Publications

Mechanistic studies using TIMING based single-cell reporter assays demonstrated that redundancy is important for efficacious T cells. Leukemic targets are sensitive to multiple granzymes whereas solid tumors are sensitive to granzyme and Fas ligand.

Hydroporation is an alternative way to deliver the CRISPR/Cas9 gene-editing machinery into the T cells. Hydroporation uses a gentle fluid force to permeabilize the cell membrane, allowing the gene-editing components to enter the T cells. Using a Time-lapse Imaging Microscopy in Nanowell Grids (TIMING) assay, scientists compared CAR-T cells from five donors after processing them with either hydroporation or nucleofection.

To improve the detection of apoptosis, researchers directly detected apoptotic bodies in a label-free manner. The trained network identified nanowells containing apoptotic bodies with 92% accuracy and predicted the onset of apoptosis with an error of one frame (5 min/frame). The apoptotic body segmentation yielded an IoU accuracy of 75%, allowing associative identification of apoptotic cells.

Immune cells such as T cells and natural killer cells, and target cells such as NALM6, K562 and EL4, were incubated in PDMS nanowell arrays and imaged using time-lapse fluorescent microscopy. The proposed cell segmentation and tracking algorithms allowed automated quantification of cell pairs, cell location, morphology, interactions and movement as well as cell viability, without the need for manual processing. The result revealed that cytotoxic T cells have higher motility than noncytotoxic T cells, both before and during synapse formation.

We found sustained activation of cytotoxicity, costimulation, oxidative phosphorylation– and proliferation-related genes, and simultaneously reduced differentiation and exhaustion. Our study identifies molecular features of TCR8 expression that can guide the development of enhanced immunotherapies.

CD19/20/22CAR T-cells killed CD19(−) blasts from patients who relapsed after CD19CAR T-cell therapy and CRISPR/Cas9 CD19 knockout primary BL-ALL both in vitro and in an animal model, while CD19CAR T-cells were ineffective. At the subcellular level, CD19/20/22CAR T-cells formed dense immune synapses with target cells that mediated effective cytolytic complex formation, were efficient serial killers in single-cell tracking studies, and were as efficacious as CD19CAR T-cells against primary CD19(+) disease. In conclusion, independent of CD19 expression, CD19/20/22CAR T-cells could be used as salvage or front-line CAR therapy for patients with recalcitrant disease.

By utilizing both traditional blob detection to generate binary mask labels from the stained channel images and the deep learning Mask RCNN model to train a detection and segmentation model, we managed to segment nuclei based only on phase images. The detection average precision is 0.82 when the IoU threshold is to be set 0.5. And the mean IoU for masks generated from phase images and ground truth masks from experts is 0.735. Without any ground truth mask labels during the training time, this is good enough to prove our hypothesis. This result enables the ability to detect nuclei without the need for exogenous labeling.

This paper proposes an efficient variant of capsule networks (CapsNets) as an alternative to CNNs. Extensive experimental results demonstrate that the proposed CapsNets achieve competitive performances in target cell apoptosis classification, while significantly outperforming CNNs when the number of training samples is small. To utilize temporal information within microscopy videos, we propose a recurrent CapsNet constructed by stacking a CapsNet and a bi-directional long short-term recurrent structure. Our experiments show that when considering temporal constraints, the recurrent CapsNet achieves 93.8% accuracy and makes significantly more consistent prediction than NNs.

Integration of transcriptomic profiling, immune phenotyping and metabolism demonstrated that motile cells are more naïve-like with higher oxidative metabolism and spare respiratory capacity. Our result also revealed that the master metabolic regulator AMP kinase (AMPK) is required for CAR+ T cells with high motility. We used a xenograft leukemia mouse model (CD19+ NALM-6) and validated that the motile cells have enhanced persistence and superior anti-cancer effect in vivo compared to the parental un-sorted population. Collectively, our multi-dimensional results demonstrated that persistent motility is a selectable biomarker of expanded CAR+ T cell bioactivity.

On a desktop computer, TIMING 2.0 takes 5 s/block/image frame, four times faster than our previous method on the same computer, and twice as fast as our previous method (TIMING) running on a Dell PowerEdge server. The cell segmentation accuracy (f-number = 0.993) is superior to our previous method (f-number = 0.821). A graphical user interface provides the ability to inspect the video analysis results, make corrective edits efficiently (one-click editing of an entire nanowell video sequence in 5–10 s) and display a summary of the cell killing efficacy measurements.

Genetically engineered T cells that express chimeric antigen receptors (CAR+) are heterogeneous and thus, understanding the immunotherapeutic efficacy remains a challenge in adoptive cell therapy. We developed a high-throughput single-cell methodology, Timelapse Imaging Microscopy In Nanowell Grids (TIMING) to monitor interactions between immune cells and tumor cells in vitro. Using TIMING we demonstrated that CD4+ CAR+ T cells participate in multi-killing and benefit from improved resistance to activation induced cell death in comparison to CD8+ CAR+ T cells. For both subsets of cells, effector cell fate at the single-cell level was dependent on functional activation through multiple tumor cells.

A comprehensive understanding of the polyfunctionality of T lymphocytes in ICI or adoptive cell transfer (ACT), at single-cell resolution, will quantify T-cell properties that are essential for therapeutic benefit. We briefly highlight several emerging integrated single-cell technologies focusing on the profiling of multiple properties/functionalities of T cells. We envision that these tools have the potential to provide valuable experimental and clinical insights on T-cell biology, and eventually pave the road for the discovery of surrogate T-cell biomarkers for immunotherapy.

CD8+BTLA- TILs could not control tumor growth in vivo as well as their BTLA+ counterpart and antigen-specific CD8+BTLA- T cells had impaired recall response to a vaccine. However, CD8+BTLA+ TILs displayed improved survival following the killing of a tumor target and heightened "serial killing" capacity. Using mutants of BTLA signaling motifs, we uncovered a costimulatory function mediated by Grb2 through enhancing the secretion of IL-2 and the activation of Src after TCR stimulation. Our data portrays BTLA as a molecule with the singular ability to provide both costimulatory and coinhibitory signals to activated CD8+ T cells, resulting in extended survival, improved tumor control, and the development of a functional recall response. Clin Cancer Res; 23(20); 6151-64. ©2017 AACR.

We used the engineered Fc domains to demonstrate in vitro and in mouse models that for therapeutic antibodies, complement-dependent cell-mediated cytotoxicity (CDCC) and complement-dependent cell-mediated phagocytosis (CDCP) by immunological effector molecules mediated the clearance of target cells with kinetics and efficacy comparable to those of the FcγR-dependent effector functions that are much better studied, while they circumvented certain adverse reactions associated with FcγR engagement. Collectively, our data highlight the importance of CDCC and CDCP in monoclonal-antibody function and provide an experimental approach for delineating the effect of complement-dependent effector-cell engagement in various therapeutic settings.

Analysis of hundreds of individual human peripheral blood NK cells profiled ex vivo revealed that CD56dimCD16+ NK cells are immediate secretors of interferon gamma (IFN-γ) upon activation by phorbol 12-myristate 13-acetate (PMA) and ionomycin (< 3 h), and that there was no evidence of cooperation between NK cells leading to either synergistic activation or faster IFN-γ secretion. Furthermore, we observed that both the amount and rate of IFN-γ secretion from individual NK cells were donor-dependent. Collectively, these results establish our methodology as an investigational tool for combining phenotyping and real-time protein secretion of individual cells in a high-throughput manner.

In aggregate, these results demonstrate the utility of our TIMING single cell methodology in uncovering not only the dynamic profile of T-cell behavior but also the ability to identify subpopulations of T-cell with enhanced polyfunctionality. Our studies support the use of motility as a surrogate and selective marker of higher CAR+ T cell bioactivity. These results also open up avenues to molecularly engineer T cells for an increased motility that could translate to better in vivo outcomes.

Fluorescently labeled human T cells, natural killer cells (NK), and various target cells (NALM6, K562, EL4) were co-incubated on polydimethylsiloxane arrays of sub-nanoliter wells (nanowells), and imaged using multi-channel time-lapse microscopy. The proposed cell segmentation and tracking algorithms account for cell variability and exploit the nanowell confinement property to increase the yield of correctly analyzed nanowells from 45% (existing algorithms) to 98% for wells containing one effector and a single target, enabling automated quantification of cell locations, morphologies, movements, interactions, and deaths without the need for manual proofreading. Automated analysis of recordings from 12 different experiments demonstrated automated nanowell delineation accuracy >99%, automated cell segmentation accuracy >95%, and automated cell tracking accuracy of 90%, with default parameters, despite variations in illumination, staining, imaging noise, cell morphology, and cell clustering. An example analysis revealed that NK cells efficiently discriminate between live and dead targets by altering the duration of conjugation. The data also demonstrated that cytotoxic cells display higher motility than non-killers, both before and during contact.