Transcription factors such as TCF-1, as well as inflammatory cytokines such as IL-12, have been shown to play a role in cytotoxic (CD8) T cell differentiation during an immune response. Here, we build a stochastic computational model of the canonical CD8 T cell immune response encompassing experimentally hypothesized transcription factor and inflammation control mechanisms. We hypothesize that the transcriptional factor control of CD8 T cells can be quantified into concrete model parameters that control various aspects of the immune response such as cell effector cell expansion, proliferation, and memory cell formation in various immune challenge conditions, such as inflammation. With a fully developed model, effects of inflammation levels and gene regulation kinetics on CD8 T cell differentiation can be measured in situ, and new hypotheses surrounding cellular differentiation kinetics can be generated. Understanding the factors that control CD8 T cell differentiation is crucial in the development of vaccines and immunotherapies, such as CAR T cell therapy.

Fluorescent biosensors are a vital tool in the goal to decipher the complexity of neural networks. Genetically encoded fluorescent indicators (GEFIs) are protein-based sensors that increase in fluorescence upon ligand binding and allow for passive monitoring of neuronal signals. However, the development of such sensors is limited by the slow throughput of traditional protein engineering which has long engineering cycles of new plasmid variants. Our project aims to tackle this problem by developing a high-throughput sensor engineering platform that can effectively generate and screen unbiased genetic libraries of GEFIs in mammalian cells. Our platform can identify high performing sensor variants on a custom microarray and effectively isolate and recover their genetic material. Our new platform will be used to develop a sensor for the μ-Opioid receptor (MOR), which is a G-protein coupled receptor that is involved in opioid addiction. Our experiments have already developed a MOR sensor that surpasses the standard in the literature and we will continue to optimize it for maximum spatial and temporal precision. The development of a MOR sensor through this iterative process allows researchers to further investigate the molecular mechanisms underlying the pathology of addiction and provides a novel platform for protein engineers to more efficiently develop a wide variety of biosensors.

Calcific Aortic Valve Disease (CAVD) is progressive osteogenic changes including calcium deposition and thickening of the aortic valve causing stiffness which impairs normal function ultimately leading to ventricular hypertrophy and death. Runx2, a transcription factor involved in osteogenic changes, has been observed to be upregulated in the diseased valves. RNA sequencing data from the lab had previously revealed several genes regulated in a Runx2 knock out mouse model of CAVD. This study aims to determine if NKX2.5, one of the genes identified in the RNA sequencing data, is an upstream regulator of Runx2. Baboon valve interstitial cells (BVICs), which can be induced to calcify and share characteristics similar to human valve cells, were used in CRISPR knockout experiments to determine molecular and biochemical changes following NKX2.5 deletion. CRISPR-Cas9 is an effective method of gene deletion that uses programmable guide RNA to induce a gene mutation, inactivating its protein product. Initial experiments were aimed at optimizing the conditions for electroporation and gRNA delivery to ensure BVIC survival and efficient gene deletion in the cell line. Gene deletion success was determined by sequencing a ~1kb region around the deletion site. After generating BVIC lacking NKX2.5, cells were cultured in media known to induce calcification. A biochemical assay indicated changes in activation. Cells were assessed for calcification changes using biochemical assay, changes in activation, and Runx2 expression determined via RT-qPCR. We expect to confirm that NKX2.5 negatively regulates Runx2 and in turn calcification.

Non-human primate (NHP) research has become an essential step in the translation of medical technologies from animal models to clinical trials. This is especially so in neural research, as there is a large discrepancy between rodent and human brains in both anatomy and size. For some techniques such as optogenetics, which requires viral transduction of neurons, traditional diffusion-based viral injection approaches are effective in rodent brains but are impractical for large NHP ones. Convection-enhanced delivery (CED), a large-scale injection approach, currently lacks a practical quantitative bench-side injection modeling method to guide neurosurgical preparation. We aim to develop a gel model of the NHP brain and replicate surgical injections of it in order to reduce the risks of directly injecting into a NHP without sufficient preparation. We are testing the validity of our model by monitoring the spread of the injection through the gel and comparing the data with those from MRI scans of the injections in NHP. Since CED can behave differently depending on the location of injection in the brain, we are testing bench-side injections at different depths to validate the versatility of our model. We are seeing that the injections in the gel model mirror that of the injections in NHP brains as expected. Our next steps are to test the effectiveness of smaller injection cannula sizes with our bench-side model to assess if injection results remain consistent. This would indicate that tissue damage could be minimized in surgeries while still achieving desired injection parameters.

Optogenetic stimulation is a technique that modulates the activity of genetically modified neurons with light of a particular wavelength. Optogenetic modulation has high temporal resolution and cell-type specificity that enables precise stimulation of cortical neurons and allows us to conduct artifact-free recording during stimulation. Using a large-scale optogenetic interface, we stimulated and recorded across the primary somatosensory (S1) and motor (M1) cortices of non-human primates (NHP). We conducted our investigation on NHPs because their cortical organization is particularly similar to that of humans. The goal of this study is to determine the effect of various spatial and temporal patterns of optogenetic stimulation on the neural response and network dynamics across the two cortical regions. Delivering stimulus pulses via two lasers placed on top of the cortical surface, we found that stimulation of one cortical region evoked neural responses across both S1 and M1, which we then classified into primary and secondary responses based on their delays. While our previous work has established that optogenetic stimulation strengthened functional connectivity between S1 and M1, we wanted to further investigate the distribution of primary and secondary neural responses after repeated stimulation. We examined two measures of neural responses, the temporal delay between the trough of evoked response and onset of light stimulation, and the distribution of power across cortical networks up to 50ms after the stimulation. Our preliminary results indicate that optogenetic stimulation changed the delay of the primary and secondary response. We also observed that different temporal patterns of paired laser pulses evoked distinct neural activity. Identifying different neural responses after complex spatiotemporal patterns of stimulation would help us predict network changes post cortical modulation and contribute significantly to the development of stimulation-based clinical therapies and rehabilitation strategies for neural disorders.

Non-human primate (NHP) research is a pivotal step in the progression of neuroscientific and neural engineering research from animal models to human trials. In most NHP neuroscience experiments, neurosurgery is required to implant devices such as head posts, recording arrays, and optical windows. Current practices for these surgeries use methods for surgical preparation that carry a degree of unavoidable uncertainty. This comes from an inability to visualize and test the physical compatibility of complex components and anatomy prior to neurosurgery. This project details methods for creating 3D printed models of a subject’s brain and skull, as well as an agarose gel model of the brain. These models can be obtained from magnetic resonance imaging (MRI) using brain extraction software for the brain model, and custom code for the skull. The preparation protocol takes advantage of state-of-the-art 3D printing technology to combine models of the brain and skull with neuroprosthesis. With the addition of a craniotomy using the custom code, the skull and brain models can visualize brain tissue inside the skull, enabling better preparation for surgeries. Using the methods outlined in the protocol, the accuracy of the 3D printed brain, skull, and craniotomy placement were successfully validated through a comparison to the original MRI scan. The gel brain was additionally used to visualize delivery of a mock viral vector through the craniotomy of a skull model. By preoperatively fitting a headpost to the physical model of the skull, we successfully shortened the implantation surgery time by 40% and greatly reduced the risk of operative complications. These methods are designed for surgeries involving neurological stimulation and recording as well as injection in NHPs, but the versatility of the system allows for future expansion of the protocol, extraction techniques, and models to a wider scope of surgeries.

Stroke, when the blood supply to the brain is reduced or disrupted, is a leading cause of disability among adults. Brain plasticity, also known as neural plasticity, can aid in the way that the brain recovers from a traumatic injury such as a stroke by creating newer, stronger synapses (connections) between neurons and thus increasing cell functionality. This literature review explores how brain plasticity informs new bioengineering solutions to stroke rehabilitation and asks, “What type of novel stroke treatments have been developed using the concept of brain plasticity?” Preliminary findings indicate that a breakthrough in this field is using optogenetics to trigger and control the neural connections that the brain can make through promoting motor function after ischemic stroke. Other innovations in this field include repetitive transcranial magnetic stimulation, transcranial direct current stimulation, and epidural cortical stimulation, which have all been shown to make permanent changes in neural synaptic transmission. These methods partially restore brain function while being less invasive and more effective in comparison to older interventions. This literature review indicates that new bioengineering treatments informed by brain plasticity are promising and could promote better rehabilitation outcomes for those suffering from stroke and potentially other traumatic brain injuries.

The locus coeruleus (LC) is a small nucleus of noradrenergic neurons in the pons, which, despite its size, has broad projections throughout the central nervous system (CNS). Functionally, the LC is believed to be involved in various critical functions, including the physiological response to stress, as well as mediating arousal. Previous investigations have demonstrated that optogenetic activation of the LC at a tonic frequency promotes wakefulness in rodents. While this observation causally implicates LC function in wakefulness, it is still not known how the LC is endogenously controlled to mediate arousal. One potential candidate in this control is the peptide nociceptin and its cognate receptor, the nociceptin opioid peptide receptor (NOPR), both of which are highly expressed around the LC. To investigate, we first conducted two pharmacological experiments using the NOPR agonist Ro64-6918 to assess its effects on locomotion and on the activity of LC noradrenergic neurons. To determine where the endogenous nociceptin signal to the LC originates, we performed an intracranial injection of a Cre-dependent retrograde virus (AAV2-DIO-eYFP) into the LC of a mouse expressing Cre recombinase in nociceptin-expressing neurons. We observed that Ro64-6198 (10 mg/kg) strongly reduced open field locomotor activity compared to vehicle treatment. Using in vivo 2-photon calcium imaging (GCaMP6s), we found that Ro64-6198 (5 mg/kg) profoundly reduced LC noradrenergic neuron activity. Wakefulness appeared reduced in both in vivo experiments. Finally, we identified nociceptin-expressing cells projecting to the LC in the peri-LC as well as a long-range projection from the bed nucleus of the stria terminalis (BNST). Together, these studies suggest that nociceptin acting on LC noradrenergic neurons reduces arousal, and that the endogenous sources of nociceptin may originate in the peri-LC and BNST. Future studies will investigate nociceptin-expressing neuron activity during sleep/wake transitions and whether this activity is sufficient to alter wakefulness.

Nanofibers have broad capabilities in biomedicine, e.g. drug delivery and tissue engineering, because of their diverse tunable properties. For these purposes, nanofibers often must be patterned to be integrated in devices or to optimize their function. Current nanofiber patterning methods lack user-control over design or alter fiber integrity. To optimize and expand the applications of nanofibers, there is a need for a versatile nanofiber patterning strategy that maintains material integrity and function and can be generalized for patterns of different complexity and dimensions. This project aims to address this need with an in situ patterning strategy that allows for complex, three-dimensional patterning at the milli/microscale with different fiber materials. The approach consists of a two-layer composite electrospinning collector with an insulative layer and conductive recessed patterns. An inexpensive collector fabrication method was designed for rapid prototyping. Collector design features predicted to affect fiber deposition were evaluated by quantifying fiber selectivity. Optimal formulations of nanofiber materials were experimentally evaluated based on reproducibility, fiber yield, and selectivity to delineate key polymer solution properties affecting patterning. This project offers a guiding foundation to adapt this patterning strategy to various applications of nanofibers by tuning fiber formulations and specific collector design features.

Implanted biomedical devices are becoming increasingly common for the treatment of tissue defects and organ failures. However, there is an ongoing issue of biocompatibility, where the host body constantly attempts to degrade the foreign object while it can’t most of the time, especially in the case of synthetic polymers. Consequently, virtually all implants will undergo an immune response named foreign body reaction and eventually get encapsulated in collagen, which can be detrimental to the device’s designated function, especially for drug delivery systems. Extensive research attempting to improve biomaterial integration has been conducted in the past decades. Recent studies suggested that the in vivo vascularization within and around a porous polymeric biomaterial is partially driven by the local phenotypic expression of macrophages, where the M1 macrophage was specifically shown to play an angiogenic role. It was also observed that the in vitro pre-vascularization of biologically-derived constructs accelerates and enhances tissue vascularization in vivo. Enlightened by these findings, I proposed to modulate both macrophage phenotypes at the material-tissue interface and scaffold pre-vascularization to restore tissue homeostasis. Two tools previously developed by my mentor’s research groups are vital for my study: the engineered M1-inducible macrophages (i-M1macs) and sphere-templated porous poly(HEMA) biomaterial scaffolds. Three main aims were constructed, where the first two aims are to optimize in vitro the scaffold pre-vascularization from endothelial cells and activation of i-M1macs in scaffolds respectively. The third aim is to determine the effect of macrophage modulation and scaffold pre-vascularization on the foreign body response to biomaterials in vivo. Findings from my research are expected to improve our understanding of the correlation of macrophage plasticity, material porosity, scaffold pre-vasculature, and tissue vascularization, which can be crucial for the development of a novel cell therapy that improves biomaterial integration and ultimately, the quality of life of people with biomedical implants.

Loss of periodontal ligament tissue (PDL) and attachment is a serious complication of periodontal diseases - the most prevalent dental health problems. PDL-degeneration leads to alveolar bone degeneration, infection, gingivitis, and eventual tooth loss. There is currently no product that can cure PDL-degeneration as regeneration requires the combinatorial process of regenerating cementum, signaling the existing relevant cells to proliferate and form PDL, and its integration into a functional system. Current restorative treatments utilize cell-based tissue regeneration, synthetic scaffolds, tissue grafts with limited, temporary success. A market product, e.g., claims to restore periodontium using harvested fetal swine periodontal tissue with highly variable clinical outcomes. Although these traditional procedures are well-established and show some success, their efficacy is limited due to the lack of structural and functional integration of a deposited layer with the underlying tooth, specifically integration into the remineralized cementomimetic layer. GEMSEC labs have developed a proprietary technology dubbed “peptide-guided remineralization” which facilitates new mineral formation using protein-derived peptides and have successfully restored dental hard tissues via several case studies including enamel, cementum, dentin under in-vitro and in-vivo conditions. Translating this technology into a daily-use product, we propose a PDL-regenerating chimeric construct which includes a biomineralizing peptide, ADP5, derived from the key enamel protein, amelogenin, with cell signaling moieties. Herein, we aim to use established bioinformatics, machine-learning tools, and high-throughput experimentation to identify peptides from proteins involved in PDL development cell-signaling towards controlled biomineralization, bioadhesion, and cell-signaling functionalities necessary for PDL regeneration. Addressing current treatment protocol limitations, the interdisciplinary approaches developed in this project are designed for the regeneration and formation of fully functional PDL. 

Every second, living human cells are executing complex functions to initiate or influence various processes such as energy production, growth, metabolism and reproduction. This also means that intracellular structures are rapidly moving and morphing as the various proteins and organelles interact. Two important goals for scientists attempting to understand and manipulate intracellular processes are to visualize structural movements within live cells and to use specifically generated proteins to adjust interactions within the cell. Recently, members of the University of Washington’s Gao Lab have made a breakthrough by creating a cholesterol tag that allows delivery of small proteins across the cell membrane. I have explored the use of this method for delivery of immunological agents such as sdAbs (single domain antibodies) and synthesized Fab fragments for fluorescent visualization of structural cellular elements. These types of proteins have very specific target-binding which makes them optimal for imaging and therapeutics. Each of the tested immunological agents was delivered into the live cells successfully and the imaging proteins provided clear pictures of cell structures. Better understanding the functions and limits of this cytosolic-delivery method will increase the accessibility of live-cell fluorescent imaging, while the success of delivery opens up new possibilities in the field of intracellular protein therapies.

Chronic kidney disease affects millions of people worldwide. Although dialysis offers a temporary treatment for these patients, it cannot effectively remove indoxyl sulfate, a naturally produced uremic toxin which can cause major illness. Our goal is to remove the precursor of indoxyl sulfate, tryptophan, from the small intestine, before it is eventually broken down by the gut microbiome in the colon and circulates as indoxyl sulfate in the blood. We propose a novel method to remove excess tryptophan from the small intestine by utilizing an orally ingested, albumin-encapsulated hydrogel microsphere. Since tryptophan has a high binding affinity for human serum albumin, we can employ albumin encapsulated hydrogel microspheres to remove excess tryptophan from the distal end of the small intestine to prevent the ultimate production of indoxyl sulfate in patients with chronic kidney disease. In this work, we provide in vitro results demonstrating the feasibility of such a system: we show that our hydrogels can remove considerable quantities of tryptophan from solution. We also propose alternative hydrogel formulations to limit the leakage of important components. Finally, we display modeling in COMSOL which can somewhat replicate the diffusion and binding conditions in the gut and discuss its implications in supporting future work.

Blood clotting plays a heavy contribution to the mortality and morbidity in patients with diabetes mellitus. In blood clotting, the interaction between platelet glycoprotein Ib (GPIb) and blood protein von Willebrand Factor (VWF) is a catch bond, a bond whose lifetime increases under tensile force. Current, identification of this single-molecule behavior of a catch bond is useful but insufficient to understand behavior with multiple molecules known as clusters. It is recognized that patients with diabetes mellitus have higher levels of VWF and GPIb but there is no existing flow assay that accurately demonstrates the difference in thrombosis flow in diabetics. In addition, there is no definitive conclusion on the influences of the amount and geometry of the catch bond between GP1b and VWF in clusters. A recent innovation in the Thomas Lab has developed a DNA origami nanostructure that allows control over the number and spacing of ligands in a cluster, facilitating the study of clusters of catch bonds. I designed a method to quantify the nanostructure-presented ligands on a surface using 96-wellplate reader. This method was used to characterize the effect of cluster size on platelet rolling behavior. The result of the method suggests using quenching low concentration of biotin-4-fluorescein had the highest accuracy and was able to be picked up by the 96 well plate reader. The method allows observation seen in platelet rolling behavior over these surfaces that are cluster-size dependent rather than concentration-dependent. This will introduce a new assay for diabetic studies and further our understand difference in platelets rolling behavior over clusters between diabetic patients and non-diabetic patients.

Environmental stress and threat influence feeding behavior in animals, but how that interaction occurs is still largely unclear. Neurons in the bed nucleus of the stria terminalis (BNST), part of the extended amygdala, have dense projections to the parabrachial nucleus (PBN) in the brainstem. We have uncovered projections in these neural circuits that link the modulation of feeding and threat assessment in mice. This project aims to investigate and characterize these previously unrecognized neural circuits with the incorporation of a variety of optogenetic, surgical, and histological techniques. We used Cre-dependent anterograde and retrograde viral tracers in order to trace the anatomy of these neural circuits, and found functional projections from inhibitory (GABA) and excitatory (glutamate) populations in the BNST to neurons in the PBN. We also used translating ribosome affinity purification (TRAP) to isolate the mRNA of these projections. This proved useful in separating and identifying the molecular expression profile of different GABAergic and glutamatergic subpopulations. Furthermore, we used a variety of behavioral assays to determine the BNST-PBN circuits’ role in feeding and threat-response behavior. We used fiber photometry to track the activity of GABAergic (vGAT) and glutamatergic (vGLUT2) populations during these behaviors, and found that vGAT and vGLUT2 populations have differing roles in threat and feeding behaviors. vGAT neurons increase their activity during feeding and decrease in response to threat, while vGLUT2 neurons decrease their activity during feeding and increase in response to threat. We also used optogenetic activation of these neurons to determine their causal role in behavior. With activation, vGAT populations drive place preference, operant positive reinforcement, and increased feeding. Conversely, vGLUT2 populations drive place aversion, operant negative reinforcement, and reduced feeding. These findings characterize the distinct nature of BNST-PBN neural circuits and the mechanism behind the evaluation of threatful stimuli and the integration of feeding.

 Stroke is the leading cause of long-term disability in the United States. Disabilities can range from a loss of sensory function like touch to motor functions like controlling arm movements due to the damage to the brain’s network. Despite the prevalence of stroke, the underlying network dynamics that lead to functional deficits are not well understood. Due to the physiological similarities between non-human primate (NHP) and human brains, an NHP model is essential for studying the effects of stroke and developing therapies. Here we used the photothrombotic (PT) stroke technique to study network dynamics in the NHP sensorimotor cortex following an ischemic lesion. Using the PT stroke technique, we induced a focal ischemic lesion on the NHP sensorimotor cortex. We collected local field potentials from both hemispheres using an electrocorticographic  array on the cortical surface. As a measure of neural activity, we calculated ipsilesional power in the low gamma band. Channels were then organized into three clusters based on their net change in power (increase, decrease, no change). As expected, the cluster with an overall decrease in power corresponded to the lesion's physical location. We also studied the network connectivity by calculating pair-wise coherence across different frequency bands: theta, beta, low gamma, and high gamma. Overall, we saw that low frequencies were associated with decreases in coherence, while higher frequencies were associated with increases following stroke. Preliminary results from the contralesional hemisphere show similar changes. In this study, we observed local neurophysiological changes up to three hours following an ischemic lesion. The observed increases in power in the perilesional region and coherence at high frequencies suggest compensatory mechanisms immediately following an injury. We can use this study's results to guide future developments in stimulation-based therapy to alleviate the functional deficits from a stroke.

Neuropeptide S (NPS) is a neuropeptide produced primarily in two regions of the hindbrain, the locus coeruleus (LC) and the Kolliker-Fuse nucleus. The LC-NPS population is particularly interesting because of the LC’s role in norepinephrine production and subsequent transmission throughout the brain. Previous work has found that when NPS is injected into the amygdala, it results in an anxiolytic phenotype, implicating NPS and its G-protein coupled receptor (NPSr1) in anxiety-related behaviors. Using fluorescent in situ hybridization, a method which allows visualization of single RNA molecules within cells via fluorescent probes, we found preliminarily, that the orbitofrontal cortex (OFC) has dense expression of NPSr1 RNA. This is significant as the OFC is involved in higher-order cognition including social, reward-learning, and anxiety-like behaviors. For example, OFC neurons respond to social interaction as well as food cues, and inactivation of the OFC results in increased anxiety-like behavior. The LC is also known to send projections to the OFC that have been largely unexplored. Therefore, to better understand and characterize the connection between the LC and OFC we utilized in vivo fiber photometry to assess endogenous OFC-NPSr1 activity during reward-learning, social interaction, and innate behaviors. Our studies aim to uncover the functional role of LC-NPS release in the OFC.

Reward is a driving force for animal and human behavior. Reinforcing behaviors with rewards leads to enhanced learning ability, which can either promote behaviors that increase survival, or lead to maladaptive behaviors. The nucleus accumbens (NAc) and ventral tegmental area (VTA) are brain regions established to be involved in reward processing and have significant neural connectivity. Recent studies have identified a long-range GABAergic neural circuit connecting these two regions, however previous studies focus primarily on dopaminergic neurons. These inhibitory GABAergic neurons synapse with cholinergic interneurons within the NAc shell (NAcSh). Further, the dorsal and ventral subdivisions within the NAcSh have been shown to have different neural connectivity. To investigate the role of this circuit, I performed fiber photometry recordings of neural activity in GABAergic terminals in the dorsal and ventral NAcSh during reward reinforced behavior in mice. The recordings show an increase in GABAergic neural during reward consumption in the ventral, but not the dorsal, NAcSh. I also recorded the activity of NAcSh cholinergic interneurons as well as acetylcholine activity in the dorsal and ventral NAc shell. These recordings show that cholinergic neural activity as well as acetylcholine activity are reduced during reward consumption in the ventral, but not dorsal, NAcSh, reflecting the inhibition by the GABA neurons during this time. I also used the inhibitory photo-activatable chloride pump JAWS to inhibit GABAergic projections during reward consumption, finding that animals made reduced reward seeking events and consumed fewer rewards when JAWS is activated. Collectively, these results indicate GABAergic projections from the VTA to specifically the ventral NAcSh function in reward reinforcement by inhibiting cholinergic activity during reward consumption. These results characterize a previously unknown neural circuit and help us better understand psychiatric disorders like depression and addiction that impact these circuits. 

 Millions abroad and at home have been devastated by the COVID-19 pandemic. The worldwide case total surpasses 110 million with a brutal death toll of 0.5 million in the US alone (nearly 2.5 million worldwide). Comprehensive characterization of SARS-CoV-2 and its impact on patient immune systems remains under-analyzed but is critically needed for the development of COVID-19 therapies. Our study presents an integrated analysis of patient clinical measurements, immune cells and plasma multi-omics of 139 COVID-19 patients. This cohort represents the entire spectrum of disease severity (as quantified via WHO) with longitudinal blood draws collected during the first week of infection following clinical diagnosis. We identify a major immunological shift between patients of mild and moderate disease, at which point elevated inflammatory signaling is accompanied by the loss of specific metabolite classes and processes. These metabolites include amino acids and lipids that may be indicative of nutrient depletion. This stressed plasma environment found in patients with moderate or severe disease is accompanied by the onset of multiple uncanonical immune cell phenotypes. These phenotypes include proliferative-exhaustive T cells, cytotoxic CD4+ T cells and dysfunctional monocytes; the presence of these unusual subtypes are amplified by increasing disease severity. Further, we condensed over 120,000 immunological features into a single axis to capture the ways in which different immune cell classes coordinate with each other in response to SARS-CoV-2 infection. This immune-response axis independently aligns with the major plasma composition changes, clinical metrics (including blood clotting), and with the sharp transition between mild and moderate COVID-19 patients. Our study offers deep immunophenotyping of COVID-19 patients that, through integration of multi-omic analyses, suggests that moderate disease may provide the most effective setting for therapeutic intervention.

Personalized medicine, enabling each patient to receive earlier diagnoses, risk assessments, and optimal treatments hold promise for improving cancer health care while also lowering costs. For example, more than 60% of breast cancers (BC) in women are diagnosed as estrogen receptor-positive/human epidermal growth factor receptor 2-negative (ER+/HER2-), which is then typically treated with adjuvant therapy that combines endocrine therapy with CDK4/6 inhibitors (CDK4/6i). In addition to high cost, CDK4/6i treatment time can increase particularly within the <50% patient population that experiences drug resistance and who must proceed to a second line of treatment. Exosomes are membrane-bound extracellular vesicles that have recently demonstrated rapid growth in BC research due to their vast array of tissue-specific surface markers and molecular contents that can be used to confirm a prognosis. For example, overexpression of exosomal TK1 is associated with CDK4/6i resistance; thus, exosomal cargo content can be analyzed to enable tailored treatment with the best response and highest safety margin for ER+/HER2- BC patients. However, exosome heterogeneity has hindered research progress due to lagging analytical techniques to effectively characterize and isolate BC-specific exosomes. This project combines an oligonucleotide hybridization reaction with temperature-sensitive polymer-oligonucleotide conjugates that can detect and rapidly isolate specific exosome subtypes depending on tissue-specific exosome surface proteins. We expect that the isolated exosome cargo will quantitatively demonstrate susceptibility or building resistance to CDK4/6i. Future work includes optimizing the oligonucleotide sequence, further pinpointing target BC-specific surface markers, testing the assay, and comparing results to current methodologies. Ultimately, enabling an exosome liquid biopsy method to tailor patient-specific BC treatments can decrease the overall time, cost, and toxicity associated with current non-specific cancer treatment methods and can also be utilized in monitoring dynamic cancer metastasis.

Reactive oxygen species (ROS) are metabolites which play a critical role in biological systems. The deregulation of ROS creation and reduction damages proteins, lipids, and DNAs. Specifically, hydrogen peroxide (H2O2) is a key molecule known for its significant role in the various physiological processes impacted by cellular redox dynamics. However, there is a lack of tools with the sensitivity, dynamic range, and kinetics necessary for researchers to measure intracellular H2O2 levels. Our lab previously developed a genetically encoded fluorescent indicator for the real-time monitoring of H2O2 dynamics to address this need. Genetically encoded fluorescent indicators are protein-based sensors that allow researchers to visualize the activity of important molecules and understand the mechanisms by which critical physiological pathways operate in living systems. These sensors have great advantages as tools for quantifying molecular activity, including high spatiotemporal resolution, subcellular specificity, and minimal invasiveness. Our green-fluorescent H2O2 sensor HRM63 demonstrated 50-times faster onset and 10-times higher sensitivity for H2O2 when compared to the current state-of-art redox sensor HyPerRed – a significant improvement. I strive to expand the fluorescent palette of H2O2 sensors based on the design of HRM63 in order to enable the multiplexed imaging necessary to capture the dynamic pathophysiological processes which involve ROS. I have taken a structure-guided protein design approach and applied molecular cloning techniques to develop and iteratively optimize red-shifted versions of HRM63 with longer excitation/emission wavelengths, which would allow for deeper penetration of light source and fluorescence. I validate the performance of prototype sensors by applying fluorescence microscopy techniques to image their responses to applied concentrations of H2O2 in models of human disease. By optimizing these sensors, we will create invaluable tools for researchers to study the pathophysiology of the wide range of diseases to which ROS are linked, including cardiovascular diseases, neurodegenerative diseases, and cancer.