phase in both APP23.ρ38γ +/+ . tau ^ and APP23.ρ38γ ~/~ . tau ^ recordings ( =0.05; F(5, 2790 ) =0.003418 ; 2-way ANOVA with Sidak' s multiple comparisons post-hoc test per phase bin) .

Figure 29 (A) to (C) is graphs showing details on effects of genetic deletion of tau on memory impairment in 12- month-old ρ38γ +/+ , ρ38γ ~7~ , APP23. ρ38γ +/+ , APP23.p38 ' ~ ,

APP23.p38gY ~7~ . tau ~7~ and APP23. p38γ +/+ . tau ~7~ mice using the Morris water maze paradigm. (A) is a graph showing time in all 4 water maze quadrants (Ql-4) during probe trials (n=6-8; **P < 0.01; *P< 0.05; ns, not significant; =0.05; F(5, 184) =0.002783; 2-way ANOVA with Sidak' s multiple comparisons post-hoc test) . (B) is a graph showing escape latencies were similar during visual cued platform

testing, confirming visual competency (n=6-8; **P<0.01 (APP23.ρ38γ ~/~ . tau ^ vs APP23.ρ38γ ~/~ in trial 1); ns, not significant (P=0.5092; APP23. 38y ~ . tau^ vs APP23.ρ38γ ~/~ in trial 3); n=6-10; =0.05; F(5, 145)=7.091; 2-way ANOVA with Sidak' s multiple comparisons post-hoc test per trial) . (C) is a graph showing Average swimming speeds were similar between APP23.ρ38γ +/+ , APP23.ρ38γ ~/~ ,

APP23.ρ38γ +/+ . tau ~/~ , APP23.ρ38γ ~/~ . tau ~/~ , ρ38γ +/+ and 38y ~ mice during MWM testing, confirming motor competence (n=6- 8; =0.05; F(5, 169 ) =0.4651 ; A OVA with Sidak' s multiple comparisons post-hoc test) .

Figure 30 (A) is a schematic of the transgene construct use for the generation of ρ38γ €Α mice by pronuclear

injection into C57B1/6 oocytes. HA-tagged ρ38γ containing the D179A mutation that renders it constitutively active was expressed under control of a neuronspecific murine Thyl.2 (mThyl.2) promoter, and followed by a bovine growth hormone poly-adenylation (pA) sequence. (B) Immunoblots of cortical (CTX) , hippocampal (HC) and cerebellar (CB) brain extracts from non-transgenic (-) and of the ρ38γ €Α .3 ( + ) transgenic mouse line confirmed expression of HA-tagged p38Y CA . HA-p38Y CA expressed in 293T cells was used as a positive control. (C) Image showing immunoprecipitation of ρ38γ from nontransgenic (-) and of p38y A ( + ) brains of transgenic p38y CA mice revealed active ρ38γ in all of p38y ' samples, as detected with an antibody to phosphorylated p38, indicating that the transgenic mice express active ρ38γ.

Figure 31 is (A) representative western blots of co- immunopreciptation of mutated tau variants with PSD- 95/tau/Fyn complexes in 293T cells, co-expressing

individual tau variants together with PSD-95 and Fyn. Only the T205E tau variant abolished complex formation with PSD-95 and Fyn.; and (B) is a graph showing quantification of 4 independent experiments as shown in (A) . The PSD- 95/tau/Fyn complex formation was only significantly disrupted in the presence of the T205E tau variant (n=4; *P < 0.05 (for WT vs T205E) ; ns, not significant; =0.05; F(18, 79) =1.003; ANOVA with Sidak' s multiple comparisons test) .

Figure 32 is (A) an image showing AAV-delivered WT tau, T205A and T205E (as indicated) is broadly expressed in the cortex of 4 month-old tau-/- mice injected intracranially at postnatal day 0. No tau was detected in tau-/- brains injected with AAV GFP (tau ~7~ . AAVGFP ) . DAPI, nuclei. Scale bar, 50 μπι; and (B) is an image showing staining of GFP or HA showing widespread neuronal AAV-mediated expression of GFP or ΗΑ-ρ38γ in brains of mice. Scale bar, 25 μπι; and (C) is an immunoblot of cortical lysates of mice intra ¬ cranially injected on postnatal day 0 with AAV carrying GFP, HA-tagged ρ38γ or HA-tagged p38ycA. HA-tagged p38y showed higher expression of ρ38γ than p38ycA. GAPDH

confirmed equal loading. Ctrl, lysate from cells

transfected with ΗΑ-ρ38γ.

Figure 33 is an immunofluorescence image showing that AAV- delivered ρ38γ αΑ is broadly expressed in murine cortex of 6 month-old APP23 mice injected intracranially with AAV at postnatal day 0. Immunofluorescence staining for HA showed expression of HA-tagged p38y CA throughout the cortex. DAPI, nuclei. Scale bar, 50 μπι.

Figure 34 is graphs showing the results of Morris Water Maze testing of WT or APP23 mice expressing AAV-delivered GFP or p38y CA , in which (A) is a graph showing time in quadrant of mice, and shows that APP23 mice expressing AAV-delivered ρ38γ αΑ (APP23.AAVp38Y CA ) show consolidated memory as compared with APP23 expressing control AAV

(APP23. AAVGFP) when performing MWM probe trials. Time in all 4 water maze quadrants (Ql-4) during probe trials (day 7) is shown for APP23.AAVp38y CA , APP23. AAVGFP and non- transgenic AAVGFP, AAVp38Y CA controls. APP23.AAVp38Y CA mice spend significantly more time in target quadrant Ql as compared with APP23.AAVGFP mice. Note that AAVp38Y CA mice show similar memory performance as AAVGFP mice (*P<0.05 (APP23.AAVp38Y CA vs APP23. AAVGFP in Ql; F(3, 84) =3.494); n=6-10; =0.05; F(3, 108) =4.454; 2-way ANOVA with Sidak' s multiple comparisons post-hoc test per quadrant); (B) shows that APP23.AAVp38Y CA , APP23.AAVGFP, AAVGFP, AAVp38y CA showed similar escape latencies after 3 visual cued trials in the MWM, indicating normal visuosensory function and motor-coordination competency (*P<0.05 (APP23.AAVp38Y CA vs APP23. AAVGFP in trial 1; F(3, 84)=3.494); ns, not

significant (P=0.5092; APP23.AAVp38y CA vs APP23. AAVGFP in trial 3); n=6-10; =0.05; F(3, 84)=0.07474; 2-way ANOVA with Sidak' s multiple comparisons post-hoc test per trial) ; and (C) shows that Average swimming speeds were similar between APP23.AAVp38Y CA , APP23.AAVGFP, AAVGFP and AAVp38Y CA mice during MWM testing, confirming motor competency (P=0.8389; n=6-10; =0.05; F(3, 68)=0.2811; ANOVA with Sidak' s multiple comparisons post-hoc test) . Figure 35 shows (A) reduced spontaneous spikes in EEG recording from APP23.AAVp38Y CA mice compared to

APP23. AAVGFP mice, and no spikes in EEG recording from AAVp38Y CA or AAVGFP-treated wild-type mice (n=5-6;

**P<0.01, *P<0.05; one-way ANOVA: F(3, 68)=301.1; a=0.05; Sidak post-hoc) ; and (B) is a graph showing the spikes/min for AAV.GFP, AAVp38y CA , APP23. AAV . GFP and APP23. AAVp38y CA mice, showing reduced spikes/min for APP23. AAVp38y CA compared to APP23. AAVGFP mice.

Figure 36 shows Theta oscillation power changes of APP23 mice at 4-8 (B) and 8-12Hz (C) were not affected by

AAVp38Y CA expression, with comparable levels in

APP23.AAVp38Y CA and APP23. AAVGFP recordings (*P<0.05; n=5- 6; =0.05; F(3, 47)=3.038; ANOVA with Sidak' s multiple comparisons post-hoc test) .

Figure 37 shows gamma oscillation power (25-100Hz) in APP23.AAVp38Y CA mouse recordings was significantly reduced compared with APP23.AAVGFP mouse recordings. (**P<0.01;

*P<0.05; n=5-6; =0.05; F(3, 26)=6.930; ANOVA with Sidak' s multiple comparisons post-hoc test) .

Figure 38 is a graph showing that AAV-delivered ρ38γ αΑ results in normal cross-frequency coupling in EEG

recordings of APP23 mice. Phase-amplitude correlation showed strong coupling of gamma amplitude along theta phase in APP23.AAVp38Y CA and in AAVGFP and AAVp38y CA recordings, yet not in recordings from APP23. AAVGFP mice (**P<0.01 (APP23.AAVp38Y CA vs APP23.AAVGFP) ; =0.05; F(3, 20) =4.793; 2-way ANOVA with Sidak' s multiple comparisons post-hoc test per phase bin) .

Figure 39 A to E shows active neuronal ρ38γ protects APP23 mice from developing impaired memory function as tested by Morris water maze (MWM) . (A) is representative traces of swim paths of WT (non tg) , p38y CA .3, APP23 and APP23. p38y CA .3 mice in the MWM test showing that APP23.ρ38γ αΑ mice swim shorter paths in the Morris water maze test (day 5) as compared with APP23 mice, indicative of non-impaired learning/memory in these mice. ρ38γ αΑ single transgenic mice showed similar swim path lengths as non-transgenic controls, suggesting that active neuronal ρ38γ does not affect learning functions on a wild-type background.

Representative swim path traces are shown (n=6-12) . (B) is a graph of escape latencies over 6 days, and shows that, consistent with shorter swim paths, escape latencies in APP23.ρ38γ αΑ mice were significantly lower than in APP23 mice, and similar to escape latencies seen in p38y CA and non-transgenic mice (*P<0.05; n=6-12; =0.05; F(3, 168) =4.454; 2-way ANOVA with Sidak' s multiple comparisons post-hoc test) . (C) is a graph of time in quadrant, and shows that APP23.ρ38γ αΑ mice spent significantly more time in the target quadrant during probe trials than APP23 mice, indicating consolidated memory in APP23.ρ38γ αΑ mice, yet not in APP23 mice. APP23.ρ38γ αΑ mice, single

transgenic ρ38γ αΑ and non-transgenic mice spent similar time in the target quadrant (*P<0.05 F(3, 116) =7.028);

n=6-12; =0.05; 2-way ANOVA with Sidak' s multiple

comparisons post-hoc test per quadrant) . (D) is a graph showing that escape latencies converged to similar levels after 3 visual cued trials in all experimental groups, indicating normal visuo-sensory function and motor- coordination in APP23.p38y CA , APP23, p38y CA and non- transgenic mice (*P<0.05 (APP23.ρ38γ αΑ vs APP23 in trial 1; F(3, 87)=3.690); ns, not significant (P=0.7190;

APP23.p38y CA vs APP23 in trial 3); n=6-12; =0.05; F(3, 87) =0.369; 2-way ANOVA with Sidak' s multiple comparisons post-hoc test per trial) . (E) is a graph showing that average swimming speeds were similar in APP23.p38y CA ,

APP23, p38y CA and nontransgenic mice during MWM testing, confirming motor competency (P=0.3221; n=6-12; =0.05;

F(3, 62) =1.187; ANOVA with Sidak' s multiple comparisons post-hoc test) .

Figure 40 is (A) EEG recordings from WT (non-tg) , APP23 mice, single transgenic p38y CA .3, and APP23.p38y CA .3 mice, showing that APP23.p38y CA .3 mice exhibited markedly lower epileptiform activity than APP23 recordings. (n=4-5) . (B) is a graph showing that significantly fewer

hypersynchronous epileptiform discharges were found in recordings from APP23.p38y CA mice compared with APP23 recordings. Single transgenic p38y CA and non-transgenic control recordings did not show hypersynchronous activity (**P<0.01; n=4-5; =0.05; F(3, 22)=11.38; ANOVA with

Sidak' s multiple comparisons post-hoc test) . (C and D) are graphs showing increased theta oscillation power of APP23 was reduced to levels of ρ38γ αΑ and nontransgenic recordings in APP23.ρ38γ αΑ recordings. Specifically, the spectral distribution peak at 4-8Hz in APP23 power spectra was significantly lower in APP23.ρ38γ αΑ spectra (**P<0.01; *P<0.05; n=4-5; =0.05; F(3, 26)=6.930; ANOVA with Sidak' s multiple comparisons post-hoc test) . (E and F) are graphs showing increased gamma oscillation power (25-100Hz) of APP23 was reduced to levels of ρ38γ αΑ and non-transgenic recordings in APP23.ρ38γ αΑ recordings (**P<0.01; *P<0.05; n=4-5; =0.05; F(3, 47)=4.761; ANOVA with Sidak' s multiple comparisons post-hoc test) .

Figure 41 (A) is an image showing comodulogram analysis of cross-frequency coupling showed unaffected coupling of theta oscillations to gamma amplitude in recordings of APP23.ρ38γ αΑ mice in contrast to APP23 recordings.

Representative comodulograms are shown (n=4-5) (B) is a graph in which phase-amplitude correlation showed strong coupling of gamma amplitude along theta phase in

APP23.ρ38γ αΑ and in ρ38γ αΑ and non-transgenic recordings, yet not in recordings from APP23 mice. (**P<0.01

(APP23.AAVp38Y CA vs APP23. AAVGFP ) ; n=4-5; =0.05; F(3, 162) =3.238; 2-way ANOVA with Sidak' s multiple comparisons post-hoc test per phase bin) . (C) is a graph showing average modulation index was significantly higher in

APP23.ρ38γ αΑ recordings as compared with APP23 recordings and reached similar levels as in recordings from single transgenic ρ38γ αΑ and non-transgenic control mice (*P<0.05; n=4-5; =0.05; F(3, 9)=6.370; ANOVA with Sidak' s post-hoc test) . Figure 42 is an immunoblot of extracts from dendritic spines of hippocampal neurons showing p38y enriched with NR1 and PSD-95 in PSD fractions of p38y+/+ synaptosome preparations, yet not in non-PSD fractions (a-syn; a- synuclein) .

Figure 43 shows (A) the nucleic acid sequence (SEQ ID NO: 1) and (B) the amino acid sequence (SEQ ID NO: 2) of full length human wild-type ρ38γ.

Figure 44 shows the amino acid sequence of p38y CA (SEQ ID NO: 3) . The location of the mutation from D to A (D179A) is underlined.

Figure 45 shows the amino acid sequence of full length human tau (top) (SEQ ID NO: 4) and tau T205E (SEQ ID NO: 5) (bottom) . The location of the mutation from T to E in tau T205E is underlined.

Figure 46 is a map of adeno-associated viral vector pAM-

CAG containing wild-type ρ38γ coding sequence. 1. Position 200-1120, CAG-promoter; 2. Position 1176-2372, 3xHA-p38y-wt coding sequence; 3. Position 2395-2970, WPRE; 4. Position 3011-3238, bGH PA; 5. Position 3344-3453, ITR; 6. Position 3655-3459, SV40 promoter; 7. Position 3608-3531, SV40 ORI; 8. Position 4644-4016, ColEl origin; 9. Position 5455- 4796, AmpR; 10. Position 5723-5695, Amp prom; 11. Position 6546-6563, SP 6.

Figure 47 is a map of adeno-associated viral vector pAM- CAG containing the coding sequence of p38y CA

(D179A) (constitutively active variant of ρ38γ) . 1. Position 200-1120, CAG-promoter; 2. Position 1176-2372, 3xHA-p38y CA coding sequence; 3. Position 2395-2970, WPRE; 4. Position 3011-3238, bGH PA; 5. Position 3344-3453, ITR; 6. Position 3655-3459, SV40 promoter; 7. Position 3608-3531, SV40 ORI; 8. Position 4644-4016, ColEl origin; 9. Position 5455- 4796, AmpR; 10. Position 5723-5695, Amp prom; 11. Position 6546-6563, SP 6.

Figure 48 is the nucleic acid sequence of adeno-associated viral vector pAM-CAG containing wild-type ρ38γ coding sequence (SEQ ID NO: 6) .

Figure 49 is the nucleic acid sequence of adeno-associated viral vector pAM-CAG containing the coding sequence of p38y CA (D179A) (constitutively active variant of ρ38γ) (SEQ ID NO: 7) .

Detailed Description

The present invention relates to a method of treating or preventing a neurological condition mediated by a tau- dependent signalling complex in neurons of a subject. The inventors have found that promoting phosphorylation of one or more amino acid residues of tau, wherein the

phosphorylation of the amino acid residues causes

disruption of the tau-dependent signalling complex in neurons of the subject, or introducing a variant of tau that causes disruption of the tau-dependent signalling complex in neurons of the subject, can be used to treat or prevent neurological conditions mediated by a tau- dependent signalling complex, such as AD.

A tau-dependent signalling complex is a post-synaptic signalling complex, typically associated with the N- methyl-D aspartate receptor (NMDA receptor) , which can mediate excitotoxicity in neurons. A signalling complex is a complex of proteins which are involved in

transduction of a signal in a cell. A tau dependent signalling complex requires tau in order to transduce the signal. The tau-dependent signalling complex typically comprises tau as a component of the complex.

Excitotoxicity refers to the process by which neurons are damaged or killed by excessive stimulation of

glutamatergic receptors, such as NMDA receptors, and is mediated via signalling complexes in the postsynaptic space. Neural damage from excitotoxicity is associated with a number of neurological conditions. Neural damage in stroke patients is believed to be caused, at least in part, by overactivation of glutamatergic receptors and associated signalling complexes by excessive amounts of extracellular glutamate that are released immediately following ischaemic stroke. Neural damage in epilepsy is also thought to result from excitotoxicity caused by overactivation of glutamatergic receptors and associated signalling complexes following release of glutamate during epileptic events.

The tau-dependent signalling complex is also thought to mediate amyloid-β (Αβ) toxicity in Alzheimer's disease (AD) . In Alzheimer's disease (AD), amyloid-β (Αβ) has been shown to induce toxicity in neurons through a signalling complex comprising NMDA receptors, PSD-95, tau and FYN.

The tau-dependent signalling complex typically comprises tau. In one embodiment, the tau-dependent signalling complex comprises PSD-95 and tau. In one embodiment, the tau-dependent signalling complex comprises PSD-95, FYN and tau. Typically, the tau-dependent

signalling complex comprises NMDA receptors, PSD-95, tau and FYN.

The neurological condition may be any neurological condition mediated by a tau-dependent signalling complex. Typically, the neurological condition is caused by

neuronal damage from overactivation of the tau-dependent signalling complex. Examples of such conditions include, for example, Alzheimer's disease, frontotemporal dementia, amyotrophic lateral sclerosis, Huntington's disease,

Parkinson's disease, neural damage from stroke and neural damage from epilepsy.

In one embodiment, the neurological condition is Alzheimer's disease. In one embodiment, the neurological condition is stroke .

In one embodiment, the neurological condition is epilepsy .

In one embodiment, the method comprises treating the subject to promote phosphorylation of one or more amino acid residues of tau, wherein the phosphorylation of the amino acid residues causes disruption of the tau-dependent signalling complex. As used herein, "disruption of the tau-dependent signalling complex" refers to an effect which prevents the tau-dependent signalling complex from mediating excitotoxicity and Αβ toxicity, and includes destabilising, dismantling or preventing formation of, the signalling complex. In one embodiment, the one or more amino acid residues of tau that are phosphorylated to cause disruption of the tau-dependent signalling complex are one or more amino acid residues that would be

phosphorylated by the MAP kinase ρ38γ. In one embodiment, the one or more amino acid residues of tau that are phosphorylated to cause disruption of the tau-dependent signalling complex is threonine at position 205 (T205) . In one embodiment, the one or more amino acid residues of tau that is phosphorylated to cause disruption of the tau- dependent signalling complex is threonine at position 205 (T205) and one or more amino acid residues selected from the group consisting of serine at position 199 (S199), serine at position 396 (S396) and serine at position

(S404) . In various embodiments, the amino acid residues of tau that are phosphorylated to cause disruption of the tau-dependent signalling complex are: (a) T205; (b) T205, S199; (c) T205, S199, S396; (d) T205 , SI 99, S396 , S404 ; (e) T205, S199, S404; (f) T205, S396, S404; (g) T205,S396; or (h) T205, S404.

In one embodiment, the subject is treated to promote phosphorylation of tau at one or more amino acid residues, wherein phosphorylation of the amino acid residues causes disruption of the tau-dependent signalling complex in neurons of the brain of the subject.

In one embodiment, the subject is treated by

administering an agent that elevates tau that has been phosphorylated at one or more amino acid residues, wherein the phosphorylation of the amino acid residues causes disruption of the tau-dependent signalling complex.

The agent may comprise, for example, a nucleic acid sequence, a nucleic acid analogue, a protein, a peptide, or a small molecule. Typically, administration of the agent introduces the agent into neurons of the subject.

More typically, administration of the agent introduces the agent into neurons of the brain of the subject.

In some embodiments, the agent comprises a nucleic acid sequence which is introduced into neurons of the subject. The nucleic acid is then transcribed and

translated in the neurons.

In some embodiments, the agent can cross the blood- brain barrier, or can be formulated to cross the blood- brain barrier.

As used herein, a "subject" is a mammal. The mammal can be a human, non-human primate, sheep, mouse, rat, dog, cat, horse, cow, pig, or any other mammals which can suffer from a neurological condition mediated by a tau-dependent signalling complex in neurons. Typically, the subject is a human.

In one embodiment, the subject is treated by

administering an agent that elevates ρ38γ activity, or activity of a variant of ρ38γ, in neurons of the subject. ρ38γ, also known as ERK6, SAPK3 and MAPK12, is a mitogen activated protein kinase (MAP Kinase) . In one embodiment, the ρ38γ is from a mammal. For example, the ρ38γ may be from a human, mouse, dog, cat, pig, cow, rat, non-human primate, goat, sheep. Typically, the ρ38γ is human ρ38γ. Wild type ρ38γ is activated through phosphorylation of tyrosine and threonine residues in the motif TGY. Wild type ρ38γ phosphorylates tau following activation.

Activation of ρ38γ is carried out by the MAP kinase kinases MKK3 and MKK6, which are in turn activated upon

phosphorylation by the MAPK kinase MAP3K.

As described in the Examples, the inventors have found that phosphorylation of tau by ρ38γ results in disruption of NR/PSD-95/tau/FYN complexes in cultured neurons and in a mouse model of Alzheimer's disease;

limits Αβ-induced toxicity in cultured neurons in a mouse model of Alzheimer's disease; and reduces the severity of pentylenetetrazole (PTZ) induced seizures in a mouse model of excitotoxicity and epilepsy. The inventors have shown that by introducing ρ38γ, or a constitutively active variant of ρ38γ, into neurons of mice, NR/PSD-95/tau/FYN complexes in neurons are disrupted and Αβ-induced

excitotoxicity is reduced in a mouse model of Alzheimer's disease, and the severity of pentylenetetrazole (PTZ) induced seizures in a mouse model of excitotoxicity and epilepsy is reduced.

An agent that elevates ρ38γ activity, or the activity of a variant of ρ38γ, in a neuron may be an agent that: (a) elevates the amount of ρ38γ, typically the amount of active ρ38γ, in the neuron; and/or (b) elevates the amount of a variant of ρ38γ, typically the amount of an active variant of ρ38γ, in the neuron; and/or (c) elevates the amount of ρ38γ activation in the neuron; and/or (d) elevates the amount of activation of the variant of ρ38γ in the neuron, if the variant if not an active variant. As used herein, "ρ38γ activity" is an activity of activated p38ythat causes disruption of the tau-dependent signalling complex.

Typically, the activity of activated ρ38γ that causes disruption of the tau-dependent signalling complex is phosphorylation of tau at T205, and optionally

phosphorylation of tau at one or more amino acid residues selected from the group consisting of, for example, S199, S396, and S404. The "activity of a variant of ρ38γ" refers to an activity of a variant of ρ38γ which is the same as, or substantially similar to, ρ38γ activity. The variant of ρ38γ may be capable of ρ38γ activity without activation (for example, an active variant, such as a constitutively active variant), or may exhibit ρ38γ activity following activation. ρ38γ activity is elevated in a neuron when the amount of ρ38γ activity in the neuron after treatment is increased relative to the amount of ρ38γ activity in the neuron prior to treatment. The activity of a variant of ρ38γ is elevated in a neuron when the amount of activity of the variant in the neuron after treatment is increased relative to the amount of activity of the variant in the neuron prior to treatment. The ρ38γ activity, or the activity of a variant of ρ38γ, may be elevated by

administering an agent which elevates: (a) the amount of endogenous ρ38γίη the neurons, such as increasing expression (transcription and/or translation) of endogenous ρ38γ; and/or

(b) the amount of exogenous ρ38γ in the neurons; and/or (c) the amount of a variant of ρ38γ in the neurons;

and/or

(d) the activation of endogenous ρ38γ, exogenous ρ38γ and/or variant of ρ38γ, in the neurons.

In one embodiment, the ρ38γ activity, or the activity of a variant of ρ38γ, is elevated by administering an agent which elevates the amount of exogenous ρ38γ, or a variant thereof, in neurons. The amount of exogenous ρ38γ, or a variant thereof, may be elevated by introducing into neurons ρ38γ, or a variant thereof, or by introducing into neurons a nucleic acid capable of expressing ρ38γ, or a variant thereof.

Thus, in one embodiment, the agent which elevates ρ38γ activity, or the activity of a variant of ρ38γ, in neurons of the subject, may comprise the ρ38γ protein or variant thereof, or a nucleic acid that is capable of expressing ρ38γ, or a variant thereof, in neurons of the subject. The nucleic acid sequence encoding full-length wild-type human ρ38γ, together with the amino acid sequence of full-length wild-type human ρ38γ, used in the Examples described herein is shown in Figure 43. Naturally

occurring isoforms and variants of human ρ38γ are also known (e.g. Genbank accession nos. NP_001290181,

CR456515) . It is envisaged that natural isoforms or variants of ρ38γ that phosphorylate tau at an amino acid residue of tau which causes disruption of the tau- dependent signalling complex could be used in the methods described herein.

In one embodiment, the agent which elevates ρ38γ activity, or the activity of a variant of ρ38γ, comprises a nucleic acid that encodes ρ38γ or a variant thereof. Those skilled in the art will be able to determine the

appropriate nucleic acid sequence which encodes the amino acid sequence of the ρ38γ or variant thereof. For example, a nucleic acid sequence which encodes ρ38γ may comprise a nucleic acid sequence that is in the range of from about 60% to 100% identical to the wild-type coding sequence of human ρ38γ ( SEQ ID NO: 1). For example, the nucleic acid encoding ρ38γ may have a sequence that has at least 60%,

61%, 62%, 63%, 64%, 65%, 66%, 67%, 68%, 69%, 70%, 71%,

72%, 73%, 74%, 75%, 76%, 77%, 78%, 79%, 80%, 81%, 82%,

83%, 84%, 85%, 86%, 87%, 88%, 89%, 90%, 91%, 92%, 93%,

94%, 95%, 96%, 97%, 98%, 99% or 100% sequence identity to the wild-type coding sequence of ρ38γ using one of the alignment programs described herein using standard

parameters. Those skilled in the art will recognize that these values can be appropriately adjusted to determine corresponding identity of proteins encoded by a nucleotide sequence by taking into account codon degeneracy, reading frame positioning, and the like.

In one embodiment, the agent which elevates ρ38γ activity, or the activity of a variant of ρ38γ, comprises a variant of ρ38γ. In one embodiment, the agent which elevates ρ38γ activity, or the activity of a variant of ρ38γ, comprises a nucleic acid that encodes a variant of ρ38γ. As used herein, a variant of ρ38γ is a protein which differs from the wild-type human ρ38γ protein by one or more amino acid substitutions, additions or deletion, and which is capable of phosphorylating an amino acid residue of tau which causes disruption of the tau-dependent signalling complex. Typically, the variant of ρ38γ phosphorylates tau at residue T205, and optionally one or more residues selected from the group consisting of S199, S396, S404. In one embodiment, the variant of ρ38γ comprises an amino acid sequence that is at least 60%, 65%, 70%, 75%, 80%, 85%, 90%, 95%, or 99% identical to the amino acid sequence of wild-type human ρ38γ. In one embodiment, the variant of ρ38γ comprises an amino acid sequence that is at least 60%, 65%, 70%, 75%, 80%, 85%, 90%, 95%, or 99% identical to the amino acid sequence represented by SEQ ID NO: 2.

As used herein, "% identity" with reference to a polypeptide, or "% identical to the amino acid sequence of a polypeptide", refers to the percentage of residues in the two sequences that are the same when aligned for maximum correspondence over a specified comparison window, as measured by sequence comparison algorithms or by visual inspection .

Sequence comparison algorithms for determining % identity between two polypeptides are known in the art. Examples of such algorithms are the algorithm of Myers and Miller (1988) ; the local homology algorithm of Smith et al . (1981); the homology alignment algorithm of Needleman and Wunsch (1970); the search-for-similarity-method of Pearson and Lipman (1988); the algorithm of Karlin and Altschul (1990), modified as in Karlin and Altschul

(1993) . Computer implementations of these algorithms for determining % identity between two polypeptides include, for example: CLUSTAL (available from Intelligenetics , Mountain View, Calif.) (Pearson et al . (1994)) .; the ALIGN program (Version 2.0) and GAP, BESTFIT, BLAST, FASTA, and TFASTA in the Wisconsin Genetics Software Package, Version 8 (available from Genetics Computer Group (GCG) , 575

Science Drive, Madison, Wis., USA) .

In some embodiments, the variant of ρ38γ may

comprise a part of ρ38γ. In one embodiment, the variant of ρ38γ comprises a PDZ interaction motif. PSD-95 comprises a PDZ motif, and ρ38γ is believed to interact with PSD-95, at least in part, through the PDZ interaction motif. The PDZ interaction motif of ρ38γ is a short amino acid sequence in the C-terminal portion of the ρ38γ molecule (see Figure 1A) . Typically, the PDZ interaction motif comprises the amino acid sequence ETPL or ETAL. In various embodiments, the variant of ρ38γ comprises an amino acid sequence selected from the group consisting of: ETPL (SEQ ID NO: 8), KETPL (SEQ ID NO: 9), SKETPL (SEQ ID NO: 10), VSKETPL (SEQ ID NO: 11), RVSKETPL (SEQ ID NO: 12), ARVSKETPL (SEQ ID NO: 13), GARVSKETPL (SEQ ID NO: 14), LGARVSKETPL (SEQ ID NO: 15), QLGARVSKETPL (SEQ ID NO: 16), RQLGARVSKETPL (SEQ ID NO: 17), PRQLGARVSKETPL (SEQ ID NO: 18),

PPRQLGARVSKETPL (SEQ ID NO: 19), KPPRQLGARVSKETPL (SEQ ID NO: 20), FKPPRQLGARVSKETPL (SEQ ID NO: 21),

SFKPPRQLGARVSKETPL (SEQ ID NO: 22), LSFKPPRQLGARVSKETPL (SEQ ID NO: 23), VLSFKPPRQLGARVSKETPL (SEQ ID NO: 24), EVLSFKPPRQLGARVSKETPL (SEQ ID NO: 25),

KEVLSFKPPRQLGARVSKETPL (SEQ ID NO: 26),

YKEVLSFKPPRQLGARVSKETPL (SEQ ID NO: 27),

TYKEVLSFKPPRQLGARVSKETPL (SEQ ID NO: 28),

VTYKEVLSFKPPRQLGARVSKETPL (SEQ ID NO: 29),

RVTYKEVLSFKPPRQLGARVSKETPL (SEQ ID NO: 30),

KRVTYKEVLSFKPPRQLGARVSKETPL (SEQ ID NO: 31), ETAL (SEQ ID NO: 32), KETAL (SEQ ID NO: 33), PKETAL (SEQ ID NO: 34), VPKETAL (SEQ ID NO: 35), RVPKETAL (SEQ ID NO: 36),

ARVPKETAL (SEQ ID NO: 37), GARVPKETAL (SEQ ID NO: 38), LGARVPKETAL (SEQ ID NO: 39), QLGARVPKETAL (SEQ ID NO: 40), RQLGARVPKETAL (SEQ ID NO: 41), PRQLGARVPKETAL (SEQ ID NO: 42), PPRQLGARVPKETAL (SEQ ID NO: 43), KPPRQLGARVPKETAL (SEQ ID NO: 44), FKPPRQLGARVPKETAL (SEQ ID NO: 45),

SFKPPRQLGARVPKETAL (SEQ ID NO: 46), LSFKPPRQLGARVPKETAL (SEQ ID NO: 47), VLSFKPPRQLGARVPKETAL (SEQ ID NO: 48), EVLSFKPPRQLGARVPKETAL (SEQ ID NO: 49),

KEVLSFKPPRQLGARVPKETAL (SEQ ID NO: 50),

YKEVLSFKPPRQLGARVPKETAL (SEQ ID NO: 51),

TYKEVLSFKPPRQLGARVPKETAL (SEQ ID NO: 52),

VTYKEVLSFKPPRQLGARVPKETAL (SEQ ID NO: 53),

RVTYKEVLSFKPPRQLGARVPKETAL (SEQ ID NO: 54), and

KRVTYKEVLSFKPPRQLGARVPKETAL (SEQ ID NO: 55) ..

In some embodiments, the variant of ρ38γ may

comprise a part of ρ38γ but otherwise differ from the wild- type ρ38γ. In this regard, the inventors envisage that variants of ρ38γ may include protein in which the PDZ interaction motif of ρ38γ is fused to the carboxy-terminus of other kinases, such as MAP kinase or other

serine/threonine kinases, or variants of other kinases that carry mutations to modify their activity. For example, the variant of ρ38γ may comprise the PDZ

interaction motif of ρ38γ fused to the carboxy-terminus of a kinase selected from the group consisting of ρ38α, ρ38β and ρ38δ, or variants of ρ38α, ρ38β and ρ38δ that carry mutations that modify their activity.

In one embodiment, the variant of ρ38γ is an active variant of ρ38γ. An active variant of ρ38γ is a variant which does not require activation by the MAP kinase

kinases MKK3 and MKK6 in order to exhibit ρ38γ activity. In one embodiment, the active variant of ρ38γ is a

constitutively active variant of ρ38γ. A constitutively active variant of ρ38γ is a variant of ρ38γ which is continuously active and therefore does not require

activation by the MAP kinase kinases MKK3 and MKK6.

Typically, a constitutively active variant comprises one or more amino acid substitutions which result in

continuous activity. In one embodiment, the

constitutively active variant of ρ38γ comprises the amino acid substitution of D179A. The amino acid sequence of an example of a constitutively active variant of ρ38γ is shown in Figure 44 (SEQ ID NO: 3) . In another embodiment, the constitutive active variant of ρ38γ may comprise the amino acid substitution of F330L/S. The substitution of F330L/S in ρ38γ corresponds to the substitution of the constitutive active variant of p38a F327L/S.

In one embodiment, there is provided a method of treating Alzheimer's disease in a subject, comprising administering a nucleic acid sequence which expresses ρ38γ or a variant thereof, typically a constitutively active variant of ρ38γ, in neurons of the subject.

In one embodiment, there is provided a method of treating stroke in a subject, comprising administering a nucleic acid sequence which expresses ρ38γ or a variant thereof, typically a constitutively active variant of ρ38γ , in neurons of the subject.

In one embodiment, there is provided a method of treating epilepsy in a subject, comprising administering a nucleic acid sequence which expresses ρ38γ or a variant thereof, typically a constitutively active variant of ρ38γ, in neurons of the subject.

In another embodiment, the subject is treated by administering an agent that introduces into neurons of the subject a variant of tau that causes disruption of the tau-dependent signalling complex. As used herein, a

"variant of tau" is a tau protein comprising one or more amino acid substitutions, insertions, or deletions, of the full length wild-type tau, wherein the one or more

deletions is not more than 100 contiguous amino acids, typically not more than 90, 80, 70, 60, 50, 40, 30, 20, or 10 contiguous amino acids. In one embodiment, the variant of tau comprises one or more amino acid substitutions or insertions of the wild-type tau. In one embodiment, the variant of tau comprises one or more amino acid

substitutions of the wild-type tau. In one embodiment, the variant of tau is a phosphomimetic of tau that causes disruption of the tau-dependent signalling complex. As used herein, a phosphomimetic of tau is a variant of tau comprising one or more amino acid substitutions, and which functions in a manner that is the same as, or

substantially the same as, that of unsubstituted tau following phosphorylation of the unsubstituted tau at a particular amino acid. A phosphomimetic comprises a phosphomimetic substitution.

As described in the Examples, the inventors have shown that introduction of a T205E variant of tau into hippocampal neurons lowered Αβ-induced toxicity in the neurons. The T205E variant of Tau is a phosphomimetic of Tau phosphorylated at T205. A phosphomimetic substitution is an amino acid substitution in a protein which results in the protein functioning in a manner which is the same as, or substantially the same as, the unsubstituted protein following phosphorylation of the unsubstituted protein. A phosphomimetic substitution of tau is an amino acid substitution at a site of tau which results in a tau protein that functions in the same, or substantially the same, manner to the wild-type tau following

phosphorylation of the wild-type tau, typically at that site .

In one embodiment, the method comprises treating the subject to introduce a phosphomimetic of tau comprising a phosphomimetic substitution of tau that causes disruption of, or reduces formation of, the tau-dependent signalling complex. In one embodiment, the one or more

phosphomimetic substitutions are at amino acid residues of the tau protein that are phosphorylated by ρ38γ. In one embodiment, the phosphomimetic substitution of tau is threonine to glutamic acid at position 205 of tau (T205E) , with amino acid numbering based on the longest human tau isoform comprising 441 amino acids. The amino acid

sequence of full-length wild-type human tau (SEQ ID NO: 4) and tau T205E (SEQ ID NO: 5) is shown in Figure 45.

Typically, the variant of tau is a variant of human tau. In other embodiments, the variant of tau may be a variant of tau from a non-human mammal. For example, the variant of tau may be a variant of tau from a mouse, dog, cat, pig, cow, rat, non-human primate, goat, sheep.

In one embodiment, there is provided a method of treating Alzheimer's disease in a subject, comprising administering a nucleic acid sequence which expresses tau which differs from wild-type tau in an amino acid

substitution of threonine to glutamic acid at position 205 (T205E), in neurons of the subject. In one embodiment, there is provided a method of treating stroke in a subject, comprising administering a nucleic acid sequence which expresses tau which differs from wild-type tau in an amino acid substitution of threonine to glutamic acid at position 205 (T205E) , in neurons of the subject.

In one embodiment, there is provided a method of treating epilepsy in a subject, comprising administering a nucleic acid sequence which expresses tau which differs from wild-type tau in an amino acid substitution of threonine to glutamic acid at position 205 (T205E) , in neurons of the subject.

In embodiments in which the agent comprises a nucleic acid that is capable of expressing ρ38γ or a variant thereof, or the variant of tau, in neurons of the subject, a nucleic acid sequence encoding ρ38γ or a variant thereof, or the variant of tau, is typically operably linked to regulatory sequence to direct expression of the ρ38γ, or variant thereof, or the variant of tau, in the neurons of the subject. A nucleic acid that is capable of expressing ρ38γ or a variant thereof, or a variant of tau, in neurons of a subject may comprise an expression cassette

comprising the coding sequence of ρ38γ or variant thereof, or the variant of tau. An expression cassette is a nucleic acid sequence comprising coding sequence and regulatory sequence which operate together to express a protein encoded by the coding sequence in a cell. "Coding sequence" refers to a DNA or RNA sequence that codes for a specific amino acid sequence. It may constitute an

"uninterrupted coding sequence", i.e., lacking an intron, such as in a cDNA, or it may include one or more introns bounded by appropriate splice junctions. The expression cassette typically includes

regulatory sequences. A "regulatory sequence" is a nucleotide sequence located upstream (5' non-coding sequences), within, or downstream (3' non-coding

sequences) of a coding sequence, and which influences the transcription, RNA processing or stability, or translation of the associated coding sequence. Regulatory sequences are known in the art and may include, for example, transcriptional regulatory sequences such as promoters, enhancers translation leader sequences, introns, and polyadenylation signal sequences. The coding sequence is typically operably linked to a promoter. A promoter is a DNA region capable under certain conditions of binding RNA polymerase and initiating transcription of a coding sequence usually located downstream (in the 3' direction) from the promoter. The coding sequence may also be

operably linked to termination signals. The expression cassette may also include sequences required for proper translation of the coding sequence. The expression cassette including the coding sequence may be chimeric. A "chimeric" vector or expression cassette, as used herein, means a vector or cassette including nucleic acid

sequences from at least two different species, or has a nucleic acid sequence from the same species that is linked or associated in a manner that does not occur in the

"native" or wild type of the species. The coding sequence in the expression cassette may be under the control of a constitutive promoter or of a regulatable promoter that initiates transcription only in a particular tissue or cell type, or when the host cell is exposed to some particular stimulus. For example, in an expression cassette comprising a nucleic acid encoding ρ38γ, the coding sequence may be operably linked to a promoter which is not native to the ρ38γ gene, such as a promoter that expresses the coding sequence in, or is inducible in, neurons. Examples of suitable neural promoters include synapsin (SYN) , calcium/calmodulin-dependent protein kinase (CaMKII), tubulin alpha I (Tal), neuron-specific enolase (NSE) , platelet derived growth factor beta chain (PDGF) , MfP, dox, GFAP, Preproenkephalin, dopamine β- hydroxylase (άβΗ) , prolactin, chicken beta actin, prion protein, murine Thyl.2, myelin basic promoter, or any of the above combined with an enhancer, such as a partial cytomegaly virus promoter. Examples of other promoters which may be used to express nucleic acid sequence in neurons include, the SV40 early promoter, mouse mammary tumor virus long terminal repeat (LTR) promoter;

adenovirus major late promoter (Ad MLP); a herpes simplex virus (HSV) promoter, a cytomegalovirus (CMV) promoter such as the CMV immediate early promoter region (CMVIE) , a rous sarcoma virus (RSV) promoter, synthetic promoters, hybrid promoters, and the like. Inducible or controllable promoters include, for example, promoters whose

transcriptional activity is modified in the presence or absence of mifepristone, doxycycline, tetracycline or tamoxifen .

A nucleic acid encoding a protein (coding sequence) is operably linked to a regulatory sequence when it is arranged relative to the regulatory sequence to permit expression of the protein in a cell. For instance, a promoter is operatively linked to a coding region if the promoter helps initiate transcription of the coding sequence.

As used herein, "expression" of a nucleic acid sequence refers to the transcription and translation of a nucleic acid sequence comprising a coding sequence to produce the polypeptide encoded by the coding sequence.

In one embodiment, the agent is a vector. In such vectors, the nucleic acid sequence encoding ρ38γ or variant thereof, or the variant of tau, or an expression cassette comprising such sequences, is inserted into an appropriate vector sequence. The term "vector" refers to a nucleic acid sequence suitable for transferring genes into a host cell, such as a neuron. The term "vector" includes plasmids, cosmids, naked DNA, viral vectors, etc. In one embodiment, the vector is a plasmid vector. A plasmid vector is a double stranded circular DNA molecule into which additional sequence may be inserted. The plasmid may be an expression vector. Plasmids and expression vectors are known in the art and described in, for

example, Sambrook et al . Molecular Cloning: A Laboratory Manual, 4 th Ed. Vol. 1-3, Cold Spring Harbor, N.Y. (2012) .

In some embodiments, the vector is a viral vector. Viral vectors comprise viral sequence which permits, depending on the viral vector, viral particle production and/or integration into the host cell genome and/or viral replication. Viral vectors which can be utilized with the methods and compositions described herein include any viral vector which is capable of introducing a nucleic acid into neurons, typically neurons of the brain.

Examples of viral vectors include adenovirus vectors;

lentiviral vectors; adeno-associated viral vectors;

Rabiesvirus vectors; Herpes Simplex viral vectors; SV40; polyoma viral vectors; poxvirus vector.

In one embodiment, the viral vector is an adeno- associated viral (AAV) vector for packaging in an adeno- associated virus. In one embodiment, the AAV vector is a serotype selected from the group consisting of AAVl, AAV2 , AAV3, AAV4 , AAV5 , AAV6 , AAV6.2, AAV7 , AAV8 , AAV9, AAVrhlO, AAVrh20, AAVrh39, AAVrh43, and AAVcy5 vector or variants thereof. In one embodiment, the viral vector is serotype AAV1, AAV9, AAVrhlO or AAVcy5. In one embodiment, the serotype of the AAV vector is AAV1. In another

embodiment, the serotype of the AAV vector is AAV9. In another embodiment, the serotype of the AAV vector is AAVrhlO. In another embodiment, the serotype of the AAV vector is AAVcy5. The use of recombinant AAV for

introducing nucleic acids into cells is known in the art and described in, for example, US20160038613; Grieger and Samulski (2005) Adeno-associated virus as a gene therapy vector: vector development, production and clinical applications, Advances in Biochemical

Engineering/Biotechnology 99: 119-145; Methods for the production of recombinant AAV are known in the art and described in, for example, Harasta et al (2015)

Neuropsychopharmacology 40: 1969-1978. An example of an adeno-associated viral vector capable of expressing ρ38γ in neuronal cells is shown in Figures 46 and 48 (SEQ ID NO: 6) . An example of an adeno-associated viral vector capable of expressing p38y CA in neuronal cells is shown in Figures 47 and 49 (SEQ ID NO: 7) . In one embodiment, the viral vector comprises SEQ ID NO: 6 or 7. In one

embodiment, the viral vector comprises SEQ ID NO: 7.

In another embodiment, the viral vector is a

lentiviral vector. Methods for production and use of lentiviral vectors are known in the art and described in, for example, Naldini et al . (1996) In vivo gene delivery and stable transduction of nondividing cells by a

lentiviral vector, Science, 272:263-267; Lois et al .

(2002) Germline transmission and tissue-specific expression of transgenes delivered by lentiviral vectors, Science, 295 : 868-872; Vogel et al (2004), A single

lentivirus vector mediates doxycycline-regulated

expression of transgenes in the brain. Hum Gene Ther.

2004; 15(2) : 157-165.

Adenoviruses are also contemplated for use in

delivery of nucleic acid agents. Thus, in another

embodiment, the viral vector is an adenoviral vector.

Adenoviral vectors are known in the art and described in, for example, Kozarsky and Wilson, Current Opinion in

Genetics and Development 3:499-503 (1993); Southgate et al . (2008) Gene transfer into neural cells in vitro using adenoviral vectors, Current Protocols in Neuroscience, Unit 4 23, Chapter 4; Akli et al . ( 1993 ) Transfer of a foreign gene into the brain using adenovirus vectors.

Nature genetics, 3(3) : 224-228.

Another aspect provides a vector as described herein, typically a viral vector as described herein.

Viral vectors are typically packaged into viral particles using methods known in the art. The viral particles may then be used to transfer cell lines,

including neural cell lines, or neural tissue, either in vitro or in vivo. Thus, another aspect provides a viral particle comprising a vector described herein.

A further aspect provides an agent as described herein .

The agent described herein may be formulated as a pharmaceutical composition. Accordingly, in another aspect, there is provided a pharmaceutical composition comprising the agent described herein. The composition comprises the agent in a pharmaceutically acceptable carrier. Methods for the formulation of agents with pharmaceutical carriers are known in the art and are described in, for example, Remington's Pharmaceutical Science, (17 th ed. Mack Publishing Company, Easton, Pa. 1985); Goodman & Gillman's: The Pharmacological Basis of Therapeutics (11 th Edition, McGraw-Hill Professional, 2005) .

Acceptable carriers, diluents and adjuvants are nontoxic to recipients and are preferably inert at the dosages and concentrations employed, and include buffers such as phosphate, citrate, or other organic acids;

antioxidants such as ascorbic acid; low molecular weight polypeptides; proteins, such as serum albumin, gelatin, or immunoglobulins; hydrophilic polymers such as

polyvinylpyrrolidone; amino acids such as glycine,

glutamine, asparagine, arginine or lysine;

monosaccharides, disaccharides , and other carbohydrates including glucose, mannose, or dextrins; chelating agents such as EDTA; sugar alcohols such as mannitol or sorbitol; salt-forming counterions such as sodium; and/or nonionic surfactants such as Tween, pluronics or polyethylene glycol (PEG) .

Administration of the agent to subject may be by intracranial, intravenous, intraperitoneal, subcutaneous, intramuscular, intranasal or intrathecal injection.

Compositions suitable for intracranial, intravenous, intraperitoneal, subcutaneous, intramuscular, intranasal or intrathecal use include sterile aqueous solutions or dispersions and sterile powders for the extemporaneous preparation of sterile injectable solutions or

dispersions. The pharmaceutically acceptable carrier can be a solvent or dispersion medium containing, for example, water, ethanol, polyol (for example, glycerol, propylene glycol, liquid polyethylene glycol and the like), suitable mixtures thereof, and vegetable oils. The proper fluidity can be maintained, for example, by the use of a coating such as lecithin, by the maintenance of the required particle size in the case of a dispersion and by the use of surfactants. The prevention of the action of

microorganisms can be brought about by various

antibacterial and antifungal agents, for example,

parabens, chlorobutanol , phenol, sorbic acid, thimerosal and the like. In many cases it will be preferable to include isotonic agents, for example, sugars or sodium chloride.

In embodiments in which the agent is packaged in a viral particle, the pharmaceutical compositions may comprise viral particles in any concentration that allows the agent to be effective. In such embodiments, the pharmaceutical compositions may comprise the virus

particle in an amount of from 0.1% to 99.9% by weight. Pharmaceutically acceptable carriers include water, buffered water, saline solutions such as, for example, normal saline or balanced saline solutions such as Hank' s or Earle's balanced solutions), glycine, hyaluronic acid etc .

Titers of viral particles to be administered will vary depending on, for example, the particular vector to be used, the mode of administration, extent of the condition, the individual, and may be determined by methods standard in the art .

The agent described herein may be formulated for introduction into neuronal cells by non-viral methods such as microinjection, electroporation, microparticle

bombardment, liposome uptake, nanoparticle-based delivery etc .

In one embodiment, the agents described herein may be formulated in one or more liposomes, lipoplexes, or lipid nanoparticles . In one embodiment, the agents described herein are formulated in liposomes. Liposomes are

unilamellar or multilamellar vesicles which have a

membrane formed from a lipophilic material and an aqueous interior. The aqueous portion contains the composition to be delivered. Liposome design may include, for example, opsonins or ligands in order to improve the attachment of liposomes to tissue or to activate events such as, for example, endocytosis.

The formation of liposomes may depend on the

physicochemical characteristics such as the agent and the liposomal ingredients, the nature of the medium in which the lipid vesicles are dispersed, the effective

concentration of the agent, any additional processes involved during the application and/or delivery of the vesicles, the optimization size, polydispersity and the shelf-life of the vesicles for the intended application, and the batch-to-batch reproducibility and possibility of large-scale production of safe and efficient liposomal products.

Methods for the production of liposomes and lipid nanoparticles for delivery of agents are known in the art, and described in, for example, US 5,264,221.

The term "administering" should be understood to mean providing a compound or agent to a subject in need of treatment .

It will be understood that the specific dose level and frequency of dosage for any particular subject may be varied and will depend upon a variety of factors

including, for example, the activity of the specific compound or agent employed, the metabolic stability and length of action of that compound or agent, the age, body weight, general health, sex, diet, mode and time of administration, drug combination, the severity of the particular condition, and the host undergoing therapy.

Also provided is a kit, comprising a container comprising the agent. The container may be simply a bottle comprising the agent in parenteral dosage form, each dosage form comprising a unit dose of the agent. The kit will further comprise printed instructions. The article of manufacture will comprise a label or the like, indicating treatment of a subject according to the present method. In one form, the article of manufacture may be a container comprising the agent in a form for parenteral dosage. For example, the agent may be in the form of an injectable solution in a disposable container.

As used herein, "treating" means affecting a subject, tissue or cell to obtain a desired pharmacological and/or physiological effect and includes inhibiting the

condition, i.e. arresting its development; or relieving or ameliorating the effects of the condition i.e. cause reversal or regression of the effects of the condition.

As used herein, "preventing" means preventing a condition from occurring in a cell or subject that may be at risk of having the condition, but does not necessarily mean that condition will not eventually develop, or that a subject will not eventually develop a condition.

Preventing includes delaying the onset of a condition in a cell or subject.

The inventors envisage that ρ38γ or variants of tau can be used in transgenic animals to assess whether a

neurological disease can be treated with the methods described herein.

Accordingly, a further aspect provides a transgenic non-human animal comprising a transgenic nucleic acid sequence which is capable of expressing in neurons of the transgenic animal ρ38γ or a variant thereof, or a variant of tau that causes disruption of the tau-dependent signalling complex .

In one embodiment, the transgenic nucleic acid sequence is a nucleic acid sequence capable of expressing ρ38γ or a variant thereof. In one embodiment, the

transgenic nucleic acid sequence is capable of expressing an active variant of ρ38γ. In one embodiment, the active variant of ρ38γ is a constitutively active variant of ρ38γ. In one embodiment, the constitutively active variant of ρ38γ is p38y CA .

The regulatory sequences for expressing the transgene in neurons of the animal are described above.

In one embodiment, the transgenic animal is a mouse. However, it will be understood that the transgenic animal may be any animal, including, for example, a rat, cow, sheep, pig or goat.

Another aspect provides a method of assessing whether a neurological condition can be treated or prevented by a method described herein, comprising the steps of:

(a) providing a test animal suffering from the

neurological condition or exhibiting a phenotype which is a model for the neurological condition;

(b) crossing the test animal with a transgenic animal to obtain progeny, the transgenic animal comprising a transgenic nucleic acid sequence which is capable of expressing in neurons of the animal ρ38γ or a variant thereof, or a variant of tau that causes disruption of the tau-dependent signalling complex; and

(c) assessing the severity of the neurological

condition or the phenotype which is a model for the neurological condition in progeny expressing the transgenic nucleic acid sequence.

In the claims which follow and in the preceding description of the invention, except where the context requires otherwise due to express language or necessary implication, the word "comprise" or variations such as "comprises" or "comprising" is used in an inclusive sense, i.e. to specify the presence of the stated features but not to preclude the presence or addition of further features in various embodiments of the invention.

All publications mentioned in this specification are herein incorporated by reference. It will be appreciated by persons skilled in the art that numerous variations and/or modifications may be made to the invention as shown in the specific embodiments without departing from the spirit or scope of the invention as broadly described. The present embodiments are, therefore, to be considered in all respects as illustrative and not restrictive.

In order to exemplify the nature of the present invention such that it may be more clearly understood, the following non-limiting examples are provided.

Examples

Materials and Methods

Mice. APP23 mice expressing human K670N/M671L mutant APP in neurons (C Sturchler-Pierrat et al . , Proc Natl Acad Sci U S A 94, 13287-92 (1997)), Alzl7 mice expressing human non-mutant tau in neurons (A Probst et al . , Acta

Neuropathol 99, 469-81 (2000)), neuron-specific Thyl.2-cre transgenic mice (I Dewachter et al . , J Neurosci 22, 3445- 53 (2002)), tau _/~ (KL Tucker, M Meyer, YA Barde, Nat

Neurosci 4, 29-37 (2001)), p38a loxP/loxP (FB Engel et al . ,

Genes Dev 19, 1175-87 (2005)), p38/T /~ and p38f^ (AR

Pogozelski et al . , PLoS One 4, e7934 (2009)), and 38δ ~ mice (G Sumara et al . , Cell 136, 235-48 (2009)) were previously described. Knockouts for ρ38β, ρ38γ and ρ38δ were global without overt phenotypes, while p38a deletion had to be limited to the CNS due to embryonic mortality of global p38a knockout mice. To obtain p38o Aneu mice, we crossed p38a loxP/loxP with Thyl.2-cre strain. All lines were maintained on a C57B1/6 background. Animal experiments were approved by the Animal Ethics Committee of the

University of New South Wales. Mice were genotyped by polymerase chain reaction using isopropanol-precipitated DNA from tail biopsies as template. Oligonucleotide primers for genotyping targeted alleles and transgenes by PCR are listed in the following Table 1:

Table 1:

Forward primer (5'-3') SEQ Reverse primer (5'-3') SEQ

ID ID

NO: NO::

APP23 GTTCTGCTGCATCTTGGACA 56 GAATTCCGACATGACTCAGG 57

Alz17 GGGTGTCTCCAATGCCTGCTTCTTCAG 58 AAGTCACCCAGCAGGGAGGTGCTCAG 59 p38a\ox TCCTACGAGCGTCGGCAAGGTG 60 AGTCCCCGAGAGTTCCTGCCTC 61 ρ38β AGAAGATGAAGGTGGAGGAGTACAAGC 62 TAACCCGGATGGCTGACTGTTCCATTTA 63 AAG G

Generation of transgenic Thyl.2-p38y mice.

The human ρ38γ coding sequence carrying the D179A mutation and an N-terminal hemagglutinin (HA) -tag was amplified by PCR and inserted into the Xhol site of the plasmid pEX12 (Ittner, et al . Proc. Natl. Acad. Sci . U.S.A. 105, 15997- 16002) (2008)) carrying the mThyl.2 promoter for neuronal expression using Gibson assembly (Gibson, et al . Nat.

Methods 6, 343-345 (2009) (Fig. 30A) . The construct was excised by restriction digest and transgenic founder mice were generated on a congenic C57B1/6 background by

pronuclear injection (Ittner et al . Nat. Protoc. 2, 1206- 1215 (2007)) . Tail DNA from founder mice was screened by PCR for genomic transgene insertion and 2 founder lines (ρ38γ ΟΑ .3 and p38y CA . ) were established by crossing to C57B1/6 mice. Normal fertility, survival and Mendelian transgene transmission was observed for both ρ38γ αΑ .3 and ρ38γ αΑ . lines. Both lines show no overt phenotype.

Immunoblots of cortical, hippocampal, and cerebellar brain extracts from transgenic mice confirmed expression of HA- tagged p38y CA (Fig. 30B) .

Seizures .

Seizures were induced with pentylenetetrazole (PTZ, Sigma- Aldrich) as previously described (LM Ittner et al . , Cell 142, 387-97 (2010)) . Briefly, PTZ was injected i.p. at 30 or 50 mg/kg body weight. Seizures were graded as: 0, no seizures; 1, immobility; 2, tail extension; 3, forelimb clonus; 4, generalized clonus; 5, bouncing seizures; 6, full extension; 7, status epilepticus. Spatial learning/memory testing.

Spatial learning/memory was tested in the Morris Water maze paradigm (CV Vorhees, MT Williams, Nat Protoc 1, 848- 58 (2006)) . Briefly, a custom-built water tank for mouse Morris Water maze (122 cm diameter, 50 cm height) with white non-reflective interior surface in a room with low- light indirect lighting was filled with water (19 - 22 °C) containing diluted non-irritant white dye. Four different distal cues were placed surrounding the tank at

perpendicular positions reflecting 4 quadrants. In the target quadrant, a platform (10cm ) was submerged 1 cm below the water surface. Videos were recorded on CCD camera and analyzed using AnyMaze Software. For spatial acquisition, four trials of each 60 seconds were performed per session. The starting position was randomized along the outer edge of the start quadrant for all trials. To test reference memory, probe trials without platform were performed for a trial duration of 60 seconds, and

recordings were analyzed for time spent within each quadrant. For visually-cued control acquisition (to exclude vision impairments), a marker was affixed on top of the platform and four trials (60 s) per session were performed. All mice were age and gender-matched and tested at 4 months of age. Mice that displayed continuous

floating behavior were excluded. Genotypes were blinded to staff recording trials and analyzing video tracks.

Tracking of swim paths was done using the AnyMaze software (Stolting) . Average swimming speed was determined to exclude motor impairments. Touchscreen operant chambers (Campden Instruments) were used with 2 different paradigms to address spatio-temporal memory and learning (differential paired-associates learning, dPAL) or recognition memory/discrimination learning (pairwise discrimination task, PD) . Previously described touchscreen chamber protocols were used (Horner et al. Nat. Protoc. 8, 1961-1984 (2013)) . Mice in dPAL schedule underwent pre-testing procedures and training as follows: food deprivation (to 85 - 90% of initial body weight) and adaptation to handling (day 0 - 4), adaptation to touchscreen boxes (day 4), collect reward (strawberry milk shake, Nippy's) (day 5 -8), panel-pushing to collect reward training (day 9), initial stimulus-dependent touch training (day 10), must touch stimulus training (day 11 - 16), must initiate trial

training (day 17 - 22), punish incorrect touches (day 23 - 26) . Followed by either dPAL acquisition for 21

consecutive days (day 27 - 49) or pairwise discrimination task acquisition (day 27 - 31) . Maximum time of sessions was set to 60 minutes. Maximum number of trials was set to 36. All training sessions were repeated until mice reached criterion before next training paradigm was started.

Criterion was defined as 36 trials within 60 minutes

(initial touch training, must touch training, must

initiate training) or 27 out of 36 correct trials (punish incorrect touches) . Mice with excessive body weight loss were excluded from the protocol.

Behavior and motor testing

Novelty-induced locomotion and anxiety-related behavior was assessed in the open field test paradigm as previously described (Ke, et al . Acta Neuropathol . 130, 661-678

(2015)) . Briefly, mice were placed individually in 40 x 40 cm 2 boxes in dimly lit sound-insulated enclosures and movements were recorded for 15 minutes. Mice had not been exposed to open field paradigm before. Boxes were wiped with 70 % ethanol between recordings. Movements were tracked using the AnyMaze software (Stolting) . Analysis was either accumulated over entire recording period or split in 1 -minute bins.

Motor performance was tested on a 5-wheel Rota-Rod

treadmill (Ugo Basile) in acceleration mode (5-60rpm) over 120 (aged) or 180 (young) seconds (van Eersel, et al .

Neuropathol. Appl . Neurobiol . 41, 906-925 (2015)) . The longest time each mouse remained on the turning wheel out of 3 attempts per session was recorded. Grip strength was determined as previously described (Ke et al . (2015)) . Briefly, the force required to pull mice off a metal wire was measured using s grip strength meter (Chatillon, AMETEK) . Mice were placed such that they had a double grip on a thin metal wire attached to the meter, and they were pulled away from the meter in a horizontal direction until they let go, and a peak force (N) was recorded at the moment when the mice let go. The highest force from three attempts was recorded.

Calcineurin activity assay

Calcineurin activity in cortical extracts of p38y _/~ and ρ38γ + + littermates was determined by following the manufacturer' s instructions (Abeam) .

Electroencephalography .

Hippocampal EEG recording in freely moving mice was carried out as previously described (AA Ittner, A

Gladbach, J Bertz, LS Suh, LM Ittner, Acta Neuropathol Commun 2 , 149 (2014)) . Briefly, wire EEG electrodes of remote telemetric transmitters (DSI) were implanted in mice anesthetized with ketamine/xylazine . The head was fixed in a stereotactic frame (Kopf instruments) and the bregma was located. Bone openings were drilled using a bone micro-drill (Fine Science Tools, F.S.T.) at positions previously described for the hippocampus (x 2.0, y -2.0, z -2 with reference to bregma) . Electrodes were inserted at this position with reference electrode placed above the cerebellum (x 0, y -6.0, z 0 from bregma) . Electrodes were fixed in place by polyacrylate followed by wound closure and rehydration. Following 10 days of recovery from the surgery, EEGs were recorded with a DSI wireless receiver setup (DSI) with amplifier matrices using the Dataquest A.R.T. recording software at 500 Hz sampling rate (M

Weiergraber, M Henry, J Hescheler, N Smyth, T Schneider, Brain Res Brain Res Protoc 14, 154-64 (2005) ) . Two days after EEG recordings were completed, animals were

transcardially perfused with cold phosphate-buffered saline (PBS) and brains extracted for biochemical and histological analysis. Correct placement of electrodes was confirmed by serial sections of paraffin embedded brain tissue stained with hematoxylin-eosin . Only recordings from mice with correct placement of electrodes were included in further analysis.

Analysis of EEG recordings was performed using the NeuroScore software v3.0 (DSI) with integrated spike detection module, to determine spike train duration, frequency and number of spikes per train were obtained. Recordings were screened manually for movement artefacts and only artefact-free EEG passages were used for

analysis. Raw local field potentials (LFP) were noise filtered using a powerline noise filter (Neuroscore, DSI) . Spectral analysis (i.e. analysis of signal power at individual frequencies expressed as square of the fast Fourier transform (FFT) magnitude) of intra-ictal

sequences was performed using the integrated FFT spectral analysis function of NeuroScore. Frequency bands of theta and gamma wave forms were defined between 4-12 Hz and 25- 100 Hz, respectively. Gamma and theta spectral

contributions were quantified by area-under-curve (AUC) analysis across the defined frequency band in 8 artefact- and hypersynchronous spike-free sequences per recording

(each 1 min in length) . Cross-frequency coupling of theta phase and gamma amplitude was performed using

MATLAB as previously described (AB Tort, R Komorowski, H Eichenbaum, N Kopell, J Neurophysiol 104, 1195-210

(2010)) . Briefly, for cross frequency coupling analysis, raw LFP was noise filtered using a powerline noise filter (Neuroscore, DSI) . Noise-filtered LFP was filtered at two frequency ranges of interest for gamma if A ) and theta (f p ) . The phase time series for theta { f P (t)) and the amplitude envelope time series for gamma {A A (t)) were obtained by

Hilbert transformation of the filtered LFPs. The combined series [ f P (t), A /A (t)] was then generated. After phase binning, the means A A (j) of A A for each bin j were

v iff)

calculated and normalized using the sum ^ ' - 1 ' of A /A ( ) over N bins to generate phase-amplitude distribution P(j). The modulation index is based on calculating the Kullback- Leibler distance D KL between the non-uniform (i.e. coupled) phase-amplitude distribution P(j) . The modulation index is based on calculating the Kullback-Leibler distance D KL between the non-uniform (i.e. coupled) phase-amplitude distribution P(j) over all phase bins and the uniform (i.e. uncoupled) distribution U(j).

The modulation index MI is defined as

log(;V)

Phase-amplitude distributions and modulation indices were determined from artefact- and hypersynchronous spike-free 8 sequences (each 1 min) per recording.

Synaptosome and post-synapt ic density preparation

Purification of synaptosomes from cortical tissue was performed as previously described (Ittner, et al . Cell

142, 387-397 (2010)) . Briefly, cortical tissue was weighed and homogenized in ice-cold sucrose buffer (0.32M sucrose, ImM NaHC03, ImM MgCl2, 0.5mM CaCl2, protease inhibitors (EDTA-free, Roche)) at 30 mg tissue/ml using a pre-cooled dounce homogenizer. After clearing the homogenate by centrifugat ion (l,400g, 10 minutes, 4°C), pellets were resuspended in sucrose buffer and centrifuged again

(l,400g, 10 minutes, 4°C) . Combined supernatants were centrifuged again and supernatant (total brain homogenate) was spun at 13,800g for 10 minutes at 4°C. Pellet was resuspended in sucrose buffer and layered on top of 5%

Ficoll (Sigma) and centrifuged at 45,000g for 45 minutes at 4°C. Pellet was resuspended in 5% Ficoll and layered on top of 13% Ficoll and centrifuged at 45,000g for 45 minutes at 4°C. The interface (synaptosomes) was

collected, diluted in 5% Ficoll and centrifuged at 45,000g for 30 minutes at 4°C. Supernatant (non-synaptic) was collected and pellet was resuspended in pH8 buffer (20mM Tris pH8, 1% Triton- X100, lOOmM NaCl, ImM EGTA, ImM EDTA, 0.5% sodium deoxycholate, 0.1% sodium dodecyl sulfate (SDS) , protease inhibitors (EDTA-free, Roche) ) . After centrifugation at 40,000g for 30 minutes at 4°C, pellets (post-synaptic densities) were resuspended in 5% SDS. The supernatants constituted synaptic non-PSD associated proteins. Protein concentrations for different fractions was determined before preparing samples for Western blotting .

Plasmids .

Plasmids for expression of rat PSD95 (kind gift from Wei- dong Yao; Addgene plasmid #15463), Fyn kinase (kind gift from Filippo Giancotti; Addgene plasmid #16032) and NR2B (kind gift from Robert Malinow; Addgene plasmid #23998), and were obtained from the Addgene depository. For live cell fluorescence confocal imaging, PSD-95 was internally tagged with mCherry between PDZ domains 2 and 3 by

megaprime PCR (Bryksinet al . Biotechniques 48, 463-465 (2010)) and tau variants were tagged with eGFP by cloning into peGFP-Cl (Clontech) .

Coding sequences for human p38 , human ρ38β and human ρ38γ were cloned into pcDNA3.1 with an N-terminal HA-tag.

Coding sequence for human ρ38δ was cloned in peGFP-Cl .

Mutations in p38 coding sequences for generation of active variants (M Avitzour et al . , FEES J " 274, 963-75 (2007)) and variants of ρ38γ lacking the PDZ motif (APDZm) were generated using the Q5 site-directed mutagenesis kit

(NEB) . Coding sequence for human tau (441 amino acids) was cloned into pcDNA3.2 /V5-DEST (Invitrogen) .

Phosphorylation-site mutants of tau were generated using the Q5 site-directed mutagenesis kit (NEB) .

Oligonucleootide primers for molecular cloning are listed in Table 2. Table 2

Forward primer (5'- SEQ Reverse primer (5'- SEQ

ID ID

3' ) 3' )

NO : NO: tauS46D CCTGAAAGAAgatCCCCTG 76 CCAGCGTCCGTGTCACCC 77

CAGACCCCC

tauT50E TCCCCTGCAGgaaCCCACTGAGG 78 GATTCTTTCAGGCCAGCG 79 tauT52E GCAGACCCCCgaaGAGGACGGAT 80 AGGGGAGATTCTTTCAGG 81

C

tauT69E TGCTAAGAGCgaaCCAACAGCGG 82 TCAGAGGTTTCAGAGCCC 83 tauT71E GAGCACTCCAgaaGCGGAAGATG 84 TTAGCATCAGAGGTTTCAG 85 tauTlll CATTGGAGACgaaCCCAGCCTGG 86 CCTGCTTCTTCAGCTGTG 87 E

tauT153 GAAGATCGCCgaaCCGCGGGGAG 88 GTTTTACCATCAGCCCCC 89 E

tauT181E CGCTCCAAAGgaaCCACCCAGCT 90 GGCGGGGTTTTTGCTGGA 91

C

TauS199 CGGCTACAGCGCCCCCGGCT 92 CTGCGATCCCCTGATTTTGGA 93

A CCC G

TauS199 CGGCTACAGCGACCCCGGCT 94 CTGCGATCCCCTGATTTTGGA 95

D CCC G tauS202A CAGCCCCGGCgccCCAGGCACTC 96 CTGTAGCCGCTGCGATCCCCTG 97 tauS202D CAGCCCCGGCgacCCAGGCACTC 98 CTGTAGCCGCTGCGATCC 99 tauS208D CACTCCCGGCgacCGCTCCCGCA 100 CCTGGGGAGCCGGGGCTG 101

C

tauT212E CCGCTCCCGCgaaCCGTCCCTTC 102 CTGCCGGGAGTGCCTGGG 103

CAAC

tauS235D TCCACCCAAGgacCCGTCTTCCG 104 GTACGGACCACTGCCACC 105

C TauS404 TGGGGACACGGCTCCACGGC 106 GACACCACTGGCGACTTGTAC 107 A ATC ACG

TauS404 TGGGGACACGGATCCACGGC 108 GACACCACTGGCGACTTG 109 D ATC

TauT205 CTCCCCAGGCGCTCCCGGCA 110 CCGGGGCTGCTGTAGCCGC 111 A GCC

TauT205 CTCCCCAGGCGAACCCGGCA 112 CCGGGGCTGCTGTAGCCG 113 E GCCG

TauS199 CCCAGGCGCTCCCGGCAGCC 114 GAGCCGGGGGCGCTGTAGCCG 115 AT205A GCTCCCGC CTGCGATCCCC

TauS199 CCCAGGCGAACCCGGCAGCC 116 GAGCCGGGGTCGCTGTAGCCG 117 DT205E GCTCCCGC CTGCGATCCCC

TauS396 CGTGTACAAGGCGCCAGTGG 118 ATCTCCGCCCCGTGGTCTG 119 A TGT

TauS396 CGTGTACAAGGACCCAGTGG 120 ATCTCCGCCCCGTGGTCT 121 D TGTCTGGGG

TauS396 TGGGGACACGGCTCCACGGC 122 GACACCACTGGCGCCTTGTAC 123 AS404A ATCTCAGCAAT ACGATCTCCGC

TauS396 TGGGGACACGGACCCACGGC 124 GACACCACTGGGTCCTTGTAC 125 DS404D ATCTCAGCAAT ACGATCTCCGC

tauS422D CATGGTAGACgatCCCCAGCTCG 126 TCGATGCTGCCGGTGGAG 127

CCAC

tauS199A CCCAGGCGCTCCCGGCAGCCGCT 128 GAGCCGGGGGCGCTGTAGCCGCTG 129 T205A CCCGC CGATCCCC

tauS199D CCCAGGCGAACCCGGCAGCCGCT 130 GAGCCGGGGTCGCTGTAGCCGCTG 131 T205E CCCGC CGATCCCC

tauS396A TGGGGACACGGCTCCACGGCATC 132 GACACCACTGGCGCCTTGTACACG 133 S404A TCAGCAAT ATCTCCGC

tauS396D TGGGGACACGGACCCACGGCATC 134 GACACCACTGGGTCCTTGTACACG 135 S404D TCAGCAAT ATCTCCGC

mCherry CAAGCCCAGCΑΑΪGCCTAGCTGA 136 137

GTGACGTGAGCAAGGGCGAGGAG A ' AGCTC 11 GTA.CAGC 1 CGTCCAT PSD-95

Adeno-associated virus vectors (von Jonquieres, et al . PLOS ONE 8, e65646 (2013)) for neuronal expression (pAM- CAG) of wildtype (Figure 46, Figure 48 (SEQ ID NO: 6)) and constitutively active (D179A) p38y (Figure 47, Figure 49 (SEQ ID NO: 7) ) or variants of tau were cloned by

conventional restriction enzyme cloning. All plasmids were amplified in E. coli DH5 or XL-lblue. AAV vectors were propagated in E. coli Stbl3 to avoid recombination events. Constructs were verified by sequencing.

Adeno-associated viruses.

Packaging of rAAVl vectors was performed as described (AE Harasta et al . , Neuropsychopharmacology 40, 1969-78

(2015)) . Titres were determined by Quantitative polymerase chain reaction (qPCR) . One μΐ (lxlO 9 viral particles) of either AAV-SGl-shR or AAV-ctr-shR vector was injected at 3 sites each bilaterally into the brains of

cryoanaesthetized neonatal mice as described (G von

Jonquieres et al . , PLoS One 8, e65646 (2013)) .

Cell culture .

Primary hippocampal neurons from E16.5 mouse embryos were cultured, using our standard protocol (T Fath, YD Ke, P Gunning, J Gotz, LM Ittner, Nat Protoc 4, 78-85 (2009)) . Cytotoxicity was determined by measuring LDH release, using a commercial assay (Promega) , or by visualization of EthDl (Thermo Fisher Scientific) added to the cell culture medium 5 min before fixation with 4% PFA/PBS . 293T cells were cultured in DMEM/10%FBS/l%Glutamate/l%P/S (Life

Technologies) and transfected by calcium precipitation (A Ittner et al . , J Exp Med 209, 2229-46 (2012)) . Primary neurons were transduced by AAV infection (AE Harasta et al . , Neuropsychopharmacology 40, 1969-78 (2015)) . Live cell confocal imaging and FLIM/FRET analysis

FLIM/FRET measurements were performed using a time

resolved, inverted confocal fluorescence microscope

(Microtime200 , PicoQuant GmbH) . Excitation of the donor GFP was via a single-photon fiber coupled pico-second- pulsed diode 473 nm laser (20 MHz repetition rate, 2 ms dwell time, 256 x 256 pixel array) using a 63x water objective (1.25 NA) . Fluorescence emission was collected through a 510/32 Semrock BrightLine band pass emission filter onto a single-photon avalanche diode (SPAD) coupled to high speed timing electronics for time-correlated single-photon counting (TCSPC) .

Fluorescence images were analysed by phasor plot using the SimFCS software (Globals Software, USA) . Briefly, Fourier transformation of the decay curve at each pixel was performed and the resulting transforms were plotted as a 2D histogram. The phasor position for the donor only was determined by measuring the donor in the absence of the acceptor. The FRET samples were measured and the phasor position along the quenching trajectory is calculated according to classical FRET efficiency calculation:

where E is FRET efficiency, tD is the fluorescence lifetime of the Donor in absence of acceptor, and tDA is the

fluorescence lifetime in the presence of acceptor.

Cell immunofluorescence staining and microscopy

Cell staining was done as previously described ( (LM Ittner et al. f Cell 142, 387-97 (2010))) . Briefly, cells were fixed with 4% PFA for 10 min, washed with phosphate buffered saline (PBS), permeabilised with 0.02% NP-40 and blocked with blocking buffer (3% horse serum/1% bovine albumin in PBS) . Primary antibodies diluted in blocking buffer were incubated over-night at 4°C or for 1 hour at room temperature. After washing with PBS, secondary antibodies diluted in blocking buffer with or without addition of DAPI to visualize cell nuclei were incubated for 1 hour at room temperature. Cells were then washed and mounted using anti-fade mounting medium (Prolong Gold, Life Technologies) . Secondary antibodies used were coupled to Alexa 488, 555, 568 or 647 dyes (Molecular Probes) . Confocal images were acquired on a Zeiss LSM780 confocal microscope with a Plan-Apochromatic lOOx 1.4 NA objective or on a Zeiss LSM880 Airyscan confocal microscope with a Plan-Apochromatic lOOx 1.4 NA objective using the Zen software (Zeiss) . Epifluorescence imaging was done on a BX51 bright field/epifluorescence microscope (UPlanFL N lenses [¥/0.17/FN26.5] : lOx/0.3, 20x/0.5, 40x/0.75,

60x/1.25oil and 100x/1.3oil) equipped with a DP70 color camera (Olympus) using CellSens software (Olympus) .

Human brain samples

Human entorhinal cortex tissue samples were received from the New South Wales Brain Tissue Resource Centre at the University of Sydney and the Sydney Brain Bank at

Neuroscience Research Australia, which are supported by The University of New South Wales, Neuroscience Research Australia and Schizophrenia Research Institute. Frozen tissue was lysed in phosphate buffered saline (20%w/v) using a rotating dounce homogeniser followed by five Is sonication bursts at 20% power (Vibra Cell, Sonics) .

Lysates were centrifuged at 3,000xg for 10 minutes at 4 2 C and supernatants were used for analysis. Details on patients are provided in Table 3. Use of human brain samples was approved by the Human Research Ethics

Committees of the University of New South Wales and

University of Sydney.

Table 3

Group Age y) gender PMI C'SJD APOE genotype Break

0 93 F 21 cardiac failure E3/E3 0

0 S5 F 23 respiratory failure E3¾3 0

0 79 M S respiretoiv failure E2/E3 0

0 S9 F 23 metastatic ndenocatciaoni3 E3/E4 0

0 S6.5±3.0 IS.S±3.fS

ΓΠ 78 F 11 respiratory failure E3 E3 I

Ι/Π SO M 12 respiratory failure E3'E3 I

I/II 103 M 20 cardiorespiratory fail-ire E3 E3 Π vu 101 F 9 cardiorespiratory failure E3/E3 11

I/II S8 F 31 cardiorespiratory failure E3/E3 11

I II 9<W±5.2 M.6±4.1

III IY 93 F 7 cardiorespiratory failure Ε2;Έ3 III ffi IV 102 F 5 acute renal failure E2 E3 IV in iv 92 F iirieciioi": E3/E3 IV

ΙΠ/TV 7 F 3 cardiac failure E3/E4 IV

1H/IV W).&±5.4

V VI 9S F 1 1 stroke E3 E3 VI

Wl S5 1 cardiac £«iure E3/E3 VI

Y/Vi 100 4 pn umoni E3/E4 VI TVI 100 F aspii afioi■ pneunicnta E3/E3 VI

V/YI 1 p S cardiorespiiaiory failure E3'E3 VI

V/VI 6.8±1.6

PMI, post mortem interval; CoD. cause of death; bold values, mean ± SEM of group

Histological Sections and staining

Mice were transcardially perfused with phosphate-buffered saline followed by 4% paraformaldehyde (PFA) and post- fixing in 4% PFA overnight. Tissue was processed in an Excelsior tissue processor (Thermo) for paraffin

embedding. Thioflavin S staining to visualize amyloid plaques were performed following a standard protocol (LM Ittner et al . , Cell 142, 387-97 (2010)). Muscle cross- sections were stained with primary antibodies to laminin (Sigma) as previously described (Ke, et al . Acta

Neuropathol . 130, 661-678 (2015) ) . Brain sections from AAV-injected mice were stained with primary antibody to tau (Taul3; Abeam) or HA-tag (HA-7; Sigma-Aldrich) to visualize viral transgene expression. Serial paraffin sections of human entorhinal cortex samples were obtained from the NSW Brain Bank and stained with a standard Nissl protocol for counting. Neuronal counting was done on an Olympus BX51 microscope equipped with agraticulated ocular (U100H6; Olympus) . Neurons with the nucleolus, nucleus and cytoplasm visible within a single plane of the section were considered for counting. Forthe CA fields (CA4-1), three random and non-overlapping fields of view were selected. For the entorhinal cortex, three non- overlapping strips of cortex extending from the pial surface and into the grey-white matter junction were marked for counting. Subsequent cortical counts were then performed across three adjacent graticule fields spanning perpendicularly to the pial surface. Mean cell counts across the section were then normalised into cell density values of neurons per mm2. All tissue sections were imaged on a BX51 bright field/epifluorescence microscope (UPlanFL N lenses [ ¥/ 0.17 /FN26.5]: lOx/0.3, 20x/0.5, 40x/0.75, 60x/1.25oil and 100x/1.3oil) equipped with a DP70 color camera (Olympus) . Western Blotting

Western blotting was performed as previously described (A Ittner et al . , J Exp Med 209, 2229-46 (2012)) . Bands were visualized by chemiluminescence on X-ray films or ChemiDoc MP (Biorad) . Densitometric quantification of Western blot results was performed using ImageJ 2.0. O-rc-49/1.51d

(NIH) . Antibodies used in this study were: anti-NRl

(Chemicon) , anti-NR2B (Santa Cruz),

antiphosphoTyrosinel473-NR2B (Affinity BioReagents ) , anti- PSD95 (Millipore) , anti-Fyn (Santa Cruz), anti-phospho- Y418 Fyn ( Invitrogen) , anti-phospho-Y529 Fyn ( Invitrogen) , anti-APP (22C11), anti-Αβ (6E10), anti-tau (DAKO) , anti- tau (tau-1, Millipore), anti-tau (Taul3, Abeam), anti- phospho-Serinel 99 tau (Abeam), anti-phospho-Serine202 tau (Abeam), anti-phospho-Threonine205 tau (Abeam),

antiphospho-Threonine212 tau (Abeam), anti-phospho- Serine214 tau (Millipore), anti-phospho-Threonine231 tau (Abeam) , anti-phospho-Serine235 tau (Abeam) , anti-phospho- Serine356 tau (Abeam), anti-phospho-Serine396 tau (Abeam), anti-phospho-Serine404 tau (Millipore), anti-phospho- Serine422 tau (Millipore), PHF-1 (phospho-Serine396- phospho-Serine404 tau; kind gift by P. Davies) , anti- p38alpha (Cell Signaling), anti-p38beta (Santa Cruz), anti-p38gamma (R&D) , anti-p38delta (R&D) , anti- phosphoThreoninel80/Tyrosinel82-p38 (Cell Signaling

Technologies), anti-Flag (M2, Sigma), anti-HA7 (Sigma), anti-V5 (Invitrogen), anti-MAP2 (mouse Abeam), anti-MAP2 (chicken: Abeam), anti~ 3 tubulin (Covance) , anti-NeuN (Abeam), anti-Debrin (Sigma), anti-Synaptophysin (Abeam), anti-asynuclein (Sigma), anti-glyceraldehyde dehydrogenase (anti-GAPDH, Millipore) .

Immunoprecipitation .

Immunoprecpitation was performed from cell or tissue lysates as previously described (LM Ittner et al . , Cell 142, 387-97 (2010)) . Briefly, cells were lysed in pTNN buffer (20mM Tris pH7.4, 150mM NaCl, ImM EDTA, ImM Na3V04, ImM NaF, ImM glycerophosphate, 2.5mM Na 2 H 2 P 2 0 7 , ImM PMSF, protease inhibitors (Complete, Roche), 1% NP-40 substitute ( Sigma-Aldrich) ) on ice. Lysates were cleared by

centrifugation (16,000xg/10 min/4°C) . Protein concentration was determined (DC Protein Assay, BioRad) and 200 g of lysate incubated with antibody (1:400) for 3 h on a rotator at 4°C. Equilibrated and blocked protein G- beads (Life Technologies) were incubated with lysates for 45 min on a rotator at 4°C. Beads were then washed 3 times and incubated in sample buffer for 5 min at 95°C before SDS-PAGE. Cortical or hippocampal tissues were homogenized in RIPA buffer (20mM Tris pH8.0, 150mM NaCl, ImM EDTA, ImM Na3V04, ImM NaF, ImM glycerophosphate, 2.5 mM sodium pyrophosphate, ImM PMSF, protease inhibitors (Complete, Roche), 1% NP-40 substitute (Sigma-Aldrich) , SDS, sodium deoxycholate ) and subjected to immunoprecipitation as outlined above. Quantitative densitometric analysis was performed using Image J2.0. O-rc-49/1.51d (NIH) and levels for immunoprecipiations of PSD-95/tau/Fyn complexes were expressed relative to immunopreciptateed PSD-95 protein levels .

Microscale thermophoresis (MST)

Tau variants were purified as GST-fusion proteins from E. coli BL21DE3pLys (Promega) using glutathione resin (GE Healthcare) followed by concentration and buffer exchange using ultrafiltration spin columns (10,000 molecular weight cut-off; Vivaspin, Sartorius) . eGFP-PSD-95 was expressed in 293T cells and lysates were prepared in TNN buffer (20 mM Tris pH7.4, 150 mM sodium chloride, 1% NP40 substitute, sodium orthovanadate, sodium pyrophosphate, glycerophosphate, sodium fluoride, protease inhibitors (Complete; Roche)) 48 h after transfection . Concentrations of fusion proteins were determined by absorbance

measurements (Nanodrop 2000C; Thermo-Fisher) using molar extinction coefficients. Thermophoresis of GFP-PSD-95 was measured on a Monolith NT115 (Nanotemper technologies) using 50 % LED power and 20 % MST power with 5 s pre-MST and 30 s MST-time with serials dilutions (1:1) of GST-tau (starting concentration 9 uM) . Thermophoresis and

temperature- jump normalized fluorescence curves from three independent experiments were expressed as fraction of the bound state of the fluorophores-tagged protein (Wienken et al . Nat. Commun. 1, 100 (2010) ) . Thermophoresis was plotted as a function of tau concentration and non-linear curves fitting to determine experimental equilibrium dissociation constants (KD) was performed using sum-of- squares minimization (Marquardt method; Graphpad Prism 6) .

Kinase assay.

Recombinant proteins were expressed in bacteria and purified as previously described (A Ittner et al . , J Exp Med 209, 2229-46 (2012) ) . Purity of proteins was assessed by SDS-PAGE and Coomassie staining. Kinase assay reactions were performed as previously described (A Ittner et al . , J Exp Med 209, 2229-46 (2012)) . Briefly, 0.5iq recombinant ρ38γ was mixed with l\iq of recombinant human tau in kinase reaction buffer (Promega) and incubated for 30 min at 30°C. Kinase reactions were stopped by addition of sample buffer and incubation for 5 min at 95°C. Mass spectrometry

Phospho-peptide mapping of tau after in vitro ρ38γ kinase reactions was done as previously described (Dolai, et al . Cancer Res. 76, 2766-2777 (2016), Thingholm, et al . Nat. Protoc. 1, 1929-1935 (2006)) . Briefly, kinase treated protein extracts containing tau were reduced with 3mM tris (2-carboxyethyl) phosphine (TCEP, 56oC, lOmin),

alkylated with 6 mM iodoacetamide (ambient temp, 30min) , buffer exchanged and concentrated using lOOmM ammonium bicarbonate and 3kDa spin-filters (Amicon Ultra-4

centrifugal filters, Merck KGaA, Darmstadt, Germany) followed by trypsin digest (25:1 w/w protein : trypsin ratio, 16h, 37oC) . A portion of the material was enriched for phosphopept ides using Titansphere Phos-TiO kit, with Ti02 Spin tips (GL Sciences, Tokyo, Japan), following the manufacturer's protocol. Phosphopept ide enriched and non- enriched samples were analysed by LC-MS/MS using Orbitrap mass spectrometers (LTQ-Orbitrap Velos with CID and ETD activation modes and HCD on the QExactive Plus: Thermo

Electron, Bremen, Germany) to maximize identification of phosphopept ides . Chromatography was carried out by nano-LC (Dionex UltiMate 3000 HPLC, Thermo Scientific, Waltham, USA) with autosampler system (Dionex, Amsterdam,

Netherlands) . Peptides {l-l\i injected) were initially captured on a C18 cartridge (Acclaim PepMap 100, 5μπι 100 A, Thermo Scientific Dionex, Waltham, USA) , switching to a capillary column (10cm) containing C18 reverse phase packing (Reprosil-Pur , 1.9 μπι, 200 A, Dr. Maisch GmbH, Ammerbuch-Entringen, Germany), supported within a column heater (45°C, Sonation GmbH, Germany) . Peptides were eluted using a 40min gradient of buffer A (H 2 0:CH 3 CN of 98:2 containing 0.1% formic acid) to 45% buffer B (H 2 0:CH 3 CN of 20:80 containing 0.1% formic acid) at 200nL/min, with high voltage applied at the column inlet. Mass spectrometer settings were: electrospray voltage 2000V, capillary temperature 275 - 300°C, positive ion mode, data dependent acquisition mode with a survey scan acquired {m/z 375- 1750) and up to ten multiply charged ions (charge state ≥ 2+) isolated for MS/MS fragmentation (counts > 2500 for

CID, >5000 for ETD and intensity threshold of 8.0x104 for HCD) . Nitrogen was used as HCD collision gas and

fluoranthene anion reagent for ETD. Peak lists were generated from the raw data using MASCOT Distiller (Matrix Science, London, England) and searched using the MASCOT search engine (version 2.5, Matrix Science) and the NCBInr database (downloaded 24-10-15) using homo sapiens

taxonomy. Search parameters were: peptide tolerance of ± 4ppm and MS/MS tolerances of ± 0.4 Da for CID and ETD or ± 0.05 Da for HCD, variable modifications were

carbamidomethyl cys, met oxidation, phospho (ST) and phospho (Y) , peptide charge of 2+, 3+, and 4+, enzyme specificity trypsin with up to three missed cleavages allowed .

Αβ preparation .

Αβ42 (Bachem) was prepared and pre-aggregated at a

concentration of 100 μΜ as described (MP Lambert et al . , Proc Natl Acad Sci U S A 95, 6448-53 (1998)) . Briefly, hexafluoro-2-propanol (Sigma) dissolved and evaporated Αβ was reconstituted in dimethyl sulfoxide (Sigma) at 5mM and then diluted in phenol-red free F-12 medium (Invitrogen) to a final concentration of ΙΟΟμΜ, followed by brief vortexing and incubation at 4 2 C for 24 hours. Further dilutions were done in culture medium.

Αβ levels and pathology.

Αβ40 and Αβ42 and levels were determined by ELISA as previously described (LM Ittner et al . , Cell 142, 387-97 (2010)) . Plaque load was determined as previously

described (LM Ittner et al . , Cell 142, 387-97 (2010)) . Statistical analysis .

Statistical analysis was performed using Graphpad Prizm

Version 6.0 (Student's t test or ANOVA) . Linear regression and correlation analysis was done by sum of-squares minimization. Survival data were analyzed by log-rank Mantel-Cox testing. All values are presented as mean ± standard error of the mean (SEM) . Results

To understand the molecular contributions of p38 kinases to AD, we first challenged mice with individual deletion of p38a, ρ38β, ρ38γ or ρ38δ (Fig. 1) by inducing

excitotoxic seizures with pentylenetetrazole (PTZ), an approach that has been instrumental in understanding excitotoxicity in AD mouse models (7, 8) . The results are shown in Figs. 2A, 6A, 6B and 6C. Surprisingly, neither neuronal deletion of p38a {p38a neu) , nor knockout of ρ38β or ρ38δ changed seizure latency and severity after PTZ administration, suggesting they have no modulatory role in acute excitotoxicity . In contrast, ρ38γ depletion (ρ38γ ~/~ ) markedly enhanced sensitivity to PTZ-induced seizures (Fig. 2A and fig. 6A, B and C) . Pan-p38 inhibition

increased severity and reduced latency of PTZ-induced seizures in wild-type mice similar to changes in ρ38γ ~/~ , suggesting ρ38γ but not ρ38 /β/δ contribute to acute excitotoxicity. Consistent with a role in post-synaptic signaling, only ρ38γ localized to dendritic spines and post-synaptic densities of cultured neurons (Fig. 2B) . p38 and ρ38β were found in soma and dendrite shafts, while ρ38δ was not detectable in neurons. Taken together, only ρ38γ localizes to the post-synaptic compartment and limits PTZ-induced excitotoxicity.

To test whether the effects of ρ38γ depletion on PTZ- induced seizures would also impact on Αβ-induced deficits in AD mouse models, we crossed ρ38γ ~/~ mice with mutant APP expressing APP23 mice. These APP23. p38γ _ ~ mice were assessed for seizure sensitivity by administering PTZ. The results are shown in Fig. 7. The increased

sensitivity of APP23 mice to PTZ-induced seizures was further augmented in APP23.ρ38γ ~/~ mice (fig. 7A-C) . APP23 mice are characterized by premature mortality, memory deficits, neuronal circuit aberrations with epileptiform brain activity, and Αβ plaque pathology (Ittner et al . , Cell 142, 387-397 (2010); ) Ittner, et al . , Acta

Neuropathol. Commun. 2, 149 (2014); Sturchler-Pierrat et al., Proc. Natl. Acad. Sci. U.S.A. 94, 13287-13292

(1997)) . While Αβ formation and plaque pathology were comparable in brains of APP23 ,ρ38γ ~/~ and APP23.p38y + + mice, deletion of ρ38γ aggravated the premature mortality of APP23 mice, and 82% of APP23.ρ38γ ~/~ mice died by 8 months of age (Fig. 2C) . ρ38γ ~/~ mice showed normal survival (Fig. 2C) . Memory deficits in APP23 ,ρ38γ ~/~ were significantly more severe compared to those of

APP23 ,ρ38γ +/+ mice, as assessed in the Morris-water maze paradigm (Fig. 2D-F, fig. 8A-C) , in differential paired associate learning (dPAL) (Fig. 21) and in a pairwise discrimination task (Fig. 22) . In contrast, ρ38γ ~/~ mice showed wild-type-like memory performance and motor function. Memory deficits were associated with neuronal circuit aberrations and hypersynchronous epileptiform brain activity in APP transgenic lines (10), including

APP23 (14) . Electroencephalography (EEG) of APP23.ρ38γ ~/~ showed more frequent spontaneous seizure spike trains and interictal hypersynchronous discharges than APP23 ,ρ38γ +/+ recordings (Fig. 2G-I) . As can be seen from Fig. 2G, virtually no spike activity was found in ρ38γ ~/~ and ρ38γ +/ mice. Theta (4-8Hz) and gamma (25-100Hz) oscillations, both critical measures of hippocampal network activity related to learning and memory {18, 19), are altered in APP transgenic mice {14) . Accordingly, theta spectral power was shifted to lower frequencies (4-8Hz) in

APP23.ρ38γ +/+ and more so APP23.ρ38γ mice, while gamma spectral power was increased compared to ρ38γ ~/~ and p38y + + mice (fig. 9A-G) . Hippocampal cross frequency coupling (CFC) through theta-phase modulation of gamma power (18) correlates with memory performance in rodents and humans {20, 21), and is impaired in APP23 mice {14) . Interictal EEG traces showed CFC of similar magnitude at ~8Hz in ρ38γ +/+ and p38y ~ mice, but a marked impairment in

APP23.p38y +/+ and virtual depletion 1 in APP23.ρ38γ ~/~ littermates (Fig. 2J) , suggesting ρ38γ depletion further exacerbates compromised CFC in APP23 mice. Similarly, synchrony of phase-amplitude distribution and theta phase was markedly reduced in APP23.p38y + + , and virtually absent in APP23 ,ρ38γ ~/~ mice compared to p38y + + and ρ38γ ~/~ mice (fig. 9) . Consequently, the modulation index, a robust measure of CFC {21), was significantly lower in APP23.p38y /_ recordings as compared with p38y + + and ρ38γ ~/~ or even APP23.ρ38γ +/+ (Fig. 2K) .

p38y levels were determined in extracts from brains of humans without Alzheimer's disease (Braak 0) and from humans with different neuropatholocial disease stages ranging from Braak I to Braak VI (Table 3) . The results are shown in Fig. 23A and 23B. As can be seen from Fig. 23A and 23B, p38y levels were markedly reduced in humans as AD advances.

In summary, ρ38γ modulates excitotoxicity, neuronal circuit synchronicity, premature mortality and memory deficits in APP23 mice, without changes in Αβ . In

addition, ρ38γ levels are reduced in APP23 mice and humans suffering from AD. To determine if levels of tau affect the excitotoxicity- limiting effects of ρ38γ in vivo, we crossed non-mutant human tau-expressing Alzl7 mice (22) with ρ38γ ~/~ mice, to challenge these mice with PTZ. The results are shown in Fig. 3A-C. As can be seen from Fig. 3A-C, while tau expression did not affect seizure thresholds in

Alzl7.p38y + + mice, Alz 17 ,ρ38γ ~/~ mice presented with significantly enhanced seizure progression and severity compared to ρ38γ ~/~ mice (Fig. 3A-C) . Conversely, crossing ρ38γ ~/~ with tau-deficient tau ~/~ mice, revealed similar protection from PTZ-induced seizures in tau / .ρ38γ / and tau _/" .ρ38γ +/+ mice (Fig. 3D-F) .

To determine whether the Αβ toxicity-limiting effects of ρ38γ were tau-dependent , APP23.p38 γ _ ~ mice were crossed with tau mice, and the resulting crosses assessed for survival, memory deficit and neuronal network disfunction. The results are shown in Fig. 3G-3I, 24, 25, 26A, 26B, 27,

28A and B, and 29) . The exacerbating effects of ρ38γ loss on reduced survival, memory deficits, and neuronal network dysfunction of APP23 mice were virtually abolished in APP23. p38 γ . tau mice. These data also show that, compared with APP23 mice, APP23. p38 γ _ ~ animals had

aggravated memory deficits that persisted with aging. In contrast, as noted above, increasing tau levels in p38 γ mice (brought about by crossing with non-mutant tau- expressing Alzl7 mice) significantly enhanced PTZ-induced seizures in Alzl7.p38 γ _/~ mice. Conversely, when compared to tau _/~ .ρ38γ +/+ mice, tau _/~ .ρ38γ ~/_ animals showed similar protection from PTZ-induced seizures. Taken together, the effects of ρ38γ on excitotoxicity and Αβ toxicity are tau- dependent . Tau resides in a post-synaptic signaling complex with Fyn and PSD-95 that mediates Αβ-induced excitotoxicity (8) . Interaction of tau, Fyn and PSD95 in Alz 17 ,ρ38γ ~/~ brains was enhanced compared to Alz 17.p38y + + mice (Fig. 4A, B) , consistent with their increased sensitivity to PTZ-induced seizures. Conversely, no PSD-95/tau/Fyn complexes could be isolated from tau ~/~ and tau ~/~ ρ38γ ~/~ brains.

Strikingly, increased ρ38γ levels compromised, and

expression of a constitutive active variant of ρ38γ

(ρ38γ αΑ ) completely disrupted, PSD-95/tau/Fyn interaction in cells (Fig. 4C and D) . Pan-p38 inhibition stopped ρ38γ and p38Y CA -induced disruption of PSD-95/tau/Fyn complexes, furthermore indicating that ρ38γ activity is required (Fig. 4E and F) . PSD-95 co-purified more tau and Fyn from ρ38γ ~/~ than ρ38γ +/+ brains, suggesting increased PSD-95/tau/Fyn complex formation in the absence of ρ38γ (Fig. 4G and H) . PTZ transiently increased PSD-95/tau/Fyn complex formation in ρ38γ +/+ animals, and even further in ρ38γ ~/~ mice.

Similarly, PSD-95/tau/Fyn complex formation was markedly increased in APP23 ,ρ38γ ~/~ compared to APP23 ,ρ38γ +/+ and ρ38γ ~/~ brains (Fig. 41 and J) . Consistent with increased PSD-95/tau/Fyn complex formation, Fyn-mediated

phosphorylation of NR2B at Y1472, that facilitates

interaction of PSD-95 and NR2B {23, 24) , was increased in ρ38γ ~/~ brains. Similarly, ρ38γ and p38Y CA expression reduced Y1472-phosphorylation of NR2B.

Importantly, neither p38 CA, p38βCA nor p385CA reduced NR2B phosphorylation, indicating that regulation of PSD- 95/tau/Fyn complexes is a non-redundant function of ρ38γ. Interestingly, both ρ38γ and p38Y CA interacted with PSD-95 (Fig. 4C) , which was abolished by deleting the C-terminal PDZ interaction motif from ρ38γ and ρ38γ εΑ (Fig. 10A) . Both ρ38γ and ρ38γ εΑ also interacted with tau (Fig. 10B) . Since, ρ38γ and more so ρ38γ εΑ disrupted PSD-95/tau interaction in the absence of Fyn overexpression (Fig. IOC), but neither disrupted tau/Fyn interaction (Fig. 10D), ρ38γ appears to regulate PSD-95/tau/Fyn complexes at the level of PSD- 95/tau interaction.

While ρ38γ phosphorylates tau at multiple epitopes during long-term in vitro kinase assays, possibly contributing to tau hyperphosphorylation (25) , the temporal profile of p38y-induced tau phosphorylation in acute signaling, including excitotoxicity, remained unknown. Using

recombinant tau for short-term in vitro kinase reactions, we tested phosphorylation of a range of SP and TP sites, using available phosphorylation site-specific antibodies (Fig. 11A) . Tau was phosphorylated strongly at serine (S) 199 and threonine (T) 205, and less at S396 and S404, but not at other sites tested (Fig. 11B) . Site-specificity was confirmed by individually mutating S199, T205, S396 and S404 to alanine, which abolished p38Y ~ induced tau

phosphorylation tau at the mutated sites in vitro (Fig. 11C) . Mass spectrometric analysis of tau in kinase

reactions confirmed these 4 sites, and an additional fourteen low abundant sites. Co-expression of ρ38γ or ρ38γ εΑ and tau revealed that ρ38γ predominantly

phosphorylated tau at T205 and to a lesser degree at S199, but barely at S396 and S404 in cells (Fig. 5A) .

Similarly, T205 (and less so S199 and S396) were

phosphorylated in p38y CA transgenic mice. Phosphorylated T205 (pT205) increased after PTZ treatment of ρ38γ +/+ animals but was virtually abolished in p38 y ^ mice, whereas pS199, pS396 and pS404 were induced in both p38 γ + + and p38 γ _ ~ mice. Similarly, pT205 was markedly reduced in APP23.p38 animals compared with APP23. p38 γ +/+ mice.

Consistently, phosphorylation of T205 in primary neurons was markedly reduced by pan-p38 inhibition, while S199 phosphorylation remained unaffected. Taken together, these data indicate that T205 is a primary site in tau

phosphorylation by ρ38γ. To determine the functional relevance of tau

phosphorylation by ρ38γ at S199 and T205, we generated phosphorylation-mimicking (S199D and T205E) and - preventing (S199A and T205A) tau variants. We also prepared phosphorylation mimicking mutants of all other sites identified by mass spectrometry and assessed all mutants for their ability to co-purify with PSD-95, tau and Fyn. The results are shown in Fig. 5B, 5C and Fig. 31. PSD-95 co-purified with Fyn and all mutants except T205E. In this regard, T205E coprecipitated significantly less with PSD-95 as compared with PSD-95 as compared with non-mutant and T205A tau, while all other phosphorylation mimicking mutants of all other identified sites had no effect on PSD-95/tau/Fyn interaction. Microscale

thermophoresis and glutathione S-transferase-pulldown in vitro and fluorescence lifetime imaging microscopy (FLIM) - fluorescence resonance energy transfer (FRET) analysis in live cells confirmed the markedly compromised interaction of T205E tau with PSD-95. The T205E mutation did not hinder tau/Fyn interaction. These data suggests that phosphorylation of tau at T205 is sufficient to disrupt interaction with PSD-95. T205E and T205A mutations did not compromise tau/Fyn interaction. Importantly, ρ38γ εΑ disrupted PSD-95/tau/Fyn complexes in the presence of non- mutant tau, but had no effects when T205A tau was co- expressed (Fig. 5D and E) . In contrast, phospho-mimicking and -preventing S396 or S404 variants of tau had no effect on PSD-95/tau/Fyn interaction (Fig. 5C) . Taken together, this suggests that ρ38γ regulates PSD-95/tau/Fyn complexes via phosphorylation of tau at T205.

Disruption of NR/PSD-95/tau/Fyn complexes prevented exitotoxicity and Αβ-induced toxicity in primary neurons and APP23 mice (8) . Hence, phosphorylation of tau at T205 should mitigate or reduce Αβ-induced neurotoxicity. To test this, we used AAV-mediated gene transfer to express wild-type, T205A or T205E tau at similar levels in primary neurons (fig. 12) . Challenge with Αβ induced cell death in wild-type and T205A, but virtually not in T205E human tau- expressing hippocampal neurons, as indicated by increased LDH release (Fig. 5F) or EthDl uptake (fig. 12A) . H 2 0 2 - treatment exerted the same level of cytotoxicity in neurons irrespectively of the tau variant expressed. To test whether increasing levels or activity of ρ38γ in neurons similarly confer protection from Αβ toxicity, we expressed ρ38γ, p38y CA or a GFP control in primary neurons (Fig. 5G and Fig. 13) . Both, expressed ρ38γ and p38y CA enriched in dendritic spines, similar to endogenous ρ38γ. Neurons expressing ρ38γ and more so p38y CA were

significantly more resistant to Αβ-induced cell death compared to controls (Fig. 5H) . Neither expression of ρ38γ nor p38yCA limited H202-induced cell death. In summary, expression of site-specific phosphorylat ion-mimicking

T205E tau or increasing ρ38γ activity mitigated the toxic effects of Αβ in hippocampal neurons. Remaining Αβ toxicity in the presence of T205E tau or p38y CA was possibly due to endogenous tau, or alternative pathways (9) .

To determine if increased neuronal ρ38γ levels and/or activity limits excitotoxicity in vivo, we used AAV- mediated gene transfer to express ρ38γ, p38y CA or a GFP control in forebrains of newborn wild-type mice (fig. 14) and challenged them with PTZ at 2 months of age.

Expression of ρ38γ in vivo moderately, but significantly decreased progression of PTZ-induced seizures in 2 month- old mice, with a trend towards reduction of mean seizure severity, compared to GFP expressing mice (Fig. 51, J and fig. 15) . p38y CA expression profoundly increased the latency to develop severe seizures in response to PTZ administration, and significantly decreased the mean seizure severity as compared with control mice (Fig. 51, J and fig. 15) . Expression levels of ρ38γ and p38y CA varied between mice as expected from AAV-mediated gene

expression, with levels of ρ38γ being on average higher than those of p38y CA (Fig. 14B) . Interestingly, levels of both ρ38γ and p38y CA and seizure latency slopes showed positive linear correlation (ρ38γ: R2 = 0.483, P = 0.0832, s = 65.23 ± 30.20; p38yCA: R2 = 0.707, P = 0.0023, s = 215.1 ± 48.96), with a significantly pronounced level- dependent protective effect of ρ38γ εΑ over ρ38γ expression (F = 6.8407, P = 0.0214) (Fig. 5K) . Thus, levels of active ρ38γ kinase in vivo determine susceptibility to

excitotoxic signals.

Memory deficits in APP23. AAV p38rCA were significantly less severe compared to those of APP23.AAV GFP mice, as assessed in the Morris-water maze paradigm (Figs. 16-18) . APP23. AAV p38rCA mice showed memory performance similar to wild-type memory performance (AAV ■GFP AAV - 1 p38YCA

Adeno-associated virus (AAV) - mediated expression of WT and T205A, but not T205E tau or green fluorescent protein (GFP), in the forebrains of tau ^ mice enhanced PTZ- induced seizures (Fig. 19) . In contrast, expression of p38g CA in WT mice using AAV or in Thy1.2-p38 y CA transgenic mice decreased PTZ-induced seizures. AAV-mediated p38 y CA expression in APP23 mice rescued memory deficits and network aberrations (Fig. 34-38) ; the same was true for crossing APP23 with Thy .1.2-p38 y CA transgenic mice (Figs. 39-41) . In summary, the levels of active p38 γ kinase and tau phosphorylation at T205 determined susceptibility to excitotoxicity and Αβ toxicity.

Tau is a key mediator of deficits in APP transgenic mice (7, 8) , and tau has been suggested to transmit detrimental signals of Αβ in neurons by becoming aberrantly

phosphorylated {4, 27) . Here, we show that tau is part of an intrinsic molecular pathway involving phosphorylation at T205 mediated by ρ38γ to inhibit excito- and Αβ

toxicity. While we formally cannot exclude further non- tested sites being phosphorylated by ρ38γ, our data with T205A/E tau suggest that phosphorylation at T205 is key to modulating post-synaptic PSD-95/tau/Fyn complexes. Tau is required for the toxicity-limiting effects of ρ38γ, as ρ38γ depletion failed to exacerbate seizures in tau ~/~ . ρ38γ " /_ mice. Although other kinases might target T205 on tau in disease or physiologically {28-30) , the very distinct localization of PSD-95, tau and ρ38γ in a complex at the post-synapse indicates a specific and spatially compartmentalized role of ρ38γ downstream of synaptic NR activation .

While different roles have been characterized for other p38 kinases, the function of ρ38γ remained understudied. Here, our study revealed an unprecedented function of ρ38γ in the brain, by showing its involvement in tau-mediated Αβ toxicity, memory deficits and survival in AD mice. Its distinct spatial expression in post-synapses and unique sequence features, when compared to neuronally expressed ρ38 /β, likely contribute to this non-redundant function of ρ38γ in neurons. ρ38 /β have been described as

downstream mediators of excito- (11) and Αβ toxicity (12, 13) . Therefore and importantly, the ρ38γ function in excito- and Αβ toxicity we describe here is distinct from and opposite to ρ383/β.

In summary, our work suggests that phosphorylation of tau at T205 is part of an Αβ toxicity-inhibit ing response. This is contrary to the current view that tau

phosphorylation downstream of Αβ toxicity is a purely pathological response (27) . However, it is in line with the idea that tau is involved in normal physiologic signaling events in neurons likely involving NR signal transduction (9) . Finally, we have identified ρ38γ as an unprecedented Αβ-toxicity limiting signaling factor, which modulates tau-dependent excitotoxicity by site-specific phosphorylation of tau and controlling post-synapt ic PSD- 95/tau/Fyn complexes. This provides new insight into post- synaptic processes involved in early AD pathogenesis and may contribute to future drug development. References

I. C Ballatore, VM Lee, JQ Trojano ski, Nature reviews.

Neuroscience 8, 663-72 (2007) .

2. C Haass, DJ Selkoe, Nature reviews. Molecular cell biology 8, 101-12 (2007) .

3. K Iqbal, F Liu, CX Gong, C Alonso Adel, I Grundke-Iqbal , Acta Neuropathol 118, 53-69 (2009) .

4. EM Mandelko , E Mandelko , Cold Spring Harb Perspect Med 2, a006247 (2012) .

5. ES Musiek, DM Holtzman, Nat Neurosci 18, 800-6 (2015) .

6. M Rapoport, HN Dawson, LI Binder, MP Vitek, A Ferreira, Proc Natl Acad Sci U S A 99, 6364-9 (2002) .

7. ED Roberson et al . , Science 316, 750-4 (2007) .

8. LM Ittner et al . , Cell 142, 387-97 (2010) .

9. L Mucke, DJ Selkoe, Cold Spring Harb Perspect Med 2,

a006338 (2012) .

10. JJ Palop, L Mucke, Nat Neurosci 13, 812-8 (2010) .

II. GE Hardingham, H Bading, Nature reviews. Neuroscience 11, 682-96 (2010) .

12. Q Wang, DM Walsh, MJ Rowan, DJ Selkoe, R Anwyl, J Neurosci 24, 3370-8 (2004) .

13. S Li et al., J Neurosci 31, 6627-38 (2011) .

14. AA Ittner, A Gladbach, J Bertz, LS Suh, LM Ittner, Acta Neuropathol Commun 2, 149 (2014) .

15. MA Fabian et al . , Nat Biotechnol 23, 329-36 (2005) .

16. MB Menon, S Dhamija, A Kotlyarov, M Gaestel, Autophagy, 0 (2015) .

17. C Sturchler-Pierrat et al . , Proc Natl Acad Sci U S A 94, 13287-92 (1997) .

18. G Buzsaki, EI Moser, Nat Neurosci 16, 130-8 (2013) .

19. R Goutagny, J Jackson, S Williams, Nat Neurosci 12, 1491-3 (2009) .

20. RT Canolty et al . , Science 313, 1626-8 (2006) .

21. AB Tort, RW Komorowski, JR Manns, NJ Kopell, H Eichenbaum, Proc Natl Acad Sci USA 106, 20942-7 (2009) . 22. A Probst et al . , Acta Neuropathol 99, 469-81 (2000) .

23. Y Rong, X Lu, A Bernard, M Khrestchatisky, M Baudry, J Neurochem 79, 382-90 (2001) .

24. M Aarts et al . , 1 Science 298, 846-50 (2002) .

25. M Goedert et al . , FEBS Lett 409, 57-62 (1997) .

26. S Mondragon-Rodriguez et ai . , J Biol Chem 287, 32040-53 (2012) .

27. LM Ittner, J Gotz, Nature reviews. Neuroscience 12, 65-72 (2011) .

28. JZ Wang, Q Wu, A Smith, I Grundke-Iqbal, K Iqbal, FEBS Lett 436, 28-34 (1998) .

29. V Buee-Scherrer, M Goedert, FEBS Lett 515, 151-4 (2002) .

30. A Cavallini et al . , J Biol Chem 288, 23331-47 (2013) .