Humans and other primates can reliably distinguish between actions performed by themselves or someone else, even when the actor’s identity is stripped of all visual cues. Interestingly, the effect cannot purely be attributed to visual familiarity as it dissociates from recognizing familiar others from strangers, and we rarely see ourselves performing actions in third person. One hypothesis is that action-related neural systems may drive the ability to recognize one’s own actions, by being engaged more for one’s own than someone else’s actions. Here we aim to explore whether network-based functional connectivity during task or at rest is linked to performance on an action based self-recognition task. Further, we investigate whether connectivity predictive of task performance is localized within areas with motor properties that are also active during action observation. We propose two empirically testable models in an attempt to explain how self-action recognition works computationally. The first is a biologically plausible artificial neural network model which can perform an abstraction of the task, learning to discriminate between actions performed by themselves or someone else based on motor and visual cues, and allowing for us to examine the cognitive architecture of self-action recognition. This is complemented by the second data driven empirically estimated neural network model, based on resting- and task-based fMRI data, which aims to show how self-action recognition is implemented neurally through predicting and classifying empirically observed BOLD activations in cortical hubs, such as the IPL or IFG, based on neural activity from other parts of the cortex.

Microglial cells are resident immune cells in the central nervous system (CNS) that help with immunity and shaping neural connection like synaptic pruning. Microglia make contacts with both presynaptic and postsynaptic neurons to sense their synaptic activities, and prune underused synaptic connections. Although microglial cells are known to be involved in neural function modulations, the temporal and spatial information about microglia-neuron interactions in memory processing still remain largely unknown. The purpose of this study is to reveal interactions between microglial cells and neurons by directly visualizing their connections through an art tool called enhanced GFP Reconstitution Across Synaptic Partners (eGRASP) during memory linking processes. This technique relies on complementary fragments of two non-fluorescent GFP fragments which fluorescent when connected. According to this method, an eGRASP system which labels microglia-neuron interaction with yellow fluorescence was designed. The purpose of this study is to validate the performance of this system in vitro by using HEK-293 cells. DNA constructs from the eGRASP system were transfected into different groups of HEK cells, then they were co-cultured to form cell-cell interactions. Our results demonstrated that yellow fluorescence successfully revealed the contact interface between HEK-293 cells, indicating that our plasmids can be used to reflect cell-cell interactions. In future experiments, this eGRASP system will be packed into AAVs and applied to the mouse hippocampus to reveal microglial-neuron interactions in vivo during learning and memory processes, which might contribute to therapies of brain diseases.

Heteroplasmy is the presence of both mutated and wild-type (WT) mitochondrial DNA (mtDNA) in a cell. Increased percentages of mutated mtDNA correspond to more severe states of heteroplasmy. If these percentages exceed the biological threshold for cell survival, severe disease pathology develops, as in Parkinson’s and Alzheimer’s disease. Humans have an innate mtDNA damage control system that prevents severe heteroplasmy where mutated mtDNA is degraded and released into the cytoplasm. Reversing the development of severe heteroplasmy could reverse severe disease pathophysiology. The Guo lab aims to establish a mouse myoblast cell culture model that can be used in high-throughput screening for drugs that reduce severe heteroplasmy. The Guo lab previously created myoblast cell lines that conditionally express restriction enzymes and ligases to create deletions in mtDNA (mtDNAΔ) that simulate mutated mtDNA. The model utilizes MGME shRNA to prevent the degradation of a linear mtDNAΔ intermediate. I aim to further increase detectable levels of mtDNAΔ and raise baseline heteroplasmy severity in this model by preventing the release of this intermediate. The developed cell lines were treated with reagents (FEN1-IN-1, Visomitin, TH10785, FEN1-IN-4, or VBIT-4) that inhibit the release of mtDNAΔ from the mitochondria. The amount of WT mtDNA and mtDNAΔ were evaluated using PCR. FEN1-IN-4 increased mtDNAΔ in cells when compared with the control group, suggesting that FEN1 inhibition preserves damaged mtDNA in mitochondria. No other reagents tested had an effect on mtDNAΔ concentration. Future steps include incorporating FEN1 inhibition into the model and determining if cells contain sufficient measurable mtDNAΔ.