Consistent with this idea, earlier

work has shown that localized downregulation of dendritic A-type potassium channels can occur during induction of long-term potentiation (Frick et al., 2004). In both cases, downregulation of dendritic A-type potassium channels has been shown to require activation of NMDA receptors. Earlier work indicated that A-type potassium channels have a range of effects on dendritic integration CP-868596 cost in CA1 pyramidal neurons, acting to either linearize or suppress excitatory postsynaptic potential summation (Cash and Yuste, 1999; Hoffman et al., 1997). One of the most interesting findings in the paper is that the capacity of recurrent inhibition to reduce the amplitude of dendritic glutamate-evoked depolarizations that are subthreshold for generation of dendritic spikes is weaker in dendritic branches that generate

strong dendritic spikes. This result is even more surprising given that much of the recurrent inhibitory input recruited by stimulation of the alveus will be located at the soma. Application of GABA to these dendritic branches suggested that the difference in the impact of recurrent inhibition on different dendritic branches is not due to differences in the density of GABA receptors or the reversal potential for GABA. These data suggest that the number or release Depsipeptide ic50 probability of GABAergic inputs recruited during recurrent inhibition is lower in dendritic branches that generate strong dendritic spikes. How this occurs is unclear, but it may involve the release of a retrograde signal, possibly in response to generation of dendritic spikes.

Because the conversion of weak dendritic branch Thymidine kinase spikes to strong dendritic branch spikes did not influence the capacity of recurrent inhibition to reduce the amplitude of subthreshold glutamate-evoked depolarizations, this process presumably takes time to develop and occurs subsequent to downregulation of A-type potassium channels in these dendritic branches. Whether this is associated with similar, or perhaps opposite, changes in feedforward inhibition on these dendritic branches is unclear. Finally, it is worth commenting on the impact of the findings on the overall excitability of CAI pyramidal neurons. Earlier work has shown that pairing dendritic spikes with action potentials can convert weak dendritic spikes to strong dendritic spikes (Losonczy et al., 2008), thereby enhancing dendritic excitability. The current work by Müller and colleagues (Müller et al., 2012) adds to this data, showing that dendritic branches that generate strong dendritic spikes are also associated with weaker recurrent inhibition. This would be expected to further enhance dendritic excitability.