WO2011045765A1 - TEMPORALLY PATTERNED OVEREXPRESSION OF Foxg1 AND Emx2 TRANSCRIPTION FACTOR GENES IN NEURAL PRECURSORS FOR BRAIN TISSUE REPAIR - Google Patents

Abstract

The present invention relates to a recombinant expression vector suitable for gene therapy and able to effectively express in a differentially state specific manner Emx2 and/or Foxg polypeptides or functional fragments thereof in neural stem cells (NSCs) and/or in neuron- restricted progenitors (NPs)as well as a neural stem cell (NSCs), or a neuron-restricted progenitor (NPs) cell or a mixed cell population thereof transformed with such10 recombinant expression vector.

Description

TEMPORALLY PATTERNED OVEREXPRESSION OF Foxgl AND Emx2 TRANSCRIPTION FACTOR GENES IN NEURAL PRECURSORS FOR BRAIN

TISSUE REPAIR Field of the invention

Neural stem cells (NSCs), both endogenous and exogenously delivered, are a promising tool for repairing the brain parenchyma, damaged upon trauma, ischemia, neurodegeneration. However their massive and customary employment for brain repair still requires the solution of a variety of problems, including but not limited to, adequate expansion of the neural proliferating pool, its preferential channelling into neuronogenic rather than gliogenic differentiative pathways, promotion of neuronal differentiation and survival.

By integrated use of a variety of state of the art technologies (lentiviral transgene delivery, simultaneous use of promoters firing in different neural precursors and fluoroproteins with distinct spectral properties, multiple immunopanelling, multichannel fluorocytometry), the authors dissected the main biological activities displayed by two developmentally regulated transcription factors, Emx2 and Foxgl, in modulation of cortico-cerebral neural stem cells proliferation and differentiation. The authors found that these two factors regulate kinetics of such processes at multiple levels, impinging on: (a) histogenetic choice of neural stem cells (neuronal vs glial), (b) equilibrium between precursor proliferation and differentiation, (c) promotion of postmitotic neuronal maturation. All that makes Emx2 and Foxgl two highly promising molecular tools, to be exploited for transgene-promoted brain tissue repair. Technical background

Cortico-cerebral neuronogenesis is mainly restricted to embryonic life. It has been shown in some experimental systems (rodents and primates) that, after birth and in adulthood, this process may start again, following experimental hypoxia and/or ischemia of cerebral parenchima, as well as following delivery of specific chemo -physical insults. In this last case newborn neurons may differentiate consistently with their laminar-areal location, establishing appropriate synaptic connections with far targets. A reactivation of neuroblast proliferation has been demonstrated in patients affected by epilepsy and/or neurodegeneration as well. All this means that, in particular circumstances, the post-natal brain may restore neuroblast proliferation and replace its damaged and/or dead neurons. However the amplitude of these processes is almost always modest, so not eliciting any evident functional advantage.

Moreover, exogenous NSCs, derived from dissociation of neural tissue and transplanted into the damaged brain, may give rise to new neurons, especially if originating from embryonic central nervous system (CNS). The efficacy of such therapeutic approach is - however - very scarce and functional recovery is modest. Moreover, the employment of human heterologous embryonic tissues rises serious concerns, as for ethics and immunocompatibility. However, it seems that such problems may be overcome. In last years, several teams (23, 24, 25, 26, 27, 28) have shown that cortico-cerebral-like NSCs may be alternatively obtained in vitro, from embryonic stem cells (ESCs), with increasing degrees of purity. In addition, nowadays, surrogates of ESCs (normally obtained from the embryonic inner cell mass, ICM) may be obtained from common adult somatic cells, these surrogates are usually referred to as iPS (29, 30, 31,32, 33).

The co-exploitation of these discoveries makes the production of NSCs for transplantations independent of the availability of isogenic tissue, neural and/or embryonic. All this makes repair of brain damages ex vitro a concretely pursuable approach.

Now, whatever the selected approach be, relying on exploitation of endogenous NSCs or on transplantation of exogenous neural precursors, what are the main problems to deal with, to make such approach therapeutically effective? Basically, at least four: (a) getting a sufficient expansion of the starting proliferating pool; (b) channelling NSCs towards appropriate differentiation programs, preferentially neuronal rather than glial; (c) guiding newborn neurons to damaged areal and laminar districts; (d) integrating such neurons in pre-existing circuits. However, adequately modulating each of the such basic processes with ad hoc treatments, based on gene therapy or relying on complex pharmacological stimulations (that is the approach followed by many operators), may be tremendously complex. Moreover it may be even more complex to coordinate all of such manipulation in the frame of a unique therapeutic intervention. A radically new approach - the one the authors follow to approach the solution of problems (a) and (b) - might consist in re-activating some special genes, normally involved in early patterning of the cortico- cerebral primordium. These genes - in fact - act as true "gene hub" and, as such, should be in principle able to harmonically co-modulate several morphogenetic subroutines, so allowing to achieve - from a reduced number of NSCs - a sustained rate of neuronogenesis, in front of negligible gliogenesis.

The two transcription factors Emx2 and Foxgl, expressed in the cortico-cerebral primordium during early phases of its development, are crucial to the proper progression of a number of morphogenetic subroutines, including: specification of the cortico-cerebral field, its subsequent regionalization and arealization, laminar specification of cortico- cerebral neurons, radial migration of such neurons from their birthplace to the cortical plate. An involvement of these factors in control of proliferation/differentiation kinetics, as promoters of neural precursors proliferation and inhibitors of their post-mitotic neuronal differentiation, has been also reported. In the case of Emx2, this conclusion, achieved "in vivo", in two independent European labs (including the authors' lab), has been overturned "in vitro" by two others labs, European as well. In a similar way, Foxgl, although mainly known for its pro -proliferative properties, also promotes - in specific contexts - the activation of molecular markers linked to piramidal-glutamatergic differentiation. Finally, nothing is known about any capabilities of these genes to modulate gliogenesis.

Patent documents referring to the same technical field are listed below. However none of these documents disclose the modulation of Emx2 and/or Foxgl expression for neuro- therapeutical applications, namely for gene therapy. title Pat Appl No.

APPLICATION OF NEURAL STEM CELL NSCS-NT3 IN

CN101278942

REPAIR OF NEURON DAMAGE

NERVE REGENERATION PROMOTERS EP1685832

CELL THERAPY FOR CHRONIC STROKE EP 1949904

METHOD AND MATERIALS RELATING TO

EP1597355

NEUROGENESIS

USE OF LUTEINIZING HORMONE (LH) AND

CHORIONIC GONADOTROPIN (HCG) FOR

EP 1740202

PROLIFERATION OF NEURAL STEM CELLS AND NEUROGENESIS

REGENERATION AND REPAIR OF NERVE BY JP2005287512 INDUCTION OF HISTOGENESIS FACTOR

AMELIORATING AGENT FOR BRAIN DAMAGE US2009076019

ANTIPSYCHOTIC AGENTS STIMULATE

US2003064082 NEUROGENESIS IN BRAIN

METHOD AND SYSTEM FOR BIASING CELLULAR

US2006110440 DEVELOPMENT

HMG COA REDUCTASE MEDIATED MODULATION OF

US2008103105 NEUROGENESIS

MELANOCORTIN RECEPTOR MEDIATED

US2008108574 MODULATION OF NEUROGENESIS

IN VIVO AMPLIFICATION OF NEURAL PROGENITOR

US2008153167 CELLS

NEUROGENESIS BY MODULATING ANGIOTENSIN US2008167291

PROLACTIN INDUCED INCREASE IN NEURAL STEM

US2008181873 CELL NUMBERS

MODULATION OF NEUROGENESIS WITH BIGUANIDES

US2008188457 AND GSK3-BETA AGENTS

METHODS OF USING DEACETYLASE INHIBITORS TO PROMOTE CELL DIFFERENTIATION AND US2008248994 REGENERATION

4-ACYLAMINOPYRIDINE DERIVATIVE MEDIATED

US2009088449 NEUROGENESIS

ALISKIREN MODULATION OF NEUROGENESIS US2009197823

MEDICINAL PRODUCT AND METHOD FOR TREATMENT OF CONDITIONS AFFECTING NEURAL US6797264 STEM CELLS OR PROGENITOR CELLS

ANTIPSYCHOTIC AGENTS STIMULATE

WO03028651 NEUROGENESIS IN BRAIN

VHL PEPTIDE WO2006126263

METHOD OF TREATING BLINDNESS WITH HNT WO 9834485 HUMAN NEURONAL CELLS

Description of the invention

Authors readdressed functional characterization of Emx2 and Foxgl (GenBank accession numbers: human foxgl cDNA: NM 005249.3; human foxgl cds: CCDS9636.1; murine foxgl cDNAs: NM_008241.2, NM 001160112.1; murine foxgl cds: CCDS25899.1; human emx2 cDNA: NM 004098.3; human emx2 cds: CCDS7601.1; murine emx2 cDNA: NM_010132.2; murine emx2 cds: CCDS29937.1), the sequence of the polypeptides are reported below , following a new analytical approach. Human, hsa-emx2-polypeptide [ENSP00000358202 (252 aminoacids)], SEQ ID No. 1

MFQPAPKRCFTIESLVAKDSPLPASRSEDPIRPAALSYANSSPINPFLNGFHSAAAAAAGRGVYSN PDLVFAEAVSHPPNPAVPVHPVPPPHALAAHPLPSSHSPHPLFASQQRDPSTFYPWLIHRYRYLGH RFQGNDTSPESFLLHNALARKPKRIRTAFSPSQLLRLEHAFEKNHYWGAERKQLAHSLSLTETQV KVWFQNRRTKFKRQKLEEEGSDSQQKKKGTHHINRWRIATKQASPEEIDVTSDD

Mouse, mmu-emx2-polypeptide [ENSMUSP00000053361 (253 aminoacids)], SEQ ID No. 2

MFQPAPKRCFTIESLVAKDSPLPASRSEDPIRPAALSYANSSPINPFLNGFHSAAAAAAAGRGVYS NPDLVFAEAVSHPPNPAVPVHPVPPPHALAAHPLPSSHSPHPLFASQQRDPSTFYPWLIHRYRYLG HRFQGNDTSPESFLLHNALARKPKRIRTAFSPSQLLRLEHAFEKNHYVVGAERKQLAHSLSLTETQ VKVWFQNRRTKFKRQKLEEEGSDSQQKKKGTHHINRWRIATKQASPEEIDVTSDD

The mmu-vs-hsa sequence have an homology of 99.6%. Human,hsa-foxgl -polypeptide [ENSP00000371975 (489 aminoacids)], SEQ ID No. 3

MLDMGDRKEVKMI PKSSFSINSLVPEAVQNDNHHASHGHHNSHHPQHHHHHHHHHHHPPPPAPQPP PPPQQQQPPPPPPPAPQPPQTRGAPAADDDKGPQQLLLPPPPPPPPAAALDGAKADGLGGKGEPGG GPGELAPVGPDEKEKGAGAGGEEKKGAGEGGKDGEGGKEGEKKNGKYEKPPFSYNALIMMAIRQSP EKRLTLNGIYEFIMKNFPYYRENKQGWQNSIRHNLSLNKCFVKVPRHYDDPGKGNYWMLDPSSDDV FIGGTTGKLRRRSTTSRAKLAFKRGARLTSTGLTFMDRAGSLYWPMSPFLSLHHPRASSTLSYNGT TSAYPSHPMPYSSVLTQNSLGNNHSFSTANGLSVDRLVNGEI PYATHHLTAAALAASVPCGLSVPC SGTYSLNPCSVNLLAGQTSYFFPHVPHPSMTSQSSTSMSARAASSSTSPQAPSTLPCESLRPSLPS FTTGLSGGLSDYFTHQNQGSSSNPLIH Mouse, mmu-foxgl -polypeptide [ENSMUSP00000021333 (481 aminoacids)], SEQ ID No. 4

MLDMGDRKEVKMI PKSS FS INSLVPEAVQNDNHHASHGHHNSHHPQHHHHHHHHHHPPPPAPQPPP PPPQQQQQQPPPAPQPPQARGAPAADDDKGPQPLLLPPSTALDGAKADALGAKGE PGGGPAELAPV GPDEKEKGAGAGGEEKKGAGEGGKDGEGGKEGDKKNGKYEKPPFSYNALIMMAIRQS PEKRLTLNG I YE FIMKNFPYYRENKQGWQNS I RHNLSLNKCFVKVPRHYDDPGKGNYWMLDPSS DDVFI GGTTGK LRRRSTT SRAKLAFKRGARLT STGLT FMDRAGSLYWPMSPFLSLHHPRAS STLSYNGTT SAYPSHP MPYSSVLTQNSLGNNHS FSTANGLSVDRLVNGE I PYATHHLTAAALAASVPCGLSVPCSGTYSLNP CSVNLLAGQTSYFFPHVPHPSMT SQT ST SMSARAAS SSTS PQAPSTLPCE SLRPSLPSFTTGLSGG LSDYFTHQNQGSS SNPL I H

The mmu-vs-hsa sequences have an homology of 95.7%.

Almost all functional studies performed till now relied on drastic perturbation of the expression level of the gene in order (straight knock-out and/or constitutive overexpression). This approach is not adequate to a fine dissection of biological processes subject of investigation. The generation of the distinct neuronal and glial types forming the central nervous system (CNS) is - in fact - a very complex process, starting from primitive elements, able to self-regenerate and multipotent, and continuing via intermediate elements, provided with more limited proliferating abilities and more restricted histogenetic properties, the neuron-restricted- and glia-restricted progenitors. It is possible that Emx2 and Foxgl play in each of these neural precursors distinct functions, even opposite ones. Thereby, there is the need to perturbate their expression levels, selectively and systematically, in each of these precursors. Thereby, the present invention was not limited to the final "post-mitotic" readout of the system, but consist in continuously monitoring the dynamics of its proliferative compartments, starting just after the genetic manipulation and for sufficiently long times.

This is what the authors did, by integrated use of lentiviral somatic transgenesis,

TetON technology, state-specific promoters, combined usage of several spectrally distinct fluoroproteins, multichannel cytofluorometry, conventional multiple immunoprofiling. This study allowed the authors to clarify several obscure points listed above. Some gene manipulations used in this study may be advantageously used for purposes of gene therapy. Therefore it is an object of the instant invention a recombinant expression vector suitable for gene therapy and able to effectively express in a differentially state specific manner Emx2 and/or Foxgl polypeptides or functional fragments thereof in neural stem cells (NSCs) and/or in neuron-restricted progenitors (NPs).

Preferably, the Emx2 polypeptide is a molecule consisting of at least 95 % identity with SEQ ID No. 1 and the Foxgl polypeptide is a molecule consisting of at least 95 % identity with SEQ ID No. 3.

Still preferably, the recombinant expression vector is for use as gene therapy of brain damages.

It is a further object of the invention a neural stem cell (NSCs), or a neuron-restricted progenitor (NPs) cell, or a mixed cell population thereof being transformed with the recombinant expression vector of the invention.

Preferably, the neural stem cell (NSCs) and/or a neuron-restricted progenitor (NPs) cell or a mixed cell population thereof is for use as gene therapy of brain damages.

In the present invention, the brain damages may induced by the following events: trauma, ischemia, neurodegeneration. Therefore the neural stem cell (NSCs) and/or a neuron- restricted progenitor (NPs) cell or a mixed cell population thereof is for use as gene therapy of brain damages induced by the following events: trauma, ischemia, neurodegeneration.

It is a further object of the invention a method for the modulation of the differentiation of NSCs and/or of neuron-restricted progenitors (NPs) comprising overexpressing the Emx2 polypeptide or a functional fragment thereof in said NSCs and/or of neuron-restricted progenitors (NPs).

Preferably, the modulation of differentiation consists in a decrease of NSCs glial output and in a stimulation of survival and terminal differentiation of NPs.

It is a further object of the invention a method for the modulation of the differentiation of NSCs and/or of neuron-restricted progenitors (NPs) comprising overexpressing the Foxgl polypeptide or a functional fragment thereof in said NSCs and/or of neuron-restricted progenitors (NPs).

Preferably the modulation of differentiation consists in a decrease of NSCs glial output and in a stimulation of survival and acceleration of neuritic growth of NPs. It is a further object of the invention a method for the modulation of the differentiation of NSCs and/or of neuron-restricted progenitors (NPs) comprising overexpressing the Emx2 and Foxgl polypeptides or functional fragments thereof in said NSCs and/or of neuron- restricted progenitors (NPs).

Preferably the modulation of differentiation consists in a decrease of NSCs glial output and in a stimulation of survival and terminal differentiation of NPs.

Still preferably the methods of the invention comprise the step of transforming said NSCs and/or neuron-restricted progenitors (NPs) with a vector able to overexpress Emx2 and/or Foxgl polypeptides or functional fragments thereof.

Yet preferably the vector is the recombinant vector as defined above.

In the present invention a Emx2 polypeptide is a molecule consisting of at least 95 % of identity with SEQ ID No. 1 or allelic variant thereof. Preferably it consists of at least 99% of identity with SEQ ID No. 1 or allelic variant thereof.

A Foxgl polypeptide is a molecule consisting of at least 95 % of identity with SEQ ID No.3 or allelic variant thereof.

In the present invention, NSCs (i.e., neural stem cells) are neural precursors provided with long-term self-renewing abilities and able to give rise to the three main neural cell types, neurons, astrocytes and oligodendrocytes. NPs (i.e., neuronal progenitors) are neural precursors provided with short-term renewing abilities and able to give rise to only neurons

The expression "state specific expression" or " able to express in a differentially state specific manner" means "expression restricted to, or especially strong in defined neural precursor types".

The present invention comprises within its scope the following three genie manipulations, alone or combined, to be delivered to CNS neural precursors (including cortico-cerebral, but not limited to), through lentiviral vectors (or whatsoever other, less genotoxic, suitable technology), in order to quantitatively and qualitatively ameliorating their neuronal outputs:

(1) overexpression of Foxgl and Emx2 in neural stem cells, to down-regulate the glial (both astro- and oligodendroglial) output of these cells, while not impairing the amplitude of their neuronal progenies;

(2) overexpression of Foxgl in neuron-biased/restricted progenitors and/or early generated neurons, to enlarge the neuronal output of such progenitors and to overstimulate the maturation of their neuronal progenies, which usually follows the shutting off of the transgene;

(3) overexpression of Emx2 in neuron-biased/restricted progenitors and/or early generated neurons, to accelerate the maturation of neuronal progenies of such progenitors.

The present invention consists in the overexpression of the transcription factor genes Emx2 and Foxgl in neural stem cells (NSCs) and neuron-restricted progenitors (NPs), single or combined, aimed at increasing the neuronal output obtainable from pools of engineered cortico-cerebral precursors and reducing their astro- and oligodendrocytic outputs. This procedure is intended to be employed for purposes of cell-based therapy of brain damages of different origin, including - but not restricted to - trauma, ischemia, neurodegeneration.

The present invention more specifically relates to the following programs of artificial gene overexpression:

(1) overexpressing Foxgl or Emx2 in NSCs, to downregulate the glial output of these cells, without impairing their neuronal output;

(2) overexpressing Foxgl in neuron-biased progenitors and/or newly generated neurons, in order to increase the neuronal output of these progenitors and over-stimulate the subsequent neuronal maturation of their progenies, which usually follows down-regulation of the gene;

(3) overexpressing Emx2 in neuron-biased progenitors and/or newly generated neurons, in order to accelerate maturation of their more mature derivatives.

The present invention discloses:

(1) a description of the methodologies the authors used to discovery the histogenetic activities of Emx2 and Foxgl subject of patenting;

(2) the experimental proofs of these activities;

(3) the proofs-of-principle of the fruitful combinability of gene manipulations needed to evoke these activities.

Three are the main innovative features and points of strenght of this invention:

(1) the cell-type-specificity of gene manipulations

(2) the manipulation of genes involved in spatial patterning of the embryo

(3) the flexibility and combinability of the proposed manipulations.

Genes involved in developmental control display highly articulated expression profiles. So, in order to allow "macroscopic" morphogenetic effects of their action to emerge, it is often necessary to restrict their activation to defined spatio-temporal subdomains. For example, Foxgl overexpression in neuron-restricted progenitors seems to prepare a dramatically enhanced morphological maturation of their post-mitotic progenies. However this phenomenon is only observed provided that Foxgl has been subsequently switched off. The combined usage of state-specific promoters and conditional gene expression systems is a quite novel approach for managing such kind of phenomena in the most advantageous way.

Several patents applications in the area of neuroprotection and neuro -regeneration refer to employment of a molecule or to activation of a gene, each usually able to modulate one specific morphogenetic subroutine (apoptosis, neuronal differentiation, gial-neuronal choice, etc.). The two genes subject of the present invention, Emx2 and Foxgl, implicated in normal spatio-temporal patterning of the embryonic cerebral cortex and especially active when embryonic neuronogenesis is in progress, conversely work as "gene hubs". This means that they may co-modulate several of these subroutines, thus allowing sustained neuronogenesis levels with very low gliogenesis rates.

Finally, the gene manipulations the authors propose may be applied to a large spectrum of cellular substrates useful for brain repair purposes and may be combined according to different rationale designs.

Four are the most important aspects of our gene manipulations, for which improvements and generalizations can be predicted: (1) how the gene manipulation is delivered, (2) the cell-type-specific promoters employed, (3) the cell substrate of manipulations, (4) the possibility to combine more than one manipulation.

(1) the vector of the invention may be delivered for therapeutic purposes by various known technologies that are bio-safe such as adenoviral or adeno-associated vectors,

DNA-transposon-encoded recombinases, ZFNs, ZFRs, etc...

(2) The employment of the state specific promoters the authors selected is not mandatory, as they may be replaced by other ones, firing in the same spatio-temporal domain. Some of these are already known (e.g. the Hes5 promoter or the multimerized CBP-RE from the Hes5 promoter, potentially replacing the NSC-restricted promoters used in this study), some others will be reasonably discovered in the future. (3) Cell substrates of the present manipulations were early murine cortico-cerebral precursors, including still a large fraction of NSCs and many neuron-restricted progenitors. So the straight translation of the present procedures to homologous human neural precursors (of fetal origin) might create not trivial problems, both ethic and immunological. However there are ways to escape these problems. Patient-specific somatic cells (skin fibroblasts, fat adipocytes, etc.) may be converted into iPS cells (induced totipotent cells, extremely similar to the totipotent embryonic stem cells, ESCs, from which the embryo develops, 29, 30, 31, 32, 33) and these iPS may be in turn differentiated to cortico-cerebral precursors 23-28), manipulable as proposed in this invention and suitable for autologous transplantation. Alternatively, the gain-of- function manipulations of Emx2 and Foxgl might be delivered to neural precursors residing within the damaged brain, in order to enhance the endogenous self-repair processes triggered by tissue damage.

(4) The authors have already provided two proofs-of-principle of the possible combination of the three basic gene manipulations subject of the present invention, in order to ameliorate the neuronal output obtainable from cortico-cerebral neural cultures. In this respect, other potentially advantageous co-manipulations may be reasonably proposed. Among these, for example: co-expressing Emx2 and Foxgl in the NSC compartment, to further knock-down gliogenesis. Moreover, should it be appropriate for procedure optimization to diversify the temporal expression profile of the two genes in the selected compartments or should it result convenient to diversify their expression levels, that would be possible, by using another conditional gene expression system, driven by an inducer distinct from tetracyclin (e.g. PIP-ON or E-ON) in parallel with the TetON one. The invention will be now described by means of exemplificative not limiting examples by making reference to the following figures:

Figure 1. Molecular tools for manipulating distinct cortico-cerebral neural precursor types and assessing sizes of the compartments they belong to, by multi-channel fluoro cytometry. Experimental designs followed to study the neuronogenic (EXP la) and the gliogenic (EXP lb) lineages.

Figure 2. (A,B) Frequencies of distinct neural types derived from dissociated and acutely lentivirus-transduced El 1.5 cortico-cerebral precursors, at different in vitro development times (days post infection, dpi), upon pNes-driven, doxycyclin-dependent transgene expression. (C) Net frequencies of neural stem cells (NSCs) and glial elements, as assessed by integrating primary EXP la and EXP lb results. Per each gene, results from at least 3 biological replicates are averaged and s.e.m.'s are indicated. Significance of differences between results is assessed by ANCOVA. NSC, eGPs, eNPs, e(NSCb)Ns, e(eNPb)Ns, INPs, IGPs, AP, OP, GCs, e(INPb)Ns, Ins acronyms as are defined in Fig. 1.

Figure 3. (E,F) Compartment-to-compartment transition indices along the neuronogenic (E) and the gliogenic (F) lineages, following EXP la and EXP lb manipulations, respectively. These indices are calculated at each in vitro developmental time, as ratios between the sizes of the "downstream" and the "upstream" compartments. Absolute indices are provided for control (luc-treated) samples, luc-normalized ones are conversely shown for Foxgl- and Emx2-gain-of-function samples. Per each gene, results from at least 3 biological replicates are averaged and s.e.m.'s are indicated. Significance of differences between results is assessed by ANCOVA. (G) Time course of the indicated compartment death rates as evaluated by 7AAD staining. NSC, eGPs, eNPs, e(NSCb)Ns, e(eNPb)Ns, INPs, IGPs, AP, OP, GCs, e(INPb)Ns, Ins acronyms as are defined in Fig. 1.

Figure 4. Late histogenesis upon pNes-driven, doxycyclin-dependent transgene expression. Percentages of cells derived from El 1.5 cortico-cerebral precursors showing neuronal (b-tub+), astrocytic (S100b+) and oligodendrocytic (CNPase+) differentiation, following primary growth under growth factors (GFs) and doxycyclin for 14 days and secondary growth under 5% serum for 7 days. For each transgene, results from at least 3 biological replicates are averaged and s.e.m.'s are indicated. Significance of differences between results is assessed by ANOVA.

Figure 5. Molecular tools for manipulating cortico-cerebral neuronal progenitors and assessing sizes of distinctive compartments of the neuronal lineage, by multi-channel fluorocytometry (EXP2).

Figure 6. (A) Frequencies of distinct neural types derived from dissociated and acutely lentivirus-transduced El 1.5 cortico-cerebral precursors, at different in vitro development times (days post infection, dpi), upon pTccl -driven, doxycyclin-dependent transgene expression. (B) Compartment-to-compartment transition indices along the neuronogenic lineage. These indices are calculated at each in vitro developmental time, as ratios between the sizes of the "downstream" and the "upstream" compartment. (C) death rates within the neural-progenitor compartment, as assessed by 7AAD exclusion. For (B) and (C) absolute indices are provided for control (luc-treated) samples, luc-normalized ones are conversely shown for Foxgl- and Emx2-gain-of- function samples. Per each gene, results from at least 3 biological replicates are averaged and s.e.m.'s are indicated. Significance of differences between results is assessed by ANCOVA.

Figure 7. Late neuronogenesis upon pTccl -driven, doxycyclin-dependent transgene expression. Percentages of cells derived from El 1.5 cortico-cerebral precursors showing neuronal (b-tub+) differentiation, following primary growth under growth factors (GFs) and doxycyclin for 14 days and secondary growth under 5% serum for 7 days. Per each transgene, results from at least 3 biological replicates are averaged and s.e.m.'s are indicated. Significance of differences between results is assessed by ANOVA.

Figure 8. Time course average neurite length in progenies of El 1.5 cortico-cerebral precursors acutely transduced with Foxgl or luciferase, following primary growth under growth factors (GFs) and doxycyclin for 14 days and secondary growth under 5% serum for 7-14 days. Per each transgene, results from at least 2 biological replicates are averaged and s.e.m.'s are indicated.

Figure 9. Late histogenesis upon combined, doxycyclin-dependent Foxgl /Emx2 transgenesis. (Upper panels, EXP3) Shown are neuron frequencies observed among progenies of E12.5 cortico-cerebral precursors, upon chronic combined overexpression of Foxgl or luc in the pNes+ and the pTccl+ compartments for 28 days under growth factors, followed by 14 days under 5% serum. Compared with the simple pTccl -driven, the pNes/pTal -driven overexpression of Foxgl increases the neuronal output from 3.5±0.1% to 4.3±0.2%, while the pNes/pTccl -driven overexpression of luc conversely reduces such output from 2.6±0.4% to 1.8±0.5%. (Lower panels, EXP4) Shown are neuron frequencies observed among progenies of E12.5 cortico-cerebral precursors, upon pTccl-driven overexpression of Foxgl for 21 days in vitro, under growth factors, and pTccl-driven overexpression of Emx2 or luc for 4 more days, still under growth factors, all followed by 7 days under 5% serum. Secondary over-activation of Emx2 increases the neuronal frequency from 5.4±1.4 to 8.0±1.6%. (In general) Per each gene manipulation, results from at least 3 biological replicates are averaged and s.e.m.'s are indicated. Significance of differences between results is assessed by ANOVA.

Material and methods Animal handling

Wild type mice (strains CD1 and FVB/N, purchased from Harlan-Italy), pNes-rtTA- IRESβgeo^+/~ mutants [1] (obtained from EMMA, Monterotondo, Italy) used in this study were maintained at the SISSA-CBM mouse facility. Embryos were staged by timed breeding and vaginal plug inspection.

Embryo generation

pNes-rtTA-IRESβgeo^+/' FVB/N males, harboring an "rtTA-IRES- geo" cassette under the control of a synthetic "rat-nestin-intron II-enhancer/hsv-tk minimal promoter" module [1], were mated to wild type FVB/N (EXPla, EXPlb, EXP2) or CD1 (EXP3 and EXP4) females. El 1.5/12.5 heterozygous transgenic embryos (which express β-galactosidase in the neural tube) were distinguished from their wild type littermates by X-gal staining [in: IX PBS, 5mM K₃Fe(CN)₆, 5mM K₃Fe(CN)₆*3 H₂0, 2mM MgCl₂, 0.01% Na deoxycholate, 0.02% NP40; lmg/ml 5-Bromo-4-Cloro-3-Indolilb-D-Galactopiranoside (X-Gal)]. pNes-rtTA-IRESfigeo^+/- embryos were used for EXPla, EXPlb, wild type embryos for EXP2 and EXP3, embryos with both genotypes for EXP4.

Primary cortical precursors (cPCs) isolation

cPCs (cortical precursors) indicate proliferating cortico-cerebral precursors, neural stem cells (NSCs) and lineage-committed progenitors (NPs, GPs, APs, OPs), independently of their differentiation state, early (e) or late (1); they do not include post-mitotic elements, neurons (Ns) and glial cells, astrocytes (As) and oligodendrocytes (Os). cPCs were isolated from El 1.5 (EXPla, EXPlb, EXP2) or E12.5 (EXP3 and EXP4) embryonic cortices and plated onto uncoated 24 multiwell (BD Falcon) after gentle mechanical dissociation to single cells. 2.5* 10⁵ cPCs were plated for each well in 350 μΐ of serum free anti-differentiative medium [1 : 1 DMEM-F12, IX Glutamax (Gibco), IX N2 supplement (Invitrogen), 1 mg/ml BSA, 0.6%> w/v glucose, 2 μg/ml heparin (Stemcell technologies), 20 ng/ml bFGF (Invitrogen), 20 ng/ml EGF (Invitrogen), IX Pen/Strept (Gibco), 10 pg/ml fungizone (Gibco)] ; 2μg/μl doxycycline was added in GOF experiments (Clontech). Long term cPCs culture maintenance

For long term maintenance, fresh GFs (20 ng/ml bFGF and 20 ng/ml EGF) plus 2μg/μl doxycycline were added 2 days after each cell dissociation. 1.5 more days later, cPCs were harvested, dissociated to single cells by trypsin (Gibco), followed by addition of DNasel (Sigma) and soybean trypsin inhibitor (Sigma) [5]. cPCs were then replated at 10⁵ cells in 400 μΐ of anti-differentiative medium/well. These culture conditions lead to free floating neurospheres formation [6] and effective cPCs propagation. cPCs differentiation

For differentiation experiments neurospheres were harvested after 14 days or 28 days propagation in serum-free, anti-differentiative culture conditions and dissociated to single cells. cPCs were plated onto 200μg/ml poly-L-lysine-coated 24-wells, at 5* 10⁵ cells/well in 500μ1 of doxycycline- free differentiative medium [1 : 1 DMEM-F12, IX Glutamax (Gibco), IX N2 supplement (Invitrogen), IX B27 supplement (Invitrogen), 1 mg/ml BSA, 0.6% w/v glucose, 2 μg/ml heparin (Stemcell technologies), ImM N-acetylcysteine (Sigma), 5% tetracycline- free serum (Clontech), IX Pen/Strept (Gibco), 10 pg/ml fungizone (Gibco)]. Differentiative medium was replaced every 3-4 days with fresh medium and cPCs were allowed to differentiate for 7 or 14 days. Lentiviral transfer vector construction

Basic DNA manipulations (extraction, purification, ligation) as well as bacterial cultures and transformation, media and buffer preparations were performed according to standard methods. DNAs were transformed in the E.Coli Xll-blue strain.

All the plasmids used in the present study were built up starting from the pCCL-SIN- 18PPT.Pgk.EGFP-Wpre lentiviral transfer vector backbone [7] (gift from L. Naldini), by replacing the pPgk-EGFP cassette with specific modules:

(A) transgene expressing vectors (tool-sets 1 and 6) were obtained by replacing the pPgk- EGFP cassette with the TREt promoter (from the pTRE-Tight plasmid, Clontech) and a bicistronic cassette alternatively containing Foxgl (GenelD: 15228), Emx2 (GenelD: 13797) or Luciferase (from the pGEM®-/wc plasmid, Promega) coding sequences, plus the IRES2EGFP module (from the pIRES2-EGFP plasmid, Clontech);

(B) pSGC10/S2/2N-mCerulean vector (tool-set 3) was obtained by replacing the pPgk- EGFP cassette with the pSGC10/S2/2N promoter, taken from the pSGC10/S2/2N- luciferase plasmid [8] (gift from H.Kiyama) and the mCerulean coding sequence [9] (from the Cerulean plasmid, Addgene);

(C) pTal-mCherry (tool-set 2) vector was obtained by replacing the pPgk-EGFP cassette with the pTal-mCherry cassette (gift from E.S.Ruthazer). Starting from the pTal-mCherry vector, pTal-rtTA/M2 (tool-set 6) was obtained by replacing the mCherry sequence with the rtTA/M2 cassette (from the pTet-ON Avanced plasmid, Clontech);

(D) to obtain the pNes/hsp68-mCherry vector (tool set 5), the neural enhancer from the second intron of the rat nestin gene was taken from the pE/Nestin-EGFP plasmid [10] (gift from H.Okano) and cloned upstream of the minimal hsp68 murine promoter, in the pHPS.sis plasmid [11]. The resulting pNes/hsp68 promoter was then used to replace the pTal promoter in the pTal -mCherry vector. The pNes/hsp68-rtTA/M2, used for expression analysis, was obtained starting from the pNes/hsp68-mCherry vector and replacing the mCherry sequence with the rtTA/M2 cassette. Lentiviral vectors packaging and titration

Third generation self-inactivating (SIN) lentiviral vectors were produced as previously described [7] with some modifications. Briefly, 293T cells were co-lipofected (Lipofectamine 2000, Invitrogen) with the transfer vector plasmid plus three auxiliary plasmids (pMD2 VSV.G; pMDLg/pRRE; pRSV-REV). The conditioned medium was collected after 24 and 48hs, filtered and ultracentrifuged at 50000 RCF on a fixed angle rotor (JA 25.50 Beckmann Coulter) for 150 min at 4°C. Viral pellets were resuspended in PBS without BSA (Gibco).

EGFP-expressing lentiviral vectors (tool-sets 1 , 6) were titrated on HeLa TET-off cells (Clontech), by end point fluorescence titration, as previously described [7] and titer expressed as transducing units per ml (TU/ml). Other LTVs were generally titrated by Real Time quantitative PCR after infection of HEK293T cells, as previously reported [12]. One end point fluorescence-titrated LTV was included in each PCR titration session and PCR- titers were converted into fluorescence-equivalent titers throughout the study. Lentiviral infection of cPCs

Gain of function experiments. At least 7.5 * 10⁵ cPCs were dissociated to single cells and plated onto a 24 multiwell plate, at 2.5 * 10⁵ ΰε1ΐ8/350μ1, in anti-differentiative doxycycline- containing medium, and infected with the appropriate LTV mix. Lentiviral vector conditioned medium was washed away 3.5 dpi and neurospheres were dissociated and plated as already described. The following lentiviral vector mixes were used for the different experimental setups:

(EXPla) LTV pTal-mCherry, at MOI 10; LTV pSGC10/S2/2N-mCerulean, at MOI 10; alternatively, LTV Tre_t-Luc-IRE S2EGFP , Tre_t- oxgi-IRES2EGFP or Tre_t-Emx2- IRES2EGFP, at MOI 15.

(EXP lb) alternatively, LTV Tre_t-Luc-IRE S2EGFP , Tre_t- oxgi-IRES2EGFP or Tre_t- Emx2-IRES2EGFP, at MOI 15.

(EXP2) pTal-rtTA/M2, at MOI 8; LTV pNes/hsp68-mCherry, at MOI 8; LTV pSGC10/S2/2N mCerulean, at MOI 8; alternatively, LTV Tre_t-Luc-IRE S2EGFP , Tre_t- Foxgl -IRES2EGFP or Tre_t-Emx2-IRES2EGFP, at MOI 8.

(EXP3) LTV pTal-rtTA/M2, at MOI 10; alternatively, LTV Tre_t-Luc-IRES2EGFP or Tre_t- oxgi-IRES2EGFP, at MOI 10.

(EXP4) primary infection: LTV pTal-rtTA/M2, at MOI 10; LTV Tre_t- oxgi- IRES2EGFP, at MOI 10. secondary infection: alternatively, LTV Tre_t-Luc-IRES2EGFP or Tre_t-Emx2-IRES2EGFP, at MOI 10. In this case, the two LTVs were delivered to the two halves of the same primarily infected EXP4 preparation.

FACS analysis of cPCs

Every 7 days, an aliquot of 100,000-200,000 dissociated cPCs, prepared as described above, was analyzed on a three lasers-equipped Cyan ADP flow cytometer (Dakocytomation, Denmark). Multivariate data analysis was performed by Flowjo™ software (Tree Star, Ashland, OR).

Forward scatter (FSC) and side scatter (SSC) were used to gate nucleated cells and to exclude debris and cell aggregates (live gate) in every analysis. Cells belonging to the live gate were then further evaluated for the expression of the fluorochromes in order.

In case of EXPla and EXP2, cells were categorized on the basis of their EGFP, mCherry and mCerulean fluorescence profiles. To assess their viability, an aliquot of them was further labelled with a 7-Amino-actinomycin D (7-AAD), which is excluded by viable cells, but can penetrate cell membranes of dying or dead cells and intercalate into double- stranded nucleic acids.

In the case of EXP lb, cells were previously decorated with anti-A2B5-APC (Miltenyi Biotec, Germany) and anti-PS A-NCAM-PE (Miltenyi Biotec, Germany) antibodies, according to manufacturer's instructions, and subsequently categorized, based on their EGFP (EXP lb) and APC/PE fluorescence profiles.

Immunocytofluorescence

For immunocytofluorescence on differentiated cPCs, cells were fixed directly on poly-L- lysine coated 24 multiwell plates, with 4% PFA for 10 min at RT. Wells where then treated as described above.

The following primary antibodies were used: anti-P-tubulin mouse monoclonal, (clone Tujl, Covance, 1 :600), anti-CNPase mouse monoclonal (clone 11-5B, Millipore, 1 :400), anti-SlOO rabbit polyclonal (DAKO 1 : 1000) anti-EGFP chicken polyclonal (AbCam, 1 :800), anti-RFP/DsRed rabbit polyclonal (antibodies-online GmbH 1 : 1000), anti-Tbr2 rabbit polyclonal (AbCam, 1 :600), anti Foxgl rabbit polyclonal (gift from G. Corte, 1 :200), anti-Emx2 mouse monoclonal (clone 4F7, Abnova, 1 :200).

Secondary antibodies were conjugates of Alexa Fluor 488, Alexa Fluor 594, Alexa Fluor 546, Alexa Fluor 633 (Invitrogen, 1 :600). DAPI (4', 6'- diamidino-2-phenylindole) was used as nuclear counter staining.

Images acquisitions and quantifications

Differentiating cPCs were photographed on a Nikon Eclipse TS100 fluorescence microscope equipped with a DS-2MBWC digital microscope camera with a 20X objective. At least 8 fields from 3 independent biological replicates were counted for each experimental condition. Images were imported and counted with Adobe Photoshop CS3 software™. DAPI stained nuclei images were imported and counted with MacBiophotonics Image J nucleus counter plug-in (software available at http ://www.macbiophotonics. ca/) .

Analysis of mean neurite length was performed with the NeuriteTracer Image J plugin [13]. 7 to 15 fields were evaluated for each biological replicate.

Statistical analysis

FACS data: compartment sizing. Frequencies of cells with different fluorescence-profiles, evaluated on samples of at least 100,000 elements, were averaged over three biological replicates and plotted against time. Statistical significance of differences was evaluated by one-way ANCOVA. FACS data: X-to-Yi transition indexes. This index was calculated for each gene manipulation and each culturing time point, by dividing the size of the "downstream" Yi compartment by the size of the "upstream" X compartment, luc data were averaged over three independent biological replicates and plotted against time. For each biological replicate and each time, indexes of Emx2 -GOF and Foxgl -GOF cultures were normalized against /wc-transduced cultures. The resulting normalized indexes were averaged over the three biological replicates and plotted against time. Statistical significance of differences was evaluated by one-way ANCOVA.

FACS data: cell death rates and enlargement rates. The first parameter was calculated as the percentage of cells with an NSC- or an NP-specific fluorescent profile further stained by 7AAD. The second parameter was evaluated by dividing the size of compartment in order at time = t by the size of the same compartment one week earlier, luc, Emx2 and Foxgl data, collected over three independent biological replicates, were further processed like X-to-Yi transition indexes.

Immunocytofluorescence data. Frequencies of cells expressing specific antigens, evaluated on samples of at least 10,000 elements, were averaged over three biological replicates and plotted against the gene manipulation in order. Statistical significance of differences was evaluated by one-way ANOVA. Results

Assessing the roles of Emx2 and Foxgl in the NSC and NP compartments: the experimental design

To investigate consequences of gene manipulations in NSCs, the authors developed two complementary experimental designs: EXPla, to monitor the neuronogenic lineage (Fig.l); EXPlb, to monitor the glial lineage (Fig.l). [Selection/ validation of cell-type specific promoters used in this study, set-up of protocols for lentivector-mediated transgene delivery, doxycycline-controlled TF overexpression and immunoprofiling of endogenous Emx2 and Foxgl in cortical precursors are reported in Brancaccio M et al. (2010), 34.

In EXPla, the authors infected El 1.5 cortical neural precursors from pNes-rtTA- IRESbgeo ^7" mutants with a common mix containing LTVs pTal-mCherry and pSGC10/S2/2N-mCerulean plus, alternatively LTV TREt-Emx2-IRES2EGFP or TREt- Foxgl -IRES2EGFP. LTV TREt-luc-IRES2EGFP, encoding for luciferase (luc), was used as a control. Under these conditions, the TF in order is mainly overexpressed in NSCs, but also - to a minor extent - in their immediate direct progenies. Similarly, the fluoroproteins EGFP and mCherry label NSCs and NPs, respectively, but also their early descendants, which may be so distinguished from more mature elements thanks to their peculiar fluorescence patterns. In such a way, several compartments of the neuronogenic lineage, downstream of NSCs and up to late neurons (INs), may be distinguished and followed up in living cultures (Fig.lA,C). As for the glial lineage, it has been previously reported that a large fraction of glial progenies obtainable in vitro from embryonic cortico-cerebral precursors derives from bipotent glial progenitors (GPs), generating astrocytes and oligodendrocytes but not neurons. This population is distinguishable thanks to its A2B5 PSANCAM^" immunological profile [15]. In EXP lb, the authors overexpressed again TREt-Emx2-IRES2EGFP or TREt- oxgi-IRES2EGFP within NSCs and used antibodies conjugated with different fluorophores to label the bipotent GP population. So, similarly to EXP la, they were able to follow early transitions from NSCs, through GPs, up to more mature glial elements (Fig.l).

In EXP2, the authors investigated the role played by Emx2 and Foxgl along the neuronogenic lineage, by overexpressing them in NPs and early neurons (eNs) of the pTal⁺ compartment (Fig.5). They infected El 1.5 cortical neural precursors from wild type mice with a common mix containing the LTVs pTal-rtTA/M2, pNes/hsp68-mCherry, pSGC10/S2/2N-mCerulean and, alternatively, TREt-Emx2-IRES2EGFP or TREt- oxgi- IRES2EGFP. LTV TREt-luc-IRES2EGFP was used as a control. Under these conditions, the TFs and EGFP are overexpressed in NPs and their early neuronal progenies, mCherry labels NSCs and their immediate derivatives, including eNPs, mCerulean labels more mature neurons. In this way, we can distinguish early(NSC-born)-NPs from late NPs, as well as early NP-born neurons from more mature ones.

Developing molecular tools for dissecting Emx2 and Foxgl regulatory functions in cortico- cerebral precursors.

In order to get an enduring overexpression of Emx2 or Foxgl and a stable fluorescent labeling of neural precursors, the authors relied on lentiviral vectors. As already described [14], in fact, this delivery system allowed us to stably trasduce almost the totality of cortical precursors (>97%, not shown).

To specifically label distinct neural compartments with different fluoroproteins, the authors selected four promoters, firing in selected neural types: pNes/hsvtk, pNes/hsp68, pTal and pSGC10/S2/2N. The first two share the neural enhancer from the II intron of the rat nestin gene, alternatively conjugated to the minimal promoters of the murine hsp68 heat shock gene [15] or the herpes virus simplex thymidine kinase gene [16], and mainly fire in NSCs [10; 16; 17]. pTal, promoter of the rat ccl tubulin [18], is specifically active in neuron-restricted progenitors and in their young progenies [17; 19; 20]. pSCG10/S2/2N includes the 2kb rat stathmin promoter plus two neuronal restricted silencer elements (NRSEs) and specifically fires in post-mitotic neurons [8].

The authors tested the specificity of these promoters, by comparing their firing patterns with the expression domains of established markers of the neuronogenic lineage. They electroporated a pTal-EGFP plasmid into the El 4.5 murine cortex and scored the EGFP distribution 12 hours later. As expected [19], the fluoroprotein was mainly restricted to a subset of ventricular zone (VZ) precursors, colocalizing to some extent with the basal progenitor marker Tbr2 and the neuronal marker β-tubulin (not shown). Colocalization of pTal-driven EGFP with Tbr2 and β-tubulin was quantified on E12.5 cortical precursors, transduced with LTVs and kept under anti-differentiative medium for 3.5 days: among EGFP⁺ cells, about 4% coexpressed Tbr2, 30% β-tubulin, 10% both of them (not shown). Significant pTal -driven mCherry/p-tubulin colocalization (about 17%) could still be found upon prolonged in vitro culture of cortical precursors, followed by 7 days differentiation under 5%> serum (not shown). Finally, the authors scored El 2.5 cortical precursors infected with LTV pSGC10/S2/2N-mCerulean and kept 3.5 days under anti-differentiative medium. Upon this treatment, 97%> of mCerulean⁺ cells expressed β-tubulin, whereas 80%> of β- tubulin⁺ cells were mCerulean⁺, suggesting that pSGC10/S2/2N firing is limited to more mature post-mitotic neurons. Finally, the authors assessed the suitability of the four promoters, pNes/hsvtk, pNes/hsp68, pTal and pSGC10/S2/2N, for multichannel fluorescence profiling. They infected El 1.5 cortical precursors with two lentiviral mixes, each driving the expression of three different fluorescent protein genes in pNes-, and pSGC10/S2/2N-firing domains, and profiled the resulting neurospheres after 7 days of anti-differentiative in vitro culture (not shown).

To achieve reversible overexpression, the authors selected the TetON technology [21; 22]. They put the rtTA gene under the control of the state-specific promoter and the TF gene under the "Tetracycline-Response Element, tight" (TREt) (Fig. S4). As for rtTA expression in NSC cells, it was achieved by performing the experiments in neural precursors from pNes-rtTA-IRESbgeo^+/" mutants [1], harboring the rtTA cassette under the pNes/hsvtk promoter. A pTal-rtTA/M2 expressing lentivector was conversely used to engineer the neuronal progenitor (NP) compartment. The TF-overexpressing compartment was generally labelled by an IRES2EGFP module, placed downstream of the TREt-TF cassette.

Detailed results

The authors performed all their gene manipulations in cortico-cerebral precursors from El 1.5 mouse embryos. In the first experiment set (Fig.l, EXP la), they over- expressed Foxgl, Emx2 or the negative control gene encoding for luciferase (luc) in the neural stem cell (NSC) compartment, by lentiviral vectors and conditional technology TetON, driving the transactivator rtTA expression by the NSC-specific promoter pNes (harboring neuro-specific regulatory sequences form the nestin locus) and keeping the so- engineered cells under doxycyclin. The authors also differentially labelled the distinct neuronogenic compartments of the cell preparation, by putting the EGFP gene under the (indirect) control of pNes (firing in NSCs and their immediate progenies), the mCherry gene under the alpha 1 -tubulin promoter (pTal, firing in neuron-restricted progenitors and their early neuronal progenies), the mCerulean gene under the direct control of a modified version of the stathmin promoter (pSGC10/S2/2N, active in neurons) (Fig. 1, toolsets 1-3). In such a way the authors made it possible to dynamically and continuously monitor the size of the various primary and transitional compartments of the neurogenic lineage, by regular cytofluorimetric assays. Moreover (Fig. l, Exp lb), the authors selected two specific antibodies, raised again the antigens A2B5 and PSA-NCAM, which allow them to distinguish glia-restricted, bipotent A2B5(+)/ PSANCAM(-) progenitors, directly originating from NSCs (Fig. 1, toolset 4) (anti-PSA-NCAM mouse monoclonal antibody, PE conjugated, MiltenyiBiotec-MACS #130-093-274; anti-A2B5 mouse monoclonal antibody, APC conjugated, MiltenyiBiotec-MACS #130-093-582). In such a way, via integrated employment of the toolsets 1 and 4 and similar to the neuronogenic lineage, the authors made it possible to dynamically evaluate the sizes of distinct "high" compartments of the gliogenic lineage, by regular cytofluorimetric assays.

The authors transferred the so-engineered neural cultures under anti-differentiative Sato medium, containing growth factors (GFs) which promote the stem state, to grow as floating neurospheres and profiled them by cytofluorometry every 8^th day, until 28 days post-infection (28 dpi). At least 100,000 cells of each genotype were analyzed in each session. The entire experiment was repeated three times, in distinct experimental sessions and making use of physically different lentiviral preparations. Data were collected, averaged and graphically represented. Significance of observed differences was systematically evaluated by ANCOVA.

Upon Foxgl overexpression, the authors observed a conspicuous enlargement of the NSC compartment and, upon overexpression of either Foxgl or Emx2, an increase of neuron-restricted progenitors, both early and late (eNPs and INPs, respectively). However, this last effect, even if statistically significant under both genes, was much more pronounced under Foxgl . As for the glial lineage, the authors conversely found an increase of early glia-restricted progenitors (eGPs) and a conspicuous decrease of the late glia- restricted progenitors (lGPs): both these effects were far more pronounced under Foxgl than under Emx2. Finally, the sizes of the NSC compartment, evaluated by comparing data from EXPs la and lb, resulted strongly upregulated upon Foxgl and - starting from the 2^nd week - reduced under Emx2. (Fig. 2).

Subsequently, for each assay time and each gene manipulation, the authors compared the size of the manipulated compartment with those of all the adjacent ones downstream to it. Per each pair, the authors calculated the ratio among the latter and the former, assuming it as an index of the frequency at which neural elements belonging to the upstream compartment move toward the downstream one. The authors finally normalized such "transition indexes" with respect to the luciferase control, averaged the data, plotted them against the culturing times and calculated the significance of observed differences by ANCOVA.

As for the neuronal lineage, the authors found that the overactivation of Emx2 in NSCs leads these cells (and their eNP progenies) to differentiate, whereas Foxgl elicits opposite effects. As for the glial lineage, Foxgl and Emx2 promote and inhibit, respectively, glial commitment of NSCs, but both genes strongly inhibit further differentiation of eGPs to lGPs (Fig. 3). Given the amplitude and the consistency of this last effect, it seemed to the authors extremely interesting, as exploitable in principle to reduce the glial output of NSCs.

It is well known that the GF-containing Sato medium is permissive for stem and progenitor cells, conversely conter-selecting more differentiated neural elements, including neurons. As such, it is well suitable for studying population dynamics of poorly differentiated neural progenitors, but not good for characterization of the final post-mitotic output of neural cultures. So, in order to study late effects of gene manipulations delivered to NSCs, the authors engineered cultures with toolset 1 (Fig. l), let them grow for 14 days under GFs and in the presence of the trans gene inducer doxycyclin, then transferred them under 5% serum (an established promoter of neuronal differentiation, allowing neuronal survival) and kept them in these conditions for more 7 days, in the absence of doxycyclin. Upon fixation by 4% paraformaldehyde, the authors immunoprofiled these cells, looking at expression of neuronal (b-tubulin), astrocytic (SI 00) and oligodendrocyte (CNPase) markers. The experiment was repeated three times, in distinct sessions and using physically different lentiviral preparations. At least 4,000 cells per genotype by session were scored. Data were averaged and plotted against genotypes. Finally, significance of results was evaluated by ANOVA.

It resulted that Emx2 or Foxgl overexpression in NSCs apparently reprograms these cells, reducing their astrocytic output by about one half and the frequency of their oligodendrocytic progenies by at least one third (with p at least less than 0.05). The neuronal output is conversely not impaired (Fig. 4). Further neuronal differentiation assays, run by prolonging the transgene over-expression phase under GFs up to 28 days and the differentiation-permissive phase under serum up to 14 days, gave similar results.

In a second experiment set (Fig.5, EXP2), the authors over-expressed Foxgl, Emx2 or the negative control gene encoding for luciferase (luc), in the NP compartment. The authors employed again lentiviral vectors and the TetON technology, driving the rtTA transactivator by the pTal promoter and keeping the neural culture under doxycyclin. The authors also differentially "stained" the distinct compartments of the neuronal lineage, by putting the EGFP gene under the indirect control by pTal (firing in NPs and early NP-born neurons), the mCherry under the pNes promoter (firing in NSCs and their immediate progenies) and the mCerulean under the pSGC10/S2/2N promoter (firing in neurons) (Fig. 5, toolsets 3,5,6). In such a way the authors allowed the subsequent, cytofluorimetric size evaluation of the different compartments of the neuronal lineage, upon engineering of neuron-restricted progenitors.

The authors let engineered neural precursors grow in GFs-containing Sato medium, as floating neurosphere, and profiled them by cytofluorometry every 8^th day, up to 28 dpi. At least 100,000 cells per genotype by session were scored. The experiment was repeated three times, in distinct sessions and using different lentiviral stocks. Data were collected, averaged and graphically represented. Significance of observed differences was systematically assessed by ANCOVA. Upon Foxgl overexpression, the authors observed a relevant enlargement of the NP compartment. By contrast, following Emx2 overexpression, a shrinkage of the NP compartment was observed. In the former case a further pronounced shrinkage of stem, neuronal an glia fractions is detectable as well, possibly as a consequence of the dramatic absolute enlargement of the NP compartment (Fig. 6A).

Then, as in Expl, per each gene manipulation and each cytofluorimetric session, the authors compared the size of the manipulated compartment with those of all the adjacent ones downstream to it. Per each pair, the authors calculated the ratio among the latter and the former, assuming it as an index of the frequency at which neural elements belonging to the upstream compartment move toward the downstream one. The authors finally normalized such "transition indexes" with respect to the luciferase control, averaged the data, plotted them against the culturing times and calculated the significance of observed differences by ANCOVA. Results were as follows.

Both Foxgl and Emx2 stimulate maturation of NSC-born NPs (eNP-to-lNP transition) and promote NP survival. Moreover, Foxgl promotes NP self-renewal, so dramatically delaying their neuronal differentiation. Emx2 conversely gives rise to a diametrically opposite effect. Finally, as assessed by 7AAD exclusion assay, both genes consistently and specifically promote NPs' survival. So, both manipulations seem promising for therapeutic purposes, as potentially able to increase the neuronal output of the culture: Foxgl overexpression via the pre-amplification of the NP pool, Emx2 overexpression by accelerating neuronal differentiation itself, overexpression of both genes by promoting NPs' survival (Fig. 6).

As already done with NSC-manipulated cultures, to assay the final histogenic output of our manipulations, the authors then treated neural precursors with toolset 6 (Fig.5), let them grow 14 days under GFs and doxycyclin (so keeping the transgene on in the NP/early neuronal compartment), and transferred them under 5% serum (promoting neuronal differentiation and survival), keeping them in these conditions for 7 days, in the absence of doxycyclin. Finally, upon fixation, engineered cultures were immunoprofiled for the neuronal marker b-tubulin. The experiment was repeated three times, in different sessions and using different preparations of lentiviruses. At least 4,000 cells per genotype* session were scored. Data were averaged and graphically represented. Significance of observed differences was assessed by ANOVA. Results were as follows. Overexpression of Emx2 or Foxgl in NPs (and early neurons) increases the neuronal output by +25% and +300%, respectively (Fig. 7). Moreover, Foxgl overexpression also induces a dramatic increase of total length and complexity of the neuritic tree. This is evident from visual inspection of neural cultures as well as from their systematic analysis by the dedicated NeuriteTracer software. Further neuronal differentiation tests, performed by doubling the duration of serum-promoted differentiation, showed that this was probably due to an acceleration of neurite growth, rather than to an anticipation of it (Fig. 8).

The analyses described above allowed the authors to discover three simple gene manipulations, useful to enhance the relative and/or absolute neuronal output of cortico- cerebral precursors cultures:

(1) Emx2 or Foxgl overexpression in NSCs, to decrease their glial output (especially the astrocytic one) and - mainly in the case of Foxgl - to ameliorate their self-renewing abilities.

(2) Foxgl overexpression in neuron-biased progenitors, to promote their survival, their expansion prior to differentiate as well as to induce an acceleration of neuritic growth in their neuronal progenies.

(3) Emx2 overexpression in neuron-biased progenitors, to stimulate their survival and their terminal differentiation.

The authors assessed if would it be possible to get supplementary advantages from these gene manipulations, by combining them according to rational designs. To provide a proof-of-principle of the feasibility of this approach, the authors assayed two of such combined manipulation, with positive results.

First, the authors simply overexpressed Foxgl in both stem cells and neuronal progenitors for 28 days, keeping the engineered neural precursors under GFs. Then, the authors switched the gene off (by washing out doxy eye lin) and transferred cells under 5% serum for additional 14 days. The authors expected that the overactivation of the gene in the NSC compartment should enhance the generation of eNPs (destined to undergo a further expansion under Foxgl) and reduce their glial output, so increasing for two reasons the prospective relative neuronal output of the culture.

The authors measured such output following combined Foxgl overexpression in the pNes⁺ and the pTal⁺ compartments (cumulatively corresponding to NSCs, NPs and newborn neurons, eNs) and found that it was 20%> higher, as compared to that obtained overexpressing the gene in pTal elements only (p<0.05, N=3). This effect, despite of its moderate amplitude, is remarkable, as it is detectable notwithstanding the intrinsic hyper- toxicity of the very high lentiviral title employed in the composite manipulation. In fact, if the control gene luc is expressed in the pNes /pTal⁺ compartment in place of Foxgl, then the neuronal output diminishes by 20% (again p<0.05, N=3) (Fig. 9). This suggests that, if the transgene delivery would be performed by a less toxic system, advantages of the double manipulation might result far more robust than reported above.

Second, the authors overexpressed Foxgl for 25 days within the NP/eN compartment and, during the last 4 days, the authors overexpressed in this compartment Emx2 as well, keeping engineered cells all the time under GFs. Then the authors switched the two transgenes off, by washing doxyxcyclin out, and transferred infected cells under 5% serum. The authors' intention was to first, induce a preliminary expansion of the NP compartment, then, force its components to differentiate en masse as neurons. As expected, even in this case, the double manipulation was successful. The neuronal output increased by about one third as compared to Foxgl-only gain-of- function precursors and the neuritic morphology remained very rich (Fig. 9).

Bibliografic references 1. Mitsuhashi T, Aoki Y, Eksioglu YZ et al. Overexpression of p27Kipl lengthens the Gl phase in a mouse model that targets inducible gene expression to central nervous system progenitor cells. Proc Natl Acad Sci U S A 2001;98:6435-6440.

2. Pellegrini M, Mansouri A, Simeone A et al. Dentate gyrus formation requires Emx2. Development 1996;122:3893-3898.

3. Hebert JM, McConnell SK. Targeting of ere to the Foxgl (BF-1) locus mediates loxP recombination in the telencephalon and other developing head structures. Dev Biol 2000;222:296-306.

4. Muzio L, Mallamaci A. Foxgl confines Cajal-Retzius neuronogenesis and hippocampal morphogenesis to the dorsomedial pallium. J Neurosci 2005;25:4435-4441.

5. Rietze RL, Reynolds BA. Neural stem cell isolation and characterization. Methods Enzymol 2006;419:3-23.

6. Reynolds BA, Weiss S. Generation of neurons and astrocytes from isolated cells of the adult mammalian central nervous system. Science 1992;255 : 1707- 1710. 7. Follenzi A, Naldini L. HIV-based vectors. Preparation and use. Methods Mol Med 2002;69:259-274.

8. Namikawa K, Murakami K, Okamoto T et al. A newly modified SCG10 promoter and Cre/loxP-mediated gene amplification system achieve highly specific neuronal expression in animal brains. Gene Ther 2006;13:1244-1250.

9. Rizzo MA, Springer GH, Granada B et al. An improved cyan fluorescent protein variant useful for FRET. Nat Biotechnol 2004;22:445-449.

10. Kawaguchi A, Miyata T, Sawamoto K et al.. Nestin-EGFP transgenic mice:

visualization of the self-renewal and multipotency of CNS stem cells. Mol Cell Neurosci 2001;17:259-273.

11. Choi T, Huang M, Gorman C et al. A generic intron increases gene expression in transgenic mice. Mol Cell Biol 1991;11 :3070-3074.

12. Sastry L, Johnson T, Hobson MJ et al.. Titering lentiviral vectors: comparison of DNA, RNA and marker expression methods. Gene Ther 2002;9:1155-1162.

13. Pool M, Thiemann J, Bar-Or A et al.. NeuriteTracer: a novel ImageJ plugin for automated quantification of neurite outgrowth. J Neurosci Methods 2008;168: 134-139. 14. Capowski EE, Schneider BL, Ebert AD et al.. Lentiviral vector-mediated genetic modification of human neural progenitor cells for ex vivo gene therapy. J Neurosci Methods 2007;163:338-349.

15. Yamaguchi M, Saito H, Suzuki M et al.. Visualization of neurogenesis in the central nervous system using nestin promoter-GFP transgenic mice. Neuroreport 2000; 11 :1991- 1996.

16. Lothian C, Lendahl U. An evolutionarily conserved region in the second intron of the human nestin gene directs gene expression to CNS progenitor cells and to early neural crest cells. Eur J Neurosci 1997;9:452-462.

17. Sawamoto K, Yamamoto A, Kawaguchi A et al. Direct isolation of committed neuronal progenitor cells from transgenic mice coexpressing spectrally distinct fluorescent proteins regulated by stage-specific neural promoters. J Neurosci Res 2001;65:220-227.

18. Gloster A, Wu W, Speelman A et al.. The T alpha 1 alpha-tubulin promoter specifies gene expression as a function of neuronal growth and regeneration in transgenic mice. J

Neurosci 1994;14:7319-7330. 19. Gal JS, Morozov YM, Ayoub AE et al.. Molecular and morphological heterogeneity of neural precursors in the mouse neocortical proliferative zones. J Neurosci 2006;26:1045- 1056.

20. Mizutani K, Yoon K, Dang L et al. Differential Notch signalling distinguishes neural stem cells from intermediate progenitors. Nature 2007;449:351-355.

21. Gossen M, Bujard H. Tight control of gene expression in mammalian cells by tetracycline-responsive promoters. Proc Natl Acad Sci U S A 1992;89:5547-5551.

22. Gossen M, Freundlieb S, Bender G et al.. Transcriptional activation by tetracyclines in mammalian cells. Science 1995;268:1766-1769. 23. Ying QL, Stavridis M, Griffiths D, Li M, Smith A (2003): Conversion of embryonic stem cells into neuroectodermal precursors in adherent monoculture. Nat Biotechnol 21 : 183-186. 31.

24. Conti L, Pollard SM, Gorba T, Reitano E, ToseUi M, Biella G, Sun Y, Sanzone S, Ying QL, Cattaneo E, Smith A (2005): Niche-independent symmetrical self-renewal of a mammalian tissue stem cell. PLoS Biol 2005, 3:e283.

25. Watanabe, K., Kamiya, D., Nishiyama, A., Katayama, T., Nozaki, S., Kawasaki, H., Watanabe, Y., Mizuseki, K., and Sasai, Y. (2005). Directed differentiation of telencephalic precursors from embryonic stem cells. Nat. Neurosci. 8, 288-296.

26. Watanabe, K., Ueno, M., Kamiya, D., Nishiyama, A., Matsumura, M., Wataya, T., Takahashi, J.B., Nishikawa, S., Muguruma, K., and Sasai, Y. (2007). A ROCK inhibitor permits survival of dissociated human embryonic stem cells. Nat. Biotechnol. 25, 681-686.

27. Eiraku M, Watanabe K, Matsuo-Takasaki M, Kawada M, Yonemura S, Matsumura M, Wataya T, Nishiyama A, Muguruma K, Sasai Y (2008). Self-organized formation of polarized cortical tissues from ESCs and its active manipulation by extrinsic signals. Cell Stem Cell 3(5):519-532.

28. Gaspard N, Bouschet T, Hourez R, Dimidschstein J, Naeije G, van den Ameele J, Espuny-Camacho I, Herpoel A, Passante L, Schiffmann SN, Gaillard A, Vanderhaeghen P (2008). An intrinsic mechanism of corticogenesis from embryonic stem cells. Nature 455(7211):351-357.]

29. Takahashi K. and Yamanaka S (2006) Induction of pluripotent stem cells from mouse embryonic and adult fibroblast cultures by defined factors. Cell 126, 663-676 30. Wernig M, Meissner A, Foreman R, Brambrink T, Ku M, Hochedlinger K, Bernstein BE, Jaenisch R (2007) In vitro reprogramming of fibroblasts into a pluripotent ES-cell- like state. Nature 448, 318-324

31. Maherali N, Sridharan R, Xie W, Utikal J, Eminli S, Arnold K, Stadtfeld M, Yachechko R, Tchieu J, Jaenisch R, Plath K, Hochedlinger K (2007) Directly reprogrammed fibroblasts show global epigenetic remodeling and widespread tissue contribution. Cell Stem Cell 1, 55-70

32. Okita K, Hong H, Takahashi K, Yamanaka S (2008) Generation of mouse induced pluripotent stem cells without viral vectors. Science 322, 949-953

33. Stadtfeld M, Nagaya M, Utikal J, Weir G, Hochedlinger K (2008) Induced pluripotent stem cells generated without viral integration. Science 322, 945-9446.

34. Brancaccio M, Pivetta C, Granzotto M, Filippis C, Mallamaci A (2010). Emx2 and Foxgl inhibit gliogenesis and promote neuronogenesis. Stem Cells 28(7): 1206-18].

Claims

Claims

1. A recombinant expression vector suitable for gene therapy and able to effectively express in a differentially state specific manner Emx2 and/or Foxgl polypeptides or functional fragments thereof in neural stem cells (NSCs) and/or in neuron-restricted progenitors (NPs).

2. The recombinant expression vector according to claim 1 wherein the Emx2 polypeptide is a molecule consisting of at least 95 % identity with SEQ ID No. 1 and the Foxgl polypeptide is a molecule consisting of at least 95 % identity with SEQ ID No. 3.

3. The recombinant expression vector according to claim 1 or 2 for use as gene therapy of brain damages.

4. A neural stem cell (NSCs), or a neuron-restricted progenitor (NPs) cell, or a mixed cell population thereof being transformed with the recombinant expression vector according to claim 1 or 2.

5. The neural stem cell (NSCs) and/or a neuron-restricted progenitor (NPs) cell or a mixed cell population thereof according to claim 4 for use as gene therapy of brain damages.

6. A method for the modulation of the differentiation of NSCs and/or of neuron-restricted progenitors (NPs) comprising overexpressing the Emx2 polypeptide or a functional fragment thereof in said NSCs and/or of neuron-restricted progenitors (NPs).

7. The method according to claim 6 wherein said modulation of differentiation consists in a decrease of NSCs glial output and in a stimulation of survival and terminal differentiation ofNPs.

8. A method for the modulation of the differentiation of NSCs and/or of neuron-restricted progenitors (NPs) comprising overexpressing the Foxgl polypeptide or a functional fragment thereof in said NSCs and/or of neuron-restricted progenitors (NPs).

9. The method according to claim 8 wherein said modulation of differentiation consists in a decrease of NSCs glial output and in a stimulation of survival and acceleration of neuritic growth of NPs.

10. A method for the modulation of the differentiation of NSCs and/or of neuron-restricted progenitors (NPs) comprising overexpressing the Emx2 and Foxgl polypeptides or functional fragments thereof in said NSCs and/or of neuron-restricted progenitors (NPs).

11. The method according to claim 10 wherein said modulation of differentiation consists in a decrease of NSCs glial output and in a stimulation of survival and terminal differentiation of NPs.

12. The method according to any one of claims 6 to 11 comprising the step of transforming said NSCs and/or neuron-restricted progenitors (NPs) with a vector able to overexpress Emx2 and/or Foxgl polypeptides or functional fragments thereof.

13. The method according to claim 12 wherein the vector is the vector according to claim 1