HHMI-NIH Research Scholars
Learn about the HHMI-NIH Research Scholars Program, also known as the Cloister Program. More

Janelia Farm Research Campus
Learn about the new HHMI research campus located in Virginia. More

Summary: Gábor Tamás's HHMI-supported research focuses on the function of electrophysiologically and anatomically identified molecules, subcellular compartments, and small neuronal networks.

We have completed a study spanning several years on the function of chandelier cells, the most specific element in the cortical circuit. Chandelier cells release the dominant inhibitory neurotransmitter gamma-aminobutyric acid (GABA). The postsynaptic side of this synapse, the proximal part of the axon, plays a prominent role in neuronal function. This cellular compartment is thought to be crucial for the initiation of action potentials, presumably due to the high local density of Na⁺ channels in this region. Thus, for many years, such axo-axonic cells (AACs) were considered inhibitory neurons with strategically placed synapses controlling the generation of action potentials.

Our recent results show that, instead of exclusively inhibiting the axon initial segment, chandelier cells can also act as uniquely powerful excitatory neurons in the cortex. So we asked ourselves what underlies such an unorthodox function. The transmitter released synaptically by chandelier cells acts on postsynaptic GABA_A receptors permeable to Cl⁻ and, to a lesser extent, HCO₃⁻ ions. In the mature cerebral cortex, GABA-mediated inhibitory responses depend on the potassium chloride cotransporter 2 (KCC2), which dominates Cl⁻ extrusion from neurons and reduces the intracellular concentration of Cl⁻. Normally, activation of GABA_A receptors increases membrane permeability to Cl⁻, which causes a net flow of Cl⁻ into the cell, resulting in hyperpolarizing responses. To characterize the function of AACs in cortical microcircuits, we investigated the subcellular distribution of KCC2 in pyramidal cells. The density of KCC2 was about 44 times higher in somatic membranes than in the axon initial segment. The polarity of GABAergic inputs in different cell regions was in line with the local availability of Cl⁻ extrusion. Measured at the resting membrane potential of postsynaptic pyramidal cells, GABAergic synapses targeting the somatic region were hyperpolarizing or inhibitory, but synapses onto the axon initial segment were depolarizing or excitatory.

Usually, excitatory inputs arriving at pyramidal cells are glutamatergic and target dendritic spines and shafts. The efficacy of these depolarizing synapses in triggering action potentials in the axon of postsynaptic cells is relatively limited: synchronous activation of several presynaptic glutamatergic cells is required to fire postsynaptic pyramidal cells. However, synapses from chandelier cells are the closest to the first axonal branch point, which is thought to be the site of action potential initiation, thus favoring extreme spike-triggering effectiveness of local depolarizing inputs. Our results suggest that a single spike in a presynaptic chandelier cell can elicit a postsynaptic action potential in pyramidal cells. Thus, it seems that the most powerful excitatory synapses onto cortical pyramidal neurons are not glutamatergic but use only the classic inhibitory transmitter GABA to achieve spike-to-spike coupling.

The series of events initiated by chandelier cells in the cortical network does not end with second-order spikes in the postsynaptic pyramidal cells. Network activation spreads downstream and triggers third-order action potentials in neighboring interneurons and possibly in a subset of excitatory neurons. As mentioned, chandelier cells do not target GABAergic neurons; therefore, the sequence of activation in such event sequences is highly repeatable. This suggests that chandelier cells are in a crucial position to activate functionally linked assemblies of neurons that are thought to be the building blocks of complex cortical representations and cognitive processes.

A second set of experiments addressed fundamental mechanisms underlying neural signal processing. The distribution of calcium in dendritic compartments of neurons is thought to be crucial for regulating postsynaptic function. The compartmentalized distribution of calcium offers the basis for selective regulation of single synapses and synaptic plasticity. Several mechanisms are known to contribute to the clearance of calcium from neurons, but the distinct contributions of various extrusion mechanisms to calcium compartmentalization in dendritic spines versus shafts remain elusive. We used a combination of methods for the first time to explore the contribution of a Na⁺/Ca²⁺ exchanger (NCX) molecule to calcium clearance across the plasma membrane in dendrites. We combined high-resolution, ultrastructural immunocytochemistry with two-photon imaging to assess the subcellular distribution and role of NCX1 in rat CA1 pyramidal cells. NCX1 is about seven times more densely distributed in dendritic shafts than in dendritic spines. In line with these results, two-photon imaging of synaptically activated Ca²⁺ transients during NCX blockade shows that NCXs act preferentially in the dendritic shafts to regulate spine-dendrite coupling. Our results suggest a subcellular compartment-dependent distribution of NCX1 in the plasma membrane of CA1 pyramidal cells. Moreover, the distribution did not depend on the distance from the soma; our results thus also indicate an excitatory input-independent distribution of NCX1. Accordingly, the distribution is markedly different from what was found earlier for AMPA receptors and hyperpolarization-activated channels. Furthermore, NCX1 does not play a role in distance-dependent scaling of excitatory inputs.

Last updated September 2008