In Vivo Fate Analysis Reveals the Multipotent and Self-Renewal Features of Embryonic AspM Expressing Cells

Radial Glia (RG) cells constitute the major population of neural progenitors of the mouse developing brain. These cells are located in the ventricular zone (VZ) of the cerebral cortex and during neurogenesis they support the generation of cortical neurons. Later on, during brain maturation, RG cells give raise to glial cells and supply the adult mouse brain of Neural Stem Cells (NSC). Here we used a novel transgenic mouse line expressing the CreERT2 under the control of AspM promoter to monitor the progeny of an early cohort of RG cells during neurogenesis and in the post natal brain. Long term fate mapping experiments demonstrated that AspM-expressing RG cells are multi-potent, as they can generate neurons, astrocytes and oligodendrocytes of the adult mouse brain. Furthermore, AspM descendants give also rise to proliferating progenitors in germinal niches of both developing and post natal brains. In the latter –i.e. the Sub Ventricular Zone- AspM descendants acquired several feature of neural stem cells, including the capability to generate neurospheres in vitro. We also performed the selective killing of these early progenitors by using a Nestin-GFPflox-TK allele. The forebrain specific loss of early AspM expressing cells caused the elimination of most of the proliferating cells of brain, a severe derangement of the ventricular zone architecture, and the impairment of the cortical lamination. We further demonstrated that AspM is expressed by proliferating cells of the adult mouse SVZ that can generate neuroblasts fated to become olfactory bulb neurons.


Introduction
Radial Glia cells (RGs) derive from neuroepithelial cells of the early embryos. At the onset of neurogenesis, RG cells perform symmetric and asymmetric division to support either the proliferating pool of cells or the generation of cortical plate neurons. RG cells are also fated to become glial cells in the post natal brain and a subset of them is maintained and propagated in specialized germinal/neurogenic niches of the adult brain, as neural stem cells (NSCs) [1]. NSCs are located in the sub ventricular zone (SVZ) of the lateral ventricle and in the dentate gyrus of the hippocampus [2,3]. SVZ-restricted NSCs cells generate fast cycling precursors -i.e. type-C progenitors-that subsequently give raise to neural precursor cells (type-A, neuroblasts), fated to become olfactory bulb neurons [2]. NSCs, however, are multipotent cells, because in vitro [4] and in vivo they can generate, neurons, oligodendrocytes and astrocytes of the adult brain [5,6,7,8].
During neurogenesis the balance of asymmetric versus symmetric divisions regulates the number of post-mitotic cells and, possibly fosters the maintenance of a proliferating pool of cells fated to supply neural stem cell niches of proliferating progenitors. The identification of these cells in both developing and adult brains, has been obtained by using feasible genetic tools that have so far been generated to label RG and SVZ restricted NSCs [9,10,11,12,13,14]. These models give the opportunity to trace, in vivo, SVZ restricted NSCs expressing self renew features [14], and to perform high resolution long term fate mapping analysis by targeting proliferating cells during forebrain development. Based on these models, it has been established that cohorts of genetically tagged embryonic precursor cells can generate SVZ restricted NSCs [1,11,14]. The analysis of a battery of Cre transgenic mouse lines revealed that SVZ restricted NSCs derive from both cortical RG cells and cells of lateral Ganglionic Eminence [15,16]. Accordingly to this view, genes capable to regulate the balance between symmetric and asymmetric cell division may influence embryonic neurogenesis and, above all, they can extinguish/ increase the number of cells fated to acquire features of long term proliferating elements. Indeed, defects in cell division occurring in the developing brain can generate microcephaly in human. The deviation from the normal brain development has been associated to a severe reduction of brain size and cognitive defects [17]. The product of Abnormal Spindle-like Microcephaly associated gene (AspM) is an interesting candidate to control the balance of symmetric and asymmetric cell divisions. Indeed, several experimental evidence suggests that AspM can regulate the positioning of the mitotic spindle during neuroblast cell division [18] and mutations in ASPM represent a leading cause of primary microcephaly, possibly interfering with cell proliferation/differentiation equilibrium of cortical neuroblasts [19,20]. Furthermore, ASPM is highly expressed in glioblastomas. In this cells, as well as in NSCs, the inhibition of ASPM dramatically affects their in vitro cell cycle progression [21]. The inactivation of AspM in transgenic mice leads to a significant reduction of the brain size and a massive reduction of germ cells [22]. Altogether these data highlight the potential role of AspM in controlling self-renewal of RG cells, and put forward the intriguing hypothesis that a cohort of AspMexpressing cells might foster the maintenance of symmetrically dividing long term progenitors. However, functional experiments supporting this hypothesis are still missing.
Here we used a novel AspMCreER T2 inducible allele to perform in-vivo long term fate mapping of a relative earlier cohort of embryonic AspM-expressing cells during neurogenesis. Cre mediated recombination was induced at the beginning of neurogenesis in a selected cohort of proliferating RG cells that were subsequently traced during forebrain development and in post natal neurogenic niches. This cohort of early AspM-expressing cells give raise to neurons, astrocytes and oligodendrocytes of the mature brain. In addition, a subset of them was maintained in germinal niches of the post natal brain as proliferating progenitors that express several molecular features in common with aNSCs, including self renewal. We next extended fate mapping experiments to the adult brain by changing the activation paradigm of the AspMCreER T2 allele. We next tested the contribution of embryonic AspM-expressing cells to the maintenance of proliferating pools of the brain, by the selective killing of these cells.

Results
AspM-CreER T2 transgene targets Cre-mediated recombination in long term proliferating cells of the cerebral cortex In order to trace the progeny of early AspM expressing cells, we generated a transgenic mouse line, in which the inducible Cre recombinase (CreER T2 ) [23,24,25] -i.e. transiently activated by injecting mice with Tamoxifen (Tam) [26,27]-, was placed under the control of AspM cis-acting regions ( Figure S1A and [material method] section for a detailed description of the construct). AspM and Cre mRNA distributions, were compared on parallel coronal sections of both embryonic (E15.5) and adult (P30) AspMCreER T2 brains by radioactive in-situ hybridization. Both transcripts were expressed in the VZ [19] of the E15.5 brain (Figure S1B, C) and in cells of the post-natal SVZ ( Figure S1D, E). RT-PCR experiments on microdissections of peri-ventricular regions of the lateral ventricle confirmed that both AspM and Cre mRNAs are expressed in the postnatal SVZ ( Figure S1F).
To analyze the recombination efficiency and to trace AspMexpressing cells during neurogenesis, we crossed mice carrying the AspMCreER T2 allele with mice carrying the Cre-dependent reporter allele Rosa26YFP [28], in which a floxed stop codon was cloned upstream YFP to allow genetic fate mapping. Double transgenic mice were injected twice with Tamoxifen (Tam) at E10.5/11.5 and brains were collected at E12.5 to study the expression of the YFP reporter after a short Cre induction. YFP + cells were preferentially placed in germinal niches of injected brains. Within the VZ/SVZ, cluster of YFP + cells also coexpressed the AspM mRNA ( Figure S1G). Fifteen (64.1)% of AspM expressing cells recombined the Rosa26YFP locus after Tam administration ( Figure S1H). The expression of AspM and the Cre-mediated recombination of Rosa26YFP locus were also investigated in post natal brains. Although AspM is greatly reduced in adult brains [19], the Cre-mediated recombination of Rosa26YFP locus occurred in approximately 12% of SVZconfined AspM + cells of P30 double transgenic mice treated with Tam for 5 days ( Figure S1I). AspMCreER T2 /Rosa26YFP embryos were further treated with Tam at E11.5/E12.5, and brains were subsequently analyzed at E13.5. Cells expressing YFP were preferentially clustered in the germinal region of the cortical wall, the vast majority of them expressed the RG cell marker RC2 [29], ( Figure 1A). Very few, if any, of them co-localized with SVZ mitotic cells expressing the phospho-Histone 3 (pH3) marker ( Figure 1B). Tangential versus radial cell divisions were analyzed on E12.5 (pulsed with Tam at E10.5/11.5) and E13.5 (pulsed with Tam at E11.5/12.5) embryos, by staining sections for YFP and pH3 ( Figure 1C, D). Accordingly to previous results [30], more than 80% of YFP + mitosis displayed a cleavage plane perpendicularly oriented to the ventral surface ( Figure 1E, upper panel). This number, however, slightly decreased at E13.5, possibly for the presence of many mitosis with unknown orientation ( Figure 1E, lower panel). Accordingly, only 6.12% (63.6) and 25.1% (64.0) of pH3/YFP double positive mitosis exhibited an horizontal cleavage plane (radial oriented mitotic spindle) at E12.5 and E13.5 respectively. These results confirmed a previous report [18] that indicates the expression of AspM enriched in tangential mitosis ( Figure 1E). E13.5 mice, pulsed with Tam at E11.5/E12.5, received a single injection of 5-ethynyl-29-deoxyuridine (EdU) one hour before the sacrifice to label proliferating cells ( Figure 1F). 30.2% (62.2) of cortical YFP + cells incorporated the EdU tracer ( Figure 1H). On the other hand, parallel sections stained for TuJ1 and YFP ( Figure 1G) showed that only 16.9% (61.7) of YFP + cells co-expressed the neuronal post mitotic marker TuJ1 ( Figure 1H).
We next investigated the distribution of E12.7/E13.2 traced AspM-descendants in germinal niches of the post natal brain. YFP + cells were still detected in the post natal germinal niches of the P30 brain. In the SVZ we found that 39.2%(621.6) of YFP + cells coexpressed the proliferating marker Ki67 ( Figure 2I) and incorporated the S-phase tracer EdU ( Figure 2L) [32]. YFP + cells coexpressed also markers that are expressed by subsets of adult neural progenitor cells, such as: Olig2 [33] (Figure 1M), and the neuroblast markers Dcx and PSA-NCAM ( Figure 2N and not shown), [34]. We next generated whole mount preparations of lateral wall that allow the visualization of the entire SVZ [35,36]. Samples stained for YFP and GFAP, were analyzed by confocal microscopy. Several GFAP/ YFP double positive cells, displaying long cellular bundles -i.e. possibly belonging to the type B cell population [37,38]-, were detected within the SVZ ( Figure 2O). In addition, double positive cells were often placed around clusters of YFP + cells displaying the bipolar morphology of migrating neuroblasts ( Figure 2O).
Long term descendants expressing YFP, were also found within the SVZ of P90 double transgenic brains. Although their number was slightly reduced at this stage, YFP + cells co-expressed markers of neural progenitor cells such as: GFAP, Olig2 and PSA-NCAM ( Figure S2A-C).

AspM descendants located in the adult SVZ maintain features of neural progenitor cells
To determine whether AspM-CreER T2 long term descendants that are detected in the adult SVZ, display features of neural progenitor cells, we tested their multipotency and self renewal capability by using standard clonal colony formation assay [44,45]. AspMCreER T2 /Rosa26-YFP double transgenic mice were injected with Tam at E12.5 and E13.5, cells were then collected from lateral ventricles of P30 brains and primary neural stem cell cultures were established [46]. Robust YFP expression levels were detected in many cells of SVZ-derived neurospheres, over serial in vitro passages (IVP) (not shown). Starting from the IVP-4, neurospheres cultures were dissociated in single cells and used in a standard clonogenic assay [4,44]. Because each neurosphere was derived from a single cells, spheres were either entirely positive or negative for YFP ( Figure 4A). In this assay, both single YFP + and YFP 2 cells give raise to primary, secondary and tertiary neurospheres with similar frequencies ( Figure 4B, C). Furthermore, dissociated YFP + cells from tertiary neurospheres were propagated in vitro for more than 10 IVPs, maintaining features of neural stem cell cultures (not shown). Starting from in IVP-5, dissociated cells are also capable to generate neurospheres of different dimensions, in a size based Neural Colony Forming Cells (NCFC) assay [45]. In this assay, dissociated cells were plated in a mitotic-containing collagen matrix and maintained in vitro for 3 weeks before assaying colony sizes. YFP + cells give raise to neurospheres of different size, and 6.8% (61.8) of them generated neurospheres of .1 mm in diameter, that presumably derive from bona fide NSCs ( Figure 4E, F). Moreover, three independent large YFP + neurospheres were dissected from plates, treated with collagenase before plating, and then propagated as YFP + neurosphere cultures over serial IVPs (not shown). Dissociated YFP + neurospheres were also assayed for multipotency in-vitro. YFP + cells, derived from large neurospheres, propagated for five IVPs, were then plated on matrigel coated dishes and kept in culture for seven days without growth factors. Differentiated YFP + cells produced GFAP + astrocytes (85.165.6%, Figure 4G, J), TuJ1 + neurons (6.561.2%, Figure 4H, J), and O4 + oligodendrocytes (1.760.7%, Figure 4I, J). Taken together, these results suggest that YFP + cells descending from E12.5/E13.5 embryonic AspM + forerunners and derived from the P30 SVZ, maintain in-vitro molecular features of bona fide neural progenitor cells of the SVZ.   Nestin-GFP flox -TK transgenic mice express the TK gene in RG cells of the developing forebrain and in aNSCs of the SVZ In order to establish the functional role of forebrain AspMexpressing cells, we generated a further transgenic mouse line in which a floxed GFP gene was cloned downstream Nestin regulatory regions [47] and upstream to the suicide Thymidine Kinase (TK) gene ( Figure S3A). Nestin-GFP flox -TK mice were characterized for the presence of GFP + cells in both developing and post natal brains. During forebrain development, GFP-expressing cells were preferentially placed in the VZ/SVZ of the cortical field, always displaying the morphology of RG cells ( Figure S3B). Proper expression of the Nestin-GFP flox -TK cassette was assessed during forebrain development, by comparing GFP and Nestin expression patterns on coronal section of E12.5 and P0 brains ( Figure S3C, D). Virtually all GFP + cells of E12.5 and P0 brains co-expressed Nestin, thus suggesting that the expression pattern of GFP fully overlaps endogenous Nestin. The presence and the distribution of GFPexpressing cells were also assayed in the post natal brain. P30 brain coronal sections, displayed many GFP/Nestin double positive cells in the SVZ ( Figure S3E). Some of them, expressing the GFP at high levels, were also positive for the neural stem cells marker Id1 [48] ( Figure S3F). Nestin-expressing cells belong to the slow dividing cell population which displays a slow cell turnover. Accordingly, very few GFP + cells of the SVZ incorporated the S-phase tracer EdU, after acute administration (10 hours), ( Figure S3G). Altogether, these results confirm that Nestin-GFP flox -TK transgenic mice express both GFP and TK genes in RG cells of the developing brain and in slow dividing cells of the adult SVZ. By crossing these mice with AspM-CreER T2 mice, we could now kill virtually all AspM/Nestin expressing cells of the developing forebrain.
Selective killing of AspM-CreER T2 /Nestin-GFP flox -TK cells severely impairs forebrain development and Vz/SVZ cell proliferation The activation of the TK gene was induced in AspM-CreER T2 / Nestin-GFP flox -TK mice by injecting Tam at days E12.7 and E13.2. Selective killing of AspM/Nestin + cells was obtained by treating mice with Gancyclovir (GCV) from E14.5 until E18.5. One hour before the sacrifice, at E18.5, mice received the EdU tracer to study proliferating cells of the brain. Double transgenic brains injected with Tam/GCV were significantly reduced in their size, when compared with control embryos (Figure 5A, D; Figure 6A, C). In particular, the dimensions of dentate gyrus and cortical wall were severely reduced ( Figure 5A, D). The selective killing of E12.5/E13.2 AspM/Nestin + forerunners affected cell proliferation in the cortical wall ( Figure 5B, E) and the hippocampus ( Figure 5C, F) of E18.5 embryos, as measured by Ki67 + and EdU + staining. In addition, pH3 + mitosis placed at the ventricular lining were significantly reduced in double transgenic mice treated with Tam/GCV ( Figure 5I-K). TK/GCV paradigm has been successfully used in tumor cells and graft versus host disease [49]. However, the cytotoxic effects mediated by GCV administration might be extended also to TK untransduced cellsi.e. the bystander effect- [50]. To assess the specificity of GCVmediated cell death occurring in Tam/GCV treated embryos, we generated AspM-CreER T2 /Nestin-GFP flox -TK/Rosa26YFP transgenic mice. Triple transgenic mice were firstly injected with Tam at E12.7/E13.2 and then treated with GCV at E14.5 and E15.5. Mice were sacrificed at E16.5 for the detection of both the activated Caspase 3 (Casp3a) and the YFP markers. Very few Casp3a + cells were detected in control mice -i.e. embryos lacking the Nestin-GFP flox -TK allele ( Figure 5D) but receiving GCV-. On the other hand, the analysis of triple transgenic mice showed that the number of Casp3a + cells specifically increased in VZ-confined YFP expressing cells ( Figure 5H). Although we cannot exclude that some YFP 2 cells undergo to cell death during GCV treatment, these results suggest that the vast majority of dying cells belong to YFP + cell population that expresses both AspM-CreER T2 and Nestin-GFP flox -TK transgenes.

Selective killing of AspM-CreER T2 /Nestin-GFP flox -TK cells severely impairs laminar organization of the cortex
The reduction of VZ proliferating cells occurring upon Tam/ GCV treatments was also accompanied to a significant reduction of Tbr2 expressing cells -i.e. BP cells-in E18.5 double transgenic mice ( Figure 6B and D). Alterations of cell proliferation in germinal niches of the developing brain can also impact on cortical lamination. We therefore looked for changes in the cell distribution and neuronal density of cortical layers in E18.5 Tam/GCV treated AspM-CreER T2 /Nestin-GFP flox -TK brains. We performed the immuno detection of FoxP2, which is predominantly expressed by early born deep-layer neurons [51], in both transgenic and control mice. The total number of FoxP2 + cells, however, did not change between double transgenic and control mice treated with Tam/ GCV, as well as their distribution within the cortical wall ( Figure 6C-E). We next extended our investigation to other cortical layers by staining sections for Ctip2 and Cux1 which are expressed in layers V [52] and layers II-IV [53] respectively ( Figure 6F-K). Double transgenic mice showed consistent thinning of upper cortical layers that was accompanied by a preferential reduction of superficial later-born neurons, as demonstrated by a significant reduction of Ctip2 + and upper Cux1 + cells ( Figure 6F, K). Thus, starting from E14.5 the selective killing of AspM-CreER T2 /Nestin-GFP flox -TK expressing cells impaired neurogenesis and reduced the number of neurons of upper cortical layers.

AspM expressing cells of the SVZ are proliferating progenitors that generate GCL neurons
We next investigated whether post-natal SVZ-restricted AspM expressing cells ( Figure S1D) do functions as neural progenitor cells. Starting from P30, AspM-CreER T2 /Rosa26-YFP mice were injected with Tam for 5 days and then sacrificed 6 days after the last injection. Several YFP + cells, were detected along the SVZ, some of them incorporated the S-phase tracer EdU ( Figure 7A), and were Ki67 + ( Figure 7B). YFP + cells of the dorsal SVZ  Figure 7D) and PSA-NCAM ( Figure 7E) [54]. We next tested whether AspM + cells of the adult SVZ retain the capacity to generate neuroblasts migrating to the olfactory bulbs (OB) [55]. AspM-CreER T2 /Rosa26-YFP mice were injected with Tam for 5 days as above, and sacrificed 17 days after the last Tam injection. In this experimental paradigm, the vast majority of YFP + cells were detected at the end terminal of the rostral migratory stream (RMS) ( Figure 7F). These cells exhibited the morphology of migrating neuroblasts and expressed the PSA-NCAM marker ( Figure 7F, G). A restricted number of them reached the OB granular region, and displayed the morphology of differentiated neurons (not shown).
The vast majority of SVZ-derived precursors migrate along RMS to OB, where they differentiate into local interneurons of the granular (GCL) and glomerular layers (GL) [56,57]. Periglomerular neurons are an heterogeneous cell population because these cells express different neurochemical markers and calcium binding proteins [58,59], whereas interneurons in the GCL are an homogeneous population of GABAergic cells [58]. P30 AspM-CreER T2 /Rosa26-YFP mice were injected with Tam as above and brains were collected after 30 and 60 days from the last Tam injection. To examine the neurochemical phenotype of YFP + cells, OB coronal sections were labeled for Calretinin (CR), Calbinding (CB), Paravalbumin (PV) and Tyrosine hydroxylase (TH) [60,61]. Thirty days after the last Tam injection, the vast majority of YFP + cells (n = 320 cells) were placed within GCL and displayed the morphological phenotype of granular cells ( Figure 8A). None of them, however, co-expressed CR, CB, PV or TH markers (not shown). Accordingly, the GCL contained the vast majority of YFP + cells (n = 465 cells), also in mice collected after 60 days from the last Tam injection ( Figure 8B). Nevertheless, 1.2 (60.5)% of YFP + cells co-expressed CB ( Figure 8D), but none of them co-expressed CR, PV or TH markers (Figure8 C, F and G). Altogether these results suggest that AspM expressing cells of the adult SVZ differentially contribute to specific classes of OB neurons. While very few AspM descendants give raise to CB + cells, the vast majority of them contribute to the generation of GCL neurons.

AspM expression increased in growing neurospheres
We next examined whether post natal SVZ-derived NSCs maintain in vitro the expression of AspM after extensive passages. Neurospheres cultures were established from P30 SVZs and propagated for 20 IVPs. Single cells derived from these long lasting cultures were plated in neurospheres culture medium at the density of 8000 cells/mm 2 . Cells were then collected every day for the detection of AspM by real time PCR. Proliferating cells increased the size of neurospheres and rapidly increased AspM levels, that peaked at day 4 ( Figure S4A). However, when spheres reached their maximum dimensions, these levels rapidly declined ( Figure S4A). We next tested AspM expression in differentiating cultures. Cells were plated in the absence of growth factors and sampled each day to assess AspM expression levels. Accordingly to previous results [21], AspM expression rapidly dropped out in differentiating cell cultures ( Figure S4B).
In conclusion, our data indicate that AspM-expressing cells of the developing forebrain supply of proliferating cells the germinal niches of both developing and post-natal brains. Furthermore, our long term fate mapping revealed that early AspM expressing cells can also differentiate in neurons, astrocytes and oligodendrocytes of the post natal brain. Supporting the notion that AspM expression is maintained by highly hierarchical progenitors of the brain.

Discussion
The mechanisms by which proliferating cells of the developing brain are maintained within germinal niches of developing and post natal brains are largely unknown. Although, many experimental evidence supports the notion that adult neurons and glial cells derive from early neuroepithelial cells of the neural plate [62,63], the precise identification of SVZ progenitors forerunners remains elusive [1]. Before the onset of neurogenesis neuroepithelial cells acquire new molecular features and become RG cells [63]. During early neurogenesis, these cells preferentially do symmetric division to ensure the expansion the cortical anlage [64,65]. However, at later time points of the neurogenesis, dividing RG cells increase the rate of asymmetric cell divisions to generate neurons of the cortical plate [30]. Nevertheless, a subset of them is maintained into germinal niches of the brain to supply the adult SVZ of proliferating progenitor cells [1,66]. Thus, during brain development the balance between RGs self renewal and the generation of committed descendants -i.e. progenitors fated to differentiate in neurons and in glial cells-, is crucial to ensure the appropriate size of brain and to maintain proliferating progenitor in germinal niches. Accordingly to this view, several genes controlling cell division/differentiation have been involved to prevent either the aberrant growth of the cortical primordium or the depletion of long term proliferating cells [67,68,69].
The human ASPM product has been linked to this function, because mutations in its coding region lead to severe microcephaly [19]. On the other hand, aberrant ASPM expression is tightly associated with malignant progression of gliomas, and in vitro short interference of ASPM resulted in G1-phase cell cycle arrest of tumor cells [21]. Recently it has been also demonstrated that the inactivation of AspM in mice produces a significant reduction of brain size [22]. By performing a short-term fate mapping experiment on E11.5 AspMCreER T2 /Rosa26YFP embryos, we demonstrated that early forebrain AspM-expressing cells are RG cells that preferentially do symmetric cell divisions in the VZ. To  assess whether or not AspM-expressing cells can give raise to long lasting proliferating cells of the brain, we performed a series of long term fate mapping experiments. In these experiments, we analyzed a single cohort of YFP-expressing cells inducing a transient Cremediated recombination of the Rosa26YFP locus after the onset of neurogenesis [14]. Recombined cells were assayed for the expression of proliferation markers and for their capacity to incorporate S-phase tracer at different time points along neurogenesis and, above all, in post natal brains. Surprisingly, a substantial number of AspM descendants were maintained as proliferating cells in germinal niches of both developing and post natal brains. Within the adult SVZ, recombined cells derived from early AspM expressing cells start to express new molecular markers which are typically associated to SVZ restricted adult neural progenitor cells. However, SVZ progenitors display heterogeneous features in term of markers and proliferating features. Thus, we took advantage of using clonogenic assays to demonstrate, in vitro, self renewal features and multipotency of these recombined cells [4]. Two independent assays demonstrated that recombined cells obtained from the adult SVZ can generate neurospheres with the same efficiency of wild type progenitors. Once neurospheres cultures were established, recombined cells were propagated for many in vitro passages, and demonstrated their multi-potency in vitro, when growth factors were removed from these cultures [70]. These data reinforce the idea that RG cells can supply the adult SVZ of proliferating cells and extend previous findings showing that progenitor cells fated to occupy the SVZ, as proliferating cells, derived from end gestational RG cells [1,11,14].
Since recombined cells detected within the SVZ account for a relative small percentage of total recombined cells, we tested whether early AspM forerunners are multipotent also in vivo. Adult brain analysis of double mutant mice revealed that AspM descendants differentiated in neurons, astrocytes and oligodendrocytes, thus suggesting that also early AspM-expressing cells are multipotent progenitors.
To test the role of early AspM-expressing forerunners, we performed their selective killing by using a novel transgenic mouse line: Nestin-GFP flox -TK mice. These mice express the TK gene under the control of both Nestin promoter region [47] and a floxed GFP cassette. TK/GCV system may induce the killing of cells not expressing the TK transgene -i.e. the so called ''bystander effect'' [50]. To rule out this possibility we generated triple transgenic mice in which recombined cells were traced by Rosa26YFP allele. The analysis of these mice demonstrated that cell death mainly occurs in recombined cells, expressing all transgenes, of the germinal brain niches. Upon Cre mediated recombination, the excision of the GFP cassette allows the expression of the TK gene. Starting from E12.7/E13.2, we were able to kill a single cohort of Nestin/AspM-expressing cells, causing a severe depletion of proliferating cells in the end gestational brain and a significant reduction of upper cortical neurons. This result reinforced the idea that early AspM-expressing cells are high hierarchical progenitors fated to maintain a proliferating pool during neurogenesis. Indeed, the selective ablation of these cells caused the significant reduction of RG cells in end gestational brains We next investigated whether SVZ restricted adult AspM expressing cells, retain features of multipotent cells, as we have found in embryonic AspM forerunners. To achieve this goal, we performed another series of fate mapping experiments on AspM-CreER T2 /Rosa26-YFP mice, changing the paradigm of Tam administration. As we have demonstrated by radioactive in situ hybridization and immunofluorescence, AspM expressing cells are prevalently located at the ventricular lining. We firstly demonstrated that AspM expressing cells can also express the Cre recombinase and that a substantial number of them can recombine the Rosa26YFP locus. Thus, we firstly performed a short term fate mapping analysis of these cells to assess if they do functions of proliferating cells. Mice were administered with Tam at P30 for five days, and then a short washing out period was applied before assaying recombined cells for their capability to incorporate the Sphase tracer EdU. Of note, in the adult brain neither neurons nor glial cells of the parenchyma displayed any recombination of the Rosa26YFP locus. Indeed, YFP expressing cells were detected only within the adult SVZ and a significant number of them were proliferating cells. Since the vast majority of SVZ proliferating cells are fated to generate neurons of the olfactory bulbs [35], we extended our observations by performing a long term fate mapping experiments of SVZ recombined cells. By increasing the washing out period, we were able to trace SVZ descendants along the RMS. These cells expressed the type-a marker PSA-NCAM and properly reached the olfactory bulbs. Extending the washing out time we also assessed that adult AspM descendants preferentially give rise to GCL neurons. Indeed, very few YFP + cells were detected within the GL and mitral layers. This result suggests that AspM expressing cells of the adult SVZ might be commitment to generate only a specific subset of OB neurons. However, it remains to elucidate whether this feature may reflect differences in the progenitor pools of SVZ cells. Indeed, our current study cannot provide a sufficient resolution to determine the mechanism/s of AspM differentiation in vivo and, above all, molecular mechanisms that drive the expression of AspM in neural progenitor cells. AspM-CreER T2 transgenic mouse line was generated by using ''recombineering'' technology. Briefly, the BAC expressing CreER T2 gene [24] under the control of AspM regulatory regions was cloned by targeting the genomic region (approximately 140 kb) of the BAC RP-24-267-F8, which contains the entire locus and approximately 70 kb of its promoter region. A frt flanked Kanamycin gene was inserted downstream CreER T2 [24] for the selection in bacteria. Homologous recombination was obtained in EL250 E. Coli strain accordingly to ''recombineering'' protocols [71], (http://web.ncifcrf.gov/research/brb/recombineeringInformation.aspx). The CreER T2 -frt-Kana-frt cassette was inserted in the first methionine of AspM gene. Upon recombination, recombined BAC was tested for proper insertion of CreER T2 gene by direct sequencing. The frt flanked Kanamycin gene was then removed by activating the expression of the inducible Flpe recombinase in EL250 bacteria. DNA was injected in FVB zygotes and then targeted zygotes were surgically implanted into foster mothers. Founders were backcrossed in C57/BL6J mouse strain. Fate mapping experiments were performed by crossing AspM-CreER T2 transgenic mice with Rosa26YFP transgenic mice (obtained from Jackson laboratory).

Gene targeting and transgenic mouse lines
We also generated Nestin-GFP flox -TK transgenic mouse line by using lentiviral technology. We used a third generation sinlentiviral vector [72] to generate NestfloxGFPfloxTK targeting lentivirus. A conserved 1.8 Kb second intronic region of the rat Nestin [73] gene was cut out (XbaI, HindIII) from the p401ZgII plasmid (a gift of Dr. McMahon, Harvard University, Cambridge, MA) and sub-cloned upstream to the minimal promoter of the Hsp68 gene. The loxP sites were produced synthetically and were cloned upstream and downstream to the EGFP coding region. EGFP was obtained from the BamHI-SalI fragment of the #277 PGK-GFP lentivirus construct (a gift of Dr. Naldini, San Raffaele Hospital, Milan, Italy). Downstream to the EGFP sequence we sub-cloned the suicide gene Thymidine Kinase. BamHI-XbaI fragment TK coding sequence was cut out from the pBSIISK-TK construct. Finally, we inserted the IRES-lacZ fragment obtained from the pMODLacZnls plasmid downstream the TK gene. Transfer vector and the packaging plasmids: pMDLg/pRRE, pRSV-REV and pMD2.VSVG were transfected into 293T cells [74] by using the calcium phosphate precipitation method. Then, 14-16 hours later the DNA transfection, medium was replaced and 36 hours later cells supernatants were collected and filtered through a 0.22-mm pore nitrocellulose filter. To obtain high titer vector stock, the cell supernatants were concentrated by ultracentrifugation (55K g, 140 min, 20uC). Then, supernatants were discarded and the pellets were suspended in 100 ml of PBS 1x, split in 20 ml aliquots and stored at 280uC. LV stock was titrated by infecting HeLa cells with serial dilution of the viral stock and flow cytometry assay. Nestin-GFP flox -TK founder mice were generated by the NestfloxGFPfloxTK vector injection into the perivitelline space of C57Bl/6 zygotes. Two independent lines (#7457 and #7454) were generated and characterized for the presence of one copy of the inserted transgene. Experiments were conducted on #7454 transgenic mouse line. Mice were genotyped by PCR using genomic DNA and the following primers: FW, 59-AACTTTCCCCGGAGAGCATCCACGC, Rev1, 59-TAGGT-CAGGGTGGTCACGAGGGT, Rev2, TGTTGATGGCAGG-GGTACGAAGC.

Tamoxifen, Ganciclovir & EdU administrations
Tamoxifen (Sigma) was dissolved in EtOH/Sunflower oil 10%/ 90% and injected at the following concentration 160 mg/kg. For the embryonic induction of AspM-CreER T2 /Rosa26YFP double transgenic mice, Tam was intraperitoneally injected into pregnant mice at embryonic days of E10.5//E11.5 and E11.5/E12.5 for short term fate mapping; E12.7 and E13.2 for long term fate mapping. For induction in adult mice, Tam was injected intraperitoneally at the same concentration, into 4 weeks old mice once a day for 5 consecutively days. Ganciclovir (GCV, Roche) was administered in pregnant females as single daily intraperitoneal injection at a dose of 100 mg/kg starting from E14.5 until E16.5 or E18.5. Cell proliferation was assayed in vivo, in adult and embryonic forebrains, injecting mice with the S-phase tracer 5ethynyl-29-deoxyuridine (EdU, Invitrogen) at the following concentration: 100 mg/kg.
Light (Olympus, BX51 equipped with 46 and 206 objectives) and confocal (Leica, SP5 equipped with 206 and 406 objectives) microscopy was performed to analyze tissue staining. Analyses were performed by using Leica LCS lite and Adobe Photoshop CS software. Cell counts were done in the cerebral cortex of E13.5, E15.5 and P0 brains. Cells were also counted in cortical fields of P30 brain in a region encompassing the anterior bregma +1.2 and the posterior bregma 20.5.

In situ hybridization
In situ hybridization were performed as previously described [75,76]. Briefly, 10 mm-thick brain sections were post-fixed 15 min in 4% paraformaldehyde, then washed three times in PBS. Slides were incubated in 0.5 mg/ml of Proteinase K (Roche) in 100 mM Tris-HCl (pH 8), 50 mM EDTA for 10 min at 30uC. This was followed by 15 min in 4% Paraformaldehyde. Slices were then washed three times in PBS then washed in H2O. Sections were incubated in triethanolamine (Merk) 0.1 M (pH 8) for 5 min, then 400 ml of acetic anhydride (Sigma) was added two times for 5 min each. Finally, sections were rinsed in H2O for 2 min and air-dried. Hybridization was performed overnight at 60uC with a-UTP-P 33 (G.E.) riboprobes at a concentration ranging from 10 6 to 10 7 counts per minute (cpm). The following day, sections were rinsed in SSC 5 X for 5 min then washed in formamide 50% (Sigma)-SSC 2 X for 30 min at 60uC. Then slides were incubated in ribonuclease-A (Roche) 20 mg/ml in 0.5 M NaCl, 10 mM Tris-HCl (pH 8), 5 mM EDTA 30 min at 37uC. Sections were washed in Formamide 50% SSC 2X for 30 min at 60uC then slides were rinsed two times in SSC 2X. Finally, slides were dried by using ethanol series. Lm1 (G.E.) emulsion was applied in dark room, according manufacturer instructions. After 10 days, sections were developed in dark room, counterstained with Dapi and mounted with DPX (BDH) mounting solution. The following probes were used: mouse AspM riboprobe (a gift of Dr. Walsh, Beth Israel Deaconess Medical Center, Boston-USA) and Cre riboprobe that was generated by cloning the SalI-PstI fragment corresponding to the entire Cre cds, from pCDNA3-Cre vector into pBSII-SK vector. For Digoxigenin in situ hybridization was performed as previously described [77]. Microphotographs of sections were digitalized in dark field light microscopy (Olympus BX51, and 46 objective) by using a CCD camera (Leica). Images manipulations were performed by using Adobe Photoshop CS. To confirm the specificity of the different RNA probes, sense strand RNA probes (showing no signal) were used as negative controls.

Laser capture microdissection & Real Time PCR
P30 mice were deeply anesthetized and then were rapidly perfused transcardially with RNase free 0.9% saline containing 10 U/ml of heparin. Brains were removed from the skulls, OCT embedded and rapidly frozen by immersion in liquid nitrogen. Laser mediated microdissections were obtained by using a Leica AS LMD equipment (objective 206). Briefly, 25 mm coronal sections were generated starting from bregma +1.2 to bregma 20.5. Then, sections were dehydrated. From each slide both the SVZ and striatum were captured. An average of 55 (610) dissections per mouse were collected and total RNA was extracted by using RNeasy Mini Kit (Qiagen) according to manufacturer's recommendations including DNase digestion. cDNA synthesis was performed by using ThermoScriptTM RT-PCR System (Invitrogen) and Random Hexamer (Invitrogen) according the manufacturer's instructions for standard PCR by using following primers: AspM f: GCC-TCTCTCGGCCTCCATCAGCCTCCTTG AspM r: CAGG-CAGCGTTGCTTCTATCAACA GCGTCG; Cre f: GCCTGC-ATTACCGGTCGATGCAACGA Cre r: GTGGCAGATGG CGCGGC AACACCATT. Gene expression analysis was performed by using the LightCyclerR 480 System (Roche) and SYBRR Green JumpStart TM Taq ReadyMix TM for High Throughput QPCR (Sigma). Each sample was normalized by using the housekeeping gene Histone H3 with the following primers: H3 F: GGTGAAGAAACCTCATCGTTACAGGCCTGGTAC H3R: CTGCAAAGCAC CAATAGCTGCACTCTGGAAGC.

Neural stem cell cultures generation and maintenance
Neural stem cell cultures were obtained from P30 AspM-CreER T2 /Rosa26YFP mice primed with Tam at E12.7 and E13.2, as previously described [75]. Briefly, three mm-thick coronal sections were obtained from the anterior forebrain of P30 double transgenic mice (2 mm from the anterior pole of the brain). Dorsal SVZs were carefully dissected by using a fine scissor in the following dissociating medium: Earl's Balanced Salt Solution (Gibco) supplemented with 1 mg/ml Papain 27 U/mg; (Sigma), 0.2 mg/ml Cysteine (Sigma) and 0.2 mg/ml EDTA (Sigma). Then, dissected tissue was incubated in the same solution for 30 min at 37uC on a rocking platform. Finally, dissociated cells were plated in standard neuro-sphere growth medium Neurocult proliferation medium (Stem cell Technology) supplemented with EGF (20 ng/ml) and FGF2 (10 ng/ml). For each in vitro passage, single cells were obtained by incubating neurospeheres in Accumax (Sigma) for 10 minutes, then 8000 cells/mm 2 were plated on T75 plastic flasks (Nunc). Neurospheres were propagated in vitro and assayed for self-renewal and cell differentiation after 4 in vitro passages (IVP), as previously described [46].

Neurospheres formation assay
Neurospheres raised from AspM-CreER T2 /Rosa26YFP neural stem cells (n = 2 independent cell cultures) after the 4 th IVP were dissociated in single cell suspensions with Accumax as above, and plated as single cells in a 96 wells multiwell. After 10 days, the number of both YFP + and YFP 2 neurospheres with a diameter .100 mm were counted in three independent experiments. To determine the cell renewal capacity of these cells, neurospheres were then dissociated and then re-cultured under the same condition as primary cultures. Again, ten days later, we determined the number of secondary neurospheres. This experiment was repeated for the detection of tertiary neurospheres. Primary, secondary and tertiary neurospheres were measured under inverted microscope equipped with fluorescence (Axiovert S100TV) and data were expressed as percentages 6 S.E.M.

Neural colony forming cell assay
AspM-CreER T2 /Rosa26YFP neural stem cells (n = 2 independent cell cultures) at 5 th IVP were dissociated in single cells as above described, diluted in Neurocult proliferation medium at the concentration of 2.2610 5 cells/ml. Then, cells were mixed with collagen solution accordingly to manufacturer's instructions (Stem cell Technology) and approximately 2.500 cells were plated on 35 mm dishes. Dishes were incubated in a 100 mm petri dishes containing open 35 mm dishes filled with sterile water. Cultures were maintained at 37uC, 5%CO 2 for 21 days and then visually assessed by using an inverted microscope equipped with fluorescence (Axiovert S100TV). Dishes were placed on a gridded scoring dish (Stem cell Technology) and scored at low magnification. Colony were classified into one of three categories: a, less then 0.5 mm diameter; b, 0.5-1 mm diameter; and c, more than 1 mm. Data were expressed as percentages 6 S.E.M. of three independent experiments.

Neural stem cell cultures differentiation assay
To analyze multipotency in-vitro, individual spheres (e.g., tertiary established neural stem cells cultures) were mechanically dissociated, and single cells plated onto MatrigelH (BD)-coated glass coverslips (12 mm diameter) in the presence of 20 ng/ml of FGF-2. After 72 hr in vitro, cells were shifted to control medium containing 1% fetal bovine serum (FBS) and 5 days later fixed in 4% paraformaldehyde pH 7.4 and processed for immune-fluorescence. Fixed cells were then rinsed with PBS 1x and incubated for 30 min at in PBS containing 10% normal goat serum (NGS)/ 0.1% Triton X-100. Cells were then incubate in primary antibodies for 2 hours. After washing, cells were incubated for 1 hour with the appropriate secondary antibodies. Samples were rinsed three times with PBS, counter stained with Dapi, and once with distilled water. Then slides were mounted with Fluorsave (Calbiochem). The following primary antibodies and antisera were used: rabbit anti-neuron specific tubulin type III (1:1000, Tuj1, Covance), mouse a-O4 (1:1000, Millipore), rabbit a-GFAP (1:1000 Dako) and chicken a-GFP (1:1000, Abcam). Appropriate fluorophore-(Alexa-fluor 488 and 546, Molecular Probes) conjugated secondary antibodies were used. The Samples were examined and photographed using a Olympus BX51, and 206 objective fluorescence microscope. The number of cells immunoreactive (IR) for different antigens was counted in n$20 non overlapping fields per sample (up to a total of n$1000 cells per sample) and data were expressed as mean percentage of immunoreactive cell (over total counted nuclei) 6 S.D. from a total of n$3 independent experiments. Figure S1 Design and activity of AspM-CreER T2 transgene. Panel A shows the targeting construct containing the CreER T2 gene, the Kanamycin/Neomycin resistance cassette and the insertion point on AspM locus (top). In the bottom panel is shown the construct after the excision of resistance cassette. Panels B and C show forebrain coronal sections of E15.5 AspM-CreER T2 embryos probed for AspM (B) and Cre (C) detection, by radioactive in situ hybridization (n = 4). Panels D and E show coronal sections derived from P30 transgenic mice probed for AspM (D) and Cre (E) (n = 3). Panel F shows a representative P30 brain coronal section from AspM-CreER T2 mice (n = 3) that was used for laser capture microdissection of the SVZ and the striatum. Total mRNA was extracted from each pool of sections and was used for RT-PCR detection of AspM and Cre transcripts (F, right panel). AspM-CreER T2 /Rosa26YFP mice were pulsed with Tam at E10.5/11.5 and collected at E12.5 (n = 4). Panel G shows the in situ hybridization for AspM coupled with the immuno-detection of YFP in a double transgenic brain. Arrows in panel G indicate double positive cells placed within the VZ. Panel H shows confocal sectioning of an adjacent section probed for AspM and YFP detection that confirmed the presence of double positive cells in germinal niches of AspM-CreER T2 /Rosa26YFP brain.