Alliance One Tobacco (Malawi)
Academic Research Repository at the Institute of Developing EconomiesNot a member yet
52305 research outputs found
Sort by
Advancing Chemical Transport Modeling for Air Quality, Satellite Retrievals, and the Clean Energy Transition
No full textTropospheric oxidant chemistry affects air quality by controlling the formation pathways of air pollutants. It also determines the atmospheric lifetime of key greenhouse gases (GHGs) such as carbon dioxide (CO2) and methane (CH4), as well as indirect GHGs like hydrogen (H2). GEOS-Chem is a state-of-the-science atmospheric chemistry model that represents our current understanding of tropospheric oxidant chemistry. The GEOS-Chem chemical transport model (CTM) is used to support the satellite retrievals of air pollutants like nitrogen dioxide (NO2) and to assess the warming potential of GHGs.
Recent advances in satellite observations of air pollutants (e.g., NO2) have emerged with the launch of three geostationary satellites: GEMS (2020), TEMPO (2023), and Sentinel-4 (2025). GEMS is the first geostationary satellite that provides hourly NO2 data over East Asia, rather than just one observation per day. As a result, improving our understanding of geostationary satellite retrievals and interpreting hourly data observed from it is an important task.
In this work, we first evaluate the ability of GEOS-Chem to accurately simulate the tropospheric oxidant chemistry over East Asia by comparing model output with an extensive suite of aircraft measurements from the KORUS-AQ campaign. Following this validation, we use GEOS-Chem vertical profiles to support geostationary satellite retrievals and investigate how diurnal variation in NO2 profiles affects hourly NO2 satellite retrievals (Chapter 1).
Next, we examine how the diurnal variation in column NO2 observed by geostationary satellite differs from that measured in surface NO2 measurements. We leverage GEOS-Chem’s ability to separate the effects of chemistry, transport, and emissions to interpret the observed NO2 variation (Chapter 2). Lastly, during the KORUS-AQ aircraft campaign, we identified a discrepancy between observed and modeled concentrations of methyl hydroperoxide (CH3OOH), which is unexpectedly elevated over the Seoul Metropolitan Area. GEOS-Chem fails to reproduce this behavior. We show that measurement interference from methanediol, a chemical species formed via in-cloud hydration of formaldehyde, may explain part of this discrepancy. We also explore the role of methanediol in oxidant chemistry and formic acid formation (Chapter 3).
While air quality is important to human health, transitioning to cleaner energy is also essential to mitigate climate change. The Intergovernmental Panel on Climate Change (IPCC) recommends achieving net-zero anthropogenic CO2 emissions by 2050 to keep global warming to 1.5 ◦C. One proposed solution is switching from fossil fuels to hydrogen. However, hydrogen emissions can affect atmospheric abundances of methane, ozone, and water vapor, making hydrogen an indirect GHG. The global warming potential (GWP) is a commonly used metric to evaluate the climate impact of GHGs. Previous studies using models have shown that soil sink is the largest uncertainty in estimating its GWP. However, current models have known biases in their simulations of OH concentration and reactivity, and how these biases affect the evaluation of hydrogen global warming potential has not been considered. We find that these biases lead to a 20% overestimate in the GWP of hydrogen (Chapter 4).Engineering and Applied Sciences - Engineering Science
Real-World Evidence in Oncology Using Reference Trial Emulation Across Multiple Electronic Health Record Databases
No full textBackground: Oncology specialty electronic health record (EHR) databases are increasingly used to generate real-world evidence (RWE) in prognostic modeling and comparative effectiveness research. These data sources offer the potential to complement randomized clinical trials (RCTs) by reflecting routine clinical practice and including more heterogeneous patient populations. However, their use remains challenged by incomplete capture of key prognostic variables, missing data, and heterogeneity across databases. As a result, it remains uncertain under what conditions oncology EHR data are sufficiently complete and reliable to support valid prognostic assessment and causal inference. Benchmarking real-world analyses against RCTs has been proposed as an approach for assessing fitness-for-purpose of a given real-world data source for specific, closely related comparative effectiveness questions.
Objectives: The objectives of this body of work were to: (1) evaluate the generalizability and performance of a computable prognostic score for overall survival (OS) across multiple oncology specialty EHR-derived databases; and (2)/(3) explore the extent to which multiple oncology specialty EHR-derived data can replicate OS treatment effects observed in two oncology RCTs.
Methods: First, we evaluated the performance of the Real-wOrld PROgnostic (ROPRO) score, a multivariable prognostic model derived from EHR-derived data, across four independent oncology EHR databases covering multiple cancer types and disease settings. ROPRO scores were computed using routinely collected demographic, clinical, and laboratory variables, with missing covariates imputed using random forest-based methods. OS was modeled using Cox proportional hazards models, and performance was assessed using discrimination and calibration metrics.
Second, we conducted two exploratory comparative effectiveness studies emulating the MONARCH-3 and MONALEESA-2 RCTs, which evaluated cyclin-dependent kinase (CDK) 4/6 inhibitors combined with endocrine therapy as first-line treatment for hormone receptor-positive, human epidermal growth factor receptor 2 (HER2)-negative metastatic breast cancer. Across three oncology specialty EHR-derived databases, we aligned eligibility criteria, treatment definitions, and follow-up with the reference trials. OS was the outcome of interest. Missing data were addressed using multiple imputation by chained equations, and confounding was controlled using 1:1 nearest-neighbor propensity-score (PS) matching. Databases achieving covariate balance after PS matching were included in the primary analysis, which included a fixed-effects inverse-variance meta-analytic approach. All analyses were conducted under preregistered protocols and were considered exploratory due to limited statistical power.
Results: The ROPRO demonstrated consistent and generalizable prognostic performance across databases and cancer types when all or nearly all variables were available, with moderate-to-good discrimination and overall adequate calibration. Performance deteriorated in a database with limited variable availability, highlighting the sensitivity of prognostic modeling to data completeness.
In the MONARCH-3 emulation, two databases met inclusion criteria for the primary analysis, and the pooled OS hazard ratio (HR) was closely aligned with the randomized trial estimate, although database-specific estimates varied substantially in magnitude and direction. In the MONALEESA-2 emulation, only one database achieved covariate balance; the resulting effect estimate was aligned with the trial result, while additional databases were informative only in sensitivity analyses. Across both emulations, pooled estimates generally approximated trial findings, but uncertainty was substantial and individual databases produced heterogeneous results.
Conclusions: Collectively, these studies demonstrate that oncology EHR-derived real-world data can, under certain conditions, support valid prediction and approximate randomized trial estimates for OS. However, fitness-for-purpose is highly context specific and depends critically on data completeness, mortality capture, confounder availability, and line-of-therapy curation. The ROPRO score was generalizable across datasets when key variables are well captured, while confidence in real-world comparative effectiveness analyses may be increased through benchmarking against closely related RCTs. These findings underscore the need for pre-specified study design and analysis, transparent reporting, and database- and question-specific diagnostics to support credible prediction and causal inference in oncology RWE.
: First, we evaluated the performance of the Real-wOrld PROgnostic (ROPRO) score, a multivariable prognostic model derived from HER-derived data, across four independent oncology EHR databases covering multiple cancer types and disease settings. ROPRO scores were computed using routinely collected demographic, clinical, and laboratory variables, with missing covariates imputed using random forest-based methods. OS was modeled using Cox proportional hazards models, and performance was assessed using discrimination and calibration metrics.
Second, we conducted two exploratory comparative effectiveness studies emulating the MONARCH-3 and MONALEESA-2 RCTs, which evaluated CDK4/6 inhibitors combined with endocrine therapy as first-line treatment for hormone receptor-positive, human epidermal growth factor receptor 2 (HER2)-negative metastatic breast cancer. Across three oncology specialty EHR-derived databases, we aligned eligibility criteria, treatment definitions, and follow-up with the reference trials. OS was the outcome of interest. Missing data were addressed using multiple imputation by chained equations, and confounding was controlled using 1:1 nearest-neighbor propensity-score (PS) matching. Databases achieving covariate balance after PS matching were included in the primary analysis, which included a fixed-effects inverse-variance meta-analytic approach. All analyses were conducted under preregistered protocols and were considered exploratory due to limited statistical power.
Results: The ROPRO demonstrated consistent and generalizable prognostic performance across databases and cancer types when all or near-all variables were available, with moderate-to-good discrimination and overall adequate calibration. Performance deteriorated in a database with limited variable availability, highlighting the sensitivity of prognostic modeling to data completeness.
In the MONARCH-3 emulation, two databases met inclusion criteria for the primary analysis, and the pooled OS hazard ratio (HR) was closely aligned with the randomized trial estimate, although database-specific estimates varied substantially in magnitude and direction. In the MONALEESA-2 emulation, only one database achieved covariate balance; the resulting effect estimate was aligned with the trial result, while additional databases were informative only in sensitivity analyses. Across both emulations, pooled estimates generally approximated trial findings, but uncertainty was substantial and individual databases produced heterogeneous results.
Conclusions: Collectively, these studies demonstrate that oncology EHR-derived real-world data can, under certain conditions, support valid prognostic modeling and approximate randomized trial estimates for OS. However, fitness-for-purpose is highly context specific and depends critically on data completeness, mortality capture, confounder availability, and line-of-therapy curation. The ROPRO are generalizable across datasets when key variables are well captured, while comparative effectiveness analyses may benefit from routine benchmarking against closely related RCTs. These findings underscore the need for pre-specified study design and analysis, transparent reporting, and database- and question-specific diagnostics to support credible causal inference and prognostic assessment in oncology RWE.Population Health Science
From Spatial Spread to Sexual Networks: Leveraging Infectious Disease Epidemiology Tools to Monitor and Respond to Strep Throat and Gonorrhea
No full textAt a moment of intense divisions and distrust, it remains critical to highlight the ways scientific tools contribute positively to society and improve public health. This work shows how three distinct methods within infectious disease epidemiology applied to answer different questions produce results that can inform public health guidance and translate to improvements in the health of individuals. First, we show that spatiotemporal analysis of strep throat visits data in the U.S. identifies geographical and seasonal trends and points to key demographic groups that may drive transmission. Next, we show how mathematical modeling can inform public health guidance on the rollout of new antibiotics to treat gonorrhea in the face of rising antimicrobial resistance. Finally, we show that combining mathematical modeling with pathogen genomic analyses can inform gonorrhea surveillance strategies to yield insights on sexual network structures. These findings together emphasize the important role infectious disease epidemiology plays in preventing people from suffering from illness, and in continuing to interrogate the natural world.Population Health Science
Guarding Against Heedlessness: Time and Temporality Among the Early Ottomans
No full textThis dissertation explores the multifaceted nature of early Ottoman temporal culture, arguing against its characterization as a monolithic entity. Temporal culture here refers to the system of practices, beliefs, and attitudes through which Ottomans experienced time in their everyday lives as well as the cosmic sense shaped by religious frameworks. I argue ultimately that Ottoman temporal culture was characterized by a bimodal approach, balancing intersecting frameworks of time. The critical distinction I aim to draw is that bimodal, unlike binary, dichotomy, or duality, implies a more nuanced scenario in which the two parts are not necessarily opposites; there is room for overlap.
Chapter One sets the groundwork by challenging the notion of modern time as wholly divorced from natural phenomena, using the regulation of atomic clocks by leap seconds as an example. This contextualizes the history of temporal systems, focusing on variable “temporal hours” and fixed “equal hours.” The chapter argues that the Ottomans employed both systems pragmatically, illustrating that temporal (also known as seasonal) hours—aligned with natural cycles of light and dark—were not less sophisticated than equal hours associated with mechanical clocks. While clocks were present in Ottoman society from the late sixteenth century, their adoption of equal hours cannot solely be attributed to technological determinism. This chapter underscores how temporal practices served diverse societal needs.
Chapter Two examines the Ottoman use of multiple calendars, introducing the concept of “calendrical pluralism” to describe the coexistence of systems like the Hijrī, Rūmī, and Folk calendars. Each calendar had distinct origins, purposes, and contexts, reflecting the multicultural and syncretic nature of medieval Anatolian society where such calendrical pluralism could flourish. The Ottoman Rūmī calendar, derived from the Julianized Seleucid Era, was widely used well before its official adoption in 1677 and was influenced by Syriac and Greek traditions. Additionally, the Folk Calendar, linked to agricultural cycles and festivals such as Ḫıḍrellez, exemplifies the integration of Mediterranean seasonal divisions into Ottoman temporal culture. This chapter demonstrates how these calendars operated along various bimodals, such as solar-lunar and seasonal-aseasonal, emphasizing the adaptability and complexity of Ottoman timekeeping.
Chapter Three highlights the Hijrī calendar and its development as a purely lunar system. The calendar’s aseasonal nature is interpreted as a spiritual admonition against attachment to the transient, earthly realm, aligning with Islamic eschatological themes. The chapter also analyzes the Qurʾanic absence of the term zamān, often used to denote abstract time, and instead highlights terms like ʿaṣr and dahr, emphasizing the timelessness of the afterlife. Drawing on Sufi teachings, the chapter introduces the figure of the ibn-i vaḳt (son of time), who embodies the spiritual ideal of living in the eternal present. For ordinary believers, however, this ideal was tempered by anxieties about the temporal world and the afterlife, a theme further explored in subsequent chapters.
Chapter Four shifts to popular Ottoman literature, examining works by Aḥmed Bīcān and Yazıcıoğlu Meḥmed, who synthesized Sufi mysticism and Hanafī orthodoxy into vernacular texts for recent converts to Islam. Alongside these, the chapter considers melhemes (meteorological prognostication manuals), which combined cosmic prognostication with spiritual comfort, reflecting the shared goal of addressing temporal and eternal anxieties. The figure of Ḫıḍır and the festival of Ḫıḍrellez serve as case studies of Ottoman syncretism, integrating Islamic, Christian, and local traditions into a cohesive temporal-spiritual framework.
Finally, Chapter Five explores early Ottoman historiography, particularly historical almanacs embedded in astrological texts. These almanacs, employing a reverse-dating system, linked past events to cosmological patterns, aligning historical narratives with Ottoman state-building ambitions. By situating the empire within a purposeful chronology, these texts echoed the Sufi emphasis on the present as the intersection of the past and future. This chapter demonstrates how state-centered histories paralleled individual spiritual frameworks in providing continuity and meaning.
This dissertation does not claim to uncover all Ottoman temporal bimodals but instead highlights how these intersecting frameworks resisted the formation of a unified temporal culture. By examining calendrical, spiritual, meteorological, astrological, and historical traditions, it reveals how Ottoman temporality defied simple categories and instead inhabited multiple bimodal frameworks, offering insights into broader questions of time, culture, and meaning in the premodern world.Middle Eastern Studies Committe
Leasing Out Sovereignty; The proliferation of Chinese Surveillance Technologies in Africa
No full textIn the everyday lexicon, digital technology is simply an instrument, a functional conduit that automatically realizes ends. Contrary to this supposition, we can explore digital technology as a political and cultural artifact that embodies values embedded in postcolonial societies. This dissertation project, based on ethnographic, documentary, and archival research, discusses digital technology's functional architecture and how it operates as the preeminent artifact to envision developed futures. The project investigates how Chinese surveillance technology has proliferated throughout Africa, particularly in Kenya, and examines the technological fix as an aspiration for improving the capabilities of the postcolonial state to bolster social order and spur economic development.
The project's principal goal is to examine the effects of digital technology on (i) state sovereignty and the governance of local populations, including the management of crime and the enforcement of economic regulation; (ii) civil society, concerning the sense of in/security, and of un/freedom it instills in citizens; and (iii) patterns of social and material inequality. What, further, are its hidden, unintended consequences in Kenya, and what particularities, if any, emerge from its postcolonial context? And, finally, what does the spread of Chinese digital infrastructure, as exemplified by the case study here, augur for China's future in Africa? These questions have significant implications for Africa and the global order at large. This project seeks to resolve these questions by focusing on the procurement and use of digital surveillance technology.African and African American Studie
Changes in the Sky: The Rise and Fall of Weather Control in the Twentieth-Century United States
No full textThis dissertation investigates American efforts to control the weather in the mid-twentieth century, and the many debates and controversies that those efforts provoked. Focusing on the 1940s to 1970s, it traces the rise and fall of cloud seeding: a controversial technology in which people put chemicals into clouds, in hopes of controlling everything from rainfall to sunshine, hurricanes to drought, local seasons to the global climate. Developed in 1946, cloud seeding quickly spread across the United States and the world. From booster development projects in the Southwest to top-secret climate warfare in Vietnam, cloud seeding provoked fierce debates—political, legal, scientific, ethical, environmental—about the dynamic, shifting relationship between people and the skies. Some believed that cloud seeding would bring a climate utopia. Others feared a climate apocalypse. And still others maintained that weather control was and always would be impossible. By the late 1970s, weather control seemed more elusive—and controversial—than ever before. The national preoccupation peaked and began to wane.
Ultimately, this dissertation argues that the rise and fall of weather control constituted both the apotheosis and failure of modern, secular ideas about controlling nature. Grounded in archival research, it combines historical and cultural studies with theoretical frameworks from the environmental humanities, science and technology studies (STS), political theory, and religious studies. To trace the hype, ambivalence, and conflict provoked by weather control, it weaves together newspapers and popular media; government, corporate, scientific, and legal records; and the personal papers of figures involved. Through these archives, it examines how the weather became an object of rationalization (through the production of knowledge, maps, experiments), militarization (through new forms of weaponry and warfare), capitalization (through agricultural and economic development), legalization (through new laws, regulations, court cases, and property regimes), and politicization (through postwar liberalism and conservatism in the U.S., as well as capitalist-democratic and communist ideologies abroad). Together, these processes can be understood as the attempted secularization and modernization of the sky. This dissertation explores how and why they failed.
In addition to its historical significance, this research holds contemporary relevance, as climate change and geoengineering have sparked fierce debates over the wisdom of controlling the skies through technology. By narrating the strange and understudied history of weather control, “Changes in the Sky” contributes to critical, interdisciplinary studies on the politics of nature and the ethics of technology in the twentieth century, and today.American Studie
Toward Personalized Stroke Rehabilitation in the Community With a Mobile FES Neuroprosthesis
No full textAs the field of lower-limb assistive technology advances, there is growing interest in shifting gait rehabilitation from controlled clinical environments to community-based settings, aiming to enhance motor recovery through increased training dosage, intensity, and ecological validity. Functional electrical stimulation (FES) neuroprostheses targeting plantarflexion have emerged as promising tools to support gait rehabilitation by activating biological muscles directly, thus promoting neuroplasticity and yielding improvements that persist beyond active stimulation. However, existing FES systems primarily rely on preset stimulation patterns and controlled clinical conditions, limiting their real-world applicability, adaptability, and long-term effectiveness.
To address these limitations, this thesis presents the development and evaluation of a mobile FES neuroprosthesis designed to provide personalized push-off propulsion and ground clearance assistance, two deficits widely observed in individuals post-stroke. The mobile platform enables adaptive control based on biomechanical responses and user preference, supported by wearable sensing frameworks to enable longitudinal monitoring and tuning in community settings.
The first aim introduces the mobile FES system and demonstrates its capability to provide targeted stimulation during overground walking with stroke survivors. Through a series of technical pilot studies, we highlight critical design considerations for community deployment, including coordinated stimulation of dorsiflexion and plantarflexion, personalized control using biomechanical inputs, and management of physiological phenomena such as electromechanical delay and fatigue. Using this system, we explored the biomechanical effects of overground walking with FES, finding both immediate and short-term retained benefits, highlighting its potential to meaningfully enhance daily function and quality of life. Notably, optimal stimulation timing varied considerably between individuals, underscoring the necessity for personalized tuning to maximize biomechanical improvements and avoid exacerbating gait impairments.
Building on these findings, the second aim investigates strategies for personalizing FES while balancing multiple competing objectives, including biomechanical effectiveness, comfort, and long-term adherence. We developed a multi-perspective personalization framework combining user preferences and biomechanical performance metrics. Interestingly, we observed that clinicians, who are often tasked with tuning and prescribing these devices, largely favored visible kinematic changes over critical kinetic features essential for effective propulsion. This finding suggests that human perception alone may be insufficient for optimal tuning and reinforces the need for objective biomechanical guidance that can ensure better clinical outcomes and more effective gait rehabilitation.
The final aim focuses on addressing this gap by establishing estimation frameworks that use wearable sensing, such as inertial measurement units and pressure insoles, to provide accurate kinetic estimates for populations with neuromotor gait impairments outside laboratory environments. These methods lay the foundation for future expert-informed and automated device tuning, and facilitate comprehensive real-world monitoring of gait biomechanics. Furthermore, we demonstrated the feasibility of integrating these wearable-derived kinetic variables as inputs into adaptive control algorithms for wearable devices, including the mobile FES platform and an ankle exoskeleton, representing a meaningful step toward practical community deployment. Additionally, we investigated methods combining electrical stimulation with ultrasound imaging to reliably track muscle fatigue, offering an initial step toward characterizing muscle states in real-time and providing feedback to enable safe assistance that could respond to the onset of fatigue.
Collectively, this thesis lays important groundwork toward achieving scalable, personalized, and clinically impactful gait rehabilitation in everyday community environments through a multifaceted approach that combines portability, user-specific adaptation, and biomechanics-informed control.Engineering and Applied Sciences - Engineering Science
Semi-supervised and Representation Learning for Improved Classification and Stratification in EHR Data
No full textThe rapid digitization of healthcare has given rise to vast repositories of electronic health record (EHR) data, offering unprecedented opportunities for data-driven advancements in disease prediction, patient stratification, and clinical decision-making. However, the high dimensionality, sparsity, and heterogeneity of EHR data present unique statistical and computational challenges. Moreover, the scarcity of high-quality labels—due to the cost and complexity of manual annotation—further complicates supervised modeling efforts. This dissertation addresses these challenges through a unified framework of semi-supervised learning and representation learning for improved classification and stratification in EHR data, with applications to phenotyping, disability prediction, and patient subgroup discovery.
The overarching goal of this work is to develop scalable, robust, and interpretable methods that leverage both labeled and unlabeled EHR data, improve generalizability across populations, and uncover clinically meaningful structure in complex disease settings. The dissertation is composed of three interrelated papers, each tackling a key methodological bottleneck in modern EHR-based machine learning: (1) evaluating model performance under distributional shift, (2) learning rich patient representations in the presence of limited labels, and (3) stratifying heterogeneous patient populations using outcome-informed embeddings.
In Chapter 1, we consider the problem of evaluating the performance of binary classifiers when labeled data are unavailable in a target population. This setting is common in clinical phenotyping tasks, where models are trained using limited chart-reviewed labels in one cohort and then applied to other cohorts with potentially different covariate distributions. We propose STEAM Semi-supervised Transfer lEarning of Accuracy Measures), a doubly robust estimation procedure for receiver operating characteristic (ROC) parameters under covariate shift. STEAM combines calibrated density ratio weighting with robust outcome imputation, using both unlabeled source and target data to improve efficiency while protecting against model misspecification. Through theoretical guarantees and empirical results, we demonstrate that STEAM enables accurate performance assessment in unlabeled target populations, with applications to phenotyping models in rheumatoid arthritis on temporally evolving EHR cohort.
Building on the challenge of label scarcity, Chapter 2 shifts focus to semi-supervised representation learning for predictive modeling. We propose SCORE (Semi-supervised Clustering thrOugh REp-
resentation learning), a generative embedding framework that models the joint distribution of high-dimensional EHR features using a multivariate Poisson-LogNormal distribution, with pretrained code embeddings capturing semantic relationships between clinical concepts. SCORE integrates limited labeled data via a hybrid Expectation-Maximization and Gaussian Variational Approximation algorithm, enabling efficient and theoretically sound inference in large-scale, partially labeled cohorts. We show that SCORE produces informative and transferable patient embeddings, improving prediction of disability status in multiple sclerosis (MS) and outperforming conventional supervised and unsupervised methods.
Finally, Chapter 3 addresses the critical task of patient stratification in heterogeneous diseases. We focus on Alzheimer’s disease (AD), where progression and prognosis vary substantially with age. We propose SOLAR (age-Specific Outcome-guided representation Learning for pAtient clusteRing), a novel clustering framework that incorporates time-to-event outcomes and explicitly models age-group structure using a multitask learning paradigm. SOLAR jointly learns low-dimensional patient representations across age groups, encouraging shared structure while allowing age-specific flexibility. By integrating survival information and modeling age-related heterogeneity, SOLAR identifies clinically meaningful AD subtypes with distinct prognostic profiles, improving both interpretability and clinical utility over existing age-unaware or outcome-agnostic methods.
Together, these three works present a cohesive framework for semi-supervised and representation learning in EHR analysis. The methods developed here contribute new strategies for evaluating, predicting, and stratifying patient outcomes in data-scarce, high-dimensional clinical settings. In doing so, they aim to advance the broader goals of personalized medicine and evidence-based healthcare by making machine learning more robust, scalable, and clinically relevant.Biostatistic
Intrinsic and Extrinsic Regulation of Autoreactive B cell Responses in Systemic Lupus Erythematosus
No full textSystemic lupus erythematosus (SLE) is a prototypic autoimmune disease characterized by the sustained production of autoantibodies with a range of specificities that contribute to immune complex deposition and systemic inflammation. Autoantibody production signifies the breakdown of tolerance by autoreactive B cells. The positive selection of autoreactive B cell clones is succeeded by the differentiation of autoantibody-secreting cells (ASCs) through the extrafollicular (EF) and germinal center (GC) pathways. In this dissertation, we explore the complex mechanisms underlying B cell breaks of tolerance and autoantibody production using various SLE mouse models designed to interrogate distinct B cell differentiation pathways.
In Chapter 2, we utilize a TLR7-driven B cell adoptive transfer lupus model to characterize the kinetics and regulators of extrafollicular pathogenic B cell differentiation. We additionally address the molecular mechanisms of complement receptor 2 (CR2/CD21) downregulation and CD21lo B cell accrual in SLE. Finally, we delineate the rapid, extrafollicular evolution of the follicular B cell repertoire in an autoreactive environment and outline the selective expansion of autoreactive clones.
In Chapter 3, we investigate the complex roles of the integrin and complement receptor CD11c in regulating B cell tolerance. Through adoptive transfer and mixed bone marrow chimera approaches, we provide evidence for tissue-specific CD11c modulation of central and peripheral B cell tolerance, potentially supporting location-directed functions. In parallel, we demonstrate early evidence that exogenous C3 sensitization restricts the autoreactive B cell repertoire, implying a possible role for C3 ligand-receptor interactions in regulating autoreactivity.
In Chapter 4, we demonstrate that the germline human IGHV1-69 allelic leucine (L54) and phenylalanine (F54) variants exhibit distinct functional autoreactivity in vivo. Using the 564Igi-based adoptive transfer lupus model, we show that L54-expressing murine B cells have a selective advantage in 564Igi recipients over the F54-expressing B cells, indicating that an intrinsically autoreactive germline B cell pool can be triggered to expand and differentiate in an autoimmune environment.
Collectively, this work highlights multifaceted dysregulation of B cell development and responses in SLE and refines the mechanistic understanding of B cell tolerance.Virolog
Spatial Organization of Cytoplasm Studied with Xenopus Egg Extract
No full textSpatial organization is essential for biological function at all scales. Within the cytoplasm, various processes require dynamic distribution of intracellular components and establishment of spatial compartments. During mitosis, cytoplasmic organelles as well as chromatin are partitioned into daughter cells; for protein quality control, aggregated proteins are sequestered into specific aggresome structures. In my thesis, I used Xenopus laevis egg extract as a model system to study spatial organization of cytoplasm in these processes.
Cell division during early embryogenesis requires spatial organization of the cytoplasm on length scales of up to a millimeter in species with large eggs. Questions that remain unclear include how intracellular forces are generated at such enormous spatial scales, and how the abundant cytoplasmic components are partitioned prior to cytokinesis. Microtubules are thought of as the main cytoskeletal network that spans the cytoplasm to organize mitosis and cytokinesis, while research on actin filaments and myosin II (together known as “actomyosin”) has mostly been focused on their function at the actin-rich cortex. Whether bulk actomyosin within the cytoplasm plays any role in cell division remains poorly understood. Here we developed a cell-free system with all cytoskeletal networks that recapitulated early embryonic cleavages of the frog Xenopus laevis, and used this system to study the role of bulk actomyosin. Cell-free divisions exhibited multifaceted defects when cytoplasmic F-actin was perturbed. Bulk actomyosin played a major role in partitioning cytoplasm after mitosis. It mechanically strengthened cytoplasm and microtubule asters, at the same time developed spatial heterogeneity in both material properties and stress distribution by setting up the actin depletion zone at future cleavage plane. Microtubule asters behaved as actin-reinforced composite gels that integrated centrosomes, nuclei, and organelles during their partition after mitosis. Actomyosin contraction provided driving force for the astral composite to move away from future cleavage plane. A fluid dynamics model with contractile stress recapitulated cytoplasmic flows away from the midplane. These findings uncover a role of bulk actomyosin in global partitioning of cytoplasm in large embryonic cells and lead to a novel framework for understanding cytoplasm as a composite gel whose dynamics are governed by principles of active fluid mechanics.
Beyond cell division, the second part of my thesis studied how protein aggregates are selectively transported to the aggresome to maintain protein homeostasis. Using the more conventional CSF egg extract, we reconstituted MTOC-directed aggregate transport in Xenopus egg extract. High-resolution single-particle tracking revealed that dynein-mediated aggregate transport was highly episodic, with average velocity positively correlating with aggregate size. We propose that size selectivity in this process enables efficient intracellular transport of protein aggregates.Biology, Molecular and Cellula