Compared to polymerase chain reaction with nasopharyngeal swab specimens, lateral flow assays (LFA) for detecting SARS-CoV-2 antigen in saliva are rapid, inexpensive, simple-to-use, and instrument-free, but have low sensitivity (ca. 11-40%) due to limited LFA sample capacities (ca. 50µL) and preparatory saliva sample dilution. To address this challenge, I developed an osmotic processor, a 3D-printed device, to concentrate target analytes through static and spontaneous osmosis for improving salivary-based COVID-19 LFA. The specimens loaded into the device were separated from an aqueous polymer solution using a dialysis membrane. The polymer solution with high mass polymer concentration resulted in an osmotic pressure difference that caused water transport from the saliva to the polymer solution, while the target analyte – SARS-CoV-2 nucleocapsid (N) protein – remained within the membrane. As the final solution contained the same N protein amount in a lower liquid volume, the overall sample was concentrated, thus improving the antigen detection limit ca. 60-fold, from 62.5 to 1 pg/mL. Combining the osmotic processor and saliva based LFA enables rapid, sensitive, simple, and inexpensive POC testing.

The functional connectivity of the brain evolves throughout the life of every individual. These changes, often referred to as neuroplasticity, can be impacted by a wide range of variables from diseases to eating a favorite dessert. How can these changes be modulated to treat neurological diseases and disorders such as post traumatic stress and major depressive disorder? My colleagues and I are intrigued with the prospects of neuromodulation as a therapeutic for abnormal brain connectivity and network dynamics, leading me to the question “At what rate do these connections accumulate and decay with optogenetic modulation; Optogenetics, a technique that uses light to activate or inhibit genetically targeted neurons, offers high cell-specificity and temporal resolution that allows us to zoom into the network dynamics and find more finely tuned results that can help in the development of neuromodulation therapies. I plan to use both single site and paired-pulse optogenetic inhibition to gain a clearer understanding of how functional connectivity behaves during and after repeated modulation periods followed by extended recordings of spontaneous activity with no modulation. By analyzing the pairwise coherence of the local field potentials collected using electrocorticographic recordings in non-human primates, I anticipate seeing targeted changes in functional connectivity when comparing before and after each inhibition session. By analyzing the rate of change in connectivity I plan to understand the timeline of neuroplasticity following optogenetic modulation, thus informing the development of future neuromodulation therapies. This research could have a profound impact on the future therapeutic paradigms for neurological and neuropsychiatric disorders that can accelerate recovery for individuals with these conditions.

Peptide cancer vaccines have had limited clinical success despite their safety, characterization, and production advantages. We hypothesize that the poor immunogenicity of peptides can be surmounted by delivery vehicles that overcome the systemic and cellular drug delivery barriers faced by peptides. We introduce Man-VIPER, a self-assembling, pH-sensitive, mannosylated polymeric peptide delivery platform that targets dendritic cells in the lymph nodes and facilitates endosomal release of antigens through a conjugated membranolytic peptide melittin. We evaluated polymers with both releasable (Man-VIPER-R) or non-releasable (Man-VIPER-NR) D-melittin. Both Man-VIPER polymers exhibited superior endosomolysis and antigen cross-presentation compared to non-membranolytic D-melittin-free analogues (Man-AP) in vitro. In vivo, Man-VIPER polymers demonstrated an adjuvanting effect, induced the proliferation of antigen-specific cytotoxic T cells and helper T cells compared to free peptides and Man-AP. Remarkably, antigen delivery with Man-VIPER-NR generated significantly more antigen-specific cytotoxic T cells than Man-VIPER-R in vivo. As our candidate for a therapeutic vaccine, Man-VIPER-NR exerted superior efficacy in a B16F10-OVA tumor model. These results highlight Man-VIPER-NR as a safe and powerful peptide cancer vaccine platform for cancer immunotherapy.

Mutations in β-myosin heavy chain (MHC) have been implicated in the manifestation of cardiomyopathies. One such iteration of disease, dilated cardiomyopathy (DCM), results in a progressive loss of muscle in the heart, causing a dilation of the chambers that makes it more difficult for the heart to pump blood and leads to progressive heart failure. Current methods of treating cardiomyopathies are ineffective and target secondary manifestations of heart disease rather than the mutagenic causes. Our aim is to better understand the pathogenesis of cardiomyopathies by studying a particular disease-associated mutation, R369Q, that resides on loop 4 of the β-MHC structure and is implicated in actin-myosin interaction. Here, I studied the molecular mechanistic effect of the R369Q mutation on β-MHC and its role in the DCM phenotype through in silico molecular dynamics experiments comparing the mutant myosin structure to the wild type. Particularly, I examined the flexibility of loop 4, coordination of the nucleotide binding pocket, and electrostatic interactions along the actin interface. We expect to see decreased electrostatic interactions along the actin interface as the loss of a positive charge in this mutant will reduce the affinity of myosin for the negatively charged actin. This work provides a strong foundational hypothesis for the atomic-level impact of the R369Q mutation on myosin dynamics and will be used to design a therapeutic small molecule as well as corroborate in vitro experiments testing the biochemical function of mutant myosin in human induced pluripotent stem cell-derived cardiomyocytes (hiPSC-CMs). Thus, computational and experimental approaches work in synergy in the comprehensive project, informing each other’s results to get a better picture of R369Q’s role in cardiomyopathies at a molecular level and forming a framework for future analyses of β-MHC mutations.

The Stimulator of Interferon Genes (STING) pathway is a promising target for cancer immunotherapies. However, STING agonists have poor cell membrane permeability and also off-site toxicity that limit their therapeutic application. Our motivation is thus to develop a novel method for the delivery of non-cell membrane-permeable STING agonists. We propose incorporating the STING agonist in a self-assembling polymer, which then transports the drug into the cytosol, thereby avoiding membrane permeability issues. The first step is the incorporation of the STING agonist into Man-VIPER, a cytosolic drug delivery system developed by the Pun Lab. Man-VIPER mediates cytosolic delivery via endosomal escape, releasing the STING agonist cargo into the cytosol. Currently, I am working on synthesizing and purifying the small molecule STING agonist called SR-012. I tested many possible reaction pathways to make SR-012, which is unavailable commercially and has only been previously synthesized once. I developed a novel alternative reaction pathway and purification that works consistently. I am also currently synthesizing the polymer building blocks for the Man-VIPER copolymer that will deliver SR-012. After the optimized drug-polymer complex is completed, I will work with collaborators to quantify the max tolerated dose. We hope to demonstrate that specificity is improved through lower toxicity, allowing for a higher maximum tolerated dose than my control, the membrane permeable STING agonist SR-717. Next, I will assay the immunotherapeutic effectiveness of Man-VIPER delivered SR-012 using in-vivo tumor models.

Cytokines are molecules that mediate cell-to-cell communication in the immune system. Their sensing by immune cells drives their response and differentiation into different functional states, such as into memory or effector states for T cells, to form a proper immune response and combat infections or cancer. Signal transducer and activator of transcription (STAT) family proteins are transcription factors that act downstream of cytokine signaling to regulate the expression of genes driving differential cell functions. Different cytokines phosphorylate and activate different STAT family members to give rise to distinct cellular responses and phenotypic changes. However, there is substantial overlap between STAT members activated by different cytokines. For instance, type I interferon (IFN-I) activates both STAT1 and STAT3 signaling to drive genes associated with growth inhibition, cell death, and anti-microbial responses. Alternatively, Interleukin-2 (IL-2) primarily activates STAT5, but can mediate STAT3, to promote T cell effector differentiation and proliferation. Thus, how different cytokines work through a small common set of STATs to elicit distinct functional phenotypes requires further investigation. Therefore, my project seeks to probe T cell responses and STAT activity with different cytokines. To interrogate this, I will develop an assay to stimulate Jurkat cells, an immortalized T cell line, under various cytokines and analyze STAT1, STAT3, and STAT5 phosphorylation with flow cytometry. We expect Jurkats under IFN-I stimulation to have higher STAT3 and STAT1 phosphorylation than STAT5, but exhibit a more immunosuppressive phenotype, as STAT3 is known to inhibit STAT1-mediated gene expression. With IL-2 stimulation, we expect cells to have higher STAT5 activity with increased proliferation and enhanced effector function. By understanding the complexities of how our T cells differentiate and exhibit certain phenotypes from STAT activity, we can potentially tune the signal a cell interprets and the downstream cellular responses to effectively enhance effector functions, or prevent cell death.

Chimeric antigen receptor (CAR) T-cell therapy is a revolutionary cancer treatment that utilizes the body’s immune system to recognize and fight malignancies. It describes the process of extracting T cells from a patient and genetically engineering them ex vivo to express CARs that direct T cells to kill cancer cells. Multiple FDA-approved CAR T-cell products have shown promising clinical efficacy in treating cancers such as relapsed/refractory CD19+ B-cell leukemia and lymphoma. However, these products still have limitations in targeting other hematological and solid cancers. Solid tumors are difficult to treat due to their heterogeneity and ability to down-regulate or mutate antigen expression in response to treatment. Also, T-cell exhaustion and treatment associated toxicity result from the inability to precisely control CAR T-cell activity with these current treatments. Biological intermediates such as antibodies may be used to address these barriers, but they are expensive and can elicit an immune response. We aim to develop a universal CAR T-cell system that uses synthetic, high-affinity biomaterials to bind receptors expressed on cancer cells for targeted T-cell killing. For this project, we engineered CD8+ T cells, a cytotoxic subset of T cells, with lentivirus to extracellularly express universal CARs. Using flow cytometry, we confirmed strong expression of CARs on the T-cell surface and demonstrated their ability to interact with a synthetic cancer-targeting intermediate. Lastly, we show that CD8+ CAR T cells pre-armed with the targeting intermediate were capable of selectively lysing cancer cells in vitro that express the receptors of interest. For future studies, I hypothesize that we can expand this universal CAR system by utilizing other synthetic intermediates, such as a heterobifunctional small molecule peptide intermediate, for the treatment of solid tumors.

One challenge in developing point-of-care molecular diagnostics is achieving a high purity of the protein of interest. Immobilized metal affinity chromatography (IMAC) is a commonly utilized method to rapidly purify polyhistidine affinity tagged proteins. A barrier to accessing purification methods is the cost of materials such as columns manufactured by biotechnology companies at high cost. To address this issue, we have used an (RH)4-tag, a short peptide which consists of four consecutive arginine-histidine groups and can constitute an affinity fusion tag for IMAC due to the arginine interaction with the silica surface via an ion pairing mechanism. Further, the (RH)4 tag can reversibly bind with the silica surface, which allows for the elution of the fusion tag using L-lysine as the elution buffer. Therefore, the silica binding property of (RH)4 enables purification to occur without requiring high cost complex resins. To evaluate the efficiency of the purification process, we have chosen eGFP as the gene of interest of the fusion tag. The green fluorescence of eGFP can be utilized to quantify protein purification efficiency. We can conduct a chromatographic workflow using a column with a silica gel bead matrix, using purified samples of eGFP-RH4 diluted to different concentrations (from 10-100nM). The diluted protein samples are mixed with a silica particle suspension and incubated. The silica beads are removed by centrifugation at 5000xg, and the amount of each protein is derived from the difference between the initial fluorescence and the fluorescence of eGFP remaining in the supernatant. Preliminary results have shown the (RH)4 tag is functional when purifying eGFP on a silica based column and when purifying eGFP using IMAC. Further exploration includes finding an optimized buffer system and using developed methodology to purify different proteins of interest.

COVID-19 is an infectious viral disease that is caused by the SARS-CoV-2 virus entering host cells through surface spike proteins that bind to surface ACE-2 receptors. Some anti-SARS-CoV-2 antibodies allow SARS-CoV-2 to also have antibody-mediated entry (AME) into immune cells, often via Fcγ receptors. This phenomenon has been correlated with cytokine release syndrome, which occurs when the immune system has a highly inflammatory response to infection and is implicated in severe COVD-19 cases. It has also been shown that other antibodies have demonstrated inhibition of SARS-CoV-2 entry. During binding and viral fusion, all three receptor binding domains (RBD) of the spike protein fold to an upward conformation, which is necessary for binding. Inhibitory anti-SARS-CoV-2 antibodies impede this process through mechanisms such as premature cleavage of the RBD or stabilizing a three-down conformation. I hypothesize that inhibitory antibodies can stop AME from occurring, but it is not yet understood which inhibitory mechanisms are most effective at preventing AME. To understand the dynamic between AME and inhibitory antibodies, I am testing the infection levels of monocytes by SARS-CoV-2 pseudo-virus in the presence of antibodies shown to induce AME combined with varying concentrations of inhibitory anti-SARS-CoV-2 antibodies that have different mechanisms of affecting the spike RBD conformation. Preliminary results suggest that high concentrations of multiple neutralizing antibodies inhibit AME. It has also been observed that an inhibitory antibody that activates the up RBD conformation increases entry at certain concentrations, whereas an inhibitory antibody that stabilizes the down RBD conformation does not enhance AME. This work will contribute to our investigation of the connections between AME, spike conformational regulation, and immune cell inflammation. Studies of this type can aid in continued development of safe vaccines and therapeutics, as well as help understand how antibodies affect SARS-CoV-2 spike conformational regulation and therefore viral entry.

Plasmodium vivax (P. vivax) is the most widespread malaria species. P. Vivax can lay dormant in hepatocytes and cause recurring malarial infections. Tafenoquine (TQ) is an antimalarial drug approved by the FDA in 2018 for the radical cure of P. vivax. Unfortunately, 8-aminoquinolines like TQ cause hemolytic anemia in G6PD deficient patients. Due to this contraindication, genetic G6PD screening is required before TQ administration. These additional tests pose a significant challenge for broad administration in resource-constrained countries. Our polymer prodrug conjugates (drugamers) are designed to improve TQ delivery to the liver and reduce red blood cell exposure. Using RAFT polymerization, drugamers can be optimized for delivery efficiency and avoidance of hemolytic anemia. Polymer architectures can be further enhanced by utilizing a variety of enzyme cleavable linkers, monomers, and receptor binding cofactors. We previously demonstrated increased liver-to-blood area under the curve ratios in mice using a variety of different polymer architectures. Mice received an IV or subcutaneous dose of drugamer and relevant organs were harvested at previously determined time points up to 48 hours. Pharmacokinetic profiles were created using a plate reader fluorescence assay and tandem LC/MS/MS with tissue samples. These drugamers represent an ongoing effort to iteratively improve our drug-polymer design. The improvement in TQ delivery through our drugamer vehicles could allow for more liberal administration guidelines, completely removing the need for G6PD testing for the treatment of P. vivax.

Chronic kidney disease (CKD) is an incurable, progressive condition that affects up to 700 million people globally. Progression of CKD currently leads to a deteriorating quality of life on dialysis, often resulting in terminal end-stage renal disease (ESRD) and increased risk of cardiovascular disease. A hallmark feature of CKD progression is epithelial-to-mesenchymal transition (EMT), a process by which kidney cells obtain malignant properties like increased mobility and resistance to apoptosis. Despite the pressing burden of CKD, current therapeutics like angiotensin-converting enzyme (ACE) inhibitors and angiotensin-receptor blockers are unable to halt the fibrotic progression of CKD. In recent years, epigallocatechin-3-gallate (EGCG) has been shown to be a promising candidate to inhibit EMT in kidney tubular epithelial cells (TECs). In this project, we conjugated EGCG to a low molecular weight polymer previously engineered at the Pun Lab for enhanced localization to the kidney TECs. In a TGF-β1 induced fibrosis model in human kidney cells, we showed that this polymer-EGCG conjugate (poly-EGCG) could diminish the RNA and protein expression of mesenchymal markers compared to untreated controls. We also confirmed that poly-EGCG was well-tolerated across a broad range of concentrations through in vitro cell viability assays. Finally, immunohistochemistry staining of mouse kidney samples injured with anti-glomerular antibodies displayed partial fibrotic recovery when treated with unconjugated EGCG. Future in vivo studies will aim to optimize the efficacy of poly-EGCG treatments compared to unconjugated EGCG treatments by assessing histology and urine samples for markers of kidney dysfunction. Through improved delivery of EGCG to the kidney TECs, this novel polymer-EGCG conjugate has the potential to halt the progression of EMT for future patients with CKD.

An effective peptide-based cancer vaccine requires efficient intracellular delivery of antigen peptides to activate tumor-killing immune responses. However, the localization of peptide antigens during endosomal release to optimize the immune response remains under-investigated. The Pun Lab has developed the self-assembling Virus-Inspired Polymers for Endosomal Release (VIPER) that can induce endosomal escape of peptide antigens. The current VIPER formulation employs reducible disulfide bonds for antigen conjugation. We hypothesize that the controlled antigen peptides release mediated by endosomal proteases can result in more effective antigen presentation that leads to more potent tumor-killing cell responses. In this project, I replace VIPER’s antigen conjugation strategy with the pentafluorobenzyl (PFB) moiety (VIPER-NR) and introduce an enzyme-labile linker in the antigen sequence to determine the optimal peptide release kinetics and localization for optimal tumor-killing cell responses. I synthesized a library of enzyme-labile linkers for the ovalbumin antigen peptide of VIPER-NR. I utilized a dendritic cell cross-presentation assay to screen for linkers associated with that efficient antigen presentation in vitro. Subsequently, I conducted enzyme-mediated drug release assay and red blood cell lysis assay in vitro to evaluate the cleaved antigen profile and endosomal escape induction capacity of VIPER-NR with cleavable linkers compared to VIPER-NR. We expect early release of antigens during endocytosis with these modified antigen sequences in VIPER-NR in vitro and significantly elevated cytotoxic T-cell response against the model antigen in vivo. This project could improve the effectiveness of the VIPER system as a peptide vaccine delivery platform. It would also provide further understanding of optimal antigen release profiles for a better tumor-killing cell response in cancer vaccine applications.

My research project aims to combat inflammation and fibrosis caused by biomaterials and implants. Nondegradable biomaterials can provide long-term stability in the body but can elicit a foreign body response, such as inflammation. The project focuses on finding ways to repair and replace damaged tissue at the location of the implant.The project involves engineered M1 cells which were created and published by the Giachelli Lab within the Department of Bioengineering at UW. We have two groups of mice: one control group that has been injected with engineered TLR4 (Toll-Like Receptor 4) cells but not given the CID (Chemically Induced Dimerizers) drug, and another group that has been injected with both the engineered TLR4 cells and given the CID drug. The inside of the cell was isolated in both groups of mice and the extracellular domain was removed. For the group with CID, the inside domain of the cell was bound with CID which is thought to result in dimerization and activation of the M1 phenotype.At this stage in the project, we have tissue samples from both these groups, and I will analyze these samples using H&E staining thus allowing me to measure healing parameters, like how dense the collagen structure is around the material and analyze the effect of CID on the state of the tissue. For expected results, I believe the group with CID will have a denser collagen structure around the material than the group without CID.Conducting the analysis on these tissue samples will help address the effect of CID on healing parameters, such as inflammation. Furthermore, this work will help us develop a better understanding of the roles that M1 and M2 phenotypes play in the healing process which can lead to new therapies that reduce the inflammatory response elicited by biomedical devices. 

Cardiomyopathy is currently the leading cause of death for patients with Duchenne muscular dystrophy (DMD), a severe neuromuscular disease affecting young boys. With no current cure, gene therapy is a promising solution, but supplementation with drug therapies is likely inevitable to fully address the pathology seen in older patients. The use of human-induced pluripotent stem cell (hiPSC) models for drug studies is beneficial due to the direct relevance to human physiology and the potential development of personalized care. Dystrophic hiPSC cardiomyocytes have been shown to exhibit calcium reuptake delays, higher resting calcium levels, and frequent arrhythmias. The Mack Lab previously conducted a preliminary drug screen on healthy and DMD-affected cardiomyocytes and found that certain L-type calcium channel blockers (CCBs) indicated a cardioprotective effect. These drug compounds (namely Nitrendipine and Nimodipine) have been shown to alleviate cardiac fibrosis in patients through vasodilation. In this project, I am validating the beneficial aspects of the drug compounds. I initially hypothesized that treatment of DMD hiPSC cardiomyocytes with L-type CCBs will rescue resting calcium levels and normalize relaxation kinetics. To assess this, I cultured mature hiPSC cardiomyocytes on a Microelectrode Array (MEA) system capable of maintaining physiological conditions while measuring properties of cardiac electrophysiology. The cells were then matured using lactate enrichment to attain further maturity and grown in culture prior to MEA experimentation. Using the MEA, I have found that the QT interval for DMD hiPSC cardiomyocytes was significantly longer than isogenic controls. In current experiments, I am using this platform to validate the effect of L-type CCB compounds of interest in relation to DMD cardiomyopathy. The development of this novel platform may not only have broader implications for DMD drug discoveries and targeted therapies, but it can potentially serve as a powerful preclinical model for other neuromuscular disorders.

In the early stages of pathological hypertrophic cardiomyopathy, stress-induced signal transduction promotes the addition of new contractile units to maintain tensional homeostasis through poorly understood mechanisms. Our previous data shows that tension can be changed by altering the contractions within the cells. Inhibition of contraction by expression of D65A cTnC (point mutation on the calcium-binding site of troponin C) results in complete myofibrillar disarray. The mechanism behind the maintenance of myofibril and passive tension seen in these non-contractile CMs is not explained. Data has shown that microtubules provide resistance in the cell and go through rounds of rapid growth or disassembly based on the cell’s need which could be caused by the change in tensional homeostasis. With this in mind, my goal was to determine the role of microtubules in maintaining tensional homeostasis in response to changes in internal tension in CMs. Here, wild-type human induced pluripotent stem cell-derived cardiomyocytes were transduced to express cardiac troponin C with point mutations L48Q (hyper-contractile), I61Q (hypo-contractile), and D65A to study the effect of varying levels of contractility or internal load on microtubules. It was found that microtubular remodeling occurred where the noncontractile CMs had an increase in microtubule density. These noncontractile CMs were then cultures on nanopatterns that provided external tension via topological cues. In addition to the myofibrillar alignment seen previously, it was observed that microtubule density decreased further confirming that microtubules play a compensatory role in these CMs. The next step is to determine if a similar increase in microtubule density observed in the 2D culture is also seen on a tissue level by developing engineered heart tissues. This new data gives insight into how the microtubule remodeling in non-contractile and dysfunctional cardiomyocytes maintains tension in the early stages and its possible role in myofibril formation.

Hepatocellular Carcinoma (HCC) is an aggressive primary liver cancer that can be characterized with contrast-enhanced ultrasound (CEUS) by blood flow parameters. The most important blood-flow parameters are wash-in and washout, measured by the rate of contrast flow into and out of the liver and tumor. Currently blood flow is observed by eye, assessing the brightness of the contrast agent over time during the scan, and then scored on the Liver Reporting and Data System (LI-RADS) scale. This subjectivity delays diagnosis and treatment, and increases HCC mortality rate. To aid clinicians in a timely diagnosis, we are creating a quantitative Python algorithm to extract blood flow parameters from CEUS videos. I studied the existing MATLAB code our lab has created and translated its functionality into a Python executable. Since the two languages have different syntaxes and functions, I reconfigured the entire script to run in Python. This Python script will run faster than the MATLAB code, be able to accurately compute blood flow measures from CEUS scans, and be easier to distribute outside of academia. To extract parameters from CEUS liver videos of HCC patients, we processed the video using respiratory gating and motion compensation to ensure the tumor is on the scan and not moving around due to breathing. Then we analyzed the pixel brightness of the liver and tumor over the course of the scan to create a time-intensity curve (TIC). Finally, we compared various points on the curve to quantify the blood flow within the tumor and surrounding liver. Further iterations of the algorithm will include machine learning tools available in Python to increase the level of autonomy and thus the objectivity of the parameters. Using this algorithm, HCC diagnosis using LI-RADS will be timelier, allowing patients access to more effective treatment options.

Clearance of wound infections can be hindered by a bacterial biofilm; a complex extracellular matrix (EM) secreted by adherent bacteria that allows them to evade the host immune system and obviate antibiotics. A novel, synthetic peptide—known as an anti-α-sheet inhibitor—can disrupt biofilm stability by inhibiting the formation of amyloid fibrils, which contribute to the biofilm EM. This project aims to design and characterize alginate porous scaffolds that elute these synthetic peptides, for use as anti-biofilm wound dressings. The physical properties and peptide release kinetics of the scaffolds will be optimized for clinical applications, supported by in vitro efficacy studies with live bacteria. This project draws upon past work from the Bryers Research Group on engineering infection immunity and tissue scaffolds, in which biofilms are prevalent. Results of this project will provide an alternative approach to biofilm prevention, thus reducing the burden of biofilm-related infection complications.

Effective rapid tests have been developed to detect COVID-19 in nasal swabs. However, nasal swabs require a trade-off in which deeper insertion of the swab allows higher accuracy but causes greater patient discomfort. On the other hand, saliva can be collected non-invasively in large volumes. Therefore, this project aims to develop a point-of-care diagnostic device (“DiagnosDisk”) for the rapid and sensitive detection of SARS-CoV-2 nucleocapsid protein (NP) in saliva. The DiagnosDisk is a circular disk (25mm diameter) consisting of three layers, from top to bottom: a plastic adhesive capping sheet with a sample port, a hydroxylated nylon detection membrane, and an absorbent pad. Analytes are efficiently enriched by binding to anti-NP antibodies conjugated to temperature-responsive polymers (poly(N-isopropylacrylamide)) which aggregate upon heating. To prepare the sample, polymer- and gold-conjugated antibodies are mixed with the sample and bind to the target antigen, forming a sandwich immunocomplex. When flowing through the heated detection membrane (>37℃), the sandwich immunocomplexes aggregate and are captured on the membrane, producing a visual signal. I quantify the signal intensity on the membrane using ImageJ software. With this design, I hypothesize that the DiagnosDisk will have higher sensitivity than the lateral flow assay (LFA) by utilizing temperature-responsive polymers for sample enrichment and using larger sample volumes which increases the number of antigens that can be captured. From my initial testing, the DiagnosDisk enabled 2mL of buffer flow-through in 7.5 minutes, demonstrating a sample volume capacity 10-times that of most LFA. Next, I tested the DiagnosDisk with buffer samples spiked with 5ng/mL NP. Heated membrane pads produced signal intensities significantly greater than unheated membranes. I plan to apply the device to saliva samples next. Ultimately, by enabling rapid and sensitive detection of COVID-19 in saliva, the DiagnosDisk can help limit community transmission of the virus.

In the body, cells grow in the extracellular matrix (ECM), which presents biochemical and mechanical signals to the cells inside. Hydrogel biomaterials are water-laden polymer networks that can mimic the properties of the ECM, allowing controlled study of cellular behavior in vitro. Many cells are mechanosensitive, but mechanical cues other than stiffness have not been fully investigated. This project aims to develop a platform in which degradability and strain can be activated by a researcher bio-orthogonally. We have synthesized a cyclic peptide crosslinker for a synthetic poly(ethylene glycol) hydrogel that acts as a Boolean AND-gate: one half is degradable by cell-secreted enzymes, and the other half is degradable by sortase, a bacterial enzyme, added by a researcher. We quantified the degradation of hydrogels made with this crosslink via fluorescence release and demonstrated that degradation only occurs after exposure to both enzymatic inputs. We further demonstrated that cells encapsulated in this material retain strong viability. We predict that cells will be unable to spread in this material until after a researcher adds sortase. After sortase addition, we expect that contractile cells will be able to locally degrade the material, spread, and generate strain. We intend to quantify spreading and strain with encapsulated fibroblasts. We also plan to use this platform to study development, by encapsulating immature cardiac stem cells and investigating the effect of fibroblast driven strain as a model; we predict that strain will trigger further specification of these immature cells. In addition to understanding the pathways for development, this research may help identify new therapeutic targets for disease, and it will also inform new strategies to grow tissue in vitro that more closely mimic the native environment.

Across a variety of signaling pathways, soluble factors in the extracellular matrix bind to protein receptors that span the cell wall, thereby triggering an information cascade that affects cell activity or function. It follows that by controlling the binding of signaling factors to these receptors, cell behavior and activity can be guided with substantial precision. In this project, we aim to design a system that allows de novo-developed protein agonists and antagonists, referred to as binders, to be activated with a high degree of temporal and spatial control within cell-encapsulating hydrogels. Towards this end, we employ methods derived from protein semisynthesis and click chemistry to tether binders to the hydrogel polymer network and then subsequently photo-release them from the network. We expect a difference in the functionality of binders when they are bound to the network compared to when they are released through light exposure and solubilized, thus achieving light-dependent control of the binder-receptor interaction and cell activity. This system will be the first to employ de novo developed agonist and antagonist biomolecules for the interrogation and control of cellular behavior. In so doing, it will expand the tool box of biomaterial engineering to include finer control over cells grown in 3D matrices, with direct implications in fields as diverse as therapeutic development, regenerative medicine, and organ-on-a-chip engineering.

Cardiovascular disease is the leading cause of death worldwide. A key reason driving the high mortality of heart disease is that the heart is unable to regenerate any muscle that is lost due to injuries like heart attacks. Furthermore, there are no current therapeutics that promote the creation of new muscle. However, in the past decade, scientists have attempted to address this issue by using stem cell-derived cardiomyocytes (iPSC-CMs) to replace lost heart muscle. A key limitation preventing using this therapy in humans has been that cardiomyocytes derived from stem cells remain immature relative to adult cardiomyocytes, and these immature cells cause several complications when transplanted into an adult heart. Identifying cardiomyocyte maturation regulators is needed in order to further develop this technology and translate it to patients. Previous studies from our lab and others have identified the RNA binding protein Muscleblind-like protein 1 (MBNL1) as a key factor controlling muscle maturation. MBNL1 expression increases as the heart matures after birth and it controls expression of many critical regulators of cardiomyocyte maturation, however, MBNL1 has never been studied directly for promoting iPSC-CM maturation. In this project, I am testing the hypothesis that increasing MBNL1 expression will improve the maturity of iPSC-CMs. I am using a genetically engineered stem cell line in which I can overexpress MBNL1 and an isogenic control line to test my hypothesis. I have found that MBNL1 expression naturally increases over time in iPSC-CMs. Additionally, I have validated the MBNL1 overexpression system in iPSC-CMs. Finally, I have used this system to test my hypothesis that MBNL1 will increase iPSC-CM maturity by measuring well-described transcriptional and structural hallmarks of maturity. Ultimately, this project will aid in identifying MBNL1’s role in controlling cardiomyocyte maturation, helping further develop stem cell-based therapeutics to repair damaged heart tissue in humans.

Non-Human Primates (NHPs) have gained importance in neural engineering preclinical studies as their brains are relevant models to investigate and better understand neural function. The Yazdan lab uses optogenetics to control neuronal activity in order to develop stimulation-based therapies for neurological disorders such as stroke. These experiments require the implantation of various devices such as headposts, cranial chambers, electrode arrays, and optical windows. The use of head posts and cranial chambers requires customization to the curvature of the skull prior to implantation in order to prevent gaps that could introduce complications, including infection or decreased stability. Using an in-house method of NHP neurosurgery preparation that processes MRI data, we can develop 3D brain and skull models. This technique has allowed for chambers to be customized and implanted chronically in two NHPs. My project builds off of this implementation by creating custom chambers for future implantation surgeries and designing custom-fit headposts, which had never been done before. In order to design these components, I extracted the skull and brain using custom Matlab code which allowed for the craniotomy location to be determined and provided a footprint for the chamber and headpost implants. I then imported the skull extraction into a design software where the chambers and headposts could be built off of to ensure a tight fit to the skull. With the components designed, I will 3D print the brain, skull, chamber, and headpost to be assembled together. This platform will simulate the surgical and experimental setup, which provides a template for various experimental components to be modified and tested. It will provide a simple and affordable solution for neurosurgical planning, reducing in-surgery and in-experiment complications. This model's versatility, ease of use and low cost allow for further expansion to other labs and to a wider scope of surgeries.

The choice to terminate a pregnancy is rarely an easy one. It is critical that the experience of pregnancy termination is made as comfortable and convenient as possible as women grapple with the mental and physical challenges that arise from their choice to abort. The current standard protocol for at-home medical termination of pregnancy involves the patient-mediated oral misoprostol delivery 24-48 hours after mifepristone. This timed delivery ensures that misoprostol can trigger contractions after the cervix is dilated by mifepristone. This presents a challenge for patients, who have to manage the side effects of mifepristone while also timing the delayed dosage window for misoprostol. Combining these two medications into the single dose modality that instantly releases mifepristone and ensures the delayed release of misoprostol will improve the at-home medical abortion process for patients globally. The nanoparticle encapsulation of misoprostol for oral delivery offers the prospect of delayed release and release modulation through the alteration of variables such as molecular weight and co-polymer polymerization and other formulation parameters. Herein, we aim to use established protocols for nanoparticle encapsulation to fabricate and characterize misoprostol nanoparticles. By varying molecular weight and copolymerization parameters, we aim to tailor misoprostol release and evaluate the efficacy of different encapsulation approaches. We intend to investigate the particle size, entrapment efficiency, shelf-stability, and in vitro release of misoprostol nanoparticles in PBS and simulated gastrointestinal fluid. In doing so, we aim to provide a proof-of-concept of formulating misoprostol into nanoparticles to demonstrate encapsulation and sustained release. This understanding can contribute to the development of a single-dosage modality to meaningfully improve the comfort and ease of at-home medical abortions.

Alzheimer’s Disease (AD) is a neurodegenerative disorder characterized by the aggregation of the amyloid-β (Aβ) peptide into fibrillar β-sheet plaques. While Aβ plaques have been the historical focus in the study of AD, it has recently been shown that Aβ also forms oligomeric intermediates of a novel α-sheet secondary structure along its aggregation pathway. Furthermore, the formation of these α-sheet oligomers is correlated with the neuronal death and subsequent cognitive impairment seen in AD, and it begins up to 10-20 years before plaque formation or symptom development. In this project, I utilize a novel diagnostic technique known as the soluble oligomer binding (SOBA) assay to characterize the in vitro aggregation of Aβ through α-sheet detection in various timepoints of incubated Aβ. Additionally, I use SOBA to characterize AD phenotypes in transgenic mouse models using α-sheet detection in brain homogenate samples. SOBA selectively detects α-sheet oligomers using plate surfaces functionalized with in-house α-sheet peptide designs, followed by the introduction of amyloid samples and then amyloid-specific primary and secondary antibodies for detection. This research allows for the further elucidation of Aβ's structure and aggregation pathway, paving the way for future treatment methods for AD. Finally, I aim to translate my work with AD to develop SOBA for the system of Type 2 Diabetes (T2D) to detect the presence of toxic islet amyloid polypeptide (IAPP), the amyloid peptide implicated in T2D. This will be achieved through the systematic tuning of antibody systems and other procedural parameters such as plate washing methods. Preliminary results have shown the ability of SOBA to detect α-sheet in synthetic IAPP samples with concentrations as low as 10 pM. This novel method for toxic IAPP detection has the potential for T2D diagnosis earlier in disease progression.

Human-machine interfaces, which map signals measured from a user into inputs for a device, hold promise to allow efficient and individualized device usage, whether for rehabilitation or recreation. Surface electromyography (sEMG) can non-invasively measure muscle activity through the skin to provide many potential user inputs to control a computer. Despite sEMG’s promise for user-controlled programs, clinical and commercial success has been low, in part due to poor user training regimens. sEMG-based interfaces are often unintuitive to learn, and not all motions will contribute equally to control, due to inherent limitations in electrode placement and sensitivity. Training users by presenting them with visual feedback of which actions can contribute to the task may enhance learning by discouraging strategies using undetectable motions. Thus, I propose training users to learn an adaptive decoder using an sEMG radar plot which displays real-time visualizations of the user’s sEMG signals, with each channel arranged in a circle such that motions appear as unique conformations of the radar. I hypothesize that showing users the radar plot before completing a sEMG-controlled computer task will confer greater task success. To test this, I conducted a set of experiments with adult subjects using the radar plot as brief, pre-task training for a two-dimensional, cursor-control game, comparing user performance between those trained and untrained. I anticipate that users who received radar plot training will demonstrate faster task learning and lower tracking error. Such a result would shed new light on how to streamline sEMG-based interface training, and may encourage further modifications or improvements of the radar plot. Optimizing the human-machine interface training process will be integral to their path to clinical and commercial success.

Over the past 50 years, breast cancer mortality rates in high-income countries have dropped significantly while low and middle-income countries (LMICs) have made few advancements towards early detection and rapid diagnosis. A lack of infrastructure in LMIC healthcare settings causes patients to receive their diagnosis several months after testing, while some never hear back due to travel and communication barriers. We have partnered with Peace and Love Hospitals in Ghana to develop CoreView Imaging on the Needle (ION), a portable, rapid point-of-care (POC) diagnostic system for use in LMIC. Our team developed ION after discovering breast core needle biopsies (CNBs) were too delicate and sticky for transport in our original millifluidic design. Our research is exploring methods of integrating the ION system into CoreView to improve the testing reliability and image quality of biopsies for diagnosis by designing, testing, and iterating fixtures to image CNBs while still on the needle. We modeled our prototypes in Solidworks and printed them using a Prusa 3D printer. Our goal is to load the CNB under a microscope, stain the tissue nuclei, compress its surface against a clear glass window, and image the biopsy surface within 5 minutes. We are using Microscopy with UV Surface Excitation (MUSE) imaging to take panoramic images along the 20 mm biopsy length. The current prototype takes 1 minute to prepare the biopsy (stain and load into ION fixture), 0.5 minutes for each MUSE image and step to the next image (total of 6 images), and 0.5 minutes to unload the biopsy, resulting in 4.5 minutes between biopsy collection and obtaining images for diagnosis. The improved image quality and expected high reliability verifies the success of our ION design for further validation testing to provide earlier breast cancer detection and rapid treatment to rural patients.

905,700 people were diagnosed with liver cancer globally in 2020, and this number is projected to rise more than 55% by 2040. The primary treatment for liver cancer is a partial hepatectomy, in which up to two-thirds of the liver is removed and then subsequently allowed to regenerate. The liver is the only organ that has this regenerative capacity, and while the regeneration process has been evaluated, the triggers of the intracellular pathways are not clear. Further research would be beneficial to the improvement of liver disease treatment. A primary hypothesis suggests a trigger of regeneration is the increased shear stress on liver sinusoidal endothelial cells(LSECs) after a partial hepatectomy channels portal vein flow through the reduced liver. However, current in vitro models that display liver regeneration do not have the ability to model the effects of flow on the liver, due to unrepresentative vasculature, no incorporation of constant flow, and a lack of organ specific endothelial cells. The development of a representative in vitro model to display this process is key, and to address the current limitations, I use the Zheng lab’s previously developed, perfusable, vascularized collagen “vessels.” In this project we 1) create a perfusable model of liver regeneration and analyze the survival of hepatocyte aggregates within it and 2) screen donor LSECs to identify their productivity through albumin production measurements and immunofluorescent staining to identify cell purity and presence of LSEC markers, which is essential if we are to use them in our vessels. Based on previous research, we expect to see greater hepatocyte function (quantified by higher albumin production) within the vascularized vessel due to the more physically representative environment. This model provides a baseline to test the effects of flow on regeneration and has many future applications including drug testing and disease modeling.

Regenerative medicine compromises to repair and replace cells, tissues, and organs damaged by disease or aging. To control cell fate in regenerative medicine, methods enabling irreversible and spatiotemporally controlled protein activation would be beneficial, particularly to those that could be applied for both inter- and extracellular activation. Furthermore, an ideal strategy could be applied to virtually any protein and afford rapid activation. In my work, I have sought to develop and exploit such a method through protein photochemistry; in response to mild and cytocompatibile light exposure, user-specified proteins are irreversibly assembled into their bioactive form. I have optimized the processes for hydrogel formation and modifications to increase cell viability. Results further inform that I can biochemically customize the landscape both intra- and extra-cellularly with a photoactivatable mCherry construct. Moving forward, I intend to apply this technique to activate epidermal growth factors and other proteins in multiple physiological systems. Successful protein photoactivation provides a potential, less invasive mechanism for controlling cells in the extracellular matrix for tissue engineering and regenerative medicine.

Antiretroviral therapy (ART) prevents the progression of human immunodeficiency virus (HIV) by suppressing viral load, limiting transmission. Of the ~20 million people receiving ART, 30-40% do not maintain adequate medication adherence, resulting in treatment failure and drug resistance. Monitoring HIV medication adherence improves the efficacy of ART but requires bulky and expensive instruments that are not widely accessible at the point-of-need (e.g. doctor’s office or patient’s home), so a rapid and accessible diagnostic alternative is necessary. Our group developed the REverSe TRanscrIptase Chain Termination (RESTRICT) enzymatic assay to provide rapid and inexpensive measurement of HIV drug adherence by measuring antiretroviral drug activity indicated by fluorescence. However, one current limitation of RESTRICT is the need for trained operators to complete multiple precisely timed steps required in the enzymatic activity assay. We aim to create a 3D-printed microfluidic device that will automate the liquid handling steps required for RESTRICT via precisely tuned capillary action for rapid and user-friendly measurement of antiretroviral drugs. To that end, we first demonstrated a proof of concept by creating a microchip with a controlled 15 minute liquid delivery time and consistent liquid extraction. Through optimization of 3D-printing methods, channel geometry, and surface treatment, we created a microchip designed to deliver liquid in 14.64 minutes that ran experimentally in 19.12 ± 2.33 minutes. In the future, we will demonstrate the feasibility of RESTRICT run on-chip and fluorescence measured off-chip by testing clinically-relevant drug concentrations using the controlled liquid delivery time and liquid extraction methods developed. By creating an automated and rapidly fabricated microfluidic chip for therapeutic drug monitoring, we hope to achieve a hands-off device that removes external manipulation to increase the accessibility of RESTRICT-on-a-chip for point-of-need settings without specialized equipment or highly trained operators.

The extra cellular matrix (ECM) is a complex, heterogenous environment that plays an important role in cellular functions such as proliferation, signaling, movement, and differentiation. The mechanical properties of the ECM vary spatially and temporally, across and within tissues, i.e., during development and disease progression. 3D biomaterial platforms, such as hydrogels – water-swollen polymeric networks—provide a greater understanding of matrix-cell interactions and can be used to study drug delivery and basic disease mechanisms. My research works to create a double network (DN) hydrogel system that allows for spatial control of ECM mechanics in 3D. Our system contains two different polymer networks, one of which uses light polymerization. I have optimized concentrations of multiple gel components and gel light exposure conditions to allow for accurately patterned stiffnesses within the gels. Currently, I am encapsulating live cells to study the amount of cell spreading and movement in the stiff and soft regions of the gels over the course of a week. I then fix, stain, and image each gel to quantify relative cellular spreading. Additionally, I have synthesized multiple components necessary for gel formation, cultured enzyme producing bacteria to degrade formed gels, and performed western blotting to analyze cellular protein concentrations. Imaging results have shown the DNs and the patterning process are cytocompatible. Current experiments have shown differences in fibroblast spreading between stiff and soft regions; future results are expected to show differences in protein expression within mechanosensitive pathways between patterning conditions. Using multiple, intertwined hydrogel networks, I have engineered a dynamic, heterogenous model of the ECM, enabling me to study cellular responses to mechanical stimuli. Accurate modeling of the ECM will allow for a better understanding of how diseases such as breast cancer progress based on differences in environmental stiffness and provide an in vitro platform for future cellular response research.

Atrial fibrillation (AFib) is the most common sustained cardiac arrhythmia, contributing to significant morbidity and mortality worldwide. Patient-specific computational models of the left atrium are currently studied to predict characteristics of reentrant activity that promotes fibrillation. However, current models’ patient-specificity is limited to anatomical structure and the distribution of disease-related remodeling (fibrosis), whereas electrical properties of cells and tissue are based on literature values. In cases where patients are clinically known to present with either AFib or atrial flutter (AFl), this lack of personalization can lead to inaccuracies in simulation outcomes (e.g., AFib-like behavior in simulations for a patient who actually had AFl, or vice-versa). My goal was to derive parameter sets that favor the initiation of one type of arrhythmia or the other (AFib or AFl). Ten fibrotic left atria were reconstructed from late-gadolinium enhanced (LGE)-MRI scans and the bioelectric parameter space (comprising ion channel expression levels and impulse propagation rates) was explored using a Taguchi L27 Design of Experiments (DoE) approach. Arrhythmias were induced by initializing four atrial regions to different phases of the action potential under each parameter permutation. I ran 300 simulations and manually classified each arrhythmia episode as either AFib- or AFl-like based on prior definitions. I pinpointed a pro-AFl parameter set – bioelectrical conditions under which 89% of all induced arrhythmias were AFl and only 11% were AFib. The pro-AFib parameter set in these preliminary simulations was comparatively less robust (61% vs. 39% for AFib vs. AFl inductions, respectively). My future work on this project will establish stronger relationships between model configurations and simulation outcomes by probing a wider array of possible parameters in a larger population of patient-specific models. Data from the present study will guide future simulations to accurately tailor models to represent the arrhythmic state in patients predisposed to AFl.

Genetically encoded fluorescent indicators (GEFI) change fluorescence level under microscope following a conformational change when bound to a target molecule, and can be used to visualize spatiotemporally specific biological processes involving a targeted molecule. Various imaging analysis tools exist to analyze the non-temporal fluorescent cell data, however there was no industry standard for the pipeline used to analyze molecular dynamics when imaged with GEFI in time-series experiments. This project aimed to develop a computational pipeline that analyzed the fluorescent readout of single cells in spatiotemporal experiments that utilized GEFI. The pipeline included both segmentation of cells, utilizing Cellpose, an existing deep learning-based generalizable and highly efficient segmentation program, and tracking of single cells across all frames. I personally contributed to the design, implementation, and testing of the tracking component of the pipeline. The tracking algorithm was designed using unsupervised machine learning, specifically k-means clustering with convolutional neural network feature extraction techniques. The pipeline was implemented using Python and made available and open source, accessible through Google Colaboratory for a more user friendly version, as well as Github for more thorough documentation and generalizability. Ultimately, this project aimed to minimize bias to result in more accurate and efficient high-throughput investigation of molecular dynamics when using fluorescent probes for dynamic cell imaging. Preliminary results demonstrated the effectiveness of the pipeline in tracking cells across various time points and provided a foundation for future optimizations and applications.

Evaluating the risk of developing breast cancer is an important aspect of cancer care as it can allow for more tailored screening strategies and preventative therapies. Clinicians use multiple measures to determine a patient’s risk of developing breast cancer, including breast density on mammography and genetic mutations. Background parenchymal enhancement (BPE) on magnetic resonance imaging (MRI) has shown promise to improve stratification of breast cancer risk in women at high-risk of cancer development. BPE is the increase in signal intensity of normal breast tissue on dynamic contrast-enhanced (DCE) MRI after the administration of contrast agent. Despite BPE having an association with an increased risk of breast cancer development, the biological basis of this increased enhancement is unknown. The aim of this study is to investigate what biologically drives BPE by connecting quantitative MRI measurements with pathological markers from normal breast tissue. Our study cohort includes women that received prophylactic mastectomies and DCE-MRI scans acquired ≤1 year before surgery. From mastectomy specimens, pathological measures of COX-2, VEGF, and Ki-67 are used to measure inflammation, vascular recruitment, and proliferation, respectively. To quantify BPE, I used in-house software to correct pre-contrast images using N4 bias field correction and segment the whole breast. I then applied the breast mask to the pre-contrast MRI and used fuzzy c-means clustering to automatically segment fibroglandular tissue (FGT) from surrounding fat, generating an FGT mask. This mask was then applied to the DCE-MRI series, which includes pre- and post-contrast images, to calculate BPE, which is the mean percent enhancement across FGT. As part of ongoing work, I will obtain more specific measurements in quadrants of the breast from which the pathology specimen was derived. I will then correlate BPE measurements to the pathology measures to determine if any associations exist between BPE and inflammation, vascular recruitment, and proliferation.

Diffusion-weighted imaging (DWI) shows great potential for improving breast cancer detection and diagnosis. Primary findings from the ECOG-ACRIN A6702 multi-site, multi-vendor clinical trial indicate that DWI apparent diffusion coefficient (ADC) values may help reduce false positives and unnecessary biopsies. Gradient nonlinearity (GNL) correction was previously found to improve the accuracy of ADC mapping within and across MRI vendor systems. In this study, we evaluated the impact of GNL correction on breast lesion ADC measures in the A6702 dataset. The dataset comprised 81 suspicious breast lesions (28/81 malignant) in 67 women. Standardized DWI scans were acquired across 9 different MRI scanners. ADC maps were created from DWI scans, and ADC values were measured for each lesion. Direction-averaged GNL correction maps were constructed based on scanner-specific gradient specifications. ADC map correction was then performed through pixel-wise scaling by the GNL correction maps using custom software developed in MATLAB. Lesion ADCs before and after GNL correction were compared using a two-tailed z-test. ADC diagnostic performance (benign vs. malignant) was evaluated using area under the receiver-operating-characteristic-curve (AUC), and optimal ADC cutoffs were chosen to maximize specificity while maintaining 100% sensitivity. GNL-corrected lesion ADCs were significantly lower than uncorrected ADCs (1.12±0.29 vs 1.17±0.30x10-3mm2/s, p<0.001). GNL error in lesion ADCs varied across gradient systems (mean ∆ADCvendorA=0.14±0.08, ∆ADCvendorB=0.03±0.02, ∆ADCvendorC =0.004±0.01, p<0.001). GNL correction produced a slightly lower optimal ADC cutoff (1.33 vs. 1.35x10-3mm2/sec). However, no overall difference in diagnostic performance was detected: AUCuncorrected=0.78 (95% CI 0.68-0.88), AUCcorrected=0.79 (95% CI:0.69-0.89), p=0.22, and 18% potential biopsy reduction for both. This study showed GNL substantially affects lesion ADC measures, with significant variability across different vendor platforms. These findings suggest that GNL correction should be implemented to ensure uniformity and consistency in diagnostic breast lesion ADC measures across MRI platforms, especially for multi-center clinical studies.

Capillary microfluidics devices automate point-of-care diagnostic assays because of their instrument-free operation, small size, and low material cost. However, capillary microfluidics currently require expensive fabrication instruments limiting rapid prototyping and deployment in low-resource settings. State-of-the-art capillary microfluidics are fabricated by using digital light project stereolithography (DLP-SLA) 3D-printers that offer high resolution (40 µm) but are expensive ($10,000 – $20,000). Liquid crystal display (LCD) SLA printers have recently emerged with comparable resolution and much lower cost ($300 – $1000). However, our initial experiments with LCD-SLA printers exhibited defects and post-processing issues. Our goal is to optimize the fabrication process of capillary microfluidics using an inexpensive LCD-SLA printer (Anycubic Photon Mono 6K, ~$400) and calibrate performance relative to a DLP-SLA printer (CADWorks Pr 4K, ~$15,000). By varying printing parameters including UV power, exposure time, layer height, retraction speed, and resin choice, we found optimal conditions that produced microfluidic channels with comparable dimensions on the Anycubic and CADWorks printers (i.e. coefficient of variation <20%). With printer settings of 40% UV Power, 1.8s exposure time, 20 µm layer height, and 0.1 mm/s retraction speed with CADWorks clear resin, the Anycubic 3D printer produced microchannels down to 100 μm, the smallest feature size we achieved with the CADWorks printer. Optimizing 3D-printing of capillary microfluidics using inexpensive LCD-SLA printers like the Anycubic has the potential to enable rapid prototyping of point-of-care diagnostics in low-resource settings. Next steps include printing more complex microfluidic components including domino valves and trigger valves.