Helios expression coordinates the development of a subset of striatopallidal medium spiny neurons
SUMMARY
Here we unravel the mechanism of action of Helios (He) during the development of striatal medium spiny neurons (MSNs). He regulates the second wave of striatal neurogenesis involved in the generation of striatopallidal neurons that express dopamine 2 receptor (D2R) and enkephalin (ENK). To exert this effect He is expressed in neural progenitor cells (NPCs) retaining them into the G1/G0 phase of the cell cycle. Thus, the lack of He produces an increase of S-phase entry and S-phase length of NPCs which in turn impairs striatal neurogenesis and produces an accumulation of the number of cycling NPCs in the germinal zone (GZ) that end up dying at postnatal stages. Therefore, He-/- mice show a reduction in the number of Dorso-Medial Striatal MSNs in the adulthood that produces deficits in motor skills
INTRODUCTION
The mammalian striatum controls body movements through a sophisticated neuronal network that is dependent on the neurogenesis of two major classes of striatal neurons: the striatal projection neurons (or medium spiny neurons; MSNs) and the interneurons. MSNs are subdivided in two subpopulations; neurons that constitute the direct (or striatonigral) pathway and preferentially express SP and D1R, and neurons of the indirect (or striatopallidal) pathway that mainly express ENK and D2R (Gerfen, 1992). These two populations are differentially distributed within the striatal compartments. Striatal patches or striosomes mainly contain SP+ MSNs, but both MSN subpopulations, SP+ and ENK+, are located in the matrix (Gerfen, 1992).During embryonic development, radial glial cells (RGC) from the ventricle wall of the lateral ganglionic eminence (LGE) undergo successive divisions to expand the pool of neural progenitor cells (NPCs), thereby increasing the volume of the germinal zone (SVZ) [for revision: (Götz and Barde, 2005; Merkle and Alvarez-Buylla, 2006)]. At certain developmental stages, NPCs differentiate into immature neurons that migrate radially to the mantle zone (MZ) (Götz and Barde, 2005; Merkle and Alvarez-Buylla, 2006; Mérot et al., 2009). Two waves of striatal neurogenesis segregate MSNs into two principal compartments; the patches, generated during the first neurogenic wave (starting at E12.5 in mouse); and the matrix, developed during late striatal neurogenesis (starting at E14.5 in mouse) (Gerfen, 1992; Mason et al., 2005).
Within the LGE, transcription factors such as Gsx1 & 2 (formerly named Gsh1 & 2), Ascl1 (formerly named Mash1) and members of the Dlx-family display specific patterns of expression within the GZ and the MZ, and they have been implicated in LGE patterning and/or differentiation (Eisenstat et al., 1999; Rallu et al., 2002; Waclaw et al., 2009; Yun et al., 2002). In addition, the transcription factors Ebf-1, Isl1, Ctip2 (alsoknown as bcl11b), and Ikaros are mainly expressed in the MZ of the LGE where they regulate terminal differentiation of striatal projection neurons (Arlotta et al., 2008; Ehrman et al., 2013; Garcia-Dominguez et al., 2003; Garel et al., 1999; Lobo et al., 2006; Lobo et al., 2008; Martín-Ibáñez et al., 2010).Ikaros family members are transcription factors that play essential roles during lymphocyte development (Cobb and Smale, 2005; Georgopoulos, 2002; Yoshida and Georgopoulos, 2014). Ikaros is the founder member of this family of DNA-binding proteins which consists of Ikaros, Helios (He), Aiolos, Eos, and Pegasus (John et al., 2009; Rebollo and Schmitt, 2003; Yoshida and Georgopoulos, 2014). In addition, Ikaros has been implicated in CNS development (Agoston et al., 2007; Alsiö et al., 2013; Martín-Ibáñez et al., 2010). We have recently described that He is also implicated in striatal development (Martín-Ibáñez et al., 2012). Within the LGE, He is expressed from E14.5 to P15 in both the GZ and the MZ, and its expression is downstream of Gsx2 and Dlx1/2 (Martín-Ibáñez et al., 2012). However, little is known about action mechanisms of He during this developmental process.Here we demonstrate that He is expressed by NPCs at the G0/G1-phase of the cell cycle and induces neuronal differentiation by decreasing the levels of Cyclin E and blocking the progression of these NPCs into S-phase. Consequently, in the absence of He, proliferating NPCs accumulate in the GZ and the number of Ctip2 and DARPP-32+ MSNs is reduced in the striatum which disturb motor skill learning.
RESULTS
Here we demonstrated that He is expressed from E12.5 in scattered cells (Fig. S1) until P15 peaking at E18.5 (Martín-Ibáñez et al., 2012). He showed preferential expression inD2R-eGFP neurons (46.69 8.37% of He+ cells colabeled with D2R; Fig.1A andFig.S2B) and ENK+ MSNs (89.05 5.77% of He+ cells colabeled with ENK; Fig.S3). Incontrast, few D1R-eGFP+ neurons and SP+ neurons co-expressed He (3.94 2.53% and18.20 2.1% of He+ cells colabeled with D1R and SP, respectively; Fig.1A, Fig.S2A and Fig.S3B-C). We next examined striatal birth-dating in He knockout (He-/-) and wild-type (wt) mice at different embryonic developmental stages (Figs.1B-E). The first wave of striatal birth-dating at E12.5 was not altered, since no differences were found in the total number of BrdU+ cells between He-/- and wt mice (Fig.1C). However, lack of He induced a significant reduction in the second wave of striatal birth-dating at E14.5 (Fig.1D). No significant differences were found between genotypes at E16.5 (Fig.1E). This striatal birth-dating impairment disturbed MSNs generation since the density and total number of Ctip2-positive cells was decreased in He-/- mice compared to wt mice at E18.5 (Figs.1F-G), suggesting a defect in the second neurogenic wave. In agreement, we observed that He+ cells were mainly generated during the second wave of striatal neurogenesis (Fig.S4), between E14.5 (Figs.S4E-G) and E16.5 (Figs.S4H-J). Only few cells were observed to be born at earlier stages (E13.5; Figs.S4B-D).To assess whether He was expressed by proliferative cells in the LGE, we performed double staining for He and Ki67 at E16.5, BrdU or phospho-histone H3 (PH3) at E14.5.
Our results showed that He and Ki67+ areas were mainly coincident at the GZ-MZ border at E16.5 (Fig.2A). Within this area He was expressed by low Ki67 expressingNPCs (Fig.2B-C) but not by high Ki67+ cells (Fig.2D; see Fig.S5 for quantification details). However, there was a lack of co-localization between He and short pulsed BrdU NPCs (Fig.2E-F), and He and PH3+ NPCs (Fig.2G-H). Interestingly, He only colocalized with Ki67 expressing cells during the neurogenic period since we could not observe colocalization from E18.5 onwards (Fig.S6).Analysis of the number of cycling cells at different developmental stages in He-/- and wt mice (Fig.2I-L) showed that, the total number of proliferating cells in the GZ was significantly increased from E14.5 to P3 (Fig.2I-K), inducing an enlargement of the proliferative area stained with Ki67 (Fig.S7). Interestingly, this feature reverted at P7, when the number of proliferating cells in He-/- mice decreased with respect to wt mice (Fig.2L and Fig.S8). To analyze whether a specific subpopulation of progenitors was more compromised than others we counted the percentage of PH3+ basal, subapical, and apical progenitors as described by Pilz et al. (Pilz et al., 2013) (Fig.S9A-B). No differences were found between He-/- and wt mice (Fig.S9B). We also analyzed by QPCR the expression of striatal progenitor markers at E16.5. No differences were found in the levels of mRNA for these markers in He-/- compared to wt mice (Fig.S9C).To further elucidate the role of He within NPCs proliferation, we performed LOF and GOF in vitro studies using neurosphere assay (Fig.S10). There was an increase in the number of proliferating cells in the absence of He (Fig.S10A,C,E-F).
Accordingly, He over-expression significantly reduced the number of proliferating NPCs with respect to the control eGFP over-expressing NPCs (Fig.S10B,D). In addition, in the absence of He, NPCs were less prone to differentiate to β-III-tubulin+ neurons (Fig.S10H). In contrast an increase in the number of neurons was observed after He over-expression (Fig.S10I-K). Interestingly, He did not exert any change in the percentage of GFAP+ cells in the LOF or in the GOF experiments (Fig.S10H-I). Consequently, He-/- mice didnot present any defects in astrocyte differentiation compared to wt mice (Figs.S11A-D).In fact, we did not observe colocalization between He and GFAP (Fig.S11E).To understand the cellular mechanism by which He regulates NPC proliferation and neurogenesis, we next analyzed cell cycle. We observed that lack of He induced a significant increase in NPCs S-phase length that in turn increased cell cycle length measured by an accumulative exposure to BrdU [see materials and methods; (Lange et al., 2009); Figs.3A,C]. However, no differences were observed between the length of the G2/M phases in He-/- compared to wt mice derived NPCs through the analysis of the mitotic BrdU labeling index as described previously (Takahashi et al., 1995) (Figs.3B-C, S12). Representation of the percentage of cell cycle phases respect to the total cell cycle length clearly demonstrated an elongation of S-phase length when He was knocked down (Fig.3C). Consistently, He over-expression induced a severe reduction of S-phase length (GOF, Fig.3D). Our results also showed that in the absence of He more NPCs entered S-phase (punctate BrdU+/EdU+; Figs.3E-H) but the number of cells exiting S-phase was not altered [BrdU+/EdU-; see S-phase analysis in materials and methods; (Lange et al., 2009); Fig.3E-F].
In addition, no differences were found in the number of cells exiting the cell cycle [BrdU+/Ki67 negative; see Cell cycle index in materials and methods and (Urbán et al., 2010)] in LOF (Figs.S13A,B,D) or GOF experiments (Fig.S13C).In order to demonstrate the mechanism by which He controls S-phase entry, we next analyzed the protein levels of Cyclin E (Fig.4), a key regulator of the transition from G1 to S phase (Ohtsubo et al., 1995). NPCs derived from He-/- mice presented increased levels of PCNA, a marker of cell proliferation, and Cyclin E (Figs.4A-C,B-D). Accordingly, He over-expression (Fig.4E-H) produced a reduction of PCNA and CyclinE protein levels (Figs.4E-G,F-H), and a drastic reduction of Cyclin E mRNA levels (Fig.4J). Similarly, in vivo analysis showed an increased number of NPCs that entered into S-phase at the GZ of He-/- compared to wt mice (Fig.4K), which was accompanied by increased protein levels of Cyclin E in the LGE (Fig.4L-M). ChIP experiments performed by Kim and coworkers (Kim et al., 2015) demonstrated that He binds the Cyclin E gene (Ccne) promoter site and in another site downstream of the gene (Fig.4N). However, no changes of the two Cyclin E regulators, E2F1 and retinoblastoma (Rb) (Harbour, 2000; Ohtani et al., 1995), were observed in NPCs derived from He-/- mice (Fig.S14). Altogether, these results suggest that He might control cell cycle progression through regulation of Cyclin E expressionPostnatal cell death is increased in He-/- miceWe next investigated whether cell death was altered in the absence of He during embryonic and postnatal stages. Cleaved caspase-3 immunohistochemistry did not show any differences between He-/- and wt mice at embryonic stages (E14.5, E16.5 and E18.5; data not shown).
However, a significant increase in the number of apoptotic cells was detected in the GZ and the MZ at P3 in He-/- mice (Fig.5A-D), which normalizes at P7 (Fig.5E-F). To check whether cell death is related to a delay in the differentiation of NPCs, we performed an EdU pulse at E18.5 and double staining for EdU and cleaved caspase-3 (Fig.5G) or neural markers (Fig.S15) at P3. EdU+ apoptotic cells were found in the MZ of He-/- mice (Fig.5H-K) and they were positive for the neuronal markerNeuN (71.3 7.10% of cleaved caspase-3+ cells colabeled with NeuN. Fig.S15). These results suggest that in the absence of He there is a delayed differentiation of NPCs that end up dying.We next characterized He-/- mice adult striatum. First, we studied brain hemisphere volume and detected a slight decrease in He-/- mice compared to wt mice (Figs.S16A,C; 8.36% decrease). Interestingly, characterization of striatal volume revealed a larger and significant reduction in He-/- compared to wt mice (Figs.S16B-C; 20.17% decrease). The ratio of striatal vs hemisphere volume showed that striatal volume is selectively disturbed in He-/- mice (wt, 18.23±0.79%; He-/-, 15.45±0.60%), showing a 15.24% reduction of relative striatal volume. Stereological analysis of calbindin+ and DARPP-32+ neurons revealed a significant decrease in the density (Figs.S16D-E,H-I) and total number of MSNs in the striatum of He-/- compared to wt mice (Figs.6A-B). We also analyzed the density of DARPP-32+ neurons at different striatal areas including the Dorso-Medial Striatum (DMS), Dorso-Lateral Striatum (DLS), Ventro-Medial Striatum (VMS) and Ventro-Lateral Striatum (VLS) (Fig.6K).
These experiments demonstrated a significant decrease only in the DMS in He-/- mice with respect to wt mice (Figs.6E-H). Interestingly, a specific alteration of the ENK+ population was also observed in the DMS in the absence of He (Fig.6I). However, no differences were found for the SP+ population in He-/- mice with respect to wt mice (Fig.6J). In addition, no differences were observed between genotypes in the cholinergic and parvalbumin+ striatal interneurons (Figs.S16F-G and 6C-D).In order to study the direct involvement of He in the acquisition of a mature MSN phenotype, we transplanted eGFP or He over-expressing NPCs into the mouse neonatal forebrain (Fig.7A). Compared to control cells, He over-expressing cells displayed a more robust branching with respect control ones 2 weeks post-transplantation (Total neurite tree length per neuron; GFP: 168.13 ± 21.92 µm, He: 413.66 ± 98.84 µm, p=0.0046. Number of branches per neuron; GFP: 14.43 ± 1.68, He: 24.89 ± 4.08,p=0.0089 Fig.7B-E) and DARPP-32 expression was observed in few scattered cells adjacent to the striatum (Fig.7G, H). 4 weeks post-transplantation, several He over-expressing cells displayed DARPP-32 expression (Fig.7J-L), in contrast to control cells, which were all DARPP-32 negative (Fig.7I). Quantification of DARPP-32+ neurons in GFP transplanted cells demonstrated a 150-fold increase in the number of double stained cells in He-expressing cells with respect to controls.
In addition, He over-expression in striatal primary cultures significantly increased the number of calbindin+, DARPP-32+ and ENK+ cells (Fig.S17).To analyze the functional implication of He loss we performed motor tasks in wt and He-/- mice (Fig.8). In the simple swimming test, He-/- mice displayed significant abnormalities compared to wt mice in their swimming latency in the first testing trial (genotype: F2,162 = 4.08, p<0.05; post hoc trial 1: p<0.01), but these disappeared over subsequent trials (Fig.8A).In addition, wt and He-/- mice progressively improved their performance in the balance beam along four trials (Trial: F3,112 = 14.66, p<0.001). However, He-/- mice fell off more times than controls during the first trials (genotype: F2,112 = 13.52, p<0.01; post hoc trial 1: p<0.001; post hoc trial 2 p<0.01; Fig.8B).Within the rotarod test, all mice reached a stable level of performance within 6 trials Fig.8C), as measured by a decrease in the number of falls in 60 sec/mice (testing trial F5,138 = 15.87, p<0.01). However, acquisition on the rotarod task was significantly delayed in He-/- respect to wt mice (genotype F2,138 = 21.03, p<0.01). DISCUSSION Striatal MSNs are generated from NPCs located at the GZ of the LGE. Here we show that He regulates late striatal neurogenesis that give rise to D2R+ ENK neurons. He is expressed by NPCs in the G1/G0 cell cycle phase at the GZ, impairing the G1-S transition by the regulation of Cyclin E, which in turn induces neuronal differentiation. Consequently, the lack of He produces an extended S phase and cell cycle length that increases the number of proliferating NPCs at the GZ. At the beginning of the postnatal period, the number of these NPCs is reduced due to their late aberrant neurogenesis that ends up producing cell death. These abnormalities of embryonic development in He-/-mice produce a reduction of a specific subset of striatopallidal neurons of the dorsomedial striatum that control motor skill learning.NPCs located at the GZ of the LGE become postmitotic and migrate into the MZ to acquire the MSN phenotype (Brazel et al., 2003). We have previously proposed a model for the development of striatal subpopulations in which Ikaros and He are involved in the development of striatopallidal ENK+ matrix MSNs (Martín-Ibáñez et al., 2012). This hypothesis is reinforced by the localization of He in ENK+ neurons that co-express D2R (present results). Besides the apparent similar function between He and Ikaros on ENK+ neurogenesis, many evidences show that they determine different ENK+ subpopulations. They are expressed by different cells (Martín-Ibáñez et al., 2012), and their expression is not modified in the reciprocal knock-out mice (Martín-Ibáñez et al., 2010; Martín-Ibáñez et al., 2012). These results are contrary to the role of Ikaros family members in the hematopoietic system where they directly interact (Hahm et al., 1998; John et al., 2009), suggesting specific mechanisms of action in each system.Gsx2+ radial glial cells constitute the first NPCs that appear during LGE ontogeny, which differentiate with the onset of the neurogenesis from the neuroepithelial cells [for review: (Dimou and Götz, 2014)]. He expressing cells are derived from radial glial cells, since its expression disappears in Gsx2 knock out mice (Martín-Ibáñez et al., 2012). However, He loss does not compromise the number of the radial glial cell subtypes described elsewhere (Pilz et al., 2013). Radial glial cells generate the large MSNs output by a series of intermediate NPCs to amplify specific lineages, although these striatal NPCs are still poorly characterized. He is expressed by a small number of NPCs distributed in deep SVZ. Although the localization of He is mainly at the dorsal areas, it does not seem to be defining a specific SVZ domain as it has been described for other transcription factorsin the VZ (Flames et al., 2007).Some of the NPCs that express He at the GZ co-express low levels of Ki67. Considering that Ki67 labels cells during all phases of the cell cycle except G0 (Kanthan et al., 2010; Scholzen and Gerdes, 2000) and that G1 is the cell cycle phase with lower Ki67 expression levels (Lopez et al., 1991), we hypothesized that He is expressed in a subset of NPCs during G1 and G0 phases. The lack of co-localization between He and BrdU or PH3 reinforces the idea that He is not expressed by cells at S or M phases, respectively. Within G1 phase He impairs S-phase entry, reducing S phase length and arresting NPCs at G1/G0 phase to facilitate neuronal differentiation. Consequently, He-/-mice NPCs increase S-phase entry and continue proliferating in the striatal GZ impairing neurogenesis (see Fig.S18 for a representative scheme). Similarly, Lacomme and coworkers demonstrated that Ngn2 regulates G1-S phase transition, blocking S-phase entry and increasing the number of NPCs at G1/G0 phase (Lacomme et al., 2012).In addition, NPCs shorten S-phase on commitment to neuron production (Arai et al., 2011; Turrero García et al., 2015). Thus, cell cycle length and G1-S phase transition are critical processes for neurogenesis and both are regulated by He. We hypothesize that He arrests LGE-derived NPCs into phases G1/G0 to allow the accumulation of the protein machinery necessary for their differentiation to specific striatal neurons. In fact, critical aspects of neural commitment are acquired in the final division cycle of NPCs. For example, the cortical laminar fate of NPC is acquired during the final progenitor cell division (Bohner et al., 1997; Edlund and Jessell, 1999; McConnell and Kaznowski, 1991). Similarly, during motor neuron development, NPCs become Sonic hedgehog (Shh) dependent late in their final progenitor cell cycle (Ericson et al., 1996), which commits them to a motor neuronal fate (Tanabe et al., 1998).G1-S phase transition is regulated by Cdk2 and Cyclin E, which form a complex that participates in G1-S phase checkpoint [reviewed in: (Hardwick and Philpott, 2014; Ohtsubo and Roberts, 1993)]. Our results suggest that Cyclin E may be a key factor regulated by He that correlates with the G1-S phase transition impairment observed in the He-/- mice. In fact, Cyclin E gene (Ccne) has two very strong He binding domains (Kim et al., 2015) suggesting a direct regulation. Similar to our results, Pilaz and colleagues described that over-expression of Cyclin E in cortical NPCs promotes proliferation increase whereas down-regulation of Cyclin E let to a decrease in progenitor proliferation (Pilaz et al., 2009). In addition, a direct correlation between Cyclin E and S phase entry was proposed by ectopic expression of Cyclin E, that shortens the G1 interval and increases the length of S phase by advancing G1-S phase transition (Resnitzky et al., 1994). Furthermore, ectopic expression of Cyclin E can drive G1 cells into S phase under conditions in which neither pRB is phosphorylated nor E2F is activated (Leng et al., 1997; Lukas et al., 1997). This is coincident with ourresults, since we observed an increase in Cyclin E but no alterations in neither pRB nor E2F in He-/- mice. The homeostasis of NPCs in the striatum is a regulated process in which neurogenesis precedes astro-gliogenesis during development (Alvarez-Buylla et al., 2001; Ninkovic and Götz, 2013). However, contrary to the increase of astro-gliogenesis observed in Ikaros-/ - mice (Martín-Ibáñez et al., 2010), we could not detect any deffects in glial cells in He-/- mice. The role of He on neurogenesis through Cyclin E mediated G1-S transition without modifying astro-gliogenesis coincides with the effect of deferoxamine, a G1/S-phase blocker, that increases neuronal but not astrocytic NPCs differentiation (Kim et al., 2006; Misumi et al., 2008).The reduction of NPCs in He-/- mice at postnatal stages can be related to the increase in cell death during this period. Naturally occurring cell death is a critical step in redefining the final size of specific neuronal populations (Burek and Oppenheim, 1996; Kristiansen and Ham, 2014), which directly correlates to the time of prior exit from cell cycle and position during neuronal development (Gould et al., 1999). Our results point to the idea that cell death observed in He-/- mice is a consequence of the delay in NPCs exiting cell cycle around E18.5, that migrate into the MZ where they become neurons and die. Therefore, lack of He produces a dysfunction in the time and position of late generated neurons in the MZ. Dual effects have also been described for Isl1 and Ebf1 that promote differentiation of striatonigral neurons and in their absence striatal cell death is observed (Garel et al., 1999; Lu et al., 2014). Taken together all these results indicate that an aberrant neurogenesis in turn induces neuronal cell death compromising striatal development.He-mediated regulation of the NPC cell cycle correlates with the determination of a subset of striatopallidal MSNs. The events occurring during striatal development of He-/- mice cause a specific reduction of striatal MSNs at the DMS in the adulthood. Taken together, present findings demonstrate that He plays a direct role in the commitment of NPCs to MSNs. Accordingly, He overexpression is sufficient to differentiate NPCs transplanted into the striatum in MSNs expressing DARPP32.Previous published works and reviews suggest that striatal motor function is involved with the habits formation (Yin and Knowlton, 2006) and procedural learning (Kreitzer, 2009) which fits with what we see in our He-/- mice. The striatum has been classically divided into dorsal and ventral areas, being the dorsal the most involved in motor behavior (Durieux et al., 2012). Accumulating evidences show anatomical and functional differences in the striatum between the external DLS, and the internal DMS (Durieux et al., 2012; Graybiel, 2008; Voorn et al., 2004). Interestingly, the DMS is involved in the initial stages of motor skill learning (Jueptner and Weiller, 1998; Luft and Buitrago, 2005), while the DLS is required for progressive skill automatization and habit learning (Miyachi et al., 2002; Yin et al., 2004). In addition, it has been shown that the loss of D2R+ neurons in the DMS produces early motor learning impairment but the animals can improve their performances to reach control levels (Durieux et al., 2012). As He-/- mice show impairments in the acquisition of motor skills, it seems plausible that He is involved in the generation of a specific subpopulation of striatopallidal D2R+ MSNs in the DMS. The cerebellum is also involved in fine-tuning the motor agility found in procedural skills. Cerebellar lesions or dysfunctions produce permanent deficits in motor tasks. However, diseased animals never perform motorB6CBA wild-type (wt) mice (from Charles River Laboratories, Les Oncins, France), He knockout mice (He-/-) (Cai et al., 2009), pCAGs-eGFP (Okabe et al., 1997), D1R-eGFP and D2R-eGFP generated by GENSAT (Gong et al., 2003) were used. For further details of mice strains and genetyping, see the supplementary materials and methods.Birth dating experiments were performed as described elsewhere [Fig.1B; (Martín-Ibáñez et al., 2010)]. To study the generation of He+ cells, injections of ethynyl deoxyuridine (EdU) at E13.5 or E14.5, and BrdU at E16.5 into wt pregnant mice were performed and allowed to develop until E18.5, when embryos were processed for He and BrdU immunohistochemistry or EdU detection (Life Technologies S.A., Alcobendas, Madrid) (Fig.S4A).To analyze in vivo proliferation in the GZ, E14.5 pregnant mice received a single dose of EdU (50 mg/Kg). The proliferation analysis of E16.5, P3 and P7 was performed by Ki67 immunohistochemistry.In order to track the origin of dead cells in the MZ, a pulse of EdU was performed at E18.5, and immunohistochemistry was performed at P3 against EdU and Cleaved caspase 3 (Cell Signaling Technology, Inc. Danvers MA), Nestin, GFAP, or NeuN (Fig.5G).To study if the lack of He could alter the cells entering the S-phase of the cell cycle, we performed in vivo experiments with He-/- and wt mice as previously described (Lange et al., 2009) (Fig.4K).To over-express He, NPCs were transduced with the pLV-He-IRES-eGFP plasmid or the pLV-IRES-eGFP plasmid which encode human He and eGFP or eGFP alone, respectively. For further details of viral particle production, see the supplementary materials and methods.Neurosphere assayLGEs from E14.5 wt or He-/- mice were dissected out and mechanically disaggregated to culture as neurosphere and differentiate to neural cells as described previously (Martín-Ibáñez et al., 2010). For further details of neurosphere cultures, see the supplementary materials and methods.Loss of function (LOF) experiments were performed with neurospheres derived from He-/- mice whereas gain of function (GOF) experiments were performed by over-expressing He. The number of neurons (β-III-tubulin+) and astrocytes (GFAP+) were analyzed after 6 days in differentiation.BrdU incorporation assays were performed in wt and He-/- mice derived neurospheres (LOF) and neurospheres over-expressing He (GOF) as described elsewhere (Urbán et al., 2010). The number of Ki67+ cells was also analyzed in wt and He-/- mice derived neurospheres (LOF) and neurospheres over-expressing He (GOF).Cell cycle lengthAn accumulative exposure to 1µM BrdU during 36 hours was performed in wt and He-/-mice derived neurospheres (LOF) and in neurospheres over-expressing He (GOF) after2DIV in proliferation. Cells were fixed at different time points after 1µM BrdU exposure (1, 3, 6, 12, 24 and 36 hours) and processed for BrdU immunocytochemistry. Following a regression analysis as previously described by Takahashi et al (Takahashi et al., 1992; Takahashi et al., 1995), the length of the cell cycle and the length of the S phase were calculated for the NPCs.S-phase analysisTo study the cells entering and exiting the S-phase of the cell cycle, we performed in vitro experiments with neurospheres derived from He-/- and wt mice as described previously (Lange et al., 2009) (Fig.3E-H). To study the combined length of the G2/M phases, an accumulative exposure to 1µMBrdU over 5 hours was performed after 2 DIV in proliferation to analyze the mitoticBrdU labeling index as described previously (Takahashi et al., 1995).We analyzed cell cycle index as the number of cells that incorporate BrdU but leave the cell cycle (abandoned the G1-S-G2/M phases and entered into G0) as previously described (Urbán et al., 2010) (Fig.S13).We detected high and low expressing KI67 cells with the automatic intensity detection of the Cell Profiler software.For further details of cell cycle analyses, see the supplementary materials and methods.We obtained and analyzed the Big Wig file from the GEO repository of the article by (Kim et al., 2015) ), and visualized it in the Integrative Genome Viewer with the files provided aligned to the Ensembl Mouse Genome. Details of database used can be found at the supplementary materials and methods.Western blotsWe performed western blot analyses for Cyclin E and PCNA as described elsewhere (Canals et al., 2004) in wt and He-/- mice derived neurospheres (LOF) and neurospheres over-expressing He (GOF). E2F1 and retinoblastoma (Rb) were determined in LOF experiments. For further details of western blot procedure, see the supplementary materials and methods.To asses which striatal subpopulation of MSNs express He, we performed double in situ hybridization for ENK or tachykinin A [(Tac1, a precursor of Substance P(SP)], the precursor of SP, and immunohistochemistry for He as described previously (Martín-Ibáñez et al., 2010). For further details of in situ procedures, see the supplementary materials and methods.For histological preparations embryonic or postnatal brains were removed at specific developmental stages and were frozen in dry-ice cooled methyl-butane or cryoprotected depending on the procedure. Immunolabeling was performed according to the protocols described in (Bosch et al., 2004; Canals et al., 2004). For further details of theIRES-eGFP plasmid, or with the MSCV-IRES-eGFP plasmid as a control (Zhang et al., 2007). We counted the number of He or eGFP overexpressing cells that co-localized with calbindin, DARPP-32 or ENK. For further details of primary culture methods, see the supplementary materials and methods.Unilateral striatal injections of He overexpressing cells were performed using a stereotaxic apparatus (Davis Kopf Instruments, Tujunga, CA, USA); coordinates (millimeters): AP, +2.3, L, +1.4 from lambda and DV, -1.8 from dura. For further details of cell transplants, see the supplementary materials and methods.The mice were placed in an extreme of a transparent perspex extended swimming tank facing away from a visible escape platform at one end of the tank and the time required to reach the platform was recorded.Animals were allowed to walk along a horizontally placed beam of a long steel cylinder (50cm) with 15mm-round diameter. Latency to fall and number of falls were measured.Acquisition of a motor coordination task was further evaluated on the rotarod apparatus (24 rpm). Latency to fall and the number of falls during 60 sec was recorded.For further details of mouse behavior, see the supplementary materials and methods.All results are expressed as the mean of independent experiments ± s.e.m. Results were analyzed using Student’s t-test or one way or two way ANOVA, followed by the Bonferroni post-hoc DEG-35 test.