5 h, and examined the distribution of labeled profiles in relation to presynaptic terminals. The results show a good ultrastructural preservation of the tissue, notably membrane structures, allowing unambiguous recognition of pre- and postsynaptic density, synaptic vesicles, mitochondria, GPCR Compound Library etc. (Fig. 3E), comparable with that seen after traditional tissue fixation (Panzanelli et al., 2011). Gephyrin immunogold labeling was prominent in profiles forming symmetric synaptic contacts with axon terminals enriched in synaptic vesicles. This intense immunoreactivity points to excellent preservation
of antigenicity owing to the brief post-fixation. We have assessed the suitability of the ACSF perfusion protocol for RNA purification compared with fresh-frozen tissue, and tested mRNA integrity by qPCR analysis. Experiments were performed in triplicate, using tissue from 2–3 mice per condition. Figure 3F illustrates that high-quality RNA can be purified from brain samples perfused with ACSF. Furthermore, the results demonstrate that RNA extracted from ACSF-perfused mice is compatible with qPCR analysis. By comparison with fresh brain samples, the expression level of four selected genes encoding synaptic proteins remained unaltered (Table 2), giving the opportunity to
study brain morphology and gene expression in parallel. For proof-of-principle that optimal biochemical and immunohistochemical analyses can be performed using tissue blocks taken from Verteporfin the same brain following ACSF-perfusion, we performed Western blotting and immunohistochemistry with tissue from an ACSF-perfused mouse. Each method was compared Linsitinib molecular weight with standard tissue preparations [fresh tissue for Western blotting and sections from perfusion-fixed
brain (4% paraformaldehyde) for immunoperoxidase staining]. In Western blots, we investigated the expression of Tau, APP and Reelin in cerebral cortex and hippocampus. As illustrated in Fig. 4A–C, no difference in relative abundance of Tau, APP or Reelin was observed in fresh-frozen and ACSF-perfused tissue, and proteolytic fragments of Reelin were readily detected, with clear differences in abundance between cortex and hippocampus. In parallel, we stained for Reelin in the hippocampal formation in sections that were pretreated with pepsin, prior to incubation with primary antibodies (Doehner et al., 2010). Immersion-fixation (3 h) of ACSF-perfused tissue allowed detection of Reelin immunoreactivity in hippocampal interneurons and neuropils with similar intensity and high signal-to-noise ratio as in perfusion-fixed tissue (Fig. 4D and E). We have shown previously that the detection of postsynaptic proteins of GABAergic synapses, in particular gephyrin and various GABAAR subunits, is markedly improved in weakly fixed tissue, in particular when derived from living brain slices (Schneider Gasser et al., 2006).