22689 research outputs found
Sort by
EFFECT OF B CELL MODULATION VIA NON-DEPLETING ANTI-CD19 ANTIBODY ON MYOCARDIAL PHYSIOLOGY
B cells play a vital role in the adaptive immune system, and modulation of B cell signaling has yielded effective therapies for autoimmune diseases and B cell malignancies. Recently, emerging evidence also implicates B cell functions in cardiac physiology and disease progression. In this study, we investigated the impact of targeting CD19 on B cell function and cardiac physiology in C57BL6J mice. In vitro treatment of primary mouse splenic B cells with anti-CD19 monoclonal antibody (mAb) induced internalization of surface CD19 on mouse splenic B cells and led to a reduction in B cell proliferation. For in vivo analyses, C57BL6/J mice received intraperitoneal (i.p.) injection of the same antibody and were evaluated at 18- and 40-week post-treatment. Male mice exhibited a sex specific increase in cardiac output, concomitant with a decrease in circulating B220⁺CD23⁺ B cells at 18 weeks and a rise in peripheral neutrophil counts at 40 weeks. Female mice also displayed an elevated neutrophil population at the 40-week timepoint. Proteomic profiling of cardiac tissue showed that chronic treatment with anti-CD19 mAb modulated the expression of hundreds of myocardial proteins, with sex specific features. In males, Mcfd2, Myl4, and Nppa showed marked increased expression in response to treatment, whereas in females, Igkc expression was remarkably downregulated. Collectively, these findings demonstrate that anti-CD19 mAb-mediated CD19 internalization suppresses B cell proliferation in vitro. Chronic in vivo administration of the same antibody leads to sex specific changes in peripheral blood immune cell profile, cardiac function, and cardiac protein expression, underscoring a novel link between B cell modulation and cardiac function
Investigating antigen diversification in Trypanosoma brucei through mosaic Variant Surface Glycoprotein (VSG) formation
Trypanosoma brucei utilizes antigenic variation to evade the host immune system through the switching of its dense Variant Surface Glycoprotein (VSG) coat. The T. brucei genome encodes a repertoire of ~2000 VSG-encoding genes, but the majority of these genes are pseudogenes or fragments. To expand the usable VSG repertoire and maintain a chronic infection, novel mosaic VSGs can be formed through recombination of two or more VSG genes. Our lab has previously shown that DNA damage can trigger mosaic VSG formation. These experiments involved targeted double stranded DNA breaks via inducible Cas9 at various sites along an expressed VSG, AnTat1.1. Comparison of mosaic formation rates between these sites highlighted a significant increase in mosaic formation at position 694, indicating a possible “recombinational hotspot”. Here, we investigate whether Cas9 cutting efficiency with different guides could have an impact on this potential hotspot. Investigations into the intensity of DNA damage and parasite survival rate with different guides have allowed us to conclude that cutting efficiency does not explain the increase in mosaic VSG formation at position 694.
This process of mosaic VSG formation requires sequence homology between the expressed VSG and the donor VSG that is used as a repair template after a DNA break. Little is known about the exact homology requirements for mosaic formation, including the minimum length of homology required for a VSG gene to be recognized as a potential donor sequence. To investigate these homology requirements, we developed truncated forms of a donor VSG that is commonly utilized around position 694 of AnTat1.1, limiting the homology available for donor identification. The truncated constructs were stably integrated into parasites expressing AnTat1.1 and dsDNA breaks were induced at the common recombination site to promote mosaic VSG formation. We then used anchored multiplex PCR sequencing (AMP-seq) to evaluate the minimum sequence homology required to form a mosaic VSG. In the data presented here we were able to conclude that T. brucei can utilize homologous donor VSG sequences at least as short as 500 bp. Both projects presented here provide important insight into this antigen diversification mechanism used by T. brucei
SOCIAL AND PSYCHOLOGICAL DRIVERS OF CARDIOVASCULAR HEALTH AND HYPERTENSION IN WOMEN OF REPRODUCTIVE AGE
Background: Cardiovascular disease (CVD) is the leading cause of death among women in the U.S., with hypertension as its primary modifiable risk factor. Women of reproductive age face rising hypertension rates driven by social, psychological, and behavioral factors. However, research has largely overlooked how cumulative social and psychological risks shape cardiovascular health in this population, particularly among racially and socioeconomically diverse women
Objective: This dissertation examines how social and psychological risk factors contribute to cardiovascular health disparities among women of reproductive age.
Methods: This explanatory sequential mixed-methods study draws on data from the SAFE Heart Study, which recruited women aged 18-50 from Baltimore, Washington, D.C., and the national Research Goes Red registry. A systematic review identified research gaps on social determinants of hypertension. The association between cumulative social risk and cardiovascular health was examined in both local and national samples. Psychological distress (depression, anxiety, and stress) was analyzed among locally recruited participants. Semi-structured interviews explored lived experiences of social, psychological, and cardiovascular risks.
Results: The systematic review revealed limited studies on cumulative social risk and hypertension among women. Higher cumulative social risk was linked to increased hypertension, diabetes, and hyperlipidemia, with a stepwise association between social disadvantage and cardiovascular risk. Psychological distress independently predicted cardiovascular risk factors. Qualitative findings highlighted financial strain, structural barriers, and stress as key challenges to cardiovascular prevention.
Conclusion & Implications: Social and psychological risk factors significantly impact cardiovascular health, particularly hypertension, in women of reproductive age. Integrating mental health screening and referral when needed into cardiovascular assessments and addressing structural barriers through policy and community interventions is critical to reducing disparities. Future research should explore long-term impacts and targeted interventions to improve cardiovascular outcomes in this high-risk populatio
te ‘ori tahiti e te ‘ahu: Public Performances of Tahitian Dance and Costuming Innovations During the Height of French Colonialism in Tahiti, 1880-1910
In Tahiti in 1956, Madeleine Moua founded the Heiva dance troupe and made bold costuming changes, replacing the modest Western clothes worn throughout the last century of dance. Historians and cultural scholars attribute this moment as one of the central components that kickstarted the indigenous mā’ohi cultural revival of the second half of the 20th century. What do scholars miss in skipping over the cultural innovations exhibited during the colonial period? While Moua’s contributions to Tahitian dance attire and public performance are undoubtedly important, the notion that Moua was the first to lay claim to “authentic” indigenous dance practices in the “post-colonial” period obscures the creative cultural activism of the early colonial period from 1880-1910. This thesis makes this visible by examining Tahitian dance costuming, the externalization of Tahitian dance as a symbol of cultural and political identity, and the public performance of various types of ‘ori tahiti (Tahitian dance) at a time when it was not perceived as decent by authorities, many indigenous elites, and even lower-class locals. Building on debates about the invention of tradition (Hobsbawm, Linnekin, Plant, etc.) and anthropological work on Tahiti and Hawai’i (Tcherkézoff, Imada, etc.), this research problematizes the scholarly assumption that Tahitian material innovation and public dancing did not occur under French colonial rule at the turn of the 20th century. Instead, this thesis demonstrates how the Fête de Tiurai helped popularize Tahitian dance and costuming innovations decades before the dance and cultural revivals of the 1950s-1970s. Additionally, while Hawai’ian dance performances abroad were more frequent in the period from 1880-1910, a one-of-a-kind Tahitian dance troupe from 1906 centers Tahiti in a field where the mā’ohi are often lumped in with the sexualization of Hawai’ian hula or the pan-Polynesian experience. Despite Western feminization of Polynesian cultures and the theatricalization of cultural performances, Tahitians at the turn of the 20th century made their mark on their indigenous cultural traditions, their innovations paving the way for Moua and cultural “revivals” decades later
The Social-Medical Network: The Integration of Personal and Professional Healthcare Networks for Older Adults
Older adults represent a substantial and growing proportion of the population. Because of this, we have also seen an increase in the demand on the healthcare system and on family and friend caregivers. These two groups of individuals providing care to older adults create healthcare networks. The social-medical network is the larger network comprised of the personal and professional healthcare subnetworks.
To investigate the social-medical network, we conducted a cross-sectional observational pilot study that included a quantitative survey and social network inventory. The social network inventory included three sections: 1) Personal healthcare network; 2) Professional healthcare network; and 3) Integration of the social-medical network. Our population of interest was community-dwelling older adults (65 years or older) living with Type 2 diabetes in Maryland. Over a period of two years, we enrolled 128 individuals and completed data collection with 122 participants. We assessed the relationship between emergency department visits and network-level constructs, adjusting for demographic and social covariates of interest.
We found that for each additional professional alter providing care for five years or more, older adults had an inverse relationship with the incidence rate of emergency department visit. Though not marginally associated, this indicates a potential protective effect of longer care relationships. We also quantified the relationship between emergency department visit and integration of the social-medical network. For each 1-unit increase in social-medical network integration, older adults with two or more personal alters had an inverse relationship with the incidence rate of emergency department visit. The direction of this relationship was reversed for older adults with only one personal alter. Although both results were not marginally associated, these findings indicate that social-medical network integration may be important in terms of utilization outcomes and could have different impacts among subgroups.
Given that this was a pilot study, these relationships require further confirmatory inquiry. For older adults living with Type 2 diabetes, care relationships and integration of family and friends with healthcare providers may prove to be advantageous in protecting and promoting health and well-being
IMPROVING THE FAIRNESS IN LANGUAGE MODELS FOR DECISION-MAKING
Methods that seek to improve the fairness of language models for prediction tasks involve collecting annotated data with specific demographic attributes and applying loss functions designed to ensure equitable predictions. While these approaches provide a foundation, they often fall short in addressing the intricate realities of data distributions in the wild and their consequences.
As a motivating example of the nuances of biases in the performance of language models, we consider the use of language models in the mental health space to aid practitioners in the diagnosis of depression. We find that models trained on social media data for depression diagnosis are unfair towards gender and racial underrepresented groups (Men and African American/Hispanic groups.) We dive into the data and investigate how the fairness, protected attributes and the depression diagnosis task have complex relationships underscoring the need for more robust fairness methods.
Current fairness methods typically involve training models on annotated data using objectives that utilize demographic labels. This requirement is often impractical or not possible due to, e.g. privacy concerns in healthcare. We are inspired by task adaptation methods and develop a multi-task method to improve the fairness of models when demographics are not available.
Exploring more recent frameworks for prediction systems, we focus on large language models (LLMs) for predictive tasks. This represents a different paradigm for traditional fairness methods as they rely on updating model parameters, which is impractical due to the large size of LLMs or impossible as they are more often proprietary. We investigate the use of demonstrations for LLMs via In-Context Learning (ICL), finding that the choice of demonstrations matters, but concernedly, there is not a single method that is consistently better across models and datasets.
Further, as prediction models are applied to increasingly complex tasks, how fairness is conceptualized and implemented must also be updated. We propose that fairness definitions and objectives should account for underlying relationships between protected attributes and the predicted label or text inputs. We show that applying fairness methods while failing to take these relationships into account may result in models that are less fair and perform worse than if no fairness objective was applied.
Improving the fairness of prediction models for language is an ongoing challenge that requires assessing the data and model predictions in practical scenarios, addressing challenges that arise in practice and in applying new frameworks, while considering the relationship of fairness definitions, protected attributes and the task at hand. This thesis contributes to the improvement of equity and effectiveness of language models as prediction systems
GENETIC AND PHARMACOLOGIC MODELS TO ASSESS THE CONTRIBUTION OF HEPATIC FATTY ACID OXIDATION
The body is constantly adjusting to maintain homeostasis in response to innumerable outside factors, two being temperature fluctuations and metabolic fluctuations. The liver is a major contributor and coordinator of the body’s response to multiple metabolic stressors, including a prolonged fast and chronic high-fat feeding. Deleting hepatic fatty acid oxidation then, allows for the understanding of peripheral contributions to maintaining homeostasis during a prolonged fast. We have worked with a model to suppress hepatic fatty acid oxidation by targeting the two key enzymes, carnitine palmitoyltransferase 1 (Cpt1) and carnitine palmitoyltransferase 2 (Cpt2), in a liver-specific manner. Cpt1 and Cpt2 are located on the outer and inner mitochondrial membrane, respectively, and facilitate fatty acid import into the mitochondrial matrix for β-oxidation. Unexpectedly, the two genotypes had vastly differed hepatic transcriptional responses due to a full deletion of hepatic Cpt1 not being a full deletion of hepatic fatty acid oxidation in an in vivo system. There are exogenous pools of long-chain acylcarnitines from peripheral tissues that can traverse into the hepatocyte and feed into β-oxidation. Other than genetic models to understand the contribution of hepatic fatty acid oxidation, there are pharmacological means to suppress fat oxidation. Etomoxir remains the most utilized small molecule inhibitor of fatty acid oxidation, but there is much question regarding its specificity. Using an activity-dependent probe and click chemistry, we show that etomoxir is indeed a non-specific Cpt1 inhibitor and identify a whole host of proteins that it interacts with. Surprisingly, there are a large number of peroxisomal proteins and even ER-resident proteins that were bound by etomoxir, showing the need for more rigorous experiments when it comes to utilizing this small molecule inhibitor. By utilizing both genetic and pharmacological inhibitors of fatty acid oxidation, we are able to provide a much fuller understanding of hepatic fatty acid oxidation and the whole-body response coordinated by the liver
METAL ORGANIC FRAMEWORK (MOF) THIN FILM SYNTHESIS USING LIQUID BASED DEPOSITION
Metal organic frameworks (MOFs) have emerged as promising materials for diverse appli- cations including gas storage, separation, catalysis, and sensing due to their exceptional porosity, structural diversity, and tunable functionality.
Metal-organic frameworks (MOFs), particularly in patterned forms, are emerging as promising materials for microelectronic applications due to their functionality as metal-rich photoresists in next-generation lithographic processes. Among these, amorphous zeolitic imidazolate frameworks (aZIFs) offer notable advantages by avoiding issues related to crystal anisotropy and grain boundary formation—common factors that often degrade surface uniformity and pattern fidelity.
This thesis investigates the synthesis of MOF thin films via liquid deposition tech- niques such as flow coating, spin coating and speciation modelling, focusing on optimizing processing parameters to achieve precise control over film thickness, crystallinity, and mor- phology. Through systematic experimentation and with different metal-ligand system we try various precursor concentrations, deposition conditions, and post-synthesis treatments, we developed reproducible protocols for fabricating uniform and well-adhered MOF thin films on silicon wafer substrate materials. Advanced characterization methods including Atomic force microsocpy(AFM), scanning electron microscopy(SEM), and XPS measure- ments were employed to evaluate film quality and growth kinetics. Our findings demon- strate that modulating precursor choice and concentrations and flow reactor configuration significantly enhances film orientation and integrity. Additionally, we completed a detailed study on the analysis of active chemial species present in the solution during deposition that leads to film growth.The optimized thin films exhibited excellent stability and were used further to investigate their utility in metal containing resist. This work contributes valuable insights into the liquid-phase fabrication of MOF thin films and establishes a ver- satile platform for their integration into functional devices for environmental and energy applications
THE EFFECTS OF MODULATING VEGF-C SIGNALING PATHWAY ON MENINGEAL LYMPHATIC CLEARANCE FUNCTION IN A MOUSE MODEL OF HUNTINGTON’S DISEASE
The brain’s glymphatic system and its meningeal lymphatic vessels (mLVs) have emerged
as critical routes for clearing metabolites and proteins implicated in neurodegeneration. To
test whether manipulating mLVs alters Huntington’s disease (HD) pathology, we modulated
vascular endothelial growth factor-C (VEGF-C) signaling in zQ175-DN knock-in mice.
Three-month-old HD and wild-type littermates received intra-cisterna magna injections of
either AAV9-VEGF-C to expand, or AAV9-VEGF-C/D Trap to ablate, mLVs. VEGF-C enlarged
mLV area and enhanced cerebrospinal fluid (CSF) drainage to deep cervical lymph nodes,
whereas VEGF-C/D Trap reduced mLV coverage and decreased glymphatic efflux. VEGF-C/D
Trap–treated HD mice also developed earlier motor deficits on balance-beam and
accelerating-rotarod tests, while VEGF-C did not affect performance. Baseline mLV density
was similar in untreated HD and wild-type mice at this age, suggesting that meningeal
lymphatic pathology manifests later in disease progression. These findings show that intact
mLVs are required for efficient brain waste clearance and that impaired mLVs accelerates
functional decline in HD, underscoring the glymphatic-lymphatic interface as a potential
therapeutic target
EVALUATING TWO LOCATION-BASED STRATEGIES FOR IMPLEMENTING COMPUTER-AIDED DETECTION FOR TB SCREENING IN PERI-URBAN UGANDA
Abstract
Tuberculosis (TB) remains a leading cause of mortality globally despite the availability of effective diagnostic and treatment tools. This dissertation evaluates two location-based strategies—community-based and facility-adjacent—for implementing computer-aided detection (CAD) chest X-ray screening for TB in peri-urban Uganda. Conducted within a hybrid implementation-effectiveness trial, the study explores three implementation outcomes: initial reach and coverage, time investment from participants and staff, and health system costs.
Methods
The trial was implemented across eight peri-urban regions near Kampala, Uganda. Participants were screened using portable digital chest X-rays analyzed by CAD software. Those eligible received follow-up diagnostic testing for active TB or latent TB infection (TBI). Data collection included participant surveys, structured observation for time and motion, and micro-costing using a time-driven activity-based costing framework.
Results
Reach and Coverage
Community-based screening reached more men (53% vs. 32%) and individuals living in poverty (18% vs. 10%) than facility-adjacent screening. While 91% of eligible participants submitted sputum samples and 80% initiated treatment for active TB, the prevention cascade had major drop-offs: only 13% received TST placement, and 55% returned for TST reading. Barriers included cost, travel, and low perceived need.
Time Investment
Three-member teams spent 65 hours per positive Gene Xpert result and 25 hours per positive TST reading. Median participant visit time was 20 minutes. Community-based sites had steadier participant flow, while facility-adjacent sites saw early morning peaks.
Costs
Cost per positive Gene Xpert result was 1,409 (facility). TST reading costs were 665, respectively. Costs were most sensitive to TB prevalence and fixed-cost resource utilization.
Conclusions
Community-based screening was more efficient and better at reaching under-screened populations. While CAD-enabled TB screening is resource-intensive, results provide actionable insights for optimizing staffing, budgeting, and targeting in future programs. This research highlights the value of hybrid trials for identifying critical implementation determinants to guide future TB control efforts