Shannon Farris

Mitochondrial calcium signaling integrates energy needs with energy production, amplifying or suppressing mitochondrial respiration in response to activity demand. Neuronal activity is tightly ATPcoupled to increases in mitochondrial calcium uptake, which stimulate the tricarboxylic acid cycle (TCA) and activate calcium-dependent enzymes important for ATP production via oxidative phosphorylation. The mitochondrial calcium uniporter (MCU) is the predominant source of matrix calcium and is differentially expressed across neuronal cell types, suggesting cell-type-specific differences in the coupling of activity-driven calcium levels and mitochondrial respiration. Here, we investigated whether elevating MCU expression enhances mitochondrial calcium uptake and oxidative phosphorylation in the hippocampus. We report that hippocampal mitochondria overexpressing MCU take up calcium at a faster rate without increased sensitivity to calcium overload. By modeling in vivo supply and demand, we found that hippocampal mitochondria overexpressing MCU are more efficient than control mitochondria at responding to increased bioenergetic demand. These findings reveal a role for MCU in modulating mitochondrial calcium uptake and boosting mitochondrial respiration under increasing demand, which contributes to our understanding of how specific cell types may adapt to different bioenergetic demands.

Mitochondrial morphology varies by neuronal cell type and subcellular compartment; however, the functional significance of these differences is unclear. Hippocampal CA2 neurons are enriched for genes encoding mitochondrial proteins compared to CA1 neurons, suggesting a difference in metabolic demand across hippocampal circuits. However, whether CA2 neuron mitochondria are structurally or functionally distinct to support circuit-specific energy demands is unknown. Here we compared mitochondrial morphology, protein expression, and calcium levels across CA1 and CA2 circuits. We found mitochondria in CA2 dendrites were larger than mitochondria in CA1 dendrites. However, both subregions harbored larger mitochondria in the entorhinal cortex (EC)-contacting distal dendrites compared to CA3-contacting proximal dendrites. Together, these data demonstrate both cell type- and input specific regulation of mitochondrial morphology that likely influences the function of these distinct circuits. To determine whether differences in mitochondrial fission or fusion account for cell and/or layer specific differences in morphology, we immunostained for OPA1 and MFF, which showed a general enrichment in distal dendrites relative to proximal dendrites, and an unexpected increase in CA1 distal dendrites compared to CA2 distal dendrites. To show whether these morphological differences result in functionally distinct mitochondria, we measured mitochondrial calcium levels in live slices. We found a striking enrichment of mitochondrial calcium levels in CA2 distal dendrites relative to proximal dendrites, and this layer-specific effect was significantly different from that in CA1 dendrites at baseline and after activity. Collectively, these findings reveal discrete morphological and functional differences in mitochondria across hippocampal subregions and dendritic layers, which likely confer unique circuit properties and/or vulnerabilities to disease.

RNA localization to neuronal axons and dendrites provides spatiotemporal control over gene expression to support synapse function. Neuronal messenger RNAs (mRNAs) localize as ribonucleoprotein particles (RNPs), commonly known as RNA granules, the composition of which influences when and where proteins are made. High-throughput sequencing has revealed thousands of mRNAs that localize to the hippocampal neuropil. Whether these mRNAs are spatially organized into common RNA granules or distributed as independent mRNAs for proper delivery to synapses is debated. Here, using highly multiplexed single-molecule fluorescence in situ hybridization (HiPlex smFISH) and colocalization analyses, we investigate the subcellular spatial distribution of 15 synaptic neuropil localized mRNAs in the male and female rodent hippocampus. We observed that these mRNAs are present in the neuropil as heterogeneously sized fluorescent puncta with spatial colocalization patterns that generally scale by neuropil mRNA abundance. Indeed, differentially expressed mRNAs across cell types displayed colocalization patterns that scaled by abundance, as did simulations that reproduce cell-specific differences in abundance. Thus, the probability of these mRNAs colocalizing in the neuropil is best explained by stochastic interactions based on abundance, which places constraints on the mechanisms mediating efficient transport to synapses.

Neuronal mitochondria are diverse across cell types and subcellular compartments in order to meet unique energy demands. While mitochondria are essential for synaptic transmission and synaptic plasticity, the mechanisms regulating mitochondria to support normal synapse function are incompletely understood. The mitochondrial calcium uniporter (MCU) is proposed to couple neuronal activity to mitochondrial ATP production, which would allow neurons to rapidly adapt to changing energy demands. MCU is uniquely enriched in hippocampal CA2 distal dendrites compared to proximal dendrites, however, the functional significance of this layer-specific enrichment is not clear. Synapses onto CA2 distal dendrites readily express plasticity, unlike the plasticity-resistant synapses onto CA2 proximal dendrites, but the mechanisms underlying these different plasticity profiles are unknown. Using a CA2-specific MCU knockout (cKO) mouse, we found that MCU deletion impairs plasticity at distal dendrite synapses. However, mitochondria were more fragmented and spine head area was diminished throughout the dendritic layers of MCU cKO mice versus control mice. Fragmented mitochondria might have functional changes, such as altered ATP production, that could explain the structural and functional deficits at cKO synapses. Differences in MCU expression across cell types and circuits might be a general mechanism to tune mitochondrial function to meet distinct synaptic demands.

Background: Folate, an essential vitamin B9, is crucial for diverse biological processes, including neurogenesis. Folic acid (FA) supplementation during pregnancy is a standard practice for preventing neural tube defects (NTDs). However, concerns are growing over the potential risks of excessive maternal FA intake. Objectives/Methods: Here, we employed a mouse model and spatial transcriptomic and single-nucleus multi-omics approaches to investigate the impact of high maternal FA supplementation during the periconceptional period on offspring brain development. Results: Maternal high FA supplementation affected gene pathways linked to neurogenesis and neuronal axon myelination across multiple brain regions, as well as gene expression alterations related to learning and memory in thalamic and ventricular regions. Single-nucleus multi-omics analysis revealed that maturing excitatory neurons in the dentate gyrus (DG) are particularly vulnerable to high maternal FA intake, leading to aberrant gene expressions and chromatin accessibility in pathways governing ribosomal biogenesis critical for synaptic formation. Conclusions: Our findings provide new insights into specific brain regions, cell types, gene expressions and pathways that can be affected by maternal high FA supplementation.

Molecular and functional diversity among synapses is generated, in part, by differential expression of neurotransmitter receptors and their associated protein complexes. N -methyl- D -aspartate receptors (NMDARs) are tetrameric ionotropic glutamate receptors that most often comprise two GluN1 and two GluN2 subunits. NMDARs generate functionally diverse synapses across neuron populations through cell-type-specific expression patterns of GluN2 subunits (GluN2A – 2D), which have vastly different functional properties and distinct downstream signaling. Diverse NMDAR function has also been observed at anatomically distinct inputs to a single neuron population. However, the mechanisms that generate input-specific NMDAR function remain unknown as few studies have investigated subcellular GluN2 subunit localization in native brain tissue. We investigated NMDAR synaptic localization in thalamocortical (TC) neurons expressing all four GluN2 subunits. Utilizing super resolution imaging and knockout-validated antibodies, we revealed subtype- and input-specific GluN2 localization at corticothalamic (CT) versus sensory inputs to TC neurons in 4-week-old male and female C57Bl/6J mice. GluN2B was the most abundant postsynaptic subunit across all glutamatergic synapses followed by GluN2A and GluN2C, and GluN2D was localized to the fewest synapses. GluN2B was preferentially localized to CT synapses over sensory synapses, while GluN2A and GluN2C were more abundant at sensory inputs compared to CT inputs. Furthermore, postsynaptic scaffolding proteins PSD95 and SAP102 were preferentially localized with specific GluN2 subunits, and SAP102 was more abundant at sensory synapses than PSD95. This work indicates that TC neurons exhibit subtype- and input-specific localization of diverse NMDARs and associated scaffolding proteins that likely contribute to functional differences between CT and sensory synapses. HIGHLIGHTS: NMDAR subtypes and synaptic scaffolding proteins show preferential localization at specific inputs to thalamocortical neurons. GluN2B was preferentially localized to corticothalamic synapses, while GluN2A and GluN2C were more abundant at sensory inputs to thalamocortical neurons. Colocalization between synaptic scaffolding proteins with NMDARs was GluN2 subtype-dependent. NMDAR subsynaptic organization in thalamocortical neurons is input- and GluN2-subtype specific.

Localization of mRNAs to dendrites is a fundamental mechanism by which neurons achieve spatiotemporal control of gene expression. Translationally repressed neuronal mRNA transport granules, also referred to as ribonuclear proteins (RNPs), have been shown to be trafficked as single or low copy number RNPs and as larger complexes with multiple copies and/or species of mRNAs. However, there is little evidence of either population in intact neuronal circuits. Using single molecule fluorescence in situ hybridization studies in the dendrites of adult rat and mouse hippocampus, we provide evidence that supports the existence of multi-transcript RNPs with the constituents varying in amounts for each RNA species. By competing-off fluorescently labeled probe with serial increases of unlabeled probe, we detected stepwise decreases in Arc RNP number and fluorescence intensity, suggesting Arc RNAs localize to dendrites in both low- and multiple-copy number RNPs. When probing for multiple mRNAs, we find that localized RNPs are heterogeneous in size and colocalization patterns that vary per RNA. Further, localized RNAs that are targeted by the same trans-acting element (FMRP) display greater levels of colocalization compared to an RNA not targeted by FMRP. Simultaneous visualization and assessment of colocalization using highly multiplexed imaging of a dozen mRNA species targeted by FMRP demonstrates that dendritic RNAs are mostly trafficked as heteromeric cargoes of multiple types of RNAs (at least one or more RNAs). Moreover, the composition of these RNA cargoes correlates with the abundance of the transcripts even after accounting for expression. Collectively, these results suggest that dendritic RNPs are packaged as heterogeneous co-assemblies of different mRNAs and that RNP contents may be driven, at least partially, by highly abundant dendritic RNAs; a model that favors efficiency over fine-tuned control for sustaining long-distance trafficking of thousands of messenger molecules. MAIN POINTS Dendritic RNAs exist as diversely composed RNPs that contain variable quantities of an individual transcript as well as combination of multiple different transcripts. RNAs targeted by the same RBP display a bias to localize together as opposed to being colocalized with a non-target RNA. Dendritically localized RNAs, regardless of their individual abundance, colocalize more with highly abundant RNAs compared to less abundant RNAs, even after correcting for random colocalization based on the abundance of RNAs being compared. There is limited evidence of dendritic RNPs being composed of a single RNA species, suggesting that dendritic RNA cargoes are multimeric complexes in vivo and RNA abundance possibly drives the assembly of heteromeric RNA cargos, facilitating the transport of thousands of transcripts to distal compartments.

Mitochondria are dynamic organelles that are morphologically and functionally diverse across cell types and subcellular compartments in order to meet unique energy demands. Mitochondrial dysfunction has been implicated in a wide variety of neurological disorders, including psychiatric disorders like schizophrenia and bipolar disorder. Despite it being well known that mitochondria are essential for synaptic transmission and synaptic plasticity, the mechanisms regulating mitochondria in support of normal synapse function are incompletely understood. The mitochondrial calcium uniporter (MCU) regulates calcium entry into the mitochondria, which in turn regulates the bioenergetics and distribution of mitochondria to active synapses. Evidence suggests that calcium influx via MCU couples neuronal activity to mitochondrial metabolism and ATP production, which would allow neurons to rapidly adapt to changing energy demands. Intriguingly, MCU is uniquely enriched in hippocampal CA2 distal dendrites relative to neighboring hippocampal CA1 or CA3 distal dendrites, however, the functional significance of this enrichment is not clear. Synapses from the entorhinal cortex layer II (ECII) onto CA2 distal dendrites readily express long term potentiation (LTP), unlike the LTP- resistant synapses from CA3 onto CA2 proximal dendrites, but the mechanisms underlying these different plasticity profiles are unknown. We hypothesized that enrichment of MCU near ECII-CA2 synapses promotes LTP in an otherwise plasticity-restricted cell type. Using a CA2-specific MCU knockout (cKO) mouse, we found that MCU is required for LTP at distal dendrite synapses but does not affect the lack of LTP at proximal dendrite synapses. Loss of LTP at ECII-CA2 synapses correlated with a trend for decreased spine density in CA2 distal dendrites of cKO mice compared to control (CTL) mice, which was predominantly seen in immature spines. Moreover, mitochondria were significantly smaller and more numerous across all dendritic layers of CA2 in cKO mice compared to CTL mice, suggesting an overall increase in mitochondrial fragmentation. Fragmented mitochondria might have functional changes, such as altered ATP production, that might explain a deficit in synaptic plasticity. Collectively, our data reveal that MCU regulates layer-specific forms of plasticity in CA2 dendrites, potentially by maintaining proper mitochondria morphology and distribution within dendrites. Differences in MCU expression across different cell types and circuits might be a general mechanism to tune the sensitivity of mitochondria to cytoplasmic calcium levels to power synaptic plasticity. MAIN TAKE HOME POINTS: The mitochondrial calcium uniporter (MCU) regulates plasticity selectively at synapses in CA2 distal dendrites. The MCU-cKO induced LTP deficit correlates with a trending reduction in spine density in CA2 distal dendrites. Loss of MCU in CA2 results in ultrastructural changes in dendritic mitochondria that suggest an increase in mitochondrial fragmentation. These ultrastructural changes could result in functional consequences, such as decreased ATP production, that could underlie the plasticity deficit. Dendritic mitochondrial fragmentation in MCU cKO occurred throughout the dendritic laminae, suggesting that MCU is dispensable for establishing layer-specific mitochondrial structural diversity.