Dendritic branch structure compartmentalizes voltage-dependent calcium influx in cortical layer 2/3 pyramidal cells

Back-propagating action potentials (bAPs) regulate synaptic plasticity by evoking voltage-dependent calcium influx throughout dendrites. Attenuation of bAP amplitude in distal dendritic compartments alters plasticity in a location-specific manner by reducing bAP-dependent calcium influx. However, it is not known if neurons exhibit branch-specific variability in bAP-dependent calcium signals, independent of distance-dependent attenuation. Here, we reveal that bAPs fail to evoke calcium influx through voltage-gated calcium channels (VGCCs) in a specific population of dendritic branches in cortical layer 2/3 pyramidal cells, despite evoking substantial VGCC-mediated calcium influx in sister branches. These branches contain VGCCs and successfully propagate bAPs in the absence of synaptic input; nevertheless, they fail to exhibit bAP-evoked calcium influx due to a branch-specific reduction in bAP amplitude. We demonstrate that these branches have more elaborate branch structure compared to sister branches, which causes a local reduction in electrotonic impedance and bAP amplitude. Finally, we show that bAPs still amplify synaptically-mediated calcium influx in these branches because of differences in the voltage-dependence and kinetics of VGCCs and NMDA-type glutamate receptors. Branch-specific compartmentalization of bAP-dependent calcium signals may provide a mechanism for neurons to diversify synaptic tuning across the dendritic tree.

Properties and computational consequences of fast dendritic spikes during natural behavior

Dendritic membrane potential was recently measured for the first time in drug-free, naturally behaving rats over several days. These showed that neuronal dendrites generate a lot of sodium spikes, up to ten times as many as the somatic spikes. These key experimental findings are reviewed here, along with a discussion of computational models, and computational consequences of such intense spike traffic in dendrites. We overview the experimental techniques that enabled these measurements as well as a variety of models, ranging from conceptual models to detailed biophysical models. The biophysical models suggest that the intense dendritic spiking activity can arise from the biophysical properties of the dendritic voltage-dependent and synaptic ion channels, and delineate some computational consequences of fast dendritic spike activity. One remarkable aspect is that in the model, with fast dendritic spikes, the efficacy of synaptic strength in terms of driving the somatic activity is much less dependent on the position of the synapse in dendrites. This property suggests that fast dendritic spikes is a way to confer to neurons the possibility to grow complex dendritic trees with little computational loss for the distal most synapses, and thus form very complex networks with high density of connections, such as typically in the human brain. Another important consequence is that dendritically localized spikes can allow simultaneous but different computations on different dendritic branches, thereby greatly increasing the computational capacity and complexity of neuronal networks.

Co-allocation to overlapping dendritic branches in the retrosplenial cortex integrates contextual memories across time

Events occurring close in time are often linked in memory, providing an episodic timeline and a framework for those memories. Recent studies suggest that memories acquired close in time are encoded by overlapping neuronal ensembles, and that this overlap is necessary for memory linking. Transient increases in neuronal excitability drive this ensemble overlap, but whether dendritic plasticity plays a role in linking memories is unknown. Here, we show that contextual memory linking is not only dependent on ensemble overlap in the retrosplenial cortex (RSC), but also on RSC branch-specific dendritic allocation mechanisms. Using longitudinal two-photon calcium imaging of RSC dendrites, we show that the same dendritic segments are preferentially activated by two linked (but not independent) contextual memories, and that spine clusters added after each of two linked (but not independent) contextual memories are allocated to the same dendritic segments. Importantly, with a novel optogenetic tool, selectively targeted to activated dendritic segments following learning, we show that reactivation of dendrites tagged during the first context exploration is sufficient to link two contextual memories. These results demonstrate a causal role for dendritic mechanisms in memory linking and reveal a novel set of rules that govern how linked, and independent memories are allocated to dendritic compartments.

Controlled Anchoring of (Phenylureido)sulfonamide-Based Receptor Moieties: An Impact of Binding Site Multiplication on Complexation Properties

The repetition of urea-based binding units within the receptor structure does not only lead to monomer properties multiplication. As confirmed by spectroscopic studies, UV-Vis and 1H-NMR in classical or competitive titration mode, the attachment to a carrier allocates the active moieties to mutual positions predetermining the function of the whole receptor molecule. Bivalent receptors form self-aggregates. Dendritic receptors with low dihydrogen phosphate loadings offer a cooperative complexation mode associated with a positive dendritic effect. In higher dihydrogen phosphate concentrations, the dendritic branches act independently and the binding mode changes to 1:1 anion: complexation site. Despite the anchoring, the dendritic receptors retain the superior efficiency and selectivity of a monomer, paving the way to recyclable receptors, desirable for economic and ecological reasons.

Any neuron can perform linearly non-separable computations

Multiple studies have shown how dendrites enable some neurons to perform linearly non-separable computations. These works focus on cells with an extended dendritic arbor where voltage can vary independently, turning dendritic branches into local non-linear subunits. However, these studies leave a large fraction of the nervous system unexplored. Many neurons, e.g. granule cells, have modest dendritic trees and are electrically compact. It is impossible to decompose them into multiple independent subunits. Here, we upgraded the integrate and fire neuron to account for saturating dendrites. This artificial neuron has a unique membrane voltage and can be seen as a single layer. We present a class of linearly non-separable computations and how our neuron can perform them. We thus demonstrate that even a single layer neuron with dendrites has more computational capacity than without. Because any neuron has one or more layer, and all dendrites do saturate, we show that any dendrited neuron can implement linearly non-separable computations.

An Autism-Associated de novo Mutation in GluN2B Destabilizes Growing Dendrites by Promoting Retraction and Pruning

Mutations in GRIN2B, which encodes the GluN2B subunit of NMDA receptors, lead to autism spectrum disorders (ASD), but the pathophysiological mechanisms remain unclear. Recently, we showed that a GluN2B variant that is associated with severe ASD (GluN2B724t) impairs dendrite morphogenesis. To determine which aspects of dendrite growth are affected by GluN2B724t, we investigated the dynamics of dendrite growth and branching in rat neocortical neurons using time-lapse imaging. GluN2B724t expression shifted branch motility toward retraction and away from extension. GluN2B724t and wild-type neurons formed new branches at similar rates, but mutant neurons exhibited increased pruning of dendritic branches. The observed changes in dynamics resulted in nearly complete elimination of the net expansion of arbor size and complexity that is normally observed during this developmental period. These data demonstrate that ASD-associated mutant GluN2B interferes with dendrite morphogenesis by reducing rates of outgrowth while promoting retraction and subsequent pruning. Because mutant dendrites remain motile and capable of growth, it is possible that reducing pruning or promoting dendrite stabilization could overcome dendrite arbor defects associated with GRIN2B mutations.

Multibranch Formal Neuron: An Internally Nonlinear Learning Unit

The transformation of synaptic input into action potential in nerve cells is strongly influenced by the morphology of the dendritic arbor as well as the synaptic efficacy map. The multiplicity of dendritic branches strikingly enables a single cell to act as a highly nonlinear processing element. Studies have also found functional synaptic clustering whereby synapses that encode a common sensory feature are spatially clustered together on the branches. Motivated by these findings, here we introduce a multibranch formal model of the neuron that can integrate synaptic inputs nonlinearly through collective action of its dendritic branches and yields synaptic clustering. An analysis in support of its use as a computational building block is offered. Also offered is an accompanying gradient descent–based learning algorithm. The model unit spans a wide spectrum of nonlinearities, including the parity problem, and can outperform the multilayer perceptron in generalizing to unseen data. The occurrence of synaptic clustering boosts the generalization efficiency of the unit, which may also be the answer for the puzzling ubiquity of synaptic clustering in the real neurons. Our theoretical analysis is backed up by simulations. The study could pave the way to new artificial neural networks.

Any neuron can perform linearly non-separable computations

Multiple studies have shown how dendrites enable some neurons to perform linearly non-separable computations. These works focus on cells with an extended dendritic arbor where voltage can vary independently, turning dendritic branches into local non-linear subunits. However, these studies leave a large fraction of the nervous system unexplored. Many neurons, e.g. granule cells, have modest dendritic trees and are electrically compact. It is impossible to decompose them into multiple independent subunits. Here, we upgraded the integrate and fire neuron to account for saturating dendrites. This artificial neuron has a unique membrane voltage and can be seen as a single layer. We present a class of linearly non-separable computations and how our neuron can perform them. We thus demonstrate that even a single layer neuron with dendrites has more computational capacity than without. Because any neuron has one or more layer, and all dendrites do saturate, we show that any dendrited neuron can implement linearly non-separable computations.

Sleep deprivation reduces the density of individual spine subtypes in a branch-specific fashion in CA1 neurons

Sleep deprivation has a negative impact on hippocampus-dependent memory, which are thought to depend on cellular plasticity. We previously found that five hours of sleep deprivation robustly decreases dendritic spine density in the CA1 area of the hippocampus in adult male mice. However, recent work by others suggests that sleep deprivation increases the density of certain spine types on specific dendritic branches. Based on these recent findings and our previous work, we conducted a more in-depth analysis of different spine types on branches 1, 2 and 5 of both apical and basal dendrites to assess whether five hours of sleep deprivation may have previously unrecognized spine-type and branch-specific effects. This analysis shows no spine-type specific changes on branch 1 and 2 of apical dendrites after sleep deprivation. In contrast, sleep deprivation decreases the number of mushroom and branched spines on branch 5. Likewise, sleep deprivation reduces thin, mushroom, and filopodia spine density on branch 5 of the basal dendrites, without affecting spines on branch 1 and 2. Our findings indicate that sleep deprivation leads to local branch-specific reduction in the density of individual spine types, and that local effects might not reflect the overall impact of sleep deprivation on CA1 structural plasticity. Moreover, our analysis underscores that focusing on a subset of dendritic branches may lead to potential misinterpretation of the overall impact of in this case sleep deprivation on structural plasticity.