Methods of Machine Learning-Based Chimeric Antigen Receptor Immunological Synapse Quality Quantification.

Chimeric Antigen Receptor (CAR)-mediated immunotherapy shows promising results for refractory blood cancers. Currently, six CAR-T drugs have been approved by U.S. Food and Drug Administration (FDA). Theoretically, CAR-T cells must form an effective immunological synapse (IS, an interface between effective cells and their target cells) with their susceptible tumor cells to eliminate tumor cells. Previous studies show that CAR IS quality can be used as a predictive functional biomarker for CAR-T immunotherapies. However, quantification of CAR-T IS quality is clinically challenging. Machine learning (ML)-based CAR-T IS quality quantification has been proposed previously.Here, we show an easy-to-use, step-by-step approach to predicting the efficacy of CAR-modified cells using ML-based CAR IS quality quantification. This approach will guide the users on how to use ML-based CAR IS quality quantification in detail, which include: how to image CAR IS on the glass-supported planar lipid bilayer, how to define the CAR IS focal plane, how to segment the CAR IS images, and how to quantify the IS quality using ML-based algorithms.This approach will significantly enhance the accuracy and proficiency of CAR IS prediction in research.

Standardized protocol for the evaluation of chimeric antigen receptor (CAR)-modified cell immunological synapse quality using the glass-supported planar lipid bilayer.

Chimeric antigen receptor (CAR)-modified cell therapy is an effective therapy that harnesses the power of the human immune system by re-engineering immune cells that specifically kill tumor cells with tumor antigen specificity. Key to the effective elimination of tumor cells is the establishment of the immunological synapse (IS) between CAR-modified immune cells and their susceptible tumors. For functional activity, CAR-modified cells must form an effective IS to kill tumor cells specifically. The formation of the CAR-specific IS requires the coordination of many cellular processes including reorganization of the cytoskeletal structure, polarization of lytic granules, accumulation of tumor antigen, and phosphorylation of key signaling molecules within the IS. Visualization and assessment of the CAR IS quality can reveal much about the molecular mechanisms that underlie the efficacy of various CAR-modified immune cells. This chapter provides a standardized method of assessing the IS quality by quantifying the tumor antigen (defining the CAR IS formation), cytoskeleton (key component of CAR IS structure), and various molecules of interest involved in the IS formation (key molecular mechanism signatures of CAR IS function) using immunofluorescence on the glass-supported planar lipid bilayer, with a focus on tumor antigen only in this study. We provide specific insights and helpful tips for reagent and sample preparation, assay design, and machine learning (ML)-based data analysis. The protocol described in this chapter will provide a valuable tool to visualize and assess the IS quality of various CAR-modified immune cells.

Development of a novel human CD147 knock-in NSG mouse model to test SARS-CoV-2 viral infection.

BACKGROUND: An animal model that can mimic the SARS-CoV-2 infection in humans is critical to understanding the rapidly evolving SARS-CoV-2 virus and for development of prophylactic and therapeutic strategies to combat emerging mutants. Studies show that the spike proteins of SARS-CoV and SARS-CoV-2 bind to human angiotensin-converting enzyme 2 (hACE2, a well-recognized, functional receptor for SARS-CoV and SARS-CoV-2) to mediate viral entry. Several hACE2 transgenic (hACE2Tg) mouse models are being widely used, which are clearly invaluable. However, the hACE2Tg mouse model cannot fully explain: (1) low expression of ACE2 observed in human lung and heart, but lung or heart failure occurs frequently in severe COVID-19 patients; (2) low expression of ACE2 on immune cells, but lymphocytopenia occurs frequently in COVID-19 patients; and (3) hACE2Tg mice do not mimic the natural course of SARS-CoV-2 infection in humans. Moreover, one of most outstanding features of coronavirus infection is the diversity of receptor usage, which includes the newly proposed human CD147 (hCD147) as a possible co-receptor for SARS-CoV-2 entry. It is still debatable whether CD147 can serve as a functional receptor for SARS-CoV-2 infection or entry. RESULTS: Here we successfully generated a hCD147 knock-in mouse model (hCD147KI) in the NOD-scid IL2Rgammanull (NSG) background. In this hCD147KI-NSG mouse model, the hCD147 genetic sequence was placed downstream of the endogenous mouse promoter for mouse CD147 (mCD147), which creates an in vivo model that may better recapitulate physiological expression of hCD147 proteins at the molecular level compared to the existing and well-studied K18-hACE2-B6 (JAX) model. In addition, the hCD147KI-NSG mouse model allows further study of SARS-CoV-2 in the immunodeficiency condition which may assist our understanding of this virus in the context of high-risk populations in immunosuppressed states. Our data show (1) the human CD147 protein is expressed in various organs (including bronchiolar epithelial cells) in hCD147KI-NSG mice by immunohistochemical staining and flow cytometry; (2) hCD147KI-NSG mice are marginally sensitive to SARS-CoV-2 infection compared to WT-NSG littermates characterized by increased viral copies by qRT-PCR and moderate body weight decline compared to baseline; (3) a significant increase in leukocytes in the lungs of hCD147KI-NSG mice, compared to infected WT-NSG mice. CONCLUSIONS: hCD147KI-NSG mice are more sensitive to COVID-19 infection compared to WT-NSG mice. The hCD147KI-NSG mouse model can serve as an additional animal model for further interrogation whether CD147 serve as an independent functional receptor or accessory receptor for SARS-CoV-2 entry and immune responses.

Development of a Novel Human CD147 Knock-in NSG Mouse Model to Test SARS-CoV-2 Viral Infection.

Background: An animal model that can mimic the SARS-CoV-2 infection in humans is critical to understanding the rapidly evolving SARS-CoV-2 virus and for development of prophylactic and therapeutic strategies to combat emerging mutants. Studies show that the spike proteins of SARS-CoV and SARS-CoV-2 bind to human angiotensin-converting enzyme 2 (hACE2, a well-recognized, functional receptor for SARS-CoV and SARS-CoV-2) to mediate viral entry. Several hACE2 transgenic (hACE2Tg) mouse models are being widely used, which are clearly invaluable. However, the hACE2Tg mouse model cannot fully explain: 1) low expression of ACE2 observed in human lung and heart, but lung or heart failure occurs frequently in severe COVID-19 patients; 2) low expression of ACE2 on immune cells, but lymphocytopenia occurs frequently in COVID-19 patients; and 3) hACE2Tg mice do not mimic the natural course of SARS-CoV-2 infection in humans. Moreover, one of most outstanding features of coronavirus infection is the diversity of receptor usage, which includes the newly proposed human CD147 (hCD147) as a possible co-receptor for SARS-CoV-2 entry. It is still debatable whether CD147 can serve as a functional receptor for SARS-CoV-2 infection or entry. Results: Here we successfully generated a hCD147 knock-in mouse model (hCD147KI) in the NOD- scid IL2Rgamma null (NSG) background. In this hCD147KI-NSG mouse model, the hCD147 genetic sequence was placed downstream of the endogenous mouse promoter for mouse CD147 (mCD147), which creates an in vivo model that may better recapitulate physiological expression of hCD147 proteins at the molecular level compared to the existing and well-studied K18-hACE2-B6 (JAX) model. In addition, the hCD147KI-NSG mouse model allows further study of SARS-CoV-2 in the immunodeficiency condition which may assist our understanding of this virus in the context of high-risk populations in immunosuppressed states. Our data show 1) the human CD147 protein is expressed in various organs (including bronchiolar epithelial cells) in hCD147KI-NSG mice by immunohistochemical staining and flow cytometry; 2) hCD147KI-NSG mice are marginally sensitive to SARS-CoV-2 infection compared to WT-NSG littermates characterized by increased viral copies by qRT-PCR and moderate body weight decline compared to baseline; 3) a significant increase in leukocytes in the lungs of hCD147KI-NSG mice, compared to infected WT-NSG mice. Conclusions: hCD147KI-NSG mice are more sensitive to COVID-19 infection compared to WT-NSG mice. The hCD147KI-NSG mouse model can serve as an additional animal model for further interrogation whether CD147 serve as an independent functional receptor or accessory receptor for SARS-CoV-2 entry and immune responses.

In vitro machine learning-based CAR T immunological synapse quality measurements correlate with patient clinical outcomes.

The human immune system consists of a highly intelligent network of billions of independent, self-organized cells that interact with each other. Machine learning (ML) is an artificial intelligence (AI) tool that automatically processes huge amounts of image data. Immunotherapies have revolutionized the treatment of blood cancer. Specifically, one such therapy involves engineering immune cells to express chimeric antigen receptors (CAR), which combine tumor antigen specificity with immune cell activation in a single receptor. To improve their efficacy and expand their applicability to solid tumors, scientists optimize different CARs with different modifications. However, predicting and ranking the efficacy of different "off-the-shelf" immune products (e.g., CAR or Bispecific T-cell Engager [BiTE]) and selection of clinical responders are challenging in clinical practice. Meanwhile, identifying the optimal CAR construct for a researcher to further develop a potential clinical application is limited by the current, time-consuming, costly, and labor-intensive conventional tools used to evaluate efficacy. Particularly, more than 30 years of immunological synapse (IS) research data demonstrate that T cell efficacy is not only controlled by the specificity and avidity of the tumor antigen and T cell interaction, but also it depends on a collective process, involving multiple adhesion and regulatory molecules, as well as tumor microenvironment, spatially and temporally organized at the IS formed by cytotoxic T lymphocytes (CTL) and natural killer (NK) cells. The optimal function of cytotoxic lymphocytes (including CTL and NK) depends on IS quality. Recognizing the inadequacy of conventional tools and the importance of IS in immune cell functions, we investigate a new strategy for assessing CAR-T efficacy by quantifying CAR IS quality using the glass-support planar lipid bilayer system combined with ML-based data analysis. Previous studies in our group show that CAR-T IS quality correlates with antitumor activities in vitro and in vivo. However, current manually quantified IS quality data analysis is time-consuming and labor-intensive with low accuracy, reproducibility, and repeatability. In this study, we develop a novel ML-based method to quantify thousands of CAR cell IS images with enhanced accuracy and speed. Specifically, we used artificial neural networks (ANN) to incorporate object detection into segmentation. The proposed ANN model extracts the most useful information to differentiate different IS datasets. The network output is flexible and produces bounding boxes, instance segmentation, contour outlines (borders), intensities of the borders, and segmentations without borders. Based on requirements, one or a combination of this information is used in statistical analysis. The ML-based automated algorithm quantified CAR-T IS data correlates with the clinical responder and non-responder treated with Kappa-CAR-T cells directly from patients. The results suggest that CAR cell IS quality can be used as a potential composite biomarker and correlates with antitumor activities in patients, which is sufficiently discriminative to further test the CAR IS quality as a clinical biomarker to predict response to CAR immunotherapy in cancer. For translational research, the method developed here can also provide guidelines for designing and optimizing numerous CAR constructs for potential clinical development. Trial Registration: ClinicalTrials.gov NCT00881920.

Deletion of ER-retention motif on SARS-CoV-2 spike protein reduces cell hybrid during cell-cell fusion.

BACKGROUND: The novel SARS-CoV-2 has quickly become a global pandemic since the first reported case in December 2019, with the virus infecting millions of people to date. The spike (S) protein of the SARS-CoV-2 virus plays a key role in binding to angiotensin-converting enzyme 2 (ACE2), a host cell receptor for SARS-CoV-2. S proteins that are expressed on the cell membrane can initiate receptor-dependent syncytia formation that is associated with extensive tissue damage. Formation of syncytia have been previously observed in cells infected with various other viruses (e.g., HIV, Ebola, Influenza, and Herpesviruses). However, this phenomenon is not well documented and the mechanisms regulating the formation of the syncytia by SARS-CoV-2 are not fully understood. RESULTS: In this study, we investigated the possibility that cell fusion events mediated by the S protein of SARS-CoV-2 and ACE2 interaction can occur in different human cell lines that mimic different tissue origins. These cell lines were transduced with either wild-type (WT-S) S protein or a mutated variant where the ER-retention motif was removed (Δ19-S), as well as human ACE2 expression vectors. Different co-culture combinations of spike-expressing 293T, A549, K562, and SK-Hep1 cells with hACE2-expressing cells revealed cell hybrid fusion. However, only certain cells expressing S protein can form syncytial structures as this phenomenon cannot be observed in all co-culture combinations. Thus, SARS-CoV-2 mediated cell-cell fusion represents a cell type-dependent process which might rely on a different set of parameters. Recently, the Δ19-S variant is being widely used to increase SARS-CoV-2 pseudovirus production for in vitro assays. Comparison of cell fusion occurring via Δ19-S expressing cells shows defective nuclear fusion and syncytia formation compared to WT-S. CONCLUSIONS: This distinction between the Δ19-S variant and WT-S protein may have downstream implications for studies that utilize pseudovirus-based entry assays. Additionally, this study suggest that spike protein expressed by vaccines may affect different ACE2-expressing host cells after SARS-CoV-2 vaccine administration. The long-term effects of these vaccines should be monitored carefully. Δ19-S mRNA may represent a safer mRNA vaccine design in the future.

Deletion of ER-retention Motif on SARS-CoV-2 Spike Protein Reduces Cell Hybrid During Cell-cell Fusion.

The novel SARS-CoV-2 has quickly become a global pandemic since the first reported case in December 2019, with the virus infecting millions of people to date. The spike (S) protein of the SARS-CoV-2 virus plays a key role in binding to angiotensin-converting enzyme 2 (ACE2), a host cell receptor for SARS-CoV-2. S proteins that are expressed on the cell membrane can initiate receptor-dependent syncytia formation that is associated with extensive tissue damage. Formation of syncytia have been previously observed in cells infected with various other viruses (e.g., HIV, Ebola, Influenza, and Herpesviruses). However, this phenomenon is not well documented and the mechanisms regulating the formation of these syncytia by SARS-CoV-2 are not fully understood. In this study, we investigated the possibility that cell fusion events mediated by the S protein of SARS-CoV-2 and ACE2 interaction can occur in different human cell lines that mimic different tissue origins. These cell lines were stably transduced with either wild-type (WT-S) S protein or a mutated variant where the ER-retention motif was removed (Δ19-S), or human ACE2 vectors. Different co-culture combinations of spike-expressing 293T, A549, K562, and SK-Hep1 cells with hACE2-expressing cells revealed cell hybrid fusion. However, only certain cells expressing S protein can form syncytial structures as this phenomenon cannot be observed in all co-culture combinations. Thus, SARS-CoV-2 mediated cell-cell fusion represents a cell type-dependent process which might rely on a different set of parameters. Recently, the Δ19-S variant is being widely used to increase SARS-CoV-2 pseudovirus production for in vitro assays. Comparison of cell fusion occurring via Δ19-S expressing cells shows defective nuclear fusion and syncytia formation compared to WT-S. This distinction between the Δ19-S variant and WT-S protein may have downstream implications for studies that utilize pseudovirus-based entry assays. Additionally, this study suggest that spike protein expressed by vaccines may affect different ACE2-expressing host cells after SARS-CoV-2 vaccine administration. The long-term effects of these vaccines should be monitored carefully.

Greedy auto-augmentation for n-shot learning using deep neural networks.

The goal of n-shot learning is the classification of input data from small datasets. This type of learning is challenging in neural networks, which typically need a high number of data during the training process. Recent advancements in data augmentation allow us to produce an infinite number of target conditions from the primary condition. This process includes two main steps for finding the best augmentations and training the data with the new augmentation techniques. Optimizing these two steps for n-shot learning is still an open problem. In this paper, we propose a new method for auto-augmentation to address both of these problems. The proposed method can potentially extract many possible types of information from a small number of available data points in n-shot learning. The results of our experiments on five prominent n-shot learning datasets show the effectiveness of the proposed method.