EP4649134A1

EP4649134A1 - In vitro platform for cell sorting

Info

Publication number: EP4649134A1
Application number: EP24741975.7A
Authority: EP
Inventors: Susan Napier THOMAS; Camila Camargo
Original assignee: Georgia Tech Research Institute; Georgia Tech Research Corp
Current assignee: Georgia Tech Research Institute; Georgia Tech Research Corp
Priority date: 2023-01-10
Filing date: 2024-01-10
Publication date: 2025-11-19
Also published as: WO2024151779A1

Abstract

Disclosed herein is a system for sorting cells, comprising: a substrate, comprising: an input; an output; an at least partially functionalized channel between the input and the output, wherein the functionalized channel is sized to replicate a physiological vasculature system; and a pump configured to flow a sample of cells through the channel at a flow rate that replicates flow through a physiological vasculature system, thereby inducing a wall shear stress on the sample of cells that replicates wall shear stress in a physiological vasculature system; wherein the sample of cells are sorted according to their affinity for the functionalized channel. Also disclosed is methods of sorting cells and methods of adoptive cell therapy.

Description

IN VITRO PLATFORM FOR CELL SORTING

CROSS REFERENCE TO RELATED APPLICATIONS

[0001] This application claims the benefit of priority to U.S. Provisional Application No. 63/438,133, filed January 10, 2023, which is incorporated by reference herein in its entirety.

GOVERNMENT SUPPORT CLAUSE

[0002] This invention was made with government support under Grant Nos. GR10005462 and R01CA207619 awarded by the National Institutes of Health. The Government has certain rights in the invention.

BACKGROUND

[0003] Tissue infiltration by circulating leukocytes, a key checkpoint in inflammation, immunity, and immune tolerance, as well as other adoptively transferred therapeutic cells occurs via adhesive interactions of circulating cells with the local vascular microenvironment. Aberrant or absent granulocyte, monocyte, or lymphocyte homing queues are associated with chronic infection and inflammation, as well as both autoimmunity and tumor outgrowth respectively, underscoring the tight orchestration of innate and adaptive immune response in pathology progression or resolution. Likewise, trafficking of therapeutic cells including but not limited to CD8⁺ T lymphocytes to sites of inflammation regulates the potency of immunity elicited in a variety of pathologies or therapeutic effects of the adoptive cell therapy. This is because a subset of CD8⁺ T cells can clear virally infected cells, making them key players in various infectious diseases such as malaria, influenza, and hepatitis. CD8⁺ T cells also contribute to control of malignant disease, with a high density of CD8⁺ T cells within tumors correlating with reductions in disease burden and improved survival. Trafficking of autoreactive CD8⁺ T cells into inflamed tissues also plays negative regulatory roles in autoimmune diseases such as type 1 diabetes, multiple sclerosis, and rheumatoid arthritis. How and what subpopulations of CD8⁺ T cells interact with an inflamed vessel-like environment is thus critical to better understand CD8⁺ T cell localization to sites of inflammation and regulation of disease-appropriate immunity.

[0004] As with leukocytes generally, CD8⁺ T cells traffic to inflamed tissues in a highly dynamic process of receptor-ligand interactions, including selectin-mediated rolling adhesion, chemokine-triggered integrin activation, and integrin-mediated firm cell adhesion leading to transmigration through the vessel wall. As the kick-off step in this cascade, selectin-mediated rolling adhesion decelerates cells relative to blood flow, thus enabling cell sensing of locally presented chemokines as well as engagement with endothelial expressed integrins. Selectin- mediated slow rolling adhesion phenotypes exhibited in vitro by specific leukocyte and metastatic cancer cell populations have also shown similar slow rolling adhesion behavior in vivo, resulting in increased cell extravasation. Selectin-mediated interactions have also been correlatively or directly implicated in homing of T cells into inflamed skin, melanomas, and other tumor types. The formation of these rolling adhesion behaviors have been previously attributed to the spatial clustering of selectin ligands on tethers and slings of the cell microvilli under shear flow. Moreover, the step-wise peeling of neutrophils from the P-selectin substrate has been attributed to the failure of P-selectin glycoprotein ligand 1 patches under hydrodynamic forces.

[0005] How adhesive quality, indicated by rolling velocity, relates to the broad spectrum of qualities that CD8⁺ T cells exhibit, ranging in both differentiation and activation state and differing substantially in their functions and renewal capacities, has not been systematically compared. This is in large part due to the lack of appropriate platforms to screen cellular phenotypes associated with adhesive qualities. Engineered in vitro systems that mimic the microenvironment of the in vivo circulatory system permits the examination of cell adhesion to a receptor-functionalized planar substrate via video-microscopy-based tracking techniques. However, while not only time-consuming, such image-based tracking metrics yield measured properties that are not retained by cells off-chip, stymieing multidimensional analyses in conjunction with other single cell analysis techniques.

[0006] Therefore, there exists a need for a methodology to sort and label cells for their quality of adhesion in physiologically relevant flow environments, such that the relationships between rolling adhesion behavior and cellular phenotypes of CD8⁺ T cells can be assayed.

SUMMARY

[0007] In an aspect, provided is a system for sorting cells, comprising: a substrate, comprising: an input; an output; an at least partially functionalized channel between the input and the output, wherein the functionalized channel is sized to replicate a physiological vasculature system; and a pump configured to flow a sample of cells through the channel at a flow rate that replicates flow through a physiological vasculature system, thereby inducing a wall shear stress on the sample of cells that replicates wall shear stress in a physiological vasculature system; wherein the sample of cells are sorted according to their affinity for the functionalized channel. [0008] In another aspect, provided is a method of sorting cells, comprising: (a) providing a substrate comprising an at least partially functionalized channel, wherein the functionalized channel is sized to replicate a physiological vasculature system; (b) flowing a sample of cells through the at least partially functionalized channel at a flow rate that replicates flow through a physiological vasculature system, thereby inducing a wall shear stress on the sample of cells that replicates wall shear stress in a physiological vasculature system; and (c) selecting the cells that show high affinity for the functionalized channel.

[0009] In another aspect, provided is a method of adoptive cell therapy, the method comprising: (a) providing a substrate comprising an at least partially functionalized channel, wherein the functionalized channel is sized to replicate a diseased tissue or tumor vasculature system; (b) flowing a sample of cells through the at least partially functionalized channel at a flow rate that replicates flow through a diseased tissue or tumor vasculature system, thereby inducing a wall shear stress on the sample of cells that replicates wall shear stress in a diseased tissue or tumor vasculature system; (c) collecting the cells that show high affinity for the functionalized channel; and (d) administering the collected cells to a patient with a diseased tissue or tumor; wherein the functionalized channel comprises a ligand; and wherein the collected cells have a greater ability to home to and engraft with the diseased tissue or tumor compared to the cells that show low affinity for the functionalized channel.

[0010] Other systems, methods, features and/or advantages will be or may become apparent to one with skill in the art upon examination of the following drawings and detailed description. It is intended that all such additional systems, methods, features and/or advantages be included within this description and be protected by the accompanying claims.

BRIEF DESCRIPTION OF DRAWINGS

[0011] FIGURES 1A-1C depict CD8⁺ T-cells exhibiting more differentiated phenotypes are enriched within the TME compared to lymphoid tissues in day 7 B 16F 10 melanoma bearing mice. FIG. 1A shows the percent of endogenous CD8⁺ T-cells of various subtypes (Naive, CD44 CD62L⁺SCA-F; stem cell memory [SCM], CD44 CD62L⁺SCA-1⁺; central memory [CM], CD44⁺CD62L⁺; and effector cells [EFF], CD44⁺CD62L ) in the spleens or tumors of day 7 B16F10 tumor-bearing animals. FIG. IB shows the frequency of CD8⁺ T cell subtypes isolated from spleens and left untreated or treated with PMA/Ion prior to transfer. FIG. 1C shows the percent of CD8⁺ T-cell subtype of CD45.1⁺ cells recovered from spleens and tumors of day 7 B16F10 melanoma bearing animals 16 h post transfer of 106 CD45.1⁺ CD8⁺ T cells. Points represent individual animals and data represent the mean ± s.e.m. Statistics performed by two-way ANOVA with Bonferroni’s multiple comparisons test. **p<0.01, ****p<0.0001. [0012] FIGURES 2A-2G depict endogenous and donor CD8⁺ T cells recovered from the TME are enriched for P-selectin ligand expression compared with lymphoid tissues. FIGS. 2A-2B show immunohistochemistry staining for P-selectin and CD31 (FIG. 2A) or CD3 (FIG. 2B) in 8-mm-thick sections of B16F10 tumors formed in C57BL/6 mice. Scale bars: top, 400 mm; bottom, 50 mm. FIG. 2C shows representative flow cytometry data scatterplots for P- selectin ligand⁺ expression by CD8⁺ T cells recovered from the spleen, tumor, and naive skin. FIG. 2D shows the percentage of CD8⁺ T cells recovered from the LNs, spleen, tumor, or naive skin expressing P-selectin ligand. Statistical comparisons by one-way ANOVA with Dunnett’s multiple comparisons test. FIG. 2E shows the total number of P-selectin ligand-expressing (P- sel L⁺) cells recovered from tumors or naive skin. Statistical comparisons by two-tailed parametric t test. FIG. 2F shows representative flow cytometry scatterplots for P-selectin ligand expression by CD45.1⁺CD8⁺ T cells recovered from the spleen, tumor, and naive skin 16 h post-adoptive transfer of 106 untreated or PMA/Ion-treated CD45.1⁺CD8⁺ T cells into melanoma-bearing mice. FIG. 2G shows the percentage of transferred CD45.1⁺CD8⁺ T cells recovered from various tissues 16 h post-transfer that express P-selectin ligand. Statistical comparisons by two-way ANOVA with Dunnett’s multiple comparisons test. Points represent data from an individual animal. Data in all panels represent the mean ± SEM; *p < 0.05, ***p < 0.001, ****p < 0.0001.

[0013] FIGURES 3A-3B depict that tumor vasculature exhibits more P-selectin expression compared to naive skin, specifically immunohistochemistry staining for P-selectin (red) and CD31 (green) in 8 pm thick sections of day 7 B16F10 tumors (FIG. 3A) formed in C57B1/6 mice or naive skin (FIG. 3B) from C57B1/6 mice. Scale bar: 50 pm.

[0014] FIGURES 4A-4J depict the characterization of CD8⁺ T-cells prior to transfer. FIGS. 4A-4B shows the purity (FIG. 4A) and viability (FI. 4B) of cells after CD8⁺ T-cell negative isolation from murine C57BL/6 spleens. FIG. 4C shows the viability of CD8⁺ T cells after PMA/Ion treatment. Statistics analyzed by two-tailed parametric t-test. FIG. 4D shows representative flow cytometry plots of subtypes of CD8⁺ T-cells of untreated or PMA/Ion treated CD8⁺ T cells. Naive, CD44 CD62L⁺SCA-F; stem cell memory [SCM], CD44' CD62L⁺SCA-1⁺; CM, CD44⁺CD62L⁺; and EFF, CD44⁺CD62L’. FIG. 4E shows histograms of adhesion molecule expression by untreated or PMA/Ion treated CD8⁺ T-cells. FIGS. 4F-4G show data from FIG. 4E represented as frequency of CD8⁺ T-cells (FIG. 4F) and normalized to untreated cell mean fluorescence (FIG. 4G) FIGS. 4H-4J show frequency of various subtypes of CD8⁺ T-cells of P-selectin ligand +/- within CD8⁺ CD3⁺ untreated (FIG. 4H) and PMA/Ion populations (FIG. 4J). Points represent individual animals and data reflect the mean ± s.e.m. In FIGS. 4F-4J, statistical comparisons were performed by two-way ANOVA with Bonferroni’s multiple comparisons test. * p<0.05, **p<0.01,. *** p<0.001, ****p<0.0001.

[0015] FIGURES 5A-5G depict a flow cytometry gating strategy and biodistribution analysis of untreated or PMA/Ion CD8⁺ T-cells adoptively transferred into B16F10 tumor bearing animals. FIG. 5A shows a flow cytometry gating strategy for donor CD45.1⁺ CD8⁺ T- cells adoptively transferred into recipient CD45.2⁺ animals bearing day 7 B16F10 melanomas. FIG. 5B shows representative flow cytometry scatter plots of adoptively transferred CD8⁺ T- cells (CD45.1⁺CD45.2‘) recovered from various tissues 16h post transfer. FIG. 5C shows the frequency of donor cells recovered of total CD8⁺ T-cells (donor and recipient) in each tissue. FIG. 5D shows the frequency of recovered donor cells in each tissue of total adoptively transferred cells. FIG. 5E shows the viability of trafficked CD8⁺ T-cells in each analyzed tissue. FIGS. 5F-5G show the distribution of untreated (FIG. 5F) and PMA/Ion (FIG. 5G) treated CD8⁺ T-cells recovered from each analyzed tissue. Points represent individual animals and data represent mean ± s.e.m. Statistics performed by two-way ANOVA with Bonferroni’s multiple comparisons test. **p<0.01, ****p<0.0001.

[0016] FIGURES 6A-6K depict that adhesion of CD8⁺ T cells to P-selectin-functionalized substrates is not correlated to P-selectin ligand expression under conditions of continuous flow at physiological levels of wall shear stress. FIGS. 6A-6B show schematics outlining perfusion conditions for adhesion experiments conducted under static (FIG. 6A) and continuous flow (FIG. 6B) conditions. FIGS. 6C-6D show numbers of untreated (FIG. 6C) or PMA/Ion-treated (FIG. 6D) CD8⁺ T cells per field view interacting with ICAM, P-selectin, or P-selectin + ICAM functionalized substrate under static conditions. FIG. 6E shows differences in total adhesion of perfused CD8⁺ T cells to P-selectin + ICAM versus P-selectin alone under static conditions. FIGS. 6F-6G show numbers of untreated (FIG. 6F) or PMA/Ion-treated (FIG. 6G) CD8⁺ T cells per field view interacting with an ICAM, P-selectin, or P-selectin + ICAM functionalized substrate under continuous flow conditions. FIG. 6H shows differences in total adhesion of perfused CD8⁺ T cells to P-selectin + ICAM versus P-selectin alone under conditions of continuous flow. FIGS. 6I-6J show a correlation analysis between number of adherent cells to P-selectin and percentage of P-selectin ligand⁺ cells under static (FIG. 61) or continuous flow (FIG. 6J). FIG. 6K shows the p values for the correlation analysis between cells adhering to P-selectin under static conditions or continuous flow and the frequency of CD8⁺ T cells expressing P-selectin ligand; dashed line represents p = 0.05. Data in all panels represent the mean ± SEM of three independently performed experiments. Statistical comparisons performed by two-way ANOVA with Tukey’s multiple comparisons test; *p < 0.05, **p < 0.01. ***p < 0 001, ****p < 0.0001.

[0017] FIGURES 7A-7D depict that untreated and PMA/Ion activated CD8⁺ T-cells adhere to P-selectin, but only activated CD8⁺ T-cells adhere to ICAM. FIGS. 7A-7B show the number of untreated (FIG. 7A) or PMA/Ion treated (FIG. 7B) CD8⁺ T cells per field view interacting with an ICAM, P-selectin, or P-selectin + ICAM functionalized substrate under static conditions when deadhesion is initiated at the indicated levels of wall shear stress (WSS). FIGS. 7C-7D show the number of untreated (FIG. 7C) or PMA/Ion treated (FIG. 7D) CD8⁺ T cells per field view interacting with an ICAM, P-selectin, or P-selectin + ICAM functionalized substrate under continuous flow at the indicated levels of WSS at 0.5xl0⁶ cells/ml. Data represent mean ± s.e.m for three or more independently run experiments. Statistical comparisons performed by two-way ANOVA with Tukey’s multiple comparisons test. * p<0.05, **p<0.01. *** p<0.001, ****p<0.0001.

[0018] FIGURES 8A-8E depict that PMA/Ion treated CD8⁺ T-cells exhibit lower velocities of rolling adhesion to P-selectin with or without ICAM compared to untreated CD8⁺ T cells. FIGS. 8A-8D show histograms of adhesion velocity of cells interacting with substrates functionalized with P-selectin alone or in combination with ICAM at varying WSS under conditions of continuous flow. FIG. 8E shows average velocities of interacting cells calculated from FIGS. 8-8D. Data represents mean ± s.e.m. for three independently run experiments. Statistical comparisons performed by two-way ANOVA with Dunnett’s multiple comparisons test. * p<0.05, ****p<0.0001.

[0019] FIGURES 9A-9Q depict perfusion of CD8⁺ T cells through an adhesion chromatography channel functionalized with P-selectin under physiological conditions of fluid flow. FIG. 9A shows a schematic diagram of hemodynamic microenvironment-mimicking adhesion chromatography microfluidic system. FIG. 9B shows a schematic diagram of elution times of CD8⁺ T cells that do (adherent) or do not (free flow) exhibit adhesion to P-selectin. FIG> 9C shows the number of interacting CD8⁺ T cells per field of view left untreated or pretreated with P-selectin chimera and perfused through the chromatography channel. FIG. 9D shows the percentage of recovered untreated or PMA/Ion-treated CD8⁺ T cells in either the free flow (FF) or the adherent (Adh) fraction of the parent (unsorted) population or reperfused cell subpopulations immediately after perfusion. FIGS. 9E-9F show a gating strategy for fractionated CD8⁺ T cell subtypes through a P-selectin-functionalized channel. FIGS. 9G-9H show data from FIGS. 9E-9F, represented as frequency of CD8⁺ T cells. FIGS. 9I-9J show representative flow cytometry histograms of different adhesion markers’ expression on parent, FF, and Adh populations. FIGS. 9K-9L show data from FIGS. 9I-9J, represented as frequency of CD8⁺ T cells. FIG. 9M shows representative flow cytometry plots of CCR7 versus P- selectin ligand expression by free flow or adherent recovered subpopulations from perfused CD8⁺ T cells left untreated or pretreated with PMA/Ion. FIGS. 9N-9O show data from FIG. 9M represented as frequency of CD8⁺ T cells. FIGS. 9P-9Q show the frequency of subtypes of P-sel L⁺CCR7⁺CD8⁺ T cells recovered in the different fractions. Points represent results using splenocytes harvested from an individual animal. Data in all panels represent the mean ± SEM. Results represent a minimum of three independently performed experiments. Statistical comparisons performed by two-way ANOVA with Dunnett’s multiple comparisons test; *p < 0.05, **p < 0.01, ***p < 0.001, ****p < 0.0001.

[0020] FIGURES 10A-10E depict the validation of the engineered adhesion chromatography microfluidic system to investigate CD8⁺ T-cell adhesion under physiological levels of fluid flow in vitro. FIG. 10A shows a schematic outlining the top view of the channel of the adhesion chromatography microfluidic system. FIG. 10B shows the distance of individual perfused CD8⁺ T-cells from the inferior substrate of an unfunctionalized channel calculated based on measured individual cell velocity and size. FIG. 10C shows the number of interacting cells along the length of the channel during the sorting phase. Statistics for were performed by one-way ANOVA with Dunnett’s multiple comparisons test. FIGS. 10D-10E show purity (FIG. 10D) and viability (FIG. 10E) of unperfused cells and cells perfused through either an unfunctionalized (Unfxn) channel or 25 ug/ml P-selectin functionalized fractionated into free flow (FF) and adherent (Adh) fractions at 0.5 dyn/cm². Statistics for FIGS. 10D-10E were performed by two-way ANOVA with Tukey’s multiple comparison test. In FIGS. 10D-10E, each data point reflects results using splenocytes harvested from an individual animal. Data represents mean ± s.e.m. Results represent a minimum of three independently performed experiments.

[0021] FIGURES 11A-11H depict that stimulation with P-selectin in solution or through a functionalized channel in flow does not activate CD8⁺ T-cells or increase their expression of CCR7. FIGS. 11A-11D show the effects of CD8⁺ T-cell co-incubation with P-selectin chimera at varying concentrations for 1 (FIG. 11 A, FIG. 11C) or 4 (FIG. 11B, FIG. 11D) hours on adhesion marker expression (FIGS. 11A-11B) and differentiation state (FIGS. 11C-11). Mean ± s.e.m, p value from two-way ANOVA with Dunnett’ s multiple comparisons test. FIGS. 11E- 11H show the percent P-selectin ligand⁺ (FIGS. 11E-11F) and CCR7⁺ (FIGS. 11G-11H) cells of untreated (FIG. HE, FIG. 11G) or PMA/Ion pre-treated (FIG. HF, FIG. HH) CD8⁺ T- cells that were unperfused, perfused through an unfunctionalized (Unfxn) channel, or and perfused through a 25 ug/ml P-selectin functionalized channel at 0.5 dyn/cm² and sorted into FF and Adh fractions. Mean ± s.e.m, p value from one-way ANOVA with Tukey’s multiple comparisons test; * p<0.05, **p<0.01, *** p<0.001, ****p<0.0001.

[0022] FIGURES 12A-12F depict expansion of CD8⁺ T-cells enriched for adhesion to P- selectin in flow using adhesion chromatography microfluidic system. FIG. 12A shows the fold expansion of fractionated CD8⁺ T-cells cultured with Dynabeads with 100 Ul/ml IL-2. FIG. 12B shows the frequency of Ki-67⁺ of live CD8⁺ T cells throughout expansion. FIG. 12C shows the viability of fractionated CD8⁺ T cells throughout expansion. FIGS. 12D-12F show the frequency of P-sel L (FIG. 12D), granzyme B (FIG. 12E), and PD-1 (FIG. 12F) of live CD8⁺ T-cells throughout expansion. Points represent an individual expansion. Data represents mean ± s.e.m. Results represent a minimum of three independently performed experiments. Statistical comparisons performed by two-way ANOVA with Dunnett’s multiple comparisons test. * p<0.05, **p<0.01.

[0023] FIGURES 13A-13R depict correlation analysis of CD8⁺ T-cells enriched based on in vitro adhesion vs. tissue-specific trafficking. Correlation analysis of adhesion molecule expression (FIGS. 13A-13B, FIGS. 13G-13J) and subtype (FIGS. 13D-13E, FIGS. 13K- 13N) enrichment between CD8⁺ T-cells enriched for adhesion to P-selectin in flow and that traffic to the spleen (FIGS. 13A-13B, FIGS. 13D-13E), NDLN (FIGS. 13G-13H, FIGS. 13K- 13L) and TDLN (FIGS. 131-13 J, FIGS. 13M-13N) is shown 16 h post transfer. -logio(p- value) versus log2(fold change) of adhesion molecule expression (FIG. 13C, FIGS. 13O-13P) and subtype (FIG. 13F, FIGS. 13Q-13R) by CD8⁺ T cells recovered from the spleen, NDLN, and TDLN is also shown. In FIGS. 13A-13B, FIGS. 13D-13E, and FIGS. 13G-13N, each point represents results from an individual animal.

[0024] FIGURES 14A-14G depict CD8⁺ T cells recovered using the adhesion chromatography system to enrich for adhesion to P-selectin in flow exhibit enhanced tumor homing compared with cells recovered in the free flow and parent fractions. FIG. 14A shows a schematic diagram outlining the experimental design. FIGS. 14B-14C show the percentage of parent untreated and PMA/Ion CD8⁺ T cells of cells fractionated based on their adhesion to P-selectin in flow recovered in various tissues 16 h after adoptive-transfer into B16F10 melanoma-bearing animals. FIGS. 14D-14E show the frequency of single and co-expression of P-selectin ligand and CCR7 by donor CD8⁺ T cells recovered in the tumor and the spleen that were left untreated or were PMA/Ion treated. FIGS. 14F-14G show subtype distribution of donor P-sel L⁺CCR7⁺ untreated (FIG. 14F) or PMA/Ion-treated (FIG. 14G) CD8⁺ T cells recovered from the tumor. Points represent individual results from individual animal. Data in all panels represent the mean ± SEM of three or more independently run experiments; two-way ANOVA with Tukey’s multiple comparisons test; *p < 0.05, **p < 0.01, ****p < 0.0001.

[0025] FIGURES 15A-15F depict CD8⁺ T-cell adherent fraction has enhanced tumor homing and reduced accumulation within lymphoid tissues compared to free flow fraction and unsorted population independent of donor cell pre-treatment. FIGS. 15A-15B show the frequency of donor CD8⁺ T-cells of all CD8⁺ T-cells recovered from various tissues. FIG. 15C shows the frequency of adoptively transferred untreated and PMA/Ion CD8⁺ T-cells recovered from the tumor of various subtypes. FIG. 15D shows the frequency of adoptively transferred untreated and PMA/Ion CD8⁺ T-cells recovered from the spleen of various subtypes. FIGS. 15E-15F show the frequency of single- and co-expression of P-selectin ligand and CCR7 by donor CD8⁺ T-cells that were untreated or pretreated with PMA/Ion recovered from NDLN and TDLN 16 h post transfer. Each point represents one individual animal. Data represents mean ± s.e.m of 3 or more independently run experiments; two-way ANOVA with Tukey’s multiple comparisons test; * p<0.05, **p<0.01, ****p<0.0001.

[0026] FIGURES 16A-16M depict the enrichment of P-selectin ligand and CCR7 expression by Dynabead-expanded CD8⁺ T cells recovered by perfusion through a P-selectin- functionalized channel is correlated with the enrichment of donor cells trafficking to the TME. FIG. 16A shows the fold expansion of CD8⁺ T cells cultured with Dynabeads and IL-2. FIG. 16B shows the frequency of P-selectin ligand-expressing CD8⁺ T cells at different days of expansion. FIG. 16C shows the number and velocity of Dynabead-expanded CD8⁺ T cells adherent to a P-selectin-functionalized substrate. FIG. 16D shows the percentage of Dynabead- expanded CD8⁺ T cells recovered within the Adh and FF fractions by perfusion through the adhesion chromatography system. FIG. 16E shows a schematic outlining adoptive transfer experimental design. FIG. 16F shows the frequency of donor CD8⁺ T cells (CD45.1⁺CD45.2‘ ) of all CD45⁺ (CD45.1⁺CD45.2⁺) cells into CD45.2 animals bearing B16F 10 melanomas. FIG. 16G shows the frequency of total CD8⁺ donor T cells recovered from analyzed tissues. FIGS. 16H-16M show the correlation analysis of adhesion molecule expression (FIGS. 16H-16I) and subtype (FIGS. 16K-16L) enrichment between CD8⁺ T cells enriched for adhesion to P- selectin in flow and that traffic to the tumors. FIG. 16J and FIG. 16M show -logio(p value) versus log2(fold change) of adhesion molecule expression (FIG. 16J) and subtype (FIG. 16M) by CD8⁺ T cells recovered from the tumor. In FIGS. 16F-16I and FIGS. 16K-16L, points represent results from one individual animal. In FIGS. 16A-16D and FIGS. 16F-16G, data represent the mean ± SEM. All data represent results from three or more independently run experiments; one-way ANOVA (FIG. 16B) and two-way ANOVA (FIGS. 16C-16D, FIGS. 16F-16G) with Dunnett’s multiple comparisons test; *p < 0.05, **p < 0.01, ***p < 0.001, ****p < 0.0001.

[0027] FIGURES 17A-17H depict CD8⁺ T-cells at different days of Dynabead and IL-2 expansion exhibit differential adhesive behaviors to P-selectin in physiological flow. FIG. 17A shows a schematic outlining perfusion workflow for static condition. FIGS. 17B-17D show the percent of rolling adhesion (FIG. 17B), firm adhesion (FIG. 17C), and total adherent cells (FIG. 17D) per field of view on P-selectin functionalized flow chamber under static perfusion conditions at various levels of wall shear stress at 0.5xl0⁶ cells/ml. FIG. 17E shows a schematic outlining perfusion workflow for continuous flow. FIGS. 17F-17H show the number of rolling (FIG. 17F), firm (FIG. 17G), and total (FIG. 17H) interacting CD8⁺ T-cells per field view on P-selectin substrate under continuous flow condition at various levels of wall shear stress at 0.5xl0⁶ cells/ml. Data represents mean ± s.e.m of 3 or more independently run experiments; two-way ANOVA with Tukey’s multiple comparisons test; * p<0.05, **p<0.01, *** p<0.001, ****p<0.0001. n.s., not significant.

[0028] FIGURES 18A-18F depict differential adhesive ligand expression by CD8⁺ T-cells at different days of expansion enriched for adhesion to P-selectin in flow using adhesion chromatography microfluidic system. FIGS. 18A-18F show the frequency of P-selectin ligand (FIG. 18A), CCR7 (FIG. 18B) and L-selectin (FIG. 18C) expressing CD8⁺ T-cells and fraction of naive (FIG. 18D), CM (FIG. 18E) and EFF (FIG. 18F) subtypes recovered from unsorted, free flow (FF), or adherent (Adh) fractions at different days of expansion. Data represents mean ± s.e.m of 3 or more independently run experiments; two-way ANOVA with Tukey’s multiple comparisons test; * p<0.05, **p<0.01. *** p<0.001, ****p<0.0001.

[0029] FIGURES 19A-19H depict that CD8⁺ T-cells exhibit differential in vivo homing capabilities depending on day of expansion. FIG. 19A shows the frequency of donor CD8⁺ T- cells (CD45.1⁺CD45.2‘) of all CD8⁺ T-cells recovered from various tissues 16 h post transfer. FIG. 19B shows the viability of recovered donor cells in each analyzed tissue. FIGS. 19C- 19H show fold change in the fraction of recovered donor (CD45.1⁺) CD8⁺ T-cells expressing CCR7 (FIG. 19C), L-selectin (FIG. 19D), and P-selectin ligand (FIG. 19E) and of a naive (FIG. 19F), CM (FIG. 19G), and EFF (FIG. 19H) subtype within each tissue relative to their prevalence in the donor population pre-transfer. Each point represents results from an individual mouse. Data represents mean ± s.e.m of 3 or more independently run experiments. Statistical comparisons by two-way ANOVA with Tukey’s multiple comparisons test. * p<0.05, **p<0.01, *** p<0.001, ****p<0.0001. [0030] FIGURES 20A-20K depict that CD8⁺ T cells enriched for adhesion to P-selectin in physiological flow improve tumor control by ACT in combination with immune checkpoint blockade (ICB) aPD-1. FIG. 20A shows a schematic outlining the experimental design. FIG. 20B shows the frequency of donor CD8⁺ T cells in all CD8⁺ T cells recovered in the tumor. FIG. 20C shows donor CD8⁺ T cells recovered in the tumor as the frequency of the total number of cells transferred into the B16F10-OVA tumor-bearing mice. FIG. 20D shows the viability of donor CD8⁺ T cells in the tumor. FIGS. 20E-20F show the percentage of cytokineproducing or Ki-67⁺ cells of donor CD8⁺ T cells in the tumor (FIG. 20E) or TDLNs (FIG. 20F) FIGS. 20G-20H show the percentage of cytokine-producing or Ki-67⁺ cells of endogenous CD8⁺ T cells in the tumor (FIG. 20G) or TDLNs (FIG. 20H). FIG. 201 shows a schematic outlining the experimental design. FIG. 20 J shows B16F10-OVA tumor growth after treatment with aPD-1 alone or adoptive transfer (AT) of Adh OT-1 CD8⁺ T cells with and without aPD-1. FIG. 20K shows Bl 6F 10-0 VA tumor growth after therapy with aPD-1 alone or in combination AT of parent or sorted OT-1 CD8⁺ T cells (FF and Adh). All graphs represent mean ± SEM (n = 5); statistics were performed by two-way ANOVA with Tukey’s multiple comparisons test. In FIGS. 20E-20H, * indicates significance of comparison to parent population, $ indicates significance of comparison between time points; *p < 0.05, **p < 0.01. ***p < 0.001, ****p < 0.0001.

[0031] FIGURES 21A-21G depict that OT-I CD8⁺ T-cells recovered in the adherent fraction have reduced trafficking into LNs. FIG. 21A shows the frequency of donor CD8⁺ T- cells recovered in various lymphoid tissues. FIG. 21B shows donor CD8⁺ T-cells recovered in various lymphoid tissues as a frequency of the total number of cells transferred into the B16F10-OVA tumor-bearing mice. FIG. 21C shows the viability of donor CD8⁺ T-cells in various lymphoid tissues. FIGS. 21D-21E show the percent of cytokine-producing, PD-1⁺, Ki- 67 of donor CD8⁺ T-cells in the NDLN (FIG. 21D) or spleen (FIG. 21E). FIGS. 21F-21G show the percent of cytokine-producing, PD-1⁺ or Ki-67 of endogenous CD8⁺ T-cells in the NDLN (FIG. 21F) or spleen (FIG> 21G). Graphs represent mean ± s.e.m, (n=5); statistics were performed by two-way ANOVA with Tukey’s multiple comparison test. In FIGS. 21D- 21G, * indicates significance of comparison to parent population, $ indicates significance of comparison between time points. *p<0.05, **p<0.01. *** p<0.001, ****p<0.0001.

[0032] FIGURES 22A-22E depict that human CD8⁺ T cells exhibit differential adhesion to P-selectin in physiological levels of fluid flow during expansion. FIG. 22A shows expansion of human CD8⁺ T cells from healthy donors cultured with Dynabeads and IL-2. FIG. 22B shows the percentage of CD8⁺ T cells recovered in Adh and FF fractions using P-selectin- functionalized channel. FIG. 22C shows the frequency of P-selectin ligand⁺ cells of live CD8⁺ T cells of parent populations or those recovered in FF or Adh fractions after perfusion through a P-selectin-functionalized channel at different days of expansion. FIG. 22D shows histograms of P-selectin ligand expression at different days of expansion of human CD8⁺ T cells. FIG. 22E shows mean fluorescence intensity of P-selectin ligand⁺ cells normalized to parent cell mean fluorescence. For all graphs the data represent the mean ± SEM of three or more independently run experiments; two-way ANOVA with Bonferroni’s multiple comparisons test; *p < 0.05, **p < 0.01, ****p < 0.0001.

[0033] FIGURES 23A-23F depict that human CD8⁺ T-cells enriched for adhesion to P- selectin in physiological fluid flow exhibit different extents of adhesion molecule expression that varies by day of expansion. Frequency (FIGS. 23A-23C) and fold change in MFI are shown relative to parent population (FIGS. 23D-23F) of CCR7 (FIG. 23A, FIG. 23D), L- selectin (FIG. 23B, FIG. 23E), sLe^ (FIG. 23C, FIG. 23F) adhesion molecules of live CD8⁺ T-cells at different days of expansion in parent or FF or Adh fractions recovered from P-selectin functionalized adhesion chromatography channel. Data represents mean ± s.e.m of 3 or more independently run experiments; two-way ANOVA with Bonferroni’s multiple comparisons test; * p<0.05, **p<0.01. *** p<0.001, ****p<0.0001.

[0034] FIGURES 24A-24H depict an optofluidic adhesion chromatography microfluidic system for in vitro single-cell adhesive profiling of hemodynamic microenvironment-regulated mechanisms of CD8⁺ T cell homing. FIG. 24A shows a microfluidic system schematic. FIG. 24B shows a top view indicating unfunctionalized settling region, light exposure window, and functionalized substrate area of adhesion chromatography channel. FIG. 24C shows a schematic diagram of elution times of cells that do (adherent) or do not (free flow) exhibit adhesion to the functionalized substrate during perfusion and how they are fractionated based on their elution time from the perfusion channel. FIG. 24D shows a schematic of side view of perfused cells inside the hemodynamic mimicking microfluidic device exhibiting varying adhesive qualities, including slow and fast rolling adhesion, and no adhesion. FIG. 24E shows histograms of rolling velocity of CD8⁺ T cells interacting with a 10 pg/ml P-selectin or 10 pg/ml E-selectin functionalized substrates at various WSSs. Data is pooled from four independently run experiments. FIG. 24F shows a schematic of various potential rolling qualities and how they may persistent or vary along the selectin-functionalized substrate length. Cell A is a slow rolling cell that exhibits sustained adhesion along channel length. Cell B exhibits slow rolling that is not persistent, changing between slow rolling and free flow velocities along the selectin-functionalized substrate length. Cell C exhibits faster rolling adhesion that is sustained along the selectin-functionalized substrate length. Upon entering the channel region containing the selectin-functionalized substrate, Cell D is not adherent and travels at the free flow velocity but initiates and sustains rolling adhesion. Cell E does not interact with the adhesive substrate so travels through the selectin-functionalized substrate length at the free flow velocity. FIGS. 24G-24H show a side view of perfusion through a light source illuminated window to photoactivated cells in proportion to their velocity in flow on selectin-functionalized substrates in the exposure window (FIG. 24G), to allow for single-cell fluorescent labeling of cell velocities (FIG. 24G), enabling the comparison of cellular characteristics associated with velocities of rolling cell adhesion (FIG. 24H).

[0035] FIGURES 25A-25F depict that photoactivation of PA-GFP CD8⁺ T cells is spatiotemporally controlled. FIG. 25A shows fluorescent images of PA-GFP⁺ CD8⁺ T cells GFP levels with increasing 405 nm exposure time, at a power output of 208 mW; Scale bar, 20 pm. FIGS. 25B-25C show flow cytometrically measured CD8⁺ T cells unactivated signal and activated signal (GFP) (FIG. 25B) and GFP/unactivated signal ratios (FIG. 25C) per cell and at various times of 405 nm laser exposure at a power output of 208 mW. FIG. 25D shows representative flow cytometry scatter plots of GFP expression with increased exposure time at a power output of 208mW. FIGS. 25E-25F show the percent of GFP⁺ cells (FIG. 25E) and normalized mean fluorescence to unactivated signal (FIG. 25F) across various exposure times and power outputs. Data represents mean ± s.e.m. from three independently run experiments.

[0036] FIGURES 26A-26G depict that adherent CD8⁺ T cells relative to free flow cell fractions exhibit increased extent of photoactivation. FIG. 26A shows a standard curve relating GFP MFI normalized to unactivated signal to cell velocity (calculated based on exposure time and length of exposure window). FIGS. 26B-26C show flow cytometrically measured GFP/unactivated signal of FF or Adh fractions of CD8⁺ T cells recovered from a 10 pg/ml P- or 2.5 pg/ml E-selectin functionalized channel integrated with photoactivation window of 1 cm at a 208mW power output at 0.5 (FIG. 26B) or 1 dyn/cm² (FIG. 26C). FIGS. 26D-26E show data represented in panels FIGS. 26B-26C represented as GFP mean fluorescence normalized to unactivated signal. FIGS. 26F-26G show the distribution of single-cell rolling velocities on P- and E-selectin across different exposure time intervals for 10 pg/ml P-selectin (FIG. 26F) and 2.5 pg/ml E-selectin (FIG. 26G). Shaded purple regions represent the upper experimental runtime limits at a prescribed WSS. Stacked bar-graphs represent the proportion of cells below the runtime limits, within each exposure time interval. Data represents mean ± s.e.m. from three or more independently run experiments. Statistical comparisons performed by one-way ANOVA with Tukey’s multiple comparisons test. * p<0.05, **p<0.01. [0037] FIGURES 27A-27J depict that slow rolling CD8⁺ T cells perfused at higher wall shear stress exhibit increased percentage of selectin ligand⁺ cells. FIG. 27A shows an experiment schematic. FIG. 27B shows a histogram of P-selectin ligand expression of the unperfused population and flow cytometry scatter plot of P-selectin ligand vs. GFP expression of the adherent fraction recovered at 0.5 dyn/cm². FIGS. 27C-27D show the percent of P- selectin ligand⁺ cells (FIG. 27C) and normalized mean fluorescence to unperfused population (FIG. 27D) of PA-GFP⁺ CD8⁺ T cells of unperfused cells, perfused cells through an unfuctionalized channel (UnF), or sorted cells through a 10 ug/ml P-selectin functionalized channel at varying wall shear stress. FIG. 27E shows a histogram of E-selectin ligand expression of the unperfused population and flow cytometry scatter plot of E-selectin ligand vs. GFP expression of the adherent fraction recovered at 0.5 dyn/cm². FIGS. 27F-27G show the percent of E-selectin ligand⁺ cells (FIG. 27F) and normalized mean fluorescence to unperfused population (FIG. 27G) of PA-GFP⁺ CD8⁺ T cells of unperfused, perfused cells through an unfuctionalized channel, or sorted cells through a 2.5 ug/ml E-selectin functionalized channel at varying wall shear stress. FIG. 27H shows GFP⁺ cells divided into two different gates PAI (lower extent of photoactivation) and PA2 (higher extent of photoactivation). FIG. 271 shows P-selectin ligand expression of cells in gate PAI or PA2 of cells perfused through an unfunctionalized channel or sorted through a 10 ug/ml P-selectin functionalized channel. FIG. 27J shows E-selectin ligand expression of cells in gate PAI or PA2 of cells perfused through an unfunctionalized channel of sorted through a 2.5 ug/ml E- selectin functionalized channel. Data represents mean ± s.e.m. from three or more independently run experiments. Statistical comparisons performed by one-way ANOVA (FIGS. 27B-27E) and two-way ANOVA (FIGS. 27G-27H) with Tukey’s multiple comparisons test. * p<0.05, **p<0.01, *** p<0.001, ****p<0.0001.

[0038] FIGURES 28A-28F depict that more differentiated CD8⁺ T cell subtypes adhere to selectin-functionalized substrates. FIGS. 28A-28F show the percent of different CD8⁺ T cell subtypes of GFP⁺ cells perfused over an unfunctionalized channel (UnF) or sorted free flow (FF) and adherent (Adh) cells through a P-selectin (FIGS. 28A-28C) or E-selectin (FIGS. 28D-28F) functionalized channel at varying WSS. Data represents mean ± s.e.m. from four independently run experiments. Statistical comparisons performed by two-way ANOVA with Tukey’s multiple comparisons test. *p<0.05, **p<0.01, *** p<0.001, ****p<0.0001.

[0039] FIGURES 29A-29C depict that enrichment of CD8⁺ T cells varies depending on inflamed-like vasculature microenvironment. FIG. 29A shows PA-GFP CD8⁺ T cells were perfused, photoactivated, and sorted over selectin-functionalized substrates. Sorted cells were stained with fluorescently tagged antibodies, allowing for the comparison of marker expression on cells that were not photoactivated (GFP ), or had low versus high photoactivation (GFP⁺L vs. GFP⁺H). FIGS. 29B-29C show the percent of expression of different adhesion ligand/receptors of PA-GFP⁺ CD8⁺ T cells sorted through a 10 pg/ml P-selectin (FIG. 29B) or 2.5 pg/ml E-selectin (FIG. 29C) at 0.5 or 1 dyn/cm². Data represents mean ± s.e.m. from four independently run experiments. Statistical comparisons performed by two-way ANOVA with Tukey’s multiple comparisons test. * indicates significance comparison between UnF GFP', FF GFP', Adh GFP', Adh GFP⁺L, and Adh GFP⁺H. $ indicates significance of comparison between 0.5 and 1.0 dyn/cm². * p<0.05, **p<0.01, *** p<0.001, ****p<0.0001.

[0040] FIGURES 30A-30I depict that photoactivation reveals distinct relationships between single-cells velocities and adhesion ligand/receptor expression levels for CD8⁺ T cells adhering on P- and E-selectin. FIG. 30A shows PA-GFP CD8⁺ T cells were perfused over 10 pg/ml P-selectin or 2.5 pg/ml E-selectin, photoactivated in a manner proportional to their average velocity, sorted into free flow and adherent fractions, and gated using flow cytometry into unphotoactivated (PA ) and photoactivated (PA⁺), respectively. Cells collected in each fraction were stained with fluorescently tagged antibodies, allowing correlation between ligand expression and photoactivation. FIGS. 30B-30D show the extent of photoactivation (GFP/unactivated signal) of PA' free flow cells and PA⁺ rolling CD8⁺ T cells perfused over P- selectin related to the expression of P-selectin ligand (FIG. 30B), CCR7 (FIG. 30C), and CXCR5 (FIG. 30D). FIGS. 30E-30G show the extent of photoactivation (GFP/unactivated signal) of rolling and free flow CD8⁺ T cells perfused over E-selectin related to the expression of E-selectin ligand (FIG. 30E), CCR7 (FIG. 30F), and CXCR5 (FIG. 30G). FIGS. 30H-30I show P-values for non-zero slopes of the linear fit between the extent of photoactivation and various adhesion ligand and receptors of free flow and rolling cells over P-selectin (FIG. 30H) and E-selectin (FIG. 301). Flow cytometry gate-normalized data, pooled from four independent experiments and plotted with corresponding linear fit. *p<0.05, **p<0.01, *** p<0.001, ****p<0.0001 represent non-zero slopes of the linear fit with each population.

[0041] FIGURES 31A-31C depict selectin concentration and wall shear stress influence the extent of CD8⁺ T cell adhesion. FIG. 31A shows a schematic of parallel plate flow chamber. FIGS. 31B-31C show the number of CD8⁺ total adherent cells per field of view on a P-selectin (FIG. 31B) or E-selectin (FIG. 31C) functionalized channel with varying selectin concentration and wall shear stress (WSS). Statistical comparisons performed by two-way ANOVA with Dunnetf s multiple comparisons test. * p<0.05, **p<0.01, *** p<0.001. [0042] FIGURES 32A-32B depict CD8⁺ T-cell rolling velocities vary between E- and P- selectin at a fixed selectin-concentration. FIG. 32A shows histograms of rolling velocity of CD8⁺ T cells interacting with a 2.5 pg/ml P-selectin or 2.5 pg/ml E-selectin substrate at various WSS. FIG> 32B shows the average free velocity measured for 0.5 dyn/cm² using an unfunctionalized channel. Data represents mean ± s.e.m. from three independently run experiments.

[0043] FIGURES 33A-33B depict the extent of protein adsorption in the chromatography channel varies linearly with selectin-concentration in solution. Standard curves of P-selectin (FIG. 33 A) and E-selectin (FIG. 33B) adsorption per unit area in the adhesion chromatography channel are shown as versus selectin concentration in solution. Goodness of fit represented by coefficient of determination (R²), based on linear regression analysis.

[0044] FIGURES 34A-34C depict cell adhesion on chromatography channel is the same across the length of the channel. FIG. 34A shows a schematic outlining the top view of the channel of the adhesion chromatography microfluidic system. FIGS. 34B-34C show the number of adherent cells on P-selectin (FIG. 34B) and E-selectin (FIG. 34C) across the functionalized portion of the channel with various WSS. Data represents mean ± s.e.m. Results represent a minimum of three independently performed experiments. N.S. indicates a nonsignificant statistical difference.

[0045] FIGURES 35A-35F depict photoconversion of CD8⁺ T cells has no effect of selectin ligand expression or functional adhesion. FIG. 35A shows the coefficient of variation of GFP MFI alone or normalized by unactivated signal at 208mW power output at various times of exposure. FIG. 35B shows percent viability across various exposure times and power outputs. FIGS. 35C-35D show P-selectin (FIG. 35C) and E-selectin (FIG. 35D) ligand expression of PA-GFP⁺ CD8⁺ T cells at various exposure time of 405 nm laser at 208mW. FIGS. 35E-35F show rolling or firm adhesion per field of view of pre-photoconverted (PC) or non-photoconverted (No PC) PA-GFP⁺ CD8⁺ T cells through a P-selectin (FIG. 35D) or E- selectin (FIG. 35F) functionalized channel at 1 dyn/cm2. Data represents mean ± s.e.m. from three independently run experiments. Statistical comparisons performed by one-way ANOVA (FIGS. 35C-35D) and two-way ANOVA (FIG. 35B, FIGS. 35E-F).

[0046] FIGURES 36A-36D depict calculated residence time of adherent CD8⁺ T cells in a 1 cm exposure window. FIGS. 36A-36B show average rolling velocity of CD8⁺ T cells on a 10 pg/ml P- (FIG. 36A) or 2.5 pg/ml E-selectin (FIG. 36B) substrate at various wall shear stress. FIGS. 36C-36D show calculated time cells would spend in an exposure window of 1 cm based on their average rolling velocity on a P-selectin substrate (FIG. 36C) or E-selectin (FIG. 36D) Data represents mean ± s.e.m. from four independently run experiments. Statistical comparisons performed by unpaired t-test,* p<0.05.

[0047] FIGURE 37 depicts a flow cytometry gating strategy for photoactivated CD8⁺ T cells.

[0048] FIGURES 38A-38E depict characteristics of CD8⁺ T cells post-perfusion and photoactivation. FIG. 38A shows representative flow cytometry data scatter plots of GFP expression on unperfused (UnP) and perfused cells through an unfunctionalized channel (UnF), and sorted free flow (FF) and adherent (Adh) cells through a 10 pg/ml P-selectin or 2.5 pg/ml E-selectin functionalized channel at various WSS with the addition of a photoactivation exposure window of 1 cm at 208 mW power output. FIGS. 38B-38C show percent GFP⁺ cells (FIG. 38B) and viability (FIG. 38C) of unperfused cells or cell perfused through an unfunctionalized or P- and E- selectin functionalized channel at 0.5 dyn/cm². FIGS. 38D-38E show percent GFP⁺ cells (FIG. 38D) and viability (FIG. 38E) of unperfused cells or cell perfused through an unfunctionalized or P- and E- selectin functionalized channel at 1 dyn/cm². Data represents mean ± s.e.m. from three or more independently run experiments. Statistical comparisons performed by one-way ANOVA with Tukey’s multiple comparisons test. **p<0.01, ****p<0.0001.

[0049] FIGURES 39A-39D depict that among free flow and adherent cells there is no significant difference in the extent of photoactivation across different CD8⁺ T cell subtypes. FIGS. 39A-39B show flow cytometrically measured CD8⁺ T cells GFP/unactivated signal across different CD8⁺ T cell subtypes in the FF and Adh fractions that were photoactivated through a P-selectin (FIG. 39A) or E-selectin (FIG. 39B) functionalized channel at various WSS. FIGS. 39C-39D show data from FIGS. 39A-39B represented as normalized GFP mean fluorescence to unactivated signal. N.S. represents non-statistical significance.

[0050] FIGURES 40A-40H depict that photoactivation reveals distinct relationships between single-cell velocities and adhesion ligand/receptor expression levels for CD8⁺ T cells adhering on P- and E-selectin. FIGS. 40A-40D show the extent of photoactivation (GFP/unactivated signal) of rolling and free flow CD8⁺ T cells perfused over P-selectin related to the expression of L-selectin ligand (FIG. 40A), CD44 (FIG. 40B), CXCR3 (FIG. 40C), and LFA-1 (FIG. 40D). FIGS. 40E-40H show the extent of photoactivation (GFP/unactivated signal) of rolling and free flow CD8⁺ T cells perfused over E-selectin related to the expression of L-selectin ligand (FIG. 40E), CD44 (FIG. 40F), CXCR3 (FIG. 40G), and LFA-1 (FIG. 40H) Flow cytometry gate-normalized data, pooled from four independent experiments and plotted with corresponding linear fit. * p<0.05, **p<0.01, *** p<0.001, ****p<0.0001 represent non-zero slopes of the linear fit with each population.

[0051] FIGURE 41 depicts a functionalized channel setup and cell pulse loading time.

[0052] FIGURE 42A shows that Wall Shear Stress (WSS) has a minimal effect on the fractional recovery of expanded cells that are perfused on Nepmucin. However, naive cell recovery is wall-shear stress dependent. FIGURE 42B shows that CCR7⁺ enrichment of untreated adherent cells is WSS-dependent. At 1.0 dyn/cm² untreated CD8⁺ T cells that adhere to Nepmucin have increased TCF-1 expression and decreased TIM-3 expression. As such, they are more stem-like and less exhausted. FIGURE 42C shows that CCR7⁺ enrichment of expanded adherent cells is WSS-dependent. At 1.0 dyn/cm² untreated CD8⁺ T cells that adhere to Nepmucin are more PD-1⁺ and less GzmB⁺ in addition to being more like stem-like (TCF- Ihi), compared to 0.5 dyn/cm². FIGURE 42D shows that untreated CD8⁺ T cells that adhere to Nepmucin are less naive and more CD62L', CD44' compared to parent. Expanded CD8⁺ T cells have a lower proportion of Central Memory phenotype at all WSS. FIGURE 42E shows that CCR7 positivity was substantially higher in the adherent fraction of cells gated as TCF- Ihi, suggesting a minor fraction of the captured cells that are more stem-like would also be more lymph-node homing. Among TCF-lhi expanded cells, the adherent fraction is consistently lower in Ki-67 expression. Therefore, enriched cells not only have higher selfrenewal potential but also are in a less proliferative state, compared to parent and freeflow. FIGURE 42F shows that adherent cells sorted from the expanded population (1.0 dyn/cm²) are increasingly Naive among cells positive form homing receptor expression (CCR7⁺ and CXCR3⁺), antigen-experience (PD-1⁺) and exhaustion (TIM3⁺). FIGURE 42G shows that there is no difference in Central Memory subtype among marker populations, irrespective of activation or fractionation group. FIGURE 42H shows that the CCR7⁺ and TIM-3⁺ populations of adherent enriched cells at 1.0 dyn are significantly less effector-like. FIGURE 421 shows that the CD62L', CD44' subtype is higher among CCR7⁺ adherent cells enriched from the expanded population.

[0053] FIGURE 43A shows results for fixed WSS. With each WSS tested individually, firm adhesion is highest at 0.25 and 0.5 dyn/cm² and lowest at 1.0 dyn/cm². FIGURE 43B shows results for increasing WSS during perfusion (every 5 minutes, after 30 minutes at 0.25 dyn), resulting in either increased or maintained levels of firm adhesion. FIGURE 43C shows results for decreasing WSS during perfusion (every 5 minutes, after 5 minutes at 1.0 dyn).

[0054] FIGURE 44A shows E-selectin interaction (firm and rolling adhesion) increases with concentration and decreased WSS. FIGURE 44B shows no major differences in cell interaction due to L-selectin concentration. Firm adhesion dominates over rolling and L- selectin adhesion is noticeably lower than E-selectin at higher concentration. FIGURE 44C shows no major differences in cell interaction due to MAdCAM Concentration. Firm adhesion dominates over rolling and MAdCAM adhesion is lower than E-selectin at higher concentrations.

[0055] FIGURE 45 shows an adhesion chromatography microfluidic system.

[0056] FIGURE 46A shows CD8⁺ T-cell adhesion initiated from a static condition varies with increased Wall Shear Stress for Nepmucin. However, differences in WSS for other molecules are minimal. FIGURE 46B shows CD8⁺ T-cell interaction within the chromatography decreases with increased WSS, under continuous flow. Adhesion for all molecules is above background (1% BSA).

[0057] FIGURE 47 shows a workflow summary.

[0058] FIGURE 48A shows that hMSC firm adhesion (averaged along the channel) decreases with increased WSS for both untreated and IFNy-treated cells. However, IFNy- treatment results in greater de-adhesion in response to WSS. FIGURE 48B shows the number of bound vs. tethering cells per FOV. Untreated cell adhesion remains relatively unchanged over time with increased WSS, whereas IFN-y treated cells de-adhere in response to WSS. There is some variability between donors at the lowest WSS (0.125 dyn/cm²). FIGURE 48C shows the percent of bound cells remaining per FOV. Untreated cell adhesion remains relatively unchanged over time with increased WSS, whereas IFNy-treated cells de-adhere in response to WSS. Percentage is calculated based on number of settled/bound cells at time=0s, either before initiating flow (static to 0.125 dyn/cm²) or increasing WSS from the prior.

[0059] FIGURE 49 shows conditions for continuous flow.

[0060] FIGURE 50 shows a summary of experimental conditions.

[0061] FIGURE 51A shows a comparison of total adhesion on P-selectin vs. Nepmucin. In general, 2nd gen anti-MUC-16 CAR-T cells bind more to P-selectin than Nepmucin, across a range of wall shear stresses. Adhesion is higher on substrates where flow is begun from a static condition, compared to continuous. At all wall shear stresses, functional adhesion on P- selectin and Nepmucin is higher than the blocked (1% BSA) control. FIGURE 51B shows that on P-selectin, firm adhesion dominates in the static condition. However, cell rolling is mainly seen in the continuous flow conditions, where increasing WSS promotes greater rolling on P- selectin (from 0.25 to 0.5 dyn). However, there is no rolling on Nepmucin or Blocked channels. FIGURE 51C shows a comparison of fractional cell recovery on P-selectin and Nepmucin. FIGURE 51D shows the number of cells recovered (static). FIGURE 51E shows the number of cells recovered (continuous).

[0062] FIGURE 52A shows that overall, P-selectin adhesion is higher throughout the channel length (compared to Nepmucin), at both WSS. FIGURE 52B shows that rolling adhesion dominates over firm for P-selectin, whereas firm adhesion is dominant for Nepmucin. There were a few rolling cells at 0.5 dyn for Nepmucin with this experiment. FIGURE 52C shows fractional recovery of CAR-T cells enriched on Nepmucin and P-selectin (20ug/mL each).

[0063] FIGURE 53A shows CCR7⁺ of live CD4 CD8’. FIGURE 53B shows CXCR3⁺ of live CD4 CD8’. FIGURE 53C shows CD62L⁺ of live CD4 CD8’. FIGURE 53D shows adhesive ligand⁺ (P-sel L⁺ vs. Nepmucin- Ligand⁺) of live CD4⁺CD8’. FIGURE 53E shows Sialyl Lewis X⁺ of live CD4 CD8’. FIGURE 53F shows CD45RA⁺ of live CD4 CD8’. FIGURE 53G shows CD45RO⁺ of live CD4 CD8’. FIGURE 53H shows Granzyme B⁺ of live CD4 CD8’. FIGURE 531 shows Ki-67⁺ of live CD4 CD8’. FIGURE 53J shows TCF-1 ⁺ of live CD4 CD8’. FIGURE 53K shows TIM-3⁺ of live CD4 CD8’. FIGURE 53L shows PD- 1⁺ of live CD4 CD8’. FIGURE 53M shows CD45RA⁺ of Ki-67’ of live CD4⁺CD8’. FIGURE 53N shows CD45RA+ of Ki-67⁺of live CD4⁺CD8’. FIGURE 530 shows CD45RO⁺ of Ki-67’ of live CD4 CD8’. FIGURE 53P shows CD45RO⁺ of Ki-67’ of live CD4⁺CD8’. FIGURE 53Q shows CD62L vs. CCR7 of CD45RA’ of live CD4⁺CD8’ (Q2: Naive and Q4: Effector). FIGURE 53R shows CD62L vs. CCR7 of CD45RA’ of live CD4⁺CD8’ (Q2: Central Memory vs. Q4: Effector Memory).

[0064] FIGURE 54A shows CCR7⁺ of live CD4’CD8⁺. FIGURE 54B shows CXCR3⁺ of live CD4’CD8⁺. FIGURE 54C shows CD62L⁺ of live CD4’CD8⁺. FIGURE 54D shows adhesion ligand⁺ of live CD4’CD8⁺. FIGURE 54E shows CD45RA⁺ of live CD4’CD8⁺. FIGURE 54F shows CD45RO⁺ of live CD4’CD8⁺. FIGURE 54G shows Sialyl Lewis X⁺ of live CD4’CD8⁺. FIGURE 54H shows Granzyme B⁺ of live CD4’CD8⁺. FIGURE 541 shows Ki-67⁺ of live CD4’CD8⁺. FIGURE 54 J shows TCF-1⁺ of live CD4’CD8⁺. FIGURE 54K shows TIM3⁺ of live CD4’CD8⁺. FIGURE 54L shows PD-1⁺ of live CD4’CD8⁺. FIGURE 54M shows CD45RA⁺ of Ki-67’ of live CD4’CD8⁺. FIGURE 54N shows CD45RA⁺ of Ki-67⁺ of live CD4’CD8⁺. FIGURE 540 shows CD45RO⁺ of Ki-67’ of live CD4’CD8⁺. FIGURE 54P shows CD45RO⁺ of Ki-67’ of live CD4’CD8⁺. FIGURE 54Q shows CD62L vs. CCR7 of CD45RA⁺ of live CD4’CD8⁺ (Q2: Naive and Q4: Effector). FIGURE 54R shows CD62L vs. CCR7 of CD45RA’ of live CD4’CD8⁺ (Q2: Central Memory vs. Q4: Effector Memory). DETAILED DESCRIPTION

[0065] It is appreciated that certain features of the disclosure, which are, for clarity, described in the context of separate aspects, can also be provided in combination with a single aspect. Conversely, various features of the disclosure, which are, for brevity, described in the context of a single aspect, can also be provided separately or in any suitable subcombination. Unless defined otherwise, all technical and scientific terms used herein have the same meaning as commonly understood by one of ordinary skill in the art. Methods and materials similar or equivalent to those described herein can be used in the practice or testing of the present disclosure.

DEFINITIONS

[0066] In this specification and in the claims that follow, reference will be made to a number of terms, which shall be defined to have the following meanings:

[0067] Throughout the description and claims of this specification, the word “comprise” and other forms of the word, such as “comprising” and “comprises,” means including but not limited to, and are not intended to exclude, for example, other additives, segments, integers, or steps. Furthermore, it is to be understood that the terms comprise, comprising, and comprises as they relate to various aspects, elements, and features of the disclosed invention also include the more limited aspects of “consisting essentially of’ and “consisting of.”

[0068] As used herein, the singular forms “a,” “an,” and “the” include plural referents unless the context clearly dictates otherwise. Thus, for example, reference to a “cell” includes aspects having two or more such cells unless the context clearly indicates otherwise.

[0069] Ranges can be expressed herein as from “about” one particular value and/or to “about” another particular value. When such a range is expressed, another aspect includes from the one particular value and/or to the other particular value. Similarly, when values are expressed as approximations, by use of the antecedent “about,” it will be understood that the particular value forms another aspect. It should be further understood that the endpoints of each of the ranges are significant both in relation to the other endpoint, and independently of the other endpoint.

[0070] As used herein, the terms “optional” or “optionally” mean that the subsequently described event or circumstance may or may not occur, and that the description includes instances where said event or circumstance occurs and instances where it does not.

[0071] For the terms “for example” and “such as,” and grammatical equivalences thereof, the phrase “and without limitation” is understood to follow unless explicitly stated otherwise. [0072] The term “administer,” “administering”, or derivatives thereof refer to delivering a composition, substance, inhibitor, or medication to a subject or object by one or more the following routes: oral, topical, intravenous, subcutaneous, transcutaneous, transdermal, intramuscular, intra-joint, parenteral, intra-arteriole, intradermal, intraventricular, intracranial, intraperitoneal, intralesional, intranasal, rectal, vaginal, by inhalation or via an implanted reservoir.

[0073] As used herein, the term “affinity” refers to the tendency of a first moiety to associate with, bind to, or make contact with a substrate or a second moiety. The first moiety may interact with the substrate or second moiety chemically, electrostatically, magnetically, or mechanically.

[0074] As used herein, the term “cell” includes progeny. It is also understood that all progenies may not be precisely identical in DNA content, due to deliberate or inadvertent mutations. Variant progeny that have the same function or biological property, as screened for in the originally transformed cell, are included. The “cells” referred to in the present invention generally are prokaryotic or eukaryotic hosts.

[0075] As used herein, “cell adhesion" refers to the process by which cells interact and attach to neighboring cells, their environment, or functionalized surfaces through specialized molecules of the cell surface.

[0076] As used herein, the term “diseased tissue” refers to a tissue or a portion of a tissue that is damaged, inflamed, infected, abnormal, or otherwise compromised. For example, a diseased tissue can refer to an inflamed tissue or a tumor.

[0077] As used herein, the term “functional” or “functionalized” refers to the treatment or conditioning of a substrate with a functional element. A “functional element” may refer to any chemical and/or biological moiety, including, but not limited to, peptides, nucleic acids, antibodies, cells, small molecules, and the like. The functional element may be bound to the substrate or incorporated into the substrate.

[0078] As used herein, “microfluidics" refers to precise control and manipulation of fluids that are geometrically constrained to a small scale at which surface forces dominate volumetric forces.

[0079] As used herein, the term “physiological” refers to a characteristic of a tissue, organ, body part, or the like as found in a living subject. It is understood that such characteristics may vary between subjects, and the term “physiological” is intended to capture a reasonable range of said characteristics as one may expect to find in any one subject. [0080] The term “sample” as used herein means a sample of biological tissue or fluid. Such samples include, but are not limited to, tissue isolated from animals. Samples can also include sections of tissues such as biopsy and autopsy samples, frozen sections taken for histologic purposes, blood, plasma, serum, sputum, stool, tears, mucus, hair, and skin. Samples also explants and primary and/or transformed cell cultures derived from patient tissues. A sample can be provided by removing a sample of cells from an animal, but can also be accomplished by using previously isolated cells (e.g., isolated by another person, at another time, and/or for another purpose). Archival tissues, such as those having treatment or outcome history can also be used.

[0081] The term “subject” refers to any individual who is the target of administration or treatment. The subject can be a vertebrate, for example, a mammal. In one aspect, the subject can be human, non-human primate, bovine, equine, porcine, canine, or feline. The subject can also be a guinea pig, rat, hamster, rabbit, mouse, or mole. Thus, the subject can be a human or veterinary patient. The term “patient” refers to a subject under the treatment of a clinician, e.g., physician.

[0082] Substrate” refers to any rigid or semi-rigid support to which a functional element is bound and includes membranes, filters, chips, slides, wafers, fibers, magnetic or nonmagnetic beads, gels, capillaries or other tubing, plates, polymers, and microparticles with a variety of surface forms including wells, trenches, pins, channels, and pores.

[0083] As used herein, “T cell" refers to a lymphocyte produced by the thymus gland that resides in lymph nodes. T cells play a major role in cell-mediated immunity, which is mediated by their specificity toward antigens due to their T cell receptor (TCR) and cytotoxic mechanisms to eliminate infected or mutated cells. T cells play a major role in cancer immunotherapy. As used herein, “CD8⁺ T cell" refers to T cells that are MHC class I restricted and are mediator of the adaptive immunity.

[0084] “Therapeutic composition” refers to any composition that has a beneficial biological effect. Beneficial biological effects include both therapeutic effects, e.g., treatment of a disorder or other undesirable physiological condition, and prophylactic effects, e.g., prevention of a disorder or other undesirable physiological condition (e.g., a non-immunogenic cancer). The terms also encompass pharmaceutically acceptable, pharmacologically active derivatives of beneficial agents specifically mentioned herein, including, but not limited to, salts, esters, amides, proagents, active metabolites, isomers, fragments, analogs, and the like. When the terms “therapeutic composition” is used, then, or when a particular agent is specifically identified, it is to be understood that the term includes the agent per se as well as pharmaceutically acceptable, pharmacologically active salts, esters, amides, proagents, conjugates, active metabolites, isomers, fragments, analogs, etc.

[0085] The term “therapeutically effective” refers to the amount of the composition used is of sufficient quantity to ameliorate one or more causes or symptoms of a disease or disorder. Such amelioration only requires a reduction or alteration, not necessarily elimination.

[0086] The term “therapeutic purposes” refers to the execution of a method of treatment, administration of a therapeutic composition, or another such action for the amelioration of one or more causes or symptoms of a disease or disorder. Such amelioration only requires a reduction or alteration, not necessarily elimination.

[0087] The term “tumor” is used herein to refer to a group of cells that exhibit abnormally high levels of proliferation and growth. A tumor may be benign, pre-malignant, or malignant; malignant tumor cells are cancerous.

[0088] As used herein, the term “wall shear stress” refers to a shear stress applied to a moiety by a wall as the moiety moves along the wall. Wall shear stress most particularly applies to a shear stress applied on a cell by the functionalized channel as the cell travels through the functionalized channel.

SYSTEM

[0089] In an aspect, provided is a system for sorting cells, comprising: a substrate, comprising: an input; an output; an at least partially functionalized channel between the input and the output, wherein the functionalized channel is sized to replicate a physiological vasculature system; and a pump configured to flow a sample of cells through the channel at a flow rate that replicates flow through a physiological vasculature system, thereby inducing a wall shear stress on the sample of cells that replicates wall shear stress in a physiological vasculature system; wherein the sample of cells are sorted according to their affinity for the functionalized channel. In some aspects, the sample of cells comprise autologous tumorinfiltrating lymphocytes, CD8⁺ T cells, manufactured leukocytes, or any combination thereof. In some aspects, the manufactured leukocytes are engineered to possess altered functions of chimeric antigen receptors, mesenchymal stem cells, hemopoietic stem cells, or any combination thereof.

[0090] In some aspects, the substrate comprises polydimethylsiloxane (PDMS), polyethylene glycol (PEG), silicone, polymethyl methacrylate, polycarbonate, or any combination thereof. In some aspects, substrate comprises a hydrogel. In some aspects, the substrate is a microfluidic device. In some aspects, the substrate can support a cell culture. [0091] In some aspects, the functionalized channel comprises a ligand. In some aspects, the ligand is a diseased tissue ligand or tumor ligand. A “diseased tissue ligand” or a “tumor ligand” refers to a ligand that is expressed exclusively by cells within diseased or inflamed tissues or a tumor or that is more expressed by the local microenvironment of the diseased tissue or tumor than a reference (e.g. non-diseased or non-tumorous) tissue. A diseased tissue ligand or tumor ligand can particularly refer to such a ligand that is expressed in the vasculature of said diseased tissue or tumor. In some aspects, the ligand is an adhesive ligand. An “adhesive ligand” refers to a ligand that mediates, controls, or is otherwise involved in physical interactions between cells or between a cell and its environment. It is understood that a ligand can be both a diseased tissue ligand or tumor ligand and an adhesive ligand. In some aspects, the ligand comprises P-selectin, E-selectin, Nepmucin, L-selectin, MAdCAM, intercellular adhesion molecule, vascular cell adhesion molecule, CD31, CD34, or any combination thereof. [0092] In some aspects, the functionalized channel is cellularized. The “cellularized” functionalized channel refers to a plurality of cells grown on at least a portion of the walls of said channel at any degree of confluence. In some aspects, the cellularized functionalized channel comprises diseased tissue or tumor cells, endothelial cells, stromal cells, or any combination thereof.

[0093] In some aspects, the functionalized channel comprises the same elements throughout the entire length of the channel. In some aspects, the functionalized channel comprises different elements in different regions of the channel.

[0094] In some aspects, the functionalized channel has a diameter of from about 50 pm to about 300 pm, or from about 70 pm to about 280 pm, or from about 90 pm to about 260 pm, or from about 110 pm to about 240 pm, or from about 130 pm to about 220 pm, or from about 150 pm to about 200 pm, or from about 170 pm to about 180 pm, or from about 50 pm to about 200 pm, or from about 60 pm to about 190 pm, or from about 70 pm to about 180 pm, or from about 80 pm to about 170 pm, or from about 90 pm to about 160 pm, or from about 100 pm to about 150 pm, or from about 110 pm to about 140 pm, or from about 120 pm to about 130 pm, or from about 150 pm to about 300 pm, or from about 160 pm to about 290 pm, or from about 170 pm to about 280 pm, or from about 180 pm to about 270 pm, or from about 190 pm to about 260 pm, or from about 200 pm to about 250 pm, or from about 210 pm to about 240 pm, or from about 220 pm to about 230 pm.

[0095] In some aspects, the functionalized channel has a consistent diameter throughout the entire length of the channel. In some aspects, the diameter of the functionalized channel varies throughout within the bounds of the above-described ranges. [0096] In some aspects, the substrate further comprises a settling region in fluid communication with the functionalized channel. The “settling region” describes a channel or region that is not functionalized and is of sufficient length such that the cells can “settle” to the same vertical height within the substrate before entering the functionalized channel. In some aspects, the settling region is from about 1 cm to about 14 cm in length, or from about 2 cm to about 13 cm, or from about 3 cm to about 12 cm, or from about 4 cm to about 11 cm, or from about 5 cm to about 10 cm, or from about 6 cm to about 9 cm, or from about 7 cm to about 8 cm, or from about 1 cm to about 8 cm, or from about 1.5 cm to about 7.5 cm, or from about 2 cm to about 7 cm, or from about 2.5 cm to about 6.5 cm, or from about 3 cm to about 6 cm, or from about 3.5 cm to about 5.5 cm, or from about 4 cm to about 5 cm, or from about 7 cm to about 14 cm, or from about 7.5 cm to about 13.5 cm, or from about 8 cm to about 13 cm, or from about 8.5 cm to about 12.5 cm, or from about 9 cm to about 12 cm, or from about 9.5 cm to about 11.5 cm, or from about 10 cm to about 11 cm.

[0097] In some aspects, cells with high affinity for the functionalized channel pass through the functionalized channel more slowly than cells with low affinity for the functionalized channel. In such aspects, the cells with high affinity for the functionalized channel may strongly interact with the functionalized channel and be able to resist the push of fluid flow through the channel, whereas the cells with low affinity for the functionalized channel may weakly interact with the functionalized channel and not be able to resist the push of fluid flow through the channel as well as the high affinity cells, thereby traveling through the functionalized channel more rapidly. Accordingly, a cell with low affinity for the functionalized channel may travel through the functionalized channel more slowly than the fluid, and a cell with high affinity for the functionalized channel may travel through the functionalized channel much more slowly than the fluid and slower than the low affinity cell.

[0098] In some aspects, the system further comprises a light source, and wherein the light source is positioned to illuminate a portion of the functionalized channel. In some aspects, the light source emits visible light, ultraviolet light, or infrared light. In some aspects, the sample of cells further comprise a photoactive element, and wherein the light source activates the photoactive element, thereby causing the cells to fluoresce. A “photoactive element” refers to a moiety that is modified by exposure to light, specifically a moiety that fluoresces upon or after exposure to light. In some aspects, the photoactive element comprises photoactivatable green fluorescent protein (PA-GFP), Dendra2, KikRGreen, Phamret, a photoactive compound or dye, or any combination thereof. [0099] In some aspects, cells with high affinity for the functionalized channel have a greater residence time in the illuminated portion of the functionalized channel than cells with low affinity for the functionalized channel. The term “residence time” refers to the duration which a given cell is in a specified region, specifically the illuminated portion. A cell with a greater residence time in the illuminated portion would receive greater light exposure than a cell with a lesser residence time. Accordingly, in some aspects, cells with high affinity for the functionalized channel have greater fluorescence intensity than cells with low affinity for the functionalized channel.

[0100] In some aspects, cells that have substantially no affinity for the functionalized channel do not fluoresce. “Substantially no affinity” refers to a cell that does not interact with the functionalized channel or only minimally interacts with the functionalized channel such that the cell travels through the functionalized channel at approximately the same rate as the fluid.

[0101] In some aspects, the flow rate is from about 0.05 mL/min to about 0.5 mL/min, or from about 0.06 mL/min to about 0.48 mL/min, or from about 0.07 mL/min to about 0.46 mL/min, or from about 0.08 mL/min to about 0.44 mL/min, or from about 0.09 mL/min to about 0.42 mL/min, or from about 0.1 mL/min to about 0.4 mL/min, or from about 0.125 mL/min to about 0.375 mL/min, or from about 0.15 mL/min to about 0.35 mL/min, or from about 0.175 mL/min to about 0.325 mL/min, or from about 0.2 mL/min to about 0.3 mL/min, or from about 0.225 mL/min to about 0.275 mL/min, or from about 0.05 mL/min to about 0.25 mL/min, or from about 0.06 mL/min to about 0.24 mL/min, or from about 0.07 mL/min to about 0.22 mL/min, or from about 0.08 mL/min to about 0.2 mL/min, or from about 0.09 mL/min to about 0.18 mL/min, or from about 0.1 mL/min to about 0.16 mL/min, or from about 0.12 mL/min to about 0.14 mL/min, or from about 0.25 mL/min to about 0.5 mL/min, or from about 0.275 mL/min to about 0.475 mL/min, or from about 0.3 mL/min to about 0.45 mL/min, or from about 0.325 mL/min to about 0.425 mL/min, or from about 0.35 mL/min to about 0.4 mL/min.

[0102] In some aspects, the wall shear stress is from about 0.25 dyn/cm² to about 1 dyn/cm², or from about 0.3 dyn/cm² to about 0.95 dyn/cm², or from about 0.35 dyn/cm² to about 0.9 dyn/cm², or from about 0.4 dyn/cm² to about 0.85 dyn/cm², or from about 0.45 dyn/cm² to about 0.8 dyn/cm², or from about 0.5 dyn/cm² to about 0.75 dyn/cm², or from about 0.55 dyn/cm² to about 0.7 dyn/cm², or from about 0.6 dyn/cm² to about 0.65 dyn/cm², or from about 0.25 dyn/cm² to about 0.65 dyn/cm², or from about 0.3 dyn/cm² to about 0.6 dyn/cm², or from about 0.35 dyn/cm² to about 0.55 dyn/cm², or from about 0.4 dyn/cm² to about 0.5 dyn/cm², or from about 0.6 dyn/cm² to about 1 dyn/cm², or from about 0.65 dyn/cm² to about 0.95 dyn/cm², or from about 0.7 dyn/cm² to about 0.9 dyn/cm², or from about 0.75 dyn/cm² to about 0.85 dyn/cm².

METHODS

[0103] In an aspect, provided is a method of sorting cells, comprising: (a) providing a substrate comprising an at least partially functionalized channel, wherein the functionalized channel is sized to replicate a physiological vasculature system; (b) flowing a sample of cells through the at least partially functionalized channel at a flow rate that replicates flow through a physiological vasculature system, thereby inducing a wall shear stress on the sample of cells that replicates wall shear stress in a physiological vasculature system; and (c) selecting the cells that show high affinity for the functionalized channel.

[0104] In some aspects, the sample of cells comprise autologous tumor-infiltrating lymphocytes, CD8⁺ T cells, manufactured leukocytes, or any combination thereof. In some aspects, the manufactured leukocytes are engineered to possess altered functions of chimeric antigen receptors, mesenchymal stem cells, hemopoietic stem cells, or any combination thereof. In some aspects, the method is used to sort cells used for therapeutic purposes. In some aspects, the method is used to sort cells for diagnostic purposes. In some aspects, the method is used to sort cells for characterization purposes.

[0105] In some aspects, the substrate comprises polydimethylsiloxane (PDMS), polyethylene glycol (PEG), silicone, polymethyl methacrylate, polycarbonate, or any combination thereof. In some aspects, substrate comprises a hydrogel. In some aspects, the substrate is a microfluidic device. In some aspects, the substrate can support a cell culture.

[0106] In some aspects, the functionalized channel comprises a ligand. In some aspects, the ligand is a diseased tissue ligand or tumor ligand. In some aspects, the ligand is an adhesive ligand. It is understood that a ligand can be both a diseased tissue ligand or tumor ligand and an adhesive ligand. In some aspects, the ligand comprises P-selectin, E-selectin, Nepmucin, L- selectin, MAdCAM, intercellular adhesion molecule, vascular cell adhesion molecule, CD31, CD34, or any combination thereof.

[0107] In some aspects, the functionalized channel is cellularized. The “cellularized” functionalized channel refers to a plurality of cells grown on at least a portion of the walls of said channel at any degree of confluence. In some aspects, the cellularized functionalized channel comprises diseased tissue or tumor cells, endothelial cells, stromal cells, or any combination thereof. [0108] In some aspects, the functionalized channel comprises the same elements throughout the entire length of the channel. In some aspects, the functionalized channel comprises different elements in different regions of the channel.

[0109] In some aspects, the functionalized has a diameter of from about 50 pm to about 300 pm, or from about 70 pm to about 280 pm, or from about 90 pm to about 260 pm, or from about 110 pm to about 240 pm, or from about 130 pm to about 220 pm, or from about 150 pm to about 200 pm, or from about 170 pm to about 180 pm, or from about 50 pm to about 200 pm, or from about 60 pm to about 190 pm, or from about 70 pm to about 180 pm, or from about 80 pm to about 170 pm, or from about 90 pm to about 160 pm, or from about 100 pm to about 150 pm, or from about 110 pm to about 140 pm, or from about 120 pm to about 130 pm, or from about 150 pm to about 300 pm, or from about 160 pm to about 290 pm, or from about 170 pm to about 280 pm, or from about 180 pm to about 270 pm, or from about 190 pm to about 260 pm, or from about 200 pm to about 250 pm, or from about 210 pm to about 240 pm, or from about 220 pm to about 230 pm. In some aspects, the functionalized channel has a consistent diameter throughout the entire length of the channel. In some aspects, the diameter of the functionalized channel varies throughout within the bounds of the above-described ranges.

[0110] In some aspects, the substrate further comprises a settling region in fluid communication with the functionalized channel. In some aspects, the settling region is from about 1 cm to about 14 cm in length, or from about 2 cm to about 13 cm, or from about 3 cm to about 12 cm, or from about 4 cm to about 11 cm, or from about 5 cm to about 10 cm, or from about 6 cm to about 9 cm, or from about 7 cm to about 8 cm, or from about 1 cm to about 8 cm, or from about 1.5 cm to about 7.5 cm, or from about 2 cm to about 7 cm, or from about 2.5 cm to about 6.5 cm, or from about 3 cm to about 6 cm, or from about 3.5 cm to about 5.5 cm, or from about 4 cm to about 5 cm, or from about 7 cm to about 14 cm, or from about 7.5 cm to about 13.5 cm, or from about 8 cm to about 13 cm, or from about 8.5 cm to about 12.5 cm, or from about 9 cm to about 12 cm, or from about 9.5 cm to about 11.5 cm, or from about 10 cm to about 11 cm.

[OHl] In some aspects, cells with high affinity for the functionalized channel pass through the functionalized channel more slowly than cells with low affinity for the functionalized channel.

[0112] In some aspects, step (c) further comprises illuminating a portion of the functionalized channel with a light source. In some aspects, the light source emits visible light, ultraviolet light, or infrared light. In some aspects, the sample of cells further comprise a photoactive element, and wherein the light source activates the photoactive element, thereby causing the cells to fluoresce. In some aspects, the photoactive element comprises photoactivatable green fluorescent protein (PA-GFP), Dendra2, KikRGreen, Phamret, a photoactive compound or dye, or any combination thereof.

[0113] In some aspects, cells with high affinity for the functionalized channel have a greater residence time in the illuminated portion of the functionalized channel than cells with low affinity for the functionalized channel. In some aspects, cells with high affinity for the functionalized channel have greater fluorescence intensity than cells with low affinity for the functionalized channel. In some aspects, cells that have substantially no affinity for the functionalized channel do not fluoresce.

[0114] In some aspects, the flow rate is from about 0.05 mL/min to about 0.5 mL/min, or from about 0.06 mL/min to about 0.48 mL/min, or from about 0.07 mL/min to about 0.46 mL/min, or from about 0.08 mL/min to about 0.44 mL/min, or from about 0.09 mL/min to about 0.42 mL/min, or from about 0.1 mL/min to about 0.4 mL/min, or from about 0.125 mL/min to about 0.375 mL/min, or from about 0.15 mL/min to about 0.35 mL/min, or from about 0.175 mL/min to about 0.325 mL/min, or from about 0.2 mL/min to about 0.3 mL/min, or from about 0.225 mL/min to about 0.275 mL/min, or from about 0.05 mL/min to about 0.25 mL/min, or from about 0.06 mL/min to about 0.24 mL/min, or from about 0.07 mL/min to about 0.22 mL/min, or from about 0.08 mL/min to about 0.2 mL/min, or from about 0.09 mL/min to about 0.18 mL/min, or from about 0.1 mL/min to about 0.16 mL/min, or from about 0.12 mL/min to about 0.14 mL/min, or from about 0.25 mL/min to about 0.5 mL/min, or from about 0.275 mL/min to about 0.475 mL/min, or from about 0.3 mL/min to about 0.45 mL/min, or from about 0.325 mL/min to about 0.425 mL/min, or from about 0.35 mL/min to about 0.4 mL/min.

[0115] In some aspects, the wall shear stress is from about 0.25 dyn/cm² to about 1 dyn/cm², or from about 0.3 dyn/cm² to about 0.95 dyn/cm², or from about 0.35 dyn/cm² to about 0.9 dyn/cm², or from about 0.4 dyn/cm² to about 0.85 dyn/cm², or from about 0.45 dyn/cm² to about 0.8 dyn/cm², or from about 0.5 dyn/cm² to about 0.75 dyn/cm², or from about 0.55 dyn/cm² to about 0.7 dyn/cm², or from about 0.6 dyn/cm² to about 0.65 dyn/cm², or from about 0.25 dyn/cm² to about 0.65 dyn/cm², or from about 0.3 dyn/cm² to about 0.6 dyn/cm², or from about 0.35 dyn/cm² to about 0.55 dyn/cm², or from about 0.4 dyn/cm² to about 0.5 dyn/cm², or from about 0.6 dyn/cm² to about 1 dyn/cm², or from about 0.65 dyn/cm² to about 0.95 dyn/cm², or from about 0.7 dyn/cm² to about 0.9 dyn/cm², or from about 0.75 dyn/cm² to about 0.85 dyn/cm². [0116] In some aspects, the sample of cells is a first native cell population. A “native cell population” refers to a sample of cells obtained from a subject and exhibiting a physiological range of characteristics (i.e., having a variance between cells as may be expected in a given subject). In some aspects, the method is used to predict a behavior or characteristic of a second native cell population. In such aspects, the first native cell population can be used to determine what percent or amount of cells in the first native cell population exhibit the behavior or characteristic, which can then be used to predict what percent or amount of cells in the second native cell population may exhibit the same behavior or characteristic. In some aspects, the behavior or characteristic is affinity for at least one element of the functionalized channel. For example, if a given percent of cells in the first native cell population exhibit an affinity for an element of the functionalized channel, the method can be used to predict that a similar percent of cells in a second native cell population also exhibit affinity for said element.

[0117] In some aspects, the method further comprises: (d) collecting the selected cells; wherein collected cells are suitable for additional testing and/or administration to a patient. A cell that is “suitable for additional testing and/or administration to a patient” is biologically active (i.e., not dead or significantly damaged) and does not include any contaminants (e.g., dyes, markers, and the like) which may impact a test result or a therapeutic benefit.

[0118] In some aspects, the collected cells have a greater ability to home to and engraft with a tissue having the physiological vasculature system replicated by the functionalized channel compared to non-selected cells. In some aspects, the collected cells have enhanced diseased tissue or tumor trafficking compared to the non-selected cells.

[0119] In some aspects, the method further comprises: (e) administering the collected cells to a patient.

[0120] In another aspect, provided is a method of adoptive cell therapy, the method comprising: (a) providing a substrate comprising an at least partially functionalized channel, wherein the functionalized channel is sized to replicate a diseased tissue or tumor vasculature system; (b) flowing a sample of cells through the at least partially functionalized channel at a flow rate that replicates flow through a diseased tissue or tumor vasculature system, thereby inducing a wall shear stress on the sample of cells that replicates wall shear stress in a diseased tissue or tumor vasculature system; (c) collecting the cells that show high affinity for the functionalized channel; and (d) administering the collected cells to a patient with a diseased tissue or tumor; wherein the functionalized channel comprises a ligand; and wherein the collected cells have a greater ability to home to and engraft with the diseased tissue or tumor compared to the cells that show low affinity for the functionalized channel. [0121] In some aspects, the sample of cells comprise autologous tumor-infiltrating lymphocytes, CD8⁺ T cells, manufactured leukocytes, or any combination thereof. In some aspects, the manufactured leukocytes are engineered to possess altered functions of chimeric antigen receptors, mesenchymal stem cells, hemopoietic stem cells, or any combination thereof. In some aspects, the collected cells have enhanced diseased tissue or tumor trafficking compared to the non-selected cells. In some aspects, the method is used to sort cells used for therapeutic purposes. In some aspects, the method is used to sort cells for diagnostic purposes. In some aspects, the method is used to sort cells for characterization purposes.

[0122] In some aspects, the substrate comprises polydimethylsiloxane (PDMS), polyethylene glycol (PEG), silicone, polymethyl methacrylate, polycarbonate, or any combination thereof. In some aspects, substrate comprises a hydrogel. In some aspects, the substrate is a microfluidic device. In some aspects, the substrate can support a cell culture.

[0123] In some aspects, the functionalized channel comprises a ligand. In some aspects, the ligand is a diseased tissue ligand or tumor ligand. In some aspects, the ligand is an adhesive ligand. It is understood that a ligand can be both a diseased tissue ligand or tumor ligand and an adhesive ligand. In some aspects, the ligand comprises P-selectin, E-selectin, Nepmucin, L- selectin, MAdCAM, intercellular adhesion molecule, vascular cell adhesion molecule, CD31, CD34, or any combination thereof.

[0124] In some aspects, the functionalized channel is cellularized. The “cellularized” functionalized channel refers to a plurality of cells grown on at least a portion of the walls of said channel at any degree of confluence. In some aspects, the cellularized functionalized channel comprises diseased tissue or tumor cells, endothelial cells, stromal cells, or any combination thereof.

[0125] In some aspects, the functionalized channel comprises the same elements throughout the entire length of the channel. In some aspects, the functionalized channel comprises different elements in different regions of the channel.

[0126] In some aspects, the functionalized has a diameter of from about 50 pm to about 300 pm, or from about 70 pm to about 280 pm, or from about 90 pm to about 260 pm, or from about 110 pm to about 240 pm, or from about 130 pm to about 220 pm, or from about 150 pm to about 200 pm, or from about 170 pm to about 180 pm, or from about 50 pm to about 200 pm, or from about 60 pm to about 190 pm, or from about 70 pm to about 180 pm, or from about 80 pm to about 170 pm, or from about 90 pm to about 160 pm, or from about 100 pm to about 150 pm, or from about 110 pm to about 140 pm, or from about 120 pm to about 130 pm, or from about 150 pm to about 300 pm, or from about 160 pm to about 290 pm, or from about 170 pm to about 280 pm, or from about 180 pm to about 270 pm, or from about 190 pm to about 260 pm, or from about 200 pm to about 250 pm, or from about 210 pm to about 240 pm, or from about 220 pm to about 230 pm. In some aspects, the functionalized channel has a consistent diameter throughout the entire length of the channel. In some aspects, the diameter of the functionalized channel varies throughout within the bounds of the above-described ranges.

[0127] In some aspects, the substrate further comprises a settling region in fluid communication with the functionalized channel. In some aspects, the settling region is from about 1 cm to about 14 cm in length, or from about 2 cm to about 13 cm, or from about 3 cm to about 12 cm, or from about 4 cm to about 11 cm, or from about 5 cm to about 10 cm, or from about 6 cm to about 9 cm, or from about 7 cm to about 8 cm, or from about 1 cm to about 8 cm, or from about 1.5 cm to about 7.5 cm, or from about 2 cm to about 7 cm, or from about 2.5 cm to about 6.5 cm, or from about 3 cm to about 6 cm, or from about 3.5 cm to about 5.5 cm, or from about 4 cm to about 5 cm, or from about 7 cm to about 14 cm, or from about 7.5 cm to about 13.5 cm, or from about 8 cm to about 13 cm, or from about 8.5 cm to about 12.5 cm, or from about 9 cm to about 12 cm, or from about 9.5 cm to about 11.5 cm, or from about 10 cm to about 11 cm.

[0128] In some aspects, cells with high affinity for the functionalized channel pass through the functionalized channel more slowly than cells with low affinity for the functionalized channel.

[0129] In some aspects, step (c) further comprises illuminating a portion of the functionalized channel with a light source. In some aspects, the light source emits visible light, ultraviolet light, or infrared light. In some aspects, the sample of cells further comprise a photoactive element, and wherein the light source activates the photoactive element, thereby causing the cells to fluoresce. In some aspects, the photoactive element comprises photoactivatable green fluorescent protein (PA-GFP), Dendra2, KikRGreen, Phamret, a photoactive compound or dye, or any combination thereof.

[0130] In some aspects, cells with high affinity for the functionalized channel have a greater residence time in the illuminated portion of the functionalized channel than cells with low affinity for the functionalized channel. In some aspects, cells with high affinity for the functionalized channel have greater fluorescence intensity than cells with low affinity for the functionalized channel. In some aspects, cells that have substantially no affinity for the functionalized channel do not fluoresce. [0131] In some aspects, the flow rate is from about 0.05 mL/min to about 0.5 mL/min, or from about 0.06 mL/min to about 0.48 mL/min, or from about 0.07 mL/min to about 0.46 mL/min, or from about 0.08 mL/min to about 0.44 mL/min, or from about 0.09 mL/min to about 0.42 mL/min, or from about 0.1 mL/min to about 0.4 mL/min, or from about 0.125 mL/min to about 0.375 mL/min, or from about 0.15 mL/min to about 0.35 mL/min, or from about 0.175 mL/min to about 0.325 mL/min, or from about 0.2 mL/min to about 0.3 mL/min, or from about 0.225 mL/min to about 0.275 mL/min, or from about 0.05 mL/min to about 0.25 mL/min, or from about 0.06 mL/min to about 0.24 mL/min, or from about 0.07 mL/min to about 0.22 mL/min, or from about 0.08 mL/min to about 0.2 mL/min, or from about 0.09 mL/min to about 0.18 mL/min, or from about 0.1 mL/min to about 0.16 mL/min, or from about 0.12 mL/min to about 0.14 mL/min, or from about 0.25 mL/min to about 0.5 mL/min, or from about 0.275 mL/min to about 0.475 mL/min, or from about 0.3 mL/min to about 0.45 mL/min, or from about 0.325 mL/min to about 0.425 mL/min, or from about 0.35 mL/min to about 0.4 mL/min.

[0132] In some aspects, the wall shear stress is from about 0.25 dyn/cm² to about 1 dyn/cm², or from about 0.3 dyn/cm² to about 0.95 dyn/cm², or from about 0.35 dyn/cm² to about 0.9 dyn/cm², or from about 0.4 dyn/cm² to about 0.85 dyn/cm², or from about 0.45 dyn/cm² to about 0.8 dyn/cm², or from about 0.5 dyn/cm² to about 0.75 dyn/cm², or from about 0.55 dyn/cm² to about 0.7 dyn/cm², or from about 0.6 dyn/cm² to about 0.65 dyn/cm², or from about 0.25 dyn/cm² to about 0.65 dyn/cm², or from about 0.3 dyn/cm² to about 0.6 dyn/cm², or from about 0.35 dyn/cm² to about 0.55 dyn/cm², or from about 0.4 dyn/cm² to about 0.5 dyn/cm², or from about 0.6 dyn/cm² to about 1 dyn/cm², or from about 0.65 dyn/cm² to about 0.95 dyn/cm², or from about 0.7 dyn/cm² to about 0.9 dyn/cm², or from about 0.75 dyn/cm² to about 0.85 dyn/cm².

EXAMPLES

Example 1: Adhesion analysis via a tumor vasculature-like microfluidic device identifies CD8⁺ T cells with enhanced tumor homing to improve cell therapy

[0133] Adoptive cell therapy (ACT) has emerged as a powerful treatment option for patients with metastatic melanoma.^1,2 ACT including autologous tumor-infiltrating lymphocytes expanded ex vivo and transferred back into the patient in combination with interleukin (IL)-2 can boost anti -tumor immunity.^{1 3} This has yielded good clinical responses for the treatment of metastatic melanoma, but at overall low rates.^1,2 It is now known that low patient rates of response are due to poor trafficking of transferred cells to relevant tissues.¹ [0134] In various cancer types, including melanoma, tumor infiltration by CD8⁺ T cells is correlated with reduction in disease burden and improved survival.^2,4-11 In the context of ACT, cultures containing more CD8⁺ T cells also tend to yield better clinical responses.^2,4,5 Unfortunately, only a small fraction of infused T cells make it into either primary or metastatic lesions.^1,12,13 This means that infusion of an extremely high number of T cells is required to have any therapeutic effect,^12,14 thus pointing to the central role T cell homing to tumors may play in the efficacy of ACT.^11,15-17 Unfortunately, methods to predict whether ACT cultures will result in robust tumor homing are lacking, as are methods to enrich cells that exhibit these features associated with favorable responses.

[0135] CD8⁺ T cells traffic to quiescent or inflamed tissues in a highly dynamic process that involves a variety of receptor-ligand interactions. The steps involve rolling adhesion that decelerates the cell against the force of blood flow, chemokine-triggered integrin activation, and integrin-mediated firm adhesion leading to transmigration through the endothelial layer.^17- ¹⁹ Due to their role as the initial kickoff step in this adhesion cascade, selectins and their ligands have been correlatively or directly implicated in homing of T cells into melanomas, inflamed skin, and other tumor types.^20-26 In experimental animal models of lymphocyte homing, endothelial cells isolated from melanomas are shown to express E- and P-selectin, among other adhesion receptors,²⁷ mirroring human disease.^20,28,29 Melanoma growth is accelerated and CD8⁺ T cell infiltration is decreased in mice lacking P-, but not E-, selectin or its ligand P- selectin glycoprotein ligand (PSGL)-l ²⁶ Despite the apparent contribution of P-selectin and its ligand(s) in CD8⁺ T cell homing and disease progression in the context of melanoma, how P- selectin can contribute to the adhesion and homing of CD8⁺ T cells in ACT has remained explored until now.

[0136] Herein, profiles of endogenous or adoptively transferred murine CD8⁺ T cells harvested from tumors in an in vivo melanoma model were found to mirror those enriched for their capacity to mediate adhesion to P-selectin in flow in vitro using a cell-adhesion chromatography system^30-32 that recapitulates the wall shear stress (WSS) environment of the tumor vasculature.^30-32 Both murine and human CD8⁺ T cells could be enriched for adhesion to P-selectin in flow in vitro, which, when applied to a murine tumor model, resulted in enhanced tumor homing that was associated with improved tumor control of ACT in combination with therapeutic blockade of immune checkpoint programmed cell death (PD-1). These results demonstrate the utility of an engineered microfluidic device in recapitulating the hemodynamic microenvironment of the tumor vasculature in modeling the in vivo homing of lymphocytes relevant to ACT. Materials and Methods

[0137] TABLE 1 shows key resources used in the following studies.

TABLE 1. Key resources table.

[0138] Cell culture: B16F10 and B16F10-OVA cells were maintained in Dulbecco’s modified Eagle’s medium supplemented with 10% fetal bovine serum, and 1% penicillin/streptomycin/amphotericin B. Cells were passaged at -80% confluence, and maintained at 37°C with 5% CO2 in a standard incubator.

[0139] Animal tumor models: C57B1/6 or B6 CD45.1 female mice were purchased at six weeks of age from the Jackson Laboratory. All protocols were approved by the Institutional Animal Care and Use Committee (IACUC). For tumor-bearing cohorts, 0.1-0.5 x 10⁶ melanoma cells were implanted intradermally in 6- to 8-week-old mice. To evaluate therapeutic effects, tumor size was measured with calipers in three dimensions and reported as an ellipsoidal volume.

[0140] Study design: This study’s objective was to implement a microfluidic system to predict the in vivo homing by CD8⁺ T-cells by evaluating their adhesion in vitro in a tumor vasculature-like microenvironment that incorporates the effects of hemodynamic flow. Adhesion molecule expression by adherent and non-adherent cells as well as endogenous and adoptively transferred cells that home to the tumor versus lymphoid tissues including the spleen and lymph nodes was evaluated. Cells enriched for their capacity to mediate adhesion in vitro were adoptively transferred into tumor-bearing murine hosts and the resulting cellular biodistribution profiles were compared. The therapeutic effects of adherent versus nonadherent cells as ACT in a tumor immunotherapy model in combination with aPD-1 were evaluated. Sample sizes were chosen based on previously published studies. For animal studies, mice were randomized into various groups before treatment, with one cage having one mouse per group. The endpoint for survival studies was set at the humane endpoint (tumors reaching 1.5 cm in any direction or ulcerating tumors). Experiments were not done in a blinded fashion. [0141] Immunohistochemistry and imaging'. B16F10 tumors 7d post intradermal implantation in the dorsal skin with 0.5 x 10⁶ B16F10 cells were surgically excised, embedded in optimum cutting temperature embedding medium, and stored at -80°C. A cryostat was used to slice 8 mm thick tissue sections that were mounted onto histological slides and stored at - 20°C. Sections were 2% PF A fixed for 20 min at room temperature, blocked with 10% donkey serum diluted in Dulbecco’s Phosphate Buffered Saline (D-PBS) with calcium and magnesium for Ih at room temperature, and incubated overnight at 4°C with the following primary antibody: goat anti-mouse CD62P (1 :13, R&D Systems, AF737), and rat anti-mouse CD3 (1 :50, Invitrogen) or rat anti-mouse CD31 (1 :50, Invitrogen). The following day, the slides were incubated for Ih at room temperature with the following secondary antibody: donkey anti-goat Alexa Flour 555 (1 :200, Invitrogen) and donkey anti-rat Alexa Flour 647 (1 : 1000, Invitrogen). In between each staining step, slides were washed three times with gentle agitation in 0.1% Tween 20 diluted in D-PBS with calcium and magnesium. Washed slides were mounted using Vectashied mounting medium with DAPI and microscopic images were taken using a Zeiss Axi oOb server Z1 fluorescent microscope with a lOx magnification objective.

[0142] Murine CD8⁺ T cell isolation: C57B1/6, B6 CD45.1, or OT-I (purchased from Charles River Laboratories and bred in-house) animals were euthanized, and the spleens were harvested and disrupted with 18G needles followed by washing with B-PBS. Cells were passed through a sterile 70-mm cell strainer, washed, and incubated with red blood cell lysing buffer (Sigma- Aldrich) for 5 min at room temperature, quenched with D-PBS, washed, and resuspended for counting. Cells were resuspended at 10⁸ cells/ml buffer (Biolegend, MojoSort Buffer), and then incubated with a biotin-antibody cocktail for 15 min, followed by streptavidin nanobeads for another 15 min (Biolegend, MojoSort Mouse CD8a Selection Kit). Buffer was added to the mixture and placed in a magnet (STEMCELL Technologies), and the supernatant was collected. Cells were then counted and resuspended in either cell media for expansion or activation experiments, saline for adoptive transfer experiments, or 0.1% BSA for perfusion experiments. Cells were maintained in sterile conditions before adoptive transfer. Pre-transfer, purity, viability, and CD8⁺ T-cell subtypes were confirmed via flow cytometry on a customized BD LSRFortessa flow cytometer.

[0143] Human CD8⁺ T cell isolation: Human peripheral blood (buffy coats) from deidentified healthy donors were purchased (Oklahoma Blood Institute). Peripheral blood mononuclear cells (PBMCs) were isolated from whole blood via centrifugation with Lymphocyte Separation Media (Corning, 25-072-CV). Following PBMC isolation, CD8⁺ T- cells were isolated using a negative isolation kit from Stem Cell (17953). Cells were cryoed down in HI-FBS with 10% DMSO until further use.

[0144] CD8⁺ T cell activation with PMA/Ion: Isolated CD8⁺ T-cells were suspended in either cell media (RPMI 1640 with 10% fetal bovine) or cell media supplemented with 20 ng/mL PMA (Sigma-Aldrich) plus 1 pg/ml lonomycin (Invitrogen, ThermoFisher). Cells were incubated for 4 h at 37°C with 5%CO2 in a standard incubator.

[0145] CD8⁺ T cell expansion: Isolated CD8⁺ T-cells were suspended at a concentration of 106 cells/mL in culture medium (RPMI 1640-containing L-glutamine, 1% penicillin/streptomycin/amphotericin B, 1% HEPES, 1% non-essential amino acids, 1% sodium pyruvate, 0.05mM 2-mercaptoethanol, and 10% fetal bovine serum). Cells were mixed with Dynabeads (Gibco, ThermoFisher) at a bead-to-cell ratio of 1 : 1 and 100 U/ml rIL-2 (R&D Systems). Cells were incubated at 37°C with 5% CO2 in a standard incubator. Cells were examined daily, and after day 3, cells were split daily. Beads were removed on day 3 (for human cells) or day 5 (for murine cells), and then maintain in media with 100 U/ml rIL-2.

[0146] Adoptive transfer: After activation or expansion, cells were suspended at a concentration of 108 in buffer (D-PBS plus ImM CaCh) and mixed with dead cell removal (Annexin V) cocktail, biotin selection cocktail, and RapidSpheres (STEMCELL Technologies). Buffer was added to the mixture and placed in a magnet; the supernatant was collected. Cells were then counted and resuspended at the desired concentration. CD8⁺ T-cells were suspended in sterile saline at a concentration of 106 cells per 200 mL of sterile saline. After mice were anesthetized, the hair over the neck of mice was removed using depilatory cream and cleaned using warm water, then suspended cells were injected intravenously via the jugular vein.

[0147] In vivo biodistribution analysis: At 16 h post-adoptive transfer, mice were euthanized, and tumor-draining lymph nodes, non-draining lymph nodes, spleens, and tumors were collected. Cells were analyzed via flow cytometry, and transferred cells were identified by staining for CD45.1 (donor mice) and CD45.2 (host mice). The percent recovered cells was defined as the number of cells in a given tissue divided by the known number of cells injected into the animal.

[0148] Treatment of B16F10-OVA melanoma-bearing mice: B16F10-OVA cells (0: 1 x 10⁶) were implanted intradermally on day 0. After 7 days mice were treated with 1 million OT- I CD8⁺ T-cells (sorted on chromatography channel or unsorted) i.v. On day 8 and 11, mice were i.t injected with 150 pg of anti-mouse PD-1 (clone RMP1-14; BioXCell) in 30 m 1 of saline. [0149] Flow cytometry: Harvested LNs were incubated with 1 mg/mL of collagenase D (Sigma-Aldrich) in D-PBS with calcium and magnesium for 1 h at 37°C, passed through a 70- mm cell strainer, washed, and resuspended in a 96-well plate for staining. Spleens were disrupted using 18G needles, passed through a 70-mm strainer, washed and incubated with red blood cell lysing buffer (Sigma- Aldrich) for 5 min at room temperature, diluted with D-PBS, washed, and resuspended. Tumors were incubated with 1 mg/mL of collagenase D (Sigma- Aldrich) in D-PBS with calcium and magnesium for 4 h at 37°C, passed through a 70-mm cell strainer, washed, and resuspended. All antibodies for flow cytometry were from Biolegend unless otherwise stated. Cells were blocked with anti-mouse CD16/CD32 (clone, 2.4G2) (Tonbo Biosciences) for 5 min on ice, washed, and stained with fixable viability dye Zombie Aqua for 30 min at room temperature, and then washed. Cells were then incubated with 10 pg/ml P-selectin plus 10 pg/ml of FITC anti-IgG (Fc specific) diluted in D-PBS for 30 min on iced and then washed. Antibody cocktails were prepared in flow cytometry buffer (0.1% bovine serum albumin in D-PBS) following manufacturer concentration or preliminary titrations. Cells were incubated with an antibody cocktail for 30 min on ice and then washed. Cells were fixed with 4% paraformaldehyde (VWR International Inc). For cytokine staining, cells were suspended in IC Fixation Buffer (eBioscience, Thermo Fisher Scientific Inc.) for 60 min on ice in the dark. Cells were then incubated with the antibody cocktail in IC Permeabilization buffer for 60 min at room temperature in the dark. Cells were then washed and resuspended in buffer, and kept at 4°C until analyzed with customized BD LSRFortessa flow cytometer (BD Biosciences). Compensation was performed using AbC, ArC, or UltraComp compensation beads (ThermoFisher).

[0150] Ex vivo SIINFEKL stimulation: After cell isolation, LNs, spleen, and tumor samples were plated in a sterile 96-well U-bottom plate. SIINFEKL peptide (1 pg/ml) in 100 mL of IMDM (Iscove’s modified Dulbecco’s medium) with 10% heat-inactivated fetal bovine serum and 0.05 mM b-mercaptoe-thanol (Sigma-Aldrich) was added to each sample and then incubated for a total of 6 h at 37°C with 5% CO2. Three hours into the incubation period, brefeldin A (50 pg/mL) was added to each sample. Cells were then stained for flow cytometry as described above.

[0151] Flow-based cell adhesion experiments: A 0.01-inch thick silicone gasket with a 2.5 mm wide rectangular opening was assembled between an acrylic disk with an inlet and outlet ports (GlycoTech Corporation) and a functionalized polystyrene dish via vacuum suction. The chamber and inlet and outlet tubing were filled with 0.1% BSA in D-PBS perfusion medium, taking care that no bubbles formed at connections. The outlet line was connected to a syringe pump (PhD Ultra Harvard Apparatus), and the inlet line was connected to a reservoir. The chamber was placed on an Eclipse Texas Instrument (TI) optical microscope (Nikon), and medium was perfused through the chamber at the desired flow rate via syringe withdrawal. A suspension of 5 x 10⁵ CD8⁺ T-cells per ml were added to the inlet reservoir. For continuous flow experiments, cells were perfused at a flow rate to attain desired shear stress. For static condition experiments, cells were perfused into the chamber for 3 min, flow was stopped for 10 min to allow cells to settle. Flow was then restarted at the desired flow rate, and after 1 min, the flow rate was continuously increased. Image recording was done using Nikon NIS- Elements software. For continuous flow, six evenly spaced positions within the functionalized region of the dish were imaged for 30 s each, followed by checking non-specific adhesion on the non-functionalized region. For static conditions, one position was imaged for the entire experiment; as the flow rate increased, the number of cells detaching from the substrate could be visualized; at the end, the non-functionalized control was checked. For all experiments, the exposure time was 0.281 ms, the frame rate was 25 frames per second, and the objective was lOx.

[0152] Substrate functionalization: In experiments utilizing the vacuum-sealed gasket, a 1.073 0.25 cm rectangle in the center of a 35 mm non-tissue culture treated, round, polystyrene dishes were coated with anti-IgG (Fc specific) (Sigma-Aldrich) diluted in D-PBS without calcium and magnesium, at the concentration corresponding to the total P-selectin and ICAM- 1 concentration in each condition. The anti-IgG solution was incubated overnight at 4°C, and then washed with D-PBS, blocked with 1% BSA in D-PBS at room temperature for Ih, then washed, incubated at room temperature for 2 h with 10 pg/ml P-selectin (R&D Systems), 5 pg/ml ICAM-1 (R&D systems), or the combination of both dilutes in D-PBS with calcium and magnesium. The non-functionalized region, a 0.53 x 0.25 cm rectangle immediately proximal to the functionalized rectangle, was blocked with 1% BSA in D-PBS. All dishes were stored at 4°C with D-PBS until use in same-day experiments.

[0153] Chromatography channel fabrication: The microfluidic channels were made using 100 mm thick double-sided adhesive tape (3M). U-shaped channel of two 2 cm wide by 14 cm long sections connected by a 2 cm wide 1.5 cm long section was cut using a crafted cutter (Silhouette America). On one side, the adhesive tape was attached to PDMS (Ellsworth Adhesives). An inlet hole was made with a biopsy punch before attaching the other side of the adhesive tape to a non-tissue culture-treated polystyrene plate with a drilled outlet hole. PDMS was pre-made by mixing PDMS base with curing agent at a ratio of 9: 1 and curing for 4h at 90°C. [0154] Chromatography channel functionalization: The chromatography channel was functionalized by incubating 25 pg/mL anti-IgG (Fc-specific) (R&D Systems) in D-PBS without calcium and magnesium overnight at 4°C, and then washed with D-PBS, blocked with 1% BSA in D-PBS at room temperature for 1 h, then washes again. Then 25 pg/ml of P-selectin diluted in D-PBS with calcium, and magnesium was placed in the functionalized portion of the channel for 2 h at room temperature. Finally, the entire device was blocked with 1% BSA in D-PBS at room temperature for 1 h.

[0155] Chromatography channel experiments: An inlet syringe connected to tubing was filled with perfusion media (0.1% BSA in D-PBS), and was connected to a syringe pump (PhD Ultra Harvard Apparatus). The syringe pump was used to withdraw a cell pulse of 1 mL at a concentration of 2.5 x 10⁶ cells/ml into inlet tubing at a rate of 0.4 mL/min. The tubing containing the cell pulse was inserted into the inlet hole of the channel, and a 5 mL test tube was connected to the bottom of the outlet hole as the cell collection reservoir. The channel was placed on an Eclipse Ti optical microscope (Nikon), and medium was perfused through the channel at the desired flow rate; perfusion was then stopped after the free flow cells elution time had been reached. Then, the syringe and inlet tubing was replaced with a new syringe and tubing containing only perfusion media to eject the adherent cell in the channel out of the channel and into a second collection tube. The number of cells in the sorted fraction was counted using a hemocytometer, and then adoptively transferred into mice or analyzed via flow cytometry.

[0156] Ex vivo P-selectin stimulation: Isolated CD8⁺ T-cells were suspended in either cell media (RPMI 1640 with 10% fetal bovine) and incubated with either 1, 2.5, 10, and 20 pg/ml of P-selectin or with 20 ng/mL PMA (Sigma-Aldrich) plus 1 pg/ml lonomycin (Invitrogen, ThermoFisher) for Ih or 4h. Cells were incubated for 4h at 37°C with 5% CO2 in a standard incubator. After incubation time, cells were analyzed via flow cytometry.

[0157] Flow cytometry analysis: Flow cytometry data were analyzed using FlowJo software version 10, data is presented in this paper using GraphPad Prism.

[0158] Quantification of cell adhesive behavior: Videos were manually post-processed by counting the number of rolling and firmly adherent cells per 30-s video field of view. Firmly adherent cell was defined as a cell that did not move more than 1 radius in length throughout the experiment. Cell velocities were measured for 10-15 rolling cells per FOV with each experiment (with 3 replicas) using a manual particle tracking plugin in ImageJ. Velocity was determined by dividing a cell’s total translational distance over the functionalized area by the total time it took to travel that same distance. [0159] Setting distance analysis: The mean distance of the cells from the bottom of the channel at any given time was experimentally determined and calculated as previously described.³⁰ In short, a cell pulse followed by perfusion media was perfused into an unfunctionalized channel, and 5 min videos were taken with the focal point set at the bottom of the channel at 9 different positions across the channel. The videos were then analyzed using OpenCV-based Traffic Flow Analyzer cell tracking software to detect cell edges and the cell x and y positions of each detected cell. Based on the total tracked distance and video frames per second, cell velocity was calculated using these data. The metric y_ve;/0.5(r_% + r_y) (the vertical location of the cell from the bottom of the channel divided by an effective cell radius) was calculated from Equation 1, and the mean x and y radius were measured during cell tracking.

(Equation 1) [0160] Statistics: Data are represented as means accompanied by SEM, and statistics were calculated using Prism 9 software (GraphPad Software). ****p < 0.0001, ***p < 0.001,**p < 0.01, and *p < 0.05 by unpaired t-test or two-way analysis of variance (ANOVA) followers by post hoc test for multiple comparisons.

Results

[0161] P-selectin ligand⁺ CD8⁺ T cells preferentially traffic to the TME versus lymphoid tissues: The phenotypes and cellular characteristics of endogenous CD8⁺ T cells in the tumor microenvironment (TME) versus secondary lymphoid tissues from C57BL/6 mice bearing B16F10 melanomas were evaluated. Endogenous CD8⁺ T cells infiltrating the TME exhibited more central memory (CD44⁺CD62L⁺; CM) and effector (CD44⁺CD62L ; EFF) phenotypes, while those recovered from the spleens of tumor-bearing animals exhibited a more naive (CD44 CD62L ) phenotype (FIG. 1A). P-selectin has been implicated in T cell homing and is expressed in human melanomas,^28,29’³³ numerous preclinical models,²⁷’²⁸’³⁴ and melanomas that form in the skin of C57BL/6 animals (FIG. 2A) at a higher extent compared with naive skin (FIGS. 3A-3B) Furthermore, T cells can be found near P-selectin-expressing vessels via immunohistochemistry (FIG. 2B). To this end, the expression of P-selectin ligands by lymphocytes harvested from various tissues of tumor-bearing or naive animals was assessed flow cytometrically by staining with a recombinant P-selectin-Fc chimera (FIG. 2C). Recombinant P-selectin-Fc was used to quantify the extent of P-selectin ligand expression by CD8⁺ T cells instead of antibodies recognizing canonical selectin ligands, because, whereas one such ligand, PSGL-1, is constitutively expressed on lymphocytes,³⁵ it mediates selectin binding only in the context of appropriate glycosylation that cannot be distinguished by antibody-mediated staining of PSGL-l’s protein backbone.³⁵ In so doing, a low frequency of P-selectin ligand⁺ CD8⁺ T cells was found in secondary lymphoid tissues such as the spleen and lymph nodes (LNs), including non-tumor-draining (NDLNs) and tumor-draining (TDLNs) LNs (FIG. 2D). In contrast, a high frequency of P-selectin ligand⁺ CD8⁺ T cells were found in the TME and the skin of tumor-naive animals (FIG. 2D). However, overall numbers of P- selectin ligand-expressing CD8⁺ T cells were higher in tumors compared with the naive skin (FIG. 2E)

[0162] To determine whether differential phenotypes of CD8⁺ T cells recovered from tumors and skin versus lymphoid tissues reflected differences in activation state, the in vivo homing of CD8⁺ T cells negatively isolated from spleens of CD45.1 mice left untreated or activated with PMA/Ionomycin (PMA/Ion) (FIGS. 4A-4J) was assessed. As a result, infused populations of cells exhibited either primarily naive or differentiated EFF cells, respectively (FIG. IB) PMA/Ion-treated CD8⁺ T cells exhibited reduced L-selectin and increased CD44 expression (FIGS. 4A-4J). They also exhibited increased expression of lymphocyte function- associated antigen (LFA)-1 and P-selectin ligand, as measured by staining with P-selectin-Fc (FIGS. 4A-4J). However, PSGL-1 expression remained unchanged (FIGS. 4F-4G). Upon transfer into B16F10 melanoma-bearing CD45.2 animals, PMA/Ion-treated cells trafficked to LNs and the spleen less compared with the tumor (FIGS. 5A-5G). Surprisingly, there was no difference in tumor trafficking between CD8⁺ T cell treatments (FIGS. 5A-5G). About 60% of donor cells infiltrating the tumor 16 h post-transfer exhibited a CM phenotype, whereas in spleens, donor cells were largely naive, trends seen in both tissues irrespective of the donor cell population (FIG. 1C). Mirroring trends in endogenous CD8⁺ T cells in each tissue, the proportion of donor cells that were

[0163] P-selectin ligand⁺ was low versus high in spleens versus tumors (FIGS. 2F-2G), irrespective of donor cell population (untreated or PMA/Ion treated). Similarly, trends seen in the spleen with low levels of trafficked donor P-selectin ligand⁺ CD8⁺ T cells were recapitulated in LNs (FIG. 2G).

[0164] CD8⁺ T cells mediate adhesion to P -selectin-functionalized substrates n hemodynamic flow: Expression of P-selectin ligands by endogenous and donor lymphocytes in the TME raised the hypothesis that CD8⁺ T cells mediate adhesion to P-selectin in flow, as has been reported for CD4⁺ T cells.³⁶ As such, the capacity of CD8⁺ T cells to mediate interactions with inflamed vasculature-expressed adhesive ligands, P-selectin, and ICAM, alone or in combination, under conditions of physiological levels of fluid force was evaluated in vitro. The extent of adhesion of CD8⁺ T cells to a P-selectin-functionalized substrate in a parallel plate flow chamber assay was measured using videomicroscopy. These experiments were performed under two conditions: static, where after cells were loaded into the channel, the flow was paused for 5 min before restarting, allowing cells within the channel to interact with the substrate to initiate adhesion in the absence of fluid force (FIG. 6A), and continuous flow, in which the fluid perfusion was never interrupted (FIG. 6B). These experiments were also performed with untreated or PMA/Ion-treated CD8⁺ T cells to evaluate the effects of cell activation state on adhesion. Under static conditions, untreated CD8⁺ T cells exhibited adhesion to P-selectin with or without ICAM, but not to ICAM alone (FIG. 6C). On the other hand, PMA/Ion-treated CD8⁺ T cells exhibited adhesion to both ICAM and P-selectin alone, with the combination of the two having the highest extent of adhesion (FIG. 6D). However, for both untreated and PMA/Ion-treated cells, the addition of ICAM to P-selectin did not increase the extent of adhesion compared with that observed on P-selectin alone (FIG. 6E). The adhesive quality was also assessed, with untreated CD8⁺ T cells largely exhibiting more rolling adhesion, whereas PMA/Ion-treated CD8⁺ T cells exhibited more firm adhesion (FIGS. 7A-7B). Under continuous flow, similar results were observed, with CD8⁺ T cells adhering to P-selectin with or without co-presentation of ICAM, but not to ICAM alone, albeit in a shear-stress-sensitive manner (FIG. 6F). PMA/Ion-treated CD8⁺ T cells under continuous flow also exhibited adhesion to ICAM and P-selectin individually, with their combination resulting in the highest extent of adhesion that was WSS dependent (FIG. 6G). Co-presentation of P-selectin with ICAM, interactions with which in flow are known to be selectin enabled,^36,37 increased the total extent of adhesion by PMA/Ion-treated but not untreated CD8⁺ T cells compared with P- selectin alone (FIG. 6H) Whereas untreated cells interacted in flow with P-selectin- functionalized substrates largely via rolling adhesion, PMA/Ion-treated CD8⁺ T cells interacted more substantially via firm adhesion, albeit at similar total extents that were WSS dependent (FIGS. 7C-7D) Firmer adhesion to the P-selectin substrate by activated CD8⁺ T cells also resulted in overall lower average velocities of adhesion in flow (FIG. 8). Results from experiments under static and continuous flow conditions show that both untreated and PMA/Ion-treated CD8⁺ T cells engage to some extent with P-selectin but not ICAM. Correlation analyses of the number of cells adhering to P-selectin and the frequency of CD8⁺ T cells expressing P-selectin ligands in perfused populations revealed that, while predictive of the extent of in vitro adhesion to P-selectin under static conditions, adhesion under continuous flow was not predicted by P-selectin ligand expression of perfused CD8⁺ T cells (FIGS. 61- 6K). Therefore, P-selectin ligand expression does not predict the extent of adhesion to P- selectin under conditions where adhesion is initiated in the presence of flow.

[0165] Characteristics of CD8⁺ T cells that interact with P-selectin under shear flow: To evaluate the cellular characteristics of CD8⁺ T cells that do or do not adhere to P-selectin- functionalized substrates under conditions of physiological levels of shear flow, a celladhesion-based chromatography microfluidic system (FIG. 9A) that exerts an almost uniform shear stress level across the width of the adhesive substrate was implemented.^31,32’³⁸ Calculated based on Stokes flow and experimentally validated (FIG. 8B), the channel was designed to allow cells to settle to the bottom of the channel prior to their encounter with the adhesive substrate, allowing uniform substrate contact for all perfused cells. With this system, cells that can mediate adhesion can thus be separated from non-inter-acting cells based on their channel elution time (FIG. 9B) and recovered for off-chip analyses.^31,32 Loaded cells were verified to interact only with the functionalized portion of the channel and, after recovery by flow-driven unloading, maintained high purity and viability (FIGS. 10A-10E). Demonstrating that cell adhesion in the perfusion channel is P-selectin dependent, pretreatment of cell suspensions for 30 min before perfusion with saturating levels (25 pg/mL) of P-selectin-Fc chimera decreased their adhesion to P-selectin substrates at 0.5 dyn/cm² (FIG. 9C). Furthermore, upon recovery from the channel, untreated and PMA/Ion-treated CD8⁺ T cells that were reperfused retained the general characteristics of the population from which they originally eluted, e.g., cells that were recovered from the free-flow (FF) fraction exhibited a higher frequency of cells in the FF rather than the adherent (Adh) fraction after reperfusion, and vice versa (FIG. 9D). To ensure yields sufficient for statistical comparisons between representative isolated populations, a WSS of 0.5 dyn/cm² was implemented for all chromatography experiments, given the higher extents of adhesion observed for CD8⁺ T cells of either treatment type, despite their highly divergent qualities.

[0166] Upon separation of cell suspensions into Adh and FF fractions by perfusion over P- selectin-functionalized channels, fractions enriched from untreated versus PMA/Ion-treated CD8⁺ T cells were found to exhibit different cellular phenotypes, with cells in the Adh fraction being enriched in CM and EFF cells (FIGS. 9E-9H). Conversely, levels of naive CD8⁺ T cells were diminished in the Adh compared with the FF population (FIGS. 9E-9H). These results match previous flow cytometry analyses of P-selectin ligand cells, where P-selectin ligandexpressing cells were found to be composed of more CM and EFF subtypes of CD8⁺ T cells (FIGS. 4H-4J) [0167] Evaluation of adhesion molecule expression in the various cell fractions revealed tumor homing molecules such as CD44, LFA-1, and P-selectin ligands to be upregulated in the Adh compared with the FF population of both untreated and PMA/Ion-treated CD8⁺ T cells (FIGS. 9I-9L). LN homing receptor

[0168] C-C motif receptor (CCR) 7 was also enriched in the Adh fractions of cells treated in either manner (FIGS. 9I-9L). To verify that this signal was not induced by stimulation of CD8⁺ T cells from P-selectin engagement, expression of CCR7 and other adhesion receptors was evaluated and found not to be upregulated by co-incubation for 1 or 4 h with P-selectin-Fc chimera (FIGS. 11A-11B). Furthermore, the effects of mechanical forces on CCR7 expression of cells were tested by perfusing cells through unfunctionalized channels and found to be negligible (FIGS. 11G-11H). Together, this shows that increases in the frequency of CCR7⁺ cells in the Adh population result from their functional enrichment from the parent population and not a cell signaling response triggered by hemodynamic force resulting from perfusion.

[0169] The concurrent versus individual expression of CCR7 and P-selectin ligand among parent and enriched CD8⁺ T cells of either source (untreated versus PMA/Ion-treated) was evaluated. Within the parent populations of both cell sources, most cells that were P-selectin ligand⁺ were found to be CCR7 negative (FIGS. 9M-9N). However, in the Adh population, the frequency of cells that were double positive for P-selectin ligand and CCR7 was increased (FIGS> 9M-9N). Subtype analysis also revealed P-selectin ligand⁺CCR7⁺ cells enriched within the different fractionation groups were not of a specific subtype and had no difference across the various fractionation groups (FIGS. 9P-9Q). The chromatography channel thus enriches for CD8⁺ T cells of a more differentiated phenotype that are enriched for co-expression of P-selectin ligand⁺ and CCR7 from source cells that vary substantially in their initial qualities. [0170] The ex vivo expansion capabilities of CD8⁺ T cells enriched from a P-selectin- functionalized channel were evaluated by incubating with Dynabeads and IL-2 immediately after perfusion. The cells expanded ~6-fold by day 8 irrespective of fractionation group (FIG. 12A), with changes in cell viability resulting from culture that were equivalent between fractionation groups (FIG. 12B). Further, CD8⁺ T cells positive for Ki-67, a proliferation marker,³⁹ were unchanged between different fractionation groups (FIG. 13C), as were granzyme-B and PD-1 (FIGS. 12D-12E). And despite P-selectin ligand expression being higher preexpansion (day 0) for the Adh fraction compared with other groups, differences were lost with expansion (FIG. 13F). Irrespective of fractionation group, cells thus exhibit the same proliferation and differentiation capabilities. [0171] Sorted CD8⁺ T cell fractions exhibit different in vivo tissue-homing capabilities '. The in vivo homing capabilities of cells enriched for adhesion to P-selectin in vitro using the adhesion chromatography system were evaluated by adoptively transferring by intravenous administration FF and Adh fractions of untreated or PMA/Ion-treated CD45.1⁺CD8⁺ T cells into B16F10 melanoma-bearing mice (FIG. 14A). Sixteen hours post-transfer, cells isolated from various tissues, including the tumor, spleen, TDLNs, and NDLNs, were stained using fluorescent antibodies and flow cytometrically analyzed. Irrespective of pretreatment condition, recovery of total CD45.1 cells in the tumor was highest for animals into which the Adh fractions were transferred compared with animals that received the parent population or FF fraction (FIG. 14B). Correspondingly, animals that received the Adh fraction with both cell sources yielded low recoveries in the spleen (FIG. 14C). The Adh fraction of untreated CD8⁺ T cells also had low yields in the NDLNs and TDLNs compared with other groups (FIG. 14C). As reflected in the fraction of total CD8⁺ T cells in each tissue, recipient animals that received donor cells from the Adh fraction exhibited higher frequencies of donor cells in the tumor compared with the parent and FF, irrespective of parent population pretreatment (FIG. 15A). In contrast, the frequency of donor cells in lymphoid tissues was lower in mice that received donor cells from Adh fractions compared with other groups, for both untreated and PMA/Ion populations (FIG. 15B). When evaluated for their distribution of CD8⁺ T cell subtypes, donor cells that trafficked to the tumor exhibited no differences in subtype across the fractionation groups, irrespective of activation state (FIG. 15C). Donor cells from the Adh fraction recovered from the spleen exhibited a lower frequency of naive cells compared with donors from the FF and parent cells for both pretreatment conditions (FIG. 15D). Overall, distributions of CD8⁺ T cell subtypes appear to be independent of fractionation group in the tumor while dependent on fractionation group in the spleen.

[0172] When evaluated for their expression of adhesion receptors/ligands, donor cells recovered from the tumor exhibited a high frequency of double positivity for P-selectin ligand and CCR7 expression irrespective of treatment or fractionation group (FIG. 14D). This trend was also observed with the P-selectin-functionalized channel where, irrespective of treatment group, the Adh fraction enriched for P-selectin ligand⁺CCR7⁺ CD8⁺ T cells. In contrast, cells recovered from lymphoid tissues varied based on their fractionation group, with the Adh fraction having a higher frequency of trafficked P-selectin ligand⁺CCR7⁺ cells and lower P- selectin ligand CCR.7 cells in the spleen and LNs (FIG. 14E, FIGS. 15E-15F). Furthermore, the P-selectin ligand⁺CCR7⁺ cells that were enriched in the tumor were not composed of a specific subtype (FIGS. 14F-14G), showing that adhesive quality and not CD8⁺ T cell subtype can be a better indicator of tumor homing capabilities. These data demonstrate that the in vitro selection for CD8⁺ T cells that can adhere to P-selectin in physiological shear flow enriches for a subset of cells both with an enhanced tumor homing capability and that mirrors the quality of CD8⁺ T cells enriched in the tumor compared with lymphoid tissues.

[0173] Ex vivo-expanded CD8⁺ T cells that exhibit higher extents of adhesion to P-selectin in flow in vitro home to greater extents to the TME in vivo: How the capacity to adhere to P- selectin in physiological fluid flow in vitro and home to tumors in vivo varies as CD8⁺ T cells are expanded was evaluated. Primary mouse CD8⁺ T cells were incubated with Dynabeads and IL-2 to yield cultures that expanded ~30-fold by day 10 (FIG. 16A). Adhesion by CD8⁺ T cells at various days of expansion was evaluated under static versus continuous flow conditions (FIGS. 17A-17H). Adhesion by untreated CD8⁺ T cells (day 0) was far less than that of cells at later expansion stages (day 2 and 8) under both static (FIG. 17D) and continuous flow conditions (FIG. 17H). In addition, cells at day 2 of expansion predominantly mediated rolling adhesion (FIG. 17B, FIG. 17F), whereas cells at day 8 interacted at high extents via firm adhesion (FIG. 17C, FIG. 17G). Correspondingly, the frequency of cells expressing P-selectin ligand was low prior to activation and expansion but dramatically increased by days 2 and 8 of expansion (FIG. 16B). Using the adhesion chromatography channel, the extent of CD8⁺ T cell adhesion to a P-selectin-functionalized substrate at a continuous WSS level of 0.5 dyn/cm² was also found to increase at days 2 and 8 compared with unexpanded (day 0) cells (FIG. 16C). Conversely, rolling velocity, which correlates with adhesion extent and cell avisity,³² decreased during expansion (FIG. 16C).

[0174] When CD8⁺ T cells were fractionated based on adhesion to P-selectin, more CD8⁺ T cells were recovered in the Adh fractions at days 2 and 8 of expansion relative to unexpanded cells (FIG. 16D). Cells recovered in the Adh fractions also expressed higher levels of P-selectin ligand compared with both other groups (FIG. 18A). However, after expansion, the fraction of cells recovered in the Adh population that expressed P-selectin ligand was the same as that of the parent population but exceeded that of the FF fraction (FIG. 18A). Adh cells at day 2 also expressed higher CCR7 (FIG. 18B), and L-selectin was expressed at lower levels compared with FF and parent throughout expansion (FIG. 18C). Irrespective of expansion stage, the Adh fractions contained a lower proportion of naive CD8⁺ T cells compared with the FF fraction and parent population, save at day 8 (FIG. 18D). Adh fractions exhibited no distinct differences in the proportion of CM cells (FIG. 18E). However, EFF CD8⁺ T cells were enriched in Adh versus FF fraction from populations that had been expanded, but not relative to the parent population of day 8 cultures (FIG. 18F). [0175] The homing of unexpanded and day 2- and 8-expanded CD45.1⁺CD8⁺ T cells adoptively transferred into B16F10 melanoma-bearing mice was next evaluated 16 h post- intravenous (i.v.) injection (FIG. 16E). With increasing day of expansion, trafficking to lymphoid tissues decreased, as reflected by the lower frequency of donor cells of total leukocytes (CD45⁺) in these tissues (FIG. 16F). Contrastingly, day 8-expanded CD8⁺ T cells accumulated to the greatest extent in the tumor (FIG. 16F). These differences in total infiltration also altered the recovery of CD8⁺ T cells in various tissues (FIG. 19A). A similar trend was seen with respect to the fraction of transferred cells, where at later days of expansion there was decreased accumulation in lymphoid tissues and increase accumulation in the tumor (FIG. 16G). Viability of the donor CD8⁺ T cells was decreased in TME compared with lymphoid tissues, irrespective of the day of expansion (FIG. 19B), as a result of the TME being immunosuppressive and rendering T cells functionally impaired.^{40 42}

[0176] To evaluate whether there are any adhesive markers or CD8⁺ T cell subtypes enriched in lymphoid tissues or the TME throughout expansion, the fold-change expression from pre-transfer levels of different markers and subtype was analyzed. The recovered cells in the TME were enriched for CCR7⁺ cells compared with untransferred on each respective expansion day, except day 8 (FIG. 19C). Of recovered cells, the proportion that were L- selectin in the TDLNs and spleen was increased relative to the parent population in day 8 cultures (FIG. 19D). Likewise, the proportion of cells expressing L-selectin recovered from tumors was diminished compared with their proportion in the donor cell population pretransfer (FIG. 19D). Similarly, the proportion of cells recovered from LNs that express P-selectin ligand was decreased in LNs irrespective of expansion day (FIG. 19E). In the adoptive transfer model, enrichment of P-selectin ligand-expressing cells in the TME was seen only in unexpanded cells, due to the fact that day 2- and day 8-expanded CD8⁺ T cells had higher levels of P-selectin ligand expression prior to transfer, resulting in the fold change relative to parent being low (FIG. 19E). Correlation analysis was performed to compare the enrichment of CD8⁺ T cells expressing adhesion markers that distribute in vivo to particular tissues versus those that are collected in the sorted cell fractions (FIGS. 13A-14B, FIGS. 13G-13J, FIGS. 16H-16I). In the tumor, increased expression of CCR7 and P-selectin ligand correlated with the fold change in expression of these markers in the Adh fraction but not the FF fraction (FIG. 16J). On the other hand, expression of L-selectin that is diminished in donor CD8⁺ T cells recovered from the TME correlated with the fold change in L-selectin expression by cells recovered from the FF fraction (FIG. 16J). The fold change in L-selectin expression by donor CD8⁺ T cells in lymphoid tissues was similarly correlated with the fold change in L-selectin expression by CD8⁺ T cells recovered from FF fractions (FIG. 13C, FIGS. 13O-13P). The fold change in CCR7 expression in donor CD8⁺ T cells recovered from lymphoid tissues also correlated with the fold change in CCR7 expression by CD8⁺ T cells in the Adh fraction (FIG. 13C, FIGS. 13O-13P) Likewise, the fold change in P-selectin expression by donor cells recovered from the TDLNs and spleen correlated with the fold change in cells recovered in the Adh fraction (FIG. 13C, FIG. 13P). This suggests that enrichment of cells expressing P-selectin ligand and CCR7 in the chromatography channel was predictive of enrichment in the TME of cells expressing these adhesion receptors, irrespective of expansion time.

[0177] When analyzed with respect to differentiation state, naive CD8⁺ T cells were enriched in LNs upon transfer of cells expanded until day 8 (FIG. 19F). In contrast, CM cells were enriched in the tumor only when transferred without expansion (FIG. 19G). On the other hand, EFF cells were enriched in the tumor at earlier days of expansion (day 0 and 2) (FIG. 19H) Correlation analysis of CD8⁺ T cell subtype within cultures of various dates of expansion that were found to be enriched in the in vitro fractionated cells versus those that were recovered from various tissues after adoptive transfer in vivo was performed (FIGS. 13D-13E, FIGS. 13K-13N, FIGS. 16K-16L) to determine if the chromatography channel predicted subtype enrichment in the TME. The diminishments in naive CD8⁺ T cells in the tumor correlated with the fold change in naive CD8⁺ T cells in both the FF and the Adh fractions (FIG. 16M). Naive CD8⁺ T cell enrichment in the LNs also correlated with the fold change in naive CD8⁺ T cells in both Adh and FF fractions (FIG. 13F, FIGS. 13Q-13R). EFF CD8⁺ T cells were correlated with the fold change of EFF cells in the FF but not the Adh fraction (FIG. 13F, FIGS. 13Q- 13R) Therefore, the fold change relative to parent in naive and, to some extent, EFF CD8⁺ T cells in the FF and Adh fractions predicts their fold change in the TME and lymphoid tissues. Overall, the enrichment of various sub-types of CD8⁺ T cells and the in vivo homing behavior of adoptively transferred CD8⁺ T cells at various states of expansion are predicted by their extent of enrichment for P-selectin adhesion in flow using the chromatography channel.

[0178] Tumor-specific CD8⁺ T cells enriched for adhesion to P-selectin in flow home to and remodel the TME and augment the efficacy of ACT with immune checkpoint blockade: The homing and engraftment of adoptively transferred CD8⁺ T cells that are tumor specific were evaluated. OT-I CD8⁺ T cells (CD45.2⁺) were enriched for adhesion to P-selectin using the adhesion chromatography system, and the FF and Adh fractions were collected and immediately transferred intravenously into B16F10-OVA melanoma-bearing mice (CD45.1) (FIG. 20A). At 16 and 64 h post-transfer, cells were isolated from the tumor, spleen, NDLNs, and TDLNs; stimulated in vitro with SIINFEKL for 6 h; antibody stained; and flow- cytometrically analyzed. Consistent with results from analogous experiments using polyclonal CD8⁺ T cells from wild-type (WT) CD45.1 mice, OT-I CD8⁺ T cells from the Adh fraction exhibited higher tumor trafficking compared with parent and FF fractions, reflected as the fraction of all CD8⁺ T cells and the fraction of the total number of transferred cells (FIGS. 20B-20C). Trafficking to secondary lymphoid tissues, on the other hand, was lower for cells in the Adh fraction compared with other groups (FIGS. 21A-21B). Viability of donor cells in the tumor was the same between fractionation groups at both 16 and 64 h post-transfer (FIG. 20D). Similarly, viability of donor cells in lymphoid tissues was the same between fractionation groups at 16 h, but in the spleen at 64 h, FF cells had higher viability compared with other fractions (FIG. 21C). At this initial analysis time point, all donor cell groups exhibited the same frequencies of granzyme-B, IFN-g, TNF-a, and Ki-67 expression in the tumor and lymphoid tissues (FIGS. 20E-20F, FIGS. 21D-21E), demonstrating that, despite differences in the extent of total homing to various tissues, donor CD8⁺ T cells that trafficked to the tumor have the same initial cytotoxic and proliferative capabilities. However, by 64 h post-transfer, TME-infiltrating donor cells from the Adh fraction exhibit increased frequencies of granzyme- B positivity (FIG. 20E). In the TDLNs (FIG. 20F) but not the NDLNs and spleen (FIGS. 21D- 21E) at 64 h, FF donor cells exhibited lower Ki-67 expression compared with Adh donor cells, demonstrating that Adh cells proliferate to a greater extent in the TDLN than FF cells. Furthermore, Ki-67 expression of donor cells decreased in the tumor and increased in the TDLNs, indicating this to be a site of donor cell proliferation (FIGS. 20E-20F).

[0179] When endogenous CD8⁺ T cells were evaluated, no differences in activation marker expression by CD8⁺ T cells infiltrating the TME were observed between mice that received different donor cells at 16 h (FIG. 20G). Endogenous CD8⁺ T cells recovered from the TDLNs of mice into which donor CD8⁺ T cells were transferred from Adh fractions, however, exhibited a higher frequency of IFN-g and TNF-a production, an effect that was sustained for IFN-g at 64 h (FIG. 20H). Given the TDLNs’ role in facilitating priming of lymphocytes in response to tumor-derived antigen, ^{43 46} this is suggestive of an anti-tumor CD8⁺ T cell response being locally elicited. Consistent with this, despite no differences at the initial 16 h time point, by 64 h post-transfer, frequencies of host CD8⁺ T cells producing granzyme-B and TNF-a infiltrating the tumor were increased for animals that received donor cells from Adh fractions (FIG. 20G). However, the overall frequency of granzyme-B-producing cells decreased between 16 and 64 h (FIG. 20G). Increased homing to tumors by tumor-specific donor CD8⁺ T cells enabled by enrichment for adhesion to P-selectin in flow drives faster initial tumor killing. Moreover, this homing drives further priming and expansion within TDLNs to facilitate anti-tumor CD8⁺ T cell immunity.

[0180] The effects of tumor homing differences on the therapeutic potency of donor tumorspecific CD8⁺ T cells were evaluated by intravenously infusing OT-I CD8⁺ T cells fractionated in the adhesion chromatography system into B 16F 10-0 VA tumor-bearing animals. The benefit of ACT in combination with blockade of PD-1, whose efficacy in melanoma is highly correlated with CD8⁺ T cell tumor infiltration,^47,48 was evaluated (FIG. 201) with or without concurrent anti -PD-1 (aPD-1) treatment, with effects compared with aPD-1 alone or saline control. The combi-nation therapy slowed tumor growth by day 11, and by day 13 both aPD-1 and ACT using Adh donor cells slowed tumor growth as a monotherapy, albeit at more modest extents compared with the two used in combination (FIG. 20J). When donor cell quality synergies with aPD-1 were evaluated, however, ACT with parent or FF donor cells did not improve the effects of aPD-1 (FIG. 20K). Therefore, cells enriched for their in vitro adhesion to P-selectin in flow with increased tumor homing in vivo improve the therapeutic synergies of ACT with aPD-1.

[0181] Enrichment of ex vivo-expanded human CD8⁺ T cells for in vitro adhesion to P- selecin in flow: The relevance of the adhesion chromatography system to the analysis of human CD8⁺ T cells expanded using a clinically relevant expansion protocol was evaluated. Human CD8⁺ T cells were isolated from healthy donor peripheral blood mononuclear cells and incubated with Dynabeads and IL-2. Analysis of expanded cells (FIG. 22A) sorted through a P-selectin-functionalized channel revealed that CD8⁺ T cell recovery within Adh fractions increased upon expansion (FIG. 22B). Adh fractions were also enriched for cells with P- selectin ligand expression, although the extent of enrichment was dependent on expansion day (FIGS. 22C-22E). Although no difference in CCR7 expression between sorted samples was found (FIG. 23 A, FIG. 23D), in contrast to results found with murine CD8⁺ T cells, L-selectin expression was higher in cells recovered in FF cells on days 5 and 7 of expansion (FIG. 23E), in agreement with results from murine studies. Sialyl Lewis a/x (sLe^a/x), a tetra-saccharide carbohydrate P-selectin ligand that plays an important role in selectin-mediated adhesion in flow,³⁵ was measured on fractionated groups and was found to be highly expressed by cells recovered in the Adh versus FF fraction (FIG. 23F). Overall, human CD8⁺ T cells fractionated for their capacity to interact with P-selectin in flow appear enriched for high expression of tumor homing molecules such as P-selectin ligand and sLe^a/x, while CD8⁺ T cells that do not interact with P-selectin are more frequently L-selectin positive. The adhesion chromatography system is thus amenable for analysis of human bio-specimens and, in line with results with a preclinical mouse tumor immunotherapy model, enriches for cells with high expression of tumor homing ligands.

Discussion

[0182] ACT has emerged as a promising therapy for metastatic melanoma, but this treatment has low rates of response due in part to poor cell trafficking to diseased tissues.^1,2 Understanding the mechanisms underlying CD8⁺ T cell infiltration to the TME holds promise for improving the clinical outcomes of ACT. In this study, an engineered microfluidic device was implemented to characterize what adhesion and chemokine receptors, as well as differentiation states, are associated with enhanced adhesion by CD8⁺ T cells to P-selectin in physiological flow. Biodistribution analysis of adoptively transferred cells into a preclinical B16F10 melanoma tumor model revealed the predictive benefit of cell adhesion to P-selectin in in vitro and in vivo tumor homing, which led to superior therapeutic effects in potentiating combination immunotherapy with aPD-1.

[0183] Preferential trafficking of P-selectin ligand⁺ CD8⁺ T cells to tumors seen herein is consistent with previous studies demonstrating that the expression of P-selectin and its ligands plays an important role in CD8⁺ T cell homing to tumors and in disease progression.^21,26 Results from in vitro flow-based perfusion experiments suggest that physiological force from fluid flow influences the mechanisms of CD8⁺ T cell adhesion to endothelial-expressed adhesive ligands. Furthermore, P-selectin ligand expression correlates with the extent of cell adhesion under conditions under which adhesion is initiated in the absence but not in the presence of flow. This is consistent with the potential for a greater number of low-affinity receptor-ligand interactions to occur under static conditions,⁴⁹ which presumably are not able to form under continuously applied shear stress. Therefore, continuous perfusion can be considered a more physiologically relevant method to investigate mechanisms underlying CD8⁺ T cell adhesion in vivo relevant to cellular homing during ACT.

[0184] A matter of controversy in the ACT field has been which differentiation state of CD8⁺ T cells contributes the most to immediate tumor killing versus long-term control of tumor growth.^1,3 Numerous studies have shown that less differentiated CD8⁺ T cells, such naive, stem-cell memory (SCM), and CM CD8⁺ T cells, have enhanced anti-tumor activity compared with effector memory (EM) and EFF CD8⁺ T cells.^3,14,50-52 This inverse relationship between T cell differentiation and treatment efficacy is believed to result from less differentiated cells having enhanced self-renewal and multipotent capabilities. However, less is known about the differences in cell adhesion and homing behaviors between CD8⁺ T cell subtypes and how this controls their trafficking and resulting functions. What subtypes of CD8⁺ T cells have enhanced cell adhesion to a tumor-like substrate was assessed here using an adhesion chromatography system that exposed cells to endothelial-presented adhesion receptors under physiological levels of shear flow. This approach offers an advantage over in vivo analysis methods in which the influences of cytokine stimulation or antigen presentation within the TME can lead to CD8⁺ T cell activation and differentiation, confounding the effects of adhesion and migration processes alone. This in vitro method instead allows the assay of CD8⁺ T cell adhesion to be done in a controlled manner, explored here in the context of engagement to P-selectin under the in-fluence of physiological levels of fluid flow. For non-activated CD8⁺ T cells, an enrichment of CM and EFF CD8⁺ T cells was found in Adh fractions, whereas only EFF cells are enriched in PMA/Ion-treated CD8⁺ T cells. Biodistribution analysis of the channel-sorted CD8⁺ T cells revealed that cells recovered in the Adh fraction trafficked better to the tumor and that the cells in the tumor exhibited higher proportions of cells that ex-pressed both P-selectin ligand and CCR7. These results show that the in vitro chromatography channel can be used to predict homing and engraftment behavior of CD8⁺ T cells in vivo.

[0185] When expanding murine CD8⁺ T cells ex vivo, P-selectin ligand expression was found to increase throughout expansion, leading to an increase in the number of CD8⁺ T cells mediating adhesion to P-selectin in flow and reducing their velocities of rolling adhesion. Using a P-selectin-functionalized channel to fractionate CD8⁺ T cells at different days of expansion revealed that P-selectin ligand⁺ cells were only enriched prior to expansion, signifying that sorted cells would have the greatest therapeutic benefit when starting from less differentiated cell sources. Less differentiated CD8⁺ T cells that are known to exhibit very potent survival and self-renewal capabilities,⁵³ however, tend to traffic more to LNs rather than tumors.^19,54 Using this system to sort less differentiated CD8⁺ T cell subtypes with greater tumor rather than LN homing can improve their therapeutic effects, an interpretation supported by tumor therapy experiments in combination with aPD-1.

[0186] When expanded cells were transferred into B16F10 melanoma-bearing mice, later days of expansion resulted in reduced LN homing, with subtle increases in tumor homing, results that closely match the in vitro data where later days of expansion resulted in a reduction of recovered cells in the FF fraction and increased recovery in the Adh fraction. Furthermore, the enrichment of P-selectin ligand and CCR7 expression by cells in the Adh fraction correlated with the enrichment of cells expressing these markers in the tumor. These studies show that the chromatography system can be employed with ex vivo expansion protocols to predict homing of CD8⁺ T cells throughout expansion. [0187] Immune checkpoint blockade is most effective in patients with tumors that are highly infiltrated by CD8⁺ T cells.^47,48 Blockade of PD-1 can restore the activation and cytotoxic capabilities of T cells to result in tumor control.^55,56 Many studies have demonstrated that aPD-1 treatment improves the potency of ACT with a high dose of transferred T cells.⁵⁷ ⁶¹ ACT comprising cells enriched for adhesion to P-selectin in flow improved the effects of aPD-1 to reduce tumor growth in the B16F10 tumor model. This shows that ACT with cells with greater adhesion propensity transforming scarcely infiltrated, immunologically “cold” B 16F 10 tumors into ‘ ‘warm” tumors, such that aPD- 1 anti -tumor effects are enhanced, thereby leading to improved therapeutic efficacy.

[0188] In conclusion, the capacity of CD8⁺ T cells to home and engraft within tumors versus lymphoid tissues can be modeled ex vivo using an engineered microfluidic device that recapitulates the hemodynamic microenvironment of the vasculature. Adhesion-based sorting of CD8⁺ T cells prior to transfer increases tumor homing and improves the therapeutic effects of ACT. Knowing what sub-population of CD8⁺ T cells homes better to the tumor, as well as determining the minimal timeline to produce cells enriched for this homing behavior, can enable dose sparing for ACT, thus minimizing undesirable side effects. An advantage of this method is its amenability not only to preclinical studies but also to investigations using human biospecimens. This approach, therefore, can improve the delivery limitations of ACT to increase treatment safety and patient response rates.

Example 2: Single-cell adhesive profiling in an integrated optofluidic device elucidates cellular phenotypes of CD8⁺ T lymphocytes associated with adhesion in inflamed vasculature-like microenvironments

[0189] To evaluate cellular characteristics associated with rolling adhesion to selectins in physiologically relevant flow conditions, an integrated optofluidic system was developed, combining an adhesion-based chromatography chip, previously developed to fractionate cells into adhesive versus non-adhesive subpopulations based on their elution time from the perfusion system, ^{61 63} and photoactivatable protein technologies,^64,65 to fluorescently “label” single-cell velocity as a retainable property of individual cells for off-chip analysis. The photoactivatable protein can be activated in a time dependent manner allowing the differentiation of cells with different adhesive rolling velocities in a given exposure window. Recovered, fractionated cells were thus fluorescently labeled according to their adhesive quality (e.g., velocity of rolling adhesion) from the chromatography channel and could be further counterstained with fluorescently tagged antibodies against various adhesive receptors and markers of phenotype and analyzed via multicolor flow cytometry for assessment of phenotypic and adhesive behavior of single cells in tandem. These multidimensional analyses elucidated that adhesion profiles and associated cellular phenotypes vary depending on the tumor-vasculature microenvironment, specifically E- versus P-selectin and at varying levels of wall shear stress (WSS). This revealed the velocity of CD8⁺ T cell rolling adhesion on E- and P-selectin in shear flow to be dependent on adhesion ligand/receptor(s) expression and not cell memory subtype. Overall, this integrated strategy enables the high-content assessment of cellular characteristics associated with heterogeneous cell adhesive behavior at the single-cell level under physiological hemodynamic conditions relevant to mechanistic and drug screening studies.

Materials and Methods

[0190] Murine CD8⁺ T cell isolation: UBC PA-GFP (purchased from The Jackson Laboratory (strain # 022486) and bred in-house) animals 6-10 weeks of age were euthanized, and the spleens were harvested and disrupted with 18G needles followed by washing with Dulbecco’s Phosphate-Buffered Saline (D-PBS). Cells were passed through a sterile 70-pm cell strainer, washed and incubated with red blood cell lysing buffer (Sigma-Aldrich) for 5 min at room temperature, quenched with D-PBS, washed, and resuspended for counting. Cells were resuspended at 10⁸ cells/ml buffer (Biolegend, MojoSort Buffer), and then incubated with a biotin-antibody cocktail for 15 min, followed by streptavidin nanobeads for another 15 min (Biolegend, MojoSort Mouse CD8a Selection Kit). Buffer was added to the mixture and placed in a magnet (STEMCELL Technologies), and the supernatant was collected. Cells were then counted and resuspended in 0.1% BSA for perfusion experiments or D-PBS for static photoactivation experiments.

[0191] Characterization of 405 nm light source and static photoactivation: The power output of a 405 nm laser (M405L2- UV Mounted LED, 1000 nA, 410 mW, ThorLabs, Newton, NJ), over a range of power settings, was determined by measuring power output at each setting using a power meter (3A High Sensitivity Thermal Sensor, Ophir Photonics, North Logan, UT). To measure the effects of different power settings on the extent of photoactivation, 5xl0⁵ CD8⁺ T cells from UCB PA-GFP mice in suspension in D-PBS were exposed to 405nm light at a range of power settings and exposure times in a 96-well plate. The extent of photoactivation was measured via flow cytometry using a customized BD LSRFortessa flow cytometer (BD Biosciences).

[0192] Flow-based cell adhesion experiments: A 0.01-inch thick silicone gasket with a 2.5 mm wide rectangular opening was assembled between an acrylic disk with an inlet and outlet ports (GlycoTech Corporation) and a functionalized polystyrene dish via vacuum suction. The chamber and inlet and outlet tubing were filled with 0.1% BSA in D-PBS perfusion medium, taking care that no bubbles formed at connections. The outlet line was connected to a syringe pump (PhD Ultra Harvard Apparatus), and the inlet line was connected to a reservoir. The chamber was placed on an Eclipse texas instrument (Ti) optical microscope (Nikon), and the medium was perfused through the chamber at the desired flow rate via syringe withdrawal. A suspension of 5 x 10⁵ PA-GFP⁺ CD8⁺ T cells per ml was added to the inlet reservoir. Image recording was done using Nikon NIS-Elements software. Six evenly spaced positions within the functionalized region of the dish were imaged for 30 seconds each, followed by checking non-specific adhesion on the non -functionalized region. For all experiments, the exposure time was 0.281 ps, the frame rate was 25 frames per second, and the objective was lOx.

[0193] Substrate functionalization: In experiments utilizing the vacuum-sealed gasket, a 1.07 x 0.25 cm rectangle in the center of a 35 mm non-tissue culture treated, round, polystyrene dishes were coated with anti-IgG (Fc specific) (Sigma-Aldrich) diluted in D-PBS without calcium and magnesium, at the concentration corresponding to the total P-selectin or E-selectin concentration in each condition. The anti-IgG solution was incubated overnight at 4°C, and then washed with D-PBS, blocked with 1% BSA in D-PBS at room temperature for 1 h, then washed, incubated at room temperature for 2 h with 1, 2.5, or 10 pg/ml P-selectin (R&D Systems), or 2.5, 10, 25 pg/ml E-selectin (R&D systems) diluted in D-PBS with calcium and magnesium. The non-functionalized region, a 0.53 x 0.25 cm rectangle immediately proximal to the functionalized rectangle, was blocked with 1% BSA in D-PBS. All dishes were stored at 4°C with D-PBS until use in same-day experiments.

[0194] Chromatography channel fabrication: U-shaped channel of two 2 cm wide by 14 cm long sections connected by a 2 cm wide 1.5 cm long section was cut using a crafted cutter (Silhouette America). On one side, double-sided adhesive tape (100 pm thick, 3M) was attached to PDMS (Ellsworth Adhesives). An inlet hole was made with a biopsy punch before attaching the other side of the adhesive tape to a non-tissue culture-treated polystyrene plate with a drilled outlet hole. PDMS was pre-made by mixing PDMS base with curing agent at a ratio of 9: 1 and curing for 4h at 90°C.

[0195] Chromatography channel functionalization: Fabricated chromatography channels were functionalized by incubating either 2.5 or 10 pg/ml anti-IgG (Fc-specific) (R&D Systems) in D-PBS without calcium and magnesium overnight at 4°C, and then washed with D-PBS, blocked with 1% BSA in D-PBS at room temperature for 1 h, then washes again. Then 2.5 ug/ml or 10 pg/ml of E-selectin or P-selectin was diluted using D-PBS with calcium, and magnesium was placed in the functionalized portion of the channel for 2 h at room temperature. Finally, the entire device was blocked with 1% BSA in D-PBS at room temperature for 1 h.

[0196] Validation of Protein Functionalization in Chromatography channel: The amount of selectin Fc-chimera (protein) adsorbed onto the chromatography channel was quantified using standard curve analysis by measuring the fluorescent intensity of the protein in solution when adsorbed into the chromatography channel.⁹⁸ Briefly, either selectin Fc-chimera was reconstituted at 1 mg/mL and fluorescently labeled using an Alexa Fluor 647 Conjugation Kit (Abeam). To generate a standard curve, solutions of varying concentrations of the selectin Fc- chimera were prepared for fluorescence imaging as wet spots. Wet spots were pipetted as bubbles on a 12 mm wide circle-sticker, which was cut on double-sided adhesive tape (100 pm thick, 3M) using a craft cutter (Silhouette America). Each wet spot was imaged under a Nikon Microscope at 3 locations under the PE-Cy5 channel. For chromatography channel measurements of protein adsorption, the channel was functionalized with varying concentrations of the fluorescently conjugated protein. Channel functionalization was conducted as described previously for perfusion experiments, with the exception of the last blocking step with 1% BSA. The functionalized channel was imaged at two FOV along six evenly spaced locations along the functionalized region. Finally, a standard curve of the protein solution fluorescence was calculated and utilized to map the fluorescence measurements in the channel to a numerical value of adsorbed protein (pg) per unit area (mm²).

[0197] Perfusion workflow: Perfusion experiments were performed as previously described.^66,67 Briefly, an inlet syringe connected to tubing was filled with perfusion media (0.1% BSA in D-PBS), and was connected to a syringe pump (PhD Ultra Harvard Apparatus). The syringe pump was used to withdraw a cell pulse of 200 pl at a concentration of 2.5 x 10⁶ cells/ml into inlet tubing at a rate of 0.4 ml/min. The tubing containing the cell pulse was inserted into the inlet hole of the channel, and a 5 ml test tube was connected to the bottom of the outlet hole as the cell collection reservoir. The channel was placed on an Eclipse Ti optical microscope (Nikon) to acquire videos. To begin perfusion, the syringe pump was set to 0.5 or 1 dyn/cm², perfusion was then stopped after the free flow cells elution time had been reached. Then, the syringe and inlet tubing was replaced with a new syringe and tubing containing only perfusion media to eject the adherent cell in the channel out of the channel and into a second collection tube. The number of cells in the sorted fraction was counted using a hemocytometer and then analyzed via flow cytometry.

[0198] For experiments that included photoactivation a 405nm light source with a mask 1.2 cm wide and 1 cm long was placed under the channel, directly upstream of the microscope objective, at the beginning of the functionalized portion of the channel to complete the system. At the beginning of the perfusion, the 405nm was turned on at power setting 5 and stayed on throughout the entirety of the experiment.

[0199] Cell adhesion and velocity analysis: Videos from the flow-based cell adhesion experiments and chromatography channel were manually post-processed by counting the number of rolling and firmly adherent. For the context of this work, a firm adherent cell was defined as a cell that interacted with the selectin-functionalized substrate and has stopped moving in the direction of flow. A rolling cell was defined as one that interacted with the selectin-functionalized substrate but continued to translate in the direction of flow at a velocity substantially slower than cells in the free flow stream. Free flow cells, was defined as cells that do not mediate specific, sustained adhesive contact with the substrate throughout the duration of its transit through the channel. Video analysis of cell rolling velocities was performed with ImageJ (National Institutes of Health) with a manual particle tracking plugin. Velocity was determined by dividing a cell's total translational distance over the functionalized area by the total time it took to travel that same distance.

[0200] Comparison between the extent of adhesion of non-photoactivated cells and prephotoactivated cells: PA-GFP CD8⁺ T cells were either left unphotoactivated or were photoactivated in a 96-well plate, with a 405 nm laser at power setting 5 for 5 min. Unphotoactivated cells or photoactivated cells were then perfused through a 10 pg/ml P- selectin or 2.5 pg/ml E-selectin functionalized channel as described above. The extent of adhesion from both populations was subsequently analyzed and compared.

[0201] Flow cytometry analysis: Collected cell solution fractions were centrifuged at 400 x G for 5 min and resuspended in diluted anti-mouse CD16/CD32 (clone, 2.4G2) (Tonbo Biosciences) blocking solution for 5 min on ice, washed, and stained with fixable viability dye Zombie Red for 30 min at room temperature, and then washed. Cells were then incubated with 10 pg/ml P-selectin or E-selectin plus 10 pg/ml of PE anti-IgG (Fc specific) diluted in D-PBS for 30 min on iced and then washed. Antibody cocktails were prepared in flow cytometry buffer (0.1% bovine serum albumin in D-PBS) following manufacturer concentration or preliminary titrations; all antibodies for flow cytometry were from Biolegend unless otherwise stated. Cells were incubated with an antibody cocktail for 30 min on ice, washed, and then resuspended in buffer for analysis. Cells were analyzed with a customized BD LSR Fortessa flow cytometer (BD Biosciences). Compensation was performed using ArC or UltraComp compensation beads (ThermoFisher), and data were analyzed using Flow Jo software version 10. Results

[0202] Heterogeneous adhesion phenotypes of murine CD8⁺ T cells to P- or E-selectin in flow: Adhesion by CD8⁺ T cells to P- or E- selectin functionalized substrates in hemodynamic flow was assessed using a parallel plate flow chamber and measured by videomicroscopy. CD8⁺ T cells were negatively isolated from spleens of naive mice transgenic for photoactivatable (PA) green fluorescent protein (GFP) (for reasons elaborated below) and perfused at a concentration of 5 x 10⁵ cells per ml. The number of adhered CD8⁺ T cells at six evenly spaced positions within the functionalized region of the flow chamber were imaged for 30 seconds each. Adhesion toP- and E-selectin in flow was found to be highly WSS-dependent, being the highest at the lowest tested WSS (0.5 dyn/cm²) and decreasing to zero above 2.0 dyn/cm². The extent of adhesion to P-selectin was also found to be highly P-selectin concentration-dependent, with P-selectin-Fc chimera suspensions used to coat the adhesive substrates of 1 pg/ml resulting in lower extents of adhesion compared to 2.5 and 10 pg/ml. In contrast, E-selectin adhesion was found to be concentration-independent with different E- selectin-Fc chimera suspensions used to generate the adhesive substrates resulting in similar numbers of adherent T cells. Overall, adhesion quality is highly dependent on the type of presented selectin, its concentration and WSS level.

[0203] CD8⁺ T cell rolling velocities were evaluated using an adhesion chromatography channel previously developed (FIG. 24A).^61,63,66,67 Via its settling region, the channel’s validated design ensures uniform contact of all perfused cells, including CD8⁺ T cells, with the adhesive substrate (FIG. 24B), ^61,63 wherein substrate protein concentration scales linearly with the concentration of the protein solution used for functionalization, and enables the fractionation and recovery of cell subpopulations (free flow, FF, versus adherent, Adh) due to differences in elution time for off-chip analysis (FIG. 24C).^42,61,66 As such, this system is not limited to image-based analyses of conventional parallel plate studies, an important property to evaluate cellular characteristics associated with adhesion phenotypes. However, interactions by CD8⁺ T cells to selectin-functionalized substrates in hemodynamic flow is highly heterogeneous, with not only a subset of perfused cells mediating any adhesion but also wide variations in the rolling velocities within the population of CD8⁺ T cells that can adhere to P- and E-selectin substrates that vary by substrate concentration and WSS (FIG. 24F). Given this, the heterogeneity of adhesive phenotypes exhibited by CD8⁺ T cells that interact with selectins in physiological flow could not be captured using the adhesive chromatography channel alone due to its binary enrichment of all adhered versus free cells without respect to adhesive quality (e.g. velocity of rolling adhesion, firm adhesion, etc). Therefore, a photoactivation platform was integrated to photoactivate cells proportional to their time spent in the exposure window (FIGS. 24A-24B, FIG. 24G). Given that WSS is nearly uniform across the bottom width of a parallel plate flow chamber,⁷¹ the number of adherent cells does not vary across the width of the channel. As the 405 nm light spans the entire channel width, all cells perfused through the exposure window are thus exposed to photoactivating light for identical channel lengths independent of cell position across the channel width. This integrated approach allows singlecell labeling based on cellular velocity, enabling the off-chip flow cytometric analysis of cellular characteristics associated with different adhesive phenotypes (FIG. 24H).

[0204] Photoactivation of PA-GFP -expressing CD8⁺ T lymphocytes'. It has been previously reported that the PA-GFP MFI of engineered cell lines can be photoactivated with violet light (380-450 nm), increasing their GFP signal intensity in proportion to light exposure (modulated by time or power), providing a very high contrast between activated and unactivated cells.^64,65 Likewise, primary murine CD8⁺ T cells isolated from PA-GFP mice increase their GFP signal with increased exposure to 405 nm light and compared to nonexposed cells in Bright Field (BF) (FIG. 25A). All cells containing PA-GFP emit a fluorescence in the native baseline state (measured in BV510) that is distinct from the fluorescence in the activated state (GFP).⁶⁸ Thus, to account for signal variability between individual cells, the fluorescent intensity of unactivated GFP (measured in BV510) for any individual cell was used to normalize the measured signal of activated GFP, reducing the coefficient of variation from ~17 to ~7 (FIGS. 25B-25C) The extent of photoactivation measured by GFP positivity was measured on a single-cell basis using flow cytometry at different times of exposure and power outputs, showing increased fraction of GFP⁺ cells with increased power as well as exposure time (FIG. 25D). Within these GFP⁺ CD8⁺ T cells, normalized GFP/unactivated signal mean fluorescence intensity (MFI) increased greatly (2-6- fold) at high power outputs above 200 mW with increased exposure time (FIG. 25F). Moreover, the change in normalized MFI at these high-power outputs of 208 and 235 mW was time-dependent (FIG. 25F), which enables the ascribing of velocity as a retainable single-cell “label” based on residence time on the exposure window.

[0205] The possible effects that photoactivation might have on viability, selectin ligand expression and functional adhesion were evaluated. Viability was shown to decrease after 8 min of 405 nm light exposure with power outputs of 208 and 235 mW. Due to the significant loss of viable cells at 235 mW, the power output of 208 mW was chosen for all future experiments, to minimize viability effects and maximize the time-dependent changes in GFP MFI. PA-GFP⁺ CD8⁺ T cells were exposed to a 405 nm light source with a power output of 208 mW at varying times and the expression of both P- and E- selectin ligand was analyzed by staining with either recombinant P- or E-selectin-Fc chimera. Expression of both P- and E- selectin ligands did not change with increased exposure time. Adhesion to both P- and E- selectin after photoactivation was assessed as well. PA-GFP⁺ CD8⁺ T cells were either left unphotoactivated or photoactivated for 5 min with a 405 nm light source at the power output of 208 mW. Both unphotoactivated and photoactivated cells were then perfused through a P- or E-selectin functionalized channel and the extent of adhesion was analyzed. No statistically significant difference in rolling or firm adhesion to both selectins with either the unphotoactivated or photoactivated cells was observed. Among adherent cells collected for post-hoc analysis, rolling (those with non-zero velocities) and firm (those with zero-velocities), firm adherent cells would exhibit the highest level of photoactivation as a result of their residence time in the exposure window. Together, these data demonstrate photoactivation does not affect PA-GFP⁺ CD8⁺ T cell expression of P- and E-selectin ligands and adhesion to P- and E-selectin in shear flow.

[0206] To ensure that the custom designed 1 cm exposure window would provide sufficient time for activation, the residence time of cells was estimated based on average rolling velocities of cells as measured on P- and E-selectin. Using a 10 pg/ml P-selectin functionalized channel, the average rolling CD8⁺ T cell perfused at either 0.5 or 1.0 dyn/cm² would spend about 5 min on the exposure window. For a 2.5 pg/ml E-selectin functionalized channel, the estimated exposure time at 0.5 dyn/cm² is 9 min and at 1 dyn/cm² it is about 30 min. However, based on the experiment perfusion time through the functionalized portion of the channel (the time it takes for free flow cells to enter and exit the functionalized potion of the channel) having a duration of 18 and 9 min at 0.5 and 1 dyn/cm², respectively, indicated by dashed lines. The percentage of cells that would fall within a different exposure window was also analyzed. The proportion of adhesive CD8⁺ T cells that would spend 5-9 min in the planned exposure window of 1 cm was assessed for each WSS, a time-range sufficient to achieve CD8⁺ T cell photoactivation (FIG. 25D), but below the time point at which cell viability starts to decline. Specifically, of the cells rolling on P-selectin, 20% and 30% of rolling cells within perfusion times for 0.5 and 1.0 dyn/cm², respectively, fall within this time window. On E-selectin, approximately one third of rolling cells perfused at either 0.5 and 1.0 dyn/cm² fall within this time window. Thus, for perfusion times of 18 and 9 min at 0.5 and 1.0 dyn/cm², respectively, a substantial proportion of cells are sufficiently photoactivated to yield measurable changes with respect to background without declines in viability. Summarily, PA-GFP⁺ CD8⁺ T cell photoactivation is dependent on 405 nm light exposure power and time, photoactivation does not affect cell adhesion, selectin ligand expression, or viability under the perfusion conditions used herein where a sufficient proportion of cells mediating rolling adhesion could be labeled via the photoactivation method.

[0207] Photoactivation differentiates CD8⁺ T cell subpopulations with distinct selectin adhesion phenotypes: The photoactivation platform was integrated with the adhesion-based chromatography channel to fluorescently “label” cells in proportion to the length of time they spend in the exposure window interacting with selectin-functionalized surfaces in hemodynamic flow. To this end, PA-GFP⁺ CD8⁺ T cells were perfused within the channel functionalized with selectin protein chimeras (FIGS. 24A-24B) at physiological levels of WSS, 0.5 and 1.0 dyn/cm², approximating those experienced in blood and lymphatic vessels.^69,70 The 405 nm light source at 208 mW power with a 1 cm mask was placed at the beginning of the functionalized channel area upstream of the imaging field of view (FOV) (FIGS. 24A-24B). This allows photoactivation of cells mediating adhesion at velocities below a photoactivation threshold within the 405 nm light exposure window (FIG. 24F), approximately 100-200 pm/sec (FIG. 26A). Fractionation based on channel elution time also allows discrimination based on whether adhesion was sustained for prolonged times along the channel length, with cells thus eluting into the Adh fraction, versus non-persistent cells that instead elute into the FF fraction (FIG. 24F).

[0208] To functionally validate this integrated system, the extent of photoactivation of FF versus Adh cells fractionated on a 10 pg/ml P- or 2.5 pg/ml E-selectin functionalized channel was examined. For both P- and E-selectin at 0.5 and 1 dyn/cm², there was a higher percentage of GFP⁺ cells in the FF versus the Adh fractions. However, the percent of GFP⁺ cells between FF cells and cells recovered from an unfunctionalized channel (UnF) was not different, demonstrating that photoactivation in the FF fraction is due to the high sensitivity of the cells to the 405 nm light source under shear flow. Indeed, the average velocity of FF cells was -150 pm/sec, meaning a substantial portion of perfused cells not mediating adhesion will experience photoactivation. However, the normalized MFI GFP/unactivated signal was higher in the Adh fractions versus FF at 0.5 and 1 dyn/cm² (FIGS. 26D-26G), indicating that the Adh cells spend more time in the exposure window compared to FF. Additionally, the percentage of adherent cells that were photoactivated was dependent on the length and time interval in which cells reside in the exposure window. Given the 1 cm exposure window present along the 14 cm long functionalized region, the theoretical percentage of adherent cells that are photoactivated through this region is 7%. Furthermore, this calculation approximates the experimental value of %GFP⁺ of PA-GFP⁺ T-cells in the adherent fraction across different perfusion conditions of WSS and selectin. These results validate that the extent of photoactivation can be used as a proxy of residence time and therefore velocity of the cells in the exposed window. Finally, the high viability of the cells perfused through this integrative system was maintained regardless of the experimental conditions used.

[0209] Selectin ligand expression level is associated with decreased velocity of CD8⁺ T cell rolling adhesion on P- and E-selectin in shear flow: The relationship between selectin ligand expression and rolling velocity was evaluated at a single-cell level by staining perfused and photoactivated PA-GFP⁺ CD8⁺ T cells with recombinant P- or E-selectin Fc chimera (FIG. 27A). When looking at all PA-GFP⁺ cells, the Adh fraction has a higher percent P-selectin ligand⁺ cells compared to FF cells at 1 dyn/cm² (FIG. 27C). Further, for both shear stresses, P-selectin ligand MFI was higher in the Adh versus FF fraction (FIG. 27D). For cells enriched in an E-selectin functionalized channel, the Adh fraction exhibited a higher percentage of E- selectin ligand⁺ cells compared to FF (FIG. 27F). E-selectin ligand MFI was also higher in the Adh fraction compared to FF (FIG. 27G). Furthermore, selectin-ligand MFI (normalized to unperfused) for the FF and Adh fractions show values less than and greater than one, respectively, for both WSS values and selectins (FIG. 27D, FIG. 27G). Altogether, these results indicate that Adh cells have higher selectin ligand expression compared to FF.

[0210] Next, the difference in selectin ligand expression of fast (lower GFP expression, GFP⁺L) versus slow-rolling cells (higher GFP expression, GFP⁺H) was evaluated (FIG. 27H). On P-selectin functionalized channels, only slow-rolling cells (GFP⁺H) had a higher percent of P-selectin ligand⁺ cells compared to FF cells and cells recovered from an unfunctionalized channel (UnF) at 1 dyn/cm² (FIG. 271). On E-selectin functionalized channels, both fast and slow rolling cells had a higher percentage of E-selectin ligand⁺ cells compared to FF and UnF, at 1 dyn/cm² (FIG. 27J). On E-selectin functionalized channels, the percent of E-selectin ligand⁺ cells between fast (GFP⁺L) and slow (GFP⁺H) rolling cells was also significantly different at 1 dyn/cm² (FIG. 27J). Overall, this shows that adhesion to P- and E-selectin in shear flow is increased in cells exhibiting higher levels of selectin ligand expression, further validating this integrated system’s performance.

[0211] Photoactivation identifies subtypes of CD8⁺ T cells mediating slow rolling adhesion on P- or E-selectin in shear flow: CD8⁺ T cells subtypes [naive (CD62L⁺CD44 ), central memory (CM) (CD62L⁺CD44⁺), effector (EFF) (CD62L CD44⁺), and double-negative (CD62L'CD44‘)] were analyzed from recovered perfused and photoactivated cells to determine if CD8⁺ T cells subtypes adhere with different frequencies and different qualities to selectins. When looking at GFP⁺ cells, the frequency of naive CD8⁺ T cells decreased in the Adh fraction for both WSS on P-selectin channels (FIGS. 28A-28C). Further, CM CD8⁺ cells increased in the Adh fraction compared to FF of GFP⁺ cells enriched in a P-selectin functionalized channel at 1 dyn/cm² (FIG. 28C). For GFP⁺ enriched in an E-selectin functionalized channel, the frequency of naive CD8⁺ T cells decreased in the Adh versus FF fractions (FIGS. 28D-28F). Furthermore, cells enriched on E-selectin at 0.5 dyn/cm² exhibited increased frequencies of effector CD8⁺ T cells (EFF), and at 1 dyn/cm² increased frequencies of CM cells in Adh versus FF fractions (FIGS. 28E-28F). Overall, naive CD8⁺ T cells exhibited reduced frequencies of photoactivation while CM and EFF cells exhibited higher frequencies of photoactivation in Adh versus FF fractions, indicative of a higher quantity of differentiated CD8⁺ T cell subtypes mediating slow rolling adhesion (FIGS. 28A-28F).

[0212] The difference in rolling velocities between different CD8⁺ T cells subtypes was analyzed by quantifying normalized GFP expression of GFP⁺ CD8⁺ T cells of the various subtypes. CD8⁺ T cells within the FF fraction exhibited very low GFP expression, across all subtypes perfused on both P- and E- selectin at various WSS. In the Adh fraction, more differentiated CD8⁺ T cells exhibit a slightly higher normalized GFP expression compared to less differentiated CD8⁺ T cells, but at levels that are not statistically significant on both P- and E-selectin at various WSS. Overall, these data demonstrate that while the quantity of photoactivated cells differed between CD8⁺ T cell subtypes, among the photoactivated cells there was no difference in the extent of photoactivation, indicating negligible differences in adhesive velocities between various CD8⁺ T cell subtypes.

[0213] Complex relationships between adhesion behaviors and ligand/receptor expression revealed by photoactivation of perfused CD8⁺ T cells in shear flow: Given that CD8⁺ T cells express a variety of adhesive ligand/receptors with redundant, interdependent, and/or overlapping functions, the integrated chromatography and photoactivation platform was implemented to explore how these ligands and receptors contribute to cell adhesive interactions with selectins in the context of fluid flow. PA-GFP⁺ CD8⁺ T cells were perfused over selectin- functionalized substrates under exposure to 405 nm light and sorted into FF and Adh fractions. The FF and Adh fractions were subsequently labeled simultaneously with fluorophore- conjugated antibodies specific for six different adhesive ligands/receptors to analyze the expression of adhesion molecules with respect to the adhesive phenotype (FIG. 29A). Recovered cells were gated based on the level of photoactivation measured by GFP, GFP' being no photoactivation and GFP⁺ L/H being cells where photoactivated at low versus high levels (FIG. 29A). Adh GFP⁺H cells recovered from a P-selectin functionalized channel had lower frequency of L-selectin⁺ compared to cells recovered in the UnF, FF, or Adh GFP' or GFP⁺L at 0.5 but not 1 dyn/cm² (FIG. 29B). Further, the frequency of L-selectin on GFP⁺H cells was lower at 0.5 than 1 dyn/cm² (FIG. 29B). The frequency of CXCR5⁺ cells was also higher in the Adh GFP⁺H compared to other groups at 0.5 but not 1 dyn/cm² (FIG. 29B). On the other hand, the frequency of LFA-1 was higher in the Adh GFP⁺ H compared to the other groups for 1 but not 0.5 dyn/cm² (FIG. 29B). In contrast to P-selectin, channels functionalized with E- selectin resulted in enrichment of different markers in the slow rolling adherent cells (GFP⁺ H) (FIGS. 29B-29C). At 1 dyn/cm², the frequency of C-C chemokine receptor type 7 (CCR7), a receptor for lymph node homing chemokine (C-C motif) ligands 21 and 19,⁷² was higher for the Adh cells compared to UnF and FF, however it was only statistically significant with the Adh GFP- group (FIG. 29C). Moreover, similar to P-selectin functionalized channels, the frequency of L-selectin decreased in the Adh GFP⁺H compared to other groups at 0.5 but not 1 dyn/cm² on E-selectin functionalized channels (FIG. 29C). The frequency of C-X-C motif chemokine receptor (CXCR3), a receptor for C-X-C motif chemokines 9, 10 and 11 involved in tumor intravascular adhesion and extravasation,¹⁴ was higher in the Adh GFP⁺L and GFP⁺H compared to other groups at 1 but not 0.5 dyn/cm² (FIG. 29C). The frequency of CXCR5 was also higher in the Adh GFP⁺H group compared to other groups at 1 but not 0.5 dyn/cm² (FIG. 29C). Further, the frequency of CXCR5 in the Adh GFP⁺H group was higher at 1 versus 0.5 dyn/cm² (FIG. 29C). The frequency of CD44, a known ligand of E-selectin³⁵, was also higher in the Adh fraction recovered from an E-selectin functionalized channel versus other groups (FIG. 29C). At 0.5 dyn/cm², Adh GFP⁺L cells had a higher frequency of CD44 positivity compared to UnF and FF GFP' cells, while at 1 dyn/cm², Adh GFP⁺L and GFP⁺H both had higher frequencies of CD44 compared to UnF and FF GFP' cells (FIG. 29C). Finally, LFA-1 frequency was also higher in the Adh GFP⁺L and GFP⁺H compared to other groups at 0.5 dyn/cm², but only Adh GFP⁺L cells have higher LFA-1 frequency compared to other groups at 1 dyn/cm² (FIG. 29C). Overall, this data demonstrates that enrichment of adhesive markers in sorted and photoactivated cells to be highly dependent on both the biochemical and biophysical environment of the recapitulated inflamed vasculature.

[0214] The relationships between mean GFP/unact. signal fluorescence and adhesive ligand/receptor expression of GFP' FF cells and GFP⁺ Adh cells subdivided into different gates were next assessed (FIG. 30A). In this analysis, a statistically significant non-zero slope between mean GFP/unact. signal fluorescence and adhesive ligand/receptor expression reflects the relative impact that adhesive molecule expression has on the rolling velocity. PA-GFP⁺ CD8⁺ T cells perfused through a P-selectin functionalized channel showed an increase in P- selectin ligand normalized MFI with increased normalized GFP expression at both 0.5 and 1 dyn/cm² (FIG. 30B). Surprisingly, normalized MFI of CCR7 also increased with normalized GFP expression (FIG. 30C). CD8⁺ T cell expression of CXCR5, linked to stem cell memorylike properties⁷³ with enhanced anti-tumor activities⁷⁴ that are enriched within the tumor microenvironment⁷⁵ and tumor-draining lymph nodes,⁷⁶ increased with increased normalized GFP expression (FIG. 30D). L-selectin normalized MFI on the other hand decreased with increased normalized GFP expression at 0.5 dyn/cm². While LFA-1 normalized MFI increased with increased normalized GFP expression. Finally, CD44 and CXCR3 normalized MFI did not change as a function of normalized GFP expression.

[0215] PA-GFP⁺ CD8⁺ T cells perfused through an E-selectin functionalized channel showed an increase in E-selectin ligand normalized MFI with increased normalized GFP expression at 1 dyn/cm² (FIG. 30E). CCR7 normalized MFI did not change as a function of normalized GFP expression (FIG. 30F). CXCR5 normalized MFI increased with increased normalized GFP expression at 1 dyn/cm² (FIG. 30G). L-selectin normalized MFI decreased with increased normalized GFP expression at 0.5 dyn/cm². While CD44 normalized MFI increased with increased normalized GFP expression at 1 dyn/cm², CXCR3 and LFA-1 normalized MFI did not change as a function of normalized GFP expression on cells enriched in an E-selectin functionalized channel. Overall, the relationship between rolling velocity (measured by normalized GFP expression) and adhesion ligand/receptor expression varied depending on channel substrate and WSS (FIGS. 30H-30I). The velocity of rolling adhesion on P-selectin in shear flow was inversely correlated with CD8⁺ T cell expression of P-selectin ligand, CCR7, and CXCR5, and LFA-1. On the other hand, the velocity of rolling adhesion on E-selectin was inversely correlated with L-selectin expression at 0.5 dyn/cm², and E-selectin ligand, CD44 and CXCR5 expression at 1 dyn/cm². Altogether, these results reveal that the expression of adhesion receptors not directly implicated in selectin-mediated molecular recognition are nevertheless enriched in CD8⁺ T cells capable of mediating adhesion to P- and E-selectin in a manner regulated by physiological levels of shear flow.

[0216] FIGS. 31A-31C, FIGS. 32A-32B, FIGS. 33A-33B, FIGS. 34A-34C, FIGS. 35A- 35F, FIGS. 36A-36D, FIG. 37, FIGS. 38A-38E, FIGS. 39A-39D, and FIGS. 40A-40H show additional experimental results.

Discussion

[0217] Successful trafficking of CD8⁺ T cells across the inflamed vasculature is recognized as a key determinant of immunity.^28,29’³⁵ However, to date, there are limited methods to analyze the cellular phenotypes associated with the breadth of adhesive behaviors from a heterogeneous population. Most methods used to analyze cellular mediators associated with adhesion and trafficking of CD8⁺ T cells require the need to pre-sort cells, knockout models, or antibody inhibition, which are time-consuming methods and provide binary results.^4,37,77-79 Here, a methodology was implemented that integrates a microfluidic device that recapitulates components of the hemodynamic environment of the circulatory system^61,80 with a photoactivation platform^42,64,65,67 to rapidly “label” and enrich CD8⁺ T cells in a manner that directly reflects the extent and velocity in which they mediate rolling adhesion. This technique enables off-chip multidimensional analysis of adhesion ligand/receptors expression associated with adhesive phenotype, providing an insight into molecular underpinnings responsible for the diversity in adhesive phenotypes with a degree of resolution unattainable by previous methods.

[0218] This integrated photoactivation and microfluidic platform enabled the analysis of relationships between various cellular phenotypes (namely adhesion receptor expression and differentiation markers) and rolling adhesion behavior of CD8⁺ T cells on selectins in shear flow. Cells mediating slow rolling adhesion were characterized as having a higher proportion of selectin ligand⁺ cells as well as a higher density of selectin ligand expression (as measured by MFI). This is consistent with previous studies that demonstrate that selectins and their ligands play an important role in CD8⁺ T cell homing to inflamed tissues and tumors as well as in disease progression ^45,48 This work also reveals the effect that WSS on features of cells capable of sustaining adhesion under higher levels of WSS, with increased WSS not only decreasing the number of adherent cells but enriching for cells with higher selectin ligand expression, demonstrating that increased ligand density is necessary to overcome the force from fluid flow.

[0219] There is extensive debate as to which subtype of CD8⁺ T cell provides better antitumor effects in the context of both disease control by the patient’ s endogenous immune system as well as cancer immunotherapies.^81,82 The abundance of stem-like CD8⁺ T cells in tumors is correlated with better disease control.^83,84 The presence of stem-like CD8⁺ T cells in the tumor is also associated with better clinical response to immune checkpoint blockade therapy.^84-87 Furthermore, in advanced melanoma and non-small cell lung cancer, the frequency of CM cells in the blood has shown to be a positive predictor of anti-PD-1 therapy and survival.^84,88 In the context of adoptive cell therapies expanded from tumor infiltrating lymphocytes, less differentiated CD8⁺ T cells such as naive, SCM, and CM exhibit enhanced antitumor activity compared to effector memory (EM) and EFF CD8⁺ T cells.^22,82,89-92 These studies have been performed in vivo where it is hard to uncouple whether the effects are due to enhanced selfrenewal and multipotent capabilities or due to enhanced tumor homing. To address this gap, what subtypes of untreated CD8⁺ T cells derived from spleens of C57B16 mice have enhanced cell adhesion behavior to an inflamed-like substrate was assessed here using the integrated system of adhesion chromatography and photoactivation under physiological levels of shear flow. An advantage of this system is that it can assay the adhesive behavior of all CD8⁺ T cells subtypes from a heterogeneous population without the need to pre-sort the cells. These results reveal there to be a lower frequency of naive CD8⁺ T cells mediating slow rolling adhesion compared to the frequency of naive cells in FF. Moreover, within both the FF and Adh fractions, no difference in the rolling velocities of various CD8⁺ T cell subtypes that had photoactivated was found. This shows, at least in an untreated CD8⁺ T cell population derived from mouse spleens, that adhesion ligand/receptor expression to be more indicative of CD8⁺ T cell adhesive phenotype than subtype, which is relevant in how cell therapies are developed and screened both preclinically and bedside.

[0220] Leveraging the system's compatibility to enabling off-chip analysis using single cell profiling techniques, complex, multidimensional relationships between other adhesion ligand/receptors and the ability of CD8⁺ T cells to mediate rolling adhesion were interrogated. First, L-selectin expression correlated with rolling velocity on both P- and E-selectin at 0.5 dyn/cm², a result consistent with more differentiated CD8⁺ T cells mediating P- and E-selectin adhesion in shear flow. CD8⁺ T cells mediating slow rolling on P-selectin also exhibited increased expression of LFA-1, that can engage with ICAM locally presented by the inflamed vasculature to mediate cell arrest in flow and the initiation of cell extravasation.^55,93’⁹⁴ More interestingly, rolling velocity on P-selectin correlated inversely with expression of CCR7, the canonical lymphoid tissue homing receptor⁹⁵’⁹⁶ for lymphoid chemokines CCL19 and CCL21 ,⁷² This increase in CCR7 expression can be of interest for tumors that have also shown to have lymph node-like vasculature that express CCL19 and CCL21 and are correlated with T cell infiltration and positive prognosis in breast cancer and melanoma patients.⁹⁷ CXCR5 expression by CD8⁺ T cells was intriguingly correlated inversely with slow rolling on P- selectin and E-selectin at 1 dyn/cm², given CXCR5⁺ CD8⁺ T cells’ potent tumor-infiltrating capabilities⁷⁵ and antitumor activity.⁷⁴ Together, expression of adhesion receptors by cells exhibiting slow rolling velocity in shear flow differs with vasculature microenvironment, showingits role in biasing the local recruitment of CD8⁺ T cells to different extents and different qualities.

[0221] In conclusion, cellular characteristics associated with different qualities of rolling adhesion within a heterogeneous CD8⁺ T cell population were evaluated using an integrated adhesion-based chromatography channel and photoactivation. The photoactivation platform increases the resolution of the previously used chromatography channel, where cells could only be sorted into two subpopulations (adherent and non-adherent), by individually “labeling” cells based on their rolling velocities, elucidating differences between cells of different adhesion qualities. This study shows that adhesion ligand/receptor expression of slow-rolling cells varies based on inflamed-vasculature microenvironment. An advantage of this method is its amenability to multiple defined configurations, allowing for the investigation of effects different vasculature microenvironments as well as cell populations including but not limited to granulocytes, CD4⁺ T cells, B cells, and CD8⁺ T from different sources or activation states, and relevance to other disease states such as inflammation and wound healing. Therefore, the presented optofluidic microengineered platform enables high-throughput single-cell velocitylabeling for rapid screening of cells based on adhesive quality to provide insight for the development of better pharmacological approaches to modulate homing-regulated CD8⁺ T cell immunity for various viral infections and cancer.

Example 3: Murine CD8⁺ T-cell Enrichment on Nepmucin

[0222] The goal of this study was to evaluate the functional quality of CD8⁺ T-cells enriched on Nepmucin, under continuous flow conditions (concentration: 10 ug/mL Nepmucin; groups: Untreated vs. Day 4 - Expanded; WSS: 0.25, 0.5 and 1 dyn/cm²). CD8⁺ T cells were expanded until Day 4 (aCD3/aCD28 dynabeads + IL-2). A pulse of expanded cells was then perfused into chromatography channels functionalized with 10 ug/mL of Nepmucin at three wall shear stress levels: 0.25, 0.5 and 1.0 dyn/cm². Perfusion videos of cell interaction were recorded at 4 locations throughout the functionalized channel region. Parent, FF and Adherent fractions were then collected. Recovery was counted, and fractions were stained for Flow Cytometry analysis. Thes steps were then repeated for naive (not expanded) CD8⁺ T-cells. FIG. 41 shows the functionalized channel setup and cell pulse loading time. Data was of N=3 independent experiments for both expanded and naive CD8⁺ T-cells. FIGS. 42A-42I show experimental results.

Example 4: Murine CD8⁺ T-cell Adhesion with different Flow Configurations and Adhesion Molecule Concentrations

[0223] The goal of this study was to characterize CD8⁺ T cell adhesion with different concentrations of Nepmucin and WSS, at both a fixed location of the channel throughout perfusion and along the functionalized length (Nepmucin flow configurations: increasing WSS, decreasing WSS, constant WSS with 10 vs. 20 ug/mL; E-selectin, L-selectin and MAdCAM: different concentrations (2.5, 5.0, 10 and 20 ug/mL) with continuous flow from a reservoir). The following conditions were tested with both 10 and 20 ug/mL of Nepmucin.

[0224] 1) 0.25 dyn/cm² only (Constant WSS)

[0225] 2) 0.5 dyn/cm² only (Constant WSS)

[0226] 3) 1.0 dyn/cm² only (Constant WSS)

[0227] 4) Increasing WSS

[0228] 5) Decreasing WSS

[0229] FIGS. 43A-43C show experimental results.

[0230] This experiment was then repeated for continuous flow (step-wise). A continuous flow of cells starting at 1 dyn/cm² was run for 5 minutes (concentration: 500,000 cells/mL, volume: 5 mL (in a reservoir)). Interaction was recorded starting after 5 minutes. 5 positions along channel length (0, 3, 6, 9 and 12 cm from beginning of functionalized region) were recorded. WSS was reduced to 0.5 dyn and interaction was recorded again. WSS was further reduced to 0.25 dyn and interaction was recorded again. TABLE 2 shows a summary of experimental conditions. FIGS. 44A-44C show experimental results.

TABLE 2. Experimental conditions.

Example 5: Untreated Murine CD8⁺ T-cell Adhesion (Static vs. Continuous)

[0231] A study was conducted to determine how untreated CD8⁺ T cell interaction under relevant fluidic forces differ when cells are 1) loaded as a pulse into the channel and allowed to settle and interact with the adhesive substrate (prior to re-initiating flow) or 2) perfused as a pulse into the channel under a single, continuous WSS (without interruption). A cell pulse was evenly loaded throughout the channel at a fixed WSS (1.0 dyn for 3 minutes). After loading, flow was stopped. Uniform cell coverage throughout channel was then verified. Next, a location with a good number of settled cells was identified. Flow was then resumed at the desired WSS, and this fixed location was recorded for 5 minutes. Lastly, cell interaction was recorded through the entire channel. FIG. 45 shows experimental conditions, and FIGS. 46A- 46B show experimental results.

Example 6: hMSC perfusion on MAdCAM (Untreated vs. IFN-y treated)

[0232] A study was conducted to observe the impacts of a step-wise increase in wall shear stress. A cell pulse (containing l.OxlO⁶ MSCs) was loaded throughout the chromatography channel at a fixed WSS (1.0 dyn/cm² for 2 minutes). After loading, flow was stopped for 5 minutes to allow cells to settle. It was then verified that cell coverage was uniform throughout the channel. Next, a location with settled cells was identified. Flow was resumed at 0.125 dyn/cm² and recorded at this fixed location for 1 minute. The microscope camera was moved across the entire channel to record cell interaction (i.e. the number of bound cells remaining after flow was resumed at this WSS). Finally, the camera was returned to a location with a good number of bound cells. This was repeated with WSS incrementally increased to 0.25, 0.5, 1.0, 2.0 and 5.0 dyn/cm². TABLE 3 shows an overview of conditions tested. FIG. 47 shows a summary of the workflow. FIGS. 48A-48C show experimental results.

TABLE 3. Overview of conditions tested.

Example 7: CAR-T Adhesion Data

[0233] The goal of this study was to determine differences in rolling v. firm adhesion of CAR-T cells on Nepmucin and P-selectin, as a function of wall-shear stress and flow configuration (functionalization schemes: 1% BSA (Blocked), human P-selectin (20 ug/mL) and human Nepmucin (20 ug/mL); flow conditions: static vs. continuous; cells: 2nd generation anti-MUC16 CAR-T (Donor #50)). The differences in adhesion quantity between static and continuous flow configurations were assessed. The amount of adhesion that is above background was also verified (using 1% BSA blocked channels). For statis trials, a cell pulse (IxlO⁶ cells) was loaded evenly throughout the channel at a fixed WSS (1.0 dyn for 3 minutes). After loading, flow was stopped for 5 minutes. Uniform cell coverage throughout channel length was then verified. Next, a location with a good number of settled cells was identified. Flow was resumed at the desired WSS and, this fixed location was recorded for 5 minutes. Lastly, cell interaction was recorded through the entire channel. FIG. 49 shows the conditions tested for continuous flow. FIG. 50 shows a summary of conditions tested. FIGS. 51A-51E show experimental results.

Example 8: Enrichment of CAR-T cells on P-selectin vs. Nepmucin

[0234] This study used the following conditions: N=1 donor: 2nd generation anti-MUC16 CAR-T (Donor 50); substrates: P-selectin (human) and Nepmucin (human), 20 ug/mL each; flow configuration: continuous WSS: 0.5 and 1.0 dyn/cm²; flow cytometry CD4⁺CD8‘ and CD4 CD8+.

[0235] At higher wall shear stress, fewer cells were recovered in the adherent fraction. Overall, P-selectin adhesion was higher throughout the channel length (compared to Nepmucin), at both WSS. Rolling dominated for P-selectin, whereas firm adhesion dominated for Nepmucin. Anti-MUC16 CAR-T were mostly CD4 CD8‘ than CD8⁺CD4‘. Flow cytometry confirmed that adhesive ligands for P-selectin and Nepmucin were enriched in the adherent fraction for both CD4⁺ and CD8⁺. For CD4⁺ and CD8⁺ T cells, upon enrichment on P-selectin and Nepmucin, adherent fractions were notably less Granzyme B⁺ and more TIM-3⁺. For CD4⁺ and CD8⁺ T cells, P-selectin-enriched CD4 CD8‘ cells exhibited higher sLeX⁺ expression (MFI and frequency shift). For CD4⁺ and CD8⁺ T cells, CCR7⁺ enrichment was minimal but more prominent for Nepmucin than P-selectin, between adherent and free-flow. An MFI shift indicating more L-selectin positivity for adherent CD8⁺ T-cells enriched on Nepmucin was observed. For CD8⁺ T cells, based on memory gating strategy (CD62L vs. CCR7 of CD45RA⁺ and CD45RA ), adherent cells enriched from Nepmucin and P-selectin perfusion were slightly more Central Memory but noticeably less Effector Memory only for Nepmucin. FIGS. 52A- 52C show further experimental results.

Example 9: Additional Data

[0236] FIGS. 53A-53R show marker and memory subtype of CD4⁺CD8‘. FIGS. 54A- 54R show marker and memory subtype of CD4 CD8 .

[0237] The following patents, applications and publications as listed below and throughout this document are hereby incorporated by reference in their entirety herein. Reference list for Example 1

1. Wu, R., Forget, M.-A., Chacon, J., Bernatchez, C., Haymaker, C., Chen, J. Q., Hwu, P., and Radvanyi, L.G. (2012). Adoptive T-cell therapy using autologous tumor-infiltrating lymphocytes for metastatic melanoma: current status and future outlook. Cancer J. 18, 160- 175. https://doi.org/ 10.1097/PPO.0b013e31824d4465.

2. Bernatchez, C., Radvanyi, L.G., and Hwu, P. (2012). Advances in the treatment of metastatic melanoma: adoptive T cell therapy. Semin. Oncol. 39, 215-226. https://doi.Org/10.1053/j.seminoncol.2012.01.006.

3. Restifo, N.P., Dudley, M.E., and Rosenberg, S.A. (2012). Adoptive immunotherapy for cancer : harnessing the T cell response. Nat. Rev. Immunol. 12, 269-281. https://doi.org/10.1038/nri3191.

4. Prieto, P.A., Durflinger, K.H., Wunderlich, J.R., Rosenberg, S.A., and Dudley, M.E. (2010). Enrichment of CD8⁺ cells from melanoma tumor-infiltrating lymphocyte cultures reveals tumor reactivity for use in adoptive cell therapy. J. Immunother. 33, 547-556. https://doi.org/10.1097/CJI.0b013e3181d367bd.

5. Radvanyi, L.G., Bernatchez, C., Zhang, M., Fox, P.S., Miller, P., Chacon, J., Wu, R., Lizee, G., Mahoney, S., Alvarado, G., et al. (2012). Specific lymphocyte subsets predict response to adoptive cell therapy using expanded autologous tumor-infiltrating lymphocytes in metastatic melanoma patients. Clin. Cancer Res. 18, 6758-6770. https://doi.org/10.

1158/1078-0432. CCR-12-1177.

6. Pages, F., Galon, J., Dieu-Nosjean, M.C., Tartour, E., Sautes-Fridman, C., and Fridman, W.H. (2010). Immune infiltration in human tumors: a prognostic factor that should not be ignored. Oncogene 29, 1093-1102. https://doi.org/10.1038/onc.2009.416.

7. Sato, E., Olson, S.H., Ahn, J., Bundy, B., Nishikawa, H., Qian, F., Jungbluth, A.A., Frosina, D., Gnjatic, S., Ambrosone, C., et al. (2005). Intraepithelial CD8⁺ tumor-infiltrating lymphocytes and a high CD8⁺/regulatory T cell ratio are associated with favorable prognosis in ovarian cancer. Proc. Natl. Acad. Sci. USA 102, 18538-18543. https://doi.org/10.1073/pnas.0509182102.

8. Mackensen, A., Ferradini, L., Carcelain, G., Triebel, F., Faure, F., Viel, S., and Hercend, T. (1993). Evidence for in situ amplification of cytotoxic T-lymphocytes with antitumor activity in a human regressive melanoma. Cancer Res. 53, 3569-3573. 9. Kim, S T., Jeong, H., Woo, O.H., Seo, J.H., Kim, A., Lee, E.S., Shin, S.W., Kim, Y.H., Kim, J.S., and Park, K.H. (2013). Tumor-infiltrating lymphocytes, tumor characteristics, and recurrence in patients with early breast cancer. Am. J. Clin. Oncol. 36, 224-231. https://doi.org/10.1097/COC. 0b013e3182467d90.

10. Naito, Y., Saito, K., Shiiba, K., Ohuchi, A., Saigenji, K., Nagura, H., and Ohtani, H. (1998). CD8⁺ T cells infiltrated within cancer cell nests as a prognostic factor in human colorectal cancer. Cancer Res. 58, 3491-3494.

11. Galon, J., Costes, A., Sanchez-Cabo, F., Kirilovsky, A., Mlecnik, B., Lagorce-Pages, C., Tosolini, M., Camus, M., Berger, A., Wind, P., et al. (2006). Type, density and location of immune cells within human colorectal tumors predict clinical outcome. Science 313, 1960- 1964. https://doi.org/ 10.1126/science.1129139.

12. Peske, J.D., Woods, A.B., and Engelhard, V.H. (2015). Control of CD8 T-cell infiltration into tumors by vasculature and microenvironment. Adv. Cancer Res. 128, 263- 307. https://doi.org/10.1016/bs.acr.2015.05.001.

13. Bernhard, EL, Neudorfer, J., Gebhard, K., Conrad, EL, Hermann, C., Nahrig, J., Fend, F., Weber, W., Busch, D.H., and Peschel, C. (2008). Adoptive transfer of autologous, HER2- specific, cytotoxic T lymphocytes for the treatment of HER2-overexpressing breast cancer. Cancer Immunol. Immunother. 57, 271-280. https://doi.org/10.1007/s00262-007-0355-7.

14. Klebanoff, C.A., Gattinoni, L., Palmer, D.C., Muranski, P., Ji, Y., Hinrichs, C.S., Borman, Z.A., Kerkar, S.P., Scott, C.D., Finkelstein, S.E., et al. (2011). Determinants of successful CD8 + T-cell adoptive immunotherapy for large established tumors in mice. Clin. Cancer Res. 17, 5343-5352. https://doi.org/10.1158/1078-0432.CCR-l l-0503.

15. Lanitis, E., Dangaj, D., Irving, M., and Coukos, G. (2017). Mechanisms regulating T- cell infiltration and activity in solid tumors. Ann. Oncol. 28, xiil8-xii32. https://doi.org/10.1093/annonc/mdx238.

16. Kmiecik, J., Poli, A., Brons, N.H.C., Waha, A., Eide, G.E., Enger, P.0., Zimmer, J., and Chekenya, M. (2013). Elevated CD3⁺ and CD8⁺ tumor-infiltrating immune cells correlate with prolonged survival in glioblastoma patients despite integrated immunosuppressive mechanisms in the tumor microenvironment and at the systemic level. J. Neuroimmunol. 264, 71-83. https://doi.Org/10.1016/j.jneuroim.2013.08.013. 17. Sackstein, R., Schatton, T., and Barthel, S.R. (2017). T-lymphocyte homing: an underappreciated yet critical hurdle for successful cancer immunotherapy. Lab. Invest. 97, 669-697. https://doi.org/10.1038/labinv-est.2017.25.

18. Nolz, J.C., Starbeck-Miller, G.R., and Harty, J.T. (2011). Naive, effector and memory CD8 T-cell trafficking: parallels and distinctions. Immuno-therapy 3, 1223-1233. https://doi.org/10.2217/imt.l l.100.

19. Weninger, W., Crowley, M.A., Manjunath, N., and von Andrian, U.H. (2001). Migratory properties of naive, effector, and memory CD8+ T cells. J. Exp. Med. 194, 953- 966. https://doi.Org/10.1084/jem.194.7.953.

20. Afanasiev, O.K., Nagase, K., Simonson, W., Vandeven, N., Blom, A., Koelle, D.M., Clark, R., and Nghiem, P. (2013). Vascular E-selectin expression correlates with CD8 lymphocyte infiltration and improved outcome in Merkel cell carcinoma. J. Invest. Dermatol. 133, 2065-2073. https://doi. org/10.1038/jid.2013.36.

21. Stark, F.C., Gumani, K., Sad, S., and Krishnan, L. (2012). Lack of functional selectin ligand interactions compromises long term tumor protection by CD8⁺ T cells. PLoS One 7, e32211. https://doi.org/10.1371/journal. pone.0032211.

22. Fisher, D.T., Chen, Q., Skitzki, J.J., Muhitch, J.B., Zhou, L., Appenheimer, M.M., Vardam, T.D., Weis, E.L., Passanese, J., Wang, W.C., et al. (2011). IL-6 trans-signaling licenses mouse and human tumor microvascular gateways for trafficking of cytotoxic T cells.

J. Clin. Invest. 121, 3846-3859. https://doi.org/10.1172/JCI44952.

23. Adams, D.H., Yannelli, J.R., Newman, W., Lawley, T., Ades, E., Rosenberg, S. a, and Shaw, S. (1997). Adhesion of tumour-infiltrating lymphocytes to endothelium: a phenotypic and functional analysis. Br. J. Cancer 75, 1421-1431. https://doi.org/10.1038/bjc.1997.245.

24. Hirata, T., Furie, B.C., and Furie, B. (2002). P-E-and L-selectin mediate migration of activated CD8+ T lymphocytes into inflamed skin. J. Immunol. 169, 4307-4313. https://doi.org/10.4049/jimmunol.169. 8.4307.

25. Tietz, W., Allemand, Y., Borges, E., von Laer, D., Hallmann, R., Vestweber, D., and Hamann, A. (1998). CD4⁺ T cells migrate into inflamed skin only if they express ligands for E- and P-selectin. J. Immunol. 161, 963-970. 26. Yamaoka, T., Fujimoto, M., Ogawa, F., Yoshizaki, A., Bae, S.J., Muroi, E., Komura, K., Iwata, Y., Akiyama, Y., Yanaba, K., et al. (2011). The roles of P- and E-selectins and P- selectin glycoprotein ligand- 1 in primary and metastatic mouse melanomas. J. Dermatol. Sci. 64, 99-107. https://doi. org/10.1016/j.jdermsci.2011.07.005.

27. Woods, A.N., Wilson, A.L., Srivinisan, N., Zeng, J., Dutta, A.B., Peske, J. D., Tewalt, E.F., Gregg, R.K., Ferguson, A.R., and Engelhard, V.H.(2017). Differential expression of homing receptor ligands on tumor-associated vasculature that control CD8 effector T-cell entry. Cancer Immunol. Res. 5, 1062-1073. https://doi.org/10.1158/2326-6066.CIR-17-0190.

28. Shamay, Y., Elkabets, M., Li, H., Shah, J., Brook, S., Wang, F., Adler, K., Baut, E., Scaltriti, M., Jena, P.V., et al. (2016). P-selectin is a nanotherapeutic delivery target in the tumor microenvironment. Sci. Transl. Med. 8, 345ra87. https://doi.org/10.1126/scitranslmed.aaf7374.

29. Weishaupt, C., Munoz, K.N., Buzney, E., Kupper, T.S., and Fuhlbrigge, R.C. (2007). T-cell distribution and adhesion receptor expression in meta-static melanoma. Clin. Cancer Res. 13, 2549-2556. https://doi.org/10. 1158/1078-0432.CCR-06-2450.

30. Oh, J., Edwards, E.E., McClatchey, P.M., and Thomas, S.N. (2015). Analytical cell adhesion chromatography reveals impaired persistence of meta-static cell rolling adhesion to P-selectin. J. Cell Sci. 128, 3731-3743. https://doi.org/10.1242/jcs.166439.

31. Birmingham, K.G., O’Melia, M.J., Ban, D., Mouw, J., Edwards, E.E., Marcus, A.I., Mcdonald, J., and Thomas, S.N. (2019). Analyzing mechanisms of metastatic cancer cell adhesive phenotype leveraging preparative adhesion chromatography microfluidic. Adv. Biosyst. 3, el800328. https://doi.org/10.1002/adbi.201800328.

32. Birmingham, K.G., Robinson, I.E., Edwards, E.E., and Thomas, S.N.(2020). Photoconversion and chromatographic microfluidic system reveals differential cellular phenotypes of adhesion velocity versus persistence in shear flow. Lab Chip 20, 806-822. https://doi.org/10.1039/c91c00923j.

33. Nooijen, P.T., Westphal, J.R., Eggermont, A.M., Schalkwijk, C., Max, R., de Waal, R.M., and Ruiter, D.J. (1998). Endothelial P-selectin expression is reduced in advanced primary melanoma and melanoma metastasis. Am. J. Pathol. 152, 679-682. 34. Laubli, H., and Borsig, L. (2010). Selectins as mediators of lung metastasis. Cancer Microenviron. 3, 97-105. https://doi.org/10.1007/sl2307-010-0043-6.

35. Tinoco, R., Otero, D.C., Takahashi, A. A., and Bradley, L.M. (2017). PSGL-1 : a new player in the immune checkpoint landscape. Trends Immunol. 38, 323-335. https://doi.Org/10.1016/j.it.2017.02.002.

36. Atarashi, K., Hirata, T., Matsumoto, M., Kanemitsu, N., and Miyasaka, M.(2005). Rolling of Thl cells via P-selectin glycoprotein ligand-1 stimulates LFA-1 -mediated cell binding to ICAM-1. J. Immunol. 174, 1424-1432. http s : //doi . org/ 10.4049/j immunol .174.3.1424.

37. Edwards, E.E., and Thomas, S.N. (2017). P-Selectin and ICAM-1 synergy in mediating THP-1 monocyte adhesion in hemodynamic flow is length dependent. Integr. Biol. 9, 313-327. https://doi.org/10.1039/c7ib00020k.

38. Wong, A.K., LLanos, P., Boroda, N., Rosenberg, S.R., and Rabbany, S.Y.(2016). A parallel-plate flow chamber for mechanical characterization of endothelial cells exposed to laminar shear stress. Cell. Mol. Bioeng. 9, 127-138. https://doi.org/10.1007/sl2195-015- 0424-5.

39. Soares, A., Govender, L., Hughes, J., Mavakla, W., de Kock, M., Barnard, C., Pienaar, B., Janse van Rensburg, E., Jacobs, G., Khomba, G., et al. (2010). Novel application of Ki67 to quantify antigen-specific in vitro lymphoproliferation. J. Immunol. Methods 362, 43-50. https://doi.org/10. 1016/j.jim.2010.08.007.

40. Horton, B.L., Williams, J.B., Cabanov, A., Spranger, S., and Gajewski, T.F.(2018). Intratumoral CD8⁺ T-cell apoptosis is a major component of T-cell dysfunction and impedes antitumor immunity. Cancer Immunol. Res. 6, 14-24. https://doi.org/10.1158/2326-

6066. CIR- 17-0249.

41. Horton, B.L., and Gajewski, T.F. (2018). Back from the dead: TIL apoptosis in cancer immune evasion. Br. J. Cancer 118, 309-311. https://doi.org/10. 1038/bjc.2017.483.

42. Zhu, J., Powis De Tenbossche, C.G., Cane , S., Colau, D., Van Baren, N., Lurquin, C., Schmitt-Verhulst, A.M., Liljestrbm, P., Uyttenhove, C., and Van Den Eynde, B.J. (2017).

Resistance to cancer immunotherapy mediated by apoptosis of tumor-infiltrating lymphocytes. Nat. Commun. 8, 1404-1415. https://doi.org/10.1038/s41467-017-00784-l. 43. O’melia, M.J., Rohner, N.A., Manspeaker, M.P., Francis, D.M., Kissick, H.T., and Thomas, S.N. (2020). Quality of CD8 + T cell immunity evoked in lymph nodes is compartmentalized by route of antigen transport and functional in tumor context. Sci. Adv. 6, eabd7134. https://doi.org/10. 1126/sciadv.abd7134.

44. Allan, R.S., Waithman, J., Bedoui, S., Jones, C.M., Villadangos, J.A., Zhan, Y., Lew, A.M., Shortman, K., Heath, W.R., and Carbone, F.R. (2006). Migratory dendritic cells transfer antigen to a lymph node-resident dendritic cell population for efficient CTL priming. Immunity 25, 153-162. https://doi.Org/10.1016/j.immuni.2006.04.017.

45. O’Melia, M.J., Manspeaker, M.P., and Thomas, S.N. (2021). Tumor-draining lymph nodes are survival niches that support T cell priming against lymphatic transported tumor antigen and effects of immune checkpoint blockade in TNBC. Cancer Immunol. Immunother.

70, 2179-2195. https://doi.org/10.1007/s00262-020-02792-5.

46. Roberts, E.W., Broz, M.L., Binnewies, M., Headley, M.B., Nelson, A.E., Wolf, D.M., Kaisho, T., Bogunovic, D., Bhardwaj, N., and Krummel, M.F.(2016). Critical role for CD103⁺/CD141⁺ dendritic cells bearing CCR7 for tumor antigen trafficking and priming of T cell immunity in melanoma. Cancer Cell 30, 324-336. https://doi.Org/10.1016/j.ccell.2016.06.003.

47. Kiimpers, C., Jokic, M., Haase, O., Offermann, A., Vogel, W., Gratz, V., Langan, E.A., Pemer, S., and Terheyden, P. (2019). Immune cell infiltration of the primary tumor, not PD-L1 status, is associated with improved response to checkpoint inhibition in metastatic melanoma. Front. Med. 6, 27. https://doi.org/10.3389/fmed.2019.00027.

48. Tumeh, P.C., Harview, C.L., Yearley, J.H., Shintaku, I.P., Taylor, E.J.M., Robert, L., Chmielowski, B., Spasic, M., Henry, G., Ciobanu, V., et al. (2014). PD-1 blockade induces responses by inhibiting adaptive immune resistance. Nature 515, 568-571. https://doi.org/10.1038/naturel3954.

49. McEver, R.P., and Zhu, C. (2010). Rolling cell adhesion. Annu. Rev. Cell Dev. Biol. 26, 363-396. https://doi.org/10.1146/annurev.cellbio.042308. 113238.

50. Klebanoff, C. a, Finkelstein, S.E., Surman, D.R., Lichtman, M.K., Gattinoni, L., Theoret, M.R., Grewal, N., Spiess, P.J., Antony, P. a, Palmer, D.C., et al. (2004). IL-15 enhances the in vivo antitumor activity of tumor-reactive CD8⁺ T cells. Proc. Natl. Acad. Sci. USA 101, 1969-1974. https://doi. org/10.1073/pnas.030729810L 51. Klebanoff, C.A., Gattinoni, L., Torabi-parizi, P., Kerstann, K., Cardones, A.R., Finkelstein, S.E., Palmer, D.C., Antony, P.A., Hwang, S.T., Rosen-berg, S.A., et al. (2005). Central memory self/tumor-reactive CD8+ T cells confer superior antitumor immunity compared with effector memory T cells. Proc. Natl. Acad. Sci. USA 102, 9571-9576. https://doi.org/ 10.1073/pnas.0503726102.

52. Gattinoni, L., Klebanoff, C.A., Palmer, D.C., Wrzesinski, C., Kerstann, K., Yu, Z., Finkelstein, S.E., Theoret, M.R., Rosenberg, S.A., and Restifo, N.P. (2005). Acquisition of full effector function in vitro paradoxically im-pairs the in vivo antitumor efficacy of adoptively transferred CD8⁺ T cells. J. Clin. Invest. 115, 1616-1626. https://doi.org/10.1172/JCI24480.

53. Gattinoni, L., Klebanoff, C.A., and Restifo, N.P. (2012). Paths to sternness: building the ultimate antitumour T cell. Nat. Rev. Cancer 12, 671-684. http s : //doi . org/ 10.1038/ nrc3322.

54. Mueller, S.N., Gebhardt, T., Carbone, F.R., and Heath, W.R. (2013). Memory T cell subsets , migration patterns , and tissue residence. Annu. Rev. Immunol. 31, 137-161. https://doi.Org/10.l 146/annurev-immunol-032712-095954.

55. Pardoll, D.M. (2012). The blockade of immune checkpoints in cancer immunotherapy. Nat. Rev. Cancer 12, 252-264. https://doi.org/10.1038/nrc3239.

56. Sharma, P., and Allison, J.P. (2015). The future of immune checkpoint therapy. Science 348, 56-61. https://doi.org/10.1126/science.aaa8172.

57. Elia, A.R., Grioni, M., Basso, V., Curnis, F., Freschi, M., Corti, A., Mondino, A., and Bellone, M. (2018). Targeting tumor vasculature with TNF leads effector t cells to the tumor and enhances therapeutic efficacy of immune checkpoint blockers in combination with adoptive cell therapy. Clin. Cancer Res. 24, 2171-2181. https://doi.org/10.1158/1078- 0432.CCR-17-2210.

58. John, L.B., Devaud, C., Duong, C.P.M., Yong, C.S., Beavis, P.A., Haynes, N.M., Chow, M.T., Smyth, M.J., Kershaw, M.H., and Darcy, P.K. (2013). Anti-PD-1 antibody therapy potently enhances the eradication of established tumors by gene-modified T cells. Clin. Cancer Res. 19, 5636-5646. https://doi.org/10.1158/1078-0432.CCR-13-0458. 59. Rafiq, S., Yeku, O.O., Jackson, H.J., Purdon, T.J., van Leeuwen, D.G., Drakes, D.J., Song, M., Miele, M.M., Li, Z., Wang, P., et al. (2018). Targeted delivery of a PD-l-blocking scFV by CAR-T cells enhances anti-tumor efficacy in vivo. Nat. Biotechnol. 36, 847-856. https://doi.org/10.1038/nbt.4195.

60. Cherkassky, L., Morello, A., Villena-Vargas, J., Feng, Y., Dimitrov, D.S., Jones, D.R., Sadelain, M., and Adusumilli, P.S. (2016). Human CAR T cells with cell-intrinsic PD-1 checkpoint blockade resist tumor-mediated inhibition. J. Clin. Invest. 126, 3130-3144. https://doi.org/10.1172/JCI83092.

61. Li, S., Siriwon, N., Zhang, X., Yang, S., Jin, T., He, F., Kim, Y.J., Mac, J., Lu, Z., Wang, S., et al. (2017). Enhanced cancer immunotherapy by chimeric antigen receptor- modified T cells engineered to secrete checkpoint inhibitors. Clin. Cancer Res. 23, 6982- 6992. https://doi.org/10.1158/1078-0432.CCR-17-0867.

62. Chong, B.F., Murphy, J.-E., Kupper, T.S., and Fuhlbrigge, R.C. (2004). E-selectin, thymus- and activation-regulated chemokine/CCL17, and intercellular adhesion molecule- 1 are constitutively coexpressed in dermal microvessels: a foundation for a cutaneous immunosurveillance system. J. Immunol. 172, 1575-1581. https://doi.org/10.4049/jimmunol. 172.3.1575.

63. Piali, L., Fichtel, A., Terpe, H.-J., Imhof, B.A., and Gisler, R.H. (1995). Endothelial vascular cell adhesion molecule 1 expression is suppressed by melanoma and carcinoma. J. Exp. Med. 181, 811-816. https://doi.org/10. 1084/jem.l81.2.811.

64. Hendry, S.A., Farnsworth, R.H., Solomon, B., Achen, M.G., Stacker, S.A., and Fox, S.B. (2016). The role of the tumor vasculature in the host immune response: implications for therapeutic strategies targeting the tumor microenvironment. Front. Immunol. 7, 621. https://doi.org/10.3389/fimmu.2016.00621.

65. Griffioen, A.W., Damen, C.A., Blijham, G.H., and Groenewegen, G. (1996). Tumor angiogenesis is accompanied by a decreased inflammatory response of tumor-associated endothelium. Blood 88, 667-673.

66. Harlin, H., Meng, Y., Peterson, A.C., Zha, Y., Tretiakova, M., Slingluff, C., Mckee, M., and Gajewski, T.F. (2009). Chemokine expression in melanoma metastases associated with CD8+ T-cell recruitment. Cancer Res. 69, 3077-3085. https://doi.org/10.1158/0008- 5472.CAN-08-2281. 67. Kamoun, W.S., Chae, S.S., Lacorre, D.A., Tyrrell, J.A., Mitre, M., Gillissen, M.A., Fukumura, D., Jain, R.K., and Munn, L.L. (2010). Simultaneous measurement of RBC velocity, flux, hematocrit and shear rate in vascular networks. Nat. Methods 7, 655-660. https://doi.org/10.1038/nmeth.1475.

68. Siemann, D.W. (2011). The unique characteristics of tumor vasculature and preclinical evidence for its selective disruption by Tumor-Vascular Disrupting Agents. Cancer Treat Rev. 37, 63-74. https://doi.org/10. 1016/j.ctrv.2010.05.001.

69. Topper, J.N., and Gimbrone, M.A., Jr. (1999). Blood flow and vascular gene expression: fluid shear stress as a modulator of endothelial phenotype. Mol. Med. Today 5, 40-46. https://doi.Org/10.l 172/JCI83083.

Reference list for Example 2

1. He, R. et al. Follicular CXCR5 -expressing CD8+ T cells curtail chronic viral infection. Nature 537, 412-416 (2016).

2. Bogoslowski, A., Butcher, E. C. & Kubes, P. Neutrophils recruited through high endothelial venules of the lymph nodes via PNAd intercept disseminating Staphylococcus aureus. PNAS 115, 2449-2454 (2018).

3. Alon, R. et al. Leukocyte trafficking to the lungs and beyond: lessons from influenza for COVID-19. Nat Rev Immunol 21, 49-64 (2021).

4. Tietz, W. et al. CD4+ T cells migrate into inflamed skin only if they express ligands for E- and P-selectin. The Journal of Immunology 161, 963-970 (1998).

5. Hirata, T., Furie, B. C. & Furie, B. P-, E-, and L-Selectin Mediate Migration of Activated CD8+ T Lymphocytes into Inflamed Skin. The Journal of Immunology 169, 4307- 4313 (2002).

6. Nourshargh, S. & Alon, R. Leukocyte Migration into Inflamed Tissues. Immunity 41, 694-707 (2014).

7. Kolaczkowska, E. & Kubes, P. Neutrophil recruitment and function in health and inflammation. Nat Rev Immunol 13, 159-175 (2013). 8. Ingersoll, M. A., Platt, A. M., Potteaux, S. & Randolph, G. J. Monocyte trafficking in acute and chronic inflammation. Trends Immunol 32, 470-477 (2011).

9. Battistini, L. et al. CD8+ T cells from patients with acute multiple sclerosis display selective increase of adhesiveness in brain venules: A critical role for P-selectin glycoprotein ligand- 1. Blood 101, 4775-4782 (2003).

10. Sandor, A. M., Jacobelli, J. & Friedman, R. S. Immune cell trafficking to the islets during type 1 diabetes. Clinical and Experimental Immunology vol. 198 314-325 Preprint at https://doi.Org/10.l 111/cei.13353 (2019).

11. Peng, H. & Tian, Z. NK Cell Trafficking in Health and Autoimmunity: A Comprehensive Review. Clinical Reviews in Allergy and Immunology vol. 47 119-127 Preprint at https://doi.org/10.1007/sl2016-013-8400-0 (2014).

12. Bobryshev, Y. V. Monocyte recruitment and foam cell formation in atherosclerosis. Micron vol. 37 208-222 Preprint at https://doi.Org/10.1016/j.micron.2005.10.007 (2006).

13. Tan, L. Y. et al. Control of immune cell entry through the tumour vasculature: A missing link in optimising melanoma immunotherapy? Clin Transl Immunology 6, el 34-9 (2017).

14. Mikucki, M. E. et al. Non-redundant requirement for CXCR3 signalling during tumoricidal T-cell trafficking across tumour vascular checkpoints. Nat Commun 6, 1-30 (2015).

15. Lanitis, E., Dangaj, D., Irving, M. & Coukos, G. Mechanisms regulating T-cell infiltration and activity in solid tumors. Annals of Oncology 28, xiil8-xii32 (2017).

16. Elia, A. R. et al. Targeting tumor vasculature with TNF leads effector t cells to the tumor and enhances therapeutic efficacy of immune checkpoint blockers in combination with adoptive cell therapy. Clinical Cancer Research 24, 2171-2181 (2018).

17. Li, X. et al. Human CD8+ T cells mediate protective immunity induced by a human malaria vaccine in human immune system mice. Vaccine 34, 4501-4506 (2016).

18. Duan, S. & Thomas, P. G. Balancing immune protection and immune pathology by CD8+ T-cell responses to influenza infection. Frontiers in Immunology vol. 7 1-16 Preprint at https://doi.org/10.3389/fimmu.2016.00025 (2016). 19. Ray, S. J., Franki, S. N. & Pierce, R. H. The Collagen Binding alphalbetal Integrin VLA-1 Regulates CD8 T Cell-Mediated Immune Protection against Heterologous Influenza Infection. Immunity 20, 167-179 (204AD).

20. Sung, P. S., Racanelli, V. & Shin, E. C. CD8+ T-cell responses in acute hepatitis C virus infection. Frontiers in Immunology vol. 5 1-7 Preprint at https://doi.org/10.3389/fimmu.2014.00266 (2014).

21. Prieto, P. A., Durflinger, K. H., Wunderlich, J. R., Rosenberg, S. A. & Dudley, M. E. Enrichment of CD8+ Cells from Melanoma Tumor-infiltrating Lymphocyte Cultures Reveals Tumor Reactivity for Use in Adoptive Cell Therapy. Journal of Immunotherapy 33, 547-556 (2010).

22. Radvanyi, L. G. et al. Specific lymphocyte subsets predict response to adoptive cell therapy using expanded autologous tumor-infiltrating lymphocytes in metastatic melanoma patients. Clinical Cancer Research 18, 6758-6770 (2012).

23. Bematchez, C., Radvanyim, L. & Hwu, P. Advances in the Treatment of Metastatic Melanoma: Adoptive T Cell Therapy. Semin Oncol. 39, 215-226 (2012).

24. Pages, F. et al. Immune infiltration in human tumors: A prognostic factor that should not be ignored. Oncogene 29, 1093-1102 (2010).

25. Sato, E. et al. Intraepithelial CD8+ tumor-infiltrating lymphocytes and a high CD8+/regulatory T cell ratio are associated with favorable prognosis in ovarian cancer. PNAS. 102, 18538-18543 (2005).

26. Mackensen, A. et al. Evidence for in Situ Amplification of Cytotoxic T-Lymphocytes with Antitumor Activity in a Human Regressive Melanoma. Cancer Res 53, 3569-3573 (1993).

27. Kim, S. T. et al. Tumor-infiltrating Lymphocytes, Tumor Characteristics, and Recurrence in Patients With Early Breast Cancer. Am J Clin Oncol 36, 224-231 (2013).

28. Naito, Y. et al. CD8+ T cells infiltrated within cancer cell nests as a prognostic factor in human colorectal cancer. Cancer Res 58, 3491-3494 (1998).

29. Galon, J. et al. Type, Density and Location of Immune Cells Within Human Colorectal Tumors Predict Clinical Outcome. Sci. 313, 1960-1963 (2006). 30. Carvalheiro, H., da Silva, J. A. P. & Souto-Cameiro, M. M. Potential roles for CD8+ T cells in rheumatoid arthritis. Autoimmun Rev 12, 401-409 (2013).

31. Tsai, S., Shameli, A. & Santamaria, P. CD8+ T Cells in Type 1 Diabetes, in Advances in Immunology vol. 100 79-124 (2008).

32. Friese, M. A. & Fugger, L. Autoreactive CD8+ T cells in multiple sclerosis: A new target for therapy? Brain vol. 128 1747-1763 Preprint at https://doi.org/10.1093/brain/awh578 (2005).

33. Liblau, R. S., Wong, F. S., Mars, L. T. & Santamaria, P. Autoreactive CD8 T Cells in Organ-Specific Autoimmunity: Emerging Targets for Therapeutic Intervention. Immunity 17, 1-6 (2002).

34. Denic, A., Wootla, B. & Rodriguez, M. CD8+ T cells in multiple sclerosis. Expert Opin Ther Targets 17, 1053-1066 (2013).

35. Sackstein, R., Schatton, T. & Barthel, S. R. T-lymphocyte homing: An underappreciated yet critical hurdle for successful cancer immunotherapy. Laboratory Investigation 97, 669-697 (2017).

36. Nolz, J. C., Starbeck-Miller, G. R. & Harty, J. T. Naive, effector and memory CD8 T- cell trafficking: parallels and distinctions. Immunotherapy 3, 1223-1233 (2011).

37. Weninger, W., Crowley, M. A., Manjunath, N. & von Andrian, U. H. Migratory Properties of Naive, Effector, and Memory CD8+ T Cells. J Exp Med 194, 953-966 (2001).

38. Luster, A. D., Alon, R. & von Andrian, U. H. Immune cell migration in inflammation: Present and future therapeutic targets. Nat Immunol 6, 1182-1190 (2005).

39. Sarkar, D. et al. Engineered cell homing. Blood 118, el84-el91 (2011).

40. Jung, U., Norman, K. E., Scharffetter-Kochanek, K., Beaudet, A. L. & Ley, K. Transit Time of Leukocytes Rolling through Venules Controls Cytokine-induced Inflammatory Cell Recruitment //? Vivo. J. Clin. Invest 102, 1526-1533 (1998).

41. Eniola, A. O., Willcox, P. J. & Hammer, D. A. Interplay between Rolling and Firm Adhesion Elucidated with a Cell-Free System Engineered with Two Distinct Receptor- Ligand Pairs. Biophys J 85, 2720-2731 (2003). 42. Edwards, E. E. et al. Fluorometric Quantification of Single-Cell Velocities to Investigate Cancer Metastasis Article Fluorometric Quantification of Single-Cell Velocities to Investigate Cancer Metastasis. Cell Syst 7, 496-509. e6 (2018).

43. Gregory, J. L. et al. Reduced leukocyte-endothelial cell interactions in the inflamed microcirculation of macrophage migration inhibitory factor-deficient mice. Arthritis Rheum 50, 3023-3034 (2004).

44. Afanasiev, O. K. et al. Vascular E-selectin expression correlates with CD8 lymphocyte infiltration and improved outcome in Merkel cell carcinoma. J Invest Dermatol. 133, 2065-2073 (2013).

45. Stark, F. C., Gurnani, K., Sad, S. & Krishnan, L. Lack of functional selectin ligand interactions compromises long term tumor protection by CD8 + T cells. PLoS One 7, (2012).

46. Fisher, D. T. et al. IL-6 trans-signaling licenses mouse and human tumor microvascular gateways for trafficking of cytotoxic T cells. J Clin Invest 121, 3846-3859 (2011).

47. Adams, D. H. et al. Adhesion of tumour-infiltrating lymphocytes to endothelium: a phenotypic and functional analysis. Br J Cancer 75, 1421-31 (1997).

48. Yamaoka, T. et al. The roles of P- and E-selectins and P-selectin glycoprotein ligand- 1 in primary and metastatic mouse melanomas. J Dermatol Sci 64, 99-107 (2011).

49. Al Alwan, B. et al. Single-molecule imaging and microfluidic platform reveal molecular mechanisms of leukemic cell rolling. Commun Biol 4, 1-14 (2021).

50. Sundd, P. et al. ‘Slings’ enable neutrophil rolling at high shear. Nature 488, 399-403 (2012).

51. Butcher, E. C. & Picker, L. J. Lymphocyte homing and homeostasis. Science (1979) 272, 60-66 (1996).

52. Adams, D. H. & Shaw, S. Leucocyte-endothelial interactions and regulation of leucocyte migration. The Lancet 343, 831-836 (1994).

53. Gattinoni, L., Klebanoff, C. A. & Restifo, N. P. Paths to sternness: Building the ultimate antitumour T cell. Nat Rev Cancer 12, 671-684 (2012). 54. Klebanoff, C. A., Gattinoni, L. & Restifo, N. P. CD8+ T-cells memory in tumor immunology and immunotherapy. Immunology Review 211, 214-224 (2006).

55. Atarashi, K., Hirata, T., Matsumoto, M., Kanemitsu, N. & Miyasaka, M. Rolling of Thl Cells via P-Selectin Glycoprotein Ligand-1 Stimulates LFA-1 -Mediated Cell Binding to ICAM-1 1. The Journal of Immunology 174, 1424-1432 (2005).

56. Nguyen, M. et al. Dissecting Effects of Anti-cancer Drugs and Cancer-Associated Fibroblasts by On-Chip Reconstitution of Immunocompetent Tumor Microenvironments. Cell Rep 25, 3884-3893. e3 (2018).

57. Jarvis, M. et al. Microfluidic co-culture devices to assess penetration of nanoparticles into cancer cell mass. Bioeng Transl Med 2, 268-277 (2017).

58. Edwards, E. E. & Thomas, S. N. P-Selectin and ICAM-1 synergy in mediating THP-1 monocyte adhesion in hemodynamic flow is length dependent. Integrative Biology 9, 313— 327 (2017).

59. Kobzdej, M. M. a et al. Discordant expression of selectin ligands and sialyl Lewis x - related epitopes on murine myeloid cells. Blood 100, 4485-4494 (2002).

60. Edwards, E. E. et al. P-, but not E-or L-, selectin-mediated rolling adhesion persistence in hemodynamic flow diverges between metastatic and leukocytic cells. Oncotarget 8, 83585-83601 (2017).

61. Birmingham, K. G. et al. Analyzing Mechanisms of Metastatic Cancer Cell Adhesive Phenotype Leveraging Preparative Adhesion Chromatography Microfluidic. Adv Biosy st 3, 1-13 (2019).

62. Oh, J., Edwards, E. E., McClatchey, P. M. & Thomas, S. N. Analytical cell adhesion chromatography reveals impaired persistence of metastatic cell rolling adhesion to P-selectin. J Cell Sci 128, 3731-3743 (2015).

63. Camargo, C. P. et al. Adhesion analysis via a tumor vasculature-like microfluidic device identifies CD8+ T cells with enhanced tumor homing to improve cell therapy. Cell Rep 42, (2023).

64. Victora, G. D. et al. Germinal center dynamics revealed by multiphoton microscopy with a photoactivatable fluorescent reporter. Cell 143, 592-605 (2010). 65. Patterson, G. H. & Lippincott-Schwartz, J. A Photoactivatable GFP for Selective Photolabeling of Proteins and Cells. Sci. 297, 1873-1876 (2002).

66. Edwards, E. E., Birmingham, K. G., O’Melia, M. J., Oh, J. & Thomas, S. N. Fluorometric Quantification of Single-Cell Velocities to Investigate Cancer Metastasis. Cell Syst 496-509 (2018) doi: 10.1016/j.cels.2018.10.005.

67. Birmingham, K. G., Robinson, I. E., Edwards, E. E. & Thomas, S. N. Lab on a Chip phenotypes of adhesion velocity versus persistence in shear flow f . Lab Chip 20, 806-822 (2020).

68. Henderson, J. N. et al. Structure and mechanism of the photoactivatable green fluorescent protein. J Am Chem Soc 131, 4176-4177 (2009).

69. Burdick, M. M., Mccarty, O. J. T., Jadhav, S., Konstantopoulos, K. & Papadaki, M. Cellular/Tissue Engineering Cell-Cell Interactions in Inflammation and Cancer Metastasis.

70. Dixon, J. B., Zawieja, D. C., Gashev, A. A. & Cote, G. L. Measuring microlymphatic flow using fast video microscopy. J Biomed Opt 10, 064016 (2005).

71. Wong, A. K., LLanos, P., Boroda, N., Rosenberg, S. R. & Rabbany, S. Y. A Parallel- Plate Flow Chamber for Mechanical Characterization of Endothelial Cells Exposed to Laminar Shear Stress. Cell Mol Bioeng 9, 127-138 (2016).

72. Forster, R., Davalos-misslitz, A. C. & Rot, A. CCR7 and its ligands : balancing immunity and tolerance. Nat Immunol 8, 362-371 (2008).

73. Im, S. J. et al. Defining CD8+ T cells that provide the proliferative burst after PD-1 therapy. Nature 537, 417-421 (2016).

74. Chu, F. et al. CXCR5+CD8+ T cells are a distinct functional subset with an antitumor activity. Leukemia 33, 2640-2653 (2019).

75. Bai, M. et al. CXCR5+ CD8+ T cells potently infiltrate pancreatic tumors and present high functionality. Exp Cell Res 361, 39-45 (2017).

76. E, J. et al. CD8+CXCR5+ T cells in tumor-draining lymph nodes are highly activated and predict better prognosis in colorectal cancer. Hum Immunol 79, 446-452 (2018).

77. Bonder, C. S., Clark, S. R., Norman, M. U., Johnson, P. & Kubes, P. Use of CD44 by CD4+ Thl and Th2 lymphocytes to roll and adhere. Immunobiology 107, 4798-4806 (2006). 78. Xie, B. H., Lim, Y., Luscinskas, F. W. & Lichtman, A. H. Acquisition of Selectin Binding and Peripheral Homing Properties by CD4+ and CD8+ T Cells. Journal of Experimental Medicine 189, 1765-1775 (1999).

79. Alcaide, P. et al. The 130-kDa Glycoform of CD43 Functions as an E-Selectin Ligand for Activated Thl Cells In Vitro and in Delayed-Type Hypersensitivity Reactions In Vivo. Journal of Investigative Dermatology 127, 1964-1972 (2007).

80. Birmingham, K. G. et al. Lymph Node Subcapsular Sinus Microenvironment-On-A- Chip Modeling Shear Flow Relevant to Lymphatic Metastasis and Immune Cell Homing. iScience 23, 101751 (2020).

81. Wu, R. et al. Adoptive T-cell therapy using autologous tumor-infiltrating lymphocytes for metastatic melanoma: current status and future outlook. Cancer J 18, 160-75 (2012).

82. Restifo, N. P., Dudley, M. E. & Rosenberg, S. A. Adoptive immunotherapy for cancer : harnessing the T cell response. Nature Reviews 12, 269-281 (2012).

83. Brummelman, J. et al. High-dimensional single cell analysis identifies stemlike cytotoxic CD8+T cells infiltrating human tumors. Journal of Experimental Medicine 215, 2520-2535 (2018).

84. Han, J., Khatwani, N., Searles, T. G., Turk, M. J. & Angeles, C. V. Memory CD8+ T cell responses to cancer. Seminars in Immunology vol. 49 1-26 Preprint at https://doi.Org/10.1016/j.smim.2020.101435 (2020).

85. Siddiqui, I. et al. Intratumoral Tcfl + PD-1 + CD8 + T Cells with Stem-like Properties Promote Tumor Control in Response to Vaccination and Checkpoint Blockade Immunotherapy. Immunity 50, 195-211 (2019).

86. Miller, B. C. et al. Subsets of exhausted CD8+ T cells differentially mediate tumor control and respond to checkpoint blockade. Nat Immunol 20, 326-336 (2019).

87. Sade-Feldman, M. et al. Defining T Cell States Associated with Response to Checkpoint Immunotherapy in Melanoma. Cell 175, 998-1013. e20 (2018).

88. Manjarrez-Orduno, N. et al. Circulating T cell subpopulations correlate with immune responses at the tumor site and clinical response to PD1 inhibition in non-small cell lung cancer. Front Immunol 9, 1-9 (2018). 89. Klebanoff, C. a et al. IL-15 enhances the in vivo antitumor activity of tumor-reactive CD8+ T cells. PNAS 101, 1969-1974 (2004).

90. Klebanoff, C. A. et al. Central memory self/tumor-reactive CD8+ T cells confer superior antitumor immunity compared with effector memory T cells. PNAS 102, 9571-9576 (2005).

91. Klebanoff, C. A. et al. Determinants of successful CD8 + T-cell adoptive immunotherapy for large established tumors in mice. Clinical Cancer Research 17, 5343- 5352 (2011).

92. Gattinoni, L. et al. Acquisition of full effector function in vitro paradoxically impairs the in vivo antitumor efficacy of adoptively transferred CD8+ T cells. Journal of Clinical Investigation 115, 1616-1626 (2005).

93. Hogg, N. et al. How T cells use LFA-1 to attach and migrate. Immunol Lett 92, 51-54 (2004).

94. Shamri, R. et al. Lymphocyte arrest requires instantaneous induction of an extended LFA-1 conformation mediated by endothelium-bound chemokines. Nat Immunol 6, 497-506 (2005).

95. Dauner, J. G., Williams, I. R. & Jacob, J. Differential Microenvironment Localization of Effector and Memory CD8 T Cells. The Journal of Immunology 180, 291-299 (2008).

96. Potsch, C., Vohringer, D. & Pircher, H. Distinct migration patterns of naive and effector CD8 T cells in the spleen: correlation with CCR7 receptor expression and chemokine reactivity. Eur J Immunol 29, 3562-3570 (1999).

97. Peske, J. D. et al. Effector lymphocyte-induced lymph node-like vasculature enables naive T-cell entry into tumours and enhanced anti -tumour immunity. Nat Commun 6, 1-15 (2015).

98. Abuzineh, K. et al. Microfluidics-based super-resolution microscopy enables nanoscopic characterization of blood stem cell rolling. Sci. Adv vol. 4 https://www.science.org (2018).

Claims

What is claimed is:

1. A system for sorting cells, comprising: a substrate, wherein the substrate comprises: an input; an output; an at least partially functionalized channel between the input and the output, wherein the functionalized channel is sized to replicate a physiological vasculature system; and a pump configured to flow a sample of cells through the channel at a flow rate that replicates flow through a physiological vasculature system, thereby inducing a wall shear stress on the sample of cells that replicates wall shear stress in a physiological vasculature system; wherein the sample of cells are sorted according to their affinity for the functionalized channel.

2. The system of claim 1, wherein the substrate comprises poly dimethylsiloxane (PDMS), polyethylene glycol (PEG), silicone, polymethyl methacrylate, polycarbonate, or any combination thereof.

3. The system of any one of claims 1-2, wherein the substrate is a microfluidic device.

4. The system of any one of claims 1-3, wherein the functionalized channel comprises a ligand.

5. The system of claim 4, wherein the ligand comprises P-selectin, E-selectin, Nepmucin, L-selectin, MAdCAM, intercellular adhesion molecule, vascular cell adhesion molecule, CD31, CD34, or any combination thereof.

6. The system of any one of claims 1-5, wherein the functionalized channel is cellularized.

7. The system of claim 6, wherein the functionalized channel comprises diseased tissue or tumor cells, endothelial cells, stromal cells, or any combination thereof.

8. The system of any one of claims 1-7, wherein the functionalized channel has a diameter of from about 50 pm to about 300 pm.

9. The system of any one of claims 1-8, wherein the substrate further comprises a settling region in fluid communication with the functionalized channel.

10. The system of claim 9, wherein the settling region is from about 1 cm to about 14 cm in length.

11. The system of any one of claims 1-10, wherein cells with high affinity for the functionalized channel pass through the functionalized channel more slowly than cells with low affinity for the functionalized channel.

12. The system of any one of claims 1-11, further comprising a light source, and wherein the light source is positioned to illuminate a portion of the functionalized channel.

13. The system of claim 12, wherein the sample of cells further comprise a photoactive element, and wherein the light source activates the photoactive element, thereby causing the cells to fluoresce.

14. The system of claim 13, wherein cells with high affinity for the functionalized channel have a greater residence time in the illuminated portion of the functionalized channel than cells with low affinity for the functionalized channel.

15. The system of claim 14, wherein cells with high affinity for the functionalized channel have greater fluorescence intensity than cells with low affinity for the functionalized channel.

16. The system of any one of claims 14-15, wherein cells that have substantially no affinity for the functionalized channel do not fluoresce.

17. The system of any one of claims 13-16, wherein the photoactive element comprises photoactivatable green fluorescent protein (PA-GFP), Dendra2, KikRGreen, Phamret, a photoactive compound or dye, or any combination thereof.

18. The system of any one of claims 1-17, wherein the flow rate is from about 0.05 mL/min to about 0.5 mL/min.

19. The system of any one of claims 1-18, wherein the wall shear stress is from about 0.25 dyn/cm² to about 1 dyn/cm².

20. The system of any one of claims 1-19, wherein the sample of cells comprise autologous tumor-infiltrating lymphocytes, CD8⁺ T cells, manufactured leukocytes, or any combination thereof.

21. A method of sorting cells, comprising:

(a) providing a substrate comprising an at least partially functionalized channel, wherein the functionalized channel is sized to replicate a physiological vasculature system;

(b) flowing a sample of cells through the at least partially functionalized channel at a flow rate that replicates flow through a physiological vasculature system, thereby inducing a wall shear stress on the sample of cells that replicates wall shear stress in a physiological vasculature system; and

(c) selecting the cells that show high affinity for the functionalized channel.

22. The method of claim 21, wherein the substrate comprises poly dimethylsiloxane (PDMS), polyethylene glycol (PEG), silicone, polymethyl methacrylate, polycarbonate, or any combination thereof.

23. The method of any one of claims 21-22, wherein the substrate is a microfluidic device.

24. The method of any one of claims 21-23, wherein the functionalized channel comprises a ligand.

25. The method of claim 24, wherein the ligand comprises P-selectin, E-selectin, Nepmucin, L-selectin, MAdCAM, intercellular adhesion molecule, vascular cell adhesion molecule, CD31, CD34, or any combination thereof.

26. The method of any one of claims 21-25, wherein the functionalized channel is cellularized.

27. The method of claim 26, wherein the functionalized channel comprises diseased tissue or tumor cells, endothelial cells, stromal cells, or any combination thereof.

28. The method of any one of claims 21-27, wherein the functionalized has a diameter of from about 50 pm to about 300 pm.

29. The method of any one of claims 21-28, wherein the substrate further comprises a settling region in fluid communication with the functionalized channel.

30. The method of claim 29, wherein the settling region is from about 1 cm to about 14 cm in length.

31. The method of any one of claims 21-30, wherein cells with high affinity for the functionalized channel pass through the functionalized channel more slowly than cells with low affinity for the functionalized channel.

32. The method of any one of claims 21-31, wherein step (c) further comprises illuminating a portion of the functionalized channel with a light source.

33. The method of claim 32, wherein the sample of cells further comprise a photoactive element, and wherein the light source activates the photoactive element, thereby causing the cells to fluoresce.

34. The method of claim 33, wherein cells with high affinity for the functionalized channel have a greater method time in the illuminated portion of the functionalized channel than cells with low affinity for the functionalized channel.

35. The method of claim 34, wherein cells with high affinity for the functionalized channel have greater fluorescence intensity than cells with low affinity for the functionalized channel.

36. The method of any one of claims 34-35, wherein cells that have substantially no affinity for the functionalized channel do not fluoresce.

37. The method of any one of claims 33-36, wherein the photoactive element comprises photoactivatable green fluorescent protein (PA-GFP), Dendra2, KikRGreen, Phamret, a photoactive compound or dye, or any combination thereof.

38. The method of any one of claims 21-37, wherein the flow rate is from about 0.05 mL/min to about 0.5 mL/min.

39. The method of any one of claims 21-38, wherein the wall shear stress is from about 0.25 dyn/cm² to about 1 dyn/cm².

40. The method of any one of claims 21-39, wherein the sample of cells comprise autologous tumor-infiltrating lymphocytes, CD8⁺ T cells, manufactured leukocytes, or any combination thereof.

41. The method of any one of claims 21-40, wherein the sample of cells is a first native cell population.

42. The method of claim 41, wherein the method is used to predict a behavior or characteristic of a second native cell population.

43. The method of claim 42, wherein the behavior or characteristic is affinity for at least one element of the functionalized channel.

44. The method of any one of claims 21-43, further comprising:

(d) collecting the selected cells; wherein collected cells are suitable for additional testing and/or administration to a patient.

45. The method of claim 44, wherein the collected cells have a greater ability to home to and engraft with a tissue having the physiological vasculature system replicated by the functionalized channel compared to non-selected cells.

46. The method of claim 45, wherein the collected cells have enhanced diseased tissue or tumor trafficking compared to the non-selected cells.

47. The method of any one of claims 44-46, further comprising:

(e) administering the collected cells to a patient.

48. A method of adoptive cell therapy, the method comprising:

(a) providing a substrate comprising an at least partially functionalized channel, wherein the functionalized channel is sized to replicate a diseased tissue or tumor vasculature system;

(b) flowing a sample of cells through the at least partially functionalized channel at a flow rate that replicates flow through a diseased tissue or tumor vasculature system, thereby inducing a wall shear stress on the sample of cells that replicates wall shear stress in a diseased tissue or tumor vasculature system;

(c) collecting the cells that show high affinity for the functionalized channel; and

(d) administering the collected cells to a patient with a diseased tissue or tumor; wherein the functionalized channel comprises a ligand; and wherein the collected cells have a greater ability to home to and engraft with the diseased tissue or tumor compared to the cells that show low affinity for the functionalized channel.

49. The method of claim 48, wherein the substrate comprises polydimethylsiloxane (PDMS), polyethylene glycol (PEG), silicone, polymethyl methacrylate, polycarbonate, or any combination thereof.

50. The method of any one of claims 48-49, wherein the substrate is a microfluidic device.

51. The method of any one of claims 48-50, wherein the ligand comprises P-selectin, E- selectin, Nepmucin, L-selectin, MAdCAM, intercellular adhesion molecule, vascular cell adhesion molecule, CD31, CD34, or any combination thereof.

52. The method of any one of claims 48-51, wherein the functionalized channel is cellularized.

53. The method of claim 52, wherein the functionalized channel comprises diseased tissue or tumor cells.

54. The method of claim 53, wherein the diseased tissue or tumor cells are collected from the patient.

55. The method of any one of claims 48-54, wherein the functionalized channel has a diameter of from about 50 pm to about 300 pm.

56. The method of any one of claims 48-55, wherein the substrate further comprises a settling region in fluid communication with the functionalized channel.

57. The method of claim 56, wherein the settling region is from about 1 cm to about 14 cm in length.

58. The method of any one of claims 48-57, wherein cells with high affinity for the functionalized channel pass through the functionalized channel more slowly than cells with low affinity for the functionalized channel.

59. The method of any one of claims 48-58, wherein step (c) further comprises illuminating a portion of the functionalized channel with a light source.

60. The method of claim 59, wherein the sample of cells further comprise a photoactive element, and wherein the light source activates the photoactive element, thereby causing the cells to fluoresce.

61. The method of claim 60, wherein cells with high affinity for the functionalized channel have a greater method time in the illuminated portion of the functionalized channel than cells with low affinity for the functionalized channel.

62. The method of claim 61, wherein cells with high affinity for the functionalized channel have greater fluorescence intensity than cells with low affinity for the functionalized channel.

63. The method of any one of claims 60-62, wherein cells that have substantially no affinity for the functionalized channel do not fluoresce.

64. The method of any one of claims 60-63, wherein the photoactive element comprises photoactivatable green fluorescent protein (PA-GFP), Dendra2, KikRGreen, Phamret, a photoactive compound or dye, or any combination thereof.

65. The method of any one of claims 48-64, wherein the flow rate is from about 0.05 mL/min to about 0.5 mL/min.

66. The method of any one of claims 48-65, wherein the wall shear stress is from about 0.25 dyn/cm² to about 1 dyn/cm².

67. The method of any one of claims 48-66, wherein the sample of cells comprise autologous tumor-infiltrating lymphocytes, CD8⁺ T cells, manufactured leukocytes, or any combination thereof.

68. The method of claim 67, wherein the collected cells have enhanced diseased tissue or tumor trafficking compared to the non-selected cells.