br Acknowledgments br Introduction Manipulating target prote

Acknowledgments
Introduction Manipulating target protein expression via either induction or suppression of gene expression is a powerful technology that has been widely used in the recent past in the field of neurobiology, both to study the pathophysiological significance of a target gene and to create disease models resembling human disorders. Inducible transgenic mouse models allowing spatio-temporal control over transgene expression have overcome several problems associated with classical constitutive transgenics, such as possible embryonic lethality or compensatory effects due to constitutive target gene modulation (Nadeau, 2001). The study of a target of interest in neuroscience could require investigation of its role in the whole appreciate or in a specific brain nucleus or cell population in either the juvenile or adult animal. In the past years several technologies have been developed to allow more sophisticated spatio-temporal control of gene expression such as tissue specific promoters or promoters that selectively modulate transgene expression in specific neuronal populations (Beglopoulos and Shen, 2004, Cazzin and Ring, 2009, Tanahira et al., 2009). A great advance in the production of inducible transgenic mouse models has been the discovery and development of the Cre-loxP system (Zheng et al., 2000, Nagy, 2000). This system uses the bacteriophage P1 Cre recombinase, a 38kDa site-specific recombinase belonging to the integrase family. This protein catalyzes the recombination and therefore excision of the DNA region between two recognized loxP sites. This technology can been applied to produce either inducible knock-out mice models or inducible knock-in models, based on the location of the loxP sites with respect to the target gene. For the inducible knock-out model, the loxP sites are located on either side of critical exons, allowing target gene expression at their normal level until Cre-mediated site-specific recombination and exon deletion with the consequent generation of a non functional target gene. Inducible knock-in animals can be produced by positioning a termination transcription signal flanked by loxP sites downstream of the target gene promoter, preventing transgene transcription. In this model, Cre recombinase expression catalyzes the removal of the termination signal, allowing transgene transcription. Several strategies have been developed in order to create temporally controlled Cre recombinase expression in the transgenic mice, such as the inducible systems controlled by tetracycline or tamoxifen or the CamKII promoter (Mansuy and Bujard, 2000). In these systems, transgenic mice are engineered by insertion of the Cre recombinase gene under the control of drug-responsive promoters or brain cell-selective promoters. These transgenic mice with inducible Cre expression are then crossed to mice engineered with the loxP sites to allow transgene modulation in a controlled manner. These strategies, despite showing considerable promise, possess the disadvantages of being extremely time-consuming and possibly being associated with position variegation effects due to difference in the location of genomic integration. Moreover, tissue specific promoters for many specific brain regions are not available and often targets of interest in the neurological research can present different functional roles in several brain nuclei, making it necessary to limit the modulation of gene expression to a specific brain area. In this context viral vectors have been extremely useful since they are natural gene transfer vehicles that can be easily engineered to carry the target genetic sequence to a wide range of mitotic and post-mitotic cells. Both RNA and DNA viruses have been validated for their ability to deliver functional Cre recombinase in discrete regions of rodent brains (Ahmed et al., 2004) and, through intracerebral injection of the viral vector in the desired brain region, it is possible to express Cre protein in a precise spatio-temporal manner in the desired animal strain and in a less time consuming manner relative to transgenic approaches. Adenovirus has been used in the past to efficiently deliver functional Cre recombinase to the brain (Wang et al., 1996) and it presents the considerable advantage of preferentially infecting glia over neurons that is mandatory if the target gene is expressed predominantly in glia. On the other hand, adenovirus used in vivo has been limited by the association with possible immunotoxicity, inflammation and relative transient expression (Schnell et al., 2001). However, due to the immune privileged status of the central nervous system (CNS), localized adenovirus injection in the brain at moderate doses leads to long-term transgene expression and has not been related to cytological damage or inflammatory response (Xia et al., 2002, Kremer, 2005).