pi is 0896627311004806

8/6/2019 Pi is 0896627311004806

1/15

Neuron

Article

Ank3-Dependent SVZ Niche Assembly Is Requiredfor the Continued Production of New NeuronsPatricia Paez-Gonzalez, 1 , 2 Khadar Abdi, 1 ,2 Dominic Luciano, 3 Yan Liu, 1 Mario Soriano-Navarro, 4 Emma Rawlins, 1 ,8 Vann Bennett, 1 ,5 Jose Manuel Garcia-Verdugo, 4 and Chay T. Kuo 1 ,2 ,3 ,6 ,7 ,*1 Department of Cell Biology2 Brumley Neonatal-Perinatal Research Institute, Department of Pediatrics3 Department of NeurobiologyDuke University Medical Center, Durham, NC 27710, USA 4 Laboratorio de Morfolog a Celular, Unidad mixta CIPF-UVEG, CIBERNED. Valencia, Spain5 Howard Hughes Medical Institute6 Preston Robert Tisch Brain Tumor CenterDuke University Medical Center, Durham, NC 27710, USA 7 Duke University Institute for Brain Sciences, Durham, NC 27708, USA 8 Present address: The Gurdon Institute, University of Cambridge, CB2 1QN, UK*Correspondence: [email protected] 10.1016/j.neuron.2011.05.029

SUMMARY

The rodent subventricular/subependymal zone (SVZ/ SEZ) houses neural stem cells (NSCs) that generateolfactory bulb interneurons. It is unclear how theSVZ environment sustains neuronal production intoadulthood. We discovered that the adapter molecule Ankyrin-3 (Ank3) is specically upregulated in ven-tricular progenitors destined to become ependymalcells, but not in NSCs, and is required for SVZniche assembly through progenitor lateral adhesion.Furthermore, we found that Ank3 expression is con-trolled by Foxj1, a transcriptional regulator of multici-lia formation, and geneticdeletion of this pathway ledto complete loss of SVZ niche structure. Interest-ingly, radial glia continued to transition into postnatalNSCs without this niche. However, inducible dele-tion of Foxj1-Ank3 from mature SVZ ependyma re-sulted in dramatic depletion of neurogenesis. Target-ing a pathway regulating ependymal organization/ assembly and showing its requirement for newneuron production, our results have important impli-cations for environmental control of adult neurogen-esis and harvesting NSCs for replacement therapy.

INTRODUCTION

Throughout embryonic and postnatal development, neuralprogenitors/stem cells give rise to differentiated neurons, astro-cytes, and oligodendrocytes ( Go tz and Huttner, 2005; Kokovayet al., 2008; Kriegstein and Alvarez-Buylla, 2009 ). While theseprogenitors are relatively abundant during embryogenesis, theybecome restricted to specialized regions/niches in the adultbrain, including the subventricular/subependymal zone (SVZ/ SEZ) along the lateral walls of lateral brain ventricles, as well as

the subgranular zone in the dentate gyrus of the hippocampus( Miller and Gauthier-Fisher, 2009; Suh et al., 2009 ). Adult neuro-genesis in the rodent SVZ is mediated by type B astrocytesfunctioning as neural stem cells (NSCs) ( Doetsch et al., 1999 ),which in turn differentiate into neuroblasts that migrate andincorporate into the mouse olfactory bulb (OB) as interneurons( Lledo et al., 2008 ). This source of new neurons provides a keyexperimental system for studying neuronal integration into func-tional circuits ( Kelsch et al., 2010 ), as well as holding promisingtherapeutic potential. However, the exact mechanisms allowingfor continuation of neurogenesis into adulthood in this brainregion are not well understood.

NSCs in the adult SVZ exist in a dedicated environment that iscomprisedmainly of multiciliated ependymalcells on the ventric-ular surface, as well as a specialized vascular network ( Alvarez-Buylla and Lim, 2004 ). Arrangement of this niche is spatiallydened, in that ependymal cells are organized in a pinwheel-like fashion surrounding monociliated NSCs touching the ven-tricular surface ( Mirzadeh et al., 2008 ). In addition, SVZ NSCsextend basal processes that terminate on blood vessels that liebeneath the ependymal layer ( Shen et al., 2008; Tavazoieet al., 2008 ). The SVZ niche is a rich source for growth factorsand specialized cell-cell interactions that maintain NSC homeo-stasis in vivo ( Miller and Gauthier-Fisher, 2009; Kokovay et al.,2010 ), and it can respond to environmental challenges by modi-

fying the proliferative/differentiation capacities of NSCs ( Kuoet al., 2006; Luo et al., 2008; Carle n et al., 2009 ). Despite thisunderstanding, there is no direct evidence that this denedSVZ architecture is required for the continued production of new neuronsdue largely to our inability to specically eliminatethe SVZ niche. We previously generated one of the rst induciblemouse modelsto postnatally disrupt SVZ architecture viaNumb/ Numblike deletion, revealing a local remodeling capacity ( Kuoet al., 2006 ); however, the ensuing ventriculomegaly precludedfunctional assessment on adult neurogenesis, as the ow of cerebrospinal uid has been shown to regulate migration of newborn neurons ( Sawamoto et al., 2006 ). These results pointto the importance of identifying SVZ niche-specic pathways

Neuron 71 , 6175, July 14, 2011 2011 Elsevier Inc. 61
mailto:[email protected]://dx.doi.org/10.1016/j.neuron.2011.05.029http://dx.doi.org/10.1016/j.neuron.2011.05.029mailto:[email protected]

8/6/2019 Pi is 0896627311004806

2/15

to allow for direct deletion of SVZ architecture without cell intrin-sically affecting NSCs.

Little is known about the molecular mechanisms controllingSVZ generation from embryonic progenitors. Shortly before

and after birth, while most embryonic radial glia terminally differ-entiate, postnatal radial glial progenitors (pRGPs) along thelateral walls of lateral ventricles generatethe SVZ niche ( Tramon-tin et al., 2003 ). The transformation from embryonic to adultneurogenesis is mediated by a subpopulation of pRGPs dif-ferentiating into SVZ NSCs ( Merkle et al., 2004 ). A secondsubpopulation of pRGPs gives rise to ependymal cells thatform the new epithelial lining of the brain ventricles, which alsoserve as multiciliated niche cells for the SVZ NSCs ( Spasskyet al., 2005 ). We showed previously that during terminal differen-tiation of pRGPs, progenitors begin to modify their lateralmembrane contacts ( Kuo et al., 2006 ). The Ankyrin family of proteins in mammals, consisting of Ankyrin R (1, Ank1), B (2, Ank2), andG (3, Ank3), arelargeadaptormolecules that organize

membrane domains in a number of different cell types, includingerythrocytes, cardiac and skeletal muscles, epithelial cells,retinal photoreceptors, and neuronal axon initial segments ( Ben-nett and Healy, 2008 ). Although Ankyrins and their homologs inother model organisms have not been linked to stem cell nichefunctions, Ank3 is known to regulate lateral membrane biogen-esis of bronchial epithelial cells, through collaborative interac-tions with b2-Spectrin and a -Adducin ( Kizhatil and Bennett,2004; Abdi and Bennett, 2008 ).

Using in vivo-inducible genetics and newly developed in vitroassays, we revealed a function for Ank3 and its upstream regu-lator in radial glial assembly of adult SVZ niche, which upondisruption led to the complete absence of SVZ ependymal nichein vivo. The revelation of these key early molecular steps allowedus to address fundamental questions about SVZ organization oncontinued production of new neurons.

RESULTS

Selective Ank3 Upregulation in Subpopulation of pRGPsduring SVZ FormationSince the SVZ niche is formed during postnatal maturation of thebrain ventricular wall, we performed surface-scanning electronmicroscopy and transmission electron microscopy (TEM) onmouse brains from postnatal days 0, 7, and 14 to observeanatomical changes (P0, P7, and P14, respectively; see Fig-ure S1 A available online). Unlike the medial wall surface, which

showed abundant multiciliated cells throughout, at P0 the cellson the lateral wall were predominantly monociliated and gradu-ally became multiciliated over the next 2 weeks. Concurrently,there was considerable specialization of cellular lateralmembranes, as pRGP contacts transitioned from simple apicaladherens junctions at P0, to mature interdigitating membranesbetween ependymal cells with mixtures of adherens and tight junctions by P14 ( Figure S1 B). These developmental processestake place during the critical period when the SVZ neurogenicniche is assembled from pRGPs.

We wondered if specialization of progenitor lateral mem-branes is required for SVZ niche assembly. Within the Ankyrinfamily of proteins, Ank1 is largely erythroid specic; therefore,

we looked for Ank2 and Ank3 expressions within the SVZ niche.Immunohistochemical (IHC) staining with Ank2- and Ank3-specic antibodies on P14 brain lateral ventricular whole mountsdemonstrated that while Ank2 was ubiquitously expressed,

Ank3 was specically localized to the lateral cellular borders of multiciliated ependymal cells, but not to the monociliated cellsat the center of pinwheel-like niche structures containing NSCs( Figure 1 A).

To understand the selectivity of this Ank3 expression duringpostnatal SVZ formation, we performed IHC staining on ventric-ular wall whole mounts at different postnatal stages from Foxj1-GFP mice ( Ostrowski et al., 2003 ). The Foxj1-GFP reporter linehas been shown to efciently label ependymal cells and theirprogenitors ( Zhang et al., 2007; Jacquet et al., 2009 ). At P0, all Ank3 + cells colocalized with GFP, although some GFP-dim cellswere Ank3

( Figure 1 B).However, by P14, Ank3 showed specic

colocalization with Foxj1-GFP labeling, and the protein becamehighly concentrated at thelateral cellular borders ( Figure 1 B). We

next imaged in the x-z plane and conrmed that as Foxj1-GFP +pRGPs transitioned from radial glia to fullydifferentiated ependy-mal cells, Ank3 became enriched at lateral cellular borders ( Fig-ure 1 C). Since Ank3 is expressed by mature neurons and en-riched in the axon initial segment, as expected on coronalsections from P14 brains, we sawhigh-level expression in striatalneurons ( Figure 1 D). However, within the SVZ, just beneath theS100 b + Ank3 + ependymal layer, we did not detect Ank3 expres-sion ( Figure 1 D, asterisks). Costaining with GFAP and DCX anti-bodies showed that both astrocytes and newborn neuroblastswithin the SVZ were Ank3

( Figure 1 D and Figure S1 C). These

results revealed Ank3 as one of the rst molecules identiedthat is differentially expressed on the lateral membranesbetween SVZ NSCs and ependymal niche cells.

Self-Organization of pRGPs into Ependymal ClustersDue to the complexity of the ank3 genomic locus (the 190 kDasplice form alone has 43 exons spanning roughly 500 kb of genomic sequence), previous attempts to generate ank3 -de-cient mice have only resulted in splice form-specic deletionsin cerebellar neurons ( Zhou et al., 1998 ). To determine the func-tion for Ank3 during SVZ neurogenesis, we set out to developa primary cell culture assay to generate SVZ ependymal nichecells. Using a modied protocol based on existing methods( Prothmann et al., 2001; Guirao et al., 2010 ), we were able todifferentiate pRGPs from P0 mice into large numbers of multici-liated ependymal cells with beating cilia ( Movie S1 ). IHC staining

against g -tubulin showed the transition from single to multiplebasal bodies in multiciliated cells after differentiation ( Figure 2 A).These pRGPs also progressed from Glast hi /S100 b lo to Glast lo / S100 b hi during differentiation ( Figure 2 A), consistent with ourprevious observation for postnatal ependymal differentiationin vivo ( Kuo et al., 2006 ). Quantifying g -tubulin cluster stainingas a percentage of DAPI nuclear staining, 64% 6.5% standarddeviation (SD) of total cells after completion of differentiationbecame multiciliated.

We noticed that differentiated ependymal cells often clusteredin culture, and scanning electron microscopy analyses showedmulticiliated cells arranged in clusters around monociliated cells( Figure 2 B and Figure S2 A), much like their in vivo organization

Neuron Ank3 Controls Continued SVZ Neuron Production

62 Neuron 71 , 6175, July 14, 2011 2011 Elsevier Inc.
http://-/?-http://-/?-

8/6/2019 Pi is 0896627311004806

3/15

( Mirzadeh et al., 2008 ). The monociliated cells within ependymalclusters were also GFAP + in culture ( Figure 2 C and Figure S2 B).To understand theprocesses leading to pRGP clustering in vitro,we performed live cell imaging using the Foxj1-GFP reportermouse line ( Ostrowski et al., 2003 ). We observed that shortlyafter plating, there was an increase in the number of GFP +

pRGPs, followed by upregulation of GFP expression and cellularclustering ( Movie S2 ). Expansion in the number of GFP + cellswas accomplished by both progenitors starting to expressGFP as well as by cell division ( Movie S3 ). It is interesting tonote that the majority of GFP + pRGP clustering took place34days after plating, prior to theappearance of multicilia, whichbegan 78 days after plating ( Movie S2 and Figure 2 D). Once the

clustering was complete, these structures were positionallystable ( Movie S2 ). IHC staining revealed that the clusters repre-sented multiciliated Foxj1-GFP + arranged around monociliatedGFP

cells ( Figure S2 C).

Ank3 Is Required for Proper SVZ Niche AssemblyTo understand if this pRGP culture may be useful for tackling Ank3 function, we saw that during in vitro differentiation, pRGPclusters upregulated Ank3 along their cell borders, prior to multi-cilia formation ( Figure 3 A). Western blot analyses of pRGPcultures conrmed robust increase in the 190 kDa Ank3 protein(known to be an epithelial-specic splice form) ( Kizhatil et al.,2007 ) after differentiation, as well as Ank3-associated proteins

Figure 1. Ank3 Expression in SVZ Neurogenic Niche(A) IHC stainingof P14 lateral ventricular wallwhole mounts showing the specicity of Ank3expression in multiciliated ependymal cells, but not monociliated cells(arrows). Regions within yellow boxes enlarged below.(B)Ank3and GFPstaining of brainlateralventricular wall wholemounts from P0,P7, andP14 Foxj1-GFPtransgenicmice. AtP0, therewereGFP + pRGPs (smallerventricular surface area) with low Ank3 expression (*). Ph, Phalloidin.(C) IHC staining of ventricular wall whole mounts in x-z plane, showing upregulation/lateral organization of Ank3 in Foxj1-GFP + pRGPs during postnatal matu-ration (arrowheads). Note that GFP-dim cells at P3 did not have high Ank3 levels (*).(D) P14 coronal sections of SVZ stained with Ank3, S100 b , DCX, and DAPI, showing little to no Ank3 expression in DCX + neuroblasts and SVZ beneath theependyma (*). V, ventricle; St, striatum. Scale bars: 5 mm (AC) and 10 mm (D). See also Figure S1 .


http://-/?-http://-/?-http://-/?-http://-/?-http://-/?-http://-/?-http://-/?-http://-/?-

8/6/2019 Pi is 0896627311004806

4/15

b 2-Spectrin and a -Adducin ( Figure 3 B). We wanted to knowwhether Ank3 expression/localization is dependent on multiciliaformation in SVZ ependymal cells. Using a tamoxifen-induciblefoxj1-CreER t2 transgene ( Rawlins et al., 2007 ), we deletedKif3a, a critical molecular motor for cilia formation ( Marszaleket al., 2000 ), from postnatal pRGPs. Lineage-tracing analysesof the foxj1-creER t2 ; rosa26-YFP reporter mice injected withtamoxifen at birth and analyzed at P14 showed that Foxj1-CreER t2 can target multiciliated ependymal cells ( Figure S3 A).We generated clonal deletion of Kif3a in pRGPs by low-dosetamoxifen injection into newborn foxj1-creER t2 ; kif3aKO/Flox

mice. When analyzed at P14, Kif3a mutant ependymal patches,lacking surface acetylated-tubulin staining for multicilia, wereable to properly express andlocalizeAnk3 to their lateral borders( Figure S3 B).

We then generated lentiviral construct-expressing shRNA against mouse Ank3, and tested this in NIH 3T3 cells, whichknocked down greater than 95% of endogenous Ank3 after len-tiviral infection ( Figure S4 A). We next made lentivirus coexpress-ing this shRNA and GFP under control of the 1 kb human Foxj1promoter (the same promoter used to generate the Foxj1-GFPtransgene) ( Ostrowski et al., 2003 ), and infected pRGP cultures

Figure 2. In Vitro Generation of SVZ Ependymal Clusters(A) IHC staining of in vitro-cultured pRGPs from P0 mice under proliferating (3 days after plating, upper row) and differentiating conditions (10 days after plating,bottom row). Note g -tubulin ( g -Tub) staining showing single basal body during proliferation (arrows, corresponding to monociliated cells), and multiple basalbodies per cell after differentiation (arrows, corresponding to multiciliated cells).(B) Scanning electron microscopic images of differentiated pRGPs showing multiciliated ependymal cells (Ecs; *) arranged in clusters around monociliated cells(arrows). Right panels show higher magnication view of multiciliated cell (arrowhead), and monociliated cells (arrows).(C) IHC staining of ependymal clusters. Monociliated cell in the center is GFAP + (*). Multiciliated ependymal cells show clustered g -Tub staining (E), and their cellborders are represented by dash lines (traced from Phalloidin costaining).(D) Stills from real-time live imaging movie of Foxj1-GFP + pRGPs showing proliferation and differentiation of SVZ progenitors. Note the increase in GFP + cellclusteringduring proliferation/differentiation transition(arrowheads; also see Movie S2 ).Scale bars:10 mm (A),10 mm/insets2 mm(B),5 mm (C),and50 mm (D).See

also Figure S2 .


http://-/?-http://-/?-

8/6/2019 Pi is 0896627311004806

5/15

24 hr after plating. Lentiviral infection of pRGPs was highly ef-cient as more than 90% of multiciliated cells (assessed byg -tubulin/DAPI staining) became GFP + after differentiation ( Fig-ure 3 C and data not shown). While control virus-infected pRGPsupregulated Ank3 in clusters as normal, we were able to knock-down this expression with the Ank3 shRNA virus ( Figure 3 C).

As GFP expression in infected cells did not become brightenough for live imaging until 34 days after infection (too late

Figure 3. Ank3 Regulation of Postnatal SVZNiche Formation(A) IHC staining showing Ank3 upregulation inclusters during in vitro differentiation of pRGPs(hours after plating). Regions within yellow boxes

enlarged below.(B) Western blot analyses of Ank3, b2-Spectrin(Spec), a -Adducin (Addu) protein levels in pRGPculture under proliferating (prolif.) (3 days afterplating) and differentiating (diff.) conditions(8 days after plating).(C) Control and Ank3 shRNA lentiviral infection of pRGPs in culture, showing knockdown of Ank3upregulation (arrows). Note the high infectionefciency.(D) Quantication of ependymal cluster formationin culture after lentiviral infection (see text andFigure S4 B for details). *p < 0.05, Wilcoxon two-sample test; error bar, SEM; n = 5.(E)IHC staining of P5 ventricularwall whole mountsafter P0 lentiviral injection. Note that while Glastexpression incontrol sample islargelyconcentratedto the center of Ank3

pinwheel-like structures

(asterisks), afterAnk3 knockdown most pRGPsstillexpressed high level of Glast, and showed disor-ganization (seealso Figure S4 C).Scalebars:20 mm(A, C, and E). See also Figures S3 and S4 .

for following cellular clustering in realtime), we used antibody staining to quan-tify theabilityof infectedpRGPsto clusterafter differentiation ( Figure S4 B). Count-ing cells stained with GFAP, g -tubulin,and Phalloidin, we found that Ank3shRNA-infected pRGPs had signicantlyreduced numbers of clustered structureswhen compared to control virus-infectedcultures ( Figure 3 D). To conrm thesendings in vivo, we performed stereo-tactic injection of control and Ank3shRNA lentiviruses into P0 mice, speci-cally targeting pRGPs through striatalinjections( Merkle etal., 2004 ). Ventricularwhole-mount staining 5 days after lentivi-ral injection showed that control pRGPswere able to assemble into clustered

structures, with Ank3+

ependymal cellsexhibiting large apical surface areassurrounding Ank3

cells with small apical

surfaces ( Figure S4 C). In contrast, Ank3knocked-down pRGPs failed to organize

into clusters along the ventricular surface, and retained a smallerapical surface area(by Phalloidin staining)as compared to neigh-boring cells with intact Ank3 expression ( Figure S4 C). Further-more, whereas the Ank3 + pRGPs had largely downregulatedimmature ependymal marker Glast ( Figure 3 E), Ank3 knocked-down pRGPs retained high-level Glast expression, showeddisorganized patterning, and failed to differentiate into maturemulticiliated ependymal cells ( Figure 3 E).



8/6/2019 Pi is 0896627311004806

6/15

Absence of Postnatal Ank3 Upregulation in ConditionalFoxj1 Mutant pRGPsSince striatal lentiviral injection can only target a small number of pRGPs, we would like to remove Ank3 function in vivo. Onestrategy is to delete its upstream regulator in pRGPs to prevent Ank3 expression. To our knowledge, transcriptional regulationof ank3 (or any of the other Ankyrins) is not known. One candi-date for such control, since its expression appears before Ank3 in pRGPs ( Figure 1 ), is the transcription factor Foxj1. It isa well-established regulator of motile-cilia formation ( Yu et al.,2008 ), as well as multicilia formation in ependymal cells ( Brodyet al., 2000; Jacquet et al., 2009 ). Due to severe defects inmultiple organ systems, including the lung, most foxj1 null

Figure 4. Ank3 Defects in Conditional foxj1Mutant Mice(A) Schematic representation of the Foxj1-oxtargeting strategy. N, NotI; E, EcoRI. Germlinetransmission of Flox allele was veried by PCR

genotyping using primers p1 (external to targetingconstruct), p2, and p3. Allele was crossed to rosa26-p , removing neomycin cassette for con-ditional Cre experiments; crossing to b -actin-Cregenerated KO allele.(B) IHC staining of P3 ventricular wall wholemountsfromcontrolandFoxj1cKOmice: note thelack of Ank3 expression in cKO pRGPs.(C) Western blot analyses of Ank3, b2-Spectrin,a -Adducin protein levels in differentiated pRGPsfrom control and cKO mice.(D) Lateral ventricular wall whole-mount IHCstaining from P6 control and cKO mice.(E) IHC staining of lateral ventricular wall sectionfrom P6 cKO; Foxj1-GFP mice showing that GFP +

cells at the ventricular surface have low S100 b ,but high Glast expression. Scale bars: 5 mm (B)and 20 mm (D and E). See also Figure S5 .

mice die within 3 days after birth ( Brodyet al., 2000 ). Despite a previous report( Jacquet et al., 2009 ), we did not obtainany null mutants surviving past P7 inmore than ten litters from crosses usingthe same foxj1 -heterozygous mice( Brody et al., 2000 ).

To address nervous system-specicquestions, we generated a conditionaloxed allele of the foxj1 gene ( Figure 4 A).We crossed our foxj1-ox ( foxj1 Flox/+ ) lineto germline b -actin-cre mice to generatea knockout allele ( foxj1-KO ). We thencrossed this foxj1-KO ( foxj1 KO/+ ) alleleto a nestin-cre driver ( Tronche et al.,1999 ) and foxj1 Flox/Flox mice, andcompared phenotypes between nestin-cre; foxj1 KO/Flox (cKO) and nestin-cre;foxj1 +/Flox (control) littermates. At birth,we could not detect histological differ-ences in brain sections between controland cKO littermates, and lateral ventricle

sizein P3 cKO mice was comparable tocontrols ( Figure S5 A anddata not shown). The cKO mice lived without obvious signs of defect until after P7, when hydrocephalus appeared from thelack of multicilia on maturing ependymal cells ( Figure S5 B).Staining of P5 cKO brain sections conrmed the removal of Foxj1 protein, normally expressed by the ependymal layer incontrol animals ( Figure S5 C).

IHC staining on brain ventricular wall whole mounts from P3control and cKO mice showed that while Ank2 was normally ex-pressed, Ank3 expression was absent from the developing SVZniche in cKO mice ( Figure 4 B). This loss was conrmed bywestern blot analyses of differentiated pRGPs, also showingconcurrent reduced levels for b2-Spectrin and a -Adducin



8/6/2019 Pi is 0896627311004806

7/15

( Figure 4 C). IHC staining on ventricular wall whole mounts fromP6 mice with antibodies against S100 b and Glast showed thatwhile pRGPs from control mice had matured into S100 b hi /Glast lo

ependymal cells, those from mutant mice remained largely

S100 blo

/Glasthi

, resembling immature pRGPs ( Figure 4 D). Todetermine if this phenotype was due to a failure of ependymaldifferentiation, or the generation of additional Glast + progenitors,we introduced by breeding the Foxj1-GFP transgenic reporterallele into the cKO background to visualize the fate of GFP +

pRGPs. The possibility that Foxj1 autoregulates the 1 kb humanFoxj1 promoter in the Foxj1-GFP transgene appeared low sincesequence analyses showed no predicted Foxj1-binding sites( Lim et al., 1997; Badis et al., 2009 ) within this promoter region(data not shown). In cKO mutant mice at P6, we detected robustFoxj1-GFP expression along the lateral ventricular surface, butthese GFP + cells continued to express Glast with little to noS100 b expression ( Figure 4 E). These results showed that incKO mice, the ventricular wall is populated by Foxj1-GFP +

progenitorsdestinedto become SVZ niche cells butfailed to fullydifferentiate into S100 b + ependymal cells.

Ank3-Mediated Lateral Membrane SpecializationDefects in Foxj1 Mutant pRGPsSince Ankyrins mediate many aspects of membrane proteinpositioning, it is likely that several lateral membrane proteinsand their molecular partners may be defective in cKO pRGPslacking Ank3 ( b 2-Spectrin and a -Adducin for example, Fig-ure 4 C). TEM analyses showed thatwhileependymalprogenitorsin P4 control mice continued to mature their lateral membranes(compared to Figure S1 B), in cKO mice both membrane interdig-itation and adherens junctions were reduced ( Figure 5 A andFigure S6 A). Ank3 binds to E- and N-cadherin, and in epithelialcells is known to limit membrane diffusion of E-cadherin in thelateral cell borders ( Kizhatil et al., 2007 ). We found that N-cad-herin protein level was greatly upregulated during in vitro pRGPsdifferentiation ( Figure 5 B), suggesting that a function for Ank3upregulation in Foxj1 + pRGPs could be to anchor newly synthe-sized N-cadherin at cell membranes. cKO pRGPs upregulatedN-cadherin expressionpostnatally, and showed onlymild reduc-tion in protein level after in vitro differentiation compared tocontrols ( Figure 5 B and data not shown). Unsurprisingly, whileventricular whole-mount staining from P4 control mice showedlateral organization of N-cadherin in many pRGPs duringongoing niche formation, this organization was difcult to detectin cKO littermates ( Figure 5 C).

To seeif Ank3 canrescue this N-cadherin localization defectinFoxj1 mutant pRGPs, we generated lentiviral construct express-ing the 190 kDa splice form of Ank3. In control pRGPs afterin vitro differentiation, N-cadherin was colocalized with Ank3 tothe lateral membranes ( Figure 5 D). However, in differentiatedFoxj1 cKO pRGPs, N-cadherin became diffusely distributedthroughout the cytoplasm ( Figure 5 D and Figure S6 B). Lentiviralexpression of Ank3 allowed previously cytoplasmic N-cadherinto locate to the lateral borders in Foxj1 cKO pRGPs ( Figures5D and 5E). The reintroduced Ank3 protein in mutant pRGPswas less evenly distributed at the lateral membranes ( Figure 5 D),reecting perhaps the heterogeneous nature of lentiviral-medi-ated Ank3 expression in these cells, as well as the possibility

that additional molecules regulated by Foxj1 may work togetherwith Ank3 to specialize progenitor lateral membranes.

To determine if Foxj1 can directly activate Ank3 expression inpRGPs, we rst infected Foxj1 cKO pRGPs with a lentiviral

construct expressing Foxj1 with a C-terminal Myc-tag. Bothwestern blots and IHC staining showed that Foxj1-Myc virus-in-fectedcKO pRGPs were able to upregulate Ank3 protein expres-sion ( Figure 5 F). Looking for direct Foxj1-binding sites within 1.4million base pairs (bp) of genomic sequence surrounding the ank3 locus (mm9. chr10:68,740,000-70,150,000), we searchedfor consensus DNA-binding motifs based on published data( Lim et al., 1997; Badis et al., 2009 ) ( Figure S6 C). Using positionfrequency matrix on the predicted A / G TAAACA-binding motif forFoxj1 ( Bejerano et al., 2005 ), we identied 790 positions in thisregion that had a 90% or better sequence match (data notshown). Filtering by evolutionary conservation to orthologouspositions in at least 20 other species ( Karolchik et al., 2007 ),we narrowed this list down to 29 potential binding sites ( Table

S1 ). Looking for clustering of at least 2 sites close by, we identi-ed a 500 bp region roughly 200 kb upstream of the rst exonthat contained 4 putative Foxj1-binding sites conserved in 21,26, 31, and 25 species, respectively (5 0 Enh, Figure S6 C). Wealso identied 2 putative binding sites spaced 540 bp apart inthe 3 0UTR of ank3 that were conserved in 29 and 32 species(30 Enh, Figure S6 C).

Using puried GST-Foxj1 DNA-binding domain fusion protein( Lim et al., 1997 ), we showed via oligonucleotide gel shift assaythat there was specic Foxj1 binding to each of these predictedelements, which can be disrupted by mutating the T and Cpositions in the binding motifs to As ( Figure S6 E). Since the700 kb fragment (from putative 5 0 enhancer to 3 0UTR) of ank3genomic DNA was not readily suitable for BAC/YAC transgene-sis, we next transfected pGL3 dual Firey and Renilla luciferasereporters into pRGP cultures, which showed that both the 5 0 and3 0 genomic fragments can function as transcriptional enhancersin pRGPs ( Figure 5 G). To determine if Foxj1 binds directly tothese ank3 sites, we performed chromatin immunoprecipitation(ChIP) experiments with Myc antibody on pRGPs infected withlentivirus expressing Foxj1-Myc. Using DNA primers specic tothe ank3 5 0 and 3 0 enhancers, we showed by PCR after ChIPthat Foxj1-Myc was able to bind to these genomic sequencesin pRGPs ( Figure 5 H).

Transitioning of RC2 + Radial Glia into Postnatal NSCswithout Ependymal Niche Formation

With the absence of Ank3 expression, we wondered if a primarydefect in Foxj1 cKO pRGPs was an inability to self-cluster and,thus, unable to assemble proper SVZ architecture. Under time-lapselive imaging of Foxj1-GFP + pRGPs in ependymalclusteringassays, we saw that while pRGPs cultured from littermatecontrols were indistinguishable from wild-type cells ( Movie S2and data not shown), the Foxj1 cKO pRGPs uniformly failed toorganize into clusters, and remained positionally static afterexpansion ( Movie S4 and Figure 6 A). We employed standardizedthresholding ( Figure S7 A) to quantify this signicant clusteringdefect in cKO pRGPs ( Figure 6 B).

To our knowledge, it is notknown whether radialglialtransitioninto postnatal SVZ NSCs is niche dependent, or cell intrinsically


http://-/?-http://-/?-http://-/?-http://-/?-

8/6/2019 Pi is 0896627311004806

8/15

Figure 5. Foxj1 Regulation of Ank3 during SVZ Niche Formation(A) TEM of P4 ventricular wall lateral junctions between immature ependymal progenitors (iEp). Note the lack of interdigitation (*)/extension of apical adherens junctions (arrowheads) in cKO lateral borders.(B) Western blot analyses comparing N-cadherin (N-cadh) expression levels during in vitro proliferation (prolif.) and differentiation (diff.) of pRGPs.(C) N-cadherin (N-cadh) IHC staining of P4 ventricular wall whole mounts in x-y and x-z planes. Note the ongoing lateral organization of N-cadh in manypRGPs (*). Cell borders visualized by Phalloidin (Ph) (traced by dashed lines for clarity). Ph and DAPI were used as landmarks to ensure cytoplasmic scanning inx-z planes.(D) IHC analyses of N-cadherin expression in Foxj1 cKO pRGPs infected with lentivirus expressing 190 kDa Ank3. Note the clearance of N-cadherin protein (*)from the cytoplasm of Foxj1 cKO pRGPs expressing Ank3 (arrowheads).(E) Percentage (%)of differentiated pRGPin culturewith lateralizedN-cadherin staining.In Foxj1cKO rescue experiments,we assessed N-cadherin status in cellsexpressing infected Ank3. *p < 0.05, Wilcoxon two-sample test; error bar, SD; n = 5.(F) Western blot and IHC analyses of Ank3 expression in cKO pRGPs infected with lentivirus expressing Foxj1-Myc.



8/6/2019 Pi is 0896627311004806

9/15

regulated. Since these primary defects in ependymal maturationand assembly have prevented formation of the SVZ niche post-

natally in Foxj1 cKO mice, we next wanted to see if pRGPs des-tined to becoming SVZ NSCs can still make their transition. IHCstaining for radialglial markerRC2 showedthatwhileit washighlyexpressed by the developing SVZ at P2 in both the control andcKO mice, this expression was properly downregulated by P6in both genotypes ( Figure 6 C). Ki67 staining on P6 brain sectionsshowed no measurable differences on cellular proliferation at the

(G) Transcriptional activity of 5 0 and 3 0 enhancer elements assayed by luciferase constructs in pRGP cultures. *p < 0.001, Wilcoxon two-sample test; error bar,SEM; n = 6.(H) PCR primer amplication of 5 0 Enh and 3 0 Enh genomic DNA fragments after ChIP with Myc antibody from differentiated pRGPs. Scale bars: 0.5 mm (A) and5 mm (C, D, and F). See also Figure S6 .

Figure 6. Radial Glial Transition to SVZStem Cells in Conditional Foxj1 MutantMice(A) Stills from time-lapse live imaging moviesof Foxj1-GFP + cKO pRGPs during in vitro differ-

entiation. Note the general lack of clusteringbetween GFP-bright cells (also see Movie S3 ).(B) Quantication of cellular clustering from liveimaging movies (see Figure S6 A for details).*p < 0.05, Wilcoxon two-sample test; error bar,SEM; n = 5.(C) RC2 IHC staining of lateral ventricular wallsections, showing the presence (P2) and down-regulation (P6) of RC2 expression in both controland cKO mice.(D) IHC staining and quantication of Ki67 + cellson theventricularwall surface from P6 control andcKO mice. Error bar, SD; n = 9.(E) In vitro differentiation of SVZ adherent NSCcultures from P6 control and cKO mice, showingproduction of Tuj1 + neurons, GFAP + astrocytes,and CNPase + oligodendrocytes. Scale bars:50 mm (A, C, and D) and 15 mm (E). See alsoFigure S7 .

ventricular surface between control andcKO mice ( Figures 6 D). Furthermore,anti-Caspase-3 staining at the same agealso showed no noticeable increase inapoptosis in the cKO mice ( Figure S7 B),suggesting that RC2 + radial glia hadtran-sitioned into postnatal NSCs withoutependymal niche formation.

To validate the presence of NSCs inthis environment, we employed theadherent SVZ NSC culture assay ( Schef-er et al., 2005 ) by harvesting primarycells from both P6 control and cKOmice. NSC cultures from the cKO lateralventricular wall expanded in proliferationmedia like those from control mice(data not shown). Differentiation of pass-age 2 primary cultures showed abundantproduction of Tuj1 + neurons, GFAP +

astrocytes, and CNPase + oligodendrocytes from both thecontrol and cKO cultures, with quantication detectingno appre-

ciable differences in lineage-restricted differentiation potential( Figure 6 E and data not shown). These results are consistentwith the notion that SVZ ependymal niche formation was notrequired for RC2 + radial glia to transition into RC2

postnatal

NSCs. However, the onset of hydrocephalus after 1 week of age in cKO mice prevented us from drawing further conclusionsabout SVZ niche function on neurogenesis.


http://-/?-http://-/?-

8/6/2019 Pi is 0896627311004806

10/15

Inducible Disruption of SVZ Ependymal Niche Leads toDepletion of Neuroblast ChainsPreviously, to our knowledge, it had not been possible to induci-bly remove SVZ ependymal structure in vivo to directly demon-strate its functional signicance on neurogenesis. Since weshowed that the Foxj1-Ank3 pathway was required for SVZformation, our strategy was to use the foxj1-CreER t2 transgene( Figure S3 A) to disrupt this pathway in the SVZ after it hadassembled properly. We generated foxj1-CreER t2 ; foxj1KO/Flox

(inducible KO, iKO) mice by crossing foxj1-CreER t2 ; foxj1KO/+

and foxj1 Flox/Flox animals. We did not observe histologic or

phenotypic differences between foxj1-CreERt2

; foxj1Flox/+

litter-mate controls injected with tamoxifen and noninjected iKOmice (data not shown). For experiments, we administeredsingle-dose tamoxifen at P14, and harvested brains at P28 tostudy the effects on SVZ and new neuron production. IHC stain-ing on coronal sections from tamoxifen-injected control and iKOlittermates showed that we were able to inducibly remove Ank3expression from the ventricular surface ( Figure S8 A). Consistentwith our previous observations, after in vivo Ank3 knockdown( Figure 3 E), aswell asin nestin-Cre; foxj1 KO/Flox mice ( Figure 4 D),inducible removal of Foxj1 and Ank3 also resulted in increasedGlast and decreased S100 b expression on the ventricularsurface ( Figure S8 B). To conrm tamoxifen-mediated deletion

Figure 7. Inducible SVZ Niche DisruptionResults in Neuroblast Chain DefectsIHC stainingof ventricularwall wholemountsfromP28 mice injected with tamoxifen at P14. (A) Incontrol animal, DCXantibodystaining showed the

abundanceof neuroblasts assembled intochains.R, rostral; C, caudal; D, dorsal; V, ventral.(B) foxj1-CreER t2 ; foxj1KO/Flox (iKO) mouse injected with tamoxifen showed disorganization inDCX+ neuroblast chains. Corresponding invertedimage is marked with red showing areas of substantial Ank3 disruption on the ventricularsurface.(C) Another example from iKO animal withtamoxifen injection. Corresponding invertedimage shows large patches of ventricular Ank3disruption and ventricular enlargement.(D) Close-up Z stack images of ventricular wholemount from iKO animal showing neuroblast chainboundary (arrows) and areas of missing Ank3(asterisks). Acetylated-tubulin (A-tub) showssurface cilia.(E) Nissl staining of OB sections from control andiKO mice, showing defects in size (dashed areas,quantied) and cell density (insets) within therostral migratory stream. *p < 0.01, Wilcoxon two-sample test; error bar, SD; n = 5. Scale bars:500 mm (AC),20 mm (D),and200 mm/insets50 mm(E). See also Figure S8 .

of foxj1 , we crossed into the iKO back-ground a Rosa26-tdTomato reporterline ( r26r-tdTomato ).IHC staining of brainsections from foxj1-CreER t2 , foxj1KO/Flox ; r26r-tdTomato Flox/+ mice, 5 days aftertamoxifen injection, showed that while

there were rare tdTomato + ependymal cells retaining someFoxj1 (likely due to transient nature of CreER-mediated recombi-nation), most tdTomato + cells had downregulated Foxj1 expres-sion ( Figure S8 C).

To see how these structural disruptions in the mature nichemay affect SVZ neurogenesis, we performed whole-mount IHCstaining using antibodies against DCX, Ank3, and acetylatedtubulin. We used coordinate-stitching confocal software toacquire Z stack images over the entire ventricular surface, whichallowed us to simultaneously assess Ank3/multicilia status andtheir relationships to newborn neuroblasts traveling in chains

beneath the ventricular surface. Confocal images of DCX stain-ing from control P28 mouse ventricular surface revealed robustmigratory chains of neuroblasts ( Figure 7 A). In contrast, iKOmice injected with tamoxifen at P14 and sacriced at P28showed signicant defects in the coverage of neuroblast chainsalong the ventricular wall ( Figures 7 B and 7C). Since the Foxj1-CreER t2 -targeting strategy generated mosaic populations of mutant and unaffected ependymal cells, we were able to largelyavoid the appearance of hydrocephalus harvesting brains2 weeks after tamoxifen injection ( Figure 7 B). In some animalswe did observe hydrocephalus, as indicated by the enlargementof ventricular surface during tissue harvesting, and this pheno-type correlated with extensive removal of ependymal Ank3



8/6/2019 Pi is 0896627311004806

11/15

expression as conrmed by IHC staining and confocal analysis( Figures 7 C and 7D).

We inverted the dark-eld whole-mount DCX neuroblastimages and noted in red, areas where we observed continuous

patches of Ank3 defects (accompanying Figures 7 B and 7C). After analysis in several tamoxifen-injected iKO mice, we couldnot nd intact DCX + migratory chains in areas that showedextensive ependymal Ank3 loss ( Figures 7 B and 7C and datanot shown). We observed that on the borders between unaf-fected ependymal regions and cells with depleted Ank3 expres-sion, DCX + neuroblast chains became disrupted ( Figure 7 D andFigure S8 D). Predictably, these defects along the ventricular wallled to signicant decrease in cellularity/size of the rostral migra-tory streamin P28 OBsafter P14 tamoxifen induction( Figure 7 E).It is interesting to note that 2 weeks after tamoxifen injection, Ank3 expression was often more affected from Foxj1 deletionthan surface multicilia, perhaps reecting the relative turnoverrates of each in mature ependymal cells ( Figure 7 D and Fig-

ure S8 C). Consistent with the dramatic reduction in DCX + neuro-blasts, Ki67 staining on coronal sections where large areas of ependyma were targeted showed decreased SVZ proliferation( Figure S8 E).

To understand whether the iKO phenotypes may be partly dueto inducible targeting of SVZ NSCs, we performed lineage-tracingexperiments in foxj1-CreER t2 ; r26r-tdTomato mice. We reasonedthat if Foxj1-CreER t2 can mediate signicant recombination inmatureSVZNSCs after niche formation,weshould seetdTomato +

lineage-traced neuroblast chains along the ventricular wall. Aftertamoxifen induction at P14, whole-mount IHC staining from P28foxj1-CreER t2 ; r26r-tdTomato lateral ventricular walls showedabundant tdTomato + ependymal cells ( Figure S8 F). However,we could not detect signicant coexpression of tdTomato inDCX+ neuroblasts in these mice ( Figure S8 F), in contrast to identi-cally treated nestin-CreER tm ; r26r-tdTomato animals. We per-formed 3D colocalization analyses on whole mounts frommultipleanimals to quantify this signicant difference in tdTomato + DCX+

labeling ( Figure S8 H). Any colocalization in foxj1-CreER t2 ; r26r-tdTomato whole mounts was observed mostly from the dorsalventricular edge, where individual cells within dense neuroblastchains were difcult to resolve. We therefore performed liveimaging of these dense neuroblast chains from P28 ventricularwhole mounts after P14 tamoxifen induction, which revealed littletdTomato + migrating cells from foxj1-CreER t2 ; r26r-tdTomatowhole mounts, in contrast to identically treated and imaged nestin-CreER tm ; r26r-tdTomato samples( MovieS5 ).These results

showed that our Foxj1-CreERt2

line does not signicantly targetthe mature SVZ NSC population along the ventricular wall (thisspecicity may differ between CreER transgenes). Our resultsalso demonstrated that ependymal niche cells, although derivedfrom radial glia, once mature do not contribute signicantly tonew neuron production. Thus, the role for Foxj1 appears to belimited to periods after the radial glia have committed to a nichecell fate and are no longer in the stem cell lineage.

Ank3 Expression in Niche Clusters Is Required for NewNeuron ProductionWithin targeted ependymal regions downregulating Ank3expression in iKO mice, concurrent with the loss of neuroblast

chains, we also saw formation of GFAP + clusters ( Figure 8 A).To explore the possibility that ependymal Ank3 expression isdirectly required for SVZ NSCs to generate new neurons, we rstwanted to address whether Ank3 is cell intrinsically required by

SVZ NSCs to make neuroblasts. We again used the adherentSVZ NSC culture assay from wild-type P5 mice, and infectedearly-passage proliferating cultures with lentivirus expressing Ank3 shRNA ( Figure 3 C and Figure S4 A) and GFP driven bya ubiquitous EF1 a promoter. There was no noticeable differencein the proliferative capacities of SVZ NSCs between cultures in-fected with control versus Ank3 shRNA lentiviruses (data notshown), since these cells are Ank3

( Figure 1 ). Upon in vitro

differentiation, we saw abundant GFP + DCX+ neuroblasts thatpersisted for the duration of culture in both the control and Ank3 shRNA-infected cultures ( Figure 8 B and data not shown),showing that Ank3 knockdown does not cell intrinsically inhibitneuroblast production from SVZ NSCs. Consistent with thein vitro ndings, we transplanted these Ank3 shRNA-infected

NSCs back in vivo, and observed 7 and 28 days posttransplan-tation GFP + cells within the SVZ as well as neuroblasts andneurons in the rostral migratory stream and OB ( Figure 8 C).Quantication of GFP + cell numbers showed no obvious differ-ences between control and Ank3 shRNA virus-treated transplan-tation experiments (data not shown).

We next tested the relationship between Ank3 and neuroblastproduction in our pRGP niche culture assay. Although no exog-enous growth factors (EGF and bFGF, required for SVZ NSCrenewal ex vivo) were added at any time to these primarycultures, we reasoned that perhaps thepresenceof Ank3 + epen-dymal niche cells may support NSCs and allow them to makeneuroblasts during differentiation. IHC staining of pRGP culturein differentiation media 5 days after plating showed largenumbers of DCX + neuroblast clusters, with most in close prox-imity to Ank3 + niche clusters (arrows, Figure 8 D). To determineif Ank3 expression by these niche clusters was required for neu-roblast production, we used the same shRNA strategy to ef-ciently remove Ank3 protein expression from differentiatingpRGPs ( Figure 3 C and Figure S4 B). This resulted in a dramaticreduction of DCX + neuroblast clusters seen in Ank3 shRNA-treated versus control virus-treated cultures ( Figure 8 E). Har-vesting the Ank3 shRNA-treated pRGP cultures earlier or laterduringdifferentiation also didnot show formation of DCX + neuro-blast clusters (data not shown), revealing that the defects in neu-roblast production were not due to DCX + cells dying or a delay indifferentiationprogram. These results arein support of our in vivo

observations that postnatal Ank3-mediated SVZ ependymalniche organization is required for the continued production of new neurons.

DISCUSSION

To study the functional signicance of SVZ niche on new neuronproduction, we rst showed that pRGPs have an intrinsic abilityto cluster into structures of the adult SVZ neurogenic niche. Wediscovered that the lateral membrane adaptor protein Ank3 isspecically upregulated in pRGPs destined to become SVZniche cells, but not in stem cells, and that this Foxj1-regulatedexpression is necessary for pRGP assembly into mature SVZ


http://-/?-http://-/?-

8/6/2019 Pi is 0896627311004806

12/15

structures. Disruption of this Foxj1-Ank3 pathway in vivo specif-ically removed SVZ architecture, allowing us to demonstrate, toourknowledge, for therst time that themature ependymalnicheis required to maintain continued production of new neurons inthe postnatal brain.

Ank3 and Postnatal Radial Glial Lateral MembraneSpecializationOur results showing that Ank3 functions in pRGPs destinedto become SVZ ependymal cells, but not in future stemcells, revealed selective Ankyrin usage by a subpopulation of

Figure 8. Ank3 Expression in SVZ Niche Is Required for Neuroblast Production(A) IHC staining of ventricular whole mounts from P28 control and iKO mice injected with tamoxifen at P14, showing abnormal GFAP + patches in targeted areas.Regions within white boxes enlarged to the right.(B) SVZ NSC adherent culture from wild-type mice infected with lentivirus expressing Ank3 shRNA and GFP driven by ubiquitous EF1 a promoter, showingabundant GFP + DCX+ neuroblasts 4 days after in vitro differentiation.(C) GFP staining of brain sections from mice transplanted with Ank3 shRNA-infected SVZ NSC culture, 7 and 28 days posttransplantation. SCJ, striatal cortical junction; RMS, rostral migratory stream; GCL, granular cell layer; MCL, mitral cell layer; LV, lateral ventricle.(D) IHC staining of wild-type pRGP niche cultures: 5 days after plating, large numbers of DCX + neuroblast clusters (arrows) can be seen as well as Ank3 + nicheprogenitor clusters.(E) pRGPniche culturesinfected withcontrolversusAnk3 shRNA lentivirus, showing a dramaticreduction in thenumber of DCX + neuroblastclusters (arrows),andquantiedbelow (each cluster has greater thanve DCX + cells per DAPIstaining, using same softwareacquisitionas described in Figure S4 B). Close-up views of DCX+ clusters shown in insets. *p < 0.01, Wilcoxon two-sample test; error bar, SD; n = 5. Scale bars: 50 mm (A), 25 mm (B), OB 500 mm and all others 50 mm (C),50 mm (D), and 100 mm/insets 25 mm (E).



8/6/2019 Pi is 0896627311004806

13/15

progenitors to establish brain ventricular wall organization. Thispreviously, to our knowledge, undescribed function for Ankyrinis exciting both for SVZ neurogenesis and Ankyrin biology. The ankyrin gene family was rst discovered over 30 years ago, but

until this study, to our knowledge, no transcription factor hadbeen linkedto these proteins. While we showed here that pRGPsupregulate the 190 kDa isoforms of Ank3, mature neuronsexpress the larger 480 and 270 kDa Ank3 isoforms ( Kordeliet al., 1995 ). Thus far, we have not detected Foxj1 expressionin mature Ank3 + neurons, suggesting separate transcriptionalnetworks regulating Ank3 expression in neurons versus pRGPs.Foxj1 is a well-known regulator of multicilia formation, but ourresults here, as well as those from previous studies ( Lin et al.,2004; Jacquet et al., 2009 ), show that there are other importantfunctions for this protein. A microarray study by Ghashghaeiand colleagues in search of additional downstream moleculesregulated by Foxj1 did not identify Ank3 or other transcriptionfactors ( Jacquet et al., 2009 ). We believe further experiments

using expression proling will be needed to understand whichproteins may work together with Foxj1 to regulate Ank3 expres-sion in pRGPs generating the adult SVZ neurogenic niche.

Beyond pRGP cell-cell adhesion, are there other moleculesanchored by Ank3 that facilitate SVZ niche generation? Prior togrowing multicilia at their apical surfaces, pRGPs begin to inter-digitate their lateral membranes with neighboring niche progen-itors. Although the signicance of this elaborate cellular transfor-mation, to our knowledge, is currently unknown, it is possiblethat it goes beyond simple sealing of the epithelium. Oneintriguing class of Ankyrin-associated proteins that warrantsfurther investigation is voltage-gated ion channels. Ank3 isknown to physically associate with both voltage-gated sodium,as well as voltage-gated potassium channels ( Bennett andHealy, 2008 ). It would be of interest to examine whether thesechannels are involved in the formation/maintenance of adultSVZ niche.

SVZ Niche Architecture and Production of New NeuronsTo demonstrate functional signicance of SVZ architecture onthe production of new neurons in postnatal and adult rodentbrains, it is necessary to specically disrupt niche cell functionwithout targeting SVZ NSCs. These experiments may seemconceptually straightforward but have been technically chal-lenging due to our lack of understanding of how the SVZ nicheis generated. They are made more difcult by the fact that epen-dymal niche cell defects can result in signicant secondary

phenotypes that preclude direct assessment of niche functionon neurogenesis, such as what we observed previously withpostnatal Numb deletion ( Kuo et al., 2006 ). Our identication of Ank3 expression regulated by Foxj1 in SVZ niche progenitorsgave us new traction, as we were aided by the relatively normalventricular size, and structurally intact ventricular wall surfaces inour mutant mice postnatally. The delay in onset of hydroceph-alus also proveduseful fordemonstrating the roles for SVZ archi-tecture on radial glial transition, as well as new neuron produc-tion after inducible removal of niche organization.

It had been shown that SVZ stem cells are derived from RC2 +

embryonic radial glia ( Merkle et al., 2004 ), although little iscurrently known about the molecular pathways regulating this

developmental switch. These pRGPs transition from a Nestin +

RC2 + GFAP

phenotype at P0 to a Nestin + RC2

GFAP + pheno-type by P14 ( Tramontin et al., 2003 ). As Nestin is expressed bymature ependymal cells as well as SVZ niche progenitors,

currently, there is a lack of better biomarkers to assist in theidentication of pRGPs destined to becoming SVZ NSCs. Weshowed that in the absence of postnatal ependymal maturation,there was still proper downregulation of RC2 expression, sug-gesting that embryonic radial glia did not persist in the cKOSVZ postnatally. Normal SVZ proliferation in vivo, the lack of increased cell death, and proper differentiation of SVZ stemcell cultures in vitro from P6 cKO mice are all consistent withthe notion that radial glial differentiation into SVZ NSCs doesnot require proper ependymal niche assembly. Work inDrosophila neural progenitors has demonstrated that the controlof progenitor maturation over time can be precisely controlledby cell-intrinsic transcriptional programs ( Isshiki et al., 2001;Maurange et al., 2008 ). It will be of interest to understand

whether similar mechanisms exist during pRGP differentiationinto SVZ NSCs.

Once the SVZ neurogenic niche is formed, our results showedthat the continued production of new neurons migrating inchains along the ventricular wall required intact ependymal orga-nization. To our knowledge, this represents the rst demonstra-tion of such functional requirement forthe SVZ ependymalniche.Ideally, this would allow us to address the effects of newbornneuron depletion from the SVZ on OB circuitry and brain homeo-stasis. However, those experiments are difcult to perform dueto the nature of inducible CreER method, generating mosaiccell populations in vivo. In our study, we struck a balancebetween targeting enough SVZ niche cells to show neuroblastproduction decits, and targeting too many resulting in hydro-cephalus. One way to study long-term consequences of depleting new neuron from the SVZ may be to rethink strategiesfor generating Ank3 mutant mice. Our identication of a criticalrole for Ank3 and the ependymal niche in maintaining newneuron production from adult NSCs should synergize with futurestudies on generating new neurons in the adult brain in healthand disease.

EXPERIMENTAL PROCEDURES

Cell CulturepRGP differentiation from P0 mice: lateral ventricular wall was dissected, trit-urated in DMEM-High Glucose 4.5 (GIBCO), 10% FBS (Hyclone), 1% L-gluta-mine, 1% Pen/Strep, and plated at 450,000500,000 cells/ml in the same

media on Poly-D-Lysine (Sigma)-coated surface, incubated under normalcell culture conditions. When cells reach 90%100% conuence (34 daysafter plating), media were switched to 2% FBS (other ingredients same asabove) and not changed again for the duration of experiment. Adherent SVZNSC culture was performed as described ( Schefer et al., 2005 ).

Electron Microscopy and IHC StainingDetails on electron microscopy, IHC staining, and antibodies used can befound in Supplemental Experimental Procedures .

Imaging and AnalysisTime-lapse cultureexperimentswere acquiredon inverted Zeiss CellObserverSystem under standard environmental control. All IHC images were acquiredon Leica TCS SP5 confocal microscope, with control and experimental



8/6/2019 Pi is 0896627311004806

14/15

samples imaged under identical settings. Details can be found inSupplemental Experimental Procedures .

Lentiviral ExperimentsLentiviruses were produced in 293T17 cells through triple transfection of transfer plasmids (pLVTHM or pWPXL), psPAX2, and pMD2.G. Mouse Ank3was targeted for shRNA knockdown through the sequence GAGAGAGAACTTATCGTCC cloned into pSuper-GFP, then subcloned into lentiviral plasmidpLVTHM. EF-1 a promoter in pLVTHM driving GFP expression was replacedby 1 kb human Foxj1 promoter ( Ostrowski et al., 2003 ). A human Ank3 shRNA ( Kizhatil and Bennett, 2004 ) cloned into pLVTHM was used as off-target effectcontrol for pRGP culture experiments. 3 0 Myc-tag Foxj1 cDNA and 190 kDa Ank3 cDNA were independently cloned into pWPXL lentiviral plasmid anddriven under EF-1 a promoter.

Generation of foxj1-ox MiceDetails of gene targeting strategy and PCR genotyping protocol can be foundin Supplemental Experimental Procedures .

In Vivo InjectionsIn vivo injection of lentivirus was performed as described ( Merkle et al., 2004 ).SVZ transplantation of passage 2 primary adherent SVZ stem cells was per-formed as described ( Merkle et al., 2007 ) with modications that can be foundin Supplemental Experimental Procedures . Tamoxifen (10 mg/ml, freshly dis-solved in cornoil) was injected intraperitoneally at 0.15 mg/g of body weight toinduce CreER t2 -mediated recombination.

Biochemical Assays and Sequence AnalysisDetails on western blots, gel shifts, ChIP assays, and enhancer analyses canbe found in Supplemental Experimental Procedures .

SUPPLEMENTAL INFORMATION

Supplemental Information includes Supplemental Experimental Procedures,eight gures, one table, and ve movies and can be found with this articleonline at doi:10.1016/j.neuron.2011.05.029 .

ACKNOWLEDGMENTS

We thank G. Bejerano and A. Wenger (Stanford) for help with ank3 sequenceanalysis; F. Merkle (Harvard), Z. Mirzadeh (Barrow Neuro. Inst.), T. Ghashghaei(N.C. State), G. Michelotti, T. Oliver, and S. Johnson for helpful discussions; L.Ostrowski (U.N.C. Chapel Hill) for Foxj1-GFP reporter and S. Brody (Washing-tonU.) forFoxj1-nullmice; F.Wangfor tdTomatoreportermice; N. Pershing,A.Gaines, D. Fromme, and E. Weston for project assistance; and M. Bagnat, A.Buckley, C.Eroglu, B. Hogan,T. Lechler, andA. West forhelpful commentsonmanuscript. This work was supported by Jean and George Brumley Jr.Endowment (to P.P.-G., K.M.A., and C.T.K.); Ruth K. Broad Foundation (toP.P.-G.); and Sontag Foundation, David & Lucile Packard Foundation, BasilOConnor Starter Scholar Research Award 5-FY09-34 from the March of Dimes, and N.I.H. Directors New Innovator Award 1 DP2 OD004453-01 (toC.T.K.).

Accepted: May 10, 2011Published: July 13, 2011

REFERENCES

Abdi, K.M., and Bennett, V. (2008). Adducin promotes micrometer-scale orga-nization of beta2-spectrin in lateral membranes of bronchial epithelial cells.Mol. Biol. Cell 19 , 536545.

Alvarez-Buylla, A., and Lim, D.A. (2004). For the long run: maintaining germinalniches in the adult brain. Neuron 41 , 683686.

Badis, G.,Berger,M.F., Philippakis,A.A.,Talukder,S., Gehrke,A.R., Jaeger,S.A.,Chan, E.T., Metzler, G., Vedenko, A., Chen, X., et al. (2009). Diversity and com-plexity in DNA recognition by transcription factors. Science 324 , 17201723.

Bejerano, G., Siepel, A.C., Kent, W.J., and Haussler, D. (2005). Computationalscreening of conserved genomic DNA in search of functional noncodingelements. Nat. Methods 2, 535545.

Bennett, V., and Healy, J. (2008). Organizing the uid membrane bilayer:diseases linked to spectrin and ankyrin. Trends Mol. Med. 14 , 2836.

Brody, S.L., Yan, X.H., Wuerffel, M.K., Song, S.K., and Shapiro, S.D. (2000).Ciliogenesis and left-right axis defects in forkhead factor HFH-4-null mice. Am. J. Respir. Cell Mol. Biol. 23 , 4551.

Carle n, M., Meletis, K., Go ritz, C., Darsalia, V., Evergren, E., Tanigaki, K., Amendola, M., Barnabe -Heider, F., Yeung, M.S., Naldini, L., et al. (2009).Forebrain ependymal cells are Notch-dependent and generate neuroblastsand astrocytes after stroke. Nat. Neurosci. 12 , 259267.

Doetsch, F., Caille , I., Lim, D.A., Garc a-Verdugo, J.M., and Alvarez-Buylla, A.(1999). Subventricular zone astrocytes are neural stem cells in the adultmammalian brain. Cell 97 , 703716.

Go tz,M., andHuttner, W.B. (2005). Thecellbiology of neurogenesis. Nat. Rev.Mol. Cell Biol. 6, 777788.

Guirao, B., Meunier, A., Mortaud, S., Aguilar, A., Corsi, J.M., Strehl, L., Hirota,Y., Desoeuvre,A., Boutin, C., Han,Y.G., et al.(2010). Coupling betweenhydro-

dynamic forces and planar cell polarity orients mammalian motile cilia. Nat.Cell Biol. 12 , 341350.

Isshiki, T., Pearson, B., Holbrook, S., and Doe, C.Q. (2001). Drosophila neuro-blasts sequentially express transcription factors which specify the temporalidentity of their neuronal progeny. Cell 106 , 511521.

Jacquet, B.V., Salinas-Mondragon, R., Liang, H., Therit, B., Buie, J.D.,Dykstra, M., Campbell, K., Ostrowski, L.E., Brody, S.L., and Ghashghaei,H.T. (2009). FoxJ1-dependent gene expression is required for differentiationof radial glia into ependymal cells and a subset of astrocytes in the postnatalbrain. Development 136 , 40214031.

Karolchik, D., Bejerano, G., Hinrichs, A.S., Kuhn, R.M., Miller, W.,Rosenbloom, K.R., Zweig, A.S., Haussler, D., and Kent, W.J. (2007).Comparative genomic analysis using the UCSC genome browser. MethodsMol. Biol. 395 , 1734.

Kelsch, W., Sim, S., and Lois, C. (2010). Watching synaptogenesis in the adult

brain. Annu. Rev. Neurosci. 33 , 131149.Kizhatil, K., and Bennett, V. (2004). Lateral membrane biogenesis in humanbronchial epithelial cells requires 190-kDa ankyrin-G. J. Biol. Chem. 279 ,1670616714.

Kizhatil, K., Davis, J.Q., Davis, L., Hoffman, J., Hogan, B.L., and Bennett, V.(2007). Ankyrin-G is a molecular partner of E-cadherin in epithelial cells andearly embryos. J. Biol. Chem. 282 , 2655226561.

Kokovay, E.,Shen, Q.,and Temple, S. (2008). Theincredibleelasticbrain: howneural stem cells expand our minds. Neuron 60 , 420429.

Kokovay, E., Goderie, S., Wang, Y., Lotz, S., Lin, G., Sun, Y., Roysam, B.,Shen, Q., and Temple, S. (2010). Adult SVZ lineage cells home to and leavethe vascular niche via differential responses to SDF1/CXCR4 signaling. CellStem Cell 7 , 163173.

Kordeli, E., Lambert, S., and Bennett, V. (1995). AnkyrinG. A new ankyrin genewith neural-specic isoforms localized at the axonal initial segment and nodeof Ranvier. J. Biol. Chem. 270 , 23522359.

Kriegstein, A., and Alvarez-Buylla, A. (2009). The glial nature of embryonic andadult neural stem cells. Annu. Rev. Neurosci. 32 , 149184.

Kuo, C.T., Mirzadeh, Z., Soriano-Navarro, M., Rasin, M., Wang, D., Shen, J.,Sestan, N., Garcia-Verdugo, J., Alvarez-Buylla, A., Jan, L.Y., and Jan, Y.N.(2006). Postnatal deletion of Numb/Numblike reveals repair and remodelingcapacity in the subventricular neurogenic niche. Cell 127 , 12531264.

Lim, L., Zhou, H., and Costa, R.H. (1997). The winged helix transcription factorHFH-4 is expressed during choroid plexus epithelial development in themouse embryo. Proc. Natl. Acad. Sci. USA 94 , 30943099.

Lin,L., Spoor, M.S., Gerth, A.J., Brody, S.L.,and Peng,S.L. (2004). Modulationof Th1 activation and inammation by the NF-kappaB repressor Foxj1.Science 303 , 10171020.


http://dx.doi.org/doi:10.1016/j.neuron.2011.05.029http://dx.doi.org/doi:10.1016/j.neuron.2011.05.029

8/6/2019 Pi is 0896627311004806

15/15

Lledo, P.M., Merkle, F.T., and Alvarez-Buylla, A. (2008). Origin and function of olfactory bulb interneuron diversity. Trends Neurosci. 31 , 392400.

Luo, J., Shook, B.A., Daniels, S.B., and Conover, J.C. (2008). Subventricularzone-mediated ependyma repair in the adult mammalian brain. J. Neurosci. 28 , 38043813.

Marszalek,J.R., Liu,X., Roberts,E.A., Chui,D., Marth, J.D., Williams, D.S., andGoldstein, L.S. (2000). Genetic evidence for selective transport of opsin andarrestin by kinesin-II in mammalian photoreceptors. Cell 102 , 175187.

Maurange, C., Cheng, L., and Gould, A.P. (2008). Temporal transcriptionfactors and their targets schedule the end of neural proliferation inDrosophila . Cell 133 , 891902.

Merkle, F.T., Mirzadeh, Z., and Alvarez-Buylla, A. (2007). Mosaic organizationof neural stem cells in the adult brain. Science 317 , 381384.

Merkle, F.T., Tramontin, A.D., Garc a-Verdug o, J.M., and Alvarez-Buylla, A.(2004). Radial glia give rise to adult neural stem cells in the subventricularzone. Proc. Natl. Acad. Sci. USA 101 , 1752817532.

Miller, F.D., and Gauthier-Fisher, A. (2009). Home at last: neural stem cellniches dened. Cell Stem Cell 4, 507510.

Mirzadeh, Z., Merkle, F.T., Soriano-Navarro, M., Garcia-Verdugo, J.M., and

Alvarez-Buylla, A. (2008). Neural stem cells confer unique pinwheel architec-ture to the ventricular surface in neurogenic regions of the adult brain. CellStem Cell 3, 265278.

Ostrowski, L.E., Hutchins, J.R., Zakel, K., and ONeal, W.K. (2003). Targetingexpression of a transgene to the airway surface epithelium using a ciliatedcell-specic promoter. Mol. Ther. 8, 637645.

Prothmann, C., Wellard, J., Berger, J., Hamprecht, B., and Verleysdonk, S.(2001). Primary cultures as a model for studying ependymal functions:glycogen metabolism in ependymal cells. Brain Res. 920 , 7483.

Rawlins, E.L., Ostrowski, L.E., Randell, S.H., and Hogan, B.L. (2007). Lungdevelopment and repair: contribution of the ciliated lineage. Proc. Natl. Acad. Sci. USA 104 , 410417.

Sawamoto, K., Wichterle, H., Gonzalez-Perez, O., Choln, J.A., Yamada, M.,Spassky, N., Murcia, N.S., Garcia-Verdugo, J.M., Marin, O., Rubenstein,

J.L., etal. (2006). Newneuronsfollow theowof cerebrospinal uid in theadultbrain. Science 311 , 629632.

Schefer, B., Walton, N.M., Lin,D.D., Goetz, A.K., Enikolopov, G., Roper, S.N.,and Steindler, D.A. (2005). Phenotypic and functional characterization of adultbrain neuropoiesis. Proc. Natl. Acad. Sci. USA 102 , 93539358.

Shen, Q., Wang, Y., Kokovay, E., Lin, G., Chuang, S.M., Goderie, S.K.,Roysam, B.,and Temple, S.(2008). AdultSVZ stemcells lie ina vascularniche:a quantitativeanalysisof niche cell-cell interactions.Cell StemCell 3, 289300.

Spassky, N., Merkle, F.T., Flames, N., Tramontin, A.D., Garc a-Verdugo, J.M.,and Alvarez-Buylla, A. (2005). Adult ependymal cells are postmitotic and arederived from radial glial cells during embryogenesis. J. Neurosci. 25 , 1018.

Suh, H., Deng, W., and Gage, F.H. (2009). Signaling in adult neurogenesis. Annu. Rev. Cell Dev. Biol. 25 , 253275.

Tavazoie, M., Van der Veken, L., Silva-Vargas, V., Louissaint, M., Colonna, L.,Zaidi, B., Garcia-Verdugo, J.M., and Doetsch,F. (2008). A specialized vascularniche for adult neural stem cells. Cell Stem Cell 3, 279288.

Tramontin, A.D., Garc a-Verdugo, J.M., Lim, D.A., and Alvarez-Buylla, A.(2003). Postnatal development of radial glia and the ventricular zone (VZ):a continuum of the neural stem cell compartment. Cereb. Cortex 13 , 580587.

Tronche, F., Kellendonk, C., Kretz, O., Gass, P., Anlag, K., Orban, P.C., Bock,R., Klein, R., and Schu tz, G. (1999). Disruption of the glucocorticoid receptorgenein the nervous system resultsin reduced anxiety.Nat. Genet. 23 , 99103.

Yu, X., Ng, C.P., Habacher, H., and Roy, S. (2008). Foxj1 transcription factorsare master regulators of the motile ciliogenic program. Nat. Genet. 40 , 14451453.

Zhang, Y., Huang, G., Shornick, L.P., Roswit, W.T., Shipley, J.M., Brody, S.L.,and Holtzman, M.J. (2007). A transgenic FOXJ1-Cre system for gene inactiva-tion in ciliated epithelial cells. Am. J. Respir. Cell Mol. Biol. 36 , 515519.

Zhou, D., Lambert, S., Malen, P.L., Carpenter, S., Boland, L.M., and Bennett,V. (1998). AnkyrinG is required for clustering of voltage-gated Na channels ataxon initial segments and for normal action potential ring. J. Cell Biol. 143 ,12951304.


pi is 0896627311004806

Documents