Specific Aim 1 (SA1): Assessment of LPS injury in microglial monoculture. 

Study 1: Induce LPS injury and assess using imaging with Nile Blue and Lysotracker Green

Excess lipopolysaccharide (LPS) can be added to the microglia monoculture to trigger oxidative stress. To assess the cultures, we will image the cells post-injury with multimodal TPEF and then stain the lipofuscin, a product of oxidative stress, in the cytoplasm of the cells with Nile Blue and Lysotracker Green and image it using Confocal. Imaging will be done after 24 hours of injury induction. 

The success measure is to validate that the optical readouts trend (i.e. redox ratio) is consistent with glycolysis, as it is well-known that oxidative stress causes a shift to glycolytic metabolism.

Specific Aim 2 (SA2):  Assessment of glutamate injury in microglial monoculture.

Study 1: Induce glutamate injury and assess using imaging and ROS assay. 

We will test glutamate concentration at a short exposure (30 minutes) as well as a medium (6h) exposure and a long exposure (24 hrs). The glutamate concentrations range between 10 and 1000 uM, inversely proportional to incubation time. This is according to relevant papers, which are summarized in the Appendix. 

We will replace the cell media with a magnesium-free minimal medium for one day prior to injury because magnesium can occupy AMPA receptors and prevent glutamate activation. After glutamate exposure, we will rinse scaffolds with minimal medium and replace it with the normal neurobasal medium. We will image three regions for each of the three wells per experimental condition, including vehicles (media only and media + ROS dye), positive control (media + ROS TBHP), and untreated.

For a success measure, we will verify optical readouts with fluorescence intensity of ROS via the ROS assay kit and the microplate reader, as glutamate excitotoxicity triggers downstream activation of ROS production. 

Specific Aim 3 (SA3): Develop a computational metabolic model that predicts injury pathway activation in microglia based on biochemical readouts. 

Study 1: Develop a basic computational metabolic model for TBI cultures 

We will obtain relevant brain metabolism computational models from the literature (microglia metabolism, oxidative stress models, injury models) and modify them by adjusting concentration conditions. In the case where there are insufficient relevant computational models specific to microglia, we will investigate papers detailing the secondary injury pathways of microglia in human or animal brain tissue. Molecules involved in the model but not present in our spectrometric results will be treated as assumed constants based on literature values. The completed model should contain central metabolism, detoxification of reactive oxygen species, and the glutamate-glutamine cycle. Within the BME8 timespan, the computational model should be able to at least follow key pathways such as glycolysis and the TCA cycle, and take into account varying initial levels of LPS. It will be able to predict the relative level of pathway activation (ex. glycolytic vs. oxidative metabolism) based on the input concentrations of downstream metabolites obtained from mass spectrometry. Mass spectrometry will not be completed in BME8 but for future work.

Study 2: Use the metabolic model to predict injury pathway activation 

From SA 1 and 2, we will have imaging data from secondary injury at multiple time points and corresponding assays data from the final time point. We can use the metabolite concentrations and the metabolic model to predict levels of pathway activation under injury conditions, and then correlate those pathway activations with the optical readouts from those same conditions. While the metabolic model does not output pathway activations, it predicts concentrations of upstream effectors that would cause observed downstream metabolite concentrations. By associating upstream effectors with particular injury pathways, we can estimate the pathway activations under different injury conditions.

The most viable model would be a kinetic/stoichiometric model similar to a michaelis-menten model. Due to limitations for experimental results, the goal BME8 is to create a framework for microglia first instead of neurons and astrocytes. 

Currently, there are no models that precisely cover our specific aims. However, the papers have a detailed outline of the differential equation sets that they used to derive their model, which we can replicate and modify. Our goal is to develop our own models for brain metabolite analysis using the existing models and mathematical equations.