RESEARCH ARTICLE

In vivo cerebellar circuit function is disrupted in an mdx mousemodel of Duchenne muscular dystrophyTrace L. Stay1,2,3,*, Lauren N. Miterko1,3,4, Marife Arancillo3, Tao Lin3 and Roy V. Sillitoe1,2,3,4,‡

ABSTRACTDuchenne muscular dystrophy (DMD) is a debilitating and ultimatelylethal disease involving progressive muscle degeneration andneurological dysfunction. DMD is caused by mutations in thedystrophin gene, which result in extremely low or total loss ofdystrophin protein expression. In the brain, dystrophin is heavilylocalized to cerebellar Purkinje cells, which control motor and non-motor functions. In vitro experiments in mouse Purkinje cells revealedthat loss of dystrophin leads to low firing rates and high spikingvariability. However, it is still unclear how the loss of dystrophin affectscerebellar function in the intact brain. Here, we used in vivoelectrophysiology to record Purkinje cells and cerebellar nuclearneurons in awake and anesthetized female mdx (also known asDmd) mice. Purkinje cell simple spike firing rate is significantly lower inmdxmice compared to controls. Although simple spike firing regularityis not affected, complex spike regularity is increased in mdx mutants.Mean firing rate in cerebellar nuclear neurons is not altered inmdxmice,but their local firing pattern is irregular. Based on the relatively well-preservedcytoarchitecture in themdx cerebellum, our datasuggest thatfaulty signals across the circuit between Purkinje cells and cerebellarnuclei drive the abnormal firing activity. The in vivo requirements ofdystrophin during cerebellar circuit communication could help explainthe motor and cognitive anomalies seen in individuals with DMD.

This article has anassociated First Person interviewwith the first authorof the paper.

KEY WORDS: Duchenne muscular dystrophy, mdx mice,Cerebellum, Purkinje cell, Cerebellar nuclei, Circuitry, In vivoelectrophysiology

INTRODUCTIONDuchenne muscular dystrophy (DMD) is a devastating X-linkeddisease that affects ∼1 in 5000 boys (Guiraud et al., 2015). DMD iscaused by mutations in the dystrophin gene (DMD) (Burghes et al.,

1987; Hoffman et al., 1987; Monaco et al., 1986). The mutationseliminate the expression of the 427 kDa protein dystrophin, or lowerit to less than 5% of normal. DMD mutations also cause the milderdisease, Becker muscular dystrophy, as well as X-linked dilatedcardiomyopathy. Interestingly, although heterozygous femalecarriers of DMD mutations are typically asymptomatic, up to ∼8%of these carriers are considered as manifesting carriers, who developsymptoms ranging from mild muscle weakness to a rapidlyprogressive DMD-like muscular dystrophy (Birnkrant et al., 2018;Moser and Emery, 1974; Norman and Harper, 1989; Taylor et al.,2007). Female carriers have also been reported to have cognitiveabnormalities (Imbornoni et al., 2014; Mercier et al., 2013; Papaet al., 2016). Dystrophin functions as a tether for stabilizing proteincomplexes, and, in the brain, it also interacts with membrane proteinsthat mediate neuronal communication (Pilgram et al., 2010; Waiteet al., 2009). Accordingly, loss of dystrophin can impair brainfunction (Anderson et al., 2002; Mirski and Crawford, 2014; Pereirada Silva et al., 2018; Snowet al., 2014). Cognition andmovement areoften affected, although the neural bases of these behavioral defectsare unclear. In this study, we sought to gain a deeper understandingof how neuronal signals are altered in the DMD brain. Towards this,we used an mdx (also known as Dmd) mouse model to test how theloss of dystrophin (Dp427 isoform) alters cerebellar function bymeasuring neuronal activity in vivo (Grady et al., 2006; Ryder-Cooket al., 1988; Sicinski et al., 1989; Sillitoe et al., 2003).

Dystrophin protein complexes are heavily expressed in thecerebellum, where they are localized predominantly to Purkinjecells (Blake et al., 1999; Lidov et al., 1990, 1993; Sillitoe et al.,2003). Purkinje cells are the principal cell type of the cerebellumand the computational center for executing all cerebellar-dependentbehaviors (Reeber et al., 2013). In mice, the loss of dystrophindramatically alters Purkinje cell microcircuit organization (Sillitoeet al., 2003). Such structural alterations are consistent with theabnormal behaviors in mdx mutant mice, including uncoordinatedmovement (Grady et al., 2006). These molecular and behavioraldefects are also consistent with defects in neuronal activity. In vitroelectrophysiology experiments demonstrated that dissociatedPurkinje cell firing activity is compromised in mdx mice (Snowet al., 2014). They found that Purkinje cells from themdxmice firedmore irregularly than those from control mice and that themembrane potential was hyperpolarized (Snow et al., 2014). Theauthors also reported a lower-than-normal Purkinje cell firingfrequency in dissociated mdx Purkinje cells. Their results areconsistent with other reports that showed a reduction in the numberof GABA synapses on Purkinje cells (Kueh et al., 2011), aberrantGABA release and uptake in cerebellar synaptosomes (Pereira daSilva et al., 2018), and a reduction in postsynaptic long-termdepression (LTD) in mdx Purkinje cells [although Sesay et al.(1996) found no LTP changes in mdx urethane-anesthetizedhippocampal cells, which could be related to the anesthetic effectsof urethane (Hara and Harris, 2002)]. Interestingly, homosynapticReceived 10 June 2019; Accepted 30 October 2019

1Department of Pathology and Immunology, Baylor College of Medicine, Houston,TX 77030, USA. 2Department of Neuroscience, Baylor College of Medicine,Houston, TX 77030, USA. 3Jan and Dan Duncan Neurological Research Institute ofTexas Children’s Hospital, 1250 Moursund Street, Suite 1325, Houston, TX 77030,USA. 4Program in Developmental Biology, Baylor College of Medicine, Houston,TX 77030, USA.*Present address: Department of Neurobiology, Stanford University, Stanford,CA 94305, USA.

‡Author for correspondence (sillitoe@bcm.edu)

T.L.S., 0000-0001-9849-8924; L.N.M., 0000-0001-9229-1834; M.A., 0000-0003-0367-6180; R.V.S., 0000-0002-6177-6190

This is an Open Access article distributed under the terms of the Creative Commons AttributionLicense (https://creativecommons.org/licenses/by/4.0), which permits unrestricted use,distribution and reproduction in any medium provided that the original work is properly attributed.

1

© 2019. Published by The Company of Biologists Ltd | Disease Models & Mechanisms (2020) 13, dmm040840. doi:10.1242/dmm.040840

Disea

seModels&Mechan

isms

LTD at parallel fiber–Purkinje cell synapses is enhanced in mdxmutant mice (Anderson et al., 2010), but how altering these intrinsicmembrane properties of Purkinje cells affects circuit function in theintact cerebellum in vivo remains largely unsolved.The cerebellum controls a variety of motor behaviors, including

coordination, learning, balance and posture (Reeber et al., 2013;Thach, 2014). It may also play a pivotal role in non-motor functionssuch as cognition, language, emotion, reward, social interactionsand spatial working memory (Buckner, 2013; Carta et al., 2019;D’Angelo and Casali, 2013; McAfee et al., 2019; Stoodley, 2012).The execution of all these behaviors requires proper Purkinje cellfunction (Heiney et al., 2014; Tsai et al., 2012; White et al., 2014).Importantly, motor and non-motor Purkinje cell functions could berelevant to DMD (Anderson et al., 2002). In either case, thecanonical Purkinje cell circuit is likely involved. Purkinje cellsreceive direct excitatory input from climbing fibers and granule cellparallel fiber axons, and inhibitory inputs from stellate and basketcell interneurons. Purkinje cells are the only output of the cerebellarcortex, and inhibit neurons in the cerebellar and vestibular nuclei.The cerebellar nuclei are located in the inner core of the cerebellumand provide the main output of the cerebellum.We hypothesize that,in DMD, Purkinje cells fire abnormally and, as a consequence,behavioral output is negatively impacted. In this study, we addresshow the loss of dystrophin affects Purkinje cell function andspecifically assess how Purkinje cell firing activity is affectedin vivo in mice (Arancillo et al., 2015; White et al., 2014).We performed in vivo extracellular recordings in both

anesthetized and awake adult control and mdx mutant mice. Themain objective of this study was to investigate whether functionalchanges occur in vivo in Purkinje cells and in their downstreamtarget cerebellar nuclear cells when dystrophin is absent. Ourexperiments revealed three main findings: (1) Purkinje cell simplespike firing rate is significantly lower in mdx mutant mice than incontrols; (2) the pattern of firing inmdx cerebellar nuclear neurons ismore irregular than that in controls; and (3) the overall firing featuresof the mdx Purkinje cells are highly reminiscent of the defectsobserved in several mutant mouse strains that model cerebellardisease, including ataxia and dystonia (Fremont et al., 2014; Gaoet al., 2012; White and Sillitoe, 2017; White et al., 2014, 2016a).Altogether, our findings provide a functional correlate to the well-studied anatomical and molecular pathogenesis of DMD (Guiraudand Davies, 2019), and provide a valuable set of cell activity data forevaluating exactly how different brain functions might be affectedin vivo in the different muscular dystrophies.

RESULTSCircuit architecture is unaltered in mdx miceMouse models of disease often exhibit changes in the anatomy ofcerebellar microcircuits (Cendelin, 2014; Lalonde and Strazielle,2007; Landis and Landis, 1978), which could affect the normalproperties of cerebellar function (Barmack and Yakhnitsa, 2008;Davie et al., 2008; De Zeeuw et al., 2011; Ruigrok et al., 2011;Schmolesky et al., 2002). To address whether this is the case inC57BL/10ScSn-Dmdmdx/J homozygous mutant mice (hereafterreferred to as mdx mice), we examined the regional localizationand layered tri-laminar distribution of cerebellar cells in mdx miceby immunostaining with a panel of cell-specific molecular markers(Fig. 1). As expected from this mouse model, we confirmed that themdx Purkinje cells lack the dystrophin protein (Dp427; Fig. 1A).Wefirst examined the inhibitory neurons of the cerebellar cortex usingmarkers that identify each cell type (Reeber and Sillitoe, 2011;White et al., 2014, 2016b). Using CAR8 staining (Miterko et al.,

2019a; White et al., 2016a), we found that mutant Purkinje cellswere arranged into the characteristic monolayer and exhibited theirsignature, wide-spanning dendritic architecture, as also observed inC57BL/10ScSnJ controls (hereafter referred to as controls; Fig. 1B).Overall inhibitory interneuron distribution was also similar betweengenotypes when revealed with parvalbumin staining (Fig. 1C).Basket cell projections were identified in the mdx mice, as seenthrough staining for NFH (also known as Nefh) (Fig. 1D).Additionally, the overall distribution of Golgi cells in the granulecell layer of mdx mice was equivalent to that in controls, as shownby neurogranin staining (Fig. 1E). These data suggest that inhibitoryneurons in the cerebellar cortex of mdx mice have similardorsoventral patterning and regional distributions to those in thesame region in control mice.

We next examined the excitatory neurons and the excitatoryafferent inputs to the cerebellar cortex through immunohistochemistry(White et al., 2014, 2016b). Adult differentiated granule cellswere stained with GABAaRα6; they were correctly accumulatedin the granule cell layer in mdx mice (Fig. 1F). Unipolar brushcells are a specialized type of neuron localized predominantly tolobules IX and X (Diño et al., 1999), with a secondary site oflocalization in lobules VI and VII (Sillitoe et al., 2003). In mdxmice, the unipolar brush cells were heavily immunoreactive forcalretinin, and, as in control mice, they were located mainly inlobules IX and X (Fig. 1G). We used VGLUT1 (also known asSlc17a7) expression to demonstrate that the terminals of granulecell projections, found on their parallel fibers, were represented intheir typical abundance in the molecular layer in mdx mice(Fig. 1H). Mossy fiber inputs to the granular layer and climbingfiber inputs to the molecular layer were clear in mdx mice, withboth classes of terminals revealed by VGLUT2 (also known asSlc17a6) (Fig. 1I). The climbing fibers were also visualized withCART (encoded by the Cartpt gene) (Fig. 1J), which showed thenormal climbing trajectory of axons in the molecular layer of mdxmice. Overall, the normal cellular layering, cell distribution,neuronal patterning and afferent terminal field localization in thecerebellar circuit of mdx mice argues against any majordevelopmental morphogenetic rearrangements as a majorcontributor to the neurological deficits observed in the mdx mice.

Lack of overt neurodegeneration in mdx miceCerebellar size and molecular layer thickness can also be altered indisease models, with variations from the normal range indicativeof improper development and function (Hansen et al., 2013).These two measures are also useful anatomical readouts ofneurodegeneration and cell loss. If Purkinje cells succumb todegeneration, regression of their large dendritic trees, the mostnotable entity of the molecular layer, results in a significant decreasein the size of the cerebellar cortex (Miterko et al., 2019a). Towardsthis, we analyzed cerebellar size and molecular layer thickness inmdx and control mice to provide an overall impression of cerebellararchitectural integrity. The distance we measured was from theapical edge of the Purkinje cell soma to the pial boundary ofthe molecular layer region directly above (Brown et al., 2019;White et al., 2014, 2016a). Measurements from three mice ofeach genotype did not reveal a significant difference in totalcerebellar size or molecular layer thickness [Fig. 2; cerebellar size:control=30.1×106±2.8×106 μm2, mdx=29.2×106±1.9×106 μm2,t(4)=0.280, P=0.79; molecular layer thickness: control=152.3±3.7 μm, mdx=155.7±4.0 μm, t(4)=−0.623, P=0.57]. A moredetailed analysis of Purkinje cell numbers, cerebellar nuclear areasand cerebellar nuclear densities additionally confirmed that overt

2

RESEARCH ARTICLE Disease Models & Mechanisms (2020) 13, dmm040840. doi:10.1242/dmm.040840

Disea

seModels&Mechan

isms

neurodegeneration does not occur in mdx mice at the ages examined(Fig. S1). The number of Purkinje cells in mdx mice does not differfrom that in controls when counted from all vermis lobules or whencounts are taken from specific zones [Fig. S1A-D; control=569.2±32.7 (whole), 256.0±12.9 (anterior), 169.4±7.4 (central), 143.8±14.3(posterior/nodular), mdx=578.7±6.7 (whole), 260.6±3.6 (anterior),181.4±5.1 (central), 136.7±2.3 (posterior/nodular); t(4)=−0.286,P=0.79, t(4)=−0.340, P=0.75, t(4)=−1.345, P=0.25, t(4)=0.489,P=0.65, respectively]. Likewise, the areas and densities of thecerebellar nuclei in the mdx mice are comparable to those of controlmice [Fig. S1E-J; areas: control=1242.5±154.9 μm2 (fastigial, FN),4051.9±385.4 μm2 (interposed, IN), mdx=1483.5±287.7 μm2 (FN),3015.8±567.9 μm2 (IN); densities (number of nuclei/50 μm):control=17.36±1.61 (FN), 36.11±3.12 (IN), mdx=21.26±1.20(FN), 33.19±1.14 (IN); t(4)=−0.737, P=0.50, t(4)=1.510, P=0.21,t(4)=−1.9646, P=0.12, t(4)=0.880, P=0.43, respectively]. These datasuggest that the main cellular components of the cerebellar circuitmaintain their anatomical integrity in the mdx mice.

In vivo Purkinje cell function is abnormal in mdx miceHaving established that the gross anatomy and basic circuit features ofthe cerebellum are undisturbed in the mdx mice, we next testedfunctional properties (Fig. 3). We first recorded from anesthetized

control andmdxPurkinje cells (Fig. 3A,B) to provide a large cell yield.Recordings revealed similar membrane voltages between thegenotypes (e.g. example voltage traces in Fig. 3C,D), and, based ontheir firing signatures, both genotypes had recognizable Purkinje cells,with clear complex spikes confirming their cell identity (indicated bygreen asterisks in Fig. 3C,D). Additionally, individual simple spikeand complex spike waveforms were comparable between control andmdxmice (Fig. 3E-H). These data provided us with confidence that ifPurkinje cells in the mdx mice are functionally abnormal, it is due tochanges in their spike firing patterns rather than to a change in or lossof the different spike types themselves [e.g. after conditional deletionof complex spike activity (White and Sillitoe, 2017)]. Therefore, inorder to examine how the Purkinje cells behave in vivo, we analyzedand compared their mean firing summary statistics.

Genotype had a significant effect on the combined set ofanesthetized simple spike mean firing rate, mean coefficient ofvariance (CV) and mean coefficient of variance of adjacentinterspike intervals (CV2) [one-way MANOVA, Wilk’sλ(1,92)=0.9028, P=0.026]. However, even though simple spikefiring was ∼13% lower in mdx animals compared to controls(Fig. 3I; 26.8±1.9 Hz vs 30.8±1.6 Hz; Wilcoxon rank sum test,z=1.99, P=0.047), this change was not significant when correctedfor multiple comparisons. Neither overall irregularity (CV) nor local

Fig. 1. The molecular expression profile of cerebellar neurons is similar in mdx and control mice. (A) Control mice express dystrophin in Purkinje cells,whereasmdxmice do not. (B-J) Control andmdxmice have similar expression patterns for markers of Purkinje cells (B), inhibitory neurons (C-E), and excitatorycells and inputs (F-J). (B) CAR8 labels Purkinje cell somata and dendrites (axons are also labeled with CAR8 but are not seen in this tissue orientation).(C) Parvalbumin (PV) labels inhibitory Purkinje cells, molecular layer interneurons and granular layer interneurons. (D) Neurofilament heavy (NFH) labels Purkinjecells and basket cell axons. (E) Neurogranin labels Golgi cells in the adult mouse. (F) Gamma-amino-butyric-acid receptor alpha 6 (GABAaRα6) labelsgranule cells and their parallel fiber axons. (G) Calretinin labels a subset of unipolar brush cells. (H) Vesicular glutamate transporter 1 (VGLUT1) labels parallelfiber and mossy fiber terminals. (I) Vesicular glutamate transporter 2 (VGLUT2) labels climbing fiber and mossy fiber terminals. (J) Cocaine- and amphetamine-related transcript peptide (CART) labels climbing fiber axons and terminals mainly in lobules IX and X. Scale bar: 200 µm. gl, granular layer; ml, molecularlayer; pcl, Purkinje cell layer. N=3 mice of each genotype.

3

RESEARCH ARTICLE Disease Models & Mechanisms (2020) 13, dmm040840. doi:10.1242/dmm.040840

Disea

seModels&Mechan

isms

irregularity (CV2) showed significant differences (z=1.76, P=0.079and z=−1.48, P=0.14, respectively). Genotype also did not have asignificant effect on complex spike firing [Fig. 3J; Wilk’sλ(1,92)=0.9929, P=0.89]. Given the trend toward lower simplespike firing in mdx mice, our previous data demonstrating thatanesthesia can suppress Purkinje cell activity in vivo (Arancilloet al., 2015), and multiple reports of mdx Purkinje cell changes

in vitro (Anderson et al., 2002; Pereira da Silva et al., 2018; Snowet al., 2014), wewere next motivated to test whether cerebellar circuitdysfunction inmdxmicemight bemore robust and pronouncedwhenanalyzed in an awake and behaving context (Fig. 4).

Visual inspection of the raw traces shows more complex spikes inawake compared to anesthetized animals, as reported in previousstudies (Arancillo et al., 2015). Upon further examination of the raw

Fig. 3. Extracellular spontaneous firing properties of Purkinje cells in ketamine–dexmedetomidine-anesthetized mdx and control mice. (A) Schematicshowing the recording approach. (B) Schematic illustrating the electrode placement for targeting the Purkinje cells. (C) Example raw voltage trace from acontrol Purkinje cell, with complex spikes indicated by green asterisks. (D) As in C, for mdx mice. (E) Average simple spike waveform for the cell shown in C.(F) Average complex spike waveform for the cell in C. (G,H) Simple spike and complex spike waveforms for the mdx cell shown in D. (I) Mean simple spike firingrate (y-axis) was not significantly lower in mdx mice (red) compared to control mice (blue) when corrected for multiple comparisons. No statistically significantdifference in mean coefficient of variance (CV) (x-axis) was found, although there was a trend toward lower variability in themdxmice. (J) No significant differencesin mean firing rate or mean CV for complex spikes were found. Circles are individual data points; crosses represent mean±s.e.m. n=48 control cells, n=46mdx cells,N=12 mice of each genotype. Please refer to Table S1 for a list of the statistical tests used in this figure.

Fig. 2. Gross cerebellar morphology doesnot differ between mdx mice and controls.(A,B) Representative mid-sagittal sectionsshowing Purkinje cells stained with calbindin(magenta) and granule cells stained withfluorescent Nissl (green). (C) No significantdifference in cerebellar size was found betweengenotypes [control=30.1×106±2.8×106 μm2,mdx=29.2×106±1.9×106 μm2, two-tailed Student’st-test, N=3 mice of each genotype, n=3-8 sectionsper animal, t(4)=0.280, P=0.79]. (D,E) Higher-power images from lobules III-IV illustrate howmolecular layer thickness was measuredperpendicular to the Purkinje cell monolayeradjacent to the primary fissure. (F) No significantdifference in molecular layer thickness was foundbetween genotypes [control mean=152.3±3.7 μm,mdx mean=155.7±4.0 μm, two-tailed Student’st-test, N=3 mice of each genotype, n=4-8 sectionsper animal, t(4)=−0.623, P=0.57]. Scale bars:500 µm (A); 50 µm (D). ml, molecular layer; ns,non-significant; pcl, Purkinje cell layer.

4

RESEARCH ARTICLE Disease Models & Mechanisms (2020) 13, dmm040840. doi:10.1242/dmm.040840

Disea

seModels&Mechan

isms

firing traces betweenmdxmice and controls, we determined that thestructure of the spikes was similar between genotypes (Fig. 4C,D).Mean spike waveform shape was identical for both simple spikesand complex spikes between genotypes (Fig. 4E-H). In awakePurkinje cells, genotype had a significant effect on the set of simplespike mean firing rate, mean CV and mean CV2 [one-wayMANOVA, Wilk’s λ(1,46)=0.8266, P=0.0371]. Interestingly,mean simple spike firing rate was significantly different, at ∼19%lower inmdx Purkinje cells than in control cells (Fig. 4I; 69.1±3.4 vs85.3±4.3 Hz; Wilcoxon rank sum test z=2.53, P=0.012). No effectswere observed in the regularity of spikes (CV z=−0.03, P=0.98,CV2 z=−0.13, P=0.89). In addition, complex spikes also differedbetween genotypes in the awake state [Fig. 4J; Wilk’s λ(1,46)=0.7772, P=0.0106]. Although overall complex spike rate wasnot changed (Hz z=−0.55, P=0.58), overall complex spikevariability was lower in mdx mice than in controls (CV=0.76±0.02 vs 0.92±0.04; z=3.17, P=0.0016). Complex spike CV2 was notaffected in Purkinje cells that lack dystrophin (z=0.75, P=0.45).Together, these data demonstrate that, in awake animals, simplespike firing frequency is lower in mdx mice than in controls. It alsoreveals less variable complex spike firing in awake mdx micecompared to awake control mice.We accounted for the expression of other dystrophin isoforms (i.e.

Dp71, Dp116, Dp140 and Dp260) in the mdx Purkinje cells byprobing for expression of the common C-terminal domain (Fig. S2).We confirmed antibody specificity by demonstrating heavyexpression in the control hippocampus, as previously reported

(Blake et al., 1999), and antibody localization to the plasmamembrane of Purkinje cells (Fig. S2). We also evaluated the numberof days since surgery (Fig. S3) and the recording depth [Fig. S4;evaluated with respect to dorsoventral and mediolateral coordinatesrelative to Paxinos and Franklin (2001) via labs.gaidi.ca/mouse-brain-atlas] as additional potential predictor variables. Through stainingthe C-terminal domain of the dystrophin protein, we found that thePurkinje cells ofmdxmice have comparable dystrophin expression tocontrol Purkinje cells (Fig. S2). This suggests that the mdx mutationand our electrophysiological findings are specific to the loss of theDp427 isoform. To assess the impact of the days post-surgery and therecording depth on our data, we binarized each factor into low or highconditions relative to the median value (Figs S3 and S4), and foundthat neither factor had significant effects on awake simple spike firingin one-way MANOVAs [respective Wilk’s λs: λ(1,46)=0.9613,λ(1,46)=0.9985; respective P-values: 0.63, 0.99]. We likewiseexamined the effect of predictor variables on complex spike firingin awake Purkinje cells. Again, the recording days post-surgery andrecording depth did not show significant effects [λ(1,46)=0.8533,P=0.070; λ(1,46)=0.9943, P=0.97, respectively]. Moreover, becauserecording depth did not differ significantly between cells recorded incontrol compared to mdx mice (Wilcoxon rank sum test, z=0.58,P=0.56), it indicated that the differences in firing between genotypeswere not attributable to differential sampling. These data suggest that,based on the factors we tested, the only factor showing a significanteffect on simple spike and complex spike firing properties in awakemice was the genotype of the mice; namely, the loss of dystrophin,

Fig. 4. Awake extracellular spontaneous firing properties of Purkinje cells in mdx and control mice. (A) Schematic showing the recording approach.(B) Schematic illustrating the electrode placement for targeting the Purkinje cells. (C) Example raw voltage trace from a control Purkinje cell, with complexspikes indicated by green asterisks. (D) As in C, for mdx mice. (E) Average simple spike waveform for the cell shown in C. (F) Average complex spike waveformfor the cell in C. (G,H) Simple spike and complex spike waveforms for the mdx cell shown in D. (I) Mean simple spike firing rate (y-axis) was significantly lowerin mdx mice (red) compared to controls (blue). No statistically significant difference in mean CV (x-axis) was found. (J) Complex spikes in mdx mice havesignificantly lower firing variability compared to those in controls, but firing rates do not differ between the genotypes. Circles are individual data points; crossesrepresent mean±s.e.m. n=24 cells of each genotype, N=4 mice of each genotype. Please refer to Table S1 for a list of the statistical tests used in this figure.

5

RESEARCH ARTICLE Disease Models & Mechanisms (2020) 13, dmm040840. doi:10.1242/dmm.040840

Disea

seModels&Mechan

isms

isoform Dp427, in the mdx mice. We therefore postulated that thesimple spike and complex spike firing defects in themdxmice shouldalter the ability of Purkinje cells to communicate their inhibitoryoutput signals to their target neurons that are located in the threedivisions of the cerebellar nuclei.

Cerebellar nuclei output function is compromised inmdx miceHaving observed changes in Purkinje cell spike firing, we sought todetermine what effect changes in their output would have oncerebellar nuclei firing (Fig. 5A,B). Raw voltage traces showedclear spike trains in the mutant mice (Fig. 5C,D). Days post-surgery,recorded position (dorsoventral, anteroposterior or mediolateral)and targeted nuclei all had no significant effect on firing propertiesin our recorded sample [λ(1,49)=0.9785, P=0.79; λ(1,49)=0.9022,P=0.18; λ(1,49)=0.9174, P=0.25; λ(1,49)=0.8891, P=0.13;λ(1,49)=0.9019, P=0.18, respectively; Fig. S5]. Genotype,however, did have a significant effect [λ(1,49)=0.7494,P=0.0033]. In pairwise testing, cerebellar nuclear firing rate andoverall variability did not differ (z=−1.47, P=0.14, z=−0.71,P=0.48), but CV2 was ∼19% higher in mdx mice than in controls(Fig. 5E; CV2=0.46±0.02 vs 0.39±0.02; z=−2.36, P=0.018). Thissuggests that the net effect that lower Purkinje cell simple spikefiring output and lower complex spike variability has on cerebellarnuclear cells inmdxmice is to increase the local variability of firing,or the variability in timing between one spike and the next. Thiseffect is not likely due to a change in full-length dystrophinexpression in the cerebellar nuclei themselves, because the verymodest expression in a small population of cells is comparablebetween control and mdx mice (Fig. S6). It should be noted that, inmany cells, the staining is barely beyond the level of background.Therefore, the changes we find in Purkinje cell activity appear toaffect cerebellar nuclei output activity with little-to-no impact fromdystrophin in the cerebellar nuclei. Activity changes in the

cerebellar nuclei are predicted to alter the manner in which thecerebellum communicates with the rest of the brain and spinal cordduring different behaviors.

DISCUSSIONDMD causes progressive muscle degeneration and fibrosis, whichleads to loss of motor strength and respiratory insufficiency(Falzarano et al., 2015). At later stages of disease, cardiomyopathydevelops, with heart failure or respiratory complications leading todeath, on average in the mid-30s, even with modern treatmentregimens. Therapy for DMD has benefited significantly from studiesof genetic mouse models of muscular dystrophy, including the mdxmodel developed in the late 1980s (Crone and Mah, 2018). Currentinvestigational therapies for DMD with ongoing research in DMDmouse models include exon skipping, micro-dystrophin or surrogategene therapy, gene editing, inflammation blockers and stem celldelivery (Long et al., 2016; Min et al., 2019). Canine models alsooffer immense insight into therapeutic approaches (Amoasii et al.,2018), although for studies of how DMD impacts brain anatomy andcircuit function, the mouse still has advantages with respect to in vivoelectrophysiology and complex genetics experiments. Therefore, inaddition to their utility for measuring muscle pathology, DMD micecould be a promising avenue for uncovering the neural impairmentsoften seen in individuals with DMD (Anderson et al., 2002; Rae andO’Malley, 2016; Vaillend et al., 2004). In this regard, the cerebellumhas been a major target of interest (Cyrulnik and Hinton, 2008).Previous studies of mdx electrophysiology have reported changes inGABA release and uptake from cerebellar synaptosomes (Pereira daSilva et al., 2018), and lower firing rates and higher variability in thefiring of dissociated Purkinje cells (Snow et al., 2014). However,whether these deficits are seen in awake mice has not previously beentested. Here, we tested this possibility and found that the loss ofdystrophin alters Purkinje cell and cerebellar nuclei communicationin behaving mdx mice.

Fig. 5. Spontaneous firing properties of mdx and controlcerebellar nuclear neurons in awake behaving mice.(A) Schematic showing the recording approach. (B) Schematicillustrating the electrode placement for targeting the cerebellarnuclear neurons. (C) Example raw voltage trace from a controlcerebellar nuclear neuron. (D) Example raw voltage trace froman mdx cerebellar nuclear neuron. (E) In awake mice, meansimple spike firing rate was not significantly different betweenthe mdx mice (red) and controls (blue), but CV2 (variability)was higher in the mdx mice than in the controls. Circles areindividual data points; crosses represent mean±s.e.m. n=28control cells, n=23 mdx cells, N=4 mice of each genotype.Please refer to Table S1 for a list of the statistical tests usedin this figure.

6

RESEARCH ARTICLE Disease Models & Mechanisms (2020) 13, dmm040840. doi:10.1242/dmm.040840

Disea

seModels&Mechan

isms

The main goal of this paper was to investigate cerebellar functionin vivo. The rationale was to determine whether the neurologicaldefects reported in humans with DMD are reflected by changes inspecific circuit functions in an alert state. Towards this, we recordedand examined the electrophysiological properties of cerebellarneurons in the intact brain, in an mdx mouse model of DMD. Wechose to analyze only female mice to dissociate cerebellarfunctional output from potential compensation (Xu-Wilson et al.,2009; Zwergal et al., 2013) due to higher muscle degradation inmales; however, it would be informative in future work to extendthese analyses to the more severely affected male mice (Hakim andDuan, 2012; Hourdé et al., 2013; Salimena et al., 2000), or toheterozygous mice that express a relevant phenotype. Notably, wedid not observe any overt behavioral deficits in the mdx mice at theages we examined, suggesting that changes in the firing propertieslikely precede the onset of pathological motor symptoms.Before interrogating circuit function, we first determined that

several anatomical properties of the cerebellar microcircuit areunaffected in mdx mice, including cell type distribution, molecularlayer thickness, molecular patterning and afferent terminallocalization. We next recorded cerebellar activity in vivo andfound a significant decrease of ∼19% in mean Purkinje cell simplespike firing rate in mdx mice compared to controls (Fig. 4). Twomeasures of simple spike variability showed no difference bygenotype, but complex spikes were ∼17% less variable in the mdxmice compared to those in the controls (Fig. 4). The centrality ofPurkinje cells in the cerebellar cortical circuits place these findingsin a broader context of cerebellar function in disease (Reeber et al.,2013). Inputs from climbing fibers, inputs from mossy fiber–granule cell–parallel fiber pathways, and local modulation byinhibitory molecular layer interneurons all converge upon thePurkinje cell dendrites. Therefore, whatever the disease-causinginjury (genetic or physical) or the ultimate impact on the circuit, thePurkinje cells are almost always involved in executing the responseof the cerebellum. In mouse models of ataxia, in vitroelectrophysiological analyses of Purkinje cell spiking revealalterations as a consequence of molecular pathological defectsand degeneration [SCA1 transgenic (Inoue et al., 2001)], which canstart before the onset of obvious pathology [Purkinje cell-specificSCA2 transgenic (Hansen et al., 2013)], or due to changes inPurkinje cell intrinsic excitability [SK3-1B-GFP (Shakkottai et al.,2017)]. These data are supported by in vivo recordingsdemonstrating that Purkinje cell activity is dramatically altered inataxia, as revealed in the Car8wdlmodel that exhibits a severe ataxiawithout neurodegeneration (Miterko et al., 2019a; White et al.,2016a). The significant tremor in Car8wdl also implicates thePurkinje cell firing defects as a source of the 4-12 Hz oscillationsthat might drive tremor pathophysiology (Kuo et al., 2019; Whiteet al., 2016a). Purkinje cell firing defects are also the source of thehigh-stepping gait, abnormal twisting of the trunk and limbs, andpoor mobility that is observed in different models of dystonia(Calderon et al., 2011; Fremont et al., 2015, 2017; Luna-Cancalonet al., 2014; White and Sillitoe, 2017). Here, we show that a mousemodel of DMD, which is primarily a muscle degenerative diseasebut with clear neurological and neuropsychiatric components,contains cerebellar circuit deficits that fall into an expanding themeof activity-related firing disruptions. Defective Purkinje cell spikingcould therefore mediate a wide range of behavioral changes in allconditions that involve cerebellar dysfunction.Notably, Purkinje cells do not function in isolation: they synapse

directly onto cerebellar nuclei neurons, which communicatecerebellar information to regions such as the thalamus, red

nucleus, vestibular nuclei and the inferior olive. Because defectsin Purkinje cell spike firing can have significant effects on theirtarget neurons (Todorov et al., 2012; White et al., 2014), withresulting problems in motor control and learning (Schonewille et al.,2010; White et al., 2014), we also recorded from cerebellar nuclearneurons. Cerebellar nuclear neurons in mdx mice had a ∼19%higher local variability in firing (CV2) compared to those in controlmice. Again, this result is reminiscent of the irregular cerebellarnuclei firing in ataxia (Hoebeek et al., 2008; White et al., 2014) anddystonia (Fremont et al., 2014; White and Sillitoe, 2017). However,it is unclear whether irregular firing of the cerebellar nuclei in onedisease is exactly equivalent, at the behavioral level, to irregularfiring in other diseases. Part of the complication is that the cerebellarnuclei have multiple targets, and it is unclear whether the sametarget neurons are implicated in different diseases. For example, it ispostulated that the thalamic nuclei and connections to the basalganglia are central to dystonia (Chen et al., 2014;White and Sillitoe,2017), but is this the same pathway that also mediates motor defectsin DMD? Perhaps, but DMD also involves non-motor dysfunctions,which are likely served through a different set of circuits. Oneargument is that the cerebellum might process all functions througha ‘universal cerebellar transform’ (Schmahmann et al., 2019). Thistheory postulates that regardless of whether the cerebellum ismodulating action or cognition, its role as a coordinator of severalneural tasks is the same. Certainly, there is strong clinical evidencefor the cerebellum in emotion-based disorders (Schmahmann,2019), and experimentally its contribution to behaviors such asreward and social behavior are emerging (Carta et al., 2019).Indeed, during non-motor function, the cerebellum mustcommunicate with the cerebral cortex (Wagner et al., 2017;Wagner et al., 2019). Multiple anatomical pathways linking thecerebellum to cortical sites likely subserve these different non-motor functions (Bostan et al., 2013). The number of cerebralcortical sites that communicate with the cerebellum is potentiallywidely distributed (Marek et al., 2018), perhaps a prerequisite to thecerebellum’s involvement in non-motor function in different andperhaps unexpected cases, such as DMD. It stands to reason thatpowerful cerebellar computations are necessary to drive suchcritical behaviors. One possibility is that the cerebellum promotescoherent activity between cortical brain regions; an example couldbe an adjustment in hippocampal–prefrontal cortex coherence uponPurkinje cell activation (McAfee et al., 2019). Interestingly, thehippocampus (Hoogland et al., 2019; Krasowska et al., 2014;Miranda et al., 2015, 2016) and prefrontal cortex (Suzuki et al.,2017) exhibit abnormalities in human DMD and mdx mice. It istherefore tantalizing to hypothesize that the decreased Purkinje cellsimple spike output and the more variable cerebellar nuclear outputin mdx mice could contribute to the motor or cognitive deficitsreported in individuals with DMD.

Previous reports have noted changes in the distribution ofGABAA receptors on Purkinje cells in mdx mice (Grady et al.,2006). Although this molecular change could potentially contributeto the altered firing we report inmdx Purkinje cells, it is currently notclear what effect it might have. Specifically, because Purkinje celldendrites have active rather than passive conduction properties, theexact distribution of GABAA receptors is critical to predicting theireffect on Purkinje cell output. For instance, in biophysicalsimulations, distal dendritic inhibition has a larger effect on firingrate than proximal inhibition (Solinas et al., 2006). Therefore, it isdifficult to predict how the changes in mdx GABAA receptordistribution would ultimately impact Purkinje cell firing propertieswithout a more detailed and quantitative description of the exact

7

RESEARCH ARTICLE Disease Models & Mechanisms (2020) 13, dmm040840. doi:10.1242/dmm.040840

Disea

seModels&Mechan

isms

composition of dendritic channels in the mdx mice, and across thedifferent cerebellar lobules. We also note that dystrophin isexpressed in at least seven isoforms (Muntoni et al., 2003), andwhile the full-length Dp427 isoform is the one most stronglyexpressed in Purkinje cells (Górecki et al., 1992), smaller isoformslike Dp71 and Dp140 could potentially compensate to some degreefor the mdx mutation and loss of Dp427. Nevertheless, our resultspoint to an important role for Dp427 in cerebellar function thatcannot completely be compensated for by other isoforms.The loss of dystrophin expression in mdx mice does not prevent

specific cell populations from differentiating, eliminate certainclasses of neurons or prevent any of the main cerebellar cell typesfrom attaining their proper position within the tri-laminar cerebellarcortex (Fig. 1). However, despite the correct location of cells andafferent fibers, cerebellar zonal topography is altered. Thecerebellum is organized into a series of sagittal zones that aredefined by neuronal birth date, lineage restriction, embryonic geneexpression, afferent topography, neuronal spike properties andmolecular marker expression (Apps et al., 2018; Miterko et al.,2019b). Zones are organized around the Purkinje cells, whichexpress the most striking of all zonal patterns. Previous workdemonstrated that dystrophin is not expressed at equal levels in allPurkinje cells, but instead in a pattern of zones that fall within thefundamental cerebellar map (Blake et al., 1999). In both the muscleand brain, dystrophin binds to a complex of proteins including thedystrobrevins and a coiled-coil protein called dysbindin (Bensonet al., 2001). Dysbindin is heavily expressed in mossy fiberterminals, where it also forms an array of zones that respect thetopography of Purkinje cells (Sillitoe et al., 2003). Loss ofdystrophin in mdx mice alters the terminal field distribution ofdysbindin-expressing mossy fibers (Sillitoe et al., 2003). Thechanges in Purkinje cell function and the accompanied defects inmossy fiber patterning in the mdx model is reminiscent of ataxicmice that have a lack of Purkinje cell neurotransmission with poorlydefined mossy fiber zones (White et al., 2014). Moreover, there isno overt neurodegeneration in either model. It therefore is temptingto speculate that the behavioral deficits in mdx mice are, at least inpart, due to zonal miswiring. Based on the distribution of firingproperties, we likely recorded from Purkinje cells in zebrinII+ aswell as zebrinII− zones (Figs 3 and 4) (Xiao et al., 2014; Zhou et al.,2014). If there are functional zonal defects in mdx mice, we predicttheir involvement in a number of different behaviors. However, howmotor and non-motor information are simultaneously encoded inzones is unclear, although the convergence of Purkinje cell zonalinformation within the cerebellar nuclei (Person and Raman, 2011)would have an impact on how signals are altered in mdx mice(Fig. 5). Specifically, changes in complex spikes would be predictedto alter cerebellar output (Tang et al., 2019), especially since theclimbing fiber projections also adhere strictly to the boundaries ofthe Purkinje cell zonal map (Gravel et al., 1987; Sugihara and Quy,2007) and contribute to behavior (Horn et al., 2010). Within eachzone, loss of dystrophin could, in theory, impact learningmechanisms via zone-dependent changes in neurotransmissionand plasticity (Kostadinov et al., 2019; Paukert et al., 2010;Wadiche and Jahr, 2005). Similar to the synchronous activation ofcell ensembles that control individual muscles (Welsh et al., 1995),cerebellar firing properties and circuit patterning could also instructnon-motor communication across higher-order brain centers. Ourextracellular recordings included cells in lobules VI-VII of thevermis and CrusI and CrusII of the hemispheres, regions known tobe involved in motor behaviors, with additional and growingevidence that they contribute to various non-motor behaviors

(Koziol et al., 2014; McAfee et al., 2019; Shipman and Green,2019). Based on the abnormalities we uncovered, we speculate thatthe function of cerebellar dystrophin complexes could mediate suchfunctions in vivo, during behavior.

ConclusionDMD is a debilitating disease that results in death. Boys are affectedin the severe forms, which start with muscle degeneration, althoughcardiac and respiratory problems often determine the endpoint of thedisease. Unfortunately, during the course of disease, afflictedindividuals also have to deal with neurological deficits that may bemotor as well as non-motor in nature. A number of brain regions,such as the hippocampus and prefrontal cortex, could mediate thecognitive defects in humans and in animal models such as the mdxgenetic model. In addition, the cerebellum, a region involved inmotor and non-motor functions, has been implicated in DMD.Behavioral studies as well as in vitro electrophysiology experimentsinmdxmice support a contribution of the cerebellum to DMD. Here,we used mdx mice to show that Purkinje cells fire abnormally invivo, and that their altered spiking properties are communicateddownstream to cerebellar nuclear neurons, which are also defective.We demonstrate that cerebellar output is more variable, a circuitmodification that could have signaling consequences similar todiseases such as ataxia, tremor and dystonia. We postulate thaterroneous cerebellar output could mediate the motor and non-motorabnormalities in DMD.

MATERIALS AND METHODSAnimal maintenanceMouse husbandry and experiments were performed under an approvedInstitutional Animal Care and Use Committee (IACUC) protocol at BaylorCollege of Medicine. Female C57BL/10ScSnJ control (stock #000476;N=22) and C57BL/10ScSn-Dmdmdx/J homozygous mutant (stock #001801;N=22; mdx/mdx) mice were obtained from The Jackson Laboratories (BarHarbor, ME, USA). The use of this specific control mouse line is based onknown difficulties in breeding the mutant strain and is supported byprevious work demonstrating the validity of their use (Rybalka et al., 2014;Xu et al., 2019; Yoon et al., 2019). We used only one sex to avoid anypotential sex differences in cerebellar function (Mercer et al., 2016).Specifically, we chose to use female mice to try to dissociate the effect of theloss of dystrophin as much as possible from potential compensatory changesin cerebellar output; changes in muscle function could potentially inducecompensatory changes in cerebellar activity (Xu-Wilson et al., 2009;Zwergal et al., 2013), and female mutants have more modest musculardeficits than male mutants when aged less than 6 months (Hakim and Duan,2012; Hourdé et al., 2013; Salimena et al., 2000). Of note, though, ourprevious work using female homozygous (mdx/mdx) and male hemizygous(mdx/Y) mice did not reveal differences in cerebellar functional anatomy orsensitivity to molecular patterning changes between the sexes (Sillitoe et al.,2003). Regardless, in this work, we analyzed cerebellar function in adult, 2-to 5-month-old female mice.

SurgeryFor anesthetized recordings, the mice were anesthetized with 3% isofluraneand then with a cocktail containing 75 mg/kg ketamine and 0.5 mg/kgdexmedetomidine. We then transferred the mice from the anesthesiachamber to a stereotaxic platform (David Kopf Instruments, Tujunga, CA,USA) that is integrated with an in vivo recording rig. Throughout theexperiment, mice were connected to a breathing tube and isofluraneconcentration was maintained at 0.15-0.25%. The head was fixed in placewith metal ear bars. Fur in the back of the head was removed usingdepilatory cream. An anteroposterior incision was madewith a scalpel bladeto expose the skull. Depending on the age, the muscle was cut and reflectedaway if the area for the craniotomy was covered. All craniotomies wereperformed 0-1.5 mm to the right of the midline with reference to bregma

8

RESEARCH ARTICLE Disease Models & Mechanisms (2020) 13, dmm040840. doi:10.1242/dmm.040840

Disea

seModels&Mechan

isms

(Paxinos and Franklin, 2001), and above the region that approximatelycorresponds to lobules VI-VII of the vermis or CrusI and CrusII of thehemispheres. A hole ∼1 mm in diameter was drilled into the skull using aDremel handheld rotary drill (model 4000). Once the skull was drilled totranslucence, the remaining bone and cartilage were etched with an 18-gauge needle until a circular flap could be removed to expose the brain. Thecraniotomy was performed carefully because even minimal tissue damagecauses brain and blood vessel pulsations, which inhibit stable single-unitrecordings. Moisture within the craniotomy was maintained by keeping theopening full with drops of 0.9% w/v saline solution. During anesthesia,Purkinje cells were recorded between 0.2 mm and 2.4 mm(median=1.1 mm) ventral to the surface of the brain.

For awake recordings, surgical procedureswere similar to thosewe previouslyreported (White et al., 2016b; Stay et al., 2019), with the followingmodifications: mice were anesthetized with 3% isoflurane at induction,followed by 1-2% isoflurane for maintenance. After the craniotomy, a custom2-mm circular plastic recording chamber was secured to the skull with dentalacrylic, along with a metal head plate for restraint. A small metal post wasimplanted over bregma for stereotaxic reference during recordings. Aftersurgery, the mice were transferred to a warmed box chamber until they returnedto sternal recumbency, at which point they were returned to their home cages tocontinue recovering. Recordings began after three or more days post-surgeryrecovery. Mice were head-fixed on a foam running wheel, where they were freeto make limb or whisking movements, though higher speeds of walking orrunning were inhibited by a high coefficient of friction of wheel rotation.Purkinje cells were recorded 5.5-7.7 mm (median=6.8 mm) posterior and 0.5-2.0 mm (median=1.1 mm) lateral to bregma, at 0.3-2.9 mm (median=1.7 mm)deep from the surface of the brain. Cerebellar nuclei were targeted between1.8 mm and 4.0 mm (median=2.8 mm) ventral from the brain surface, andbelow surface coordinates of 5.8-7.6 mm (median=6.2 mm) posterior and 0.5-2.0 mm (median=1.4 mm) lateral to bregma. The precise stereotaxic coordinatesrelative to bregma were used to reconstruct the recording sites post hoc.

Electrophysiology: spike properties and data analysisSingle-unit recordings were attained with 2-5 MΩ Tungsten electrodes(Thomas Recording, Germany) that are controlled by a motorizedmicromanipulator (MP-225, Sutter Instrument). The signals were band-pass filtered at 0.3-13 kHz, amplified with an ELC-03XS (NPI, Tamm,Germany) amplifier, and digitized into Spike2 (CED, Cambridge, UK).Well-isolated units were analyzed over 200-300 s in anesthetized animals(median=200 s), and 40-120 s in awake animals (median=71 s). Theanesthetized mice were secured in a stereotaxic unit during the recordings,whereas the awake animals were recorded during quiet wakefulness whileon a foam wheel (White et al., 2016b). Analysis of the raw traces wasperformed with Spike2, Excel (Microsoft) and MATLAB (MathWorks).Purkinje cells were identified by the presence of simple spikes and complexspikes. Simple spikes are generated intrinsically within Purkinje cells andare modulated by mossy fiber to granule cell afferent inputs, whereascomplex spikes are a Purkinje cell-specific action potential that resultexclusively from the excitatory climbing fiber input. Complex spikes cause asubsequent pause of ∼20 ms in simple spike firing. Simple spikes andcomplex spikes were sorted independently. Cerebellar nuclear neurons wereidentified by their significant and characteristic depth within the cerebellum,as well as their general firing profile that typically sounds distinct fromPurkinje cells when isolated with the aid of an audio amplifier. Based on thestereotaxic coordinates that were used for targeting the electrodes, we predictthat most cerebellar nuclear neurons were recorded from the fastigial andinterposed nuclei, with some potentially isolated from deeper within thevestibular nuclei. The spike trains for both cell types were analyzed forfrequency (Hz=spikes/s). To quantify the average variability in their firingpattern, the interspike interval (ISI) CV [(standard deviation of ISIs)/(meanof ISIs)] was calculated. Tomeasure the variability in firing patterns within ashort period of two interspike intervals, the CV2 [CV2=2*|ISIn+1−ISIn|/(ISIn+1+ISIn)] was calculated (Holt et al., 1996). Throughout the study,numerical results are reported as mean±s.e.m. Pairwise statisticalcomparisons of the different firing properties were performed with thenon-parametric Wilcoxon rank sum test (Table S1). Significance was set atα=0.05 threshold. Sample size was not determined using a priori power

analysis, but was based on the statistical criteria for significance inobservations. Animals were not randomized into analysis groups, and noanimals or samples were excluded. Data are available upon request.

Statistics of electrophysiological recordingsWe examined several different predictor variables and their relationship toseveral response variables. Specifically, we tested the null hypotheses thatthe predictor variables [genotype, days post-surgery at recording andrecorded depth (also indicates the putative lobule)] had no significant effectson the means of the response variables (mean firing rate, mean CV andmeanCV2). We tested these hypotheses through an initial omnibus MANOVAtest for each predictor variable. If the MANOVA result showed likelihood ofP<0.05 that the means of each response variable were equal between thelevels of the predictor variable, we rejected the null hypothesis andproceeded with pairwise comparisons to determine which responsevariable(s) differed. Since we did not know the true underlying variancelevels for each response variable, we used non-parametric Wilcoxon ranksum tests rather than Student’s t-tests, to avoid assuming equal variance. Thesignificance levels of all statistical tests were corrected for multiplecomparisons with the Benjamini–Hochberg correction. A P-value less thanthe Benjamini–Hochberg critical value with false discovery rate set to 0.2was interpreted to reject the null hypothesis of equal means in a givenresponse variable between the predictor groups. Please refer to Table S1 fora list of the statistical tests used in this study.

Tissue preparation and cuttingFor perfusion-fixation, animals were deeply anesthetized with avertin (2,2,2-tribromoethanol), and then perfused through the heart with 0.1 M phosphate-buffered saline (PBS; pH 7.2), followed by 4% paraformaldehyde (PFA)diluted in PBS. After incubation in 4% PFA overnight, the tissue was placedinto 70% ethanol before embedding in wax for paraffin tissue processing(Sillitoe et al., 2008). Paraffin tissue sectionswere cut at 10 µm on amicrotomeand ribbons collected on positively charged glass slides from a warm waterbath. Alternatively, for free-floating tissue processing, the brains were post-fixed in 4%PFA for 24-48 h after perfusion, cryoprotected stepwise in sucrosesolutions (15% and 30% diluted in PBS), embedded in Tissue-Tek OCTCompound (Sakura Finetek, Torrence, CA, USA), and then frozen at −70°.The tissue was cut at 40 µm on a cryostat and collected into 24-well plates.

ImmunohistochemistryImmunohistochemistry was carried out as described previously (Reeber andSillitoe, 2011; Sillitoe et al., 2003, 2010; White and Sillitoe, 2013). Briefly,tissue sections were thoroughly washed, blocked with 10% normal goat serum(NGS; Sigma-Aldrich, St Louis, MO, USA) for 1 h at room temperature andthen incubated in 0.1 M PBS containing 10% NGS, 0.1% Tween-20 andprimary antibodies for 16-18 h at room temperature, shaking gently. The tissuesections were then washed three times in PBS and incubated in goat anti-mouse horseradish peroxidase (HRP)-conjugated secondary antibodies(1:200; DAKO, Carpinteria, CA, USA) for 2 h at room temperature, againshaking gently. The cerebellar tissue was subsequently rinsed, andimmunoreactivity was analyzed using SG substrate (Vector Laboratories,Burlingame, CA, USA) as a chromogen. After mounting the tissue onto glassslides, sectionswere coverslipped using Cytosealmountingmedium (Richard-Allen Scientific, San Diego, CA, USA). We tested the specificity of thesecondary antibodies by processing the tissue sections without primaryantibodies. No signal was detected in such control experiments, indicating thatthe staining we observed was not due to non-specific signals from the HRPantibodies (data not shown). Please refer to our previous work for additionaltests of reliable signals from the antibodies used in this study (Reeber andSillitoe, 2011; White and Sillitoe, 2017; White et al., 2014, 2016b).

AntibodiesAnti-dystrophin antibody (1:20, NCL-DYS1), purchased from LeicaBiosystems (Newcastle upon Tyne, UK), was used to verify that full-length dystrophin was eliminated inmdxmice. To account for the expressionof the other dystrophin isoforms (i.e. Dp71, Dp116, Dp140 and Dp260) incontrol and mdx Purkinje cells, we used an anti-dystrophin (MANDRA1)antibody (1:100; NB120-7164) from Novus Biologicals (Centennial, CO,

9

RESEARCH ARTICLE Disease Models & Mechanisms (2020) 13, dmm040840. doi:10.1242/dmm.040840

Disea

seModels&Mechan

isms

USA), targeted to the C-terminus. Mouse monoclonal anti-calbindin-D28K(calbindin; 1:10,000; CD38), rabbit polyclonal anti-parvalbumin (1:1000;PV25) and rabbit polyclonal anti-calretinin (1:500; CR7699/3H) werepurchased from Swant (Marly, Switzerland). Calbindin specifically labeledPurkinje cells, parvalbumin labeled Purkinje cells and GABAergicinterneurons, and calretinin labeled unipolar brush cells. We also usedanti-CAR8 primary antibodies (1:3000; Santa Cruz Biotechnology, SantaCruz, CA, USA) to label Purkinje cells (Miterko et al., 2019a). To assessexcitatory synapses, we used mouse monoclonal anti-vesicular glutamatetransporter 2 (VGLUT2; 1:500; MAB5504), purchased from Chemicon(Millipore, Billerica, MA, USA), for visualizing climbing and mossy fiberterminals (Gebre et al., 2012; Hisano et al., 2002; Reeber and Sillitoe, 2011),and rabbit polyclonal anti-vesicular glutamate transporter 1 (VGLUT1;1:500; 135 302), purchased from Synaptic Systems (Goettingen, Germany),for visualizing parallel fiber synapses and mossy fiber terminals (Gebreet al., 2012). The granule cell layer was assessed by using a granule cellmarker, rabbit anti-GABARα6, which was purchased from Millipore(1:500; AB5610). Mouse monoclonal anti-NFH (also called anti-SMI-32;1:1500) was purchased from Covance (Princeton, NJ, USA). Anti-SMI-32recognizes the non-phosphorylated form of NFH (see manufacturer productdatasheet for details), which on tissue sections labels the soma, dendritesand axons of adult Purkinje cells, and also basket cell axons and pinceaux(Demilly et al., 2011). Rabbit anti-neurogranin (1:500) was raised againstfull-length recombinant rat neurogranin protein (Chemicon, Temecula, CA,USA; AB5620). Neurogranin recognizes Purkinje cells in the perinatalcerebellum and Golgi cells in the adult cerebellum (Larouche et al., 2006;Singec et al., 2003). Rabbit polyclonal anti-cocaine- and amphetamine-related transcript peptide (CART 55-102; H-003-62) was used at aconcentration of 1:250 to detect climbing fibers (Reeber and Sillitoe,2011) and was purchased from Phoenix Pharmaceuticals (Burlingame, CA,USA). To fully visualize the fibers, the CART signal was amplified using abiotinylated secondary antibody and the Vectastain Elite ABC method fromVectorLabs (Burlingame, CA, USA; Reeber and Sillitoe, 2011).

Fluorescent Nissl histologyFor dual staining of calbindin and NeuroTrace, 10 µm or 40 µm tissuesections were immunolabeled with primary antibody as described above(calbindin; 1:10,000; CD38). We then added NeuroTrace fluorescentNissl stain (Molecular Probes, Eugene, OR, USA), diluted to 1:1000,directly to the secondary antibody cocktail. Alexa-Fluor®-conjugatedsecondary antibodies (1:1500; Life Technologies; A31570) were used toamplify and detect the calbindin signal. After three PBS rinses, wemounted the stained tissue sections with FLUORO-GEL mountingmedium (Electron Microscopy Sciences, Hatfield, PA, USA) onto glassslides before imaging.

Imaging and data analysisPhotomicrographs of tissue sections were captured with a Zeiss AxioCamMRc5 camera mounted on a Zeiss Axio Imager.M2 microscope. Images oftissue sections were acquired with the Zeiss Apotome.2 system and analyzedusing Zeiss ZEN software (2012 edition). After imaging, the raw data wereimported into Adobe Photoshop CS5 and corrected for brightness andcontrast levels. Schematics were drawn in Adobe Illustrator CS5.

Quantification of anatomical featuresCerebellar size, molecular layer thickness, Purkinje cell number, andcerebellar nuclei areas and densities were quantified from three control andthree mdx mice using ImageJ software (https://imagej.nih.gov/ij/).Cerebellar size and Purkinje cell numbers were determined by calculating,then averaging, the areas or the soma numbers of three to eight sections oftissue, per animal, respectively. All of the cerebellar tissue used in thesecalculations was previously stained with calbindin and/or Nissl and wastaken from within 200 µm from the midline. We used the tissue area,inclusive of all ten vermal lobules, measured from sagittal sections, as aproxy for examining cerebellar size. The number of Purkinje cells wascalculated for the whole sagittal cerebellum (lobules I-X), the anterior zoneof the cerebellum (lobules I-V), the central zone of the cerebellum (lobulesVI-VIII) and the posterior/nodular zones of the cerebellum (lobules IX-X).

To complement our quantification of Purkinje cell number, molecular layerthickness and cerebellar nuclei areas and densities were additionallymeasured. Molecular layer thickness was measured in regions of cerebellarcortex flanking the fissure between lobules III and IV, on sagittal sectionsthat were co-stained with calbindin and Nissl. We chose to quantify thesespecific regions because the cerebellar cortex in these lobules has substantialstretches without curves, which makes quantification more systematic whencomparing between mice. For each mouse, the molecular layer thicknessmeasurements were averaged from four to eight sections separated by∼200 μm around the midline. For additional details on the quantification ofmolecular layer thickness, see Brown et al. (2019). Finally, we quantifiedthe areas and densities of the fastigial and interposed cerebellar nuclei inthree control and three mdx mice, each using three sagittal tissue sectionsstained with calbindin (the Purkinje cell axons in the cerebellar nucleiprovide an ideal outline for assessing general nuclei anatomy) and Nissl.The fastigial nuclei measurements were calculated from sagittal tissuesections that were ±0.48 μm to ±1.08 μm from the midline, whereas theinterposed nuclei measurements were calculated from sagittal tissue sectionsthat were ±1.08 μm to ±1.92 μm from the midline. We did not quantify thearea of the dentate nuclei because our recordings in this study are predictedto be from the fastigial and interposed cerebellar nuclei. Nuclear densitieswere determined in ImageJ software by first converting the TIFF imagescontaining the cerebellar nuclei to a 16-bit resolution, then adjusting thethreshold so that the Nissl staining of the nuclei was accurately detected bythe ‘Analyze Particles’ plug-in. The subsequent nuclear counts were thendivided by 50 μm to obtain the average number of cell nuclei present in anygiven 50-μm area. For all of the anatomical features quantified, numericalresults are reported as the mean±s.e.m. Sample size was determined basedon the statistical criteria for significance in observations. As previous reportshave described morphological features of cells in the cerebellum as beingwell approximated by normal distributions (Ristanovic et al., 2011),pairwise comparisons of the quantifications of these anatomical featureswere performed with two-tailed Student’s t-tests.

This article is part of a special collection ‘A Guide to Using Neuromuscular DiseaseModels for Basic and Preclinical Studies’, which was launched in a dedicated issueguest edited by Annemieke Aartsma-Rus, Maaike van Putten and James Dowling.See related articles in this collection at http://dmm.biologists.org/collection/neuromuscular.

AcknowledgementsWe thank the members of the Sillitoe Laboratory for helpful discussions and criticalreading of the manuscript. We also thank the IDDRC Neuropathology Core forhelp with tissue staining and histology experiments, and for providing advice andmicrocopy expertise during the project.

Competing interestsThe authors declare no competing or financial interests.

Author contributionsConceptualization: T.L.S., L.N.M., M.A., R.V.S.; Methodology: T.L.S., L.N.M., M.A.,R.V.S.; Formal analysis: T.L.S., L.N.M.; Investigation: T.L.S., L.N.M., M.A., T.L.; Datacuration: T.L.S., L.N.M., M.A., T.L.; Writing - original draft: T.L.S., L.N.M., M.A.,R.V.S.; Writing - review & editing: T.L.S., L.N.M., M.A., R.V.S.; Funding acquisition:T.L.S., M.A., R.V.S.

FundingThis work was supported by funds from Baylor College of Medicine (BCM) and TexasChildren’s Hospital (to R.V.S.); a grant to R.V.S. from the Mrs. Clifford Elder WhiteGraham Endowed Research Fund; the Hamill Foundation; a BCM IDDRC ProjectDevelopment Award; a BCM IDDRC grant [U54HD083092] from the EuniceKennedy Shriver National Institute of Child Health and Human Development; theNational Center For Research Resources [C06RR029965]; and the National Instituteof Neurological Disorders and Stroke [R01 NS089664 and R01 NS100874] (toR.V.S.). T.L.S. was supported by a Ruth L. Kirschstein National Research ServiceAward [F31 NS095491] from the National Institute of Neurological Disorders andStroke. M.A. was supported by a postdoctoral award from the National AtaxiaFoundation (NAF).

Supplementary informationSupplementary information available online athttp://dmm.biologists.org/lookup/doi/10.1242/dmm.040840.supplemental

10

RESEARCH ARTICLE Disease Models & Mechanisms (2020) 13, dmm040840. doi:10.1242/dmm.040840

Disea

seModels&Mechan

isms

ReferencesAmoasii, L., Hildyard, J. C. W., Li, H., Sanchez-Ortiz, E., Mireault, A., Caballero,D., Harron, R., Stathopoulou, T.-R., Massey, C., Shelton, J. M. et al. (2018).Gene editing restores dystrophin expression in a canine model of Duchennemuscular dystrophy. Science 362, 86-91. doi:10.1126/science.aau1549

Anderson, J. L., Head, S. I., Rae, C. and Morley, J. W. (2002). Brain function inDuchenne muscular dystrophy. Brain 125, 4-13. doi:10.1093/brain/awf012

Anderson, J. L., Morley, J. W. and Head, S. I. (2010). Enhanced homosynapticLTD in cerebellar Purkinje cells of the dystrophic mdx mouse. Muscle Nerve 41,329-334. doi:10.1002/mus.21467

Apps, R., Hawkes, R., Aoki, S., Bengtsson, F., Brown, A. M., Chen, G., Ebner,T. J., Isope, P., Jorntell, H., Lackey, E. P. et al. (2018). Cerebellar modules andtheir role as operational cerebellar processing units: a consensus paper.Cerebellum 17, 654-682. doi:10.1007/s12311-018-0952-3

Arancillo, M., White, J. J., Lin, T., Stay, T. L. and Sillitoe, R. V. (2015). In vivoanalysis of Purkinje cell firing properties during postnatal mouse development.J. Neurophysiol. 113, 578-591. doi:10.1152/jn.00586.2014

Barmack, N. H. and Yakhnitsa, V. (2008). Functions of interneurons in mousecerebellum. J. Neurosci. 28, 1140-1152. doi:10.1523/JNEUROSCI.3942-07.2008

Benson, M. A., Newey, S. E., Martin-Rendon, E., Hawkes, R. and Blake, D. J.(2001). Dysbindin, a novel coiled-coil-containing protein that interacts with thedystrobrevins in muscle and brain. J. Biol. Chem. 276, 24232-24241. doi:10.1074/jbc.M010418200

Birnkrant, D. J., Bushby, K., Bann, C. M., Alman, B. A., Apkon, S. D., Blackwell,A., Case, L. E., Cripe, L., Hadjiyannakis, S., Olson, A. K. et al. (2018).Diagnosis and management of Duchenne muscular dystrophy, part 2: respiratory,cardiac, bone health, and orthopaedic management. Lancet. Neurol. 17, 347-361.doi:10.1016/S1474-4422(18)30025-5

Blake, D. J., Hawkes, R., Benson, M. A. and Beesley, P. W. (1999). Differentdystrophin-like complexes are expressed in neurons and glia. J. Cell Biol. 147,645-658. doi:10.1083/jcb.147.3.645

Bostan, A. C., Dum, R. P. and Strick, P. L. (2013). Cerebellar networks with thecerebral cortex and basal ganglia. Trends Cogn. Sci. 17, 241-254. doi:10.1016/j.tics.2013.03.003

Brown, A. M., Arancillo, M., Lin, T., Catt, D. R., Zhou, J., Lackey, E. P., Stay, T. L.,Zuo, Z., White, J. J. and Sillitoe, R. V. (2019). Molecular layer interneuronsshape the spike activity of cerebellar Purkinje cells. Sci. Rep. 9, 1742. doi:10.1038/s41598-018-38264-1

Buckner, R. L. (2013). The cerebellum and cognitive function: 25 years of insightfrom anatomy and neuroimaging. Neuron 80, 807-815. doi:10.1016/j.neuron.2013.10.044

Burghes, A. H. M., Logan, C., Hu, X., Belfall, B., Worton, R. G. and Ray, P. N.(1987). A cDNA clone from the Duchenne/Becker muscular dystrophy gene.Nature 328, 434-437. doi:10.1038/328434a0

Calderon, D. P., Fremont, R., Kraenzlin, F. andKhodakhah, K. (2011). The neuralsubstrates of rapid-onset Dystonia-Parkinsonism. Nat. Neurosci. 14, 357-365.doi:10.1038/nn.2753

Carta, I., Chen, C. H., Schott, A. L., Dorizan, S. and Khodakhah, K. (2019).Cerebellar modulation of the reward circuitry and social behavior. Science. 363,eaav0581. doi:10.1126/science.aav0581

Cendelin, J. (2014). From mice to men: lessons from mutant ataxic mice.Cerebellum Ataxias 1, 4. doi:10.1186/2053-8871-1-4

Chen, C. H., Fremont, R., Arteaga-Bracho, E. E. and Khodakhah, K. (2014).Short latency cerebellar modulation of the basal ganglia. Nat. Neurosci. 17,1767-1775. doi:10.1038/nn.3868

Crone, M. and Mah, J. K. (2018). Current and emerging therapies for Duchennemuscular dystrophy. Curr. Treat. Options. Neurol. 20, 31. doi:10.1007/s11940-018-0513-6

Cyrulnik, S. E. and Hinton, V. J. (2008). Duchenne muscular dystrophy: acerebellar disorder? Neurosci. Biobehav. Rev. 32, 486-496. doi:10.1016/j.neubiorev.2007.09.001

D’Angelo, E. and Casali, S. (2013). Seeking a unified framework for cerebellarfunction and dysfunction: from circuit operations to cognition. Front. NeuralCircuits 6, 116. doi:10.3389/fncir.2012.00116

Davie, J. T., Clark, B. A. and Hausser, M. (2008). The origin of the complex spike incerebellar Purkinje cells. J. Neurosci. 28, 7599-7609. doi:10.1523/JNEUROSCI.0559-08.2008

De Zeeuw, C. I., Hoebeek, F. E., Bosman, L.W. J., Schonewille, M., Witter, L. andKoekkoek, S. K. (2011). Spatiotemporal firing patterns in the cerebellum. Nat.Rev. Neurosci. 12, 327-344. doi:10.1038/nrn3011

Demilly, A., Reeber, S. L., Gebre, S. A. and Sillitoe, R. V. (2011). Neurofilamentheavy chain expression reveals a unique parasagittal stripe topography in themouse cerebellum. Cerebellum 10, 409-421. doi:10.1007/s12311-010-0156-y

Din o, M. R., Willard, F. H. and Mugnaini, E. (1999). Distribution of unipolar brushcells and other calretinin immunoreactive components in the mammaliancerebellar cortex. J. Neurocytol. 28, 99-123. doi:10.1023/A:1007072105919

Falzarano, M. S., Scotton, C., Passarelli, C. and Ferlini, A. (2015). Duchennemuscular dystrophy: from diagnosis to therapy. Molecules 20, 18168-18184.doi:10.3390/molecules201018168

Fremont, R., Calderon, D. P., Maleki, S. and Khodakhah, K. (2014). Abnormalhigh-frequency burst firing of cerebellar neurons in rapid-onset dystonia-Parkinsonism. J. Neurosci. 34, 11723-11732. doi:10.1523/JNEUROSCI.1409-14.2014

Fremont, R., Tewari, A. and Khodakhah, K. (2015). Aberrant Purkinje cell activityis the cause of dystonia in a shRNA-based mouse model of rapid onset dystonia-Parkinsonism. Neurobiol. Dis. 82, 200-212. doi:10.1016/j.nbd.2015.06.004

Fremont, R., Tewari, A., Angueyra, C. and Khodakhah, K. (2017). A role forcerebellum in the hereditary dystonia DYT1. eLife 6, e22775. doi:10.7554/eLife.22775

Gao, Z., Todorov, B., Barrett, C. F., van Dorp, S., Ferrari, M. D., van denMaagdenberg, A. M. J. M., De Zeeuw, C. I. and Hoebeek, F. E. (2012).Cerebellar ataxia by enhanced CaV2.1 currents is alleviated by Ca2+-dependentK+-channel activators in Cacna1aS218L mutant mice. J. Neurosci. 32,15533-15546. doi:10.1523/JNEUROSCI.2454-12.2012

Gebre, S. A., Reeber, S. L. and Sillitoe, R. V. (2012). Parasagittalcompartmentation of cerebellar mossy fibers as revealed by the patternedexpression of vesicular glutamate transporters VGLUT1 and VGLUT2. BrainStruct. Funct. 217, 165-180. doi:10.1007/s00429-011-0339-4

Gorecki, D. C., Monaco, A. P., Derry, J. M. J., Walker, A. P., Barnard, E. A. andBarnard, P. J. (1992). Expression of four alternative dystrophin transcripts in brainregions regulated by different promoters. Mol. Genet. 1, 505-510. doi:10.1093/hmg/1.7.505

Grady, R.M.,Wozniak, D., Ohlemiller, K. andSanes, J. (2006). Cerebellar synapticdefects and abnormalmotor behavior inmice lacking alpha- and beta-dystrobrevin.J. Neurosci. 26, 2841-2851. doi:10.1523/JNEUROSCI.4823-05.2006

Gravel, C., Eisenman, L. M., Sasseville, R. and Hawkes, R. (1987). Parasagittalorganization of the rat cerebellar cortex: direct correlation between antigenicpurkinje cell bands revealed by mabQ113 and the organization of theolivocerebellar projection. J. Comp. Neurol. 265, 294-310. doi:10.1002/cne.902650211

Guiraud, S. and Davies, K. E. (2019). Regenerative biomarkers for Duchennemuscular dystrophy. Neural Regen. Res. 14, 1317-1320. doi:10.4103/1673-5374.253534

Guiraud, S., Aartsma-Rus, A., Vieira, N. M., Davies, K. E., van Ommen, G.-J. B.and Kunkel, L. M. (2015). The pathogenesis and therapy of musculardystrophies. Annu. Rev. Genomics Hum. Genet. 16, 281-308. doi:10.1146/annurev-genom-090314-025003

Hakim, C. H. and Duan, D. (2012). Gender differences in contractile and passiveproperties of mdx extensor digitorum longus muscle. Muscle Nerve 45, 250-256.doi:10.1002/mus.22275

Hansen, S. T., Meera, P., Otis, T. S. and Pulst, S. M. (2013). Changes in Purkinjecell firing and gene expression precede behavioral pathology in a mouse model ofSCA2. Hum. Mol. Genet. 22, 271-283. doi:10.1093/hmg/dds427

Hara, K. and Harris, R. A. (2002). The anesthetic mechanism of urethane: theeffects on neurotransmitter-gated ion channels. Anesth. Analg. 94, 313-318.doi:10.1213/00000539-200202000-00015

Heiney, S. A., Kim, J., Augustine, G. J. andMedina, J. F. (2014). Precise control ofmovement kinematics by optogenetic inhibition of Purkinje cell activity.J. Neurosci. 34, 2321-2330. doi:10.1523/JNEUROSCI.4547-13.2014

Hisano, S., Sawada, K., Kawano, M., Kanemoto, M., Xiong, G., Mogi, K., Sakata-Haga, H., Takeda, J., Fukui, Y. and Nogami, H. (2002). Expression of inorganicphosphate/vesicular glutamate transporters (BNPI/VGLUT1 and DNPI/VGLUT2)in the cerebellum and precerebellar nuclei of the rat. Brain Res. Mol. Brain Res.107, 23-31. doi:10.1016/S0169-328X(02)00442-4

Hoebeek, F. E., Khosrovani, S.,Witter, L. andDe Zeeuw, C. I. (2008). Purkinje cellinput to cerebellar nuclei in tottering: ultrastructure and physiology. Cerebellum 7,547-558. doi:10.1007/s12311-008-0086-0

Hoffman, E. P., Brown, R. H. and Kunkel, L. M. (1987). Dystrophin: the proteinproduct of the duchennemuscular dystrophy locus.Cell 51, 919-928. doi:10.1016/0092-8674(87)90579-4

Holt, G. R., Softky, W. R., Koch, C. and Douglas, R. J. (1996). Comparison ofdischarge variability in vitro and in vivo in cat visual cortex neurons.J. Neurophysiol. 75, 1806-1814. doi:10.1152/jn.1996.75.5.1806

Hoogland, G., Hendriksen, R. G. F., Slegers, R. J., Hendriks, M. P. H., Schijns,O. E. M. G., Aalbers, M. W. and Vles, J. S. H. (2019). The expression of the distaldystrophin isoforms Dp140 and Dp71 in the human epileptic hippocampus inrelation to cognitive functioning. Hippocampus 29, 102-110. doi:10.1002/hipo.23015

Horn, K. M., Pong, M. and Gibson, A. R. (2010). Functional relations of cerebellarmodules of the cat. J. Neurosci. 30, 9411-9423. doi:10.1523/JNEUROSCI.0440-10.2010

Hourde, C., Joanne, P., Noirez, P., Agbulut, O., Butler-Browne, G. and Ferry, A.(2013). Protective effect of female gender-related factors on muscle force-generating capacity and fragility in the dystrophic mdx mouse. Muscle Nerve 48,68-75. doi:10.1002/mus.23700

Imbornoni, L., Price, E. T., Andrews, J., Meaney, F. J., Ciafaloni, E. and Cunniff,C. (2014). Diagnostic and clinical characteristics of early-manifesting females withDuchenne or Becker muscular dystrophy. Am. J. Med. Genet. Part A 164,2769-2774. doi:10.1002/ajmg.a.36728

11

RESEARCH ARTICLE Disease Models & Mechanisms (2020) 13, dmm040840. doi:10.1242/dmm.040840

Disea

seModels&Mechan

isms

Inoue, T., Lin, X., Kohlmeier, K. A., Orr, H. T., Zoghbi, H. Y. and Ross, W. N.(2001). Calcium dynamics and electrophysiological properties of cerebellarPurkinje cells in SCA1 transgenic mice. J. Neurophysiol. 85, 1750-1760. doi:10.1152/jn.2001.85.4.1750

Kostadinov, D., Beau, M., Pozo, M. B. and Hausser, M. (2019). Predictive andreactive reward signals conveyed by climbing fiber inputs to cerebellar Purkinjecells. Nat. Neurosci. 22, 950-962. doi:10.1038/s41593-019-0381-8

Koziol, L. F., Budding, D., Andreasen, N., D’Arrigo, S., Bulgheroni, S., Imamizu,H., Ito, M., Manto, M., Marvel, C., Parker, K. et al. (2014). Consensus paper: thecerebellum’s role in movement and cognition. Cerebellum 13, 151-177. doi:10.1007/s12311-013-0511-x

Krasowska, E., Zabłocki, K., Gorecki, D. C. and Swinny, J. D. (2014). Aberrantlocation of inhibitory synaptic marker proteins in the hippocampus of dystrophin-deficient mice: Implications for cognitive impairment in Duchenne musculardystrophy. PLoS ONE 9, e108364. doi:10.1371/journal.pone.0108364

Kueh, S. L. L., Dempster, J., Head, S. I. and Morley, J. W. (2011). Reducedpostsynaptic GABAA receptor number and enhanced gaboxadol induced changein holding currents in Purkinje cells of the dystrophin-deficient mdx mouse.Neurobiol. Dis. 43, 558-564. doi:10.1016/j.nbd.2011.05.002

Kuo, S.-H., Louis, E. D., Faust, P. L., Handforth, A., Chang, S.-Y., Avlar, B., Lang,E. J., Pan, M.-K., Miterko, L. N., Brown, A. M. et al. (2019). Current opinions andconsensus for studying tremor in animal models. Cerebellum. doi:10.1007/s12311-019-01037-1

Lalonde, R. and Strazielle, C. (2007). Brain regions and genes affecting posturalcontrol. Prog. Neurobiol. 81, 45-60. doi:10.1016/j.pneurobio.2006.11.005

Landis, D. and Landis, S. (1978). Several mutations in mice that affect thecerebellum. Adv. Neurol. 21, 85-105.

Larouche, M., Che, P. M. and Hawkes, R. (2006). Neurogranin expressionidentifies a novel array of Purkinje cell parasagittal stripes duringmouse cerebellardevelopment. J. Comp. Neurol. 494, 215-227. doi:10.1002/cne.20791

Lidov, H. G. W., Byers, T. J., Watkins, S. C. and Kunkel, L. M. (1990). Localizationof dystrophin to postsynaptic regions of central nervous system cortical neurons.Nature 348, 725-728. doi:10.1038/348725a0

Lidov, H. G.W., Byers, T. J. andKunkel, L. M. (1993). The distribution of dystrophinin the murine central nervous system: An immunocytochemical study.Neuroscience 54, 167-187. doi:10.1016/0306-4522(93)90392-S

Long, C., Amoasii, L., Mireault, A. A., McAnally, J. R., Li, H., Sanchez-Ortiz, E.,Bhattacharyya, S., Shelton, J. M., Bassel-Duby, R. and Olson, E. N. (2016).Postnatal genome editing partially restores dystrophin expression in a mousemodel of muscular dystrophy. Science 351, 400-403. doi:10.1126/science.aad5725

Luna-Cancalon, K., Sikora, K. M., Pappas, S. S., Singh, V., Wulff, H., Paulson,H. L., Burmeister, M. and Shakkottai, V. G. (2014). Alterations in cerebellarphysiology are associated with a stiff-legged gait in Atcayji-hes mice. Neurobiol.Dis. 67, 140-148. doi:10.1016/j.nbd.2014.03.020

Marek, S., Siegel, J. S., Gordon, E. M., Raut, R. V., Gratton, C., Newbold, D. J.,Ortega, M., Laumann, T. O., Adeyemo, B., Miller, D. B. et al. (2018). Spatial andtemporal organization of the individual human cerebellum. Neuron 100,977-93.e7. doi:10.1016/j.neuron.2018.10.010

McAfee, S. S., Liu, Y., Sillitoe, R. V. and Heck, D. H. (2019). Cerebellar lobulussimplex and crus I differentially represent phase and phase difference of prefrontalcortical and hippocampal oscillations. Cell Rep. 27, 2328-2334. doi:10.1016/j.celrep.2019.04.085

Mercer, A. A., Palarz, K. J., Tabatadze, N., Woolley, C. S. and Raman, I. M.(2016). Sex differences in cerebellar synaptic transmission and sex-specificresponses to autism-linked Gabrb3 mutations in mice. eLife 5, e07596. doi:10.7554/eLife.07596

Mercier, S., Toutain, A., Toussaint, A., Raynaud, M., De Barace, C., Marcorelles,P., Pasquier, L., Blayau, M., Espil, C., Parent, P. et al. (2013). Genetic andclinical specificity of 26 symptomatic carriers for dystrophinopathies at pediatricage. Eur. J. Hum. Genet. 21, 855. doi:10.1038/ejhg.2012.269

Min, Y.-L., Li, H., Rodriguez-Caycedo, C., Mireault, A. A., Huang, J., Shelton,J. M., McAnally, J. R., Amoasii, L., Mammen, P. P. A., Bassel-Duby, R. et al.(2019). CRISPR-Cas9 corrects Duchenne muscular dystrophy exon 44 deletionmutations in mice and human cells. Sci. Adv. 5, eaav4324. doi:10.1126/sciadv.aav4324

Miranda, R., Nagapin, F., Bozon, B., Laroche, S., Aubin, T. and Vaillend, C.(2015). Altered social behavior and ultrasonic communication in the dystrophin-deficient mdx mouse model of Duchenne muscular dystrophy. Mol. Autism 6, 60.doi:10.1186/s13229-015-0053-9

Miranda, R., Laroche, S. and Vaillend, C. (2016). Reduced neuronal density in theCA1 anterodorsal hippocampus of the mdx mouse. Neuromuscul. Disord. 26,775-781. doi:10.1016/j.nmd.2016.08.006

Mirski, K. T. and Crawford, T. O. (2014). Motor and cognitive delay in Duchennemuscular dystrophy: implication for early diagnosis. J. Pediatr. 165, 1008-1010.doi:10.1016/j.jpeds.2014.07.006

Miterko, L. N., White, J. J., Lin, T., Brown, A. M., O’Donovan, K. J. and Sillitoe,R. V. (2019a). Persistent motor dysfunction despite homeostatic rescue ofcerebellar morphogenesis in the Car8 waddles mutant mouse. Neural Dev. 14, 6.doi:10.1186/s13064-019-0130-4

Miterko, L. N., Sillitoe, R. V. and Hawkes, R. (2019b). Zones and stripes:Development of cerebellar topography. In Handbook of the Cerebellum andCerebellar Disorders (ed. M.Manto, D. Gruol, J. Schmahmann, N. Koibuchi and F.Rossi), pp. 1-23. Cham: Springer. doi:10.1007/978-3-319-97911-3_3-2

Monaco, A. P., Neve, R. L., Colletti-Feener, C., Bertelson, C. J., Kurnit, D. M. andKunkel, L. M. (1986). Isolation of candidate cDNAs for portions of the Duchennemuscular dystrophy gene. Nature 323, 646-650. doi:10.1038/323646a0

Moser, H. and Emery, A. E. H. (1974). The manifesting carrier in Duchennemuscular dystrophy. Clin. Genet. 5, 271-284. doi:10.1111/j.1399-0004.1974.tb01694.x

Muntoni, F., Torelli, S. and Ferlini, A. (2003). Dystrophin and mutations: one gene,several proteins, multiple phenotypes. Lancet Neurol. 2, 731-740. doi:10.1016/S1474-4422(03)00585-4

Norman, A. and Harper, P. (1989). A survey of manifesting carriers of Duchenneand Becker muscular dystrophy in Wales. Clin. Genet. 36, 31-37. doi:10.1111/j.1399-0004.1989.tb03363.x

Papa, R., Madia, F., Bartolomeo, D., Trucco, F., Pedemonte, M., Traverso, M.,Broda, P., Bruno, C., Zara, F., Minetti, C. et al. (2016). Genetic and early clinicalmanifestations of females heterozygous for Duchenne/Becker musculardystrophy. Pediatr. Neurol. 55, 58-63. doi:10.1016/j.pediatrneurol.2015.11.004

Paukert, M., Huang, Y. H., Tanaka, K., Rothstein, J. D. and Bergles, D. E. (2010).Zones of enhanced glutamate release from climbing fibers in the mammaliancerebellum. J. Neurosci. 30, 7290-7299. doi:10.1523/JNEUROSCI.5118-09.2010

Paxinos, G. and Franklin, K. (2001). The Mouse Brain in Stereotaxic Coordinates.Elsevier Academic Press.

Pereira da Silva, J. D., Campos, D. V., Nogueira-Bechara, F. M., Stilhano, R. S.,Han, S. W., Sinigaglia-Coimbra, R., Lima-Landman, M. T. R., Lapa, A. J. andSouccar, C. (2018). Altered release and uptake of gamma-aminobutyric acid inthe cerebellum of dystrophin-deficient mice. Neurochem. Int. 118, 105-114.doi:10.1016/j.neuint.2018.06.001

Person, A. L. and Raman, I. M. (2011). Purkinje neuron synchrony elicits time-locked spiking in the cerebellar nuclei. Nature 481, 502-505. doi:10.1038/nature10732

Pilgram,G. S. K., Potikanond, S., Baines, R. A., Fradkin, L. G. andNoordermeer,J. N. (2010). The roles of the dystrophin-associated glycoprotein complex at thesynapse. Mol. Neurobiol. 41, 1-21. doi:10.1007/s12035-009-8089-5

Rae, M. G. and O’Malley, D. (2016). Cognitive dysfunction in Duchenne musculardystrophy: a possible role for neuromodulatory immune molecules.J. Neurophysiol. 116, 1304-1315. doi:10.1152/jn.00248.2016

Reeber, S. L. and Sillitoe, R. V. (2011). Patterned expression of a cocaine- andamphetamine-regulated transcript peptide reveals complex circuit topography inthe rodent cerebellar cortex. J. Comp. Neurol. 519, 1781-1796. doi:10.1002/cne.22601

Reeber, S. L., Otis, T. S. and Sillitoe, R. V. (2013). New roles for the cerebellum inhealth and disease. Front. Syst. Neurosci. 7, 83. doi:10.3389/fnsys.2013.00083

Ristanovic, D., Milosevic, N. T. and Maric, D. L. (2011). On the classification ofnormally distributed neurons: An application to human dentate nucleus. Biol.Cybern. 104, 175083. doi:10.1007/s00422-011-0426-x

Ruigrok, T. J. H., Hensbroek, R. A. and Simpson, J. I. (2011). Spontaneousactivity signatures of morphologically identified interneurons in thevestibulocerebellum. J. Neurosci. 31, 712-724. doi:10.1523/JNEUROSCI.1959-10.2011

Rybalka, E., Timpani, C. A., Cooke, M. B., Williams, A. D. and Hayes, A. (2014).Defects in mitochondrial ATP synthesis in dystrophin-deficient mdx skeletalmuscles may be caused by complex I insufficiency. PLoS ONE 9, e115763.doi:10.1371/journal.pone.0115763

Ryder-Cook, A. S., Sicinski, P., Thomas, K., Davies, K. E., Worton, R. G.,Barnard, E. A., Darlison, M. G. and Barnard, P. J. (1988). Localization of themdx mutation within the mouse dystrophin gene. EMBO J. 7, 3017-3021. doi:10.1002/j.1460-2075.1988.tb03165.x

Salimena, M. C., Lagrota-Candido, J. and Quırico-Santos, T. (2000). Genderdimorphism influences extracellular matrix expression and regeneration ofmuscular tissue in mdx dystrophic mice. Histochem. Cell Biol. 122, 435-444.doi:10.1007/s00418-004-0707-8

Schmahmann, J. D. (2019). The cerebellum and cognition. Neurosci. Lett. 688,62-75. doi:10.1016/j.neulet.2018.07.005

Schmahmann, J. D., Guell, X., Stoodley, C. J. andHalko,M. A. (2019). The theoryand neuroscience of cerebellar cognition. Annu. Rev. Neurosci. 42, 337-364.doi:10.1146/annurev-neuro-070918-050258

Schmolesky, M., Weber, J., De Zeeuw, C. and Hansel, C. (2002). The making of acomplex spike: ionic composition and plasticity. Ann. N. Y. Acad. Sci. 978,359-390. doi:10.1111/j.1749-6632.2002.tb07581.x

Schonewille, M., Belmeguenai, A., Koekkoek, S. K., Houtman, S. H., Boele,H. J., van Beugen, B. J., Gao, Z., Badura, A., Ohtsuki, G., Amerika, W. E. et al.(2010). Purkinje cell-specific knockout of the protein phosphatase PP2B impairspotentiation and cerebellar motor learning. Neuron 67, 618-628. doi:10.1016/j.neuron.2010.07.009

Sesay, A. K., Errington, M. L., Levita, L. and Bliss, T. V. P. (1996). Spatial learningand hippocampal long-term potentiation are not impaired in mdx mice. Neurosci.Lett. 211, 207-210. doi:10.1016/0304-3940(96)12747-6

12

RESEARCH ARTICLE Disease Models & Mechanisms (2020) 13, dmm040840. doi:10.1242/dmm.040840

Disea

seModels&Mechan

isms

Shakkottai, V. G., Batla, A., Bhatia, K., Dauer, W. T., Dresel, C., Niethammer, M.,Eidelberg, D., Raike, R. S., Smith, Y., Jinnah, H. A. et al. (2017). Currentopinions and areas of consensus on the role of the cerebellum in dystonia.Cerebellum 16, 577-594. doi:10.1007/s12311-016-0825-6

Shipman, M. L. and Green, J. T. (2019). Cerebellum and cognition: does the rodentcerebellum participate in cognitive functions? Neurobiol. Learn. Mem. S1074-7427(19)30027-9. doi:10.1016/j.nlm.2019.02.006

Sicinski, P., Geng, Y., Ryder-Cook, A. S., Barnard, E. A., Darlison, M. G. andBarnard, P. J. (1989). The molecular basis of muscular dystrophy in the mdxmouse: a point mutation. Science. 244, 1578-1580. doi:10.1126/science.2662404

Sillitoe, R. V., Benson, M. A., Blake, D. J. and Hawkes, R. (2003). Abnormaldysbindin expression in cerebellar mossy fiber synapses in the mdx mousemodelof Duchenne muscular dystrophy. J. Neurosci. 23, 6576-6585. doi:10.1523/JNEUROSCI.23-16-06576.2003

Sillitoe, R. V., Stephen, D., Lao, Z. and Joyner, A. L. (2008). Engrailed homeoboxgenes determine the organization of Purkinje cell sagittal stripe gene expressionin the adult cerebellum. J. Neurosci. 28, 12150-12162. doi:10.1523/JNEUROSCI.2059-08.2008

Sillitoe, R. V., Vogel, M. W. and Joyner, A. L. (2010). Engrailed homeobox genesregulate establishment of the cerebellar afferent circuit map. J. Neurosci. 30,10015-10024. doi:10.1523/JNEUROSCI.0653-10.2010

Singec, I., Knoth, R., Ditter, M., Frotscher, M. and Volk, B. (2003). Neurograninexpression by cerebellar neurons in rodents and non-human primates. J. Comp.Neurol. 459, 278-289. doi:10.1002/cne.10600

Snow, W. M., Anderson, J. E. and Fry, M. (2014). Regional and genotypicdifferences in intrinsic electrophysiological properties of cerebellar Purkinjeneurons from wild-type and dystrophin-deficient mdx mice. Neurobiol. Learn.Mem. 107, 19-31. doi:10.1016/j.nlm.2013.10.017

Solinas, S. M. G., Maex, R. and De Schutter, E. (2006). Dendritic amplification ofinhibitory postsynaptic potentials in a model Purkinje cell. Eur. J. Neurosci. 23,1207-1218. doi:10.1111/j.1460-9568.2005.04564.x

Stay, T. L., Laurens, J., Sillitoe, R. V. and Angelaki, D. E. (2019). Geneticallyeliminating Purkinje neuron GABAergic neurotransmission increases theirresponse gain to vestibular motion. Proc. Natl. Acad. Sci. USA 116, 3245-3250.doi:10.1073/pnas.1818819116

Stoodley, C. J. (2012). The Cerebellum and cognition: evidence from functionalimaging studies. Cerebellum 11, 352-365. doi:10.1007/s12311-011-0260-7

Sugihara, I. and Quy, P. N. (2007). Identification of aldolase C compartments in themouse cerebellar cortex by olivocerebellar labeling. J. Comp. Neurol. 500,1076-1092. doi:10.1002/cne.21219

Suzuki, Y., Higuchi, S., Aida, I., Nakajima, T. and Nakada, T. (2017). Abnormaldistribution of GABAA receptors in brain of duchenne muscular dystrophypatients. Muscle Nerve 55, 591-595. doi:10.1002/mus.25383

Tang, T., Blenkinsop, T. A. and Lang, E. J. (2019). Complex spike synchronydependent modulation of rat deep cerebellar nuclear activity. eLife 8, e40101.doi:10.7554/eLife.40101

Taylor, P. J., Maroulis, S., Mullan, G. L., Pedersen, R. L., Baumli, A., Elakis, G.,Piras, S., Walsh, C., Prosper-Gutierrez, B., De La Puente-Alonso, F. et al.(2007). Measurement of the clinical utility of a combined mutation detectionprotocol in carriers of Duchenne and Becker muscular dystrophy. J. Med. Genet.44, 368-372. doi:10.1136/jmg.2006.047464

Thach, W. T. (2014). Does the cerebellum initiate movement? Cerebellum 13,139-150. doi:10.1007/s12311-013-0506-7

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. et al. (2012). Purkinje cell-specific ablation of CaV2.1 Channels is sufficient to cause cerebellar ataxia inmice. Cerebellum 11, 246-258. doi:10.1007/s12311-011-0302-1

Tsai, P. T., Hull, C., Chu, Y. X., Greene-Colozzi, E., Sadowski, A. R., Leech, J. M.,Steinberg, J., Crawley, J. N., Regehr, W. G. and Sahin, M. (2012). Autistic-like

behaviour and cerebellar dysfunction in Purkinje cell Tsc1 mutant mice. Nature488, 647-651. doi:10.1038/nature11310

Vaillend, C., Billard, J.-M. and Laroche, S. (2004). Impaired long-term spatial andrecognition memory and enhanced CA1 hippocampal LTP in the dystrophin-deficient Dmdmdxmouse.Neurobiol. Dis. 17, 10-20. doi:10.1016/j.nbd.2004.05.004

Wadiche, J. I. and Jahr, C. E. (2005). Patterned expression of Purkinje cellglutamate transporters controls synaptic plasticity. Nat. Neurosci. 8, 1329-1334.doi:10.1038/nn1539

Wagner, M. J., Kim, T. H., Savall, J., Schnitzer, M. J. and Luo, L. (2017).Cerebellar granule cells encode the expectation of reward. Nature 544, 96-100.doi:10.1038/nature21726

Wagner, M. J., Kim, T. H., Kadmon, J., Nguyen, N. D., Ganguli, S., Schnitzer,M. J. and Luo, L. (2019). Shared cortex-cerebellum dynamics in the executionand learning of amotor task.Cell 177, 669-82.e24. doi:10.1016/j.cell.2019.02.019

Waite, A., Tinsley, C. L., Locke,M. andBlake, D. J. (2009). The neurobiologyof thedystrophin-associated glycoprotein complex. Ann. Med. 41, 344-359. doi:10.1080/07853890802668522

Welsh, J. P., Lang, E. J., Suglhara, I. and Llinas, R. (1995). Dynamic organizationof motor control within the olivocerebellar system. Nature 374, 453-457. doi:10.1038/374453a0

White, J. J. and Sillitoe, R. V. (2013). Postnatal development of cerebellar zonesrevealed by neurofilament heavy chain protein expression. Front. Neuroanat. 7, 9.doi:10.3389/fnana.2013.00009

White, J. J. and Sillitoe, R. V. (2017). Genetic silencing of olivocerebellar synapsescauses dystonia-like behaviour in mice. Nat. Commun. 8, 14912. doi:10.1038/ncomms14912

White, J. J., Arancillo, M., Stay, T. L., George-Jones, N. A., Levy, S. L., Heck,D. H. and Sillitoe, R. V. (2014). Cerebellar zonal patterning relies on Purkinje cellneurotransmission. J. Neurosci. 34, 8231-8245. doi:10.1523/JNEUROSCI.0122-14.2014

White, J. J., Arancillo, M., King, A., Lin, T., Miterko, L. N., Gebre, S. A. andSillitoe, R. V. (2016a). Pathogenesis of severe ataxia and tremor without thetypical signs of neurodegeneration. Neurobiol. Dis. 86, 86-98. doi:10.1016/j.nbd.2015.11.008

White, J. J., Lin, T., Brown, A. M., Arancillo, M., Lackey, E. P., Stay, T. L. andSillitoe, R. V. (2016b). An optimized surgical approach for obtaining stableextracellular single-unit recordings from the cerebellum of head-fixed behavingmice. J. Neurosci. Methods 262, 21-31. doi:10.1016/j.jneumeth.2016.01.010

Xiao, J., Cerminara, N. L., Kotsurovskyy, Y., Aoki, H., Burroughs, A., Wise,A. K., Luo, Y., Marshall, S. P., Sugihara, I., Apps, R. et al. (2014). Systematicregional variations in Purkinje cell spiking patterns. PLoS ONE 9, e105633.doi:10.1371/journal.pone.0105633

Xu, L., Lau, Y. S., Gao, Y., Li, H. and Han, R. (2019). Life-long AAV-mediatedCRISPR genome editing in dystrophic heart improves cardiomyopathy withoutcausing serious lesions in mdx mice. Mol. Ther. 27, 1407-1414. doi:10.1016/j.ymthe.2019.05.001

Xu-Wilson, M., Chen-Harris, H., Zee, D. S. and Shadmehr, R. (2009). Cerebellarcontributions to adaptive control of saccades in humans. J. Neurosci. 29,12930-12939. doi:10.1523/JNEUROSCI.3115-09.2009

Yoon, S. H., Grynpas, M. D. and Mitchell, J. (2019). Growth hormone increasesbone toughness and decreases muscle inflammation in glucocorticoid-treatedmdx mice, model of Duchenne muscular dystrophy. J. Bone Miner. Res. 34,1473-1486. doi:10.1002/jbmr.3718

Zhou, H., Lin, Z., Voges, K., Ju, C., Gao, Z., Bosman, L. W. J., Ruigrok, T. J. H.,Hoebeek, F. E., De Zeeuw, C. I. and Schonewille, M. (2014). Cerebellarmodules operate at different frequencies. eLife 3, e02536. doi:10.7554/eLife.02536

Zwergal, A., La Fougere, C., Lorenzl, S., Rominger, A., Xiong,G., Deutschenbaur,L., Schoberl, F., Linn, J., Dieterich, M., Brandt, T. et al. (2013). Functionaldisturbance of the locomotor network in progressive supranuclear palsy. Neurology80, 634-641. doi:10.1212/WNL.0b013e318281cc43

13

RESEARCH ARTICLE Disease Models & Mechanisms (2020) 13, dmm040840. doi:10.1242/dmm.040840

Disea

seModels&Mechan

isms