Deficiency of the miR-29a/b-1 cluster leads to ataxic features and cerebellar alterations in mice

1

2

3Q5

4

5

6

789101112131415161718

1 9

2021222324

2526272829303132

43

4445

46

47

Neurobiology of Disease xxx (2014) xxx–xxx

YNBDI-03341; No. of pages: 14; 4C: 5, 6, 7, 8, 9, 10, 12

Contents lists available at ScienceDirect

Neurobiology of Disease

j ourna l homepage: www.e lsev ie r .com/ locate /ynbd i

Deficiency of the miR-29a/b-1 cluster leads to ataxic features andcerebellar alterations in mice☆,☆☆

D P

RO

OFAikaterini S. Papadopoulou a,b, Lutgarde Serneels a,b, Tilmann Achsel a,b, Wim Mandemakers g,

Zsuzsanna Callaerts-Vegh c, James Dooley d, Pierre Lau a,b, Torik Ayoubi a,b,f, Enrico Radaelli a,b,Marco Spinazzi a,b, Melanie Neumann e, Sébastien S. Hébert h,i, Asli Silahtaroglu j, Adrian Liston d,Rudi D'Hooge b,c, Markus Glatzel d, Bart De Strooper a,b,k,⁎a VIB Center for the Biology of Disease, VIB Leuven, 3000 Leuven, Belgiumb Center for Human Genetics and Leuven institute for neurodegenerative disorders (LIND), KU Leuven and Universitaire Ziekenhuizen, 3000 Leuven, Belgiumc Laboratory of Biological Psychology, Department of Psychology, University of Leuven, 3000 Leuven, Belgiumd Autoimmune Genetics Laboratory, Department of Microbiology and Immunology, VIB and KULeuven, 3000 Leuven, Belgiume Institute of Neuropathology, University Medical Center Hamburg-Eppendorf, UKE, 20246 Hamburg, Germanyf Saint James School of Medicine, Plaza Juliana, Kralendijk, Bonaire, Dutch Caribbean, The Netherlandsg Department of Clinical genetics, Erasmus MC Faculty Building, 3015GE Rotterdam, The Netherlandsh Centre de Recherche du CHUQ de Quebec, CHUL, Axe Neurosciences, Quebec, Canadai Université Laval, Département de psychiatrie et de neurosciences, G1V 4G2 Quebec, Canadaj Wilhelm Johannsen Centre for Functional Genome Research, Department of Cellular and Molecular Medicine, Faculty of Health and Medical Sciences, University of Copenhagen,DK-2200 Kobenhavn, Denmarkk Department of Molecular Neuroscience, UCL Institute of Neurology, WC1N 3BG London, UK

UNC

Abbreviations:miR,microRNA;BACE1, beta secreataseneuronalnavigator3;SCA17,spinocerebellarataxia17;FISHtion; LNA, locked nucleic acid; APP, amyloid precursor proprotein 22.☆ Author contributions: ASP and BDS conceived the pro

MS, MN, BDS performed research or analysed data; ASP, LAL, RD,MG, BDS designed research; ASP, BDSwrote thepaped input on the manuscript.☆☆ The authors declare no direct competing financial in

⁎ Corresponding author at: Laboratory for theResearchVIB Center for the Biology of Disease, KU Leuven CentHerestraat 49 box 602, 3000 Leuven, Belgium.

E-mail address: [email protected] online on ScienceDirect (www.sciencedi

http://dx.doi.org/10.1016/j.nbd.2014.10.0060969-9961/© 2014 Published by Elsevier Inc.

Please cite this article as: Papadopoulou, A.Smice, Neurobiol. Dis. (2014), http://dx.doi.or

TE

a b s t r a c t
a r t i c l e i n f o
33

34

35

36

37

38

39

40

41

42

Article history:Received 26 June 2014Revised 5 September 2014Accepted 1 October 2014Available online xxxx

Keywords:miR-29a/b-1 cluster knockoutAtaxiaLocomotor behaviorCerebellumPurkinje cellsDendritesVoltage gated potassium channel

RREC

miR-29 is expressed strongly in the brain and alterations in expression have been linked to several neurologicaldisorders. To further explore the function of this miRNA in the brain, we generated miR-29a/b-1 knockout ani-mals. Knockout mice develop a progressive disorder characterized by locomotor impairment and ataxia. The dif-ferent members of the miR-29 family are strongly expressed in neurons of the olfactory bulb, the hippocampusand in the Purkinje cells of the cerebellum.Morphological analysis showed that Purkinje cells are smaller and dis-play less dendritic arborisation compared to their wildtype littermates. In addition, a decreased number of par-allel fibers form synapses on the Purkinje cells. We identified several mRNAs significantly up-regulated in theabsence of the miR-29a/b-1 cluster. At the protein level, however, the voltage-gated potassium channel Kcnc3(Kv3.3) was significantly up-regulated in the cerebella of the miR-29a/b knockout mice. Dysregulation ofKCNC3 expression may contribute to the ataxic phenotype.

© 2014 Published by Elsevier Inc.

O

48

49

50

51

52

53

54

55

56

57

58

59

60

61

; AD,Alzheimer'sdisease;NAV3,,fluorescenceinsituhybridiza-tein; PMP22, peripheral myelin

ject. ASP, LS, TA,WM, JD, PL, ER,S, TA, TA, WM, ZCV, PL, SSH, AS,er. All authors read andprovid-

terests.of NeurodegenerativeDiseases,er for Human Genetics, O&N4

e (B. De Strooper).rect.com).

., et al., Deficiency of the miRg/10.1016/j.nbd.2014.10.006

Introduction

ThemiR-29 family consists of three relatedmiRNAs, miR-29a,−29band −29c, which are transcribed from two different genes as clusters;miR-29a/b-1 and miR-29b-2/c located on chromosome 6 and chromo-some 1 of the mouse genome, respectively. miR-29b-1 and miR-29b-2are identical in sequence, while miR-29a and miR-29c differ in one nu-cleotide (Chang et al., 2008; Hwang et al., 2007). miR-29 is ubiquitouslyexpressed and implicated in biological processes such as fibrosis (Ottet al., 2011; van Rooij et al., 2008; Xiao et al., 2012) myogenesis(Kapinas et al., 2010), metabolism (Long et al., 2011; Silva et al., 2011)and immune responses (Ma et al., 2011; Papadopoulou et al., 2012;Smith et al., 2012).

We became interested in the role of miR-29 in the brain because ofits high expression there and its association with neurodegenerative

-29a/b-1 cluster leads to ataxic features and cerebellar alterations in

http://dx.doi.org/10.1016/j.nbd.2014.10.006

mailto:[email protected]


http://www.sciencedirect.com/science/journal/09699961

www.elsevier.com/locate/ynbdi


T

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

100

101

102

103

104

105

106

107

108

109

110

111

112

113

114

115

116

117

118

119

120

121

122

123

124

125

126

127

128

129

130

131

132

133

134

135

136

137

138

139

140

141

142

143

144

145

146

147

148

149

150

151

152

153

154

155

156

157

158

159

160

161

162

163

164

165

166

167

168

169

170

171

172

173

174

175

176

177

178

179

180

181

182

183

184

185

2 A.S. Papadopoulou et al. / Neurobiology of Disease xxx (2014) xxx–xxx

UNCO

RREC

diseases. miR-29 controls post-transciptionally the expression of β-secretase (BACE1), an enzyme involved in the generation of theamyloid-β peptide in Alzheimer's disease (AD) (Hebert et al., 2008).miR-29 is also proposed to control neuronal navigator 3 (NAV3) expres-sion, which was found to be up-regulated in Alzheimer's disease pa-tients (Shioya et al., 2010). miR-29b targets progranulin and controlsits production/secretion in vitro, supporting a role ofmiR-29 in the path-ogenesis of Frontotemporal Dementia (Jiao et al., 2010). miR-29 isdown-regulated in Huntington's disease and in the R6/2 mouse modelof Huntington's disease (Johnson et al., 2008; Packer et al., 2008), andis up-regulated in Amyotrophic Lateral Sclerosis (ALS) patients(Shioya et al., 2010), but also in amousemodel, the SOD1G93A, of thedis-ease (Nolan et al., 2014). Prion infection in the brains of mice reveals anup-regulation of miR-29a amongst others, in the CA1 region of the hip-pocampus, suggesting a protective, survival role in neurons (Majer et al.,2012). In Simian Immunodeficiency Virus (SIV) infected macaques,miR-29 is transported from astrocytes to neurons through exosomesto down-regulate the levels of PDGFβ, and combined with opiate ad-ministration, is leading to neuronal apoptosis (Hu et al., 2012). miR-29participates in the regulation of the titetraprolin (TTP), an RNA-binding protein, during SIV CNS infection (Liu et al., 2013). A recentstudy suggests that miR-29 in astrocytes has neuroprotective functionsin a forebrain ischemia model, as it is elevated in the CA1 region of thehippocampus and accounts for down-regulating the proapototic BH3-only PUMA (p53 up-regulated modulator of apoptosis) (Ouyang et al.,2013). In a cellular model of spinocerebellar ataxia 17 (SCA17), down-regulation of miR-29a/b correlates with up-regulated BACE1 levels,however, the molecular mechanism linking BACE1 to SCA17 was notaddressed (Roshan et al., 2012).

miR-29a and miR-29b are up-regulated in differentiatingglioblastoma-multiforme-initiating cells, which act like stem cells.Targeting Mcl-1 in these cells enhances apoptosis (Aldaz et al., 2013);which can have translational application. miR-29 involvement in glio-blastomas is further supported by an independent study, (Zhao et al.,2014).

up-regulation of miR-29 occurs during aging (Ugalde et al., 2011;Zhang et al., 2013) and during neuronal maturation (Kole et al., 2011).A miR-29 neuroprotective role is further supported in ischemia modelswhere miR-29 down-regulation contributes to brain damage (Khannaet al., 2013; Pandi et al., 2013). Recently, miR-29b and miR-29a wereshown to be down-regulated under hypoxic conditions or activationof δ-opioid receptor, suggesting a protective role in cortical neurons(Yang et al., 2012). miR-29 targets Arpc3, and is by that way involvedin spinemorphogenesis and synaptic plasticity in hippocampal neuronsin vitro, while inhibition of miR-29a/b results in decreased thresholdsensitivity to neurotransmission in the postsynaptic membrane (Lippiet al., 2011).

A key limitation of previous studies on miR-29 biology is the depen-dence on in vitromodels.We therefore generated amiR-29a/b-1 knock-out mouse (Papadopoulou et al., 2012) and investigate here theimplications of loss of function of this miRNA in the brain. We findthat the knockout mice develop a progressive neurological impairment.In particular, motor coordination becomes affected early and can betraced back to abnormalities in the Purkinje cell layer of the cerebellum.We demonstrate that loss of miR-29a/b-1 expression in the cerebellumleads to ~60% up-regulation of the voltage-activated potassium channelKCNC3 (Kv3.3), which is important for the rapid firing of Purkinje cellsand motor function (Hurlock et al., 2008, 2009; Joho et al., 2006).

Materials and methods

Mice

Mir29-aMir29b-1tm1Bds animals used in this study are hereby namedmiR-29a/b-1 knockout animals and were generated as previously de-scribed (Papadopoulou et al., 2012).

Please cite this article as: Papadopoulou, A.S., et al., Deficiency of the miRmice, Neurobiol. Dis. (2014), http://dx.doi.org/10.1016/j.nbd.2014.10.006

ED P

RO

OF

Fluorescence in situ hybridization

Fluorescence in situ hybridisation (FISH) was performed as previ-ously describedwithminormodifications (Silahtaroglu et al., 2007). Ex-tracted brains were flash-frozen in iso-pentane immersed in liquidnitrogen and then embedded for sagittal sectioning at 12–20 μm thick-ness (cryostatMicromHM550). After fixation, sectionswere acetylatedand prehybridised for 1–3 hours at 55 °C, followed by hybridisationwith locked nucleic acids (LNA) and 2-O′-methyl- modified oligonucle-otides conjugated with FITC probes (Ribotask, Denmark) (Aronica et al.,2010). After hybridisation at the appropriate temperature as indicatedbelow for 20–30 minutes, sections were washed in SSC (saline sodiumcitrate) and then incubated with 3% H2O2 for 10 minutes and furtherwashed and blocked before incubation with anti-FITC/HRP (1:75)(Roche) antibody followed by washes. Amplification of the signal wasachieved by using the tyramide signal amplification technique (TSA,Perkin-Elmer) (1:60 dilution in amplification buffer) followed by thefinalwashes andmounting of the sections in ProLongGoldmedium con-taining DAPI (Invitrogen). Hybridisation time for miR-29a was 20 mi-nutes. Hybridisation temperature for miR-29a was 69 °C, miR-29b was55 °C, miR-219 was 49 °C and the scramble probe was 55 °C. Anti-FITC/HRP antibody was diluted 1:100. Incubation time with TSA was4 minutes for miR-29a and 6 minutes for miR-29b. Washes after hy-bridization with the miR-29a were performed with 0.1xSSC, 6 timesand for the other probes with 0.1xSSC, 3 times followed by 3 morewashes in 0.5xSSC. Imageswere acquired on aNikon A1R Eclipse Ti con-focal microscope using Plan Apo 4x (0.20 dry) and Plan Apo 10x (0.45dry) objectives and image analysis was performed using the ImageJsoftware.

Histopathology of the CNS

Animals of 6 months old were sacrificed with ketamine hydro-chloridum (115 mg/ml (DECHRA)), atropine sulphate (0.25 mg/ml(STEROP)) and xylazine (2% (VMD)), followed by intracardial perfusionwith ice-cold 4% formaldehyde (Sigma-Aldrich) solution. The entirehead and spine were collected and further fixed for 48 hours at 4 °C.The brain was removed from the skull and serially sliced using the cor-onal brain matrix system. The spine was decalcified in a supersaturatedsolution of tetra-sodium EDTA and cross sectioned at the level of cervi-cal, thoracic, lumbar and sacral segments. Samples were then processedfor histopathological examination and embedded in paraffin blocks. Ac-cording to the protocols described by Prophet et al. (1992) 5 and 10 μmthick sections obtained, were stained with Haematoxylin and Eosin(H&E) and Luxol Fast Blue (LFB) respectively. Additionally, 5 μm thickserial sections were also immunostained with a rabbit polyclonal anti-body raised against glial fibrillary acidic protein (GFAP) as previouslydescribed (Radaelli et al., 2009). For the morphometrical analyses ofthemuscles, 6 months oldmicewere sacrificedwith xylazine and keta-mine, followed by intracardial perfusionwith ice-cold 4% formaldehydesolution. Skeletal muscle tissue from the hind leg (musculus quadricepsfemoris) was removed and further fixed for at least 48 hours. Sampleswere then embedded in O.C.T. and snap frozen in 2-methyl-butanecooled with liquid nitrogen. After embedding, samples were cut in8 μm thick sections using a cryostat, and were stained with H&E solu-tions using standard protocols. Pictures were taken with an OlympusXT microscope with 20× objective. The diameter, the ferret max andthe ferret min values of at least 100 muscle fibers per animal (n = 3)were calculated using Axioscope software.

Behavioral tests

All behavioral analyses were donewith female mice and blind to thegenotype. Afirst group ofmicewas tested between 9 and 13 weeks afterbirth, while a second (different batch of mice) was tested between 22and 26 weeks old. All mice were backcrossed for at least 7 generations



T

186

187

188

189

190

191

192

193

194

195

196

197

198

199

200

201

202

203

204

205

206

207

208

209

210

211

212

213

214

215

216

217

218

219

220

221

222

223

224

225

226

227

228

229

230

231

232

233

234

235

236

237

238

239

240

241

242

243

244

245

246

247

248

249

250

251

252

253

254

255

256

257

258

259

260

261

262

263

264

265

266

267

268

269

270

271

272

273

274

275

276

277

278

279

280

281

282

283

284

285

286

287

288

289

290

291

292

293

294

295

296

297

298

299

300

301

302

303

304

305

306

307

3A.S. Papadopoulou et al. / Neurobiology of Disease xxx (2014) xxx–xxx

UNCO

RREC

to the C57Bl6/J background.We used aminimumof 9 animals per geno-type per assay. Mice were in advance habituated to their new environ-ment for a week before tests were conducted during the light period.Health and weight of the mice were recorded routinely during thewhole period of testing.

Balance Beam TestMice had to traverse elevated wooden beams of graded and increas-

ing difficulty (Carter et al., 1999). The beams were 50 cm elevated fromthe groundand 1m long. Square beamshad a cross-section of 28, 12 and5mmand round beams had a diameter of 28, 17 and 11mmrespective-ly. Mice were trained 4 times a day for 3 consecutive days on the squarebeam of the largest cross-section, until they were able to cross within20 seconds. On the testing day, the animals started from low to high de-gree of difficulty, first on the squared and then on the circular beams.The time needed to traverse each beam as well as the number of slipsof the hindlimbs was recorded.

Gait Analysis TestLocomotion and gaiting was assessed using the DigiGait™ analysis

system (Mouse Specifics Inc, Boston) (Nori et al., 2011; Stroobantset al., 2013). A digital camera beneath the animal was used to recordthe pawmovements on a transparent belt rolling at 16 cm/s. Gait indi-ces were provided by the Mouse Specifics program software. The rightand left limb values obtained were averaged for the front and hindlimbs. Stride length was calculated as the ratio between the speed ofthe belt and the stride frequency. Paw angle was measured as the de-gree of deviation of the paw from the longitudinal body axis.

Accelerated Rotarod TestMice were subjected to the accelerating rotarod (Med Associates, St

Albans) as previously described (Callaerts-Vegh et al., 2006). Trainingwas first conducted at constant speed of 4 rpm for 2 minutes. Therotarod was then gradually accelerated (4 to 40 rpm) in 4 consecutivetrials of 5 minutes each. Trials were 10 minutes apart. The animal's la-tency to fall was recorded with a cut-off time of 5 minutes.

RNA isolation and RT-qPCR

Total RNA and protein isolation was conducted using the miRVanamiRNA isolation kit (Ambion), according to themanufacturer's protocolwith slightmodifications. After extraction, tissueswere homogenized in1ml of TRIzol (Invitrogen)using a 22G syringe and 0.2ml of chloroform:isoamyl alcohol (Invitrogen) (24: 1) was added before vortexing for1 minute and centrifuging at 14,000 rpm for 3 minutes. Then, to theRNA-containing phase, 1.25 volumes of 100% ethanol were added fol-lowing the manufacturer's protocol. RNA was finally eluted in 20 μl ofRNase-free H2O and quantified using the ND-1000 Nanodrop. TotalRNA was subjected to DNase I treatment (RQ1, Roche) according tomanufacturer's protocol and subsequently repurified. Reverse transcrip-tionwas performed on 2 μg of total RNA using the Superscript II ReverseTranscriptase (Invitrogen), following the company's protocol and usingeither random hexamers. The cDNA was diluted in water and quantita-tive PCR was performed using the SYBR Green Dye I (Roche), accordingto the manufacturer's protocol in the Roche Lightcycler 480 system. Theresults were normalized to beta-actin (ACTB) and further analysedaccording to the 2-ΔΔCt method or taking into account the amplificationefficiency (Livak and Schmittgen, 2001; Pfaffl, 2001, 2010). Student's t-test was used for statistical analysis and the results were further con-trolled for multiple testing by applying the Benjamini-Hochberg correc-tion; false discovery rate (FDR) (Benjamini and Hochberg, 1995).

Protein isolation and Western Blotting

Protein was extracted in RIPA buffer (150mMNaCl, 1% (v/v) Triton-X 100, 0.5% (w/v) sodium deoxycholate, 0.1% (v/v) SDS, 50 mM Tris–


ED P

RO

OF

HCl, pH 8.0) with protease (Roche) and phosphatase inhibitors(Sigma) on ice. The tissue was sonicated every 30 seconds for 10 sec-onds and the lysates were centrifuged at 14,000 rpm for 30 minutes at4 °C. The supernatant was subjected to protein quantification usingthe BCA assay (Pierce Biotechnology), and after denaturation, proteinswere loaded on precast 10% Bis-Tris SDS-PAGE gels (Invitrogen) andtransferred to a 0.2 μm nitrocellulose membrane (Whatman). Themembranes were blocked for 1 hour in TBS (10 mM Tris–HCl, pH 7.6,150 mM NaCl,) containing 0.1%Tween-20 and either 1% BSA (w/v) or5% (w/v) dry skimmed milk and probed with the indicated primary an-tibodies andwith either HRP-conjugated secondary antibodies followedby enhanced chemiluminescence reagents (PerkinElmer Life Sciences)or fluorescent-conjugated antibodies. Antibodies against BACE1(D10E5 Cell Signaling, 1:2,000), ATXN1 (Cell Signaling, 1:1,000),ARPC3 (Abcam, 1:1,000), HIP1 (Novus, 1:1,000), CALNEXIN (BD Biosci-ence, 1:500), ARRB1 (BD Bioscience, 1:250), KCNC3 (Invitrogen,1:1,000), GAPDH (HyTest, 1:10,000) and ACTB (AC15, Sigma, 1:5,000)were used. Secondary antibodies were goat-anti-mouse and anti-rabbit polyclonal antibodies (Biorad, 1:5,000), rat anti-goat polyclonalantibody (Dako, 1:2,000), goat anti-mouse-AlexaFluor680 (Invitrogen,1:15,000) and goat anti-rabbit-IR800 (Rockland, 1:5,000). Proteinswere visualized using the Image Reader LAS-4000-mini digital detector(Fujifilm) or the Odyssey Infrared system and quantified by the AIDAsoftware. Equal loadingwas evaluated byusingACTBor GAPDHprimaryantibody as internal control.

Morphological analysis of the cerebellum

Immunostaining on floating sections was performed as previouslydescribed (Duvick et al., 2010) with slight modifications. Animalswere perfused with 4% formaldehyde solution and brains were isolatedand post-fixed for an additional 2 hours at RT, and eitherwashed in PBS,embedded in 4% low-melting point agarose in PBS in a sagittal orienta-tion and cut at 50 μm thickness on a vibratome (Leica VT1000S) or cryo-preserved in 18% sucrose at 4 °C until sunk to the bottomof the tube andcut on a sliding microtome at 50 μm thickness for stereological pur-poses. After unmasking the epitopes and blocking the sections at RTfor 1 hour in 2% normal goat serum (Dako), 0.3% Triton X-100 (Sigma)in PBS, floating sections were incubated with the primary antibody(rabbit calbindin (E300, Swant Laboratories 1:500), mouse vGLUT1(Chemicon, 1:500) and mouse vGLUT2 (MAB 5504, Millipore, 1:500))for 48–60 hours at 4 °C and frozen sections for 1 hour at RT. Sectionswere extensively washed in PBS and further incubated for 1–2 hoursat RT with appropriate secondary antibodies (goat-anti-mouse or anti-rabbit-Alexa-Fluor-555 (Invitrogen, 1:500), goat-anti-mouse or anti-rabbit-Alexa-Fluor-488 (Invitrogen 1:500) or sheep-anti-mouse IgG/HRP-conjugated (Roche 1:200) followed by PBS washes. For stereolog-ical purposes, DAB was used according to the manufacturer's protocol(Invitrogen). Floating sections were transferred to Superfrost Plus(Thermo Scientifc) slides. All sections were mounted in Prolong Goldcontaining DAPI (Invitrogen). The thickness of molecular layer of thecerebellum, defined as the distance between the external border ofthe molecular layer and the apex of the Purkinje cells, was measured.The vGLUT1 and vGLUT2 immunoreactive boutons in the Purkinjecells of the cerebellar cortex were counted in the primary fissure, andtheir size areawasmeasured using ImageJ software; using particle anal-ysis plugin, after using watershed plugin in binary images. All imageswere treated under the same conditions. Stereological analysis was con-ducted using the StereoInvestigator Software. Images were acquired ona Nikon A1R Ti Eclipse confocal microscope using Plan Apo VC 20x(0.75) or Plan Fuor 60x (1.40 oil) objectives. All measurements wereperformed using the ImageJ software.

Golgi stainingThe Golgi impregnation was performed with the FD Rapid Golgi

stain kit (FD Neurotechnologies) according to the manufacturer's



308

309

310

311

312

313

314

315

316

317

318

319

320

321

322

323

324

325

326

327

328

329

330

331

332

333

334

335

336

337

338

339

340

341

342

343

344

345

346

347

348

349

350

351

352

353

354

355

356

357

358

359

360

361

362

363

364

365

366

367

368

369

370

371

372

373


protocol, with slightmodifications. After anaesthetizing themice, brainswere isolated, rinsed and treated according to themanufacturer's proto-col. Sections were cut on a vibratome (Leica VT1000S) at 100–150 μmthickness and after further treatment as recommended, they weretransferred on Superfrost Plus slides (Thermo Scientific) and mountedwith Mowiol. Soma diameter of Purkinje cells, Sholl analysis (Todorovet al., 2012) and spine density on tertiary dendritic branches70–90 μm away from the soma (Kusnoor et al., 2010) were measured.Images were taken on a Leica TCS SP5 II microscope using HCX PL Apo63x (1.4 oil) objective. All analysis was performed using the ImageJsoftware.

374

375

376

377

378

379

380

381

382

383

384

Statistics

Statistics were performed using one, two or three-way ANOVAwithrepeated measures when applicable and as independent factors geno-type and/or age and/or gender and/or side, or ANCOVA with weight asa covariate. Statistical analysis was corrected for multiple comparisonsusing the Bonferroni method. Student's t-test was used when appropri-ate. All statistical analysis was performed in the SPSS software orGraphPad Prism software.

T385

386

387

388

389

390

391

392

393

394

395

396

397

398

399

400

401

402

403

404

405

406

407

408

409

410

ORREC

Results

miR-29 expression in the brain

miR-29a/b-1 is highly expressed in the central nervous system,however the relative expression in different regions remains poorlycharacterized.We performed FISH (Fig. 1) using LNA and 2′-O-meth-yl- modified oligonucleotides against the mature miRNAs (Aronicaet al., 2010; Silahtaroglu et al., 2007). miR-29a (Fig. 1A) and miR-29b (Fig. 1B) showed widespread neuronal expression in the brain.To control for the specificity of the probes, miR-29a/b-1 knockoutsections (Papadopoulou et al., 2012) were used in parallel. Therewas also no signal when a scrambled miR sequence was used(Fig. 1C). Furthermore, miR-219 (Fig. 2D) was expressed only by ol-igodendrocytes, as previously described (Dugas et al., 2010), furtherconfirming the specificity of our technique (Fig. 1D). These datashow that miR-29 is strongly expressed in the glomeruli as well asthe mitral layer of the olfactory bulb, the cortex, all the CA1-CA4 re-gions of the hippocampus and the dentate gyrus. miR-29 is alsoexpressed in the cerebellum with a high signal in the Purkinje celllayer. There is lower expression in the granule and molecular celllayers (Fig. 1). The residual miR-29b signals observed in the olfactorybulb, cortex, hippocampus and cerebellum likely account for themiR-29b expressed from the miR-29b-2/c cluster.
C 411
412

413

414

415

416

417

418

419

420

421

422

423

424

425

UNGeneral observations on miR-29a/b-1 knockout mice

Previously, we have generated a full miR-29a/b-1 cluster knockout(KO) mousemodel (Papadopoulou et al., 2012). Thesemice showMen-delian inheritance, but display a progressive disorder ultimatelyresulting in death at about 7±2 months of age (data not shown). Knock-out mice weigh less than their wildtype (WT) littermate controls (fe-males: 3mo: p = 0.001; 6mo: p b 0.001males: 3mo: p = 1; 6mo:p b 0.001), and the difference between the genotypes becomes largeras they age, since the KOanimals do not regain their bodyweight (geno-type x age, F1, 106= 15.613, p b 0.0010) (Fig. 2A). Themice show sever-al neurological disease features. For instance when suspended by thetail, they show progressive flexor contraction (Fig. 2B). Moreover, themiR-29a/b-1 KO mice have kyphosis which becomes more pronouncedat an older age (data not shown). They also display several motoric fea-tures, which are further analyzed quantitatively below.


ED P

RO

OF

miR-29a/b-1 knockout mice show impaired motor coordination andimbalance

We tested the mice using the balance beam, the DigiGait™ and theaccelerating rotating rod assays. The balance beam test is a sensitivetest of balance and coordination, where mice were trained to cross ele-vatedwooden beams to reach an escape platform.miR-29a/b-1 KOmicedisplayed coordination impairments, as measured by time needed tocross different shapes of the beam (difficulty x genotype: squares,F1.97, 94.57 = 6.313, p = 0.003; circles, F1.148, 55.096 = 13.651,p b 0.0001) (Fig. 3A). In addition, the miR-29a/b-1 KO mice slippedmore often than the WT littermates with increasing beam difficulty(difficulty x genotype, squares, F1.044, 50.133 = 6.8, p = 0.011; circles,F1.40, 67.286 = 32.925, p b 0.0001) (Fig. 3B). Aged animals performedworse (time: difficulty x age: squares, F1.97, 94.57 = 4.533, p = 0.014;circles, F1.148, 55.096 = 7.601, p = 0.006; slips: difficulty x age:squares, F1.044, 50.133 = 4.242, p = 0.043; circles, F1.40, 67.286 = 11.928,p b 0.0001), but more so the miR-29a/b-1 KO (time: difficulty xgenotype x age: squares F1.97, 94.57 = 5.318, p = 0.007; circles, F1.148,55.096 = 4.047, p = 0.044; slips: difficulty x genotype x age: squaresF1.044, 50.133 = 4.649, p = 0.034; circles, F1.40, 67.286 = 6.935, p =0.005). At the highest degree of difficulty, i.e. the narrowest circle, the6 month old knockout animals were unable to balance on the beam[Supplementary material videos 1 (6 months old WT mouse) and 2(6 months old miR-29a/b-1 KO mouse)].

Abnormal gait is a hallmark of ataxia and often associated with cer-ebellar dysfunction.We evaluated the gait performance of miR-29a/b-1KO andWTmice by recordingwalking behaviour on a transparentmov-ing runaway (DigiGait™) at a constant speed (16 cm/s). miR-29a/b-1KO mice took smaller steps (reduced stride length), (genotype,frontlimbs, F1,44 = 17.448, p b 0.0001; hindlimbs, F1,44 = 31.043,p b 0.0001) but their stride frequency was significantly increased(Fig. 3C), (genotype, frontlimbs, F1,44 = 17.448, p b 0.0001; hindlimbs,F1,44 = 32.722, p b 0.0001). Furthermore, the miR-29a/b-1 KO micewalk with a significantly splayed angle between the hind paws andthe body axis (genotype, hind limbs, F1,44 = 66.217, p b 0.0001)(Figs. 3D, E). Gait disturbances were not affected by age, suggestingthat gait performance of the miR-29a/b-1 knockout mice was alreadyimpaired as early as 3 months of age.

The motor impairment of miR-29a/b-1 knockout mice was furtherconfirmed by the accelerating rotarod, where mice were tested fortheir ability to stay on it. The miR-29a/b-1 KO mice fell off quickerthan the WT littermates (genotype, F1,56 = 17.998, p b 0.0001). Agingworsened their performance, (age, F1,56 = 8.248, p = 0.006), butthis was independent of the genotype (genotype x age, F1,56 = 0.030,p = 0.863). Furthermore, both miR-29a/b-1 KO and WT mice learnedto stay longer on the rod over the trials (Fig. 3F). In conclusion, these be-havioral data confirmed deficits in the motor skills of the miR-29a/b-1KO mice resulting in an ataxic phenotype.

miR-29a/b-1 knockout mice show altered cerebellar morphology andpathologic features

In order to explain the disturbed locomotion phenotype of themice,we compared themuscles, spinal cords and Purkinje cells of the cerebel-lar cortex. miR-29 loss has been previously involved inmuscle degener-ation and dystrophy (Wang et al., 2011, 2012). However, skeletalmuscles of the hind limbs of 6 months old miR-29a/b-1 KO andWT an-imals were not altered (Figs. 4A, B).

Furthermore, miR-29 is expressed in the spinal cord (Bak et al.,2008; Hohjoh and Fukushima, 2007). We analyzed therefore 6 monthsold spinal cords of miR-29a/b-1 KO and WT animals and did not findmajor abnormalities (Figs. 4C–E). We finally turned to the cerebellum,where we showed that miR-29 is highly expressed in the Purkinjecells (Fig. 1).



UNCO

RRECTED P

RO

OF

426

427

428

429

430

431

432

433

434

435

436

437

438

439

440

441

WT/29a/DAPIOlfactory bulb Cortex Hippocampus Cerebellum

KO/29a/DAPI

KO/29b/DAPI

WT/29b/DAPI

WT DAPI

A

B

C /Scr/DAPI

WT/219/DAPI

GL

MCL

HIP

DG

ML

PCLGCLGCL

Fig. 1.miR-29 expression inmouse brain tissue. A, B. miR-29 is expressed in themitral cell layer as well as in the glomeruli of the olfactory bulb. There is clear staining in all CA regions ofthe hippocampus and in the dentate gyrus. Fluorescent staining is also present in the cortex. The cerebellum shows strong staining in the Purkinje cell layer aswell as some staining in thegranule cell layer. The knockout ofmiR-29a/b-1 brain tissuewas used as negative control. C. A scrambled (SCR) sequence is used as a negative control. The oligodendrocyticmiR-219 usedto control for signal specificity in all brain areas. miRNAs and scrambled sequence is depicted in green and DAPI in blue. GL: glomeruli; MCL: mitral cell mayer; HIP: hippocampus; DG:dentate gyrus; ML: Molecular layer; GCL: Granule cell layer; PCL: Purkinje cell layer. Scale bar at 10 μm. Background staining for miRNA29b in the miR29a/b ko is probably caused by re-sidual activity of the 29b/c cluster.


The altered motor behavior of the miR-29a/b-1 KO mice and theataxic phenotype is compatible with altered cerebellar function. Histo-logical examination of the cerebellum did not reveal any gross abnor-malities (Figs. 5A, B). The molecular layer thickness of the cerebellarcortex of the miR-29a/b-1 KO mice was decreased at 6 months com-pared to WT (genotype: F1, 9 = 5.834, p = 0.039) which may be asign of atrophy of the dendrites of the Purkinje cells (Figs. 5C, D). Wetherefore analyzed the Purkinje cells of the cerebellum in miR-29a/b-1


KO mice and WT littermates at 7 months of age by stereologicalmethods. miR-29a/b-1 KO mice demonstrated a trend towards a mildreduction of the Purkinje cell population by ~20%, but this did notreach significance (F1, 8 = 1.411, p = 0.269) (Fig. 5E). Therefore theloss of Purkinje cells is likely not accounting for the ataxic phenotypeof the miR-29a/b-1 KO mice. We also examined the morphological in-tegrity of the two major excitatory inputs to the Purkinje cells i.e. theparallel and the climbing fibers. The vGLUT1 immunoreactivity was



T

OF

442

443

444

445

446

447

448

449

450

451

452

453

454

455

456

457

458

459

460

461

462

463

464

465

466

467

468

469

470

471

472

473

474

475

476

477

478

479

480

481

482

483

484

485

486

487

488

489

490

491

492

493

494

495

A B WT KO

Fig. 2.General observations ofmiR-29a/b-1 knockoutmice. A. Bodyweight of wildtype andmiR-29a/b-1 knockoutmice (mean±SEM) at 3 months (KOmice: females: n=14;males: n=6;WTmice: females: n=21;males: n= 13) and 6 months (KOmice: females: n= 14;males: n=12;WTmice: females: n=18;males: n= 12) of age. Theweight of KOmice is overalldecreased (p b 0.0001) in comparison to theWT littermates aswell as that of females vsmales (p b 0.0001). Groupswere compared using threeway ANOVA, and Bonferonni correction aspost-hoc analysis. * represents pairwise comparisonswith p b 0.05, p b 0.01 and p b 0.001 for one, two or three symbols respectively. Three-way ANOVA. B. Representative pictures of aWTand a KO mouse of 6 months old, when hung by the tail. miR-29a/b-1 KO mice display hind limb clasping in contrast to the age-matched WT, which shows a normal plantar posture.


significantly decreased in the molecular layer of the miR-29a/b-1-defi-cient cerebellum, indicating that miR-29a/b-1 KO Purkinje cells receiveless innervation by the parallel fibers (F1, 7 = 5.755, p = 0.048)(Fig. 6A). The average bouton size was also slightly but significantly in-creased (F1, 7 = 6.298, p = 0.040) (Fig. 6B). Representative pictures ofthe vGLUT1 immunoreactivity are given in Fig. 6C. In contrast, vGLUT2immunoreactivity in the molecular layer of the cerebellum of the miR-29a/b-1 KO mice showed no significant difference when compared tothat of the WT littermates (F1, 8 = 2.513, p = 0.152) (Fig. 6D). Alsothe size of the vGLUT2 boutons was not different between the miR-29a/b-1 KO and WT controls (F1, 7 = 0.046, p = 0.836) (Fig. 6E).Based on these resultswe conclude that themiR-29a/b-1 KOcerebellumshows different characteristics versus the control, promoting a role ofmiR-29a/b-1 cluster in the maintenance of Purkinje cells.
C 496
497

498

499

500

501

502

503

504

505

506

507

508

509

510

511

512

513

514

515

516

517

518

519

520

521

522

523

524

525

526

UNCO

RREMorphology of Purkinje cells is altered in the miR-29a/b-1 knockout mice

The lower vGLUT1 immunoreactivity in themiR-29a/b-1 KOanimalssuggested a decreased number of synapses formed between parallelfibers and Purkinje cells. A lower number of synapses can result eitherfrom a decreased spine density in the dendrites or by less dendriticarborisation of the Purkinje cells. Therefore, miR-29a/b-1 KO andWT cerebella were analyzed by Golgi impregnation (Fig. 7A). Wemeasured the spine density at 70-90 μm from the soma, on the distalparts of the dendrites,where the Purkinje cells synapsewith the parallelfibers. At 7 months of age, we also did not find any difference in thespine density between WT and miR-29a/b-1 KO Purkinje cells (geno-type, F1, 4 = 0.113, p = 0.754) (Fig. 7B). However, Sholl analysis ofthe intersections of the dendritic tree of Purkinje cells indicated loss ofarborisation in the KO mice with respect to the WT animals as the dis-tance from the soma increases (distance x genotype x age, F2.54,25.4 = 3.457, p = 0.037) (Fig. 7D). At 10 μm away from the soma, the7 month oldmiR-29a/b-1 deficient Purkinje cells show less arborisation(p= 0.008). Themost important finding of the Sholl quantificationwasthat themiR-29a/b-1 KOPurkinje cells were characterized by decreasedcomplexity of their dendritic arbors over age, which did not occur inthe WT cells, especially at 80 (p = 0.034) and 100 μm (p = 0.029)away from the cell body. At 90 μmwe did not get any significant differ-ence in arborisation of the KO cells over age, (p= 0.070). The diameterof the miR-29a/b-1 KO Purkinje cells was decreased (genotype: F1, 9 =35.256, p b 0.0001). Aging had an effect (age: F1, 9 = 107.172,p b 0.0001), and the difference became larger between the two geno-types (genotype x age, F1, 9 = 5.256, p = 0.048) (Fig.7C). All together,the data suggest indeed that there are subtle structural abnormalities


ED P

RO

of the Purkinje cells in the cerebella of the miR-29a/b-1 KO mice com-pared to the WT controls.

Bioinformatic identification and experimental validation of miR-29a/b-1targets potentially related to the ataxic phenotype

Two individual bioinformatic approaches were used to identify rele-vant miR-29 targets. In a first approach, we identified potential miR-29targets using themicro-RNA body map (Mestdagh et al., 2011), which isa repository of all available bioinformatic algorithms used to identifygenes possibly targeted by miRNAs. We filtered these genes with acut-off of being identified by at least 4 different algorithms as miR-29potential targets. The following criteria were used to further filterthe obtained candidate targets (1). Relevance to nervous system-associated diseases (DAVID Bioinformatics) (Huang da et al., 2007,2009), (2). Applicability to the top 100 scoring targets in theMicrocosm(Griffiths-Jones et al., 2008) and (3). Expression in the cerebellum ac-cording to the HPA (Human Protein Atlas) (Ponten et al., 2008; Uhlenet al., 2010) (Fig. 8A). In the second approach we used the IngenuityPathway Analysis (IPA) tool to identify all the genes related to ataxia.We then retained only those that were predicted to be targets ofmouse miR-29 by Targetscanmouse 6.2 (Grimson et al., 2007) or PicTaralgorithms (Krek et al., 2005).

We identified the following genes as of interest for further evalua-tion: Dnmt3a, Ifi30, Hbp1, Atxn1, Plp1, Slc16a2, Hip1, Bace1, Ireb2, Pitpna,Plp1, Scn3b, Ank3, Atxn1, CamkIV, Dicer1, Ercc6, Kcnc3, Slc1a2 and Canx.We also included additional genes that were previously suggested tobe targets of miR-29 and/or could play a role in neuronal function, i.e.Bace1 (Hebert et al., 2008), Arpc3 (Lippi et al., 2011), Bax (Kole et al.,2011), Dnmt3a (Fabbri et al., 2007; Garzon et al., 2009; Pandi et al.,2013; Takada et al., 2009), Rora (Gold et al., 2003; Zoghbi and Orr,2009) and Arrb1. Arrb2was used as a negative control as it is not a pre-dicted miR-29 target (DeWire et al., 2007; Shi et al., 2007). These geneswere validated for changes at the mRNA expression level by quantita-tive RT-PCR analysis at ~3 months of age, as at that time point themice display behavioural deficits already.

Thirteen out of the 22 shortlisted genes showed a significant in-crease at the mRNA level in the miR-29a/b-1 KO cerebella, after correc-tion formultiple testing (Thissen et al., 2002) (Fig. 8A), whichwere Ifi30,Kcnc3, Dnmt3a, Gsta4, Ireb2, Arrb1, Camk4, Hbp1, Bax, Rora, Pitpna, Dicer1and Scl1a2. The rest of the genes tested showed no significant up-regu-lation at themRNA level in themiR-29a/b-1 KO cerebella, namely Bace1,Ank3, Ercc6, Arpc3, Hip1, Atxn1, Canx, Slc16a2, Plp1 and Scn3b. The candi-date genes were subjected to further validation at the protein levelsbased on antibody availability, as miRs mostly but not exclusively affect



UNCO

RRECTED P

RO

OF

527

528

529

530

531

532

533

534

535

536

537

538

539

540

541

WTEKO

C D

F

BA

WT KO

Fig. 3.miR-29a/b-1 knockoutmice show alteredmotor behavior. A, B. Balance beam test involved a series of beams of decreasing diameters (1 N 2 N 3). Small diameter beamsweremoredifficult to cross. The beams differ in diameter and with decreasing diameter (1 N 2 N 3) there is increased difficulty. A. Latency of mice to cross the cylindrical beams. Cut-off time to tra-verse is 60 seconds. Overall,WTmice cross the beam faster than KOmice (p b 0.001) and oldermice are slower than young ones (p b 0.0001). SeveralmiR-29a/b-1 KOmicewere not ableto finish the task or fell off the beam, particularly on the most difficult beam #3. WT mice traversed within the time limit (60s) of the test (p b 000.1). B. Number of slips (and/or falls)during the beam test. KO mice showed more slips than the WT littermates (p b 0.001) and older mice showed an increased number of slips (or falls) in comparison to younger mice(p b 0.001). The smaller the diameter, the more difficulties the miR-29a/b-1 KO mice had to cross and slipped or fell more often than WT (p b 000.1). Maximal number of slips formice that failed to cross was set at 100 (n = 9 WT; n = 10 KO 3 months old; n = 20 WT; n = 13 KO mice of 6 months old). C. D. E. Gait performance on a moving belt (DigiGait™).C. The stride frequency for both fore- and hindlimbs of KO mice was increased in relation to the age-matched KO littermates (p b 000.1). D. In WT mice, the absolute paw angle of thefore limbs decreased with age (p = 0.011). In contrast, in KO the absolute paw angle of the fore limb was smaller but did not change over time. Hind paw angle of KO mice was largerthan in WT, but was not affected by age (p b 0.0001). Combination of these features can be the result of gait incoordination (n = 12 WT; n = 14 KO mice of 3 months old; n = 12WT; n = 10 KO mice of 6 months old). E. Photographic representation of the gait performance of a WT and a KO mouse of 6 months old. (blue: left front paw; pink: right front paw;red: left hindpaw; green: right hindpaw). F. Balancing on the rotarod. At 3 months of age, KOanimals initially fell off the rotating rodmuch faster, but eventually reachedWTperformance.In contrast, at 6 months of age, both genotypes needed consecutive trials to improve their performance (and never reached the 3 month levels) (p= 0.006), but KO performed overallworse thanWT (p b 0.0001). (n= 14WT; n=17 KOmice of 3 months old; n= 14WT; n= 15 KOmice of 6 months old). Data are expressed as means± SEMs. Groups were comparedusing two-way repeated measures ANOVA, and Bonferroni correction as post-hoc analysis. * represents pairwise comparisons with p b 0.05, p b 0.01 and p b 0.001 for one, two or threesymbols respectively.


themRNA levels of their targets genes (Guo et al., 2010). Representativeblots are shown in Fig. 8B. Kcnc3 and Hip1 proteins are significantly up-regulated in the miR-29a/b-1 deficient cerebella of 3 months old micein comparison to theWT littermates. On the other hand, Bace1, in agree-ment with Roshan et al. (Roshan et al., 2014), as well as Atxn1, Arrb1,Arpc3 and Clnx proteins did not reach significance (Fig. 8C). Thus,only Kcnc3 was consistently shown up-regulated relative to the WTlevels at both the mRNA (56%) and protein levels (70%) (Figs. 8B, C).


Discussion

Mapping the expression of miR-29 family members in the brain re-veals that miR-29 is mainly in neuron occupied regions. Anotherin vivo study confirms that miR-29 is mainly expressed in neurons (Liet al., 2014). These observations contrast with previous work demon-strating expression of miR-29 in astrocytes (Hu et al., 2012; Ouyanget al., 2013; Smirnova et al., 2005), and microglia (Fenn et al., 2013).



ORRECTED P

RO

OF

542

543

544

545

546

547

548

549

550

551

552

553

554

555

556

557

558

559

560

561

562

563

564

565

566

567

568

569

570

571

572

573

574

575

A

C D EWT

KO

WT

KO

WT

KO

BWT KO

Fig. 4.Nomajor abnormalities inmuscle and spinal cord of the miR-29a/b-1 knockout mice. A. Representative picture of hematoxylin and eosin (H&E) staining of skeletal muscle. (n= 7WT; n= 9 KO animals of 7 months old). Scale bars: upper panels:100 μm; lower panels:50 μm. B. Morphometrical analysis of the hind limbmuscles shows no difference in diameter be-tween themiR-29a/b-1 KOandWT littermates (n=7WT; n=9KOanimals of 7 months old). C. D. Representative pictures of hematoxylin and eosin (H&E) and Luxol fast blue staining ofthe spinal cords show nomajor differences between the genotypes at 6 months of age. C. Cervical region of the spinal cord. D. Sacral region of the spinal cord. E. Representative picture ofGFAP staining of the cervical spinal cord show no difference in astrogliosis. C. D. E. n= 4/genotype of 6 months old. Dorsal horn is depictedwith “D”, ventral horn is depictedwith “V” andspinocerebellar tracts are shown with asterisks (*). Scale bars at 200 μm.


UNC

Obviously in vitro versus in vivo, and stress conditions versus normalconditions can explain some of the discrepancies in this regard.

In the present studywe find that the genetic inactivation of themiR-29a/b-1 cluster in mice results in a progressive neurological disorder,one of themain features being ataxia. Ataxia is lack of muscle coordina-tion and usually a consequence of neurological damage. Themain struc-ture of the brain responsible for proper coordination and balance is thecerebellum (Ito, 1982; Marr, 1969; Thach and Bastian, 2004). We showthat miR-29 is highly expressed in the Purkinje cell layer of the cerebel-lum. The Purkinje cells provide the sole output of neuronal signals fromthe cerebellum. We found several behavioral alterations that were pre-viously observed in other models for cerebellar ataxia (Becker et al.,2009; Duvick et al., 2010; Perkins et al., 2010; Sausbier et al., 2004). Al-though we obviously cannot exclude the possibility that alterations inother organs or brain areas contribute to the ataxic phenotype observedin these animals, morphological analyses of gastrocnemiusmuscles andof brains and spinal cords of miR-29a/b-1 knockout animals did not


show gross abnormalities. This, together with the abundant expressionof miR-29a/b-1 in the Purkinje cells prompted us to analyse more indepth the effects of miR-29 ablation on those cells.

Although a recent study showed an overall increase in programmedcell death of miR-29 down-regulated neurons, as measured by TUNELassay (Roshan et al., 2014), we used stereological analyses and foundthat the number of Purkinje cells was not significantly affected in themiR-29a/b-1 knockout mice. Our data do not definitively rule out thatcell loss contributed to the phenotype, aswe noticed a strong variabilityin the number of Purkinje cells over the individual knockout animals.This variability could be due to patterned Purkinje cell loss, as hasbeen suggested for cerebellar ataxia mice (Baader et al., 1998; Sachset al., 2009; Sarna and Hawkes, 2011; Sawada et al., 2009) or differentdegrees of penetrance of the phenotype, as shown for a number ofother neurological diseases, such as Huntington's disease (McNeilet al., 1997), Parkinson's disease (Kruger, 2008) and Spinocerebellarataxia 8 (Ikeda et al., 2004).



UNCO

RRECTED P

RO

OF

576

577

578

579

580

581

582

583

A

B

WT

WT

KO

KO

C WT/CB

ML

GL

PCL

D E

KO/CB

Fig. 5. Cerebellar morphological alterations in miR-29a/b-1 knockout animals. A, B. Representative pictures of the histological examination of the cerebellum (n=/genotype of 6 monthsold). Cerebellar cortex is abbreviated as “CC”, cerebellar nuclei as “CN” and arbor vitae as “av”. Scale bars at 400 μm.A.Hematoxylin and eosin combinedwith Luxol fast blue staining revealsnomajor abnormalities in the architecture of the cerebellumor themyelin between the twogenotypes. B.GFAP staining shows nodifferences betweenWTandmiR-29a/b-1KOanimals. C.Representative picture of calbindin immunostaining of the primary fissure ofWT and KO animals of 6 months old. Themolecular layer is indicated as “ML”, the Purkinje cell layer as “PCL”and the granule cell layer as “GL”. Scale bar at 100 μm.D. Molecular layer (ML) thickness of WT and KOmice of 3 and 6 months old. Six measurements of each section are made and theaverage is representative of the individual animal (n= 3/genotype, total of 14 (WT) and 12 (KO) sections of 3 months oldmice; n= 3/genotype, total of 14 (WT) and 17 (KO) sections of6 months oldmice,with no less than three sections/animal). Overall the knockoutmicehave a thinnermolecular layer than theWT(p= 0.024). There is a general increase of the thicknessin aged animals (p= 0.031). Groupswere compared using two-way ANOVA, and Bonferroni correction as post-hoc analysis. * represents pairwise comparisonswith p b 0.05, p b 0.01 andp b 0.001 for one, two or three symbols respectively. E. Stereological analysis of 7 months oldmice. There is a tendency ofmild decrease in the Purkinje cell population in the KO animals,however insignificant (n = 5/genotype). Groups were compared using one-way ANOVA.


The ataxia in the miR-29a/b-1 background might also reflectperturbed function of the cerebellum (Matsushita et al., 2002; Walteret al., 2006). Purkinje cells are the major players in the main cerebellarcircuitry. The parallel fibers (PF) from the granule cells and the climbing


fibers (CF) from the inferior olive synapse to Purkinje cells and excitethem. The Purkinje cells send information for balance and coordinationto the deep cerebellar nuclei (DCN)which project their axons to the ap-propriate centers (i.e. motor cortex). miR-29a/b-1-deficient Purkinje



ECTED P

RO

OF

584

585

586

587

588

589

590

591

592

593

594

595

596

597

598

599

600

601

602

603

604

605

606

607

608

609

610

611

612

613

614

615

616

617

618

619

620

621

622

623

624

625

626

627

628

629

630

631

632

633

634

635

C WT/CB/vGLUT1/DAPI KO/CB/vGLUT1/DAPI

A

D

B

E

Fig. 6. The cerebellar circuitry is altered in the molecular layer of the miR-29a/b-1 knockout cerebellum. The estimated number of vGLUT1 (A) and vGLUT2 (E) varicosities per mm2

peridendritic area of the Purkinje cells at 6 months of age and the average size of the vGLUT1 (B) and vGLUT2 (F) bouton is measured (mice used for vGLUT1 measurements: n = 4WT and n = 5 KOmice, with no less than 3 areas/animal; for vGLUT2: n = 5WT and n = 5 KOmice, with no less than 3 areas/animal). A. There is a small reduction in the vGLUT1 im-munoreactivity of the KOmice when compared to theWT littermates (p= 0.048). B. The average size of the vGLUT1 boutons is significantly larger in the miR-29a/b-1 KO animals (p=0.040). C. Representative pictures of the molecular layer of the WT and miR-29a/b-1 KO cerebellum. Calbindin (CB) is represented in red, vGLUT1 in green and DAPI in blue. Scale bar at20 μm. D. The estimated number of vGLUT2 varicosities per mm2 dendrite of Purkinje cells at 6 months of age. No changes are observed (p= 0.152). E. The boutons of vGLUT2 immuno-reactive boutons have similar average size between the miR-29a/b-1 KO and WT animals (p = 0.040).


UNCO

RRcells show a decreased dendritic arborisation with age but not a de-

crease in spine density. Less arborisation of the Purkinje cells has beennoticed in other models of ataxia (Galliano et al., 2013; Todorov et al.,2012). Interestingly, and in line with a defect in excitatory input onthe Purkinje cells we found a slight but significant drop in the numberof boutons and consequently synapses between granule and Purkinjecells. We also found an increase in the size of the boutons in the miR-29a/b-1 knockout cerebellum (Daniels et al., 2004; Herzog et al., 2006).

Bioinformatic tools were used to identify possible target genes ofmiR-29, which could explain the ataxia phenotype and the Purkinjecell atrophy (Bargaje et al., 2012). We prioritised 22 genes for furtheranalysis, based on what is known about their expression, their relationto ataxia, and the probability that they represent true targets of miR-29. Recent data suggest that an important number of miRNAs controlthe expression of their targets by lowering their mRNA level (Baeket al., 2008; Guo et al., 2010; Selbach et al., 2008). Suitable antibodiesare not always available to analyse protein level alterations. We there-fore set out to check which genes of the list showed alterations inmRNA expression. We found that the following candidate target genesdisplayed increased mRNA expression in the cerebellum of the knock-out animals: Hbp1, Ifi30, Ireb2, Gsta4, Rora, Pitpna, Bax, Arrb1, Kcnc3,Dicer1, Camk4, Scl1a2 and Dnmt3a. However, of the few genes tested atthe protein level only Kcnc3 andHip1 showed significant up-regulation.The reason for some discrepancy between mRNA and protein levels isunknown for now but can involve different transcript stability or pro-tein turnovermechanisms, and thus does not entirely exclude identified


candidates as “genuine”miR-29a/b targets in vivo. Global transcriptomeand proteome analyses will now be required to evaluate the precisenumber of affected genes (and proteins) in the cerebellum as well asin other brain regions expressing miR-29a/b. Furthermore, a potentialcompensation of miR-29c in this model cannot be ruled out, and willneed further investigation using for instance miR-29a/b/c triple knock-out mice.

Hip1 protein level is elevated in ourmiR-29a/b-1 knockout cerebella.HIP1 is important in clathrin-mediated endocytosis and AMPA receptortrafficking in the hippocampus (Metzler et al., 2003) as well as NMDAreceptor function (Metzler et al., 2007). HIP1 deficiency in neurons hasshown partial protection against NMDA-dependent excitotoxicity. Onecould speculate that Hip1 up-regulation will lead to overactivation ofNMDAR and increased excitotoxicity, which could fit our observationof increased size of vGLUT1 immunoreactive boutons. However, Hip1mRNA level was not affected (failed to reach significance) and the pro-tein level shows only a modest up regulation. In contrast Kcnc3(Kv3.3) is significantly up-regulated both at the mRNA and proteinlevels. KCNC3 encodes for the voltage gated potassium channel subunitKv3.3, important for fast spiking of the Purkinje cells and associatedwithspinocerebellar ataxia 13 (Espinosa et al., 2008; Figueroa et al., 2010,2011; Joho and Hurlock, 2009; Waters et al., 2006). Mutations ofKCNC3 can render the channel either less active or cause altered channelkinetics leading to longer opening of the pore (Figueroa et al., 2010,2011; Waters et al., 2006). It is unclear whether an increase in expres-sion would be sufficient to cause similar problems, but one can



UNCO

RRECTED P

RO

OF

636

637

638

639

640

641

642

643

644

645

646

647

648

649

650

651

652

653

654

655

A

B

KOWT

C

D

Fig. 7. Atrophic Purkinje cells in miR-29a/b-1 knockout mice. A. Representative pictures of Golgi staining of WT and miR-29a/b-1 KO mice of 7 months old. Scale bar at 20 μm. B. Spinedensity of 7 month old mice reveals no differences between theWT and the miR-29a/b-1 KO animals (n= 3/genotype total of 17 (WT) and 23 (KO) cells, with no less than 5 PC/animaland no less than 5 dendritic segments/PC). C. The soma diameter of KO animals is smaller than that of the WT littermates (p b 0.0001) At 7 months of age both genotypes have smallersomata (p b 0.0001), but KOmice have a significantly smaller soma than the age-matchedWT (p= 0.001) (n=3WT and 4 KO animals of 3 months old,with a total of 37 cells/genotype;n = 3 WT and 4 KO animals of 6 months old, with a total of 23 WT and 29 KO cells). Groups were compared using two-way ANOVA, and Bonferroni correction as post-hoc analysis. *represents pairwise comparisons with p b 0.05, p b 0.01 and p b 0.001 for one, two or three symbols respectively. D. Sholl analysis of the dendritic trees of the Purkinje cells in WT andKO animals of 3 (n = 3 WT and 4 KO animals of 3 months old, with a total of 14 cells/genotype; n = 3 WT and 4 KO animals of 7 months old, with a total of 14 WT and 28 KO cells).There is a reduction in the number of intersections at a distance of 10 μm (p = 0.008). Interestingly, at 80-100 μm away from the soma the dendritic trees of the Purkinje cells of theKO animals show a significant reduction in arborisation, in contrast to the wildtype cells, over age (p= 0.032 (80 μm); p= 0.025 (90 μm); p= 0.018 (100 μm)). Groups were comparedusing two-way repeated measures ANOVA, and Bonferroni correction as post-hoc analysis. * represents pairwise comparisons with p b 0.05, p b 0.01 and p b 0.001 for one, two or threesymbols respectively.


hypothesize that in case of up-regulated levels of Kcnc3mRNA and pro-tein, more channels will be formed leading to increased activity of thechannel. In support to this hypothesis, previous studies have shownthat overexpression of genes can lead to neurodegenerative disorders,i.e. duplication of APP (Rovelet-Lecrux et al., 2006) in Alzheimer's dis-ease and PMP22 in Charcot Marie Tooth Disease Type 1A (CMT1A)(Roa et al., 1996; Warner et al., 1996).

In this study, we have provided evidence for the role of miR-29 infine-tuningmotor function,which is in linewith a recently published in-dependent study (Roshanet al., 2014). Further,wehavemorphologically


documented the possible involvement of the miR-29a/b-1 cluster in theproper Purkinje cell synaptic communication. We have analysed severalcandidate genes that could be regulated by miR-29a/b-1 and their al-tered expression could have implications in cerebellar functions, eitherexplaining the ataxia component and/or the Purkinje cell atrophy. Lossof miR-29 and subsequent dysregulation of a number of these genescould result in perturbed cerebellar functions that will eventually leadto ataxia.

Supplementary data to this article can be found online at http://dx.doi.org/10.1016/j.nbd.2014.10.006.





UNCO

RRECTED P

RO

OF

656

657

658

659

660

661

662

663

664

665

666

667

668

669

670

671672

A

B

HIP1ACTB

KO WT

ARPC3ACTB

BACE-1ACTB

ATXN1ACTB

KCNC3ACTB

CANXACTB

ARRB1GAPDH

C

Fig. 8. Changes in miR-29a/b-1 knockout cerebella of candidate target genes. A. Fold-change in the mRNA levels in the cerebellum of 3 months old mice of the selected candidate genespotentially regulated by the miR-29a/b-1 cluster. Values are normalized to mRNA levels of ACTB and further normalized to the mean of the wildtype levels, which is set as 1 (n= 6/ge-notype). The mean ± SEM is reported. p values are corrected for multiple testing (FDR) (*p b 0.05; **p b 0.01;***p b 0.001). B. Representative western blot of KO and WT age-matchedlittermates of 4 months old for some of the candidate genes. ACTB or GAPDH is used as loading control. C. Graphical representation of western blot quantification shows fold changesat the protein level of the candidate genes tested. Values are normalized to the intensity of ACTB or GAPDH and further normalized to the mean of theWT values, set as 1 (n = 4–5/ge-notype). The mean ± SEM is reported (*p b 0.05; **p b 0.01;***p b 0.001).


Acknowledgements

Wewould like to thank Veronique Hendrickx, Jonas Verwaeren, andGreet Marien for mouse colony support; Anouk Roberfroid for genotyp-ing help; Zhiyong Zhang andKathleen Craessaerts for ES cells injections;Dr. Pietro Fazzari, Dr. Annette Gärtner, Dr. Diego Sepulveda-Falla, andDr. Annerieke Sierksma for discussions; the KULeuven Inframouse facil-ity for themouse generation and theUKEMouse pathology Core-Facilityfor technical help. This work was supported by the Fund for ScientificResearch Flanders (FWO); research fund KU Leuven; the Hercules


Foundation, Federal Office for Scientific Affairs (IAP P7/16); aMethusalem grant of the Flemish Government, VIB, IWT, the EuropeanResearch Council (ERC AdG to BDS), the Queen Elisabeth Foundation,Stichting Alzheimer Onderzoek (SAO). BDS is the Arthur Bax and AnnaVanluffelen chair for Alzheimer's disease.

References

Aldaz, B., Sagardoy, A., Nogueira, L., Guruceaga, E., Grande, L., Huse, J.T., Aznar, M.A., Diez-Valle, R., Tejada-Solis, S., Alonso, M.M., Fernandez-Luna, J.L., Martinez-Climent, J.A.,



T

673674675676677678679680681682683684685686687688689690691692693694695696697698699700701702703704705706707708709710711712713714715716717718719720721722723724725726727728729730731732733734735736737738739740741742743744745746747748749750751752753754755756757758

759760761762763764765766767768769770771772773774775776777778779780781782783784785786787788789790791792793794795796797798799800801802803804805806807808Q3809810811812813814815816817818819820821822823824825826827828829830831832833834835836837838839840841842843844


UNCO

RREC

Malumbres, R., 2013. Involvement of miRNAs in the differentiation of human glio-blastoma multiforme stem-like cells. PLoS One 8, e77098.

Aronica, E., Fluiter, K., Iyer, A., Zurolo, E., Vreijling, J., van Vliet, E.A., Baayen, J.C.,Gorter, J.A., 2010. Expression pattern of miR-146a, an inflammation-associatedmicroRNA, in experimental and human temporal lobe epilepsy. Eur. J. Neurosci.31, 1100–1107.

Baader, S.L., Sanlioglu, S., Berrebi, A.S., Parker-Thornburg, J., Oberdick, J., 1998. Ectopicoverexpression of engrailed-2 in cerebellar Purkinje cells causes restricted cell lossand retarded external germinal layer development at lobule junctions. J. Neurosci.18, 1763–1773.

Baek, D., Villen, J., Shin, C., Camargo, F.D., Gygi, S.P., Bartel, D.P., 2008. The impact ofmicroRNAs on protein output. Nature 455, 64–71.

Bak, M., Silahtaroglu, A., Moller, M., Christensen, M., Rath, M.F., Skryabin, B., Tommerup, N.,Kauppinen, S., 2008.MicroRNA expression in the adult mouse central nervous system.RNA 14, 432–444.

Bargaje, R., Gupta, S., Sarkeshik, A., Park, R., Xu, T., Sarkar, M., Halimani, M., Roy, S.S., Yates,J., Pillai, B., 2012. Identification of novel targets for miR-29a using miRNA proteomics.PLoS One 7, e43243.

Becker, E.B., Oliver, P.L., Glitsch, M.D., Banks, G.T., Achilli, F., Hardy, A., Nolan, P.M., Fisher,E.M., Davies, K.E., 2009. A point mutation in TRPC3 causes abnormal Purkinje cell de-velopment and cerebellar ataxia in moonwalker mice. Proc. Natl. Acad. Sci. U. S. A.106, 6706–6711.

Benjamini, Y., Hochberg, Y., 1995. Controlling the false discovery rate: a practical andpowerful approach to multiple testing. J. R. Stat. Soc. Ser. B 289–300.

Callaerts-Vegh, Z., Beckers, T., Ball, S.M., Baeyens, F., Callaerts, P.F., Cryan, J.F., Molnar, E.,D'Hooge, R., 2006. Concomitant deficits in working memory and fear extinction arefunctionally dissociated from reduced anxiety in metabotropic glutamate receptor7-deficient mice. J. Neurosci. 26, 6573–6582.

Carter, R.J., Lione, L.A., Humby, T., Mangiarini, L., Mahal, A., Bates, G.P., Dunnett, S.B.,Morton, A.J., 1999. Characterization of progressive motor deficits in mice transgenicfor the human Huntington's disease mutation. J. Neurosci. 19, 3248–3257.

Chang, T.C., Yu, D., Lee, Y.S., Wentzel, E.A., Arking, D.E., West, K.M., Dang, C.V., Thomas-Tikhonenko, A., Mendell, J.T., 2008. Widespread microRNA repression by Myc con-tributes to tumorigenesis. Nat. Genet. 40, 43–50.

Daniels, R.W., Collins, C.A., Gelfand, M.V., Dant, J., Brooks, E.S., Krantz, D.E., DiAntonio, A.,2004. Increased expression of the Drosophila vesicular glutamate transporter leadsto excess glutamate release and a compensatory decrease in quantal content. J.Neurosci. 24, 10466–10474.

DeWire, S.M., Ahn, S., Lefkowitz, R.J., Shenoy, S.K., 2007. Beta-arrestins and cell signaling.Annu. Rev. Physiol. 69, 483–510.

Dugas, J.C., Cuellar, T.L., Scholze, A., Ason, B., Ibrahim, A., Emery, B., Zamanian, J.L., Foo, L.C.,McManus, M.T., Barres, B.A., 2010. Dicer1 and miR-219 Are required for normal oligo-dendrocyte differentiation and myelination. Neuron 65, 597–611.

Duvick, L., Barnes, J., Ebner, B., Agrawal, S., Andresen, M., Lim, J., Giesler, G.J., Zoghbi, H.Y.,Orr, H.T., 2010. SCA1-like disease in mice expressing wild-type ataxin-1 with a serineto aspartic acid replacement at residue 776. Neuron 67, 929–935.

Espinosa, F., Torres-Vega, M.A., Marks, G.A., Joho, R.H., 2008. Ablation of Kv3.1 and Kv3.3potassium channels disrupts thalamocortical oscillations in vitro and in vivo. J.Neurosci. 28, 5570–5581.

Fabbri, M., Garzon, R., Cimmino, A., Liu, Z., Zanesi, N., Callegari, E., Liu, S., Alder, H.,Costinean, S., Fernandez-Cymering, C., Volinia, S., Guler, G., Morrison, C.D., Chan,K.K., Marcucci, G., Calin, G.A., Huebner, K., Croce, C.M., 2007. MicroRNA-29 family re-verts aberrant methylation in lung cancer by targeting DNA methyltransferases 3Aand 3B. Proc. Natl. Acad. Sci. U. S. A. 104, 15805–15810.

Fenn, A.M., Smith, K.M., Lovett-Racke, A.E., Guerau-de-Arellano, M., Whitacre, C.C.,Godbout, J.P., 2013. Increasedmicro-RNA 29b in the aged brain correlateswith the re-duction of insulin-like growth factor-1 and fractalkine ligand. Neurobiol. Aging 34,2748–2758.

Figueroa, K.P., Minassian, N.A., Stevanin, G.,Waters, M., Garibyan, V., Forlani, S., Strzelczyk,A., Burk, K., Brice, A., Durr, A., Papazian, D.M., Pulst, S.M., 2010. KCNC3: phenotype,mutations, channel biophysics-a study of 260 familial ataxia patients. Hum. Mutat.31, 191–196.

Figueroa, K.P., Waters, M.F., Garibyan, V., Bird, T.D., Gomez, C.M., Ranum, L.P., Minassian,N.A., Papazian, D.M., Pulst, S.M., 2011. Frequency of KCNC3 DNA variants as causesof spinocerebellar ataxia 13 (SCA13). PLoS One 6, e17811.

Galliano, E., Gao, Z., Schonewille, M., Todorov, B., Simons, E., Pop, A.S., D'Angelo, E., vanden Maagdenberg, A.M., Hoebeek, F.E., De Zeeuw, C.I., 2013. Silencing the majorityof cerebellar granule cells uncovers their essential role in motor learning and consol-idation. Cell Rep. 3, 1239–1251.

Garzon, R., Liu, S., Fabbri, M., Liu, Z., Heaphy, C.E., Callegari, E., Schwind, S., Pang, J., Yu, J.,Muthusamy, N., Havelange, V., Volinia, S., Blum, W., Rush, L.J., Perrotti, D., Andreeff,M., Bloomfield, C.D., Byrd, J.C., Chan, K., Wu, L.C., Croce, C.M., Marcucci, G., 2009.MicroRNA-29b induces global DNA hypomethylation and tumor suppressor genereexpression in acute myeloid leukemia by targeting directly DNMT3A and 3B andindirectly DNMT1. Blood 113, 6411–6418.

Gold, D.A., Baek, S.H., Schork, N.J., Rose, D.W., Larsen, D.D., Sachs, B.D., Rosenfeld, M.G.,Hamilton, B.A., 2003. RORalpha coordinates reciprocal signaling in cerebellar devel-opment through sonic hedgehog and calcium-dependent pathways. Neuron 40,1119–1131.

Griffiths-Jones, S., Saini, H.K., van Dongen, S., Enright, A.J., 2008. miRBase: tools formicroRNA genomics. Nucleic Acids Res. 36, D154–D158.

Grimson, A., Farh, K.K., Johnston, W.K., Garrett-Engele, P., Lim, L.P., Bartel, D.P., 2007.MicroRNA targeting specificity in mammals: determinants beyond seed pairing.Mol. Cell 27, 91–105.

Guo, H., Ingolia, N.T., Weissman, J.S., Bartel, D.P., 2010. Mammalian microRNAs predomi-nantly act to decrease target mRNA levels. Nature 466, 835–840.


ED P

RO

OF

Hebert, S.S., Horre, K., Nicolai, L., Papadopoulou, A.S., Mandemakers,W., Silahtaroglu, A.N.,Kauppinen, S., Delacourte, A., De Strooper, B., 2008. Loss of microRNA cluster miR-29a/b-1 in sporadic Alzheimer's disease correlates with increased BACE1/beta-secretase expression. Proc. Natl. Acad. Sci. U. S. A. 105, 6415–6420.

Herzog, E., Takamori, S., Jahn, R., Brose, N., Wojcik, S.M., 2006. Synaptic and vesicular co-localization of the glutamate transporters VGLUT1 and VGLUT2 in the mouse hippo-campus. J. Neurochem. 99, 1011–1018.

Hohjoh, H., Fukushima, T., 2007. Expression profile analysis of microRNA (miRNA) inmouse central nervous system using a new miRNA detection system that examineshybridization signals at every step of washing. Gene 391, 39–44.

Hu, G., Yao, H., Chaudhuri, A.D., Duan, M., Yelamanchili, S.V., Wen, H., Cheney, P.D., Fox,H.S., Buch, S., 2012. Exosome-mediated shuttling of microRNA-29 regulates HIV Tatand morphine-mediated neuronal dysfunction. Cell Death Dis. 3, e381.

Huang da, W., Sherman, B.T., Tan, Q., Kir, J., Liu, D., Bryant, D., Guo, Y., Stephens, R., Baseler,M.W., Lane, H.C., Lempicki, R.A., 2007. DAVID Bioinformatics Resources: expandedannotation database and novel algorithms to better extract biology from large genelists. Nucleic Acids Res. 35, W169–W175.

Huang da, W., Sherman, B.T., Lempicki, R.A., 2009. Systematic and integrative analysis oflarge gene lists using DAVID bioinformatics resources. Nat. Protoc. 4, 44–57.

Hurlock, E.C., McMahon, A., Joho, R.H., 2008. Purkinje-cell-restricted restoration of Kv3.3function restores complex spikes and rescues motor coordination in Kcnc3 mutants.J. Neurosci. 28, 4640–4648.

Hurlock, E.C., Bose, M., Pierce, G., Joho, R.H., 2009. Rescue of motor coordination byPurkinje cell-targeted restoration of Kv3.3 channels in Kcnc3-null mice requiresKcnc1. J. Neurosci. 29, 15735–15744.

Hwang, H.W., Wentzel, E.A., Mendell, J.T., 2007. A hexanucleotide element directsmicroRNA nuclear import. Science 315, 97–100.

Ikeda, Y., Dalton, J.C., Moseley, M.L., Gardner, K.L., Bird, T.D., Ashizawa, T., Seltzer, W.K.,Pandolfo, M., Milunsky, A., Potter, N.T., Shoji, M., Vincent, J.B., Day, J.W., Ranum, L.P.,2004. Spinocerebellar ataxia type 8: molecular genetic comparisons and haplotypeanalysis of 37 families with ataxia. Am. J. Hum. Genet. 75, 3–16.

Ito, M., 1982. Experimental verification of Marr-Albus' plasticity assumption for the cere-bellum. Acta Biol. Acad. Sci. Hung. 33, 189–199.

Jiao, J., Herl, L.D., Farese, R.V., Gao, F.B., 2010. MicroRNA-29b regulates the expression levelof human progranulin, a secreted glycoprotein implicated in frontotemporal demen-tia. PLoS One 5, e10551.

Johnson, R., Zuccato, C., Belyaev, N.D., Guest, D.J., Cattaneo, E., Buckley, N.J., 2008. AmicroRNA-based gene dysregulation pathway in Huntington's disease. Neurobiol.Dis. 29, 438–445.

Joho, R.H., Hurlock, E.C., 2009. The role of Kv3-type potassium channels in cerebellarphysiology and behavior. Cerebellum 8, 323–333.

Joho, R.H., Street, C., Matsushita, S., Knopfel, T., 2006. Behavioral motor dysfunction inKv3-type potassium channel-deficient mice. Genes Brain Behav. 5, 472–482.

Kapinas, K., Kessler, C., Ricks, T., Gronowicz, G., Delany, A.M., 2010. miR-29 modulatesWnt signaling in human osteoblasts through a positive feedback loop. J. Biol. Chem.285, 25221–25231.

Khanna, S., Rink, C., Ghoorkhanian, R., Gnyawali, S., Heigel, M., Wijesinghe, D.S., Chalfant,C.E., Chan, Y.C., Banerjee, J., Huang, Y., Roy, S., Sen, C.K., 2013. Loss of miR-29b follow-ing acute ischemic stroke contributes to neural cell death and infarct size. J. Cereb.Blood Flow Metab.

Kole, A.J., Swahari, V., Hammond, S.M., Deshmukh, M., 2011. miR-29b is activated duringneuronal maturation and targets BH3-only genes to restrict apoptosis. Genes Dev. 25,125–130.

Krek, A., Grun, D., Poy, M.N., Wolf, R., Rosenberg, L., Epstein, E.J., MacMenamin, P., daPiedade, I., Gunsalus, K.C., Stoffel, M., Rajewsky, N., 2005. Combinatorial microRNAtarget predictions. Nat. Genet. 37, 495–500.

Kruger, R., 2008. LRRK2 in Parkinson's disease - drawing the curtain of penetrance: a com-mentary. BMC Med. 6, 33.

Kusnoor, S.V., Parris, J., Muly, E.C., Morgan, J.I., Deutch, A.Y., 2010. Extracerebellar role forCerebellin1: modulation of dendritic spine density and synapses in striatal mediumspiny neurons. J. Comp. Neurol. 518, 2525–2537.

Li, H., Mao, S., Wang, H., Zen, K., Zhang, C., Li, L., 2014. MicroRNA-29a modulates axonbranching by targeting doublecortin in primary neurons. Protein Cell 5, 160–169.

Lippi, G., Steinert, J.R., Marczylo, E.L., D'Oro, S., Fiore, R., Forsythe, I.D., Schratt, G., Zoli, M.,Nicotera, P., Young, K.W., 2011. Targeting of the Arpc3 actin nucleation factor bymiR-29a/b regulates dendritic spine morphology. J. Cell Biol. 194, 889–904.

Liu, J., Sisk, J.M., Gama, L., Clements, J.E., Witwer, K.W., 2013. Tristetraprolin expressionand microRNA-mediated regulation during simian immunodeficiency virus infectionof the central nervous system. Mol. Brain 6, 40.

Livak, K.J., Schmittgen, T.D., 2001. Analysis of relative gene expression data using real-time quantitative PCR and the 2(−Delta Delta C(T)) Method. Methods 25, 402–408.

Long, J., Wang, Y., Wang, W., Chang, B.H., Danesh, F.R., 2011. MicroRNA-29c is a signaturemicroRNA under high glucose conditions that targets Sprouty homolog 1, and itsin vivo knockdown prevents progression of diabetic nephropathy. J. Biol. Chem.286, 11837–11848.

Ma, F., Xu, S., Liu, X., Zhang, Q., Xu, X., Liu, M., Hua, M., Li, N., Yao, H., Cao, X., 2011. ThemicroRNA miR-29 controls innate and adaptive immune responses to intracellularbacterial infection by targeting interferon-gamma. Nat. Immunol. 12, 861–869.

Majer, A., Medina, S.J., Niu, Y., Abrenica, B., Manguiat, K.J., Frost, K.L., Philipson, C.S.,Sorensen, D.L., Booth, S.A., 2012. Early mechanisms of pathobiology are revealed bytranscriptional temporal dynamics in hippocampal CA1 neurons of prion infectedmice. PLoS Pathog. 8, e1003002.

Marr, D., 1969. A theory of cerebellar cortex. J. Physiol. 202, 437–470.Matsushita, K., Wakamori, M., Rhyu, I.J., Arii, T., Oda, S., Mori, Y., Imoto, K., 2002. Bidirec-

tional alterations in cerebellar synaptic transmission of tottering and rolling Ca2+channel mutant mice. J. Neurosci. 22, 4388–4398.



T

845846847848849850851852853854855856857858859860861862863864865866867868869870871872873874875876877878879880881882883884885886887888889890891892893894895896897898899900901902903904905906907908909910911912913914915916917918919920921922923924925926927

928929930931932933934935936937938939940941942943944945946947948949950951952953954955956957958959960961962963964965966967968969970971972973974975976977978979980981982983984985986987988989990991992993994995996Q499799899910001001100210031004100510061007


UNCO

RREC

McNeil, S.M., Novelletto, A., Srinidhi, J., Barnes, G., Kornbluth, I., Altherr, M.R., Wasmuth,J.J., Gusella, J.F., MacDonald, M.E., Myers, R.H., 1997. Reduced penetrance of theHuntington's disease mutation. Hum. Mol. Genet. 6, 775–779.

Mestdagh, P., Lefever, S., Pattyn, F., Ridzon, D., Fredlund, E., Fieuw, A., Ongenaert, M.,Vermeulen, J., De Paepe, A., Wong, L., Speleman, F., Chen, C., Vandesompele, J.,2011. The microRNA body map: dissecting microRNA function through integrativegenomics. Nucleic Acids Res. 39, e136.

Metzler, M., Li, B., Gan, L., Georgiou, J., Gutekunst, C.A., Wang, Y., Torre, E., Devon, R.S., Oh,R., Legendre-Guillemin, V., Rich, M., Alvarez, C., Gertsenstein, M., McPherson, P.S.,Nagy, A., Wang, Y.T., Roder, J.C., Raymond, L.A., Hayden, M.R., 2003. Disruption ofthe endocytic protein HIP1 results in neurological deficits and decreased AMPA re-ceptor trafficking. EMBO J. 22, 3254–3266.

Metzler, M., Gan, L., Wong, T.P., Liu, L., Helm, J., Georgiou, J., Wang, Y., Bissada, N., Cheng, K.,Roder, J.C., Wang, Y.T., Hayden, M.R., 2007. NMDA receptor function and NMDAreceptor-dependent phosphorylation of huntingtin is altered by the endocytic proteinHIP1. J. Neurosci. 27, 2298–2308.

Nolan, K., Mitchem, M.R., Jimenez-Mateos, E.M., Henshall, D.C., Concannon, C.G., Prehn,J.H., 2014. Increased expression of microRNA-29a in ALS mice: functional analysisof its inhibition. J. Mol. Neurosci. 53, 231–241.

Nori, S., Okada, Y., Yasuda, A., Tsuji, O., Takahashi, Y., Kobayashi, Y., Fujiyoshi, K., Koike, M.,Uchiyama, Y., Ikeda, E., Toyama, Y., Yamanaka, S., Nakamura, M., Okano, H., 2011.Grafted human-induced pluripotent stem-cell-derived neurospheres promotemotor functional recovery after spinal cord injury in mice. Proc. Natl. Acad. Sci. U. S.A. 108, 16825–16830.

Ott, C.E., Grunhagen, J., Jager, M., Horbelt, D., Schwill, S., Kallenbach, K., Guo, G., Manke, T.,Knaus, P., Mundlos, S., Robinson, P.N., 2011. MicroRNAs differentially expressed inpostnatal aortic development downregulate elastin via 3′ UTR and coding-sequencebinding sites. PLoS One 6, e16250.

Ouyang, Y.B., Xu, L., Lu, Y., Sun, X., Yue, S., Xiong, X.X., Giffard, R.G., 2013. Astrocyte-enriched miR-29a targets PUMA and reduces neuronal vulnerability to forebrain is-chemia. Glia 61, 1784–1794.

Packer, A.N., Xing, Y., Harper, S.Q., Jones, L., Davidson, B.L., 2008. The bifunctionalmicroRNA miR-9/miR-9* regulates REST and CoREST and is downregulated inHuntington's disease. J. Neurosci. 28, 14341–14346.

Pandi, G., Nakka, V.P., Dharap, A., Roopra, A., Vemuganti, R., 2013. MicroRNA miR-29cdown-regulation leading to de-repression of its target DNAmethyltransferase 3a pro-motes ischemic brain damage. PLoS One 8, e58039.

Papadopoulou, A.S., Dooley, J., Linterman,M.A., Pierson,W., Ucar, O., Kyewski, B., Zuklys, S.,Hollander, G.A., Matthys, P., Gray, D.H., De Strooper, B., Liston, A., 2012. The thymic ep-ithelial microRNA network elevates the threshold for infection-associated thymic in-volution via miR-29a mediated suppression of the IFN-alpha receptor. Nat. Immunol.13, 181–187.

Perkins, E.M., Clarkson, Y.L., Sabatier, N., Longhurst, D.M., Millward, C.P., Jack, J., Toraiwa, J.,Watanabe, M., Rothstein, J.D., Lyndon, A.R., Wyllie, D.J., Dutia, M.B., Jackson, M., 2010.Loss of beta-III spectrin leads to Purkinje cell dysfunction recapitulating the behaviorand neuropathology of spinocerebellar ataxia type 5 in humans. J. Neurosci. 30,4857–4867.

Pfaffl, M.W., 2001. A new mathematical model for relative quantification in real-time RT-PCR. Nucleic Acids Res. 29, e45.

Pfaffl, M.W., 2010. The ongoing evolution of qPCR. Methods 50, 215–216.Ponten, F., Jirstrom, K., Uhlen, M., 2008. The Human Protein Atlas–a tool for pathology. J.

Pathol. 216, 387–393.Prophet, E.B., Mills, B., Arrington, J.B., Sobin, L.H. (Eds.), 1992. Armed Forces Institute of

Pathology. Laboratory methods in Histotechnology. American Registry of Pathology,Washington, DC.

Radaelli, E., Ceruti, R., Patton, V., Russo, M., Degrassi, A., Croci, V., Caprera, F., Stortini, G.,Scanziani, E., Pesenti, E., Alzani, R., 2009. Immunohistopathological and neuroimagingcharacterization of murine orthotopic xenograft models of glioblastoma multiformerecapitulating the most salient features of human disease. Histol. Histopathol. 24,879–891.

Roa, B.B., Greenberg, F., Gunaratne, P., Sauer, C.M., Lubinsky, M.S., Kozma, C., Meck, J.M.,Magenis, R.E., Shaffer, L.G., Lupski, J.R., 1996. Duplication of the PMP22 gene in 17ppartial trisomy patients with Charcot-Marie-Tooth type-1 neuropathy. Hum. Genet.97, 642–649.

Roshan, R., Ghosh, T., Gadgil, M., Pillai, B., 2012. Regulation of BACE1 by miR-29a/b in acellular model of Spinocerebellar Ataxia 17. RNA Biol. 9, 891–899.

Roshan, R., Shridhar, S., Sarangdhar, M.A., Banik, A., Chawla, M., Garg, M., Singh, V.P., Pillai, B.,2014. Brain-specific knockdown of miR-29 results in neuronal cell death and ataxia inmice. RNA 20, 1287–1297.

Rovelet-Lecrux, A., Hannequin, D., Raux, G., Le Meur, N., Laquerriere, A., Vital, A.,Dumanchin, C., Feuillette, S., Brice, A., Vercelletto, M., Dubas, F., Frebourg, T.,Campion, D., 2006. APP locus duplication causes autosomal dominant early-onsetAlzheimer disease with cerebral amyloid angiopathy. Nat. Genet. 38, 24–26.

Sachs, A.J., David, S.A., Haider, N.B., Nystuen, A.M., 2009. Patterned neuroprotection in theInpp4a(wbl) mutant mouse cerebellum correlates with the expression of Eaat4. PLoSOne 4, e8270.

Sarna, J.R., Hawkes, R., 2011. Patterned Purkinje cell loss in the ataxic sticky mouse. Eur. J.Neurosci. 34, 79–86.

Sausbier, M., Hu, H., Arntz, C., Feil, S., Kamm, S., Adelsberger, H., Sausbier, U., Sailer, C.A.,Feil, R., Hofmann, F., Korth, M., Shipston, M.J., Knaus, H.G., Wolfer, D.P., Pedroarena,C.M., Storm, J.F., Ruth, P., 2004. Cerebellar ataxia and Purkinje cell dysfunction causedby Ca2 + −activated K + channel deficiency. Proc. Natl. Acad. Sci. U. S. A. 101,9474–9478.


ED P

RO

OF

Sawada, K., Kalam Azad, A., Sakata-Haga, H., Lee, N.S., Jeong, Y.G., Fukui, Y., 2009. Strikingpattern of Purkinje cell loss in cerebellum of an ataxic mutant mouse, tottering. ActaNeurobiol. Exp. (Wars) 69, 138–145.

Selbach, M., Schwanhausser, B., Thierfelder, N., Fang, Z., Khanin, R., Rajewsky, N., 2008.Widespread changes in protein synthesis induced by microRNAs. Nature 455, 58–63.

Shi, Y., Feng, Y., Kang, J., Liu, C., Li, Z., Li, D., Cao, W., Qiu, J., Guo, Z., Bi, E., Zang, L., Lu, C.,Zhang, J.Z., Pei, G., 2007. Critical regulation of CD4+ T cell survival and autoimmunityby beta-arrestin 1. Nat. Immunol. 8, 817–824.

Shioya, M., Obayashi, S., Tabunoki, H., Arima, K., Saito, Y., Ishida, T., Satoh, J., 2010. Aber-rant microRNA expression in the brains of neurodegenerative diseases: miR-29a de-creased in Alzheimer disease brains targets neurone navigator 3. Neuropathol. Appl.Neurobiol. 36, 320–330.

Silahtaroglu, A.N., Nolting, D., Dyrskjot, L., Berezikov, E., Moller, M., Tommerup, N.,Kauppinen, S., 2007. Detection of microRNAs in frozen tissue sections by fluorescencein situ hybridization using locked nucleic acid probes and tyramide signal amplifica-tion. Nat. Protoc. 2, 2520–2528.

Silva, V.A., Polesskaya, A., Sousa, T.A., Correa, V.M., Andre, N.D., Reis, R.I., Kettelhut, I.C.,Harel-Bellan, A., De Lucca, F.L., 2011. Expression and cellular localization ofmicroRNA-29b and RAX, an activator of the RNA-dependent protein kinase (PKR),in the retina of streptozotocin-induced diabetic rats. Mol. Vis. 17, 2228–2240.

Smirnova, L., Grafe, A., Seiler, A., Schumacher, S., Nitsch, R., Wulczyn, F.G., 2005. Regula-tion of miRNA expression during neural cell specification. Eur. J. Neurosci. 21,1469–1477.

Smith, K.M., Guerau-de-Arellano, M., Costinean, S., Williams, J.L., Bottoni, A., Mavrikis Cox,G., Satoskar, A.R., Croce, C.M., Racke, M.K., Lovett-Racke, A.E., Whitacre, C.C., 2012.miR-29ab1 deficiency identifies a negative feedback loop controlling Th1 bias thatis dysregulated in multiple sclerosis. J. Immunol. 189, 1567–1576.

Stroobants, S., Gantois, I., Pooters, T., D'Hooge, R., 2013. Increased gait variability in micewith small cerebellar cortex lesions and normal rotarod performance. Behav. BrainRes. 241, 32–37.

Takada, S., Berezikov, E., Choi, Y.L., Yamashita, Y., Mano, H., 2009. Potential role of miR-29b in modulation of Dnmt3a and Dnmt3b expression in primordial germ cells of fe-male mouse embryos. RNA 15, 1507–1514.

Thach, W.T., Bastian, A.J., 2004. Role of the cerebellum in the control and adaptation ofgait in health and disease. Prog. Brain Res. 143, 353–366.

Thissen, D., Steinberg, L., Kuang, D., 2002. Quick and easy implementation of theBenjamini-Hochberg procedure for controlling the false positive rate in multiplecomparisons. J. Educ. Behav. Stat. 27, 77–83.

Todorov, B., Kros, L., Shyti, R., Plak, P., Haasdijk, E.D., Raike, R.S., Frants, R.R., Hess, E.J.,Hoebeek, F.E., De Zeeuw, C.I., van den Maagdenberg, A.M., 2012. Purkinje cell-specific ablation of Cav2.1 channels is sufficient to cause cerebellar ataxia in mice.Cerebellum 11, 246–258.

Ugalde, A.P., Ramsay, A.J., de la Rosa, J., Varela, I., Marino, G., Cadinanos, J., Lu, J., Freije, J.M.,Lopez-Otin, C., 2011. Aging and chronic DNA damage response activate a regulatorypathway involving miR-29 and p53. EMBO J. 30, 2219–2232.

Uhlen, M., Oksvold, P., Fagerberg, L., Lundberg, E., Jonasson, K., Forsberg, M., Zwahlen, M.,Kampf, C., Wester, K., Hober, S., Wernerus, H., Bjorling, L., Ponten, F., 2010. Towards aknowledge-based Human Protein Atlas. Nat. Biotechnol. 28, 1248–1250.

van Rooij, E., Sutherland, L.B., Thatcher, J.E., DiMaio, J.M., Naseem, R.H., Marshall, W.S., Hill,J.A., Olson, E.N., 2008. Dysregulation of microRNAs after myocardial infarction revealsa role of miR-29 in cardiac fibrosis. Proc. Natl. Acad. Sci. U. S. A. 105, 13027–13032.

Walter, J.T., Alvina, K., Womack, M.D., Chevez, C., Khodakhah, K., 2006. Decreases in theprecision of Purkinje cell pacemaking cause cerebellar dysfunction and ataxia. Nat.Neurosci. 9, 389–397.

Wang, X.H., Hu, Z., Klein, J.D., Zhang, L., Fang, F., Mitch, W.E., 2011. DecreasedmiR-29 sup-presses myogenesis in CKD. J. Am. Soc. Nephrol. 22, 2068–2076.

Wang, L., Zhou, L., Jiang, P., Lu, L., Chen, X., Lan, H., Guttridge, D.C., Sun, H., Wang, H., 2012.Loss of miR-29 in myoblasts contributes to dystrophic muscle pathogenesis. Mol.Ther. 20, 1222–1233.

Warner, L.E., Roa, B.B., Lupski, J.R., 1996. Absence of PMP22 coding region mutations inCMT1A duplication patients: further evidence supporting gene dosage as a mecha-nism for Charcot-Marie-Tooth disease type 1A. Hum. Mutat. 8, 362–365.

Waters, M.F., Minassian, N.A., Stevanin, G., Figueroa, K.P., Bannister, J.P., Nolte, D., Mock,A.F., Evidente, V.G., Fee, D.B., Muller, U., Durr, A., Brice, A., Papazian, D.M., Pulst,S.M., 2006. Mutations in voltage-gated potassium channel KCNC3 cause degenerativeand developmental central nervous system phenotypes. Nat. Genet. 38, 447–451.

Xiao, J., Meng, X.M., Huang, X.R., Chung, A.C., Feng, Y.L., Hui, D.S., Yu, C.M., Sung, J.J., Lan,H.Y., 2012. miR-29 Inhibits Bleomycin-induced Pulmonary Fibrosis in Mice. Mol.Ther.

Yang, Y., Zhi, F., He, X., Moore, M.L., Kang, X., Chao, D., Wang, R., Kim, D.H., Xia, Y., 2012.delta-opioid receptor activation and microRNA expression of the rat cortex in hypox-ia. PLoS One 7, e51524.

Zhang, Q., Liu, H., McGee, J., Walsh, E.J., Soukup, G.A., He, D.Z., 2013. IdentifyingmicroRNAsinvolved in degeneration of the organ of corti during age-related hearing loss. PLoSOne 8, e62786.

Zhao, D., Jiang, X., Yao, C., Zhang, L., Liu, H., Xia, H., Wang, Y., 2014. Heat shock protein 47regulated by miR-29a to enhance glioma tumor growth and invasion. J. Neurooncol.118, 39–47.

Zoghbi, H.Y., Orr, H.T., 2009. Pathogenic mechanisms of a polyglutamine-mediated neuro-degenerative disease, spinocerebellar ataxia type 1. J. Biol. Chem. 284, 7425–7429.



Documents

Deficiency of the miR-29a/b-1 cluster leads to ataxic features and cerebellar alterations in mice