So far, direct electrophysiological recordings from granule cell dendrites have not been possible, due to the small diameter of these processes (approximately 0.8 μm in distal and medial molecular layer, Hama et al., 1989). We used combined two-photon excitation fluorescence and infrared-scanning gradient contrast (IR-SGC, Figure 1A), suitable for recording from thin neuronal processes (Nevian et al., 2007), to obtain MDV3100 purchase dual somatodendritic recordings from granule cells. We first studied the attenuation of action potentials evoked by somatic

current injection that back-propagated into granule cell dendrites (bAPs, Figures 1B and 1C, see insets for magnifications). The bAP amplitudes decreased strongly toward more remote dendritic recording sites (n = 20, Figure 1D), with an attenuation length constant of 86.0 ± 8.5 μm. This corresponds to a much steeper attenuation than reported either for pyramidal cell main apical (Golding et al., 2001) or basal dendrites (Nevian et al., 2007). When the bAP amplitudes were plotted over the somatodendritic distance normalized to the total length

of the dendrite, the attenuation length constant was 0.31 ± 0.04 BIBW2992 clinical trial (n = 14, Figure 1E). Using this analysis, attenuation normalized to total dendritic distance seemed to be similar to both pyramidal cell apical and basal dendrites (cf. Figure S3B in Nevian et al., 2007). Concomitantly,

the delay of the action potential peaks increased (Figure 1F). The average conduction velocity of action potentials back-propagating into granule cell dendrites was calculated from the action potential peak delays, yielding 149.3 ± 2.0 μm·ms−1. This conduction velocity is markedly different when compared with pyramidal cell apical dendrites (approximately 500 μm·ms−1; Stuart et al., 1997a) and is also lower than the estimates for basal dendrites (approximately 200 μm·ms−1; Antic, 2003 and Nevian et al., 2007). We also observed pronounced distance-dependent next broadening of bAPs, which manifested in a decrease of the maximal rate of rise of bAPs (δV/δt, Figure 1G) together with an increase in bAP half width (Figure 1H). We next studied how bursts of action potentials back-propagate into granule cell dendrites. In our recordings of bAPs during repetitive firing induced by prolonged somatic current injections, we had already observed that the amplitudes of individual bAPs stayed constant during trains of action potentials (Figure 1C, see red dendritic voltage recording). This is in contrast to pyramidal cells, in which a pronounced amplitude reduction during a train of action potentials due to slow inactivation of dendritic voltage-gated Na+ channels was described (Colbert et al., 1997, Jung et al., 1997 and Spruston et al., 1995).