UBQLN2-HSP70 axis reduces poly-Gly-Ala aggregates and alleviates behavioral defects in the C9ORF72 animal model

•UBQLN2 is recruited to reduce poly-GA aggregation and alleviate its neurotoxicity•UBQLN2 can recognize HSP70 ubiquitination facilitate degradation•The UBQLN2-HSP70 interaction required for degradation•Pharmacologically enhancing mitigates behavioral defects in mice Expansion of a hexanucleotide repeat GGGGCC (G4C2) the intron C9ORF72 gene most common cause amyotrophic lateral sclerosis (ALS) frontotemporal dementia (FTD) (C9-ALS/FTD). Transcripts carrying G4C2 expansions generate neurotoxic dipeptide (DPR) proteins, including poly-Gly-Ala (poly-GA), which tends form protein aggregates. Here, we demonstrate that UBQLN2, another ALS/FTD risk factor, aggregates poly-GA-induced neurotoxicity. UBQLN2 could ubiquitination, facilitates UBQLN2-HSP70-GA complex formation promotes degradation. ALS/FTD-related mutants fail bind clear Disruption between inhibits C9-ALS/FTD iPSC-derived neurons. Finally, by chemical compound 17AAG at adult stage animals. Our findings suggest critical role axis aggregate clearance C9-ALS/FTD. The accumulation notorious feature neurodegenerative disorders, (Blokhuis et al., 2013Blokhuis A.M. Groen E.J. Koppers M. van den Berg L.H. Pasterkamp R.J. Protein sclerosis.Acta Neuropathol. 2013; 125: 777-794Crossref PubMed Scopus (337) Google Scholar; Ross Poirier, 2004Ross C.A. Poirier M.A. disease.Nat. Med. 2004; 10: S10-S17Crossref (2350) Scholar). ALS adult-onset fatal motoneuron disease characterized progressive disability motor system (Andersen Al-Chalabi, 2011Andersen P.M. Al-Chalabi A. Clinical genetics sclerosis: what do really know?.Nat. Rev. Neurol. 2011; 7: 603-615Crossref (511) Geevasinga 2016Geevasinga N. Menon P. Özdinler P.H. Kiernan M.C. Vucic S. Pathophysiological diagnostic implications cortical dysfunction ALS.Nat. 2016; 12: 651-661Crossref (111) Taylor 2016Taylor J.P. Brown Jr., R.H. Cleveland D.W. Decoding ALS: from genes mechanism.Nature. 539: 197-206Crossref (997) FTD second type presenile degeneration frontal temporal lobes brain (Graff-Radford Woodruff, 2007Graff-Radford N.R. Woodruff B.K. Frontotemporal dementia.Semin. 2007; 27: 48-57Crossref (98) Guerreiro 2015Guerreiro R. Brás J. Hardy SnapShot: FTD.Cell. 2015; 160: 798-798.e1Abstract Full Text PDF (57) co-occurrence symptoms observed up 15%–50% cases (Guerreiro Nguyen 2018Nguyen H.P. Van Broeckhoven C. der Zee genomic era their FTD.Trends Genet. 2018; 34: 404-423Abstract (152) Excessive expansion chromosome 9 open reading frame 72 (C9ORF72) was discovered be genetic (C9-ALS/FTD) (DeJesus-Hernandez 2011DeJesus-Hernandez Mackenzie I.R. Boeve B.F. Boxer A.L. Baker Rutherford N.J. Nicholson Finch N.A. Flynn H. Adamson al.Expanded noncoding region causes 9p-linked ALS.Neuron. 72: 245-256Abstract (3230) Renton 2011Renton A.E. Majounie E. Waite Simón-Sánchez Rollinson Gibbs J.R. Schymick J.C. Laaksovirta Swieten Myllykangas L. al.ITALSGEN ConsortiumA 9p21-linked ALS-FTD.Neuron. 257-268Abstract (2953) potential pathological mechanisms underlying include haploinsufficiency, RNA foci, proteins (DPRs) repeat-associated non-ATG translation (RAN) (Freibaum Taylor, 2017Freibaum B.D. repeats C9ORF72-related ALS-FTD.Front. Mol. Neurosci. 2017; 35Crossref (133) Gao 2017Gao F.B. Richter J.D. Rethinking unconventional neurodegeneration.Cell. 171: 994-1000Abstract (36) Haeusler 2016Haeusler A.R. Donnelly C.J. Rothstein expanding biology C9orf72 nucleotide 17: 383-395Crossref (129) Hao 2020Hao Z. Wang Ren G. Role pathogenesis dementia.Neurosci. Bull. 2020; 36: 1057-1070Crossref (2) Paul Gitler, 2014Paul J.W. Gitler A.D. Cell biology. Clogging information flow ALS.Science. 2014; 345: 1118-1119Crossref (10) Rohrer 2015Rohrer Isaacs Mizielinska Mead Lashley T. Wray Sidle K. Fratta Orrell R.W. al.C9orf72 sclerosis.Lancet 14: 291-301Abstract (156) C9-DPR inclusions are abundant cerebral neocortex, cerebellum, hippocampus but less so spinal cord (Davidson 2014Davidson Y.S. Barker Robinson A.C. Thompson Harris Troakes Smith B. Al-Saraj Shaw al.Brain distribution lobar neurone associated with C9ORF72.Acta Commun. 2: 70Crossref (83) 2013Mackenzie Arzberger Kremmer Troost D. Lorenzl Mori Weng S.M. Haass Kretzschmar H.A. Edbauer Neumann Dipeptide pathology mutation cases: clinico-pathological correlations.Acta 126: 859-879Crossref (234) Schludi 2015Schludi M.H. May Grässer F.A. Rentzsch Küpper Klopstock German Consortium Lobar DegenerationBavarian Brain Banking AllianceDistribution cellular models suggests link transcriptional silencing.Acta 130: 537-555Crossref (121) Among five species (poly-GP, poly-PA, poly-GR, poly-PR, poly-GA), believed DPR brains patients (Mackenzie Scholar, 2015Mackenzie Frick Gendron T.F. Petrucelli Cashman Prudlo Quantitative analysis correlations different pathologies carriers.Acta 845-861Crossref (142) 2013Mori Schmid Cruts al.The translated into aggregating dipeptide-repeat FTLD/ALS.Science. 339: 1335-1338Crossref (832) Importantly, compact neurotoxicity, faithfully mimicking neurocytoplasmic suggesting one driving forces neurotoxicity during chronic (Lee 2017Lee Y.B. Baskaran Gomez-Deza Chen H.J. Nishimura B.N. Adachi Y. Stepto poly GA RAN-translated plays key via toxicity.Hum. 26: 4765-4777Crossref (44) 2014May Hornburg Schwenk B.M. Banzhaf-Strathmann FTLD/ALS-associated Gly-Ala neuronal toxicity Unc119 sequestration.Acta 128: 485-503Crossref (205) Yamakawa 2015Yamakawa Ito Honda Kubo Noda Nakajima Suzuki Characterization molecular c9FTD/ALS.Hum. 24: 1630-1645Crossref (103) Zhang 2016Zhang Y.J. Grima Sasaguri Jansen-West Xu Y.F. Katzman R.B. Gass Murray M.E. Shinohara al.C9ORF72 poly(GA) sequester impair HR23 nucleocytoplasmic transport proteins.Nat. 19: 668-677Crossref (197) Mutations cell display incomplete penetrance (Cooper-Knock 2014Cooper-Knock P.J. Kirby widening spectrum disease; genotype/phenotype modifiers clinical phenotype.Acta 127: 333-345Crossref (117) Galimberti 2014Galimberti Arosio Fenoglio Serpente Cioffi Bonsi Rossi Abbate Mari Scarpini Incomplete expansions: frequency cohort geriatric non-demented subjects.J. Alzheimers Dis. 39: 19-22Crossref (26) Murphy 2017Murphy Arthur K.C. Tienari Houlden Chiò Traynor B.J. Age-related expansion.Sci. Rep. 2116Crossref (60) Zucchi 2019Zucchi Ticozzi Mandrioli Psychiatric beyond neuron disorder.Front. 2019; 13: 175Crossref (37) Scholar), other factor(s) converge trigger factor (Deng 2011Deng H.X. W. Hong S.T. Boycott K.M. Gorrie G.H. Siddique Yang Fecto F. Shi Zhai al.Mutations dominant X-linked juvenile ALS/dementia.Nature. 477: 211-215Crossref (838) has been implicated degradation misfolded (Hjerpe 2016Hjerpe Bett J.S. Keuss M.J. Solovyova McWilliams T.G. Johnson Sahu I. Varghese Wood Wightman al.UBQLN2 mediates autophagy-independent proteasome.Cell. 166: 935-949Abstract (181) Renaud 2019Renaud Picher-Martel V. Codron Julien Key dementia.Acta 103Crossref Samant 2018Samant R.S. Livingston C.M. Sontag E.M. Frydman Distinct proteostasis circuits cooperate nuclear cytoplasmic quality control.Nature. 563: 407-411Crossref (86) ?entürk 2019?entürk Lin Zuo Mao Watson Mikos A.G. Bellen Ubiquilins regulate autophagic flux through mTOR signalling lysosomal acidification.Nat. Biol. 21: 384-396Crossref (77) possesses an N-terminal ubiquitin-like (UBL) domain binds proteasome C-terminal ubiquitin-associated (UBA) polyubiquitin chains substrates Marín, 2014Marín ubiquilin family: evolutionary patterns functional insights.BMC Evol. 63Crossref (43) Between UBL UBA domains, there conserved set STI-like binding nearby unique proline-X-X-proline (PXXP) domain. Interestingly, ALS/FTD-linked mutations or near PXXP Osaka 2016Osaka Disturbance proteasomal pathways sclerosis-linked 2.Biochem. Biophys. Res. 472: 324-331Crossref (50) UBQLN2-positive exist not only also patients, they colocalize (Bennion Callister 2016Bennion Ryan Sim Pickering-Brown Modelling physiologically relevant size.Hum. 25: 5069-5082PubMed Brettschneider 2012Brettschneider Deerlin V.M. J.L. Kwong Lee E.B. Ali Y.O. Safren Monteiro Toledo J.B. Elman al.Pattern FTLD indicates presence expansion.Acta 2012; 123: 825-839Crossref (145) Scotter 2017Scotter E.L. Smyth Bailey Wong C.H. de Majo Vance Synek Turner Pereira Charleston New Zealand: pathologic study using banked human tissue.Neurobiol. Aging. 49: 214.e1-214.e5Crossref (12) However, roles regulating remain unclear. show forms colocalizes postmortem tissues C9-FTD patients. Furthermore, lysine-free help HSP70. Enhancing treatment rescues uncover mechanism We first examined samples In healthy individuals, no GA-positive signal cortex (seven samples), while (six were (Figure 1A, black arrow; Figure S1A, Table S1), confirming present (Mori occur small fraction familial cases, inclusion bodies have found wide control samples, distributed evenly cytoplasm nucleus 1A; S1). frequently lost even formed red S1, arrow). dystrophic neurites arrowhead; arrowhead). Some colocalized 1B), these two might actively interact vivo. transfected N2a cells, GFP (or GFP-GA5) diffusely 1C; S2A). GFP-GA100 (referred as GFP-GA if indicated) 1C, white Endogenous some aggregates, indicating recruitment Aggregates high weight (HMW) detergent resistant stacking gel SDS-PAGE, monomers resolved separating addition being gel, portion (HMW-GFP-GA; 1D). (HMW-UBQLN2), associates complexes A filter trap assay showed insoluble GFP-GA-expressing controls next used antibodies pull down HMW-GFP-GA coimmunoprecipitated 1E). To examine sgRNAs against increased sgUBQLN2/Cas9. This effect sgUBQLN2-nonexpressing cells (Figures 1F 1G). Consistently, immunoblotting assays levels significantly sgUBQLN2/Cas9-transfected 1H 1I). Compared poly-DPRs, S3A). sgUBQLN2 enhanced HMW-poly-GA rather than poly-DPRs S3B), prefers target Taken together, our results clearance. decreased exogenous compared 2A–2C). alone had little on GFP-GA5 expression 2F; S2B). There change mRNA UBQLN2-transfected excluding possibility promoter interference cotransfection 2D). further asked whether toxicity. GFP-, GFP-GA5-, MTT values GFP-GA-transfected decreased. GFP-GA-reduced rescued overexpression 2E; S2C). GA-induced active caspase-3 inhibited 2F). GFP-transfected neurons (GFP+/Tuj1+) nontransfected (GFP?/Tuj1+), neurite number (arrow, GFP-GA+/Tuj1+) dramatically reduced 2G 2H), supporting idea impairs morphogenesis (May 2014Zhang Bieniek K.F. W.L. Caulfield Hubbard Daughrity al.Aggregation-prone c9FTD/ALS inducing ER stress.Acta 505-524Crossref (212) effectively reduction poly-GA. expressed universally mouse brains, cortex, hippocampus, olfactory bulb 3A). more abundantly primary cultured glia 3B). performed intracerebroventricular administration AAVs P0 neonatal pups S4A S4B). After 1 month, fluorescence regions, bulb, cerebellum AAV-GFP-injected S4C) rare (data shown). Most GFP-positive pyramidal S4C S4D). contrast S2D S4D), numerous puncta (arrowhead) AAV-GFP-GA-infected similar S1A). These rarely AAV-GFP-infected S4D S4E). infected coinfected kinds viruses Mice subjected 3 months age followed histological biochemical analyses 3C). (human UBQLN2) ~64% endogenous 3D). Exogenous (arrow), size much smaller those 3E). amount HMW-GA 3F). comparable AAV-UBQLN2-infected 3G), occurs level. Motor coordination learning measured rotarod test. mice, AAV-GFP-GA group latency fall day, mild increase third days, severe balance poly-GA-expressing 3H). Strikingly, + AAV-UBQLN2 performance test behavior affected either upper motoneurons lower cord, balance, learning, neuromuscular coordination, muscle strength/stamina (Jones Roberts, 1968Jones Roberts D.J. quantiative measurement inco-ordination naive acelerating rotarod.J. Pharm. Pharmacol. 1968; 20: 302-304Crossref (418) Luh 2017Luh L.M. Das Bertolotti qMotor, rules sensitive, robust quantitative mice.Nat. Protoc. 1451-1457Crossref (8) ChAT-positive groups S4F), loss major mice. probably due low infection efficiency AAV after injection (Chew 2015Chew Prudencio Castanedes-Casey C.W. Kurti al.Neurodegeneration. TDP-43 pathology, loss, deficits.Science. 348: 1151-1154Crossref (260) AAV-GFP groups, exploration distance (~66.9%) time (~53.6%) new zone, delivery 3I 3J). significant difference Y-maze S2F–S2H). decrease 3K 3L). AAV-GA TDP43 cytoplasm, absent 3M). PSD95 expressing GFP-GA, indicated excitatory synapses. 3N). data AAV-delivered alleviates cognitive acts shuttl