Strikingly, a primary decrease in the firing of MD neurons selectively disrupted this task-specific increase in MD-PFC beta-synchrony. In conclusion, our results demonstrate a causal relationship between decreased MD activity and deficits in executive function. They further suggest thalamofrontal beta synchrony as a potential mechanism contributing to working memory. A modified human muscarinic receptor, hM4D, was coexpressed along with humanized renilla green fluorescent protein (hrGFP) selectively in the MD using an adenoassociated virus expression system (AAV2-hM4D-IRES-hrGFP) (Figure 1A). The hM4D receptor is activated solely by a pharmacologically

inert compound, clozapine-N-oxide (CNO), and not by endogenous acetylcholine or any other neurotransmitter. Upon CNO activation, hM4D hyperpolarizes neurons through a G protein mediated activation of inward-rectifying selleck chemicals llc potassium channels ( Armbruster et al., 2007). Stereotactic injection of AAV2-hM4D-IRES-hrGFP virus into the MD induced coexpression of both GFP (green) and hM4D as assessed

by anti-HA immunostaining (red) ( Figure 1B). At higher magnification, we observed that the hM4D is localized in the plasma membrane as well as in neuronal processes ( Figures 1B and 1D). Costaining with this website anti-NeuN antibodies revealed that GFP expression was exclusively neuronal ( Figure 1C, top panel). Due to the absence of interneurons in rodent MD ( Kuroda et al., 1998), all infected neurons should be relay projection neurons. Using AAV2.2, an average of 27% ± 6% of the MD neurons expressed the GFP with a peak at 66% ± 9.6%

at the site of injection ( Figure 1C, bottom panel). The virus spread almost entirely among the anterioposterior axis of the MD whereas it stayed constrained within the second dorsoventral and lateromedial axis of the MD ( Figure 1C and see Figure S1A available online). Consistent with the known anatomy of the MD ( Kuroda et al., 1998), infected neurons projected to layers I and III/V of prelimbic and orbitofrontal cortices ( Figure 1D). To determine whether activation of hM4D will hyperpolarize thalamic neurons as has been shown for hippocampal neurons (Armbruster et al., 2007), we performed whole-cell patch-clamp recordings from thalamic slices. Fluorescently-identified neurons expressing hM4D were significantly hyperpolarized after CNO bath application, while the resting membrane potential of cells infected with a control virus expressing only hrGFP were not (Figure 2A). To determine whether activation of hM4D reduces the activity of MD neurons in vivo, we performed multiple single-unit recordings from the MD of freely moving mice injected with AAV2-hM4D-IRES-hrGFP (MDhM4D mice). For this experiment, mice were exploring a familiar environment, specifically a T-maze to which they had been previously habituated.