One of the possible disadvantages of
One of the possible disadvantages of this methodology is a possible cytotoxic effect resulting from either adenoviral vector injection or to Cre recombinase expression. Adenovirus has previously been used to efficiently mediate transgene expression in mouse Nepafenac receptor (Thomas et al., 2000, Xia et al., 2002) and in glial cells in vitro (Smith-Arica et al., 2000, Southgate et al., 2008) without any associated toxic effects. In our study, we observed cytotoxicity mediated by Ad5-Cre virus in cultured DRG cells derived from Rosa26 mice only at the highest dose (1000iu) and only after prolonged exposure to Ad5 vector. At doses of 10 and 100iu/ml we did not observe any sign of cellular toxicity at the time points investigated, while at 1000iu/ml signs of cellular toxicity were detected starting at 3 days after inoculation of the viral particles, as observed previously (Smith-Arica et al., 2000). No sign of adenoviral-induced toxicity was observed in the injected thalamic nuclei, as detected by morphological analysis and by evaluating astrocytic density. In agreement, the amount of Ad5 virus injected was previously reported not to induce immunological or inflammatory response in the brain (Xia et al., 2002, Kremer, 2005) while only at high multiplicity of infections (>5×108iu/injection) cytotoxic effect in glial cells in rodent brain have been reported (Kajiwara et al., 2000). Several authors have reported inflammation and cytotoxic effects associated with viral-mediated expression of Cre recombinase in cell cultures and in the brain (Silver and Livingston, 2001, Loonstra et al., 2001). However, viral vector-mediated Cre expression has been extensively used over the past years as strategy for spatio-temporal modulation of genes, without any reported signs of Cre-derived cytotoxicity in vitro and in brain regions (Prost et al., 2001, Ahmed et al., 2004). Also in our study, we did not observe any signs of Cre-mediated toxicity both in vitro, in DRG cultures, and in vivo, in injected thalamic nuclei. It has been reported that the Cre-mediated toxic effects is dose-dependent and that limited amount of Cre protein expression are not associated with adverse effects (Loonstra et al., 2001). Moreover, site specific viral vector delivery allows Cre recombinase expression in a discrete region, limiting the possibility of Cre-derived toxicity. After injection in the thalamic nuclei, we observed a certain level of anatomical damage around the area of needle penetration, in particular 1 week post injection, while at 2 weeks post injection no evident signs of morphological damage were seen. We confirmed that this anatomical lesion was not associated with inflammatory process, since the levels of GFAP staining was comparable in all the thalamic area and was not increased in the injected region. Our in vitro and in vivo studies confirm the ability of Ad5 virus to infect both dividing and non-dividing cells and specifically, both glia and neurons. A higher degree of co-localization between Cre protein and GFAP was observed in the injected thalamic slices, with lower but consistent degree of Cre-MAP2 co-localization in the same area, confirming the specific glial tropism of adenovirus as previously reported in rodent brain (Kremer, 2005). In first instance, we validated the Ad5-Cre vector for its ability to efficiently express Cre recombinase and mediate site-specific recombination in the reporter mouse strain Rosa26. We observed β-galactosidase staining in DRG cultures derived from Rosa26 mice. A dose- and time-dependent increase in β-gal staining was observed, with a larger number of cells expressing β-gal, in the absence of toxic effects at 100iu/ml after 48–72h. Similarly, when the Ad5-Cre virus was injected in the thalamic nuclei, we measured a widespread β-gal staining only in the area of virus diffusion, confirming the ability of Ad5-Cre to mediate functional Cre expression only in the infected cells. In the area of viral injection, both glial and neuronal cells showed β-gal staining.