More Like this

1. A Deep Learning Framework for Sickle Cell Disease Microfluidic Biomarker Assays

   https://doi.org/10.1182/blood-2020-141693  

   Praljak, Niksa ; Shipley, Brandon ; Gonzales, Ayesha ; Goreke, Utku ; Iram, Shamreen ; Singh, Gundeep ; Hill, Ailis ; Gurkan, Umut A. ; Hinczewski, Michael ( November 2020 , Blood)

   Introduction: Vaso-occlusive crises (VOCs) are a leading cause of morbidity and early mortality in individuals with sickle cell disease (SCD). These crises are triggered by sickle red blood cell (sRBC) aggregation in blood vessels and are influenced by factors such as enhanced sRBC and white blood cell (WBC) adhesion to inflamed endothelium. Advances in microfluidic biomarker assays (i.e., SCD Biochip systems) have led to clinical studies of blood cell adhesion onto endothelial proteins, including, fibronectin, laminin, P-selectin, ICAM-1, functionalized in microchannels. These microfluidic assays allow mimicking the physiological aspects of human microvasculature and help characterize biomechanical properties of adhered sRBCs under flow. However, analysis of the microfluidic biomarker assay data has so far relied on manual cell counting and exhaustive visual morphological characterization of cells by trained personnel. Integrating deep learning algorithms with microscopic imaging of adhesion protein functionalized microfluidic channels can accelerate and standardize accurate classification of blood cells in microfluidic biomarker assays. Here we present a deep learning approach into a general-purpose analytical tool covering a wide range of conditions: channels functionalized with different proteins (laminin or P-selectin), with varying degrees of adhesion by both sRBCs and WBCs, and in both normoxic and hypoxic environments. Methods: Our neuralmore »networks were trained on a repository of manually labeled SCD Biochip microfluidic biomarker assay whole channel images. Each channel contained adhered cells pertaining to clinical whole blood under constant shear stress of 0.1 Pa, mimicking physiological levels in post-capillary venules. The machine learning (ML) framework consists of two phases: Phase I segments pixels belonging to blood cells adhered to the microfluidic channel surface, while Phase II associates pixel clusters with specific cell types (sRBCs or WBCs). Phase I is implemented through an ensemble of seven generative fully convolutional neural networks, and Phase II is an ensemble of five neural networks based on a Resnet50 backbone. Each pixel cluster is given a probability of belonging to one of three classes: adhered sRBC, adhered WBC, or non-adhered / other. Results and Discussion: We applied our trained ML framework to 107 novel whole channel images not used during training and compared the results against counts from human experts. As seen in Fig. 1A, there was excellent agreement in counts across all protein and cell types investigated: sRBCs adhered to laminin, sRBCs adhered to P-selectin, and WBCs adhered to P-selectin. Not only was the approach able to handle surfaces functionalized with different proteins, but it also performed well for high cell density images (up to 5000 cells per image) in both normoxic and hypoxic conditions (Fig. 1B). The average uncertainty for the ML counts, obtained from accuracy metrics on the test dataset, was 3%. This uncertainty is a significant improvement on the 20% average uncertainty of the human counts, estimated from the variance in repeated manual analyses of the images. Moreover, manual classification of each image may take up to 2 hours, versus about 6 minutes per image for the ML analysis. Thus, ML provides greater consistency in the classification at a fraction of the processing time. To assess which features the network used to distinguish adhered cells, we generated class activation maps (Fig. 1C-E). These heat maps indicate the regions of focus for the algorithm in making each classification decision. Intriguingly, the highlighted features were similar to those used by human experts: the dimple in partially sickled RBCs, the sharp endpoints for highly sickled RBCs, and the uniform curvature of the WBCs. Overall the robust performance of the ML approach in our study sets the stage for generalizing it to other endothelial proteins and experimental conditions, a first step toward a universal microfluidic ML framework targeting blood disorders. Such a framework would not only be able to integrate advanced biophysical characterization into fast, point-of-care diagnostic devices, but also provide a standardized and reliable way of monitoring patients undergoing targeted therapies and curative interventions, including, stem cell and gene-based therapies for SCD. Disclosures Gurkan: Dx Now Inc.: Patents & Royalties; Xatek Inc.: Patents & Royalties; BioChip Labs: Patents & Royalties; Hemex Health, Inc.: Consultancy, Current Employment, Patents & Royalties, Research Funding.« less
2. Determining Kinetic Parameters for a Mathematical Model to Optimize Growth of Cancer-Fighting T Cells in a Novel Bioreactor

   Moore, Z ( November 2021 , Annual Biomedical Research Conference for Minority Students (ABRCMS) Virtual)

   T cell transfer immunotherapy is a highly effective cancer treatment in which the immune system’s inherent ability to fight cancer is amplified by increasing the amount of T cells that are deemed most active within a patient. T cells are a lymphocyte produced as an immune response to cancerous cells. Despite this advanced form of biological therapy, current T cell expansion methods are inefficient, resulting in high manufacturing costs, which brings question to the efficacy of T cell therapies. To address this issue, the recent development of a centrifugal bioreactor aims to rapidly expand T cells for cancer immunotherapy treatments at higher cell densities and in a shorter amount of time compared to current systems on the market. We hypothesize that by producing a mathematical model of a proof-of-concept T cell line to determine substrate consumption and metabolite production over time, we will be able to optimize growth of the cell line in the bioreactor. A series of three studies were performed to produce the growth model: (1) measuring yield coefficients of lactate, ammonium ion, and glucose, (2) determining the Monod constant and maximum specific growth rate, and (3) finding critical metabolite concentrations. To measure yield coefficients, T cells weremore »grown in a 6-well plate at 1 x 105 cells/mL in 4 mL of medium with 100 uL samples taken and frozen each day over a 5-day period. At the end of the study, samples are thawed and used with lactate and ammonium assay kits for microplate reading to determine metabolite levels over time. To determine the Monod constant and maximum specific growth rate, T cells were grown in 12-well plates at pre-calculated varying glucose concentrations in 4 mL of medium in triplicates. Cells were counted for a minimum of six days to determine expansion over time to develop a linearized growth plot. To find critical metabolite concentrations, ammonium and lactate were added to glucose-free T cell medium at four different concentrations in triplicates utilizing a 12-well plate with a seeding density of 1 x 105 cells/mL in 4 mL of medium. The T cells then remained undisturbed in culture and were counted on day three. Once all parameters are determined, we can apply them to the growth model to determine levels of glucose, lactate, and ammonium as the T cells grow to high densities in the bioreactor and, as a result, optimize the manufacturing process for cancer immunotherapy treatments.« less
3. Identification of Cellular Genes Involved in Baculovirus GP64 Trafficking to the Plasma Membrane

   https://doi.org/10.1128/jvi.00215-22  

   Hodgson, Jeffrey J. ; Buchon, Nicolas ; Blissard, Gary W. ( June 2022 , Journal of Virology)

   Sandri-Goldin, Rozanne M. (Ed.)

   ABSTRACT The baculovirus envelope protein GP64 is an essential component of the budded virus and is necessary for efficient virion assembly. Little is known regarding intracellular trafficking of GP64 to the plasma membrane, where it is incorporated into budding virions during egress. To identify host proteins and potential cellular trafficking pathways that are involved in delivery of GP64 to the plasma membrane, we developed and characterized a stable Drosophila cell line that inducibly expresses the AcMNPV GP64 protein and used that cell line in combination with a targeted RNA interference (RNAi) screen of vesicular protein trafficking pathway genes. Of the 37 initial hits from the screen, we validated and examined six host genes that were important for trafficking of GP64 to the cell surface. Validated hits included Rab GTPases Rab1 and Rab4 , Clathrin heavy chain , clathrin adaptor protein genes AP-1-2β and AP-2μ , and Snap29 . Two gene knockdowns ( Rab5 and Exo84 ) caused substantial increases (up to 2.5-fold) of GP64 on the plasma membrane. We found that a small amount of GP64 is released from cells in exosomes and that some portion of cell surface GP64 is endocytosed, suggesting that recycling helps to maintain GP64 atmore »the cell surface. IMPORTANCE While much is known regarding trafficking of viral envelope proteins in mammalian cells, little is known about this process in insect cells. To begin to understand which factors and pathways are needed for trafficking of insect virus envelope proteins, we engineered a Drosophila melanogaster cell line and implemented an RNAi screen to identify cellular proteins that aid transport of the model baculovirus envelope protein (GP64) to the cell surface. For this we developed an experimental system that leverages the large array of tools available for Drosophila and performed a targeted RNAi screen to identify cellular proteins involved in GP64 trafficking to the cell surface. Since viral envelope proteins are often critical for production of infectious progeny virions, these studies lay the foundation for understanding how either pathogenic insect viruses (baculoviruses) or insect-vectored viruses (e.g., flaviviruses, alphaviruses) egress from cells in tissues such as the midgut to enable systemic virus infection.« less
4. Administration of FVIII-Expressing Human Placental Cells to Juvenile Sheep Yields Multi- Organ Engraftment, Therapeutic Plasma FVIII Levels and Alter Immune Signaling Pathways to Evade FVIII Inhibitor Induction

   Trevisan, Brady ; Rodriguez, Martin ; Dizon, Jacqueline ; George, Sunil ; Shields, Jordan ; Lankford, Shannon ; Combs, Rebessa ; Owen, John ; Atala, Anthony ; Doering, Christopher ; et al ( December 2021 , ASH)

   Administration of FVIII-Expressing Human Placental Cells to Juvenile Sheep Yields Multi-Organ Engraftment, Therapeutic Plasma FVIII Levels and Alter Immune Signaling Pathways to Evade FVIII Inhibitor Induction 63rd ASH Annual Meeting and Exposition, December 11-14, 2021, Georgia World Congress Center, Atlanta, GA Program: Oral and Poster Abstracts Session: 801. Gene Therapies: Poster III Hematology Disease Topics & Pathways: Bleeding and Clotting, Biological, Translational Research, Hemophilia, Genetic Disorders, Immune Mechanism, Diseases, Gene Therapy, Therapies, Adverse Events, Biological Processes, Transplantation Monday, December 13, 2021, 6:00 PM-8:00 PM We have previously reported that normal juvenile sheep that received weekly intravenous (IV) infusions of human (n=3) or an expression/secretion-optimized, bioengineered human/porcine hybrid (ET3) FVIII protein (n=3) for 5 weeks (20 IU/kg) developed anti-FVIII inhibitory antibodies (10-116 BU, and IgG titers of 1:20–1:245) by week 3 of infusion. By contrast, the IV infusion, or IP administration, of human placental mesenchymal cells (PLC) transduced with a lentiviral vector encoding a myeloid codon-optimized ET3 transgene (PLC-mcoET3) to produce high levels of ET3 protein (4.9-6IU/10^6 cells/24h) enabled the delivery of FVIII without eliciting antibodies, despite using PLC-mcoET3 doses that provided ~20-60 IU/kg ET3 each 24h to mirror the amount of FVIII protein infused. In addition, we showed that themore »route of PLC-mcoET3 administration (IP vs IV) did not impact the resultant plasma FVIII levels, with animals in these two groups exhibiting mean increases in FVIII activity (quantified by aPTT) of 30.9% and 34.2%, respectively, at week 15 post-treatment. Here, we investigated whether the sites and levels of PLC-mcoET3 engraftment were dependent upon the route of administration and performed s sheep-specific multiplexed transcriptomic analysis (NanoString) to define the immune signaling pathways that thwarted FVIII/ET3 protein immune response when ET3 was delivered through PLC. Tissue samples were collected from various organs at euthanasia and RT-qPCR performed using primers specific to the mcoET3 transgene, to the human housekeeping transcript GAPDH, and to sheep GAPDH, to quantify PLC-mcoET3 tissue engraftment, and normalize the results. RT-qPCR demonstrated PLC-mcoET3 engrafted, in both IP and IV groups, in all the organs evaluated (liver, lung, lymph nodes, thymus, and spleen). Animals that received PLC-mcoET3 via the IP route displayed higher overall levels of engraftment than their IV counterparts. The spleen was the preferential organ of engraftment for both IP and IV groups (IP:2.41±1.97%; IV: 0.64±0.54%). The IP group exhibited significantly higher engraftment in the left lobe of the liver (IP: 1.36±0.35%; IV: 0.041±0.022%), which was confirmed by immunohisto-chemistry (IHC) with an antibody to the human nuclear antigen Ku80 and ImageJ analysis (IP:5.24±3.36%; IV: 0±0). Of note is that the IP route resulted in higher levels of engraftment in the thymus, while IV infusion yielded higher levels of PLC-mcoET3 in lymph nodes. Analysis of H&E-stained tissues demonstrated they were devoid of any abnormal histologic changes and exhibited no evidence of hyperplasia or neoplasia, supporting the safety of the cell platform, irrespective of the route of administration. To date, NanoString analysis of PBMC collected at day 0, week 1, and week 5 post-infusion demonstrated that animals who received FVIII protein had upregulation of UBA5 and BATF, genes involved in antigen processing and Th17 signaling pathways, respectively. Although both IV and IP recipients of PLC-mcoET3 also had an increase in BATF, the IV group exhibited upregulation of BTLA, a gene involved in immune-tolerance, and downregulation of NOTCH and DDL1, involved in T cell differentiation, as well as MAPK12 and PLCG1, genes involved in proinflammatory cytokine regulation and T signaling within the Th17 signature. In IP recipients, BTLA, NOTCH, and DLL1 were all downregulated. Since ET3-reactive Th1 cells were not present in any of the treated animals, it is possible that the Th17 cells are responsible for the inhibitory antibodies seen in the juvenile sheep treated with FVIII/ET3 protein, while in animals receiving PLC-mcoET3, downregulation of genes involved in T cell differentiation and proinflammatory cytokine signaling keeps the immune system in check to avoid an immune response. Disclosures: Doering: Expression Therapeutics: Divested equity in a private or publicly-traded company in the past 24 months. Spencer: Expression Therapeutics: Divested equity in a private or publicly-traded company in the past 24 months.« less
5. A microfluidic electrochemical flow cell capable of rapid on-chip dilution for fast-scan cyclic voltammetry electrode calibration

   https://doi.org/10.1007/s00216-020-02493-z  

   Delong, Lauren M. ; Li, Yuxin ; Lim, Gary N. ; Wairegi, Salmika G. ; Ross, Ashley E. ( February 2020 , Analytical and Bioanalytical Chemistry)

   Here, we developed a microfluidic electrochemical flow cell for fast-scan cyclic voltammetry which is capable of rapid on-chip dilution for efficient and cost-effective electrode calibration. Fast-scan cyclic voltammetry (FSCV) at carbon-fiber microelectrodes is a robust electroanalytical technique used to measure subsecond changes in neurotransmitter concentration over time.Traditional methods of electrode calibration for FSCV require several milliliters of a standard. Additionally, generating calibration curves can be time-consuming because separate solutions must be prepared for each concentration. Microfluidic electrochemical flow cells have been developed in the past; however, they often require incorporating the electrode in the device, making it difficult to remove for testing in biological tissues. Likewise, current microfluidic electrochemical flow cells are not capable of rapid on-chip dilution to eliminate the requirement of making multiple solutions. We designed a T-channel device, with microchannel dimensions of 100 μm × 50 μm, that delivered a standard to a 2-mm-diameter open electrode sampling well. A waste channel with the same dimensions was designed perpendicular to the well to flush and remove the standard. The dimensions of the T-microchannels and flow rates were chosen to facilitate complete mixing in the delivery channel prior to reaching the electrode. The degree of mixing was computationally modeledmore »using COMSOL and was quantitatively assessed in the device using both colored dyes and electrochemical detection. On-chip electrode calibration for dopamine with FSCV was not significantly different than the traditional calibration method demonstrating its utility for FSCV calibration. Overall, this device improves the efficiency and ease of electrode calibration.« less