Electronic Theses and Dissertations

Permanent URI for this collection

https://ir.vanderbilt.edu/handle/1803/9599

Electronic theses and dissertations of masters and doctoral students submitted to the Graduate School.

Browse

Now showing 1 - 20 of 7416

The Role of Extracellular Matrix Alterations Induced by Radiation Therapy in Modulating Breast Cancer Recurrence
(2025-03-24) Zhu, Tian; Rafat, Marjan
Triple-negative breast cancer (TNBC) is an aggressive subtype of breast cancer with high metastatic potential, and radiation therapy (RT) is a crucial treatment strategy. However, despite its effectiveness in reducing recurrence overall, many patients with TNBC experience relapse, suggesting that irradiation may play a significant role in promoting a pro-tumor microenvironment. This dissertation investigates the impact of RT-induced extracellular matrix (ECM) remodeling on tumor recurrence, with a focus on structural, molecular, and mechanical properties of the ECM. We employed a combination of in vivo and ex vivo irradiated models, including non-tumor bearing and tumor-bearing models, to assess how RT affects ECM properties that regulate cancer and immune cell behavior. Our analysis of murine mammary fat pads revealed significant ECM alterations following RT, including increased collagen deposition, enhanced fiber density, increased tissue stiffness, and changes in the expression of key ECM components such as collagen I, IV, VI, and fibronectin. Utilizing decellularized ECM (dECM) hydrogels, we successfully replicated the in vivo irradiated microenvironment and studied its effects on tumor and immune cell proliferation, invasion, and function. Our results demonstrate that RT-induced ECM structure and composition modifications create a permissive platform for breast cancer cell growth, invasion, and immune modulation, especially in an immunocompromised context known to correlate with post-RT relapse. Notably, TNBC cells co-cultured with macrophages in irradiated dECM hydrogels led to macrophage polarization toward an immunosuppressive phenotype, which further enhanced tumor cell proliferation in both ex vivo irradiated and in vivo irradiated tumor-bearing models. Additionally, RT-induced ECM remodeling promoted cancer and immune cell metabolic reprogramming and facilitated ECM uptake. In conclusion, our study underscores the critical role of ECM remodeling in the RT response, highlighting its contribution to tumor growth and invasion as well as immune suppression in TNBC. These findings provide new insights into how targeting ECM-immune cell interactions may improve RT outcomes and reduce the likelihood of cancer recurrence.
Developing and Validating Imaging Techniques for Accurate Quantification of Whole-Body Skeletal Muscle Architecture
(2025-03-22) Zhou, Xingyu; Damon, Bruce M.; Damon, Bruce M.
Muscle architecture, defined as “the number and orientation of the muscle fibers within a muscle”, has an important influence on a muscle’s ability to produce force and shorten or lengthen under load. Briefly, muscle fiber arrangements that result in more sarcomeres aligned in series to each other and parallel with the muscle’s mechanical line of action allow greater length excursions and contraction velocities, while muscle fiber arrangements that result in more sarcomeres aligned in parallel to each other and obliquely to the muscle’s mechanical line of action allow greater force production. In addition, the curved fiber geometry that is necessary for mechanical stability results in heterogeneous patterns of strain development and the formation of intramuscular fluid pressure gradients that may restrict perfusion. However, our tools to study the structure-function relationship of human skeletal muscle, including ultrasound (U/S) and diffusion-tensor magnetic resonance imaging (DTMRI), need further development. The overall goal of this dissertation research is to develop and validate imaging techniques for accurate quantification of whole-body muscle architecture. First, we developed a numerical simulation framework and systematically examined the effects of condition variables on the accuracy and precision of DTMRI-based muscle architecture estimates. Based on the simulation predictions, we made practical recommendations for the implementation of in vivo skeletal muscle DTMRI experiments. In the second part of our study, we implemented a fiducial-based co-registration framework which allowed the direct comparison between U/S- and DTMRI-reconstructed muscle fiber architectures in vivo. As previous studies showed that U/S had good agreement with direct anatomical inspection in muscle architecture measurements, our newly developed U/S-MR cross-modality registration framework demonstrated its potential to work as a validation tool for DTMRI-based in vivo muscle architecture measurements. In the third part of our study, we designed and implemented an in vivo experiment protocol for the characterization of human skeletal muscle. We focused on the muscle groups that act on the lower body, optimized the image acquisition and analysis protocols, and measured the architecture parameters of representative muscles. Overall, this work has produced an improved understanding of skeletal muscle architecture, and new tools for its validation and in vivo characterization.
Metabolic Engineering of Phaeodactylum tricornutum for Triacylglycerol Production
(2025-03-18) Zheng, Amy Olivia; Young, Jamey; Young, Jamey
Culturing microalgae for biodiesel production is one of the few viable methods to completely replace our reliance on petroleum for transportation vehicles. Phaeodactylum tricornutum (Pt) produces lipids called triacylglycerol (TAG) that can be easily converted into biodiesel. Despite the recent advancements in TAG production, a clear understanding of how environmental conditions and genomic engineering enhance TAG production. In this work, 13C isotopic nonstationary metabolic flux analysis (INST-MFA) was utilized to investigate how light intensity and malic enzyme overexpression increase TAG production. First, wild type Pt cells were cultured under two different light conditions, 60 μE and 250 μE. Increased light intensity was found to increase growth and TAG production. As light intensity increases, the excess fixed carbon was directed towards TAG synthesis instead of biomass. Next, a malic enzyme (ME) overexpressing strain was developed through electroporation. The mutation increased TAG production while maintaining equivalent growth rate to wild type. Using INST-MFA, the primary mechanism of ME was shown to direct the citrate and pyruvate pool towards TAG synthesis and decrease TCA cycle flux. Finally, a systematic analysis of the effect of pool size measurements on flux estimates was conducted. Pool size measurements increase sensitivity to errors in the model. Therefore, inclusion of pool size measurements results in the accurate flux estimates even when there are certain model errors. Overall, these studies show the utility of INST-MFA to elucidate the underlying mechanisms for TAG synthesis. Furthermore, this work highlights the potential of INST-MFA to develop gene targets and optimize cell culture conditions to increase TAG production.
Molecular Basis for Ligand Activation of NR4A-RXR
(2025-01-06) Yu, Xiaoyu; Kojetin, Douglas J; Sanders, Charles R
Two members of the orphan nuclear receptor (NR) NR4A family, Nurr1 and Nur77, are essential for the development, regulation, and maintenance of several important aspects of mammalian brain development and homeostasis and have been shown to be associated with pathological development and progression, including Parkinson’s disease (PD) and Alzheimer’s disease (AD). Although NRs are considered to be ligand-dependent transcription factors, Nurr1 and Nur77 are thought to function independent of binding an endogenous ligand that is produced and present in cells. It also remains unknown in the field whether there are functional endogenous ligands that bind to the ligand-binding domains (LBD) of Nurr1 and/or Nur77 via their putative canonical ligand-binding pocket, a region within the core of the LBD conserved across the NR superfamily where endogenous ligands bind to other NRs. An alternative way to target NR4A activity is through ligands that bind to the retinoid X receptor (RXR), which forms heterodimers with Nurr1 and Nur77. Interestingly, although RXR activates transcription of most NR-RXR heterodimers including PPARγ-RXRα, RXR heterodimerization represses Nurr1 and Nur77 transcription. Additionally, synthetic RXR ligands can activate transcription of Nurr1- and Nur77-RXR heterodimers and have been implicated in treating neurodegenerative disorders. However, the mechanism by which RXRα represses Nurr1 and Nur77 transcription, and the mechanism of action of ligands that activate NR4A-RXR transcription, remain poorly understood. Here, I use a multidisciplinary approach combining structural biology, biochemistry, biophysics, and molecular and cellular pharmacology approaches to characterize the structure of Nurr1-RXRα LBD heterodimer, which uncovered a novel structural mechanism of ligand-dependent activation of NR4A-RXR heterodimers: RXR ligands can function as allosteric protein-protein interaction inhibitors that bind to the canonical orthosteric ligand-binding pocket within the RXR LBD and influence NR4A-mediated transcription via heterodimer dissociation. This dissertation provides fundamental insights into the structural mechanisms governing NR4A-RXR heterodimer function and ligand-dependent activation, which may inform the development of targeted therapeutic strategies for PD, AD, and other aging-associated neurodegenerative disorders.
RSK2 in Homeostasis
(2024-12-04) Wright, Eric Brandon; Lannigan, Deborah A; Duvall, Craig; Brunger, Jonathan; Blind, Ray; Stafford, John
The p90 ribosomal S6 kinase 2 (RSK2) is a serine/threonine protein kinase activated by extracellular signal-regulated kinase 1/2 (ERK1/2) whose physiological functions are not fully understood. We found that loss of RSK2 results in decreased litter size and underweight pups. In an effort to understand these fertility issues serum lipid levels were analyzed and serum triacylglyceride (TAG) levels were found to be decreased in female RSK2 knockout (RSK2KO) mice compared to wild type (WT). TAG levels are important regulators of fertility with the liver being the major organ involved in TAG metabolism. In support of a defect in liver lipid metabolism, TAG levels were also found to be reduced in the livers of female RSK2KO mice. Transcriptomic analysis of livers from female RSK2KO mice, staged during high estrogen levels, showed decreased expression of genes involved in de novo lipogenesis (DNL). These results were validated by qrtPCR. DNL is regulated by the transcription factor sterol response element binding protein 1c (SREBP1c), whose mRNA expression levels did not change. SREBP1c activity is tightly regulated and studies on the mechanism of RSK2 regulation of DNL are ongoing. DNL is regulated in an estrogen-dependent manner and interestingly, loss of RSK2 eliminated the increase in DNL that occurs in response to estrogen signaling. Transcriptomic and qrtPCR analysis showed that RSK2 increases the expression of ESR1, the gene encoding estrogen receptor alpha (ER), during high estrogen levels. We propose that RSK2 drives estrogen signaling in the liver through regulation of ER mRNA levels by a mechanism that is currently being investigated. TAG levels are essential for fertility and we hypothesize that RSK2 regulates fertility through control of DNL in the liver.
Data-centric AI for Small Molecule Drug Discovery
(2025-03-24) Wang, Xin; Derr, Tyler; Berger, Matthew
Data-centric AI focuses on improving the quality and utility of data to enhance model performance rather than solely emphasizing innovations in model architecture. The shift from a model-driven to a data-driven paradigm has gained attention in drug discovery, where data quality issues and data quantity limitations present significant challenges for model training. Therefore, we introduced WelQrate, a high-quality benchmark dataset for small molecule drug discovery, supported by a professional curation pipeline and standardized evaluation framework aimed at bridging the gap between the biochemistry and AI communities. We found that dataset quality greatly influences model evaluation during benchmarking, stressing the importance of enhancing data quality. Our experiments revealed two significant challenges in real-world drug discovery data that must be carefully considered when developing new algorithms: extreme class imbalance due to low percentages of active compounds and structural distribution shift resulting from the unexplored chemical space compared to known drugs. Examining imbalance issues in WelQrate datasets, we developed ScaffAug, a novel data-driven augmentation framework that utilizes a diffusion model to generate novel training samples based on molecular scaffolds with under-represented structures and minor classes. We demonstrated that incorporating these generated molecules through a self-training strategy can significantly augment the performance of activity predictors.
A Synthetic Receptor Platform to Engineer Mesenchymal Stromal Cells for Osteoarthritis Detection and Guided Articular Cartilage Regenerative Medicine
(2025-03-05) Walton, Bonnie Leigh; Brunger, Jonathan M; Brunger, Jonathan M
Osteoarthritis (OA) treatment remains a significant clinical challenge due to the absence of disease-modifying therapies. Current approaches are primarily palliative, providing symptom relief or temporary efficacy without addressing the underlying pathology, including the degradation of articular cartilage. Mesenchymal stromal cells (MSCs) have shown promise for cartilage repair, but their therapeutic effects remain inconsistent and insufficiently targeted to OA pathology, limiting their clinical utility. To address these shortcomings, we engineered a synthetic Notch (synNotch) receptor that allows precise, spatially controlled activation of MSCs in response to type II collagen (CII), an extracellular matrix component exposed during OA-associated cartilage degradation. This system enables receptor activation and transgene expression only upon direct engagement with CII-coated substrates, including both adsorbed CII and enzymatically degraded, primary cartilage tissue. Notably, engineered CII-synNotch MSCs produce supraphysiologic levels of pro-anabolic transforming growth factor-beta 3 (TGF-β3) and anti-catabolic interleukin-1 receptor antagonist (IL-1Ra), two key modulators of OA pathology. Sustained transgene production facilitated both anabolic gene expression of MSCs and modulation of IL-1-associated inflammation in chondrocyte co-cultures. However, in vivo, CII-synNotch cells exhibited ligand-independent activation, prompting further investigation into underlying proteolytic and mechanical mechanisms of receptor activation in the joint. We investigated receptor-mediated activation in arthritic joints using canonical synNotch receptors and alternative receptor designs. While these studies offered insights into factors contributing to off-target activation, further optimization is required to enhance receptor specificity and achieve controlled activation that is aligned with OA progression. The findings presented here establish a foundation for synthetically regulated, engineered cell therapies that respond to OA-specific pathological features for programmed therapeutic effects.
Integrating Large-Scale Human Genetic and Regulatory Genomic Data to Functionally Annotate ctcf Binding Variation
(2025-03-11) Tubbs, Colby; Ruderfer, Douglas M; Gamazon, Eric
CCCTC binding factor (CTCF) regulates gene expression through DNA binding at thousands of genomic loci. Genetic variation in these CTCF binding sites (CBSs) are important drivers of phenotypic variation, yet extracting those that are likely to have functional consequences in whole genome sequencing (WGS) remains challenging. Through this dissertation, I explore conceptual frameworks to identify and prioritize CBS variants in gnomAD, a WGS database consisting of 76,156 individuals. First, I integrate computational and experimental predictions of CTCF binding into an empirical false-positive measure that can be applied to the score distribution of a precision-weight matrix. I then synthesize CTCF’s binding patterns at 1,063,878 genomic loci across 214 biological contexts into a summary of binding activity. This measure correlates with both conserved nucleotides and sequences that contain high-quality CTCF binding motifs. Finally, I use binding activity to evaluate high confidence allelic binding predictions for 1,253,329 SNVs in gnomAD that disrupt a CBS. I find a strong, positive relationship between the mutability adjusted proportion of singletons (MAPS) metric and the loss of CTCF binding at loci with high in vitro activity. Together, this body of work nominates thousands of rare, noncoding variants that disrupt CTCF binding for further functional studies while providing a blueprint for synthesizing large-scale genomic data to better prioritize noncoding variation in human disease studies.
Neuromodulation of Cognitive Flexibility in the Frontal-Striatal Network
(2025-03-24) Treuting, Robert Louis; Womelsdorf, Thilo
Cognitive flexibility, a pivotal component of learning and decision-making, enables human and nonhuman primates to adapt their behavior in response to changing environmental conditions. Neuromodulation techniques provide insights into how stimulation causally influence and facilitate adaptive behavior during complex learning tasks. Microstimulation studies in nonhuman primates (NHPs) have contributed substantially to the advancement of our understanding of the functional organization and behaviorally roles of neuronal circuits pertinent to attention and decision-making. This modality holds immense potential for augmenting cognitive performance, particularly in complex cognitive tasks. However, it has remained unclear if microstimulation can facilitate quick, flexible adjustments during learning. Here, we present the development and application of a gaze-contingent stimulation protocol targeting the anterior cingulate cortex (ACC) and the anterior striatum (aSTR), areas implicated to support fast learning of rewarded stimuli and attentional control, in the frontal-striatal network. The protocol is designed to achieve pro-cognitive neuromodulation during adaptive goal-directed behavior. Furthermore, we investigate the role of oscillatory burst events in mediating cognitive flexibility and assess the feasibility of using transient periods with oscillatory bursting events to improve behavioral performance. These methods hold clinical significance as such techniques could provide critical insights in establishing stimulation parameters for brain-computer interfaces (BCIs) that hold promise to be versatile treatment tools for subjects with neuropsychiatric disorders suffering from reduced cognitive flexibility including Parkinson’s disease, severe major depression, and obsessive-compulsive disorder. This research aims to make a transformative contribution to our understanding of the neural mechanisms underlying cognitive flexibility. In conclusion, this thesis provides fundamental insights to our understanding how neuromodulation of the frontal-striatal network affects cognitive flexibility.
Molecular Mechanisms of Mental Illness: Quantum Dot-based Single Molecule Investigation of the Dopamine Transporter and Its Protein Partners
(2025-03-22) Torres, Ruben; Rosenthal, Sandra J; Rosenthal, Sandra J
The human dopamine transporter (DAT) is a presynaptic transmembrane protein that facilitates the reuptake of synaptically released dopamine. Several lines of evidence indicate that DAT dysfunction is linked to neuropsychiatric disorders, such as autism spectrum disorder (ASD), bipolar disorder (BD), and attention deficit hyperactivity disorder. The lateral membrane diffusion and clustering propensity of DAT are emergent properties of functional dopamine signaling. This thesis work will be focused on interrogating DAT membrane diffusion dynamics utilizing a suite of independent labeling strategies. Using fluorescence microscopy, single particle tracking (SPT) of ASD/BD-associated DAT missense mutant A559V will be carried out by labeling them with antagonist-conjugated quantum dots (Qdots). The effects of the D2 dopamine receptor (D2R) modulation on aberrant DAT A559V diffusion will be investigated, including functional D2R antagonism and simultaneous tracking of both DAT and D2R. In addition, the propensity and extent of DAT and DAT A559V colocalization with D2R into membrane microdomains will be evaluated using a multiplexed Qdot-SPT strategy. While heterologous expression systems prove to be valid platforms in neuroscience, translating SPT studies to a native 3D neuronal architecture will be necessary for application in diagnosis and therapy for mental disease. To assess the degree to which DAT and its variants diffuse and cluster in vivo, our group has developed a novel antagonist-conjugated Qdot-SPT strategy that capitalizes on a Qdot design that yields exceptional photophysical properties and access to sterically hindered spaces in deep tissue to track native DAT in physiologically relevant environments such as brain tissue slices. This thesis work will be focused on probe optimization to reduce non-specific binding to maximize DAT membrane tracking confidence.
Crafting as Grafting‽ Investigating the Hybrid Aesthetics in Lotte Reiniger’s Films
(2025-03-24) Timmons, Wendy Cassandra Ellen; Koepnick, Lutz; Koepnick, Lutz
This dissertation is an investigation into the hybrid aesthetics of the pre-exilic films of the stop-motion silhouette puppet animator, Lotte Reiniger. Recent research has posited that Reiniger was a marginalized figure who worked in isolation and in the shadow of contemporary, male animators and filmmakers like Walt Disney and others. This has resulted in a self-fulfilling prophecy, in which the myth that Reiniger was unknown in her own time is perpetuated, shifting focus from the innovations of her art to the apparently necessary rediscovery of the filmmaker. The hybridization of media and genres belies the centrality of Reiniger’s work. Most of Reiniger’s films are adaptations, grafting fairy tales, theater pieces, and operas onto film. In these adaptations, self-referential scenes that show Reiniger’s hand at work or that play with sound design reveal the possibilities and limitations of film as a medium for adaptation. The result of this experimentation is a series of hybridized films that exhibit qualities of both the source material and analog film. Three body chapters provide examples of these hybrid films. The first is about Aschenputtel (1922), a film that thematizes the in/visibility of women’s labor. The second examines the ludic qualities of Harlekin (1931). The playful spirit of the commedia dell’arte genre is revived in the aesthetics of the film, suggesting that Reiniger’s creative process was a joyful engagement with crafting and hobbying. Finally, the third chapter discusses Papageno (1935), a film that puts the bird-catching side character in the spotlight. Together, these three films demonstrate that Reiniger was an artist who by no means worked in isolation, but rather, worked collaboratively with some of the biggest names in animation and composition of her time. In addition to this, her films demonstrate a deep awareness of the art historical context of her various sources as well as her own time. By deploying hybrid aesthetics, Reiniger participated in multiple artmaking practices and movements, which challenges present-day categories and forces researchers to think differently about the relationship between media.
Developmental Programming of Hypothalamic Neural Circuits Integrating Fluid and Energy Homeostasis
(2025-03-21) Sweet, Serena R; Simerly, Richard B.; McGuinness, Owen
Drinking and feeding are coordinated homeostatic events, but our understanding of the development of their converging hypothalamic neural circuits remains rudimentary. Many environmental factors developmentally program neural circuits during critical periods in early postnatal life. Agouti-related peptide (AgRP) neurons are substrates of developmental programming, responding to nutritional cues during a critical period to reach downstream targets. The paraventricular nucleus of the hypothalamus (PVH) receives inputs from both AgRP neurons and the median preoptic nucleus (MEPO) to regulate energy and fluid homeostasis, representing a possible node of integration. We used TRAP2;Ai14 double transgenic mice to show a population of active neurons responding to water deprivation (Thirst-TRAP) overlaps with fast-refeed Fos induction (Hunger-Fos) in the PVH. To determine the age at which MEPO projections reach the PVH, we used DiI axonal labeling as well as Fos-labeling in response to hypertonic saline (HS). Our results indicate neurons in the MEPO project to the PVH and respond to HS by the end of the first week of life, and densities of Fos-labeled nuclei in the PVH do not peak until the second postnatal week, preceding innervation of the PVH by AgRP neurons. Based on these observations, we hypothesized perturbations to fluid homeostasis in neonatal mice may impact the formation of AgRP circuitry with sustained changes in ingestive behavior. Adult male mice exposed to HS treatment daily from postnatal day (P) 5 to P15 (HSPN) displayed significantly increased densities of AgRP axons in the MEPO and PVH, while female HSPN mice experienced a decrease in the MEPO. Short-term high fat diet (HFD) exposure leads to greater water intake in adult HSPN male mice, while long-term HFD leads to a significantly lower body weight gain in HSPN females. Moreover, a dehydration-anorexia challenge results in a sustained anorexic response HSPN males after rehydration, while a fast-refeed challenge results in a sustained decrease in water intake in HSPN males after refeeding. Together, these results suggest a sexually dimorphic effect of early perturbations to fluid homeostasis on the development of feeding circuits, with context-specific consequences for ingestive behavior.
Diagnostics of Soft X-ray Emission from the Sun: Energetic Solar Flares and Gentle Quiescence
(2025-03-26) Suarez Bustamante, Crisel; Holley-Bockelmann, Kelly; Moore, Christopher S.; Holley-Bockelmann, Kelly
The Sun is complex and enchanting. It produces all types of electromagnetic radiation and releases the energy needed to sustain life on Earth. The variability of the solar soft X-ray (SXR) emission can answer: (1) How plasma is transported within the solar atmosphere during solar flares? (2) What mechanisms could drive Quasi-Periodic Pulsations (QPP) signatures in solar flare radiation? (3) What are the characteristic temperatures of the quiescent Sun? During periods of solar activity, the dim and hot solar corona becomes a blazing ball emitting flaring phenomena. Solar flares are capable of releasing over 10^32 erg, radiating SXR light, and heating the ambient plasma to excess of 20 MK. The SXR inferred elemental abundance of low first-ionization potential (FIP) elements are observed to deviate from their nominal values during solar flares. The Miniature X-ray Solar Spectrometer CubeSat–1 (MinXSS-1) conducted solar SXR spectral observations of ~30 flares C to M class flares. The flares were spectrally examined to determine the time evolution of temperature, volume emission measure, and elemental abundances of Fe, Ca, Si, S, Mg, and Ar. The low FIP elements were depleted near the SXR peak which is consistent when lower atmospheric plasma fills the coronal loops through chromospheric evaporation. A key observational signature in flaring plasma is the presence of low-amplitude pulsations known as QPP. These small flux variations could be driven by periodic or "bursty" energy releases and place important constraints on the energy release processes and heating mechanisms. Using multithreaded hydrodynamic models we synthesized SXR and EUV emission to understand the generation of QPP signatures. Results showed that the successive reconnection of the flare loops could generate the ~26 second QPP observed in SXR. In the tranquility of the Sun, the solar corona has minimal flaring events. The MinXSS-2 CubeSat made solar daily quiescent SXR spectra observations and were analyzed to determine the time evolution of temperature, volume emission measure. The average isothermal temperature ranged from 1~2 MK being consistent with other SXR observations during the solar minimum.
The Role of Prostanoid Signaling in Inflammatory Responses Relevant to Diabetic Retinopathy
(2025-03-13) Stark, Amy Kathryn; Penn, John S; Davies, Sean S
Diabetic retinopathy (DR) is the leading cause of blindness in working-age Americans, and its disease burden continues to rise as the global prevalence of diabetes grows. Current therapies for DR address only late-stage disease relevant to retinal neovascularization when irreparable retinal damage from aberrant blood vessel growth has already begun. Herein, we investigated prostanoid signaling as a contributor to the inflammatory responses that occur early in DR pathology to identify novel, targeted therapeutic options addressing disease before the most severe risks to vision ensue. In vitro analyses were performed using primary human Müller glia (hMG), primary human retinal microvascular endothelial cells (hRMEC), and primary human retinal pigment epithelial cells (RPE) or a relevant RPE cell line. After stimulating retinal cells in conditions modeling hyperglycemia, dyslipidemia, and chronic inflammation occurring systemically in patients with diabetes, we found that hMG produced highly elevated levels of one prostanoid, PGE2, whereas hRMEC produced elevated levels of PGF2α most consistently. The receptor-specific effects of prostanoid signaling were then evaluated in cell behavior assays relevant to early DR. Proinflammatory cytokine production in hMG was amplified by both PGE2-EP2 and PGF2α-FP signaling. Leukocyte adhesion to hRMEC was increased by PGF2α-FP signaling. Inner blood-retina barrier function modeled by hRMEC was enhanced by PGE2-EP4 signaling, yet, in contrast, outer blood retina barrier function modeled by cultured RPE was weakened by PGE2-EP2 signaling. Together, these results provide initial characterizations of the inflammatory effects of prostanoid signaling pathways in single cell type behaviors that model early DR progression. Modulation of individual prostanoid signaling pathways via selective agonists and/or antagonists may provide a valuable therapeutic strategy for DR management at the earliest stages of disease.
Molecular and Microenvironmental Drivers of Gastrointestinal Diseases
(2024-12-09) Spencer, Paige Nicole; Lau, Ken; Macara, Ian
The gastrointestinal tract consists of interconnected organs responsible for digestion and nutrient absorption, with the small intestine and colon being the largest components of this system. Although these organs share many similarities, their specialized functions in maintaining homeostasis contribute to varying disease susceptibilities. This dissertation examines the molecular and microenvironmental factors, such as upregulated genes, emergent cell types, and microbial influences, in the pathogenesis of colorectal cancer and Crohn’s disease. In colorectal cancer, the tumor microenvironment, composed primarily of epithelial, immune, and mesenchymal cells, plays a crucial role in disease progression. This work investigated the role of cancer-associated fibroblasts in colorectal tumorigenesis. Through single-cell RNA-sequencing and multiplex immunofluorescence imaging, we characterized cancer-associated fibroblast subpopulations across two mouse models. In advanced lesions, cancer-associated fibroblasts exhibited pro-inflammatory gene expression and were spatially associated with tumor cells. Computational analysis of ligand-receptor interactions revealed significant communication between cancer-associated fibroblasts and tumor cells, suggesting cancer-associated fibroblasts are key drivers of tumor progression. This dissertation also explores how regionalization of the gut influences susceptibility to chronic intestinal inflammation, such as Crohn’s disease. Although the gastrointestinal tract varies in structure and function across regions, the factors driving inflammation in specific areas remain unclear. Using a TNF-overexpressing Crohn's disease mouse model (TnfΔARE/+), we investigated the role of genetic and environmental factors in region-specific disease, focusing on the terminal ileum and ascending colon, the two most affected regions in human Crohn’s disease. We identified Chlamydia muridarum as necessary and sufficient for disease manifestation in the ascending colon, with inflammation driven by goblet cell expression of indoleamine 2,3-dioxygenase (IDO1) in genetically susceptible hosts. Finally, we investigated the role of Paneth cells in terminal ileal inflammation. While Paneth cells upregulate IDO1 in the inflamed ileum, three independent ablation models demonstrate that Paneth cells are dispensable for the development of ileal inflammation in the TnfΔARE/+ model. Overall, these findings provide new insights into the molecular and environmental factors driving colorectal cancer and Crohn’s disease, highlighting the roles of cancer-associated fibroblasts, microbes, upregulation of IDO1, and regional gut differences in disease susceptibility and progression.
Investigating Lipid-siRNA Transport and Gene Silencing Activity in the Rodent Central Nervous System
(2025-03-23) Sorets, Alexander Gustavo; Lippmann, Ethan S; Duvall, Craig L; Lippmann, Ethan S; Schrag, Matthew S
The clinical neurosciences are in the midst a resurgence spurred by the development of new therapeutic modalities. Short interfering RNAs (siRNA) in particular are gaining clinical traction owing to their ability to mediate selective and sustained gene inhibition. Now, a key objective in this quest of developing disease modifying therapies for neurodegenerative disorders is to get siRNA to specific sites and cells in the brain. Direct injection into cerebrospinal fluid (CSF) is the most common route of administration, however, the limited interface between CSF and brain parenchyma poses a considerable challenge for achieving therapeutic effects beyond the superficial layers of the brain. As such, the overarching objective of this dissertation is to enhance delivery to deep brain structures using lipid-siRNA conjugates injected into CSF. We investigated how lipid conjugate structure influences CSF to brain transport mechanisms, regional gene silencing, and cell-specific knockdown. After identifying a suitable candidate in mice, the goal of aim two was to deeply characterize this compound after intrathecal delivery in rats, a model that closely mimics administration in humans. To determine if this construct is truly the most effective, the final aim investigates how structural properties such as linker length and lipid valency influence delivery and knockdown after rat intrathecal injection. Collectively, this work examined properties of lipid-siRNA conjugates that facilitate CSF to brain delivery and generated a promising platform for silencing genes implicated in CNS disorders.
Exploring Mechanisms behind Phase Formations of Metal Chalcogenide Nanoparticles
(2025-03-20) Shults, Andrey Andreyevich; Macdonald , Janet E; Macdonald , Janet E
Our understanding of nanoparticle phase control is currently limited by our knowledge of the mechanisms responsible. How molecular precursors interact with each other, the solvent, and the ligand, and how they decompose thermally can all have an effect on the resulting phase of the product. Through the use of various analytical techniques like nuclear magnetic resonance (NMR), in-situ gas phase Fourier transform infrared (FTIR) spectroscopy, and transmission electron microscopy (TEM), we shed light on the interactions that can occur between ubiquitous reagents in nanosynthesis. In the first project, we demonstrate that thiourea can interact with oleic acid during the synthesis of metal sulfide nanoparticles to produce ammonium thiocyanate, a more reluctant sulfur source, leading to the formation of metastable and sulfur poor nanoparticles, a trend seen across four different metals (Cu2+, Ni2+, Co2+, and Fe3+). In the second project, we use selenourea for the synthesis of iron selenides, and, for the first time, show that chalcogen allotropes can act as phase-determining intermediates in the synthesis of metal chalcogenides. In the presence of oleylamine, selenourea can decompose into red selenium, leading to the formation of Fe7Se8. In the presence of oleic acid, grey selenium forms instead and leads to the formation of FeSe2. In the third project, we make progress towards the synthesis of colloidal chromium telluride nanoparticles. With the use of the fast reacting didodecyl ditelluride, we show that commonly used chromium (III) halide precursors yield only unreactive Te(0). We attempt to overcome this synthetic challenge with the synthesis of a more reactive chromium (II) iodide. Finally, in the fourth project, we use the steric effects of N-heterocyclic derived selenoureas to influence not only the phase of iron selenide nanoparticles but also their morphology. We find that as the sterics of the N-heterocyclic derived selenourea increase, the nanoparticle size decreases, and the shape becomes more irregular. With these reports, we hope to deconvolute the mechanisms affecting phase control and to add onto our synthetic tool-box for repeatable and reliable synthesis of nanoparticles.
STING-Activating Polymer-Drug Conjugates for Cancer Immunotherapy
(2025-02-06) Sheehy, Taylor; Wilson, John; Wilson, John
Cancer remains the second leading cause of death in the United States. Immunotherapies, such as immune checkpoint blockade, are shifting the paradigm of cancer treatment by invigorating the immune system’s natural mechanism of recognizing and eliminating the disease. However, current treatments only benefit a minority of cancer patients. Low response rates are typically correlated to immunologically “cold” tumors, meaning they lack sufficient proinflammatory cell types such as activated macrophages, activated dendritic cells, and CD8+ cytotoxic T-cells. Therefore, there is a need for therapeutic systems that will shift tumors towards an immunogenically “hot” phenotype. One promising strategy is to activate the stimulator of interferon genes (STING) pathway, which triggers a type I interferon (IFN-I)-driven inflammatory response that enhances T cell priming, activation, and tumor infiltration. Although STING agonists are currently explored in clinical trials, most are limited to intratumoral routes of administration, which are not ideal for all patients and tumor types. Small molecule STING agonists, such as GlaxoSmithKline’s diABZI, have been developed for systemic administration; however, these clear rapidly and activate STING indiscriminately, lacking tumor-specific infiltration which can lead to unwanted systemic inflammation and non-optimal efficacy. To address these drug delivery challenges, we have designed a STING-Activating Polymer-drug Conjugate (SAPCon) to enhance pharmacokinetics, biodistribution, and tumor accumulation of a modified STING agonist to enrich activation within the tumor and promote antitumor immunity. By modifying a diABZI molecule for chemical-conjugation to a clickable polymer platform and intracellular drug release through cathepsin cleavage, we demonstrate that drug circulation time and tumor accumulation increase with polymer molecular weight. We also report the unique finding that even non-cleavable diABZI-polymer constructs can activate STING. SAPCon promotes an immunogenic tumor microenvironment, resulting tumor growth inhibition, prolonged survival, and immunological memory in orthotopic breast cancer models. We also employed a liver-targeted, STING siRNA-loaded lipid nanoparticle to knockdown STING activation within the liver to further mitigate off-target activation and reduce serum cytokine release, typically attributed to concerning side effects in the clinic. These results position SAPCon as a modular and programmable platform for improving the efficacy and safety of systemically administered STING agonists for cancer immunotherapy.
Leveraging Electronic Health Record Derived Phenome to Enhance Genetic Studies of Common and Rare Disease
(2025-03-21) Scalici, Alexandra; Cox , Nancy J; Hohman, Timothy
The integration of electronic health record (EHR) derived phenotypes with genetic data has been instrumental in enhancing our understanding of gene-phenotype associations. Vanderbilt University Medical Center has an extensive EHR system that has over 3.5 million individuals that are stored in a de-identified database, the synthetic derivative (SD) that includes ICD billing codes, medications, clinical notes, medical histories, and laboratory values. A subset of the SD, BioVU, is a de-identified EHR-linked DNA repository. Having EHR-linked genetic data has allowed for the development of scalable and accurate phenotyping methods (phecodes), the expansion of analyses beyond genome-wide association studies (GWAS) to identify associations phenome, transcriptome, and laboratory-wide (PheWAS, TWAS, LabWAS). Additionally, methods such as the development of phenotype risk scores (PheRS) have allowed us to disentangle distinct phenotypic patterns of disease and clinical presentation of the complex, multisystem Mendelian diseases as well as enhance our understanding of disease comorbidities. Leveraging these data and methods, we have been able to study common genetic disorders and provide biological context to shared genetic architecture of eye disease by implementing gene and comorbidity-based analyses. These analyses were able to identify established and novel eye disease associations. Additionally, these data and resources have allowed us to study Mendelian diseases on a large scale. We implemented an approach that leverages PheRS to ask questions about Mendelian disease gene function and explore gene-phenotype associations among genes used for clinical diagnostic testing with accurate phenotyping of craniofacial congenital anomalies. The ability to implement these methods to ask questions about both rare and common genetic diseases highlights the diversity and breadth of research questions that can be asked of EHR-linked genetic data.
Maximizing Statistical Efficiency in Clinical Trials: Ordinal Longitudinal Models and Two-Phase Designs
(2025-03-25) Rohde, Maximilian Dimitrios; Harrell, Frank E; Schildcrout, Jonathan S
Clinical trials are the primary method used to establish the safety and efficacy of medicines and interventions. Results from clinical trials form the foundation of medical knowledge and improve the lives of millions of patients. However, clinical trials are often expensive to run, slow to enroll participants, and subject to unforeseen delays -- leading to trials that are underpowered to detect a clinically relevant effect. Optimizing statistical efficiency through the use of modern study designs and analysis methods is therefore an important goal. In this work, we advance two Bayesian methods to improve the efficiency of clinical trials. First, we show how Bayesian ordinal transition models (OTMs) can be used to analyze ordinal longitudinal data, an information-rich type of outcome data that has become common in COVID-19 clinical trials, and develop the statistical theory for these models from first principles. We illustrate OTMs using data from ACTT-1, a clinical trial evaluating the use of remdesivir in hospitalized patients with COVID-19. We then compare OTMs to other commonly used models and approaches for analyzing ordinal longitudinal data through a comprehensive simulation study, and demonstrate that OTMs can greatly improve statistical power. Second, we describe a Bayesian factored likelihood approach to account for missing exposure data in two-phase designs using outcome-dependent sampling and BLUP-dependent sampling. We demonstrate how this approach can improve statistical efficiency compared to simple random sampling through simulations, and provide a case study using data from the Lung Health Study clinical trial. Both of these methods can inform clinical trial design to improve statistical efficiency and advance medical knowledge more quickly.

Browse

Recent Submissions