Unit: Retroviruses and Gene Transfer - URA CNRS 1930
Director: Jean Michel Heard
Using gene therapy, we delivered the missing enzyme to the brain of deficient animals with the aim to prevent neurodegeneration in models of lysosomal storage diseases. In vitro differentiation and in vivo motor regeneration models were used to investigate the generation of motor neurons from adult neural stem cells. We deciphered the mechanisms by which the homeogene barhl2 participates to the interpretation of the cell position information in the developing neural plate.
The notion that repair strategies could be proposed for treating central nervous system (CNS) lesions emerged from observations made over the last decade, including fetal cell graft in Parkinson's disease patients, evidence for neurogenesis in adult mammals, and the isolation of neural stem cells from various parts of the adult mammalian CNS. At the same time, gene transfer vectors capable of inducing the expression of foreign genes into CNS cells became available.
A first type of repair strategy for the CNS relies on the delivery of soluble factors that could prevent cell death or stimulate axonal regeneration. Local sources of soluble factors can be created by genetically modifying resident cells with gene transfer vectors. We approach this first type of strategy by delivering gene transfer vectors coding for the missing enzyme in models of lysosomal storage diseases, which induce dramatic neurodegeneration in children.
A second type of strategy would consist in replacing lost cells. A requisite is that CNS had sufficient plasticity in the adult to incorporate exogenous cells within the pre-existing architecture. It is postulated that neural progenitors might do so, either spontaneously, or after genetic re-programming. Successful achievement will require combining cell transplantation with interventions aimed at favouring plasticity. We approach this second type of strategies in models of motor pathway repair in rodents.
Knowledge about tools and targets that would make manipulations of neural progenitors cell fate in adults possible comes from studies of the determinants of cell differentiation in the embryo. B. Durand studies mechanisms supporting the action of a homeodomain transcription factor (Barhl2), whose expression determines how cells interpret positional information upon neural plate patterning, as well as at later developmental stages of the anterior brain.
Investigation and treatment of neurodegeneration in mucopolysaccharidosis (MPS)
Results obtained by N. Desmaris and A. Cressant in mouse models of Hurler syndrome (MPSI) and Sanfilippo syndrome (MPSIIIb) indicated that a single stereotactic injection of AAV2 or AAV5 vectors in the brain was sufficient to clearing pathology from the entire encephalon, preventing neuronal cell death and improving behavioural performances.
In the perspective of human trials, we are currently examining the therapeutic efficacy and safety of vector delivery and enzyme secretion in large animal brains, according to a clinically relevant protocol. AAV2 and AAV5 vectors were used. Experiments in affected MPSI dogs allowed to draw brain maps of vector distribution, enzyme spreading and disease correction. A clinical trial in affected children will be considered when the immune response directed against the previously unseen enzyme will be controlled and long term tolerance studies have demonstrated safety. This is a collaborative program with the team of P. Moullier in Nantes, the Ecole Nationale Vétérinaire de Nantes, a neuropediatrician (Pr. M. Tardieu) and neurosurgeons (Prs. M. Tadié and Y. Lajat).
Mechanisms responsible for lysosomal enzyme spreading in the brain are not known. Using lysosomal enzyme fused to GFP, F. Chen studies the modalities of the axonal transport of lysosomal enzymes. Experiments are performed using time-lapse microscopy on cultured neurons. In addition, the in vivo production of GFP-labelled molecules in deficient mice will allow marking of brain areas in which enzyme delivery is associated to improved behaviour.
Study of the differentiation modalities of adult stem cells in motor neurons
The aim of this program is to find conditions in which neural stem cells would differentiate into motor neurons. In vitro, stem cells from rat or mouse fetuses (E13.5) are exposed to soluble factors (RA or Sonic Hedgehog), or to lentivirus vectors coding for transcription factors involved in motor neuron differentiation (Pax6, Nkx6.1, Olig2, Ngn2, HB9, Islet1, Lhx3, etc ). Proportions of cultured cells expressing motor neuron markers (HB9, Isl1-2, ChAT) will indicate which is the optimal cocktail to induce stem cell commitment to motor neuron differentiation. Cells are also marked with GFP. An enrichment method for committed progenitors is based on the expression of a cell marker from the HB9 promoter, a gene whose transcription is activated upon commitment to the motor neuron differentiation pathway.
Treated cells were transplanted in mouse and rat models of motor pathway regeneration. Models were created by S. Liu (rats) and A. Nosjean (mice), by inserting the proximal extremity of a predegenerated nerve graft in the ventral horn of the thoracic spinal cord (rats) or the facial nerve nucleus (mice). The distal extremity is fixed to a skeletal muscle. We showed that the nerve graft provided a robust stimulation of axonal growth from endogenous motor neurons. Axons colonised the nerve graft , formed end-plates and induced motricity. The transplantation of GFP+ naive neural stem cells close to the nerve insertion site led to cell survival after two months, with GFP+ axonal prolongations extending into the nerve graft, a proportion of which expressed ChAT. The morphology of GFP+ cellular bodies was however not consistent with that of motor neurons. Similar experiments are now performed with cells previously treated with the aim to induce commitment into the motor neuron differentiation pathway.
In the absence of an appropriate control from supra-spinal motor centres, artificial connection of motor neurons to muscles does not result in functional motor patterns. In collaboration with the laboratory of P. Brûlet (Department of Developmental Biology, Institut Pasteur), we study a transynaptic labelling method based on the expression of the fusion protein GFP-TTC from muscles transduced with AAV1 vectors. Our aim is to describe anatomical variations associated to the reorganisation of supra-spinal connections in response to animal motor training. The method will also indicate whether transplanted cells could integrate pre-existing neuronal networks.
Specification of neural plate cell fate during vertebrate development.
The fate of undifferentiated cells that form the prosencephalic neural plate is determined very early during development, mostly with respect to their spatial location within the plate. Position signals are provided through gradient of morphogenetic factors such as Sonic Hedgehog, Fgf8 or BMPs. These signals are interpreted by target cells through the expression of transcription factors, most of which possess a homeodomain. B. Durand isolated a new homeogene in the xenopus and the mouse, called barhl2. She showed that appropriate expression of the transcriptional repressor Barhl2 is required for the definition of the neural plate midline, as well as, later during development, that of the dorso-ventral organisation of the neural tube and the morphogenesis of the embryo. N. Duval and N. Offner showed that Barhl2 controlled cell survival and proliferation. This is the first described transcriptional factor that participates to the induction of an apoptotic process. Indirect evidence suggests that Barhl2 occupied a crucial position in signalling pathways controlled by Sonic Hedgehog and BMPs.
Photo : transplantation of GFP-labelled neural stem cells close to the insertion site of a predegenerated nerve graft in the mouse facial nucleus. GFP+ axonal prolongations are visible in the nerve graft.
Keywords: gene therapy, stem cells, nervous system, lysosomal storage diseases, motricity, homeogenes