Presentations

Chloride dynamics alter the input-output properties of neurons

Christopher B. Currin, Andrew J. Trevelyan, Tim P. Vogels, and Joseph V. Raimondo

Summary

Different populations of inhibitory interneurons change the input-output (IO) function of their targets in different ways [1]–[5]. A factor which has received less consideration is changes in the reversal potential of GABA (∆EGABA), arising from changes in neuronal chloride ion concentration ([Cl-]i). Chloride-loading of neurons is known to be a factor in various neurological conditions [6]. Here we investigate its impact on the input-output function of neurons.

We explored how local dendritic parameters affect the dynamics of [Cl-]i in response to synaptic input, and how this might affect computation depending on the location of synaptic input. By introducing a novel measure of the functional influence of Cl- on the neuron’s output, the ‘chloride index’, we quantify the difference between a model with and without Cl- dynamics. Our results highlight the importance of accounting for Cl- dynamics when designing computational models that include dendritic inhibition.

Inhibition can quickly change the computational properties of a neuron over time as Cl- accumulates inside the dendrites. Neurons with finely balanced excitatory and inhibitory synaptic input are particularly susceptible to transient changes in Cl- with shifts in firing rate being more pronounced over time, as well as over increasing frequency of input. Distal GABAergic dendritic input is more susceptible to [Cl-]i accumulation than proximal input, which impacts the amount of control inhibition has on the output of the neuron and maps to physiological findings of GABA synapse distribution in the cortex.

These results are highly relevant for assessing computational models which incorporate static inhibition and generally apply to models exhibiting inhibitory plasticity. We highlight the important computational changes neurons can undergo over short periods of time due to inhibitory input, which applies to models of dendritic processing, the behaviour of individual neurons, and also to networks of neurons.

Additional Detail

A neuron’s principle function is to transform its synaptic input into an output firing rate. Inhibition is crucial in shaping the transformation due to its strong modulatory ability, allowing for a broader dynamic range of responses [7]. Fast synaptic inhibition is mediated by GABAA receptors, which are selectively permeable to chloride (Cl-) and, to a lesser extent, bicarbonate ions. As a result, the transmembrane Cl- gradient plays a critical role in setting the properties of synaptic inhibition within neuronal networks [8]. Here we investigated the impact of transient changes to internal chloride ion concentration ([Cl-]i) on a neuron’s output, mediated by changes to the driving force of GABAARs; a phenomenon known as short-term ionic plasticity [9]. We demonstrate the importance of accounting for dynamic [Cl-]i in theoretical and computer-based models, particularly in situations where dendritic inhibition is being considered.

To explore the functional relevance of Cl- dynamics, we developed computational tools to model ionic plasticity and [Cl-]i regulation in multi-compartment models using NEURON. By utilising these computational tools together with our experimentally collected data on neuronal Cl- extrusion, we have identified novel and important effects of ionic plasticity on the input-output properties of neurons.

We show that with the onset of synaptic bombardment, there is a progressive change in Cl- that evolves over a few hundred milliseconds, that there are clear differences in the input-output structure of neurons depending on the location of input, the size of the dendrites, the relative levels of excitatory and inhibitory synaptic input, as well as the strength of Cl- extrusion (KCC2 activity). Our experimentally-based equations for Cl- dynamics, can be applied to various contexts of computational complexity, such as morphologically realistic neurons and point neurons in a network.

We are currently extending this work to determine how various patterns of steady-state synaptic drive reflect what is known about different brain states (e.g. up-states, seizures), and how dynamic changes to Cl- levels within the dendritic arbor might contribute to these various patterns of activity.

Figure 1 Cl- dynamics alter the input-output properties of neurons. (A) Accounting for dynamic Cl- markedly reduces the effect of dendritic (top) but not somatic (bottom) inhibition on the I/O function. B) The chloride index quantifies the effect of dynamic Cl- on the I/O function. A high Cl- index means that dynamic Cl- is reducing the effect of inhibition. C) Chloride index is directly proportional to the change in EGABA during synaptic input. D) The I/O function of neurons is progressively perturbed over time when accounting for Cl- dynamics.

References

  1. B. V Atallah, W. Bruns, M. Carandini, and M. Scanziani, “Parvalbumin-expressing interneurons linearly transform cortical responses to visual stimuli.,” Neuron, vol. 73, no. 1, pp. 159–70, Jan. 2012.
  2. A. J. Trevelyan and O. Watkinson, “Does inhibition balance excitation in neocortex?,” Prog. Biophys. Mol. Biol., vol. 87, no. 1 SPEC. ISS., pp. 109–143, 2005.
  3. E. T. Vu and F. B. Krasne, “Evidence for a Computational Distinction Between Proximal and Distal Neuronal Inhibition,” Science (80-. )., vol. 255, no. 5052, pp. 1710–1712, 1992.
  4. N. R. Wilson, C. A. Runyan, F. L. Wang, and M. Sur, “Division and subtraction by distinct cortical inhibitory networks in vivo,” Nature, vol. 488, no. 7411, pp. 343–348, 2012.
  5. M. Lovett-Barron et al., “Regulation of neuronal input transformations by tunable dendritic inhibition.,” Nat. Neurosci., vol. 15, no. 3, pp. 423–30, S1-3, 2012.
  6. J. V Raimondo, B. A. Richards, and M. A. Woodin, “Neuronal chloride and excitability — the big impact of small changes,” Curr. Opin. Neurobiol., vol. 43, pp. 35–42, 2017.
  7. J. S. Isaacson and M. Scanziani, “How inhibition shapes cortical activity.,” Neuron, vol. 72, no. 2, pp. 231–43, Oct. 2011.
  8. M. Farrant and K. Kaila, “The cellular, molecular and ionic basis of GABAA receptor signalling,” Prog. Brain Res., vol. 160, pp. 59–87, 2007.
  9. J. V Raimondo, H. Markram, and C. J. Akerman, “Short-term ionic plasticity at GABAergic synapses.,” Front. Synaptic Neurosci., vol. 4, no. October, p. 5, Jan. 2012.