How do individual neurons integrate excitatory and inhibitory inputs?
All CNS neurons convert the input arising from excitatory synapses to an output signal in which information is encoded as the frequency or pattern of action potentials (spikes) propagating down the axon. A preeminent topic in neuroscience has been to study the precise mechanisms underlying this integrative process.
Neuronal integration and active properties of dendrites:
How efficiently excitatory synaptic potentials (EPSPs) are converted to an action potential output is strongly affected by the presence of voltage-gated conductances in the somatodendritic compartment of neurons. Dendritic ion channels strongly influence the degree of voltage attenuation of EPSPs propagating along dendrites. In addition, dendritic voltage-gated channels strongly affect voltage attenuation and filtering of action potentials back-propagating into the dendritic tree (b-APs). Finally, voltage-gated Ca2+ and Na+ channels as well as NMDA receptors endow dendrites with the capability to generate regenerative spikes on their own (termed dendritic spikes), which can propagate to the soma and can be powerful enough to elicit axo-somatic action potentials (Spruston, 2008; Beck and Yaari, 2008). While much is known about the main apical dendrites of some types of CNS neurons, much less is known about the properties of small 2nd and 3rd order dendrites and their integrative properties, even though a large proportion of excitatory synapses terminate on these dendrites. We are using direct dendritic voltage recordings as well as focal neurotransmitter liberation by multiphoton uncaging and iontophoresis to assess integrative properties of these dendrites.
(Remy et al., 2009)
Neuronal integration and firing mode:
In addition to signal integration within dendrites, EPSP-action potential conversion is strongly affected by the firing mode of neurons. In most neurons, a brief depolarization causes generation of a single action potential, whereas a prolonged depolarization induces a series of independent action potentials. These neurons are generally termed ‘regular-firing neurons’. In some types of neurons (‘bursting neurons’), however, threshold depolarization triggers a high-frequency, all-or-none burst of action potentials. The firing mode (i.e. regular vs. burst-firing) clearly also affects input-output behaviour of single neurons strongly. Bursting can be supported by voltage-gated conductances in both dendritic and somatic compartments (Beck and Yaari, 2008; Yaari and Beck, 2002).
Because the presence of different types of voltage-gated ion channels in dendrites and soma powerfully determines the efficiency of EPSP-to-spike coupling, we are particularly interested in identifying the specific subtypes of voltage-gated channels involved in these types of integration, and in characterizing their subcellular localization and biophysical properties (Yue et al., 2005, Royeck et al. 2008, Bender et al., 2007, Müller et al. 2007, Surges et al. 2006, Uebachs et al. 2006, Riazanski et al. 2001). We approach this problem using genetic and pharmacological inhibition of ion channel subtypes. We employ different physiological and imaging techniques including pach-clamp recordings from different cellular compartments, multiphoton uncaging and Ca2+ imaging.
Neuronal integration and local inhibition:
GABAergic innervation in neurons is supported by an intricate network of GABAergic neurons. Different types of GABAergic neurons target different subcellular compartments of principal neurons. In addition, they are distinct with respect to the temporal dynamics of their activity in-vivo. Thus, different types of GABAergic neurons endow neuronal networks with the capability for GABAergic inhibition of principal neurons with exquisite spatiotemporal control. Inhibition is thought to affect neuronal EPSP-to-spike coupling dependent on the precise spatiotemporal relationship of excitation and inhibition. We are therefore also interested in defining how efficiency of inhibition is affected by this relationship in the dendritic tree. We are also interested in identifying elementary microcircuits underlying inhibition, and in determining the dynamic behaviour of their cellular components.