Back

Synaptic Input Triggers On-Demand Spine-Specific Mitochondrial ATP Production and Delivery

Paquin-Lefebvre, F.; Kushnireva, L.; Xu, Z.; Kubler, S.; Feofilaktova, T.; Laughlin, S.; Rouach, N.; Korkotian, E.; Holcman, D.

2026-03-11 neuroscience
10.64898/2026.03.09.708229 bioRxiv
Show abstract

Synaptic activity imposes acute energy demands, especially for restoring ionic gradients via pumps and exchangers that require ATP. While mitochondria are positioned near dendritic spines to meet this demand, how ATP is produced and delivered with spatial precision remains unclear. Here, using high-resolution calcium and ATP imaging, immuno-cytochemistry, and computational modeling, we demonstrate that synaptic input--but not back-propagating action potentials (bAPs)--triggers on-demand mitochondrial ATP production. This occurs only in spines containing a spine apparatus (SA), where calcium-induced calcium release (CICR) activates mitochondrial calcium uniporters (MCUs), initiating ATP synthesis. We show that ATP delivery is spatially constrained to mitochondrial regions facing the spine base, where ATP-synthase is enriched. Importantly, ATP produced elsewhere on the mitochondrial surface tends to diffuse into the dendrite. We further demonstrate that the delivery of ATP to the spine head is geometrically optimized: an intermediate spine neck length maximizes delivery efficiency. Mathematical modeling and simulations revealb that the time scale for ATP to reach and refill all head-localized exchangers is on the order of hundreds of milliseconds--fast enough to meet local metabolic needs. The present findings establish a mechanism in which nanoscale calcium signaling and mitochondrial architecture together ensure rapid, spatially targeted ATP delivery, tightly coupled to synaptic activity.

Matching journals

The top 4 journals account for 50% of the predicted probability mass.

50% of probability mass above

"Similar papers" are the closest papers from that journal in the model's embedding space. They show what the match is built on, but the ranking comes mostly from a classifier over the whole training set, not from these examples alone.