Methodology for sorting, expansion, and transcriptomics analysis of killer t-cells

Abstract

Methods are provided for activating killer T cells, sorting and collecting the activated killer T cells, expanding the activated or non-activated T cells, and analyzing gene transcription for individual cells or groups of cells. The methods utilize a first microfluidic device for co-encapsulating killer T cells and target cells, such as cancer cells, and docking them for time-lapse light microscopic analysis, and a second microfluidic device for sorting the cells based on their fluorescence. The methodology is useful for analyzing immune function at the single cell level and correlating cell behaviors and interactions with transcriptomics. It also can be used to identify, isolate, collect, and expand highly active killer T cells from an individual subject for use in immunotherapy.

Description

[0001] TITLE

[0002] Methodology for Sorting, Expansion, and Transcriptomics Analysis of Killer T-Cells

[0003] CROSS REFERENCE TO RELATED APPLICATIONS

[0004] This application claims the priority of U.S. Provisional Application Nos. 63 / 693,467 filed 11 September 2024 and entitled “Method and System for Analysis of immune Cell Interactions with Target Cells Using Fluorescence Activated Droplet Sorting”; 63 / 722,331 filed 19 November 2024 and entitled “Methodology for Sorting and Expansion of Killer T-Cells”; 63 / 722,336 filed 19 November 2024 and entitled “Transcriptomics of Activated T Cells and Uses Thereof”; 63 / 723,210 filed 21 November 2024 and entitled “Activation and Preparation of Natural Killer Cells and Killer T Cells Using Microfluidics and Multi-omic Profiling”. Each of the aforementioned applications is hereby incorporated by reference in its entirety.

[0005] STATEMENT REGARDING FEDERALLY SPONSORED RESEARCH OR DEVELOPMENT

[0006] This invention was made with government support under Grant No. 2310303 awarded by the National Science Foundation. The government has certain rights in the invention.

[0007] BACKGROUND

[0008] Bispecific antibodies (bsAbs) are a class of immunotherapeutic agents crafted to enlist and guide the immune system in the precise elimination of cancer cells [1-4], These bsAbs bring cytotoxic immune cells and tumor cells into close proximity, activating the T-cells to kill the malignant cells. T-cell activating bsAbs represent the largest subclass of bsAbs and account for over 50% of the bsAb preclinical and clinical pipeline, including clinically approved blinatumomab [5], Although the introduction of CD3xCD19 bsAbs such as blinatumomab have significantly improved the prognosis of cancers such as B-cell non-Hodgkin lymphoma (NHL), there remains a significant number of patients who do not benefit after receiving this class of therapy [6], Hence, the need arises for improved treatment strategies and approaches to identify patients most likely to respond to a particular bsAb therapeutic regimen. Multiple factors can be involved in the mechanisms of cancer treatment resistance, including loss of target tumor antigens, immunosuppressive microenvironment, and poor T- cell responses [7], Given the reliance bsAbs have on immune cells, the functionality of cytotoxic T-cells particularly as it pertains to cell exhaustion has been well-studied and considered to be an important resistance mechanism hampering T-cell fitness and hence longterm clinical activity of CD3xCD19 bsAbs [8], Gradual loss of T-cell function, including proliferation, killing capacity and cytokine secretion have been described and attempts are being made to interfere with T-cell exhaustion by targeting costimulatory and inhibitory pathways to improve bsAb antitumor activity [9], Dynamic cell-cell interactions and the effect of environmental factors further increase the level of complexity of these interactions and contribute to the diversity in T-cell function

[0010] , The fact remains that these CD3xCD19 bsAbs rely on the functional killing capacity of individual T-cells which are heterogeneous by default and often dysfunctional in the tumor microenvironment

[0011] ,

[0009] Currently, however, there is a lack of technologies that can characterize the anti-tumor activity of individual T-cells and discriminate between functional versus less functional (lazy) T-cells at the single-cell level. The effectiveness of CD3xCD19 bsAbs is being evaluated by conventional techniques such as flow cytometry and cytotoxicity co-culture assays [12, 13], These studies have shown that the extent of tumor cell lysis and the effective concentration obtained from healthy donor T-cells varies considerably using these techniques, with tumor cell lysis ranging from 25 to 95% [14, 15], Recently, studies using more sensitive techniques such as single-cell sequencing analysis of T-cells has started to unravel some of this complex biology, revealing unique transcriptional signatures associated with CD3xCD19 bsAb mediated tumor cell killing

[0016] , However, these methods are based on averaging the effects of immune cells with varying functional capacity to target and kill tumor cells and thus not capable of tracking and correlating the outcome of dynamic individual effector-target (E-T) cell interactions. These approaches also do not permit detection of key interactive features of single-cell activity, such as repeated tumor-cell interactions, kinetics of cytotoxic responses, as well as the genomic features driving T-cell killing activity. Importantly, this necessitates the implementation of in-depth characterization aimed to discover the key factors accounting for the heterogeneous response of CD3xCD19 bsAb activated cells.

[0010] SUMMARY

[0011] The present technology provides methods for activating killer T cells, sorting and collecting the activated killer T cells and / or those T cells that do not become activated, expanding the activated or non-activated T cells, and analyzing gene transcription for individual cells or groups of cells. The methods utilize droplet microfluidics and are carried out in part using a microfluidic device for co-encapsulating killer T cells and target cells, such as cancer cells and docking them for time-lapse light microscopic analysis, followed by sorting the cells based on their function as reflected in their fluorescence in a droplet sorting device. The sorting device can be optionally combined with the droplet forming and docking device. The technology is especially useful for analyzing immune function at the single cell level and correlating cell behaviors and interactions with transcriptomics, but also can be used to identify, isolate, collect, and expand highly active killer T cells from an individual subject for use in immunotherapy.

[0012] The technology can be summarized in the following listing of features.

[0013] 1 . A method for enriching a population of killer T cells with active killer T cells, the method comprising:

[0014] (a) providing

[0015] (i) a microfluidic system configured for co-encapsulating single cells in microdroplets in an oil stream and sorting the microdroplets based on fluorescence within the microdroplets;

[0016] (ii) a population of single killer T cells, optionally loaded with an intracellular Ca2+-sensitive fluorescent indicator;

[0017] (iii) a population of single target cells, such as cancer cells or tumor-derived cells, optionally loaded with a fluorescent indicator of caspase activity, wherein either the population of T cells in (ii) or the population of target cells in (iii) contains said fluorescent indicator, but not both; and

[0018] (iv) a bispecific or multispecific binding molecule having a T cell binding specificity for an antigen on a surface of the killer T cells and a target cell binding specificity for an antigen on a surface of the target cells;

[0019] (b) co-encapsulating the killer T cells, the target cells, and the bispecific or multispecific binding molecule in microdroplets in an oil stream using the microfluidic system;

[0020] (c) allowing the killer T cells and target cells to bind the bispecific or multispecific binding molecule and to form killer T cell-binding molecule-target cell complexes in the microdroplets, whereby some of the killer T cells become activated by interaction with the target cells and the activated killer T cells kill associated target cells, and whereby a fluorescence signal is produced either from the activated killer T cells or the associated target cells;

[0021] (d) sorting the microdroplets based on a level of said fluorescence signal, yielding a first group of microdroplets containing activated killer T cells and a second group of microdroplets containing non-activated killer T cells;

[0022] (e) collecting the activated killer T cells from the first group, thereby producing a population of killer T cells that is enriched in active killer cells compared to the initially provided killer T cells; and (f) optionally collecting the non-activated T cells from the second group.

[0023] 2. The method of feature 1 , wherein target cells co-encapsulated with killer T cells are killed, and the population of enriched activated killer T cells produced in (e) is essentially free of living or intact target cells.

[0024] 3. The method of feature 1 or 2, wherein the bispecific or multispecific binding molecule is a bispecific or multispecific antibody or aptamer.

[0025] 4. The method of any of the preceding features, wherein the T cell binding specificity comprises binding to CD3, or wherein the bispecific or multispecific binding molecule is a multispecific binding molecule, and wherein the multispecific binding molecule has T cell binding specificities for CD3 and CD8, or for CD3 and CD4, or for CD3 and a marker for Tregs selected from the group consisting of Foxp3, CD25, GITR, Nrp1, Helios, and CTLA-4.

[0026] 5. The method of any of the preceding features, wherein the target cell binding specificity is for binding to a tumor cell surface antigen, such as a tumor cell surface antigen selected from the group consisting of CD19, CD20, CD44, CTLA-4, GITR-A, OX-40, CLDN1, LY6G6D / F, TLR4, GPR56, and SLCO183.

[0027] 6. The method of any of the preceding features, wherein the fluorescence signal is produced using an intracellular Ca2+-sensitive fluorescent indicator in the killer T cells, or wherein the fluorescence signal is produced using a fluorescent indicator of caspase 3 / 7 activity in the target cells.

[0028] 7. The method of any of the preceding features, wherein the killer T cells and target cells are co-encapsulated in the microdroplets at a ratio of about 1 :1 killer cells to target cells, or wherein the killer T cells and target cells are co-encapsulated in the microdroplets at a ratio of less than 1:1 killer cells to target cells, such as wherein the ratio of killer cells to target cells is from about 0.1 to about 0.9, and wherein the activated killer cells collected in (e) are mixed with target cells that survived.

[0029] 8. The method of any of the preceding features wherein, prior to said transcriptomic analysis, step (g) comprises expanding the enriched population of active killer T cells by performing one or more rounds of proliferation of the enriched population of active killer T cells in culture.

[0030] 9. The method of any of the preceding features, further comprising

[0031] (g) performing transcriptomic analysis of the one or more of the activated killer T cells collected in (e) and / or the nonactivated killer T cells collected in (f); wherein said transcriptomic analysis comprises

[0032] (i) isolating mRNA from a pool of the enriched population of active killer T cells;

[0033] (ii) performing RNA sequencing on the isolated mRNA; and

[0034] (iii) analyzing the sequenced RNA from (b) to determine a level of expression of one or more genes of the sorted and enriched population of active killer T cells. 10. The method of feature 9, wherein step (iii) comprises determining whether one or more biological processes or metabolic pathways are activated or inhibited in the collected active killer T cells compared to nonactivated killer T cells or nonkiller T cells.

[0035] 11. The method of any of the preceding features, further comprising genetically modifying one or more of the collected active killer T cells.

[0036] 12. The method of feature 12, wherein the genetic modification comprises transducing the one or more cells with a viral vector, whereby the genetically modified cells’ expression of one or more genes is increased or decreased.

[0037] 13. The method of feature 11 or 12, wherein the genetic modification is performed prior to expansion of the genetically modified cells.

[0038] 14. A population of active killer T cells obtained by the method of any of the preceding features.

[0039] 15. The population of active killer T cells of feature 14, wherein the population comprises cells having one or more genes or pathways upregulated or downregulated compared to nonkiller T cells.

[0040] 16. A method of immunotherapy of a subject in need of killer T cell supplementation, the method comprising:

[0041] (a) obtaining a population of enriched and expanded active killer T cells using the method of any of the preceding features;

[0042] (b) administering at least a portion of the population of enriched and expanded active killer T cells to the subject.

[0043] 17. The method of feature 16, wherein (a) comprises: obtaining a population of killer T cells from the subject, and using the population of the subject’s killer T cells as the initial population of killer T cells in the method of expansion of active killer T cells.

[0044] 18. The method of feature 16 or 17, wherein (a) comprises: obtaining a population of single target cells from the subject, and using the population of the subject’s target cells as the target cells in the method of enriching and expanding killer T cells.

[0045] 19. The method of any of features 16-18, wherein the population of isolated, enriched, and / or active killer T cells is accepted or rejected for administration to the subject in step (b) based on results of the transcriptomic analysis.

[0046] 20. The method of any of features 16-18 wherein transcriptomic analysis of the one or more of the activated killer T cells collected in (e) and / or the nonactivated killer T cells collected in (f) is performed, and wherein, based on results of the transcriptomic analysis, one or more cells of the population of isolated, enriched, and / or active killer T cells is genetically modified to alter their gene expression prior to administration to the subject in step (b) of the genetically modified cells.

[0047] 21. The method of any of features 16-20, wherein the subject has cancer, and the method is used for cancer therapy of the subject.

[0048] 22. A kit comprising a microfluidic device and instructions for performing the method of any of the preceding features.

[0049] 23. The kit of feature 22 further comprising one or more reagents, such as a bispecific or multispecific antibody or aptamer having at least a first binding specificity for a T cell and a second binding specificity for a target cell, a reagent for detecting caspase 3 / 7 activity by fluorescence, a reagent for detecting changes in intracellular Ca2+concentration by fluorescence, one or more reagents for killer T cell expansion such as IL15 and / or IL7, one or more reagents for transcriptomics analysis, a viral vector for transduction of active killer T cells, or a reagent for specifically detecting a killer T cell biomarker.

[0050] BRIEF DESCRIPTION OF DRAWINGS

[0051] Fig. 1A is a schematic illustration of a single-cell droplet microfluidic device for characterizing cell-cell interactions and T cell killing activity. Fig.1 B is a schematic illustration of a droplet microfluidic workflow for studying CD8+ T cells and Raji cell interactions at the single cell level. CD8+ T cells with or without CD3xCD19 bispecific antibody were added to the droplet microfluidic device at an inlet separate from the Raji cells, which were labeled with calcein AM. Separate droplet microfluidic devices were used for CD3xCD19 bispecific Ab treated and untreated CD8+ T cells. Droplets were docked in a droplet docking array, and preselected droplets were imaged automatically every 30 minutes for a total of 48 h. The cytotoxicity and contact dynamics of CD3xCD19 bispecific activated CD8+ T cells were compared to untreated cell pairs.

[0052] Figs. 2A-2E show results from single-cell cytotoxicity and cell contact dynamics using a CD3xCD19 bispecific antibody. Fig. 2A. Cytotoxicity kinetics of CD3xCD19 bispecific activated CD8+ T-cell and tumor cell co-encapsulations compared to untreated cells for a total of 4 donors over 48h. Error bars represent standard deviation between different donors. n=93 treated cells, n=106 untreated cells. Fig. 2B. Cytotoxicity of CD3xCD19 bispecific activated CD8+ T-cell and tumor cell co-encapsulations compared to untreated cells at every 6h for a total of 48h. Error bars represent standard deviation between donors at each selected timepoint. Fig. 2C. Time of death for all tumor cells in each condition for all donors. Fig. 2D. Number of synaptic contacts and contact duration (minutes) for killed tumor cells. Time it takes a CD8+ T-cell to establish an initial contact with a tumor cell is shown, along with the number of contacts between CD8+ T-cell and tumor cells, proportion of CD8+ T- cells in contact with tumor cells and duration of contact. n=77 treated cells, n=79 untreated cells. Fig. 2E. Number of synaptic contacts and contact duration for tumor cells that stayed alive (n=16 treated cells, n=27 untreated cells). Time it takes a CD8+ T-cell to establish an initial contact with a tumor cell is shown, along with the number of contacts between CD8+ T-cell and tumor cells, proportion of CD8+ T-cells in contact with tumor cells and duration of contact. Statistical analysis for cytotoxicity kinetic profiles based on unpaired t test and time to death profiles and synaptic contacts based on Welch’s t test. Statistical significance was calculated by Mann-Whitney T test; ** p< 0.005, *** p< 0.0005, **** p < 0.00005.

[0053] Figs. 3A-3E show results for differential killing and cell contact dynamics of CD3xCD19 bispecific activated CD8+ T-cell killers. Fig. 3A. Time to death profiles for CD3xCD19 bispecific activated CD8+ T-cell killers for each donor (donor 1 , n= 53; donor2, n= 121 ; donor 3, n= 40, donor 4, n= 99). Fig. 3B. Number of contacts for killer and non-killer cells with tumor cells. Fig. 3C. Proportion of CD8+ T-cells with a specific contact number with tumor cells for killer and non-killer cells. Fig. 3D. The duration of contact (minutes) for non-killer and killer cells with tumor cells. Fig. 3E. Time to first contact with tumor cell for non-killer and killer cells. Statistical significance was calculated by Mann-Whitney T test;*** p< 0.0005, **** p < 0.00005.

[0054] Figs.4A-4D illustrate a droplet microfluidics function-to-omics workflow to study CD3xCD19 bispecific activated T-cell killers and non-killers. Fig. 4A. Primary human CD8+ T- cells are incubated with a CD3xCD19 bispecific antibody and co-encapsulated with fluorescently labeled caspase 3 / 7 tumor cells (Raji) for 18h 37°C at a target effector to target ratio of 1 : 1 . Fig. 4B. Droplets are introduced into the F-FADS device to separate T-cell killers from non-killers (lazy) cells and total RNA is extracted from each sample. Fig. 4C. RNA sequencing is done on each sample. Fig. 4D. Integrated transcriptom ic analysis is done on T- cell killers and non-killer (lazy) cells. Fig. 4E. The transcriptional prolife of each functional cell type is compared to identify genes and biological pathways driving tumor cell killing and genes limiting tumor cell killing in CD3xCD19 bispecific activated T-cells. Fig. 4E shows a schematic representation of a device for performing F-FADS.

[0055] Figs. 5A-5H show cytolytic, immune checkpoint, and activation gene profiles of F- FADS sorted bsAb activated CD8+ T-cells. Fig. 5A. Heatmap demonstrating FPKM values between CD3xCD19 BsAb CD8+ T-cell killers and non-killers for each donor for selected cytolytic genes (red= high, blue = low) Fig. 5B. Mean expression changes for each cytolytic gene based on Iog2 fold change between CD3xCD19 BsAb CD8+ T-cell killers and non-killers for all donors. Fig. 5C. Gene expression for each cytolytic gene based on Iog2 fold change between CD3xCD19 bsAb activated CD8+ T-cell killers and non-killers for all donors. Fig. 5D. Heatmap demonstrating FPKM values between CD3xCD19 BsAb CD8+ T-cell killers and non- killers for each donor for selected immune checkpoint and activation genes. Fig. 5E. Mean expression changes for each immune checkpoint gene based on Iog2 fold change between CD3xCD19 BsAb CD8+ T-cell killers and non-killers for all donors. Fig. 5F. Gene expression for each

[0056] immune checkpoint gene based on Iog2 fold change between CD3xCD19 bsAb activated CD8+ T-cell killers and non-killers for all donors. Fig. 5G. Gene expression for each activation gene based on Iog2 fold change between CD3xCD19 bsAb activated CD8+ T-cell killers and non-killers for all donors. Fig. 5H. Mean expression changes for each activation gene based on Iog2 fold change between CD3xCD19 BsAb CD8+ T-cell killers and non-killers for all donors. Dotted red line represents a Iog2 fold change cutoff of 1. indicates p-adj < 0.05.

[0057] Figs. 6A-6D show immune synapse sequencing profiles of F-FADS sorted CD3xCD19 bispecific activated CD8+ T-cells from different donor samples. Fig. 6A. Immune synapse profile for donor 1. Fig. 6B. Immune synapse profile for donor 2. Fig. 60. Immune synapse profile for donor 3. Fig. 6D. Immune synapse profile for donor 4. Hierarchical clustering analysis of each donor sample represent expression of immune synapse genes comparing CD3xCD19 bsAb activated CD8+ T-cell killers to non-killer cells. Bar plots compare DEGs between CD3xCD19 bsAb activated CD8+ T-cell killers and non-killers, represented as a Iog2 fold change. Pie chart shows the percentage of immune synapse genes upregulated versus downregulated for each donor.

[0058] Figs. 7A-7D show metabolic activity of F-FADS sorted CD3xCD19 bispecific activated CD8+ T-cell killers and non-killers. Fig. 7A. Metabolic activity for donor 1. Fig. 7B. Metabolic activity of donor 2. Fig. 70. Metabolic activity of donor 3. Fig. 7D. Metabolic activity of donor 4. Hierarchical clustering analysis of each donor sample represents expression of metabolic genes involved in OXPHOS, glycolysis, FAO and PPP comparing CD3xCD19 bsAb activated CD8+ T-cell killers to non-killer cells. Bar plots compare differentially expressed genes between CD3xCD19 bsAb activated CD8+ T-cell killers and non-killers, represented as a Iog2 fold change.

[0059] Figs. 8A-8E show bulk RNA-seq GO, KEGG and Reactome analysis of upregulated pathways in CD3xCD19 bsAb activated CD8+ T cell killers compared to CD8+ non-killers for Donor 1 (Fig. 8A), Donor 2 (Fig. 8B), Donor 3 (Fig. 80), and Donor 4 (Fig. 8D). The top 30 pathways listed for each donor with corresponding p-adj and gene count values are shown from each enrichment pathway analysis GO, KEGG and Reactome. Fig. 8E shows the analysis of alternative splicing for each donor.

[0060] Figs. 9A-9D shows transcriptomics results indicating the activation or inhibition of various genes or groups of genes.

[0061] DETAILED DESCRIPTION

[0062] The present technology provides a novel droplet microfluidics-based function-to-omics approach to quantify the activity of activated CD8+ T-cells and their antitumor response at the single-cell level. The data obtained can reveal the specific pathways and cellular switches for for activation of CD8+ killer T-cells and can provide insights into the functional drivers and cellular diversity in immune responses. Understanding this complex diversity in T-cell responses can lead to optimization of immunotherapies for patients with cancer.

[0063] The technology utilizes dynamic time-lapse microscopy at single-cell resolution and functional fluorescence-activated droplet sorting (F-FADS) methodology to isolate activated killer T-cell, which can then be subjected to RNA sequencing to identify the transcriptional drivers associated with effective tumor cell killing.

[0064] Currently, it is challenging to study heterotypic interaction of T-cells and tumor cells using standard microscopic methods since one or both cell types can easily drift apart over time. Furthermore, coculture assays of mixed effector and target cells do not allow any control over interaction of individual cells. For example, one effector cell activated by the CD3xCD19 bsAb may contact and kill a tumor cell, or it may interact with multiple target cells, or not interact at all. This could potentially affect the response of both effector and target cells and alter the outcome of study results. In contrast, droplet microfluidic cell-pairing bioassays provide such as their single-cell droplet platform offers improved control over heterotypic cellcell interaction. Combined with F-FADS and transcriptional profiling of CD3xCD19 bsAb activated T-cell killers, the inventors can assess the dynamics of individual T-cell responses against target cells and differentiate the mechanisms of T cell killers versus non-killers. The present methodology using CD3xCD19 bsAb activated CD8+ T-cells showed heterogeneous but effective anti-tumor activity at a low effector-to-target (E:T) ratio of 1 :1 , and the nature of the dynamic interaction of these cells resulted in transcriptional features reflecting the activity of highly functional T-cells.

[0065] An aspect of the present technology is a method for enriching a population of killer T cells with active killer T cells. The method includes the following steps:

[0066] (a) Providing (i) a microfluidic system configured for co-encapsulating single cells in microdroplets in an oil stream and sorting the microdroplets based on fluorescence within the microdroplets (see Fig. 1A for an example); (ii) a population of single killer T cells, optionally loaded with an intracellular Ca2+-sensitive fluorescent indicator; (iii) a population of single target cells, such as cancer cells or tumor-derived cells, optionally loaded with a fluorescent indicator of caspase activity, wherein either the population of T cells in (ii) or the population of target cells in (iii) contains said fluorescent indicator, but not both; and (iv) a bispecific or multispecific binding molecule having a T cell binding specificity for an antigen on a surface of the killer T cells and a target cell binding specificity for an antigen on a surface of the target cells.

[0067] (b) Co-encapsulating the killer T cells, the target cells, and the bispecific or multispecific binding molecule in microdroplets in an oil stream using the microfluidic system; alternatively the bispecific or multispecific binding molecule can be pre-bound with either the killer T cells or the target cells prior to mixing with the other cell type. (c) Allowing the killer T cells and target cells to bind the bispecific or multispecific binding molecule and to form killer T cell-binding molecule-target cell complexes in the microdroplets, whereby some of the killer T cells become activated by interaction with the target cells and the activated killer T cells kill associated target cells, and whereby a fluorescence signal is produced either from the activated killer T cells or the associated target cells;

[0068] (d) Sorting the microdroplets based on a level of said fluorescence signal, yielding a first group of microdroplets containing activated killer T cells and a second group of microdroplets containing non-activated killer T cells; the device shown in Fig. 4E can be used for fluorescence-based sorting, for example;

[0069] (e) Collecting the activated killer T cells from the first group, thereby producing a population of killer T cells that is enriched in active killer cells compared to the initially provided killer T cells; and

[0070] (f) Optionally collecting the non-activated T cells from the second group.

[0071] The bispecific or multispecific binding molecule can be any molecule having binding specificity for two, three, or more molecular targets on the surface of a cell, such as an epitope of a cell surface protein or glycoprotein. One of the targets must be found on the killer T cells to be activated and another of the targets must be found on the target cells, such as tumor cells. The bispecific or multispecific binding molecule can contain two or more individual binding agents, such as antibodies or antigen binding fragments thereof, or aptamers, having the desired binding specificities, with one or more linker molecules connecting the binding agents. The binding agents and linkers can have any chemistry and configuration known in the art. A number of suitable bispecific antibodies are commercially available and in clinical use; these can have one binding specificity for T cells and another binding specificity for a target cell, such as a specific type of cancer cell.

[0072] Following activation of killer T cells using the method described above, it may be desirable to expand the collected population of activated killer T cells. The following protocol can be used for cell expansion

[0073] Day 0: Breaking the Droplets and Initial Cell Culture o After sorting, place the collection and waste contents separately onto a PTFE filter. o Allow the oil phase to evaporate completely. The aqueous phase containing the cells should remain as a blob on top. o Once the oil has dried out, carefully collect the aqueous phase containing the cells. o Transfer the collected aqueous phase into a Il-bottom plate containing 100uL of warm, complete RPMI media . o Estimate the volume of the aqueous phase containing the cells during transfer, and then add warm, complete RPMI media to the well plate to bring the total volume to 200 pL. o Place the Il-bottom plate in an incubator set at 37°C with 5% CO2. o Let the cells rest undisturbed for 3 days to recover.

[0074] Day 3: Cytokine Addition and Media Refresh o After 3 days of incubation, gently remove 100 pL of media from the top without disturbing the settled cells. o Add 200uL of fresh warm RPMI complete media and add 5 ng / mL of IL-15 and 10 ng / mL of IL-7. o Return the plate to the incubator and continue incubation.

[0075] Day 5 and Onward: Media Refresh and Cell Expansion o Every 2 days, remove 100 pL of media from the top and replace it with same volume of fresh RPMI complete media and 5 ng / mL of IL-15 and 10 ng / mL of IL-7. o When the cells have expanded sufficiently, transfer the cells to larger well plate and increase the media volume accordingly. o Repeat this process until the cells are ready to be harvested.

[0076] Cell growth during the incubation period can be monitored by light microscopic observation.

[0077] Dynamic profiling of CD3xCD19 bispecific activated CD8+ T-cells using a single-cell droplet platform

[0078] As previously mentioned, CD3xCD19 bsAbs bring together cytotoxic CD8+ T-cells and tumor cells, resulting in T-cell dependent killing of tumor cells. However, the interaction between CD3xCD19 bsAb activated T-cells and tumor cells at the single-cell level is largely uncharacterized. Here, the inventors studied the interaction events and cytolytic activity between untreated and CD3xCD19 bsAb treated CD8+ T-cell and tumor cell coencapsulations at an E:T of 1 :1 over a period of 48h at the single-cell level (see Fig. 1B). The inventors used Raji cells, a well-studied CD19 positive lymphoma cell line as the target. The selected CD3xCD19 bsAb features a construct design similar to that of blinatumomab. The inventors selected this CD3xCD19 bsAb construct on the basis that it has been well characterized using various in-vitro and in-vivo models [4, 24, 25], First, the inventors wanted to determine if their single-cell droplet platform can be used to study the activity of bispecific antibodies, in this case a CD3xCD19 bsAb, at the single-cell level using their droplet microfluidic platform (Fig. 1A) To accomplish this, the inventors first evaluated the cytolytic activity of CD3xCD19 bsAb treated co-encapsulations continuously over a period of 48h and selected droplets for analysis based on visual confirmation of an effector to target ratio of 1 to 1.

[0079] CD3xCD19 bsAb treated co-encapsulations showed a time-dependent increase in tumor cell death compared to untreated co-encapsulations across (Fig. 2A). The inventors observed differential tumor cell killing after 6h of co-encapsulation which was sustained throughout the entire incubation period with CD3xCD19 bsAb treated cells (Fig. 2B). The tumor cell viability after 48h of incubation was significantly lower (27±13%) in CD3xCD19 bsAb treated co-encapsulations compared to untreated cells (55±1%) (Fig. 2A). CD3xCD19 bsAb treated co-encapsulations also killed Raji cells faster compared to untreated cells (Fig. 2C). Next, the inventors evaluated the synaptic contact dynamics between CD8+ T-cell and tumor cell interactions. The majority of CD3xCD19 bsAb treated CD8+ T-cells with cytolytic activity had fewer and shorter contact periods with tumor cells compared to untreated coencapsulations (Fig. 2D). CD3xCD19 bsAb treated CD8+ T-cells also were quicker to contact tumor cells compared to untreated cells (Fig. 2D). These findings are consistent with the mechanism of action of this bsAb. CD3xCD19 bsAb treated CD8+ T-cells lacking cytolytic activity had a similar number but with a wider distribution of contacts as untreated cells (Fig. 2E). These cells also stayed in contact longer with tumor cells (Fig. 2E). This likely reflects a failed attempt being made by these less functional CD3xCD19 bsAb treated CD8+ T-cells to form an immune synapse with tumor cells. CD3xCD19 bsAb treated CD8+ T-cells lacking cytolytic activity were also quicker to target and contact tumor cells (Fig. 2E), similar to CD3xCD19 bsAb treated CD8+ T-cells with cytolytic activity. Collectively, these results demonstrate that their single-cell droplet platform can be used to study the cytolytic activity and synaptic contacts of CD3xCD19 bsAb treated T-cells at the single-cell level with a high resolution of 1 to 1 effector to target ratio.

[0080] Differential killing and cell contact dynamics of CD3xCD19 bispecific activated CD8+ T-cell killers

[0081] The inventors noted that CD3xCD19 bsAb activated CD8+ T-cells differentially interacted with and killed tumor cells using their single-cell droplet platform. Given this finding, the inventors decided to study the cytotoxicity and cell contact dynamics of CD3xCD19 bsAb CD9+ T-cell killers in more detail (Fig. 3). To this end, CD3xCD19 bsAb activated CD8+ T-cell killers from each donor displayed a heterogeneous time to death (Fig. 3A). These killer cells had fewer contacts (Fig. 3B) and the majority (>60%) only needed to interact with the tumor cell once before killing it (Fig. 3C). These killer cells also stayed in contact with target cells for a significantly longer period of time (Fig. 3D). Interestingly, killer and non-killer cells had a similar time to first establish contact with tumor cells (Fig. 3E), however, the time to locate tumor cells was more variable in the killer cell population. In addition, non-killer cells from all donors were in contact with tumor cells at the start of imaging. Despite doing so, however, these cells failed to lyse the tumor cells. The inventors found these observations intriguing and decided to further characterize CD8+ T-cell killers treated with the CD3xCD19 bsAb using their function to omics workflow.

[0082] Function-to-omics technology detects well-characterized genes of CD3xCD19 bsAb activated CD8+ T-cells

[0083] The single cell droplet microfluidic analysis allowed us to characterize the CD3xCD19 bsAb effect on tumor cells at an E:T of 1 :1. The inventors observed tumor cell killing at this level of resolution along with significant heterogeneity in number of cell contacts, duration of contact and killing kinetics of co-encapsulated CD3xCD19 bsAb treated CD8+ T-cell: tumor cell pairs. The inventors hypothesized that the killing effectiveness of individual T-cells may be attributed to the functional capacity of each T-cell, and that not all CD3xCD19 bsAb treated T-cells have the same killing capabilities. Furthermore, the inventors speculated that CD3xCD19 bsAb activated CD8+ T-cell killers display unique transcriptional profiles that permit effective tumor cell lysis, and that some T-cells are inherently dysfunctional to some extent which limit their ability to engage and kill tumor cells, despite being bound to CD3xCD19 bsAb and engaged with the target cell. To investigate this further, the inventors leveraged their function-to-omics technology to separate CD3xCD19 bsAb activated CD8+ T cell killers from non-killers after co-encapsulating the cells for 18h and then subjecting these cells to bulk RNA- seq to identify mechanisms driving effective and non-effective antitumor responses (Figs. 4A- 4E). RNA-seq was performed on a total of 8 samples, 4 donors each having sorted CD3xCD19 bsAb activated CD8+ T-cell killers and non-killer samples.

[0084] To determine if the CD3xCD19 bsAb activated CD8+ T-cell killers sorted capabilities, the inventors started their analysis by assessing the gene expression of well-known cytolytic markers. Cytotoxic T-cells mediate lysis of target cells by exocytosis of specialized cytoplasmic granules containing cytotoxins perforin (PRFT), granzyme B (GZ / WB) and granulysin (GNLY), receptor-ligand binding of Fas ligand (FASL) and TNF-related apoptosisinducing ligand (TNFSF10 or TRAIL) molecules

[0026] , As shown in Figs. 5A-C, the inventors observed a significant upregulation in cytolytic genes involved in both exocytosis of lytic proteins and receptor-ligand binding with donor to donor differences. Interestingly, TNFSF10 and FASL were upregulated to a greater extent compared to cytolytic genes GZMB, GNLY and PRF1 (Fig. 5C). The inventors also observed upregulation of other granzymes (GZMK, GZMH, GZMM, and GZMA) in CD3xCD19 bsAb activated CD8+ T-cell killers. Although the main function of these granzymes are presumed to be direct elimination of target cells, the specific mechanism and type of cell death induced varies and is still an active area of investigation [27-29], For example, GZMK and GZMA have been shown to induce activation of inflammatory cytokines such as I L-1 p

[0027] , while GZMM has been demonstrated to induce morphological changes associated with microtubule disruption in target cells

[0030] , GZMH is still poorly understood given that some studies have shown that GZMH acts independently of caspases while other studies found GZMH induces caspase-dependent form of cell death

[0031] . Although the inventors sorted samples on the basis of caspase-dependent target mediated cell death, non-caspase GZMH mediated cell death of target cells cannot be ruled out since cells can simultaneously leverage caspase-dependent and caspase-independent mechanism of cell death

[0032] , Interestingly, GZMH and GZMA were both upregulated in the donors with increased TNFSF10 and FASL. Studies indicate that these molecules often work in tandem to enhance the immune response against tumors

[0033] , For instance, TNFSF10 can induce apoptosis in tumor cells, while granzymes like GZMA and GZMH can further ensure the destruction of these cells by triggering mitochondrial damage

[0033] , This coordination between TNFSF10 and granzymes may help optimize the cytotoxic potential of immune cells activated by a CD3xCD19 bsAb, and represents an example by which bispecific antibodies can be studied using their function-to-omics approach. Overall, these results suggest that the mechanism of tumor cell kill mediated by CD3xCD19 bsAb activated T-cell killers varied from donor to donor, whereby some donors may rely primarily on the release of cytotoxic granules while others more on death receptor-mediated cytotoxicity to kill tumor cells, or a combination of the two.

[0085] Immune checkpoint molecules are inhibitory receptors on the surface of immune cells that ensure appropriate regulation of the immune response. Prominent checkpoint molecules include programmed cell death receptor 1 (PD-1 / PDCD7) and CTLA-4 expressed by T-cells and other immune cells. Other checkpoint molecules include lymphocyte-activation gene 3 (LAG3), T-cell immunoglobulin and mucin domain-3 (TIM3), T-cell immunoreceptor with Ig and ITIM domains (TIGIT). The expression of immune checkpoint molecules can be used to monitor cell exhaustion as well as activation of T-cells [1 , 34], Given this, the inventors were interested in analyzing genes involved in immune checkpoint regulation. The inventors observed donor-dependent changes in gene expression of immune checkpoint genes in CD3xCD19 bsAb activated CD8+ T-cell killers compared to non-killers (Figs. 5D-F). CD244, TIGIT, and ENTPD1 were upregulated in CD3xCD19 bsAb activated CD8+ T-cell killers in 3 of the 4 donors whereas EOMES, LAG3, and HAVCR2 were upregulated in only 1 of the 4 donors (Figs. 5C, D). CD244 expression has been shown to be a mediator of CD8+ T-cell exhaustion in the setting of persistent antigen exposure but it has also been shown to be coexpressed on a subset of antigen-experienced effector and effector memory CD8+ T-cells along with other immunoregulatory receptors such as CD39 (ENTPD1) but not TIGIT

[0035] , There is ample evidence that the TIGIT pathway regulates T-cell mediated tumor recognition but also contributes to T-cell exhaustion

[0036] , Interestingly, 2 of the 4 donors with increased expression of immune checkpoint genes in CD3xCD19 bsAb activated CD8+ T-cell killers did not have any significant change in key cytokine genes such as TNF and IFNG. Given that the loss of effector functions such as functional impairment of proinflammatory cytokines TNF-a and IFN-y is an early hallmark of T-cell exhaustion, it could be that CD3xCD19 bsAb activated CD8+ T-cell killers from donors 2 and 4 are more prone to exhaustion, or these cells are primed for other functions. Whereas, in donors 1 and 3, although upregulation of immune checkpoint genes was observed in CD3xCD19 bsAb activated CD8+ T-cell killers, these donors maintained gene expression of TNF, IFNG, and IL2, suggesting that these cells are not exhausted and the upregulation of immune checkpoint genes is more reflective of an activated state. Overall, these results suggest donor to donor differences in the immune checkpoint gene expression profile of CD3xCD19 bsAb activated CD8+ T-cell killers. Additional studies are needed to discern the biological consequences of CD3xCD19 bsAb activated CD8+ T-cell killers with different upregulated immune checkpoint gene profiles.

[0086] The CD3 arm of the CD3xCD19 bsAb is a common signaling molecule complex associated with the T-cell receptor (TOR) and has a high potency for T-cell activation after binding to the CD19 antigen on the surface of tumor cells

[0015] , In addition to TCR binding, a number of secondary signals in CD8+ T-cells become activated after forming an immunological synapse with target cells

[0026] , To assess the effect of CD3xCD19 bsAb treatment on T-cell activation in CD8+ T-cell killers versus non-killers, the inventors selected a panel of well characterized activation genes involved in TCR signaling and downstream effector responses

[0037] , These genes are CD2, CD3, CD8, TCR, IFNG, TNF, IL2, CD28, TNFRSF4, TNFRSF9 and ICOS. TCR related genes (TRA, TRB, TRG, TRD, CD3D, CD3E, CD3G, CD8A, CD8B and ICOS) were significantly upregulated in CD3xCD19 bsAb activated CD8+ T-cell killers compared to non-killers in all donors with the exception of ICOS for donor 2 (Fig. 5G). Gene expression levels of CD2, TNFRSF4, IL-2, and IFNG were also higher in CD3xCD19 bsAb activated CD8+ T-cell killers compared to non-killers in two of the four donors. TNF and TNFRSF9 levels from CD3xCD19 bsAb activated CD8+ T-cell killers were upregulated in one of the four donors (Fig. 5G). Overall, CD3xCD19 bsAb-dependent T-cell activation was observed in all four donors with donor to donor differences in the extent and genes involved in activating the cells (Fig. 5H).

[0087] Table 1. Targeted panel of immune synapse genes and their functions

[0088] CD3xCD19 bsAb activated CD8+ T cell killers display unique immune synapse profiles

[0089] To determine the genes involved in immune synapse formation, the inventors selected a total of 40 family member genes (Table 1) with known roles in the formation of an immunological synapse. Each donor was analyzed on its own to better understand donor to donor differences in the ability of CD3xCD19 bsab activated CD8+ T-cell killers to target and interact with target cells. Starting with donor 1 (Fig. 6A), immune synapse genes STXBP3, VAMP2, SPN, ITGA4, ICAM2, STX16, DYNC1 I2 and WASL were among the highest expressing genes in CD3xCD19 bsab activated CD8+ T-cell killers and non-killers, demonstrating that both cell populations were interacting with target cells. The majority of these synapse genes (90%) were upregulated in CD3xCD19 bsab CD8+ T-cell killers compared to non-killers. This highlights the importance of a coordinated response to target and kill tumor cells. The downregulation of NCKAP5, VAV3, WASL, STX16, and DYNC1 H1 in CD3xCD19 bsab activated CD8+ T-cell killers might be associated with a dynamic regulation of energy conservation and reducing excessive activation, and transitioning the cells to a state of killing. For donor 2 (Fig. 6B), the highest expressing immune synapse genes in CD3xCD19 bsab activated CD8+ T-cell killers and non-killers were MYH9, TLN1 , VCL, and VAV2. Compared to non-killer cells, CD3xCD19 bsab activated CD8+ T-cell killers upregulated WIPF3, PTPN family members (PTPN3, PTPN13, PTPN14, PTPN20, and PTPN21), MAPK4, DOCK family members (DOCK1 and DOCK5), LAMP3, ACTN2, STXBP5-AS1 , NCKAP5L, and RAB27A, and downregulated DOCK11 , VCL, TLN1 , MAPK8IP3, VAV2 and MYH9. These upregulated genes reflects a critical adaptation to enhance the cytotoxic function of CD3xCD19 bsab activated CD8+ T-cells. These changes would be expected to improve actin cytoskeleton regulation, refine signal transduction, enable efficient vesicle trafficking and enhance antigen processing and presentation. The downregulated genes likely provide CD3xCD19 bsab activated CD8+ T-cells with a balanced response to forming a stable immunological synapse while maintaining the flexibility to engage multiple target cells. Compared to the other donors, donor 3 expressed more immune synapse genes (Fig. 6C). The highest expressing immune synapse genes in both cell types were LCK, PTPN22, ICAM1 , PTPN7, MAPKAPK5-AS1 , NCK1 , ITGAL, and RAB27A. CD3xCD19 bsab activated CD8+ T- cells also expressed higher levels of VAMP5, PTPN18, and PTPN2 while non-killers also expressed relatively high levels of VAV2, MAPK9 and STX18-AS1. Interestingly, the upregulation of antisense genes in CD3xCD19 bsab activated CD8+ T-cell killers and nonkillers likely represents an attempt to limit the expression of their target genes and likely ensure a balanced immune response. CD3xCD19 bsab activated CD8+ T-cell killers from this donor upregulated many immune synapse genes including genes from the same family members including DOCK, STX, PTPN, MAPK, VAMP, and RAB27. Collectively, these genes are critical for the formation of the immune synapse and likely enabled these cells to more effectively kill target cells via death receptor pathways involving TNFSF10 and FASL, given the combined expression of these cytolytic genes were upregulated to a greater extent compared to the other donors (Fig. 5A). Interestingly, donor 4 expressed significantly fewer immune synapse genes (Fig. 6D). The immune synapse genes with the highest expression were ACTN1 and ZAP70, present in both CD3xCD19 bsab activated CD8+ T-cell killers and non-killers. However, CD3xCD19 bsab activated CD8+ T-cell killers were distinguishable from non-killers by the upregulation of DYNC1I1 , ZAP70, DOCK5, UNC13B, ACTN1 and PTPN14. Collectively, their function to omics platform identified immune synapse genes in both killer and non-killer cells. Killer cells also displaying increased immune synapse activity as evidenced by the upregulation of more immune synapse genes.

[0090] CD3xCD19 bsAb activated CD8+ T cell killers display unique metabolic profiles

[0091] The metabolic demands of naive T-cells are relatively low. Upon stimulation through the TCR receptor and co-stimulatory molecules, however, T-cells activate various signaling pathways which are accompanied by changes in cellular metabolism to support their proliferation and effector function

[0038] , To this end, resting T-cells shift from oxidative phosphorylation (OXPHOS) to aerobic glycolysis which is coupled with increased activity of the pentose phosphate pathway (PPP)

[0039] , Activated T-cells, especially those transitioning to a memory state or those in low-nutrient environments, benefit from metabolic flexibility by leveraging fatty acid oxidation (FAO)

[0040] , Here, the inventors evaluated genes from each of these metabolic pathways in each of their donor samples comparing gene expression of CD3xCD19 bsAb activated CD8+ T-cell killers to non-killers (Fig. 7A). Starting with donor 1 , MT-ND6 was the highest expressing gene in killer and non-killer cells. Non-killer cells also had relatively high levels of COX4I2. Comparing these two cell populations, CD3xCD19 bsAb activated CD8+ T-cell killers mostly upregulated genes from all four metabolic pathways, particularly glycolysis, FAO and OXPHOS. A few genes (ACADL, COX4I2 and ATP6V1 E2) were downregulated in killer cells. Donor 2 had a similar metabolic profile to donor 1 , whereby all four metabolic pathways were being utilized in killer cells compared to non-killer cells (Fig. 7B). The highest expressing metabolic gene in this donor was also MT-ND6, in both killer and non-killer cells. All metabolic genes shown in Fig.7B were upregulated in killers versus nonkillers with the exception of PPP related genes H6PD and TKTL2. The metabolic profile of donor 3 was similar to the previous two donors with the exception that more genes were downregulated in killer cells (Fig. 7C). Once again MT-ND6 was the highest expressing metabolic gene in both cell populations. The downregulated FAO genes were SOD, HACD2, and ELOVL6. The downregulated OXPHOS genes were NDUFA13, ATP6V1A, ATP6V0A1 , ATP6V1G2, ATP6V1C2, and ATP6V0A4. There were only two differentially expressed PPP genes in this donor, upregulation of H6P and downregulation of G6PD. Donor 4 was the least metabolically active donor and primarily relied on the upregulation of several OXPHOS genes (MT-ATP8, MT-ND3, MT-ND4L, MT-CO3, MT-ATP6, MT-ND2, MT-ND6, MT-CYB, MT-ND5, MT-CO2 and MT-ND4) to meet energy demands (Fig. 7D). This donor may be considered less metabolically fit the lack of utilization of other pathways and reliance on mostly OXPHOS.

[0092] To further characterize CD3xCD19 bsAb activated CD8+ T-cell killers, the inventors conducted Gene Ontology (GO), Kyoto Encyclopedia of Genes and Genomes (KEGG), and Reactome enrichment analysis of their bulk RNA-seq datasets (Fig.8) and investigated the biological pathways associated with killing target cells for each donor. The top 30 upregulated pathways from each pathway enrichment analysis for each donor is shown in Figs. 8A-8D. Starting with donor 1 (Fig. 8A), the inventors noted an upregulation in pathways involving the cell membrane, chemotaxis, cell adhesion, ion channels, axons, transporter activity, G-protein coupled receptor activity, and receptor activity from GO. From KEGG, there was an increase in several signaling pathways (MAPK, PI3K-Akt, and Hippo), neuroactive ligand-receptor interaction, cytokine-cytokine receptor interaction, cell adhesion, GABAergic synapse and axon guidance. From Reactome, there was an upregulation in pathways involving GPCR ligand binding, extracellular matrix, collagen formation, transmission across chemical synapses, G-alpha signaling, GABA-A receptor activation, synaptic interactions, and integrin interactions. Collectively, these findings provide evidence that the CD3xCD19 bsAb activated CD8+ T-cell killers from this donor utilize various pathways and are primed to effectively locate, adhere to, and destroy tumor cells. Next, the inventors evaluated samples from donor 2 (Fig. 8B). CD3xCD19 bsAb activated CD8+ T-cell killers from donor 2 mainly upregulated many of the same GO pathways as donor 1 including the extracellular matrix, collagen, detection of stimulus, ion channels, and cell membrane pathways. From KEGG, similar to GO the inventors identified pathways involving ligand-receptor interactions, and extracellular matrix. In addition, pathways involving ABC transporters, phospholipase D signaling, complement and coagulation cascades, protein digestion and absorption, and metabolism involving histidine and glycerolipids. Reactome analysis also identified pathways involving the extracellular matrix, complement and collagen. Pathways involving G alpha signaling, SLC-mediated transmembrane transport, transmission across chemical synapses, GPCR ligand binding, integrins. Next, the inventors evaluated donor 3 (Fig. 8C). Again, from their GO analysis pathways involving chemotaxis, extracellular matrix, cell membrane, ion channels, and G- protein coupled receptor activity were upregulated in CD3xCD19 bsAb activated CD8+ T-cell killers. Unlike the other donors, donor 3 also upregulated the lipid transport and localization pathways, glycosaminoglycan binding, and receptor kinase and tyrosine activity in CD3xCD19 bsAb activated CD8+ T-cell killers. From KEGG, pathways involving cell membrane receptor interactions, protein digestion and absorption, and cell adhesion were also upregulated in CD3xCD19 bsAb activated CD8+ T-cell killers. Interestingly, the ABC transporter pathway was also upregulated. Reactome enrichment analysis identified extracellular matrix, collagen, integrin and non-integrin interactions, cGMP effects, immunoregulatory interactions between a lymphoid and non-lymphoid cell, PTK2 signaling involving MET, NCAM1 interactions, GPCR ligand binding, peptide ligand-binding receptors and nitric oxide stimulated guanylate cyclase. Last, the inventors evaluated CD3xCD19 bsAb activated CD8+ T-cell killers from donor 4 (Fig. 8D). The cells from this donor upregulated T-cell activation pathways, adhesion responses, immunological synapse, calcium signaling, signaling adapter activity, S100 protein binding, and dynein light chain binding according to their GO enrichment analysis. From KEGG, similar to GO, the inventors noted an upregulation in T-cell activation, extracellular matrix, and cell adhesion responses. In addition, there was an upregulation of Th1 , Th2, Th17 and hematopoietic cell lineage pathways. Antigen processing, ABC transporters, and the JAK / STAT signaling pathways were also upregulated. Reactome enrichment analysis identified many of the same pathways including T-cell activation, immune synapse, and cell surface interactions. Several interleukin signaling pathways were also upregulated including IL-4, IL-10, IL-13 and IL-35 along with clathrin-mediated endocytosis. In addition, DAP12 was upregulated along with TP53 related genes involving cell death whose specific roles in p53- dependent apoptosis remains uncertain. The results from this enrichment analysis demonstrates donor specific upregulation of pathways driving CD3xCD19 bsAb activated CD8+ T-cell killers as well as pathways shared among these donors. CD3xCD19 bsAb activated CD8+ T-cell killers from all donors upregulated the extracellular matrix (ECM) receptor interaction and ABC transporter pathways. Compared to non-killers, CD3xCD19 bsAb activated CD8+ T-cell killers upregulated integrins, lamins, collagen and syndecan related genes which have been shown to modulate leukocyte cell function and enhance various signaling pathways that play a key role in regulating T-cell responses [41-44], Another related pathway also upregulated in the majority of donors was the cell adhesion molecule pathway, which contains genes involved in TCR signaling including CD28, ICOS, CD8, CD40L, CD6, TIGIT and ITGAL and leukocyte adhesion molecules such as ITGAL, ITGAM, SELPG, PECAM1 and CD226 which are involved in cellular interactions to facilitate T-cell mediated killing. ABC transporters are a heterogeneous group of ATP- dependent transport proteins, which recently were found to regulate the development and function of different T-cell populations

[0045] , In CD8+ T-cells, ABC transporter ABCB1 was shown to regulate memory function and found to be increased in CD8+ effector memory and central memory cells

[0045] , Furthermore, ABCB1 was found to be important for mitochondrial fitness by protecting against oxidative stress during early activation of CD8+ T-cells and memory formation. The inventors found several ABC transporter genes upregulated in their data-set from CD3xCD19 bsAb activated T-cell killers including ACB1. Additional KEGG pathways upregulated in the majority of donors tested were the neuroactive ligand-receptor interaction, complement and coagulation cascade, circadian entrainment and dilated cardiomyopathy pathways. A closer look at the genes upregulated in these pathways reveals genes involved in the activation, chemotaxis and effector functions such as HRH. Histamine (HRH) has been shown to increase levels of chemokine CCL17 and CCL22 which are chemoattractants of Th2 lymphocytes

[0046] , The inventors found this observation to be particularly interesting given that Th2 cells can help facilitate primary and long-lived memory CD8+ T-cell responses

[0047] ,

[0093] Differential alternative splicing analysis for CD3xCD19 bsAb activated CD8+ T-cell killers vs. non-killers

[0094] Given the observed donor to donor variability in CD3xCD19 bsAb activated CD8+ T- cell killer profiles of cytolytic, activation / immune checkpoint, immune synapse and metabolic genes, the inventors leveraged their experience with alternative splicing analysis to dissect these samples further and compared the alternative splicing patterns in CD3xCD19 bsAb activated CD8+ T-cell killers to non-killers from each donor. To be as comprehensive as possible the inventors looked for dysregulation in all five modes of alternative splicing: skipped exon (SE), mutually exclusive exons (MXE), alternative 3’ splice site (A3SS), alternative 5’ splice site (A5SS), and retained introns (Rl) that are generated by rMATS-turbo. From all the transcripts in which alternative splicing was detected by rMATS-turbo, the inventors picked the transcripts corresponding to the genes that showed dysregulation in any of the four groups of genes: cytolytic, immune checkpoint, activation and metabolic. If difference in alternative splicing was seen in one or more of the transcripts corresponding to the dysregulated genes, the inventors selected those genes that had showed significant differential alternative splicing (FDR < 0.5) in each of the five modes of alternative splicing. As skipped exons are by far the most prevalent mode of alternative splicing, most of the significant differences arose in this category. Interestingly, CD3xCD19 bsAb activated CD8+ T-cell from donor 1 showed several significant differences in splicing patterns between killer and non-killer cells for the GK, GZMH, GNLY and GZMB amongst the cytolytic genes, LAG3 amongst the immune checkpoint genes and NQO1, TXNRD1, NRF1 amongst the metabolic genes. Significant difference in alternative splicing were also seen in the A3SS and Rl modes for GNLY. Moreover, there were significant differences in alternative splicing in GNLY, TXNRD1 for donors 3 and 4 too. While there was an upregulation in GNLY expression, in contrast, the expression of TXNRD1, was unchanged between CD3xCD19 bsAb activated CD8+ T-cell killers compared to non-killers for all donors as noted before in the DEG analysis.

[0095] Aggressive B-cell lymphomas are generally treatable but there remains a significant need to identify more targeted and non-toxic therapeutic agents as well as biomarkers that predict enhanced activity in the treatment of B-cell lymphoma patients. In their study, the inventors investigated various aspects of cytotoxic CD8+ T-cell interactions with Raji cells, a B-cell lymphoma cell-line, to characterize the effect of a CD3xCD19 bsAb using a droplet microfluidic function-to-omics approach and identify biological mechanisms driving productive anti-tumor responses. Direct physical interactions occurring between CD3xCD19 bsAb activated T-cells and target cells have the potential to determine CD8 T-cell effector outcome in the short run and the fate of T cell in the long run

[0048] , Thus, quantifying the nature and response of single CD8+ T-cells in a defined microenvironment is useful for illustrating the functional and mechanistic diversity in CD8 T-cell interactions with tumor cells. To this end, the inventors leveraged their established droplet microfluidic technology ScanDrop which has been used to study other immune cells, such as NK cells [49-51], to assess dynamic CD8+ T cell - target cell interactions. The inventors investigated the effector-target cell contact dynamics as well as determined the cytotoxicity and transcriptomics of CD3xCD19 bsAb activated CD8+ T cell killers.

[0096] Various aspects of cytotoxic T-cell interactions with target cells were investigated to characterize the effect of CD3xCD19 bsAb activation and heterogeneity of the anti-tumor response. Here, CD8+ T-cells incubated with and without the CD3 x CD19 bsAb and tumor cells were captured in the droplet microfluidic device through separate inlets in a single step loading process to ensure that interaction and activation occurred only in the droplets. Since the cells are not physically or chemically constrained in the droplets, they are highly mobile allowing us to track the cells from the point of contact initiation and conjugate formation. Effector T-cells such as CD8+ T-cells need to form immunological synapses with recognizes target cells to elicit productive cytolytic effects

[0052] , In fact, facilitating immunological synapse formation is the principle pharmacological action of most T-cell based cancer immunotherapies

[0053] , CD3 x CD19 bsAbs kill malignant cells via mechanisms such as cytolytic immune synapses formation, however, the dynamic interaction events induced by the CD3 x CD19 bsAb between CD8+ T-cells and tumor cells are largely unknown. Here the inventors characterized the initial interaction and dynamics of synaptic conjugation between untreated and CD3 x CD19 bsAb treated CD8+ T-cell and tumor cells. The inventors used Raji cells, a well-studied CD19+ lymphoma cell line used to evaluate bsAbs [54, 55], as the target model system and demonstrated significant differences in the way CD3xCD19 bsAb activated T-cells interact with and eventually kill target cells at an effector-target ratio of only 1 :1. Their study showed that CD3xCD19 bsAb activated CD8+ T-cells required fewer interactions with target cells prior to killing them and effectively lysed target cells. The inventors noted significant donor to donor differences in the interaction dynamics (number of contacts and contact duration), rate of tumor cell killing, duration to target mediated cell death and total tumor cell death. To the best of their knowledge, this is the first time that CD3xCD19 bsAb activated CD8+ T-cells has demonstrated in-vitro anti-tumor activity at this level of single-cell resolution using droplet microfluidics. Thus, the sensitivity of the microfluidic assay exceeded that of a conventional assay in this specific target cell line. The inventors surmise that increasing the proximity of effector and target cells enhanced the possibility of a higher kill at lower E:T ratio using their single-cell droplet platform. While further studies are required to confirm this response in multiple target cell lines and with CD8+ T cells from more donors, their preliminary investigation provides an approach to study the interaction and cytotoxicity kinetics of bsAbs with the desired tumor type. Their droplet microfluidic technology also enables in-depth characterization of various bsAb constructs targeting any immune cell and target cell at the single-cell level of resolution. Furthermore, their approach can also be applied to the characterization of other immunotherapies currently under development such as cell therapies

[0056] ,

[0097] To capture the full spectrum of the mechanistic effect of CD3xCD19 bsAb activation within the heterogenous CD8+ T-cell population, the inventors applied their function-to-omics approach to study the transcriptomics of F-FADS sorted CD8+ T-cell killers compared to nonkillers. First, the inventors analyzed a panel of known cytolytic genes and found that CD3xCD19 bsAb CD8+ T-cell killers compared to non-killer cells upregulated PRF1, GZMB and GNLY, as well as in death receptor genes FASL and TNFSF10. In fact, there appeared to be a reciprocal relationship between these two cytolytic modes of cell death, i.e. the donors with lower PRF1 / GZMB / GNLY had higher levels of FASL / TNFSF10 and vice versa. A recent study suggested that NK cells switch from inducing a fast GZMB-mediated cell death in their first killing events to a slow death receptor-mediated killing during subsequent tumor cell interactions

[0057] , The inventors surmised the same mechanism may be present here for CD3xCD19 bsAb activated CD8+ T-cell killers. Although this phenomenon is less studied in T-cells, it could be that those donors with higher expression of FASL / TNFSF10 on CD3xCD19 bsAb activated CD8+ T cells have greater serial killing capacity. Future studies will focus on elucidating the biological mechanisms and pathways enabling CD3xCD19 bsAb activated CD8+ T-cells to kill target cells serially. The inventors then analyzed a panel of activation and immune checkpoint genes and showed that CD3xCD19 bsAb activated CD8+ T-cell killers upregulated various genes involved in T-cell activation and immune checkpoint. Interestingly, these activated T-cell killers displayed a diversity of immune checkpoint genes despite being at a 1 :1 E:T ratio with the same tumor cells. This may be a result of killing tumor cells at different times or these cells are inherently primed to express certain immune checkpoint genes. CD3xCD19 bsAb activated CD8+ T-cell killers from all donors upregulated all four TOR chains (TRA, TRB, TRD and TRG) indicating that TCR triggering is critical for effective elimination of tumor cells. Not all CD3xCD19 bsAb activated CD8+ T-cell killers upregulated key activation genes such as IFNG and TNF, however, all the donors displayed some level of activation with donor to donor differences in the type and extent of activation. This indicates that CD8+ T-cells from different donors may not have the same level of activation needed to effectively clear tumor cells. Given that T-cells need to form a stable immune synapse with tumor cells to be effectively activated and lyse tumor cells, the inventors analyzed 40 synapse family member genes with known roles in immune synapse activity and formation. CD3xCD19 bsAb activated CD8+ T-cell killers from all donors upregulated immune synapse genes, however, the extent and level of activation varied from donor to donor. Their findings suggest donor to donor differences in the ability of CD3xCD19 bsAb activated CD8+ T-cell killers to activate genes critical for the formation of a stable immune synapse. Additional research studies are needed to discern these differences and to determine the long-term impact on T- cell fitness and cytotoxicity. Recently T-cell metabolic flexibility has emerged as a critical determinant of CD8+ T-cell anti-tumor activity yet the metabolic pathways involved in CD3xCD19 bsAb activated T-cells is not well understood. In their study the inventors focused on efforts initially on metabolic pathways that have been well-characterized in immune cells including glycolysis, OXPHOS, FAO and PPP. All but one of the donor samples the inventors tested displayed some level of metabolic flexibility, with CD3xCD19 bsAb activated CD8+ T- cell killers relying one more than one metabolic pathway. The observed differences in metabolic activity and pathway utilization warrants further investigation. Additional studies are needed to determine if a lack of T-cell metabolic flexibility in CD8+ T-cell killers limits their persistence and ability to kill tumor cells again. Finally, the inventors showed the incorporation of alternative splicing analysis into their wokflow as a means to better understand the differences in gene expression between CD3xCD19 bsAb activated CD8+ T-cell killers and non-killers. The observed splicing differences can be probed further in genes that are known to be significant in some of the important pathways.

[0098] The characterization of different CD3xCD19 bsAbs are an active area of research for various hematological malignancies and solid tumors [1 , 58-60], Despite the clinical benefit of bsAbs such as blinatumomab, certain patients still fail to respond to therapy

[0058] , The reasons for this are complex and can be patient dependent

[0061] , Most of the research conducted to date has focused primarily on the intrinsic characteristics of tumors, such as CD19 loss, which may be responsible for treatment induced resistance [7], The inventors also know that highly functional T-cells are required for successful bsAb therapy [62-64], More recently, single-cell analysis of bulk cells activated with blinatumomab has furthered their understanding on the importance of having specific T-cell responses present for effective bsAb mediated antitumor responses

[0016] , In this study, Huo, Y., et al identified different T-cell populations and the mechanism behind target cell-dependent activation in response to blinatumomab treatment, and showed that blinatumumab-induced transcriptional changes reflecting the functional immune activity of the blinatumomab-activated T cells, including upregulation of pathways such as the immune system, glycolysis, IFNa signaling, gap junctions and IFNG signaling. The challenge, however, with these studies is that T-cells are heterogeneous in their functional capabilities (i.e. , some T cells can kill while others cannot) and currently there is a lack of technologies that can specifically probe the functionally active T-cell killers. Using their novel approach permitted the investigation of CD3xCD19 bsAb activated CD8+ T cell killers from various donor samples, including the interaction dynamic events and cytolytic activity at the single-cell level between individual T-cells and tumor cells, and the identification of genes and biological pathways enriched in CD3xCD19 bsAb activated CD8+ T-cell killers. Their work has shed light into the mechanisms underlying target celldependent T-cell killing induced by a CD3xCD19 bsAb. These findings indicate that CD3xCD19 bsAb activated CD8+ T-cells can be studied at the single-cell level with high resolution (effector to target ratio of 1 :1) to better assess the in-vitro effects of this class of therapeutics and that F-FADS combined with bulk RNA-seq can be used to study the transcriptional drivers of productive antitumor activity driven by activated CD8+ T-cell killers. To the best of their knowledge, no study previously evaluated the dynamic interaction events of CD3xCD19 bsAb activated CD8+ T-cells with CD19+ tumor cells at the single-cell level using droplet microfluidics. While their study was conducted with a single CD3xCD19 bsAb format and using primary samples from healthy donors, their approach has the potential to be a dynamic, sensitive technology at low E:T ratio to evaluate other bsAb constructs and generate transcriptional data-sets to improve the design of future immunotherapies. An interesting finding in their study was the donor to donor heterogeneity observed in both their single-cell droplet platform and function to omics analysis, and suggests that a precision medicine approach may be needed to elicit effective and durable anti-tumor responses in patients treated with CD3xCD19 bsAbs. This can be achieved by pre-selecting patients most likely to respond or designing immunotherapeutics tailored to a patient’s immune system.

[0099] EXAMPLES

[0100] Example 1. Cell isolation and culture.

[0101] Raji cells (B cell lymphoma cell line) were purchased from American Type Culture Collection (Manassas, VA, USA) and maintained in RPMI-1640 medium supplemented with 10% fetal bovine serum (FBS) and 1% antibiotic-antimycotic solution (Corning Cellgro, Manassas, VA, USA). Primary human cryopreserved CD3+ CD8+ cells (> 90% purity) were purchased from Stemcell Technologies (Cambridge, MA, USA), thawed the day prior to experiments, and rested overnight in the same media as Raji cells. All cells were grown at 37°C and 5% CO2 in a humidified atmosphere. Raji cells were routinely passaged every 2-3 days and harvested at a density of 1 x 106 viable cells / mL. Primary human CD3+ CD8+ T cells were obtained from four different donors.

[0102] Example 2. Microfluidic Device Fabrication and Droplet Generation.

[0103] Microfluidic devices were fabricated using standard soft lithography techniques with PDMS and designed as previously described [17, 18], Each inlet of the device was connected to individual syringes containing aqueous (i.e. cell suspension in media) or oil-based fluids through Tygon Micro Bore PVC Tubing of the following dimensions: 0.010” ID, 0.030” OD, 0.010”wall (Small Parts Inc., FL, USA). The device was treated with AQUAPEL glass treatment (Aquapel, Pittsburg, PA, USA) for 15 min, then flushed with air immediately before experiments. The syringes were operated by individually programmable syringe pumps (Harvard Apparatus, USA). The oil to aqueous flow rates were generally maintained at a ratio of 4:1 to obtain optimal droplet sizes. The oil phase consisted of Fluorinert® FC-40 (Sigma, St. Louis, MO, USA) supplemented with 2% w / w surfactant (008-FluoroSurfactant, Ran Biotechnologies, Severely, MA, USA).

[0104] Example 3. Cell-Mediated Cytotoxicity and Dynamic Cell-Pairing Interactions using Droplet Microfluidics.

[0105] Tumor cell viability in droplets was determined by a live viability cytotoxicity assay reagent calcein AM (Life Technologies, Carlsbad, CA, USA) and changes in cell morphology. The final concentration of calcein AM was maintained at 2 pM and detected by time-lapse microscopy at excitation / emission: 494 / 517 nm. The proportion of live tumor cells was calculated as the number of live tumor cells to the total number of tumor cells calculated as a percentage, and expressed as “Target Cell Viability”. Primary human CD8+ T-cells were left unlabeled and included with the CD3xCD19 bsAb (InvivoGen, San Diego, CA, USA) at 0.5 pg / mL loaded in a syringe separate from the Raji cells at an initial concentration of 4 million / mL and 3 million / mL respectively to target an E:T ratio of 1 :1.

[0106] Dynamic cell-pairing interactions was determined by live cell-imaging analysis by counting the number of interactions between CD8+ T-cells and Raji cells as well as the duration of each contact period. The percentage of T-cells engaged with Raji cells from the onset (i.e. , start of the live cell imaging defined as T= 0 h) was determined by dividing the number of T-cells in contact with a Raji cell at T= Oh by the total number of T-cells x 100%. The time to first contact was determined by the time period (min) from the start of live cell imaging to the first time a T-cell contacts a Raji cell.

[0107] Example 4. Single-cell image acquisition and processing.

[0108] Cell images in droplets were captured using ZEISS AXIO OBSERVER Z1 Microscope (Zeiss, Germany) equipped with Hamamatsu digital camera C10600 ORCA-R2, 10x-40x objectives and standard FITC / DAPI / TRITC filters. The microfluidic device containing cell- encapsulated droplets was maintained in a humidified microscopic stage-top incubator at 37°C and 5% CO2 for the duration of the experiment. All time-lapse images were obtained by automated software control. The array was scanned to identify locations containing 1 :1 effector: target ratio and the specific x-, y-, and z- positions were programmed in the Zen imaging program (Zeiss). Images of these locations were obtained every 30 minutes for a total period of 48 h. Image analysis was done using Microsoft Office Excel. Contact periods were defined as cells forming visible conjugates. All periods of association and dissociation were counted for each cell and represented as percentage of total cells analyzed. CD3 bsAb mediated cytolysis of target cells was characterized by loss of calcein AM fluorescence from the target cells. Target cell death was further verified by membrane rupture and blebbing

[0019] , Killing time for target cell death was defined as the time elapsed from the initiation of contact to loss of fluorescence and morphological changes (as described above). All statistical analysis was performed using non-parametric T- test; p value < 0.05 was considered statistically significant.

[0109] Example 5. Functional Fluorescence-Activated Droplet Sorting (F-FADS).

[0110] To prepare the cells for sorting, CD8+ T-cells and CD3xCD19 bsAb at 0.5ug / mL were loaded in a syringe at an initial concentration of 4 million / mL. Raji cells were incubated with 24uL of diluted CELLEVENT Caspase-3 / 7 detection reagent (ThermoFisher, Waltham, MA, USA) and loaded in a separate syringe at an initial concentration of 3 million / mL. Each inlet of the device was connected to individual syringes containing either Raji cells, CD8+ T-cells treated with 0.5ug / mL of CD3xCD19 bsAb (InvivoGen, San Diego, CA, USA) or oil-based fluids through PVC Tubing of the following dimensions: 0.010” ID, 0.030” OD, 0.010”wall (US Plastic Corp., OH, USA). The F-FADS droplet generation device was filled with Novec 7500 oil containing 3% surfactant. The syringes were immediately loaded on the same programable pump and the oil to aqueous flow rates were generally maintained at a ratio of 2:1 to obtain optimal droplet sizes. Droplets were collected in a 2mL conical vial containing 1 mL of mineral oil (Thermo Fisher Scientific, Waltham, MA) for 1.5 hrs and then incubated at 37°C and 5% CO2 in a humidified atmosphere for 16 h. Following incubation, droplet emulsions were directly injected into the sorting device, and NK cells that killed K-562 cells were sorted based on fluorescent signal. Fluorescence was detected using an external 488 nm laser (Opto Engine LLC, Midvale, UT) and PMT (Hamamatsu Photonics, Hamamatsu City, Japan). Signal detection and sorting impulse were regulated by a control unit and LabView software (National Instruments, Woburn, MA), and electronic impulse to the device was amplified with a Trek 609C-6 High Voltage Amplifier (Advanced Energy, Denver, CO). The sorting device is designed with a reinjection channel and a sorting junction with a Y shaped channel leading towards outlets. Each channel is tuned so that droplets will naturally flow down the bottom channel unless manipulated by an external force to flow towards the top channel. The resistances within each channel are controlled by external syringe pumps (Harvard Apparatus, MA) withdrawing from each outlet at the same flow rate as is being perfused into the device to prevent internal pressure buildup. Upon detection of a positive kill within a droplet, the system activates the circuit within the device sending a 1 kV voltage through internally embedded circuits creating a dielectrophoretic field. This field creates a dipole within the droplets that flow past the circuit pulling them towards the upper channel. Effectively separating the populations into “killed” (upper channel) and “non-killed” (bottom channel). Detection and activation occur within a 500 ms timeframe and can assess 10,000 droplets per hour. For all sorting experiments, baseline thresholds for sorting are set at the start of the experiment, to automatically trigger sorting impulses on any fluorescent peak above baseline or background noise. Droplets containing dead tumor cells were separated from droplets containing live tumor cells into separate syringes containing complete media until sufficient number of desired droplets were collected.

[0111] Example 6. RNA extraction from sorted droplets.

[0112] After collecting droplets from the sorting device, emulsions were removed through gentle pipetting in excess cell media. The solution was allowed to settle, and the aqueous layer was pipetted put and transferred to anew vial. Cells were pelleted, and RNA extracted via RNeasy mini kit (QIAGEN, Venlo, Netherlands). Cell pellets were resuspended with buffer RLT and run through QIASHREDDER columns to homogenize the cells and remove any impurities. RNA was then precipitated and cleaned following kit instructions and eluted in 25uL of RNAse-free water. Samples were then stored at -80°C.

[0113] Example 7. RNA-sequencing analysis.

[0114] F-FADS sorted RNA samples were subjected to the ultra-low input mRNA non- directional sequencing analysis pipeline (Novogene, Beijing, China). Briefly, mRNA was purified from total RNA using poly-T oligo-attached magnetic beads. After fragmentation, the first strand cDNA was synthesized using random hexamer primers, followed by the second strand cDNA synthesis using dTTP for non-directional library. Samples were ready after end repair, A-tailing, adapter ligation, size selection, amplification and purification. The library was checked with Qubit and real-time PCR for quantitation and bioanalyzer for size distribution detection. Quantified libraries were pooled and sequenced on an illumina instrument (NQVASEQ6000) according to effective library concentration and data amount. Data quality control raw reads of fastq format were firstly processed through Novogene’s perl scripts. In this step, clean reads were obtained by removing reads containing adapter, reads containing ploy-N and low-quality reads from raw data. At the same time, Q20, Q30 and GC content the clean data were calculated. All the downstream analyses were based on clean data with high quality.

[0115] Reference genome and gene model annotation files were downloaded from the genome website directly. Index of the reference genome was built using Hisat2 v2.0.5 and paired-end clean 1 reads were aligned to the reference genome using Hisat2 v2.0.5. The inventors selected Hisat2

[0020] as the mapping tool for that Hisat2 can generate a database of splice junctions based on the gene model annotation file and thus a better mapping result than other non-splice mapping tools, featurecounts

[0021] v1.5.0-p3 was used to count the reads numbers mapped to each gene. Then Fragments Per Kilobase of transcript per Million mapped reads (FPKM) of each gene was calculated based on the length of the gene and reads count mapped to this gene.

[0116] Prior to differential gene expression (DEG) analysis, for each sequenced library, the read counts were adjusted by edgeR program package through one scaling normalized factor. Differential expression analysis of two conditions was performed using the edgeR package (3.22.5). The p-values were adjusted using the Benjamini & Hochberg method. Corrected P- value of 0.05 and absolute foldchange of 2 were set as the threshold for significant DEGs.

[0117] Gene Ontology

[0022] (GO) enrichment analysis of DEGs was implemented by the clusterProfiler R package, in which gene length bias was corrected. GO terms with corrected p-values less than 0.05 were considered significantly enriched by DEGs. Kyoto Encyclopedia of Genes and Genomes (KEGG) is a database resource for understanding high-level functions and utilities of the biological system, such as the cell, the organism and the ecosystem, from molecular-level information, especially large-scale molecular datasets generated by genome sequencing and other high-through put experimental technologies (genome.jp / kegg / ). The clusterProfiler R package was used to test the statistical enrichment of differential expression genes in KEGG

[0023] pathways. The Reactome database brings together the various reactions and biological pathways of human model species. Reactome pathways with corrected p- values less than 0.05 were considered significantly enriched by DEGs. clusterProfiler software was used to test the statistical enrichment of DEGs in the Reactome pathway.

[0118] As used herein, "consisting essentially of" allows the inclusion of materials or steps that do not materially affect the basic and novel characteristics of the claim. Any recitation herein of the term "comprising", particularly in a listing of components of a composition or elements of a device, constitutes inclusion of alternative embodiments in which “comprising” is replaced with "consisting essentially of" or "consisting of'.

[0119] While the present invention has been described in conjunction with certain preferred embodiments, one of ordinary skill, after reading the foregoing specification, will be able to effect various changes, substitutions of equivalents, and other alterations to the compositions and methods set forth herein.

[0120] References

[0121] 1. Dahlen, E., N. Veitonmaki, and P. Norlen, Bispecific antibodies in cancer immunotherapy. Ther Adv Vaccines Immunother, 2018. 6(1): p. 3-17.

[0122] 2. Ma, J., et al., Bispecific Antibodies: From Research to Clinical Application. Front Immunol, 2021. 12: p. 626616.

[0123] 3. Ellerman, D., Bispecific T-cell engagers: Towards understanding variables influencing the in vitro potency and tumor selectivity and their modulation to enhance their efficacy and safety. Methods, 2019. 154: p. 102-117.

[0124] 4. Hoffman, L.M. and L. Gore, Blinatumomab, a Bi-Specific Anti-CD19 / CD3 BiTE(®) Antibody for the Treatment of Acute Lymphoblastic Leukemia: Perspectives and Current Pediatric Applications. Front Oncol, 2014. 4: p. 63.

[0125] 5. Segaliny, A. I., et al., A high throughput bispecific antibody discovery pipeline. Commun Biol, 2023. 6(1): p. 380.

[0126] 6. Wudhikarn, K., et al., Outcomes of relapsed B-cell acute lymphoblastic leukemia after seguential treatment with blinatumomab and inotuzumab. Blood Adv, 2022. 6(5): p. 1432- 1443.

[0127] 7. Bai, R., et al., Mechanisms of Cancer Resistance to Immunotherapy. Front Oncol, 2020. 10: p. 1290.

[0128] 8. Abou Dalle, L, et al., Bi- and Tri-specific antibodies in non-Hodgkin lymphoma: current data and perspectives. Blood Cancer J, 2024. 14(1): p. 23.

[0129] 9. Lejeune, M., et al., Bispecific, T-Cell-Recruiting Antibodies in B-Cell Malignancies. Front Immunol, 2020. 11: p. 762. 10. Huseby, E.S. and E. Teixeiro, The perception and response of T cells to a changing environment are based on the law of initial value. Sci Signal, 2022. 15(736): p. eabj9842.

[0130] 11. Koh, C.H., et al., CD8 T-cell subsets: heterogeneity, functions, and therapeutic potential. Exp Mol Med, 2023. 55(11): p. 2287-2299.

[0131] 12. Martens, A.W.J., et al., CD3xCD19 DART molecule treatment induces non-apoptotic killing and is efficient against high-risk chemotherapy and venetoclax-resistant chronic lymphocytic leukemia cells. J Immunother Cancer, 2020. 8(1).

[0132] 13. Tapia-Galisteo, A., L. Alvarez-Vallina, and L. Sanz, Bi- and trispecific immune cell engagers for immunotherapy of hematological malignancies. J Hematol Oncol, 2023. 16(1): p. 83.

[0133] 14. Loftier, A., et al., A recombinant bispecific single-chain antibody, CD19 x CDS, induces rapid and high lymphoma-directed cytotoxicity by unstimulated T lymphocytes. Blood, 2000. 95(6): p. 2098-103.

[0134] 15. Hoffmann, P., et al., Serial killing of tumor cells by cytotoxic T cells redirected with a CD19- / CD3-bispecific single-chain antibody construct. Int J Cancer, 2005. 115(1): p. 98-104.

[0135] 16. Huo, Y., et al., BHnatumomab-induced T cell activation at single cell transcriptome resolution. BMC Genomics, 2021. 22(1): p. 145.

[0136] 17. Berger Fridman, I., et al., High-throughput microfluidic 3D biomimetic model enabling guantitative description of the human breast tumor microenvironment. Acta Biomater, 2021. 132: p. 473-488.

[0137] 18. Sabhachandani, P., et al., Microfluidic assembly of hydrogel-based immunogenic tumor spheroids for evaluation of anticancer therapies and biomarker release. J Control Release, 2019. 295: p. 21-30.

[0138] 19. Zhang, Y., et al., Plasma membrane changes during programmed cell deaths. Cell Res, 2018. 28(1): p. 9-21.

[0139] 20. Mortazavi, A., et al., Mapping and guantifying mammalian transcriptomes by RNA-Seg. Nat Methods, 2008. 5(7): p. 621-8.

[0140] 21. Liao, Y., G.K. Smyth, and W. Shi, featurecounts: an efficient general purpose program for assigning seguence reads to genomic features. Bioinformatics, 2014. 30(7): p. 923-30.

[0141] 22. Young, M.D., et al., Gene ontology analysis for RNA-seg: accounting for selection bias. Genome Biol, 2010. 11(2): p. R14.

[0142] 23. Kanehisa, M. and S. Goto, KEGG: kyoto encyclopedia of genes and genomes. Nucleic Acids Res, 2000. 28(1): p. 27-30.

[0143] 24. Robinson, H.R., et al., A CD19 / CD3 bispecific antibody for effective immunotherapy of chronic lymphocytic leukemia in the ibrutinib era. Blood, 2018. 132(5): p. 521-532.

[0144] 25. Zimmerman, Z., T. Maniar, and D. Nagorsen, Unleashing the clinical power of T cells: CD19 / CD3 bi-specific T cell engager (BITE9) antibody construct blinatumomab as a potential therapy. Int Immunol, 2015. 27(1): p. 31-7.

[0145] 26. Raskov, H., et al., Cytotoxic CD8(+) T cells in cancer and cancer immunotherapy. Br J Cancer, 2021. 124(2): p. 359-367.

[0146] 27. Bouwman, A.C., et al., Intracellular and Extracellular Roles ofGranzyme K. Front Immunol, 2021. 12: p. 677707.

[0147] 28. Cigalotto, L. and D. Martinvalet, Granzymes in health and diseases: the good, the bad and the ugly. Front Immunol, 2024. 15: p. 1371743.

[0148] 29. Trapani, J. A., Granzymes: a family of lymphocyte granule serine proteases. Genome Biol, 2001. 2(12): p. REVIEWS3014.

[0149] 30. Bovenschen, N., et al., NK cell protease granzyme M targets alpha-tubulin and disorganizes the microtubule network. J Immunol, 2008. 180(12): p. 8184-91.

[0150] 31. Ewen, C.L., K.P. Kane, and R.C. Bleackley, A guarter century of granzymes. Cell Death Differ, 2012. 19(1): p. 28-35.

[0151] 32. Adams, J.M., Ways of dying: multiple pathways to apoptosis. Genes Dev, 2003. 17(20): p. 2481-95. 33. Chowdhury, D. and J. Lieberman, Death by a thousand cuts: granzyme pathways of programmed cell death. Annu Rev Immunol, 2008. 26: p. 389-420.

[0152] 34. Thommen, D.S. and T.N. Schumacher, T Cell Dysfunction in Cancer. Cancer Cell, 2018. 33(4): p. 547-562.

[0153] 35. Blazkova, J., et al., Correlation Between TIGIT Expression on CD8+ T Cells and Higher Cytotoxic Capacity. J Infect Dis, 2021. 224(9): p. 1599-1604.

[0154] 36. Chauvin, J.M. and H.M. Zarour, TIGIT in cancer immunotherapy. J Immunother Cancer, 2020. 8(2).

[0155] 37. Shah, K., et al., T cell receptor (TCR) signaling in health and disease. Signal Transduct Target Ther, 2021. 6(1): p. 412.

[0156] 38. Raud, B., et al., Fatty acid metabolism in CD8(+) T cell memory: Challenging current concepts. Immunol Rev, 2018. 283(1): p. 213-231.

[0157] 39. Rangel Rivera, G.O., et al., Fundamentals of T Cell Metabolism and Strategies to Enhance Cancer Immunotherapy. Front Immunol, 2021. 12: p. 645242.

[0158] 40. Zhang, S., et al., Fatty acid metabolism of immune cells: a new target of tumour immunotherapy. Cell Death Discov, 2024. 10(1): p. 39.

[0159] 41. Teixe, T., et al., Syndecan-2 and -4 expressed on activated primary human CD4+ lymphocytes can regulate T cell activation. Mol Immunol, 2008. 45(10): p. 2905-19.

[0160] 42. Bertoni, A., et al., Integrins in T Cell Physiology. Int J Mol Sci, 2018. 19(2).

[0161] 43. Kuczek, D.E., et al., Collagen density regulates the activity of tumor-infiltrating T cells. J Immunother Cancer, 2019. 7(1): p. 68.

[0162] 44. Rocha-Perugini, V. and J.M. Gonzalez-Granado, Nuclear envelope lamin-A as a coordinator of T cell activation. Nucleus, 2014. 5(5): p. 396-401.

[0163] 45. Thurm, C., B. Schraven, and S. Kahlfuss, ABC Transporters in T Cell-Mediated Physiological and Pathological Immune Responses. Int J Mol Sci, 2021. 22(17).

[0164] 46. McIlroy, A., et al., Histamine and prostaglandin E up-regulate the production of Th2- attracting chemokines (CCL17 and CCL22) and down-regulate IFN-gamma-induced CXCL10 production by immature human dendritic cells. Immunology, 2006. 117(4): p. 507-16.

[0165] 47. Ekkens, M.J., et al., Thl and Th2 cells help CD8 T-cell responses. Infect Immun, 2007. 75(5): p. 2291-6.

[0166] 48. Uhl, L.F.K., Modes of Communication between T Cells and Relevance for Immune Responses. Int J Mol Sci., 2020. 21(8).

[0167] 49. Sarkar, S., et al., T Cell Dynamic Activation and Functional Analysis in Nanoliter Droplet Microarray. J Clin Cell Immunol, 2015. 6(3).

[0168] 50. Sarkar, S., et al., Dynamic Analysis of Human Natural Killer Cell Response at Single-Cell Resolution in B-Cell Non-Hodgkin Lymphoma. Front Immunol, 2017. 8: p. 1736.

[0169] 51. Sarkar, S., et al., Dynamic analysis of immune and cancer cell interactions at single cell level in microfluidic droplets. Biomicrofluidics, 2016. 10(5): p. 054115.

[0170] 52. Dustin, M.L., T-cell activation through immunological synapses and kinapses. Immunol Rev, 2008. 221: p. 77-89.

[0171] 53. Liu, C., et al., Population dynamics of immunological synapse formation induced by bispecific T cell engagers predict clinical pharmacodynamics and treatment resistance. Elife, 2023. 12.

[0172] 54. Dalia Pieta, A., et al., Innovative therapeutic strategy for B-cell malignancies that combines obinutuzumab and cytokine-induced killer cells. J Immunother Cancer, 2021. 9(7).

[0173] 55. Feucht, J., et al., T-cell responses against CD19+ pediatric acute lymphoblastic leukemia mediated by bispecific T-cell engager (BITE) are regulated contrarily by PD-L1 and CD80 / CD86 on leukemic blasts. Oncotarget, 2016. 7(47): p. 76902-76919.

[0174] 56. Sullivan, M.R., et al., An innovative single-cell approach for phenotyping and functional genotyping of CAR NK cells. J Immunother Cancer, 2024. 12(5).

[0175] 57. Prager, I., et al., NK cells switch from granzyme B to death receptor-mediated cytotoxicity during serial killing. J Exp Med, 2019. 216(9): p. 2113-2127. 58. Singh, A., S. Dees, and I.S. Grewal, Overcoming the challenges associated with CD3+ T-cell redirection in cancer. Br J Cancer, 2021. 124(6): p. 1037-1048.

[0176] 59. Runcie, K., et al., Bi-specific and tri-specific antibodies- the next big thing in solid tumor therapeutics. Mol Med, 2018. 24(1): p. 50. 60. Simao, D.C., et al., Bispecific T-Cell Engagers Therapies in Solid Tumors: Focusing on Prostate

[0177] Cancer. Cancers (Basel), 2023. 15(5).

[0178] 61. Aldoss, I., et al., Correlates of resistance and relapse during blinatumomab therapy for relapsed / refractory acute lymphoblastic leukemia. Am J Hematol, 2017. 92(9): p. 858-865.

[0179] 62. Duell, J., et al., Freguency of regulatory T cells determines the outcome of the T-cell-engaging antibody blinatumomab in patients with B-precursor ALL. Leukemia, 2017. 31(10): p. 2181-

[0180] 2190.

[0181] 63. Wong, R., et al., Blinatumomab induces autologous T-cell killing of chronic lymphocytic leukemia cells. Haematologica, 2013. 98(12): p. 1930-8.

[0182] 64. Philipp, N., et al., T-cell exhaustion induced by continuous bispecific molecule exposure is ameliorated by treatment-free intervals. Blood, 2022. 140(10): p. 1104-1118.

Claims

1. CLAIMSWhat is claimed is1. A method for enriching a population of killer T cells with active killer T cells, the method comprising:(a) providing(i) a microfluidic system configured for co-encapsulating single cells in microdroplets in an oil stream and sorting the microdroplets based on fluorescence within the microdroplets;(ii) a population of single killer T cells, optionally loaded with an intracellular Ca2+-sensitive fluorescent indicator;(iii) a population of single target cells, such as cancer cells or tumor-derived cells, optionally loaded with a fluorescent indicator of caspase activity, wherein either the population of T cells in (ii) or the population of target cells in (iii) contains said fluorescent indicator, but not both; and(iv) a bispecific or multispecific binding molecule having a T cell binding specificity for an antigen on a surface of the killer T cells and a target cell binding specificity for an antigen on a surface of the target cells;(b) co-encapsulating the killer T cells, the target cells, and the bispecific or multispecific binding molecule in microdroplets in an oil stream using the microfluidic system;(c) allowing the killer T cells and target cells to bind the bispecific or multispecific binding molecule and to form killer T cell-binding molecule-target cell complexes in the microdroplets, whereby some of the killer T cells become activated by interaction with the target cells and the activated killer T cells kill associated target cells, and whereby a fluorescence signal is produced either from the activated killer T cells or the associated target cells;(d) sorting the microdroplets based on a level of said fluorescence signal, yielding a first group of microdroplets containing activated killer T cells and a second group of microdroplets containing non-activated killer T cells;(e) collecting the activated killer T cells from the first group, thereby producing a population of killer T cells that is enriched in active killer cells compared to the initially provided killer T cells; and(f) optionally collecting the non-activated T cells from the second group.

2. The method of claim 1, wherein target cells co-encapsulated with killer T cells are killed, and the population of enriched activated killer T cells produced in (e) is essentially free of living or intact target cells.

3. The method of claim 1 , wherein the bispecific or multispecific binding molecule is a bispecific or multispecific antibody or aptamer.

4. The method of claim 1, wherein the T cell binding specificity comprises binding to CD3, or wherein the bispecific or multispecific binding molecule is a multispecific binding molecule, and wherein the multispecific binding molecule has T cell binding specificities for CD3 and CD8, or for CD3 and CD4, or for CD3 and a marker for Tregs selected from the group consisting of Foxp3, CD25, GITR, Nrp1, Helios, and CTLA-4.

5. The method of claim 1, wherein the target cell binding specificity is for binding to a tumor cell surface antigen, such as a tumor cell surface antigen selected from the group consisting of CD19, CD20, CD44, CTLA-4, GITR-A, OX-40, CLDN1 , LY6G6D / F, TLR4, GPR56, and SLCO183.

6. The method of claim 1, wherein the fluorescence signal is produced using an intracellular Ca2+-sensitive fluorescent indicator in the killer T cells, or wherein the fluorescence signal is produced using a fluorescent indicator of caspase 3 / 7 activity in the target cells.

7. The method of claim 1, wherein the killer T cells and target cells are co-encapsulated in the microdroplets at a ratio of about 1 : 1 killer cells to target cells, or wherein the killer T cells and target cells are co-encapsulated in the microdroplets at a ratio of less than 1:1 killer cells to target cells, such as wherein the ratio of killer cells to target cells is from about 0.1 to about 0.9, and wherein the activated killer cells collected in (e) are mixed with target cells that survived.

8. The method of claim 1 wherein, prior to said transcriptomic analysis, the enriched population of active killer T cells is expanded by performing one or more rounds of proliferation of the enriched population of active killer T cells in culture.

9. The method of claim 1, further comprising(g) performing transcriptomic analysis of the one or more of the activated killer T cells collected in (e) and / or the nonactivated killer T cells collected in (f);wherein said transcriptomic analysis comprises(i) isolating mRNA from a pool of the enriched population of active killer T cells;(ii) performing RNA sequencing on the isolated mRNA; and(iii) analyzing the sequenced RNA from (b) to determine a level of expression of one or more genes of the sorted and enriched population of active killer T cells.

10. The method of claim 9, wherein step (iii) comprises determining whether one or more biological processes or metabolic pathways are activated or inhibited in the collected active killer T cells compared to nonactivated killer T cells or nonkiller T cells.

11. The method of claim 1 , further comprising genetically modifying one or more of the collected active killer T cells.

12. The method of claim 11, wherein the genetic modification comprises transducing the one or more cells with a viral vector, whereby the genetically modified cells’ expression of one or more genes is increased or decreased.

13. The method of claim 11, wherein the genetic modification is performed prior to expansion of the genetically modified cells.

14. A population of active killer T cells obtained by the method of claim 1.

15. The population of active killer T cells of claim 14, wherein the population comprises cells having one or more genes or pathways upregulated or downregulated compared to nonkiller T cells.

16. A method of immunotherapy of a subject in need of killer T cell supplementation, the method comprising:(a) obtaining a population of enriched and expanded active killer T cells using the method of claim 1;(b) administering at least a portion of the population of enriched and expanded active killer T cells to the subject.

17. The method of claim 16, wherein (a) comprises: obtaining a population of killer T cells from the subject, and using the population of the subject’s killer T cells as the initial population of killer T cells in the method of expansion of active killer T cells.

18. The method of claim 16, wherein (a) comprises: obtaining a population of single target cells from the subject, and using the population of the subject’s target cells as the target cells in the method of enriching and expanding killer T cells.

19. The method of claim 16, wherein the population of isolated, enriched, and / or active killer T cells is accepted or rejected for administration to the subject in step (b) based on results of the transcriptomic analysis.

20. The method of claim 16, wherein transcriptomic analysis of the one or more of the collected activated killer T cells and / or the collected nonactivated killer T cells is performed, and wherein, based on results of the transcriptomic analysis, one or more cells of the population of isolated, enriched, and / or active killer T cells is genetically modified to alter their gene expression prior to administration to the subject in step (b) of the genetically modified cells.

21. The method of claim 16, wherein the subject has cancer, and the method is used for cancer therapy of the subject.

22. A kit comprising a microfluidic device and instructions for performing the method of claim 1.

23. The kit of claim 22 further comprising one or more reagents, such as a bispecific or multispecific antibody or aptamer having at least a first binding specificity for a T cell and a second binding specificity for a target cell, a reagent for detecting caspase 3 / 7 activity by fluorescence, a reagent for detecting changes in intracellular Ca2+concentration by fluorescence, one or more reagents for killer T cell expansion such as IL15 and / or IL7, one or more reagents for transcriptomics analysis, a viral vector for transduction of active killer T cells, or a reagent for specifically detecting a killer T cell biomarker.

Images