Induced pluripotent stem cells (iPSC) have a high potential, for they can be differentiated into any cell type for regenerative medicine, drug discovery, developmental biology, and disease modeling. However, iPSC’s and their differentiated progeny display an undesired variability in their shape, contractile properties, growth rates, etc. Identifying subsets of phenotypes in iPSCs and their differentiated progeny will allow us to optimize tissue models for research. Here, we generated a rainbow reporter line in iPSCs that can track individual cells as they clonally expand and differentiate while providing phenotypic information. Knocking in four copies of a cassette containing three distinct fluorescent proteins allowed the expression of up to eighteen different colors. However, not all colors were present in equal proportion, increasing the probability that distinct lineages could have the same color. To achieve an equal color distribution, colored cells were isolated by sparsely plating a culture of mixed colored cells. After a week of expansion, individual colonies were picked and imaged under a spinning disk microscope to determine the color of the colony and whether it was single lineage or mixed. Viable cell lines were isolated and frozen in stock. These cells will be examined for markers of cell proliferation, pluripotency, apoptosis and quantitative RNA expression analysis to confirm that the color barcoded iPSCs act the same as non-engineered iPSCs. To date, we were able to create eight color barcoded iPSC lines for further experimentation, increasing the concentration limit of colored cells in non-colored cells by five-fold. The next step will engineer 3D tissues by growing iPSC-derived cardiac cells in a mold to simulate in vivo tissue development. Colored coded cells will allow us to track how the initial location/physical stresses/phenotype of an iPSC-derived cardiac cell in an engineered tissue determines its tissue layer and cell type.

Of the estimated 38 million HIV infections globally, approximately 1-2 million are people living with HIV-2 (PLHIV-2). Infection with HIV-2, which is endemic in West Africa, is characterized by lower viral loads (VL) and slower disease progression to AIDS than HIV-1. However, many PLHIV-2 eventually progress to AIDS and death if left untreated. Antiretroviral therapy (ART) for HIV-2 is complicated by fewer effective drugs and significant challenges in drug resistance testing to identify appropriate therapy. This study aims to validate a novel method of HIV-2 drug resistance testing using nucleic acid from dried blood spots (DBS). One hundred and fifty DBS have been collected as part of two clinical studies of ART for HIV-2 in Senegal. HIV viral nucleic acid (DNA/RNA) is extracted and the regions encoding protease (PR) and reverse transcriptase (RT), which are the most common drug targets, are amplified by PCR, then sequenced. Sequence data are examined for evidence of drug resistance-associated mutations. To date, we have tested 115 samples with a median viral load of 84 (range: 0-24,000) and obtained drug resistance data from 43 (37.3%). Testing was most commonly successful from samples with higher viral loads; samples with viral loads >250 copies/ml were successful 75.7% of the time. We observed several drug resistance mutations, including PR V47A (5 patients), I50V (3 patients), and L90M (2 patients), and RT K65R (4 patients), Q151M (2 patients), and M184V (10 patients). Eight patients had no known resistance mutations. Once DBS testing is complete, we will select a subset of sequences to compare to genotypic resistance testing from corresponding plasma samples to evaluate the sensitivity of DBS-based testing vs. standard methods. DBS-based resistance testing has the potential to revolutionize ART for HIV-2 by allowing faster, easier, and cheaper assessment of drug resistance to optimize second-line therapy for HIV-2-infected patients.

Native mass spectrometry (MS) experiments provide direct mass measurements of intact proteins and protein complexes. Protein samples for native MS are prepared in solutions that mimic physiological conditions, which maintain a protein’s native folded state before entering the gas phase of the mass spectrometer. Ammonium acetate solution is typically used due to its volatility and relevant ionic strength. However, protein purification protocols typically require inorganic salts and detergents to maintain protein stability. Native MS experiments can be hindered or made uninterpretable by those salts and detergents. Furthermore, the presence of protein modifications or multiple proteins can make native mass spectra difficult to interpret. Anion exchange chromatography (AEX) is well suited for the requirements of native MS, as it can simultaneously desalt, remove non-ionic detergents, and separate proteins or proteoforms directly into an ammonium acetate solution. This project seeks to develop a comprehensive method for desalting, removing non-ionic detergents, and separating proteins through an ammonium acetate-based anion exchange chromatography method. Preliminary experiments in egg whites, a complex matrix with a high sodium concentration, showed separation and four distinct proteins using an AEX pH gradient from pH 10 100 mM ammonium acetate to pH 4 100 mM ammonium acetate. Native MS analysis showed low interference from sodium or other contaminants and the various modified forms of those proteins were identified. Refinement of this preparation technique can result in the improvement and efficiency of native MS analysis of proteins.

Nicotine is the addictive substance found in various tobacco products. CYP2A13 is an enzyme localized in the lungs, metabolizes tobacco-specific nitrosamine carcinogens that contribute to lung cancer. Therefore, pinpointing CYP2A13 inhibitors is an approach to lower tobacco-based lung cancer risk. Cinnamaldehyde is a common flavoring agent in the fluids of electronic nicotine vaping devices. Cinnamaldehyde was found to be a potent inhibitor of CYP2A6, another enzyme that metabolizes nicotine. Because CYP2A13 and CYP2A6 exhibit overlap in substrate/inhibitor selectivity, the goal here was to evaluate the inhibition of CYP2A13 by cinnamaldehyde. A time-dependent inhibition coumarin assay was performed to determine the kinetic parameters for cinnamaldehyde in recombinant CYP2A13. Primary incubations contained cinnamaldehyde, CYP2A13 Supersomes, and potassium phosphate buffer. Incubations were initiated with NADPH. Secondary incubations contained coumarin, NADPH, and potassium phosphate buffer. At selected time points, an aliquot of the primary incubation mixture was transferred to the secondary incubation tubes, which were terminated with trichloroacetic acid after heating at 37°C for 5.5 minutes. A linearity study was conducted to determine the appropriate termination time. CYP2A13 activity was measured by the detection of hydroxycoumarin using high-performance liquid chromatography (HPLC) and a fluorescence detector. Hydroxycoumarin formation decreased with time and inhibitor concentrations. Maximal inhibition following an 18-minute incubation was 38.3 ± 1.6 and 4.0 ± 0.6%. The maximal rate of inhibition was 0.109 per minute. The results provide evidence that cinnamaldehyde is a time-dependent inhibitor of CYP2A13. Furthermore, cinnamaldehyde appears to be a more potent inhibitor of CYP2A13 than CYP2A6, based on the maximal rate of inhibition. The results imply that cinnamaldehyde could interfere with the bioactivation of nitrosamine lung carcinogens. Additional kinetic studies are needed to confirm the results of this study and to evaluate the safety and toxicity profiles of cinnamaldehyde in complex physiological models.

Understanding cause-and-effect is key for informed decision-making. The gold standard in causal inference is performing controlled experiments, which may not always be feasible due to ethical, legal, or cost constraints. As an alternative, inferring causality from observational data has been extensively used in statistics and social sciences. However, the existing methods critically rely on a restrictive assumption that the population of study consists of homogeneous units that can be represented as a single flat table. In contrast, in many real-world settings, the study domain consists of heterogeneous units that are best represented using relational databases. We propose and demonstrate CaRL: an end-to-end system for drawing causal inference from relational data. In addition, we built a visual interface to wrap around CaRL. In the demonstration, I will use CaRL, which I have implemented, to show a live investigation of causal inference from real academic and medical relational databases.

Each year, nearly 6 million deaths worldwide are caused by lower respiratory tract infections, diarrhoeal diseases, and tuberculosis. These infectious diseases are leading causes of death worldwide. Currently, the pharmacological treatment of infection is encumbered by the presence of inaccessible intracellular pathogen reservoirs, and the need for prolonged treatment regimes. Drug delivery systems can be engineered to overcome these biological barriers for effective treatment by facilitating intracellular delivery and tailored release. Extended release of drugs alleviates the need for exhaustive treatment regimes and increases patient compliance. Furthermore, this can decrease treatment duration, reduce the cost of treatment, and improve access for disadvantaged populations. Our research utilizes modular polymeric prodrugs composed of molecular targeting agents, cleavable linkers, and antimicrobial drugs. This platform permits facile alteration of functional modalities, enabling custom tailored treatments for each disease setting. By utilizing tunable linkers, we can control the precise delivery mechanism and therefore direct the localized release of drugs. To characterize the controlled release of antimicrobial drugs from our polymeric prodrugs, we are designing high performance liquid chromatography (HPLC) and liquid chromatography mass spectrometry (LC-MS) assays. The developed assay will evaluate the release mechanism with stability and release studies. Furthermore, the robust methodology will enable the determination of the pharmacokinetics of the polymeric prodrug delivery system. The assay and results from this study will ultimately support the development of improved therapies.

Pseudomonas aeruginosa is a Gram-negative opportunistic pathogen associated with worsening disease outcomes in cystic fibrosis (CF) patients. P. aeruginosa uses quorum sensing (QS), a cell-cell signaling system, to control expression of a variety of genes including virulence factors. In P. aeruginosa, QS is mediated in part by acyl-homoserine lactone (AHL) signals that can diffuse in and out of cells. Once AHLs accumulate, they bind to a receptor regulator that activates gene transcription. P. aeruginosa has two complete AHL QS systems, LasI-LasR and RhlI-RhlR. The two systems are arranged in a hierarchy, with the las system controlling the rhl system. QS activation in P. aeruginosa is restrained by cellular proteins that dampen the QS response. These proteins, known as “anti-activators”, attenuate QS by preventing receptor activation. Three anti-activator proteins, QscR, QslA, and QteE, have been identified in P. aeruginosa. These anti-activator proteins have additive, overlapping roles in repressing expression of QS gene products in laboratory strains but their role in the QS dynamics of CF isolates is still unclear. This project used standard molecular cloning techniques to delete or overexpress anti-activator genes in a selection of clinical isolates from CF patients. A reporter plasmid with a fluorescent marker was used to track the activity of LasR and RhlR. These experiments were used to quantify differences in QS-controlled gene activation. To test the hypothesis that anti-activators decrease the amount of LasR in the cell, Western blots were used to assess the cellular levels of QS receptors. In strains with deleted anti-activator genes, LasR levels were higher and induction was earlier. Additional tests for phenotypes controlled by QS, such as protease and pyocyanin production, were also performed. Future research should focus on evaluating these effects in additional CF isolates.

Cystic fibrosis (CF) is a genetic disorder affecting the lungs, and chronic polymicrobial lung infections are responsible for decreased life expectancy and poor quality of life of CF patients. Staphylococcus aureus (SA) is a microbe commonly found in the respiratory tract of CF patients, and this organism adapts within the lung environment to establish chronic infections. Among the most common bacterial adaptations is the emergence of mutants known as small colony variants (SCVs). There are multiple subtypes of SCVs that arise from mutations in different metabolic pathways. Recent studies have demonstrated that SCVs are prevalent in the CF respiratory tract, are more difficult to treat with antibiotics, and are associated with worse lung health. SCVs are very difficult to detect in clinical laboratories, thus complicating the selection of appropriate treatment by physicians to improve the health of CF patients. The goal of this study is to determine if SCVs can be more readily detected than with standard culture by using mass spectrometry to identify proteins that distinguish these variants from normal colony S. aureus. Matrix Assisted Laser Desorption/Ionization-Time of Flight (MALDI-TOF) will be used to identify proteins unique to specific SCV subtypes by separating those proteins using ionization and Tandem Mass Spectrometry. This analysis will generate isolate-specific spectra of peaks which will subsequently be compared to each other using Principal Coordinate Analysis (PCoA). We hypothesize this technique will identify differences between proteins produced by each SCV type, which can then be distinguished from normal colony S. aureus, allowing the rapid identification of these variants. As a result, we anticipate the detection of SCV’s will improve, which will help inform physicians to select appropriate treatments to target SCVs.

Numerous interventions and genetic modifications have been shown to extend lifespan across a diversity of species. However, these studies often assume that extended lifespan is synonymous with extended healthspan. Recent research in the nematode worm, Caenorhabditis elegans, has questioned this assumption, and suggests that increasing lifespan can prolong the frailty associated with old age. This is particularly important for humans, as increasing lifespan without a corresponding increase in healthspan could spell disaster. The majority of healthcare costs are associated with aging-related pathologies, and prolonging life without prolonging health could radically inflate these costs. To parse out the genetic relationship between healthspan and lifespan, we have turned to Drosophila melanogaster, a well characterized model organism for studies on the genetics of aging. We have collected lifespan data as well as multiple measures of healthspan, such as negative geotaxis (climbing), intestinal permeability, Cold stress resistance, and metabolomics data across 16 inbred genotypes. We found a strong positive correlation between lifespan and climbing, and no correlation between cold stress resistance and lifespan. This confirms the importance of lifespan as a primary parameter in aging studies, but suggests additional measures of health are needed to accurately assess health. 

Training a nonhuman primate (NHP) for research experiments generally requires the NHP to spend large quantities of time learning experimental tasks outside their home environment, and this requires a human researcher to be present at all times during training sessions. The purpose of this project is to create a wireless, semi-autonomous, low cost, cage-side training reward system allowing NHPs to train on experimental tasks for extended periods of time without the presence of a human researcher. An ideal device would allow for wireless data collection and provide both real-time and post-training information on the NHP’s training progress. Exposing NHPs to tasks first in the low-stress environment of their home cage before exposing them to the same task in an experimental booth can potentially speed up training processes. This lets research laboratories maximize researcher time and efficiently use equipment. Our cage-side training reward system consists of an iPad displaying touchscreen tasks, a speaker supplying audial cues for the tasks, an automatic feeder administering treats to the NHP for correct performance, and a computer to control the touchscreen tasks and collect data with custom MATLAB code. The iPad and computer communicate via a Wi-Fi router and this router also communicates with a Wi-Fi receiver which runs the feeder and speaker. The connection methods give the ability for wireless communication through walls, allowing the researcher to run tasks semi-autonomously from a computer outside the animal room. Excluding the costs of the iPad, computer, and MATLAB license, the system is estimated to cost under $300. Two rhesus macaques have undergone cage-side training with this device and have subsequently transitioned smoothly to learning tasks in a traditional experimental booth. In conclusion, this device serves as a low-cost method to enhance the training process for non-human primates while saving time and resources of the research laboratory.
 

Trace metals are a vital part of primary production for phytoplankton, and are a scarce resource around the world. Iceland's unique geology allows for these metals to be recycled through geologic processes as they are released from volcanic ash and trapped in ice, frozen in time until being released again through melt. Iceland's glaciers are melting at a rapid pace due to the warming climate, and are releasing and transporting these trace metals that have not been exposed to the environment for an extended period of time. Sólheimajökull, a glacier located in Southeast Iceland, is no exception and has seen an increasing retreat as it melts at an accelerated pace. The metals being released and transported in this melt are found to be magnitudes higher in abundance relative to the open waters surrounding the island. Introducing these metals to the open ocean could cause rapid changes in the biogeochemical processes of the surface ocean. With this increased melt, we're likely to see an increase in release and transportation of these metals, some of which may reach the coast and cause ecological devastation. In this study, I aim to quantify the trace metal deposition through various freshwater streams of Sólheimajökull.

Joubert syndrome (JS) is a genetic neurodevelopmental disorder that affects ~1 in 100,000 live births. JS is diagnosed by a distinctive hindbrain malformation that manifests as the “molar tooth sign” on axial brain imaging. Remarkably, >40 genes have been associated with JS, making it one of the most genetically heterogeneous Mendelian conditions. The clinical and brain imaging features of people with JS display a broad range of severity. In fact, we have identified a substantial number of individuals without the molar tooth sign but that have imaging features suggestive of JS. It is not known whether these “JS-like” patients represent the mild end of the phenotypic spectrum associated with variants in JS genes or a different set of genetic disorders. It is also not known whether these JS-like patients are at risk for the progressive retinal, kidney and liver disease seen in some JS patients. To answer these questions, I performed targeted DNA sequencing of the JS genes in JS-like patients, and I used an in-house bioinformatics pipeline to identify predicted-pathogenic variants. We hypothesize that a large subset of JS-like patients will have genetic causes in JS genes. If this hypothesis is supported, we will expand the phenotypic spectrum associated with variants in JS genes and improve the medical care of JS-like patients by supporting monitoring of JS-associated progressive features and sequencing of JS genes in these patients. This will also be proof of concept for evaluating mild clinical presentations of other conditions to determine if they share the same genetic causes.

Every minute, Twitter users send hundreds of thousands of tweets, providing a rich resource of publicly available text data. Our goal is to use this data to learn from and imitate the sentence structure of specific accounts. To this end, we develop mRkov, a statistical tool that takes the username of a Twitter account, also known as a handle, as input and outputs fake tweets that mimic the linguistic style of the tweets from that handle. We built mRkov into an R software package as well as an interactive and user-friendly web tool that walks the user through the process of using our software. mRkov first scrapes tweets posted from the input Twitter handle, and then after processing the text and sentiments of the scraped tweets, generates new tweets using Markov chain simulation. Markov chains consist of a sequence of items, where each item is probabilistically sampled dependent only on the preceding item in the chain. By using non-independent sampling, the Markov chain method generates a sample that mimics the true distribution. In this application, the Markov chain is a sequence of words, and the distribution is the sentence structure of the tweets. mRkov also allows users to provide input that influences the sentiment of tweets in order to generate tweets that tend to be more “positive” or “negative” in sentiment. Tools such as mRkov help us better understand patterns of speech and writing. This has many useful applications, including identifying if multiple accounts are coming from the same source or writer, analyzing and comparing how the style and sentiment of different accounts change over time, and detecting bots or other fake accounts.