A self-activating intrinsic brake on bursting in CA3 neurons

In the 1960s, intracellular recordings from cortical neurons within epileptogenic foci demonstrated that the interictal spike was associated with a paroxysmal depolarizing shift (Matsumoto & Ajmonemarsan, 1964). This led to the concept of the ‘epileptic’ neuron, and it was noted that the paroxysmal depolarizing shift in these neurons was redolent of the spike afterdepolarizing potential (ADP) that had been previously described in hippocampal neurons. Indeed, it was hypothesized that the high epileptogenic potential of the hippocampus was a result of the spike depolarizing afterpotential. Since that time, there has been a resurgence of interest in the ‘epileptic’ neuron hypothesis as evidence has accumulated that neurons can change their firing patterns during the epileptogenic process (Beck & Yaari, 2008).

Since the ADP can promote pathological network activity, and it is present in normal CA3 pyramidal cells, then a mechanism by which the ADP can be controlled during excessive network activity may act as a brake preventing the transition of this activity to seizures.

In this issue of The Journal of Physiology, Brown & Randall (2009) provide a novel mechanism by which the ADP can be suppressed during network activity. They begin with a comprehensive characterization of the firing properties of CA3 pyramidal neurons, and show that, although there is much diversity in the overall spiking pattern of these cells, almost all cells fire a short burst (up to 5 APs) in response to different depolarizing stimuli. These bursts are marked by extremely reproducible timing of the first few spikes, and rely on a robust afterdepolarizing potential (ADP) that returns cells to threshold after the first action potential (see Fig. 1).

Figure 1. The role of the ADP on burst generation.

Three sweeps are overlaid following identical first spikes. The number and timing of spikes is set by the size of the ADP.

In different neurons an ADP may be driven by different channels. T-type calcium channels, persistent sodium currents from unknown sodium channels, and a ‘non-specific cation channel’ are all implicated in producing ADPs. All of these channels have a low threshold of activation, a tendency to depolarize cells when activated, and a challenging inaccessibility to pharmacological modification which makes ADPs very difficult to suppress directly.

Brown and Randall show that by activating I_M, various conditioning stimuli (CS) can dampen the ADP in CA3 neurons and consequently reduce burst firing. A variety of CS, ranging from robust depolarization (150 s at −14 mV; Brown and Randall call it pathological), to stimuli designed to mimic trains of synaptic currents that CA3 cells may experience in vivo, are capable of activating enough I_M to suppress the ADPs and disrupt the well-defined bursts of these neurons. In fact, some cells stop firing bursts altogether, and convert to firing individual action potentials. The shift away from burstiness is accompanied by a longer delay between the first and second spikes and much more jitter in when the second spike occurs.

This form of intrinsic plasticity is long lasting; Brown and Randall follow it for more than 30 min without decay, and it appears to be intrinsic to the neurons, as all the recordings are done with synaptic currents blocked. Could this plasticity be used to suppress pathological bursting?

Healthy CA1 pyramidal neurons do not generally fire bursts; however, it has been shown that during epileptogenesis there is a pronounced shift in CA1 pyramidal cells towards bursting (Beck & Yaari, 2008). Consistent with this, ADPs are increased in bursting CA1 cells, possibly through an upregulation of T-type or R-type calcium currents, increased persistent sodium currents, or some mix of these. Is it possible to block these increased ADPs by activation of I_M? Brown and Randall worked exclusively in CA3 cells; however a similar I_M-dependent form of intrinsic plasticity was recently described in healthy CA1 neurons (Wu et al. 2008). Unlike the CA3 cells, the healthy CA1 cells did not fire bursts even in the pre-conditioned state, but activation of I_M with a depolarizing CS leads to a qualitatively similar decrease in spike number during long test depolarizations.

Taken together, the data from Brown & Randall (2009), showing that intrinsic activation of I_M can suppress bursting, and from Wu et al. (2008), suggesting that I_M can be intrinsically activated in CA1 cells, raise the question of whether pathologically bursting CA1 neurons in epileptic tissue might be quenched by intrinsic plasticity of I_M. As Beck & Yaari (2008) point out, there are a growing number of neurological disorders where neurons have increased ADPs and shift to a bursting firing mode, and the same question might apply to these other disorders.

Mutations that decrease the slow subthreshold neuronal M-current, I_M, are now linked to several types of epilepsy, to peripheral nerve hyperexcitablity, and to deafness (Maljevic et al. 2008). Generally, reducing I_M, either genetically, or pharmacologically (e.g. by activating the muscarinic receptors from which I_M gets its name), leads to increased neuronal firing. In contrast, drugs that activate I_M (e.g. retigabine) suppress neuronal firing, particularly bursting. A growing number of neurological disorders, including epilepsy, neuropathic pain, stroke and migraine, are associated with aberrant bursting, and consequently are being investigated for potential treatment via activation of I_M (Maljevic et al. 2008).