Active dendrites enable strong but sparse inputs to determine orientation selectivity

Abstract

The dendrites of neocortical pyramidal neurons are excitable. However, it is unknown how synaptic inputs engage nonlinear dendritic mechanisms during sensory processing in vivo, and how they in turn influence action potential output. Here, we provide a quantitative account of the relationship between synaptic inputs, nonlinear dendritic events, and action potential output. We developed a detailed pyramidal neuron model constrained by in vivo dendritic recordings. We drive this model with realistic input patterns constrained by sensory responses measured in vivo and connectivity measured in vitro. We show mechanistically that under realistic conditions, dendritic Na+ and NMDA spikes are the major determinants of neuronal output in vivo. We demonstrate that these dendritic spikes can be triggered by a surprisingly small number of strong synaptic inputs, in some cases even by single synapses. We predict that dendritic excitability allows the 1% strongest synaptic inputs of a neuron to control the tuning of its output. Active dendrites therefore allow smaller subcircuits consisting of only a few strongly connected neurons to achieve selectivity for specific sensory features.

Glutamate indicators with improved activation kinetics and localization for imaging synaptic transmission

“…Lastly, we have now achieved sufficient SNR, kinetics, and spatial specificity to record from populations of synapses while resolving differences of a few milliseconds in the timing of individual inputs. These capabilities will be valuable for studying the synaptic dynamics underlying neuronal computation and learning (9,11,42).…”

High-fidelity dendritic sodium spike generation in human layer 2/3 neocortical pyramidal neurons

“…At these sites, dendritic excitatory input is integrated to form dendritic sodium spikes, which forward propagate to drive neuronal output across a wide range of correlated excitatory input, consistent with computational models which describe the dendritic integrative operations of rodent pyramidal neurons as two-layered neural networks 9, 32, 33 . As neuronal modeling has suggested that dendritic sodium spikes of rodent L2/3 pyramidal neurons are driven by the strongest 1% of excitatory synaptic inputs, and by the activation of spatially clustered excitatory input 28 , it is tempting to speculate that the increased number of dendritic spines in HL2/3 pyramidal neurons 3, 5, 6 provides an increased substrate for such dendritic computations.…”

“…Notably, it has recently been reported that the dendritic integrative operations of human L2/3 (HL2/3) pyramidal neurons are mechanistically and functionally divergent from those of the rodent, with direct recordings demonstrating the initiation and forward propagation of calcium channel-mediated apical dendritic spikes 22 , a dendritic spike generation mechanism not apparent in rodent L2/3 pyramidal neurons 10, 14, 28, 29 . This fundamental species-specific difference in active dendritic integration has been suggested to transform the dendritic computations implemented by HL2/3 pyramidal neurons, allowing the execution of anti-coincidence computations, as calcium-channel mediated dendritic spikes were found to be activated only within a narrow range of dendritic excitatory input 22 .…”

Dissecting cascade computational components in spiking neural networks

“…Recent evidence, using advanced experimental techniques recording the activity of single synapses in vivo , shows that single synapses could be active, while the population of synapses is rather spare [ 40 ]. This indicates that, in terms of spikes of a postsynaptic neuron, only a small subset of synapses actively contribute to the somatic firing at one time, while most of the synapses are silent.…”

“…This indicates that, in terms of spikes of a postsynaptic neuron, only a small subset of synapses actively contribute to the somatic firing at one time, while most of the synapses are silent. Experimental observations and theories utilizing this feature suggest complex scenarios of the interaction of spare synaptic firing and dendritic computation at the single-cell level [ 40 ], and spare neural coding at the level of neural circuits [ 41 ]. The STNMF may have an advantage in utilizing these shreds of evidence for understanding the computational principle of neural coding.…”