In one mouse visual cortex volume, mitochondria of layer 2/3 pyramidal cells followed dendrite caliber, synapse density, depth and a stable cell trait; in vivo activity explained little
Question: do layer 2/3 pyramidal cells that are more active in vivo have more mitochondria? Data: one public mouse V1 EM volume (MICrONS Layer 2/3, pinky100 v185; Turner et al. 2022) with 2.41 million mitochondria, skeletons of 363 pyramidal cells and two-photon calcium traces of 127 of them. Method: we cut dendrites into 10 µm segments and modelled log mitochondria volume per µm from diameter, input synapse density and path distance. A cell's 'excess' is the mean residual of its segments. We related excess to the inferred calcium event rate with Spearman partial correlations, permutations and bootstraps over cells. Two tests were preregistered. Results: in 37,463 basal segments, mitochondria per µm rose with diameter (log-log slope 0.95) and by 24% per added synapse per µm (R² 0.37). The excess was a stable whole-cell trait: split-half r 0.83; the same cells had more mitochondria in apical dendrites (ρ 0.61), soma (0.41) and axon (0.43); deeper cells had more (0.48). Within imaging scans, activity followed the excess only weakly: ρ +0.19 (95% CI 0.00 to 0.35), positive in all 5 scans, pooled +0.20 (0.01 to 0.37); 1 SD of activity matched 1.8% more mitochondria while cells differed by 11.6% (SD). With segment position in the model, ρ was +0.14 (-0.03 to 0.32). The preregistered basal test gave partial ρ +0.24 (Holm p 0.014) but needed caliber and scan covariates (raw ρ -0.04); somatic mitochondria per nucleus volume did not follow activity (ρ +0.01). By contrast, fast-spiking basket and chandelier cells had 2.3-5.1 times the perisomatic mitochondria per nucleus volume of pyramidal cells. Limitations: one animal and 127 imaged cells; calcium rates are in noise units, are comparable only within a scan, and are not firing rates; activity was recorded before fixation; skeleton diameters are noisy; the segment excess and within-scan tests were not preregistered. The package has code, derived tables, figures and a preregistration file.
mitochondria · connectomics · MICrONS · electron microscopy · calcium imaging · structure and function · pyramidal cell · layer 2/3 · mouse visual cortex · dendrite caliber · synapse density · energy budget · basket cell · chandelier cell · cortical depth · preregistration
References
9- usesdoi:10.5281/zenodo.7510511
MICrONS Layer 2/3 data tables v185.3 (CC BY 4.0): mitochondria, skeletons, mito-to-skeleton map, received synapses, calcium traces, cell types, nuclei.
- usesdoi:10.1016/j.cell.2022.01.023
Source paper of the volume, the compartment labels, the functional vignette tables and the in-connection densities.
- doi:10.1097/00004647-200110000-00001
Energy budget of grey matter that motivates the activity-mitochondria hypothesis.
- doi:10.1016/j.neuron.2012.08.019
Synaptic energy use and supply; agrees with the synapse-density term of the segment model.
- supportsdoi:10.1111/j.1460-9568.2006.04814.x
High cytochrome c in fast-spiking interneurons; our EM perisomatic mitochondria of basket and chandelier cells agree.
- supportsdoi:10.1093/brain/awq333
High mitochondrial complex I in fast-spiking interneurons; our cell-type result agrees.
- doi:10.1101/2024.04.21.590464
Hemibrain mitochondria by cell type and transmitter in the fly; related connectome mitochondria analysis.
- doi:10.1101/2024.08.21.608912
Fly mitochondria organization related to visual responses of projection neurons; closest prior structure-function test.
- In released ferret spine records, tuning diversity had rank correlations of 0.56 with positive 1 µm mitochondrial volume and −0.003 with entry status sci:sha256:c5ef19522c7b1ab8707ddc17db1a0217993fc5c78cc3609265e3645753eb8777
Related SCI work on local mitochondria and spine tuning diversity in ferret visual cortex (different data and question).
Cited by
2 publicationsPublic catalog entries on Sciscape.
- In one fly connectome, pooled mitochondrial shape models gave 68–69% balanced recall of three predicted transmitter labels in unseen cell types
Related work connects mitochondrial content to cell traits and geometry in mouse cortex. It motivated anatomy controls; no code or data were reused.
- In one mouse visual cortex EM volume, neurons sat at chance distances from vessels and vessels did not follow synapse density; glia sat nearer, and vessels above neurons lowered imaging scores
Related MICrONS energy-supply study (mitochondria vs in vivo activity); a weak activity link there agrees with our weak function link.