We have used cell-attached patch-clamp electrophysiology to characterize the activation and distribution of an 85 pS K+ channel on freshly dissociated rat striatal (caudate-putamen) neurons. In recordings from 643 cells, openings of this channel showed an absolute dependence on the presence of dopamine or the D2-like dopamine receptor agonist quinpirole in the cell-attached patch pipette, but were never seen when the D2 antagonist domperidone was applied along with quinpirole, or in the absence of drug. This channel displayed inward rectification at depolarized membrane potentials, but its activation was otherwise voltage insensitive. It was largely restricted to a subset of dissociated cells with diameters > or = 10 microns, with channel openings seen in about 25% of patches. When present, there were typically multiple channels per patch. Cells of this size were immunocytochemically stained for neuron-specific enolase but not glial fibrillary acidic protein; about 40% were also labeled for gamma-amino butyric acid (GABA) and about 60% for NADPH diaphorase, with GABAergic cells displaying a shape most similar to that of cells expressing the channel. A large number of distinct types of other channels were also present, comprising inwardly rectifying channels of 5-35 pS conductance and voltage-activated channels of 100-250 pS, but the frequencies of occurrence and fractional open times of these channels were independent of the presence or absence of dopaminergic agonists. Thus, the 85 pS K+ channel uniquely requires activation by a D2-like dopamine receptor on rat striatal neurons, and is selectively expressed by a subset of these cells, which are most likely to be GABAergic neurons.

To examine the role of G(o) in modulation of ion channels by neurotransmitter receptors, we characterized modulation of ionic currents in hippocampal CA3 neurons from mice lacking both isoforms of Galpha(o). In CA3 neurons from Galpha(o)(-/-) mice, 2-chloro-adenosine and the GABA(B)-receptor agonist baclofen activated inwardly rectifying K(+) currents and inhibited voltage-dependent Ca(2+) currents just as effectively as in Galpha(o)(+/+) littermates. However, the kinetics of transmitter action were dramatically altered in Galpha(o)(-/-) mice in that recovery on washout of agonist was much slower. For example, recovery from 2-chloro-adenosine inhibition of calcium current was more than fourfold slower in neurons from Galpha(o)(-/-) mice [time constant of 12.0 +/- 0.8 (SE) s] than in neurons from Galpha(o)(+/+) mice (time constant of 2.6 +/- 0.2 s). Recovery from baclofen effects was affected similarly. In neurons from control mice, effects of both baclofen and 2-chloro-adenosine on Ca(2+) currents and K(+) currents were abolished by brief exposure to external N-ethyl-maleimide (NEM). In neurons lacking Galpha(o), some inhibition of Ca(2+) currents by baclofen remained after NEM treatment, whereas baclofen activation of K(+) currents and both effects of 2-chloro-adenosine were abolished. These results show that modulation of Ca(2+) and K(+) currents by G protein-coupled receptors in hippocampal neurons does not have an absolute requirement for Galpha(o). However, modulation is changed in the absence of Galpha(o) in having much slower recovery kinetics. A likely possibility is that the very abundant Galpha(o) is normally used but, when absent, can readily be replaced by G proteins with different properties.

By combining biochemical experiments with computer modelling of biochemical reactions we elucidated some of the currently unresolved aspects of calciumcalmodulin-dependent protein kinase II (CaMKII) activation and autophosphorylation that might be relevant for its physiological function and provided a model that incorporates in detail the mechanism of CaMKII activation and autophosphorylation at T286 that is based on experimentally determined binding constants and phosphorylation rates. To this end, we developed a detailed state model of CaMKII activation and autophosphorylation based on the currently available literature, and constrained it with data from CaMKII autophosphorylation essays. Our model takes exact phosphorylation patterns of CaMKII holoenzymes into account, and is valid at physiologically relevant conditions where the concentrations of calcium and calmodulin are not saturating. Our results strongly suggest that even when bound to less than fully calcium-bound calmodulin, CaMKII is in the active state, and indicate that the autophosphorylation of T286 by an active non-phosphorylated CaMKII subunit is significantly faster than by an autophosphorylated CaMKII subunit. These results imply that CaMKII can be efficiently activated at significantly lower calcium concentrations than previously thought, which may explain how CaMKII gets activated at calcium concentrations existing at synapses in vivo. We also investigated the significance of CaMKII holoenzyme structure on CaMKII autophosphorylation and obtained estimates of previously unknown binding constants.

No abstract