Asymmetric voltage attenuation in dendrites can enable hierarchical heterosynaptic plasticity

Abstract

Long-term synaptic plasticity has been shown to be mediated via calcium concentration ([Ca2+]). Using a synaptic model which implements calcium-based long-term plasticity via two sources of Ca2+, NMDA receptors and voltage-gated calcium channels (VGCCs), we show in dendritic cable simulations that the interplay between these two calcium sources can result in a diverse array of heterosynaptic effects. When spatially clustered synaptic input produces an NMDA spike, the resulting dendritic depolarization can activate VGCCs at other spines, resulting in heterosynaptic plasticity. Importantly, NMDA spike activation at a given dendritic location will tend to depolarize dendritic regions that are located distally to the input site more than dendritic sites that are proximal to it. This dendritic asymmetry results in a hierarchical heterosynaptic plasticity effect in branching dendrites, where clustered inputs to a proximal branch induce heterosynaptic plasticity primarily at branches that are distal to it. We also explore how simultaneously activated synaptic clusters located at different dendritic locations synergistically affect each other as well as the heterosynaptic plasticity of an inactive synapse "sandwiched" between them. We conclude that dendrites enable a sophisticated form of plastic supervision wherein NMDA spike induction can be spatially targeted to produce plasticity at specific dendritic regions.

Ca2+-dependent phosphodiesterase 1 regulates the plasticity of striatal spiny projection neuron glutamatergic synapses

“…To provide more detail than the heat map of dendritic depolarization (Figure 6A), a plot of peak membrane potential as a function of distance from the soma was generated; it shows the strong attenuation with distance from the dendritic spike activation site (Figure 6C). However, in agreement with previous work, 66,69 distal to the site of spike generation, the peak membrane potential did not fall off significantly with distance (Figure 6C). To link this voltage measurement to the opening of CaV1.3 Ca 2+ channels (and PDE1 activation), both a dendritic heat map and a detailed distance plot of CaV1.3 channel conductance was constructed (Figures 6D-6F).…”

BTSP, not STDP, Drives Shifts in Hippocampal Representations During Familiarization

“…However, synaptic normalization is not realistic and induces some limitations in our model (see Methods). Therefore, to determine whether heterosynaptic or purely homosynaptic processes support the bidirectional changes observed in BTSP-induction experiments, future comparisons between the two classes of models should implement more realistic fast heterosynaptic rules (Abraham, 2008;Chistiakova et al, 2015;Ebner et al, 2019;Moldwin et al, 2023;Triesch et al, 2018). Additionally, experiments with longer tracks and varying speeds will be required to rigorously test each model's predictions on the effect of BTSP-triggering events occurring more than 5s away from the initial PF.…”

The Calcitron: A Simple Neuron Model That Implements Many Learning Rules via the Calcium Control Hypothesis

“…Another important feature of biological neurons not incorporated into the calcitron is the spatial distribution of synapses on a neuron’s dendrites. This is especially relevant for heterosynaptic plasticity, which depends on the location of synapses relative to each other as well as their absolute location on the dendritic tree (Chater & Goda, 2021; Chistiakova et al, 2014; Moldwin et al, 2022; Tong et al, 2021). The calcitron can be augmented to include location-dependent heterosynaptic plasticity on a single dendrite following the schemes of the clusteron (Mel, 1991) or the G-clusteron (Moldwin et al, 2021) or to explicitly include a branching dendrite to account for the branch-dependent hierarchical heterosynaptic plasticity effect we posited in previous work (Moldwin et al, 2022).…”

“…This is especially relevant for heterosynaptic plasticity, which depends on the location of synapses relative to each other as well as their absolute location on the dendritic tree (Chater & Goda, 2021; Chistiakova et al, 2014; Moldwin et al, 2022; Tong et al, 2021). The calcitron can be augmented to include location-dependent heterosynaptic plasticity on a single dendrite following the schemes of the clusteron (Mel, 1991) or the G-clusteron (Moldwin et al, 2021) or to explicitly include a branching dendrite to account for the branch-dependent hierarchical heterosynaptic plasticity effect we posited in previous work (Moldwin et al, 2022). The spatial structure of the dendrite can also influence homeostatic plasticity (Rabinowitch & Segev, 2006a, 2006b) as well as how inhibitory inputs affect plasticity at excitatory synapses (Bar-Ilan et al, 2012).…”