A dendritic mechanism for balancing synaptic flexibility and stability

Abstract

The modification of synaptic weights is critical for learning, while synaptic stability is required to maintain acquired knowledge. Single neurons have thousands of synapses within their dendritic arbors, and how the weights of specific inputs change across experience is poorly understood. Here we report that dendritic compartments receiving input from different presynaptic populations acquire distinct synaptic plasticity and integration rules across maturation. We find that apical oblique dendrites of layer 5 pyramidal neurons in adult mouse primary visual cortex receive direct monosynaptic projections from the dorsal lateral geniculate nucleus (dLGN), linearly integrate input, and lack synaptic potentiation. In contrast, basal dendrites, which do not receive dLGN input, exhibit NMDA receptor (NMDAR)-mediated supralinear integration and synaptic potentiation. Earlier in development, during thalamic input refinement, oblique and basal dendrites exhibited comparable NMDAR-dependent properties. Oblique dendrites gain mature properties with visual experience, and over the course of maturation, spines on oblique dendrites develop higher AMPA/NMDA ratios relative to basal dendrites. Our results demonstrate that cortical neurons possess dendrite-specific integration and plasticity rules that are set by the activity of their inputs. The linear, non-plastic nature of mature synapses on oblique dendrites may stabilize feedforward sensory processing while synaptic weights in other parts of the dendritic tree remain plastic, facilitating robust yet flexible cortical computation in adults.

Patch2MAP combines patch-clamp electrophysiology with super-resolution structural and protein imaging in identified single neurons without genetic modification

“…We used whole cell patch-clamp recording and rapid two-photon glutamate uncaging to measure functional AMPA and NMDA receptor properties [19][20][21][22][23] at individual spiny synapses in human neurons. The response of a given spine to glutamate uncaging relies on the uncaging laser power, as well as the uncaging location relative to the synapse of interest and the local delivery of the caged glutamate compound 24 .…”

“…; https://doi.org/10.1101/2023.03. 20.533452 doi: bioRxiv preprint acquiring human tissue and ensured that tissue was collected. M.T.H.…”

“…; https://doi.org/10.1101/2023.03. 20.533452 doi: bioRxiv preprint critical problem, as it is unclear which cells, which subcellular compartments (i.e. synapses, axons, dendrites), and which proteins produce malfunctions that mediate brain diseases like Alzheimer's 5,6 , autism 7 , or epilepsy 8,9 , preventing development of effective therapies.…”

Silences, spikes and bursts: Three‐part knot of the neural code

“…For example, bursts of high‐frequency or spatio‐temporally co‐ordinated inputs onto basal dendrites produce NMDA spikes, increasing the probability of producing a somatic burst (Polsky et al., 2009). However, similar to cells from CA1, inputs to basal and radial oblique dendrites receive recurrent and thalamic connections (pathway 1 (Yaeger et al., 2022)) whose ability to generate action potentials at a high frequency is counteracted by potent feedforward inhibition (Sermet et al., 2019). Apical inputs from thalamus or other cortical areas (pathway 2) must cross a prohibitively large electrotonic distance to affect the cell body, preventing a simple summation of inputs (Larkum, et al., 1999; Williams & Stuart, 2002; Beaulieu‐Laroche et al., 2018).…”