Firing first: compensatory changes in K+ channel knockout mice preserve excitability but not synaptic scaling

Abstract

Synaptic potentials are severely filtered as they propagate through the dendritic tree. As action potentials are usually generated in the axon, how can distal synapses influence neuronal output? In hippocampal CA1 pyramidal neurons, distal synapses cope with this situation by producing progressively larger synaptic currents at increasing distance from the soma (Magee & Cook, 2000). An increase in the number of AMPA receptors with distance from the soma is thought to be responsible for this amplitude normalization (distance-dependent synaptic scaling) (Andrasfalvy & Magee, 2001). The question then arises: how is this gradient set-up; how does the distal synapse come to know of its disadvantaged location and express more AMPA receptors? In this issue of The Journal of Physiology, Andrasfalvy and colleagues (Andrasfalvy et al. 2008) investigate the possibility that action potentials, initiated in the axon but propagating back into the dendrites (bAPs), provide feedback to the synapses about their location. In the primary apical dendrite of hippocampal CA1 pyramidal neurons, bAPs attenuate as they travel away from the soma (Spruston et al. 1995). This attenuation can by reduced by pharmacologically blocking A-type K+ channels, which increase several-fold in density with distance from the soma (Hoffman et al. 1997). Greater back-propagation is also observed upon genetic down-regulation or deletion of the A-type K+ channel α subunit Kv4.2, establishing this protein as the molecular identity of dendritic A-type currents (IA) in these neurons (Fig. 1) (Kim et al. 2005; Chen et al. 2006). In the present work, the authors compared miniature excitatory postsynaptic current (mEPSC) amplitudes in proximal and distal dendritic recordings from Kv4.2−/− and wild-type (WT) mice. Consistent with previous data from this group, in WT mice, distal mEPSCs were larger than proximal. If bAPs inform the synapse of its location, larger bAPs in Kv4.2−/− mice might then be expected to exhibit smaller mEPSCs at distal locations compared to those from WT. Indeed, distance-dependent scaling was lost in Kv4.2−/− mice, predominately through a reduction in distal mEPSC amplitudes. Figure 1 Targeted rescue of intrinsic excitability in Kv4.2−/− mice Previous studies using dominant negative Kv4.2 constructs to acutely decrease its function, in both neocortical and hippocampal pyramidal neurons, have demonstrated a prominent role of Kv4.2 in regulating membrane excitability (e.g. input resistance, AP threshold, half-width and frequency) (Kim et al. 2005; Yuan et al. 2005). Recently, however, recordings from cortical pyramidal neurons of Kv4.2−/− mice revealed no major effects of Kv4.2 elimination on firing properties compared to WT. The rescue of firing properties appears to come about through increased densities of other types of K+ channels to compensate for the loss in IA (Nerbonne et al. 2008). In hippocampal neurons, Andrasfalvy et al. also find only modest changes in membrane excitability of Kv4.2−/− mice. Although there is also some evidence for compensation via increased expression of other K+ channel subunits in Kv4.2−/− hippocampal neurons (Chen et al. 2006), in the present report, the authors also find changes in tonic and phasic GABA currents. Blocking these inhibitory currents revealed increased membrane excitability in knockout hippocampal pyramidal neurons compared to WT. In addition to demonstrating important roles for dendritic K+ channels in distance-dependent scaling, the study by Andrasfalvy et al. raises a number of questions. First, are bAPs themselves responsible for setting up the distance-dependent scaling or are other factors that also depend on Kv4.2 expression involved? If bAPs are responsible, what signalling cascades are involved? Finally, distance-dependent synaptic scaling is reminiscent of another form of plasticity that promotes network stability, synaptic homeostasis. In synaptic homeostasis, excitatory input strength throughout the neuron is negatively regulated by extended periods of high activity or positively by low activity. How are these two homeostatic mechanisms related and does synaptic homeostasis, unlike distance-dependent scaling, persist in Kv4.2−/− neurons? It is tempting to compare the acute effects of Kv4.2 down-regulation in CA1 neurons of organotypic slice cultures (Kim et al. 2005) with the more prolonged and complete elimination in the knockout presented here (Fig. 1). Interestingly, while both manipulations decrease IA, resulting in greater back-propagation and changes in mEPSC amplitudes; membrane excitability and firing patterns, altered with acute Kv4.2 knockdown, are preserved in Kv4.2−/− mice through the compensatory mechanisms detailed above. Under the same developmental and homeostatic influences, however, the distance-dependent gradient of synaptic strength is lost in knockout mice. This privileged rescue of intrinsic excitability (a cellular phenomenon lacking spatial specificity) with apparent disregard for the normal spatial distribution of synaptic weights, runs counter to the notion that compensatory mechanisms would act to preserve, as much as possible, the neuron's computational power. Rather these results indicate that the plasticity of intrinsic excitability may act as an important information storage system distinct from synaptic plasticity (Zhang & Linden, 2003).