Skip to main content
NCBI home page
NIHPA Author Manuscripts logoLink to NIHPA Author Manuscripts
. Author manuscript; available in PMC: 2026 Jun 18.
Published in final edited form as: Biosens Bioelectron. 2025 Jun 18;287:117711. doi: 10.1016/j.bios.2025.117711

Multiplexed Profiling of Single-Cell Secretion on a Hierarchical Loading Microwell Chip

Ning Shao 1,*, Yufu Zhou 1,2, Xuewu Liu 1,*
PMCID: PMC12282956  NIHMSID: NIHMS2091697  PMID: 40540973

Abstract

Protein secretion is involved in many biological processes, such as cellular communication, embryonic development, tissue homeostasis, disease pathogenesis, and execution of immune functions. Highly multiplexed detection of single-cell secretion is greatly needed but has always been challenging. Herein, we present a hierarchical loading microwell chip (HL-Chip)-based method that efficiently aligns thousands of single cells with multiple antibody-coated microbeads for highly multiplexed detection of secreted proteins from single cells. We demonstrate the applications of this platform in profiling secretion of six cytokines from single T cells and macrophages after pan- and antigen-specific stimulation. The 6-plex cytokine profiling reveals an early but transient cytokine burst and polyfunctional heterogeneity in antigen peptide-stimulated T cell receptor-engineered T (TCR-T) cells. This simple and versatile technology could find its application in single-cell measurements in both biological discoveries and clinical diagnosis.

Keywords: Microfluidics, cytokine secretion, single-cell analysis, multiplexed detection, TCR-T cell

1. Introduction

Secreted proteins, including cytokines, chemokines, growth factors, antibodies, hormones, and some enzymes, play critical roles in many biological processes, such as embryonic development, tissue homeostasis, disease pathogenesis, and execution of immune functions (Chen et al., 2019; Dinarello, 2007). For example, in the immune system, cytokines, chemokines, antibodies and cytotoxic enzymes are secreted in response to infection or inflammation, and to regulate immune effector functions (Altan-Bonnet et al., 2019; Stanley et al., 2010). Approaches for analyzing secreted proteins are important not only for understanding biological processes, but also for clinical diagnosis, prognosis, and therapy evaluation. Besides a systemic response, the functional roles of secreted proteins are closely linked to the identities of cells that secrete these proteins, their numbers, localization, and activity (Bucheli et al., 2021). Methods for bulk measurements of protein secretion, such as enzyme-linked immunosorbent assay (ELISA), mass spectrometry, antibody microarray, etc., only measure average levels of protein secretion in body fluids or from a population of cells, masking potential cellular heterogeneity (Chen et al., 2019). Therefore, analytical technologies that can detect protein secretion with single-cell resolution are desired to provide better understanding of the diverse cellular functionality and the biological processes involved. Enzyme-linked immunospot (ELISpot) and derivatives, intracellular cytokine staining (ICS) coupled with flow cytometry and derivatives are two types of widely used methods for detecting protein secretion of single cells directly and indirectly, respectively. ELISpot (Czerkinsky et al., 1983) and its derivative FluoroSpot (Gazagne et al., 2003) detect truly secreted proteins from single cells and have been widely used in clinical settings and basic research. However, they can only detect four or less proteins simultaneously. In addition, they were usually used to evaluate the relative frequency of secreting cells but not to quantify the protein secretion of each single cells (Chen et al., 2019). ICS is commonly used as another standard method for measuring single-cell secretion due to its simplicity, high throughput and multiplicity. In ICS, secretion inhibitors are used to retain the proteins within the cell, followed by cell fixation, permeabilization and antibody staining, finally the flow cytometry measurements (Jung et al., 1993; Lovelace et al., 2018). Mass cytometry, which is a derivative of the conventional flow cytometry, replaces fluorescent probes by rare earth metal probes, further improving the multiplicity significantly (Bendall et al., 2011). However, both ICS and mass cytometry detect proteins within cells and on the cell surfaces, not truly secreted proteins. In addition, the using of secretion inhibitors may result in toxicity problems, which needs careful optimization (O’Neil-Andersen et al., 2002). Notably, in both ICS and ELISpot, it’s impossible to retrieve lives cells to link the single-cell secretion with other functional measures. To keep the live-cell retrievability, secreted proteins can also be immobilized to the cell surface by pre-functionalization of the cell surface with capture reagents, followed by standard flow cytometry protocols (Manz et al., 1995). However, capture reagents may be limited for some specific cell types.

Microfluidics falls into another major category of methods for detecting single-cell protein secretion (Jammes et al., 2020; Junkin et al., 2014; Shao et al., 2018). Microfluidics provides remarkable assay miniaturization, high throughput, and well controllability. By confining single cells into nano- or picoliter volume to increase local protein concentration, microfluidics-based methods are well suitable for accurate quantification of the tiny amount of proteins secreted by single cells. Microfluidics-based methods can be categorized into three formats: microwells, microchambers and droplets. In microwell methods, the single cells are sedimented into thousands of microwells, and the protein-capture reagents are either coated on a glass slide and covered onto the microwell chip to enclose the microwells, namely microengraving (Han et al., 2012; Love et al., 2006), or coated on the bottom surfaces of- or the planar surfaces between the microwells, forming an open-well format (Jin et al., 2009; Torres et al., 2014). Microchamber methods relies on isolation of single cells into elongated microchambers and precise patterning of high-density antibody barcode arrays (Ma et al., 2011; Shi et al., 2012). Compared to reported microwell methods, which can detect no more than three analytes simultaneously, microchamber-based methods utilize different spatial barcoding strategies to increase the multiplicity to 10~42 (Ji et al., 2019; Lu et al., 2015; Zhao et al., 2018). Both microwell and microchamber methods have been applied to detection single-cell secretion of a variety of cell types, including B cells (Jin et al., 2009; Love et al., 2006), T cells (Han et al., 2012; Ma et al., 2011), macrophages (Abdullah et al., 2019; Lu et al., 2015; Zhao et al., 2018) and tumor cells (Ji et al., 2019; Li et al., 2023). However, both methods suffer from uncontrolled cell isolation, basically following the Poisson distribution, which results in a lot empty or multi-cell-occupied units, wasting the cells and assay units. We previously developed a hierarchical loading microwell chip (HL-Chip), in which single cells and cytokine-detection microbeads were precisely paired into microwells with high efficiencies up to > 90%, offers an attractive alternative for detection of single-cell secretion (Zhou et al., 2020). However, the HL-Chip only demonstrated detection of up to two cytokines from single cells. Droplet microfluidics, which could encapsulate single cells in aqueous-solution-in-oil emulsions, enabled high detection sensitivity with higher throughput (up to millions of droplets per experiment) (Chokkalingam et al., 2013; Eyer et al., 2017; Konry et al., 2011; Wei et al., 2019). Similar to microwell and microchamber methods, droplet methods also suffer from uncontrolled, Poisson distribution-determined cell and optionally protein-capture microsphere encapsulation (typically between 0.2 to 0.01 cells/droplet), thus requiring higher sample input (Bucheli et al., 2021).

Herein we present an advanced hierarchical loading microwell chip (HL-Chip) -based method for highly multiplexed single-cell secretion profiling. The advanced HL-Chip efficiently aligns thousands of single cells each with two different-sized microbeads. Each bead is coated with three different capture antibodies for capturing secreted proteins. Spatial (position) and spectral (color) barcoding of the beads adjacent to the single cell enables highly multiplexed detection. We applied this platform to 6-plex detection of cytokine secretion of single T cells and macrophages after either pan- or antigen-specific stimulation, and revealed an early cytokine burst and polyfunctional heterogeneity in peptide-pulsed T cell receptor-engineered T (TCR-T) cells.

2. Materials and methods

2.1. Reagents and materials

Details of the key reagents and materials were summarized in Table S1.

2.2. Cell culture, differentiation, and stimulation

Human CD4+ and CD8+ T cells were purchased from STEMCELL Technologies and cultured in RPMI 1640 medium supplemented with 2 mM Glutamax, 20 mM HEPES, 10% human AB serum, 1% nonessential amino acids, 1 mM sodium pyruvate, 55 mM b-mercaptoethanol, 100 IU/mL penicillin, 100 mg/mL streptomycin, and 50 IU IL-2 (T-cell Medium) in a 37 °C incubator with 5% CO2. For CD4+ T cell stimulation, the chips loaded with T cells were incubated in 5 mL AIM-V medium containing either 1 × eBioscience™ Cell Stimulation Cocktail (contains 81 nM phorbol 12-myristate 13-acetate (PMA) and 1.34 μM ionomycin, Thermo Fisher Scientific) or dimethylsulfoxide (DMSO) for 1.5 h in a 37 °C incubator with 5% CO2 before washing and in-situ detection. Human monocytic cell line THP-1 was purchased from ATCC and cultured in RPMI 1640 medium supplemented with 10% fetal bovine serum (FBS), 100 IU/mL penicillin, and 100 mg/mL streptomycin (R10 medium) in a 37 °C incubator with 5% CO2. For THP-1 cell differentiation and stimulation, THP-1 cells seeded in a 24-well plate were cultured in R10 medium with 150 nM PMA for 24 hours to promote a macrophage differentiation (Genin et al., 2015). Then the cells were harvested and loaded to the chips, incubated in 5 mL AIM-V medium either containing 1 μg/mL lipopolysaccharide (LPS) or not, in a 37 °C incubator with 5% CO2 for 8 h, followed by on-chip detection. NY-ESO-1-specific TCR-T cells were generated from CD8+ T cells from a healthy donor as described in our previous work (Zhou et al., 2020) and cultured in T-cell Medium. For peptide stimulation, the TCR-T cells were treated with NY-ESO-V157-165 peptide (SLLMWITQV, 10 μg/mL, GenScript) or DMSO in a U-shape 96-well plate at 37 °C for 1 h, 2 h, 4 h, or 6 h. The cells were then harvested, washed and loaded to the chip for cytokine detection.

2.3. Fabrication of the HL-Chip

The HL-Chip features were designed using AutoCAD software (Adobe), printed out as chrome photomasks (Photo Sciences Inc.), and fabricated using standard single-layer soft-photolithography and elastomer molding. Briefly, SU-8 3025 photoresists (MicroChem Corp.) were spin-coated onto a 4-inch silicon wafer. The HL-Chip features were patterned onto the wafer using the photomask to serve as the template for making polydimethylsiloxane (PDMS) chips. The HL-Chip with a thickness of 0.5 mm were made by pouring polydimethylsiloxane (PDMS) mixture (10A: 1B; Sylgard 184 kit, Dow Corning Corp.) over the master wafer, degassing and curing at 80°C for 2 h. Then, the PDMS replica was peeled off, cut to single chips, tape-cleaned and placed in 35-mm petri dishes until use.

2.4. Microbeads functionalization, on-chip calibration and crosstalk test

Two different-sized streptavidin-coated microbeads (mean size 18.6 μm, SVP-200-4, and mean size 22.0 μm, SVM-200-4, Spherotech) were incubated in 10 μg/mL biotinylated capture antibody cocktail I (anti-IL2, anti-TNF-α and anti-IFN-γ) and cocktail II (anti-GM-CSF, anti-MIP-1α and anti-IL-8), respectively (Table S1), at room temperature (RT) for 2 h with shaking. The antibody-conjugated microbeads were then washed 3 times with 1.5% bovine serum albumin (BSA) and stored in 1.5% BSA. To calibrate the microbeads for multiplexed cytokine secretion, microbeads of the larger size and the small size were loaded into BSA-blocked chips in turn to form pairs and incubated with recombinant human cytokine cocktails containing all the six cytokines with serially diluted concentrations for 2 h at RT. After washing 3 times with 3 mL of 1.5% BSA with shaking, 40 μL of the detection antibody cocktail containing all the six fluorescently-labelled detection antibodies (Table S1) with optimized concentrations was added onto the chip. The chip was incubated at RT for 2 h, followed by washing with 3 mL of 1.5% BSA for 5 times. In the last step, the chip was immersed in PBS and imaged using a Nikon A1 confocal microscope. Multichannel fluorescent intensities of each paired larger microbead and smaller microbead were digitalized using ImageJ software (National Institutes of Health). Calibration curves were obtained using the fluorescent intensities and corresponding cytokine concentrations. Nonlinear regression with sigmoidal dose-response (variable slope) was performed using GraphPad Prism 8 software to fit the calibration curves (GraphPad Software, CA). The limits of detection (LOD) were calculated based on the formula Y=Yblk+3*SDblk, where Yblk is the average fluorescent intensity of the blank sample and SDblk is the standard deviation of the blank sample. LOD value for each cytokine was then calculated using its Y value and its corresponding regression equation. For the crosstalk test, each microbead-loaded chips were incubated with individual recombinant cytokine standards with a concentration of 1 ng/mL, and the complete detection antibody cocktail was applied for detection. Recorded fluorescent intensities were converted to detected cytokine concentrations using the corresponding regression equations.

2.5. Multiplexed on-chip single-cell cytokine detection assay

The chip in a 35-mm petri dish was first activated with oxygen plasma for 1 min and blocked with 1.5% BSA. Then 20 μL of the larger microbeads (2 × 106 /mL) were added onto the chip. The chip was centrifuged at 40 g for 3 s to dock single beads into the largest microwells. To improve the loading efficiency, the chip was shaken to resuspend the unsettled beads, and centrifuged again at 40 g for 3 s. This step can be repeated 1–2 times to achieve optimal capture efficiency. After loading, unsettled beads were gently washed with PBS and recycled for future usage. Then, the same loading step was performed to load the smaller microbeads, which matched the size of the medium-sized microwells. Finally, the cells with the smallest size among the three objects were loaded into the smallest wells in the same manner. To remove stacked cells or beads, 50 μL PBS was added onto the chip and a horizontal flow was generated by putting a filter paper strip in contact with one edge of the chip to suck the PBS along with the upper layer of the stacked beads or cells away. This washing procedure can be repeated three to five times until no more than 5% of wells become empty due to over washing. The fully-loaded chip was incubated in 5 mL of AIM-V medium containing different stimulators as described in the Results for designated durations. After incubation, all the medium was aspirated and 3 mL 1.5% BSA was added to the dish to immerse the chip. All the following steps were the same as described in the above section for the on-chip detection. For the live-cell retrieval, the HL-Chip was loaded with unstimulated THP-1 cells stained with Calcein AM. After the on-chip assay, single live cells were retrieved from the chip using a micromanipulation setup as we have described before (Zhou et al., 2020) and transferred to a 96 well-plate preadded with 100 µL 1:1 mixture of fresh media and THP-1 cell-conditioned media supplemented with 20% FBS in each well. The retrieved single cells in the 96 well-plate were checked under a fluorescent microscope for their presence and viability. Some of the single cells were stimulated with 150 nM PMA for 48 hours for differentiation to macrophages. The remaining single cells were cultured for clonal expansion for up to 14 days.

2.6. Image acquisition and analysis

Scanning electron microscopy (SEM) images for the graded loading were obtained using a Nova Nano scanning electron microscopy 230 instrument (high vacuum, HV=5 kV). Fluorescence images were obtained using a Nikon A1 confocal microscope. Bright-field images were used to identify single cells. Multichannel fluorescent intensities of each paired larger microbead and smaller microbead were digitalized using ImageJ software (National Institutes of Health) and converted to detected cytokine concentrations using the corresponding regression equations. Values from no-cell wells were used as baseline to determine the thresholds for true positive secretion of each cytokine.

2.7. Statistical analysis:

The statistical analysis was performed using GraphPad Prism 8 software (GraphPad Software, CA). Unpaired Student’s t-tests (two-tailed) were used to compare datasets with Gaussian distributions, and Mann–Whitney tests were used for datasets without Gaussian distributions. Differences were considered statistically significant when P-values were <0.05. All estimated errors for analysis of biological replicates were standard deviation (SD) values (unless otherwise stated in figure legends).

3. Results and discussion

3.1. Principle of the HL-Chip for single-cell secretion profiling

The high-density microwell array was designed by AutoCAD software and the chip was fabricated with PDMS. Each microwell unit was composed of three connected circular wells in a row, with different sizes, i.e., 25 μm, 11 μm (or 16 μm, depending on the cell type), and 20 μm. Each chip contains up to 10,000 tri-well units. Fig. 1A shows the procedure of the hierarchical loading on the chip. Firstly, 20 μL of the large antibody-coated beads (mean diameter of 21.9 μm), with a count of 3 times the number of the wells, was added onto the microwell chip. After 2–3 times of brief centrifugation, most of the largest wells were occupied by the large beads. Due to the size limitation, the large beads couldn’t fall into the two smaller wells. After washing off all unsettled beads outside the wells, the smaller antibody-coated beads with a mean diameter of 18.6 μm, were loaded onto the chip, similar to the loading of the large beads. Since the large wells were occupied, the size of the small beads determined that the small beads could only been loaded into the wells with a diameter of 20 μm. Finally, the cells, with the smallest sizes (11 or 16 μm), were loaded into the central well similarly. The details for determination of the well sizes were described in our previous publication. The three-step hierarchical loading was also demonstrated in the scanning electron microscope images shown in Fig. S1. Fig. 1B and 1C show the principle of multiplexed cytokine capture and detection. Each size of beads was coated with three different antibodies targeting three different cytokines. Cytokines secreted by a single cell could be efficiently captured by the beads adjacent to it in the same tri-well unit. After cytokine capture, the chip with beads and cells were incubated with a detection antibody cocktail. The three antibodies targeting the three cytokines captured on the same beads were labelled with three different fluorescence to distinguish from each other. The 2-plex spatial barcoding and 3-plex spectral barcoding realized 6-plex cytokine detection. This multiplicity could be further expanded by increasing the number of spatial and spectral barcoding. Theoretically, if M different colors are used for antibody labeling and N different-sized beads are employed, the final multiplicity would be M×N.

Fig. 1.

Fig. 1.

Open in a new tab

Working principle of the HL-Chip-based multiplexed single-cell cytokine secretion profiling. (A) Working flow of the hierarchical loading of beads and cells. (B) Schematic illustration of the multiplexed single-cell cytokine capture and detection. (C) Representative images of the multichannel fluorescent-based detection of beads.

3.2. Characterization of the HL-Chip for multiplexed cytokine detection.

To evaluate the efficiency of HL-Chip loading, we hierarchically loaded the two sizes of cytokine-capture-beads and THP-1 cells (with an average diameter of ~14 um) to the chip. The corresponding well occupancies of each loading step were evaluated by imaging random areas after each loading and calculate the fractions of wells occupied correctly. Fig. 2A shows representative images of the three-step loading. Fig. 2B shows the corresponding loading efficiencies of each step. The loading efficiency of the first beads was 98.5% ±0.9%, the loading efficiency of paired large-small beads was 89.4% ±3.1%, and the correct bead-cell-bead pairing efficiency was 77.9% ±4.0%.

Fig. 2.

Fig. 2.

Open in a new tab

Characterization of the hierarchical loading of beads and cells, and the bead-based cytokine detection. (A) Representative images of the hierarchical loading of large and small beads and cells into individual units. Scale bar, 100 µm. (B) Loading efficiency of each step (i.e., large bead, large bead-small bead pair, large bead-small bead-single cell combination). Bars represent mean ± SD from three independent experiments. (C) Calibration curve of the beads-based on-chip multiplexed detection of cytokines using mixtures of the six cytokine recombinant proteins (i.e., IL-2, TNF-α, IFN-γ, GM-CSF, MIP-1α and IL-8) with a series of dilution. For TNF-α and IL-8, the signal intensity ranges were significantly larger than the other four, thus a different Y-axis (right side) was used for these two cytokines. Bars represent mean ± SD from a minimum of 50 beads after on-chip incubation and detection. (D) Crosstalk heatmap for the six cytokines. Individual cytokines were incubated with detection beads and the fluorescent intensities recorded were converted to protein concentrations using the calibration curve.

To demonstrate the multiplexed cytokine detection, we chose six cytokines that were commonly secreted by activated T cells and macrophages and play important roles in execution of their immune functions, i.e., IL-2, TNF-α, IFN-γ, GM-CSF, MIP-1α and IL-8, as our detection targets (Altan-Bonnet and Mukherjee, 2019). Commercial biotinylated capture antibodies for IL-2, TNF-α and IFN-γ were coated on the streptavidin-coated small beads, and capture antibodies for GM-CSF, MIP-1α and IL-8 were coated on the streptavidin-coated large beads. Functionalized beads were loaded onto the chip and serially diluted recombinant protein standards were used for the on-chip calibration. The LOD were determined to be 4.7 pg mL−1 (IL-2), 5.0 pg mL−1 (TNF-α), 108.8 pg mL−1 (IFN-γ), 144.6 pg mL−1 (GM-CSF), 74.6 pg mL (MIP-1α), 6.2 pg mL (IL-8), respectively (Fig. 2C), with a dynamic range of 103−104. These LOD values are fairly comparable to (for IL-2, TNF-α and IL-8), or higher than (for IFN-γ, GM-CSF and MIP-1α) the values of conventional ELISA methods for bulk assays (with LOD values of 2–10 pg/mL for commercial ELISA kits, Table S2). The LOD values are mainly determined by the affinities of the chosen antibodies. Though improvable, we found that the sensitivities were sufficient for the on-chip detection of the single-cell secretion, mainly owing to the small media volume in the microwells and the proximity between the cell and the paired beads, which could significantly increase local cytokine concentration and cytokine capture efficiency. Based on the calculation and simulation in our previous work (Zhou et al., 2020), sensitivity of the HL-Chip was estimated to be within a range of 18–600 copies/ cell (Table S2), which is comparable to other microwell/microchamber-based methods for single-cell secretion detection (Chen et al., 2019). Nevertheless, the sensitivities could be further improved by choosing antibodies with higher affinities and specificities, or employment of signal amplification methods (Xiong et al., 2020). To evaluate crosstalk among different antibodies, only one protein standards (1000 pg mL−1) and all the six detection antibodies were applied on chip each time, and signals corresponding to all the six targets were recorded and quantified (Fig. S2 and Fig. 2D). The crosstalk heatmap (Fig.2D) demonstrated good antibody specificity and quantification capability of the platform.

3.3. Multiplexed detection of single-cell secretions with HL-Chips

To validate the HL-Chip method for multiplexed detection of single-cell secretions, two different cell types, human CD4+ T cells and a human monocytic cell line THP-1, were tested. The diameter of the central well for accommodating the cell was determined to be 11 μm for T cells, which have an average diameter of ~8 μm, and 16 μm for THP-1 cells, which have an average diameter of ~14 μm. To detect single-cell secretions from CD4+ T cells, purified cells were loaded onto the chip after loading of the cytokine-detection beads, then incubated in AIM-V medium containing either PMA-ionomycin or DMSO for 1.5 h before on-chip detection of the single-cell cytokine secretions. PMA-ionomycin directly stimulates protein kinase C and Ca2+ influx, bypassing TCR stimulation, inducing rapid T cell activation and cytokine secretion (Chatila et al., 1989). Representative images of the single-cell multiplexed detection were shown in Fig. 3A and Fig. S3. Consistent with the results from bulk or single T cells in previous reports (Ai et al., 2013; Zhou et al., 2020), PMA-ionomycin induced rapid secretion of several cytokines compared to DMSO, as shown in the heatmaps in Fig. 3B. Each row in the heatmaps corresponds to a single cell, and each column corresponds to a cytokine of interest. The heatmaps also show significant heterogeneity among the CD4+ T cell population from the same source. To calculate the percentages of secretion-positive cells, signal intensities from beads in no-cell control wells were recorded to set up thresholds for discriminating positive signals. As shown in Fig. 3C and Fig. S5A, only 10.8%, 11.0% and 15.1% of the transiently-stimulated CD4+ T cells secreted IL-2, GM-CSF and MIP-1α, respectively, while 75.0% and 57.3% of the stimulated T cells secreted TNF-α and IFN-γ, respectively, suggesting different cytokine-secretion response to PMA/ionomycin stimulation. Only 2.0% of the stimulated T cells secrete IL-8. In comparison to the stimulated T cells, the vast majority of the unstimulated T cells didn’t secrete detectable level of any of the six cytokines.

Fig. 3.

Fig. 3.

Open in a new tab

Multiplexed detection of single cell cytokine secretions with HL-Chip. (A) Representative bright-field and fluorescent images of the 6-plex cytokine detection from single CD4+ T cells stimulated with PMA/ionomycin (top) or DMSO (bottom). GM-CSF, MIP-1α and IL-8 were detected using the large beads and IL-2, TNF-α and IFN-γ were detected using the small beads. Scale bar, 25 µm. (B,D) Heatmaps of single-cell cytokine secretion profiles of CD4+ T cells stimulated with PMA/ionomycin or DMSO (B), and PMA-differentiated THP-1 macrophages stimulated with LPS or DMSO (D). Each row represents a single cell’s cytokine profile, and each column is a cytokine of interest. (C,E) Vertical scatterplots comparing single-cell secretion of IL-2, TNF-α, IFN-γ, and GM-CSF from stimulated (red) and non-stimulated (blue) CD4+ T cells (C), and TNF-α, IL-8, GM-CSF and MIP-1α from stimulated (red) and non-stimulated (blue) THP-1 macrophages (E). The black dash lines indicate thresholds for the cytokines, determined by no-cell data. Each dot represents converted cytokine concentration from a single cell. Results are from three independent experiments. For (B) and (C), number of cells n=511 and 500 for the PMA/ionomycin group and DMSO group, respectively, from three independent experiments. For (D) and (E), N=539 and 524 for the LPS group and DMSO group, respectively, from three independent experiments. Bars represent median values with interquartile ranges. **** P<0.0001, Welch’s t-tests.

To detect single-cell secretions from THP-1 cells in response to stimulation, THP-1 cells were first stimulated with PMA off-chip for 24 hours to promote a macrophage differentiation, then loaded to the chips and either stimulated with LPS or DMSO on-chip for 4 h, followed by on-chip detection (Fig. S4). It is known that LPS induces inflammatory response and secretion of many proinflammatory cytokines in THP-1-differentiated macrophages (Chanput et al., 2014). As expected, significant amounts of the LPS-stimulated THP-1 macrophages secreted proinflammatory cytokines such as TNF-α (76.6%), MIP-1α (46.2%), IL-8 (38.6%) and GM-CSF (22.8%), while the percentages were 4.2%, 3.6%, 1.5% and 10.7%, respectively, without LPS stimulation (Fig. 3D, E). While the results indicated partial basal secretions in the PMA-differentiated macrophages without LPS stimulation, the secretion magnitudes were lower (except for GM-CSF), which is consistent with previous work (Zhao et al., 2018). The results also show that THP-1 macrophages barely secrete IL-2 (0.6% and 0.4%, respectively) and only small percentages (3.2% and 2.7%, respectively) of THP-1 macrophages secrete IFN-γ, whether LPS-stimulated or not (Fig. S5B).

To demonstrate the live-cell retrieval capability of the HL-Chip, the HL-Chip was loaded with unstimulated THP-1 cells. After the on-chip assay, single live cells were retrieved from the chip using a micromanipulation setup that we have described before (Zhou et al., 2020). The retrieved cells maintained their viability, indicated by the bright Calcein AM fluorescence. Functionality of the retrieved cells were demonstrated by the capability to differentiate to macrophages after PMA stimulation and the capability to proliferate and form single-cell clones (Figure S6).

3.4. Single-cell secretion profiling of human TCR-T cells

We next applied this technology to single-cell cytokine profiling of a more clinically relative cell type, the NY-ESO-1-specific TCR-T cells. NY-ESO-1, as a cancer-testis antigen, is not expressed in normal nongermline tissues but is aberrantly expressed in many different types of tumors, making it a very attractive target for cancer immunotherapy. NY-ESO-1-specific TCR-T cell-based immunotherapies have shown good safety and efficacies in clinical trials for treating solid tumors (Bethune et al., 2018; Wang et al., 2017). High-resolution single-cell assessment of functional outcomes of the TCR-T cells, such as cytokine secretion after encounter with antigen, may provide new insights into the variability of clinical responses and contribute to the development of better immunotherapies. To generate the NY-ESO-1-specific TCR-T cells, CD8+ T cells from a healthy donor were transduced with retrovirus to express the 1G4 TCR that recognizes the HLA-A*02:01-presented epitope NY-ESO-1157-165 (Zhou et al., 2020). The transduction efficiency was evaluated to be 64.0% by flow cytometry (Fig. S7).

The duration of antigen stimulation is critical in determining T cell activation and cytokine responses (Iezzi et al., 1998; Kaveh et al., 2012; Pala et al., 2000). To investigate cytokine secretion under different stimulation time, the NY-ESO-1-specific TCR-T cells were stimulated with NY-ESO-V157–165 peptide (which shows higher binding affinity to MHC and 1G4 TCR and stronger antigenicity compared to wild type NY-ESO-1157–165 peptide) (Zhou et al., 2020) for different durations (i.e., 1 h, 2 h, 4 h and 6 h). Then the T cells were harvested, washed and loaded to the chip and incubated in a standard cell-culture incubator for 5 h before in-situ staining (Fig. S8). The cytokine secretion heatmaps show significant yet very heterogeneous cytokine secretion profiles among single TCR-T cells after peptide stimulation (Fig. 4A). IL-2, TNF-α, IFN-γ and GM-CSF are the four major secreted cytokines while MIP-1α and IL-8 secretion was scarce (Fig. 4A, 4B and Fig. S9). The data also show an early burst of cytokine secretion with a stimulation duration as short as 1 h, and a significant dropping of the secretion with 4 h and 6 h stimulation (Fig. 4B and 4C). In addition, IFN-γ was shown to be the most persistent effector-cytokine compared to the other three cytokines, which is in consistency with previous report that IFN-γ secretion favors prolonged and sustained antigen stimulation (Faroudi et al., 2003). Overall, our results indicate that the cytokine secretion in T cell activation via TCR/pMHC interaction is a very timely, yet transient process followed by a refractory phase (Pala et al., 2000). Prolonged antigen stimulation is not necessary for sustained cytokine secretion. We also investigated if the secretion magnitudes also decreased as the percentages of positive cells decreased over time of antigen stimulation (Fig. S10). We found that the secretion magnitudes of TNF-α decreased significantly with prolonged antigen stimulation while those of IFN-γ were retained. The secretion magnitudes of IL-2 and MIP-1α were fluctuating.

Fig. 4.

Fig. 4.

Open in a new tab

Single-cell secretion profiling of human NY-ESO-1-specific TCR-T cells with different antigen-pulsing time. (A) Heatmaps of single-cell cytokine secretion profiles of the TCR-T cells stimulated with NY-ESO-V peptide for a series of duration or control. (B) Vertical scatterplots comparing single-cell secretion of IL-2, TNF-α, IFN-γ, and GM-CSF from the TCR-T cells pulsed with NY-ESO-V peptide for a series of duration (red) or control (blue). Results are from two or three independent experiments for (A,B). The numbers of cells were between 281 and 442. Bars represent median values with interquartile ranges. (C) Heatmap of the percentages of each cytokine-positive cells over different antigen-pulsing time. (D) Polyfunctional heatmap showing 15 TCR-T cell subsets. Heatmap intensities represent proportions of the corresponding subsets.

It has been shown that T cells capable of producing multiple cytokines, termed polyfunctional T cells, provide more effective immune responses against viral infection or cancer (Han et al., 2012; Xue et al., 2017). To investigate polyfunctionality of the NY-ESO-1-specific TCR-T cells after NY-ESO peptide stimulation, the TCR-T cells were categorized into 15 functional subsets based on all the possible combinations of the secretion of IL-2, TNF-α, IFN-γ and GM-CSF (Fig. 4D). Each row represents one pulsing time, and each column represents one functional subset. The functional heatmap shows significant functional and temporal heterogeneity in the same antigen-stimulated TCR-T cell population. With 1-h peptide pulsing, 10.63% of the whole population secreted all the four cytokines (subset 1) and 19.46% secreted IL-2, TNF-α, IFN-γ but not GM-CSF (subset 2), which were the two largest subsets. This is consistent with other reports that polyfunctional T cells are the major effector cells against viral infection or cancers. Among other subsets, 2.26%, 5.20%, 3.17% and 3.85% of the whole population solely secreted IL-2, TNF-α, IFN-γ and GM-CSF, respectively (subset 12–15). The percentages of all the other subsets were less than 3%, except for the subset 8 that secreted TNF-α and IFN-γ (4.52%). With 2-h stimulation, subset 1 and 2 were still the two largest subsets despite a slight decrease of the percentages (8.74% and 15.03%, respectively). Among other subsets, the subset 7 that secreted IL-2 and IFN-γ shows a percentage increase from 0.23% to 3.15%, the subset 8 that secreted TNF-α and IFN-γ shows a percentage increase from 4.52% to 7.69%, the subsets 12–14 that solely secreted IL-2, TNF-α, and IFN-γ also show increased percentages to 5.94%, 6.29%, 8.74%, respectively. On the contrary, the percentage of the subset 15 that solely secreted GM-CSF decreased from 3.85% to 2.80%. The percentages of all the other subsets were less than 3%. With 4-h and 6-h peptide stimulation, all the subsets show significant percentage decrease, except for the subsets 14 and 15 (solely secreted IFN-γ and GM-CSF, respectively), of which the percentages only decreased slightly (from 8.74% and 2.80% in the 2-h row to 6.88% and 2.01% in the 6-h row, respectively). Our results show that polyfunctional cytokine responses became dominating soon after short-time stimulation but are more transient, while monofunctional responses with sole IFN-γ or GM-CSF secretion were more persistent. This is consistent with previous work that investigated dynamics of IL-2, TNF-α, and IFN-γ secretion after PMA/ionomycin stimulation or anti-CD3/CD28 bead stimulation (Han et al., 2012). It has been reported that the duration of antigenic stimulation is a major determining factor for the fate of naive and effector T cells (Iezzi et al., 1998). Our results indicate that it is also a major determining factor for TCR-T cell function.

4. Conclusions

Single-cell secretion analysis, which characterizes cellular functional heterogeneity, is particularly important for evaluation of immune cells with a highly heterogeneous and dynamic nature. Although there have been some technologies targeting multiplexed single-cell protein secretion detection, such as microengraving, single-cell barcode chips and MIST, all of them suffer from uncontrolled cell sedimentation, which results in significant wasting of the cells and assay units due to a high percentage of empty or multi-cell-occupied units. On the contrary, the multiplexed HL-chip method we introduced here deterministically position single cells with multiple cytokine detection beads, resulting in a much higher single-cell occupancy (~78% successful bead-single-cell-bead combination). Beads-cell loading efficiency could be further enhanced by modifications to the microwell design, including refining the relative position and angles of the three connected wells, optimizing the spatial arrangement of the array, and adapting different well depths to better fit the different sizes of beads and cells. Alternative particle manipulation methods, such as acoustic, electrical, optical and magnetic techniques, could also be adapted to further enhance the precision of cell and bead loading (Zhang et al., 2020). In addition, by utilizing different bead sizes and fluorescent spectrum, we provided a simple and expandable approach for highly multiplexed detection. To further expand the detection capacity beyond the 6-plex detection we have achieved, three strategies could be applied. First, it is suggested to preferably choose well-validated commercial antibodies and perform an antibody crosstalk test before using on HL-Chips to minimize antibody crosstalk. Second, the numbers of the fluorescent labelling should be adjusted according to the users’ specific fluorescent detection setup to avoid spectral overlap. In addition, fluorescent labels with narrow emission spectrum, such as quantum dots, could be applied to enhance spectral multiplicity without spectral overlapping. Third, the bead-based detection is also compatible with the coding and decoding principle of MIST approach or DNA-barcoding coupled with sequencing approaches, which could further increase the multiplicity to hundreds (Peterson et al., 2017; Wu et al., 2023; Zhao et al., 2018). Given that all the beads and conjugated antibodies we used are commercial, the whole procedure is similar to standard sandwich ELISA assays, lowering the operation requirements. The capture antibody-conjugated beads stored at 4 °C for over six months didn’t show a notable decrease in binding capacity during our testing. The storage lifetime of the preloaded HL-Chip needs to be further estimated in the future. Furthermore, with adjusted microwell structure, this method can also be applied in studying cell-cell interaction in combination with multiplexed cytokine detection. Due to the open-well nature of the HL-Chip, the interested cells can be picked up after the cytokine detection assay for further culture and assaying, as we have demonstrated in this work and our previous work (Zhou et al., 2020). A detailed comparison of HL-Chip method with other single-cell secretion detection methods is presented in Table S3. Applying the multiplexed HL-chip method, we observed vast heterogeneity in cytokine secretion in T cells and macrophages. In particular, we revealed the heterogenous and dynamic polyfunctionality in antigen-stimulated TCR-T cells originated from the same donor’s CD8+ T cells. This finding implies significant functional and temporal heterogeneity in the cells used in adoptive cell therapies, which may also correlate to the inconsistency of clinical outcomes. Tumor-specific CD8+ T cells (TST) in patients with progressing tumors are always dysfunctional/exhausted, which is thought to be driven by persistent antigenic stimulation of TCR. This dysfunction/exhaustion could be surprisingly established within hours of tumor antigen encounter (Rudloff et al., 2023). Our in-vitro investigation of the TCR-T cells revealed that short antigen exposure resulted in a strong burst of cytokine secretion while prolonged antigen exposure significantly reduced the secretion, which is in line with the previous in-vivo results. Other findings emphasize the role of the affinity and kinetics of TCR-pMHC interaction in the execution of T cell function (Dustin et al., 2010; Shevyrev et al., 2022; Zhou et al., 2020). These findings collectively suggest that the kinetics of antigen-receptor interaction and the resulting functional dynamics should be considered together when designing/optimizing natural or artificial antigen receptors. Our highly multiplexed, single-cell method could be useful for comprehensive assessment of the functional diversity and dynamics of the cell products and guiding their optimization. In summary, we have developed a simple, robust and quantitative approach for highly multiplexed single-cell secretion profiling by efficiently aligning thousands of single cells with multiple antibody-coated microbeads on the HL-Chip. We profiled cytokine secretion of single macrophages and T cells after pan- and antigen-specific stimulation, revealing functional and temporal heterogeneity in those immune cells. Given its simplicity and versatility, we envision the great potential of this technology in comprehensive single-cell secretion measurements in both research and clinical applications.

Supplementary Material

1
2
3
4

Appendix A. Supplementary data

Supplementary data to this article can be found online.

Acknowledgements

We are grateful for funding support from NIH U01CA244107.

Footnotes

Publisher's Disclaimer: This is a PDF file of an unedited manuscript that has been accepted for publication. As a service to our customers we are providing this early version of the manuscript. The manuscript will undergo copyediting, typesetting, and review of the resulting proof before it is published in its final form. Please note that during the production process errors may be discovered which could affect the content, and all legal disclaimers that apply to the journal pertain.

CRediT authorship contribution statement

Ning Shao: Conceptualization, Methodology, Data curation, Writing - original draft, Writing - review & editing.

Yufu Zhou: Conceptualization, Methodology.

Xuewu Liu: Project administration, Resources, Writing - review & editing.

Declaration of competing interest

The authors declare no competing interests.

Declaration of interests

The authors declare that they have no known competing financial interests or personal relationships that could have appeared to influence the work reported in this paper.

References

  1. Abdullah MAA, Wang J, 2019. Ultrasimple Single-Cell Detection of Multiple Cytokines by a Nanowell Chip Integrated with Encoded Microarrays. ACS Sens. 4(9), 2296–2302. [DOI] [PMC free article] [PubMed] [Google Scholar]
  2. Ai W, Li H, Song N, Li L, Chen H, 2013. Optimal method to stimulate cytokine production and its use in immunotoxicity assessment. Int. J. Environ. Res. Public Health 10(9), 3834–3842. [DOI] [PMC free article] [PubMed] [Google Scholar]
  3. Altan-Bonnet G, Mukherjee R, 2019. Cytokine-mediated communication: a quantitative appraisal of immune complexity. Nat. Rev. Immunol 19(4), 205–217. [DOI] [PMC free article] [PubMed] [Google Scholar]
  4. Bendall SC, Simonds EF, Qiu P, Amir el AD, Krutzik PO, Finck R, Bruggner RV, Melamed R, Trejo A, Ornatsky OI, Balderas RS, Plevritis SK, Sachs K, Pe’er D, Tanner SD, Nolan GP, 2011. Single-cell mass cytometry of differential immune and drug responses across a human hematopoietic continuum. Science 332(6030), 687–696. [DOI] [PMC free article] [PubMed] [Google Scholar]
  5. Bethune MT, Li XH, Yu J, McLaughlin J, Cheng D, Mathis C, Moreno BH, Woods K, Knights AJ, Garcia-Diaz A, Wong S, Hu-Lieskovan S, Puig-Saus C, Cebon J, Ribas A, Yang L, Witte ON, Baltimore D, 2018. Isolation and characterization of NY-ESO-1-specific T cell receptors restricted on various MHC molecules. Proc. Natl. Acad. Sci. U. S. A 115(45), E10702–E10711. [DOI] [PMC free article] [PubMed] [Google Scholar]
  6. Bucheli OTM, Sigvaldadottir I, Eyer K, 2021. Measuring single-cell protein secretion in immunology: Technologies, advances, and applications. Eur. J. Immunol 51(6), 1334–1347. [DOI] [PMC free article] [PubMed] [Google Scholar]
  7. Chanput W, Mes JJ, Wichers HJ, 2014. THP-1 cell line: an in vitro cell model for immune modulation approach. Int. Immunopharmacol 23(1), 37–45. [DOI] [PubMed] [Google Scholar]
  8. Chatila T, Silverman L, Miller R, Geha R, 1989. Mechanisms of T cell activation by the calcium ionophore ionomycin. J. Immunol 143(4), 1283–1289. [PubMed] [Google Scholar]
  9. Chen Z, Chen JJ, Fan R, 2019. Single-Cell Protein Secretion Detection and Profiling. Annu. Rev. Anal. Chem 12(1), 431–449. [DOI] [PubMed] [Google Scholar]
  10. Chokkalingam V, Tel J, Wimmers F, Liu X, Semenov S, Thiele J, Figdor CG, Huck WT, 2013. Probing cellular heterogeneity in cytokine-secreting immune cells using droplet-based microfluidics. Lab Chip 13(24), 4740–4744. [DOI] [PubMed] [Google Scholar]
  11. Czerkinsky CC, Nilsson LA, Nygren H, Ouchterlony O, Tarkowski A, 1983. A solid-phase enzyme-linked immunospot (ELISPOT) assay for enumeration of specific antibody-secreting cells. J. Immunol. Methods 65(1–2), 109–121. [DOI] [PubMed] [Google Scholar]
  12. Dinarello CA, 2007. Historical insights into cytokines. Eur. J. Immunol 37(S1), 34–45. [DOI] [PMC free article] [PubMed] [Google Scholar]
  13. Dustin ML, Long EO, 2010. Cytotoxic immunological synapses. Immunol. Rev 235(1), 24–34. [DOI] [PMC free article] [PubMed] [Google Scholar]
  14. Eyer K, Doineau RCL, Castrillon CE, Briseno-Roa L, Menrath V, Mottet G, England P, Godina A, Brient-Litzler E, Nizak C, Jensen A, Griffiths AD, Bibette J, Bruhns P, Baudry J, 2017. Single-cell deep phenotyping of IgG-secreting cells for high-resolution immune monitoring. Nat. Biotechnol 35(10), 977–982. [DOI] [PubMed] [Google Scholar]
  15. Faroudi M, Utzny C, Salio M, Cerundolo V, Guiraud M, Muller S, Valitutti S, 2003. Lytic versus stimulatory synapse in cytotoxic T lymphocyte/target cell interaction: manifestation of a dual activation threshold. Proc. Natl. Acad. Sci. U. S. A 100(24), 14145–14150. [DOI] [PMC free article] [PubMed] [Google Scholar]
  16. Gazagne A, Claret E, Wijdenes J, Yssel H, Bousquet F, Levy E, Vielh P, Scotte F, Goupil TL, Fridman WH, Tartour E, 2003. A Fluorospot assay to detect single T lymphocytes simultaneously producing multiple cytokines. J. Immunol. Methods 283(1–2), 91–98. [DOI] [PubMed] [Google Scholar]
  17. Genin M, Clement F, Fattaccioli A, Raes M, Michiels C, 2015. M1 and M2 macrophages derived from THP-1 cells differentially modulate the response of cancer cells to etoposide. BMC Cancer 15, 577. [DOI] [PMC free article] [PubMed] [Google Scholar]
  18. Han Q, Bagheri N, Bradshaw EM, Hafler DA, Lauffenburger DA, Love JC, 2012. Polyfunctional responses by human T cells result from sequential release of cytokines. Proc. Natl. Acad. Sci. U. S. A 109(5), 1607–1612. [DOI] [PMC free article] [PubMed] [Google Scholar]
  19. Iezzi G, Karjalainen K, Lanzavecchia A, 1998. The duration of antigenic stimulation determines the fate of naive and effector T cells. Immunity 8(1), 89–95. [DOI] [PubMed] [Google Scholar]
  20. Jammes FC, Maerkl SJ, 2020. How single-cell immunology is benefiting from microfluidic technologies. Microsyst. Nanoeng 6, 45. [DOI] [PMC free article] [PubMed] [Google Scholar]
  21. Ji Y, Qi D, Li L, Su H, Li X, Luo Y, Sun B, Zhang F, Lin B, Liu T, Lu Y, 2019. Multiplexed profiling of single-cell extracellular vesicles secretion. Proc. Natl. Acad. Sci. U. S. A 116(13), 5979–5984. [DOI] [PMC free article] [PubMed] [Google Scholar]
  22. Jin A, Ozawa T, Tajiri K, Obata T, Kondo S, Kinoshita K, Kadowaki S, Takahashi K, Sugiyama T, Kishi H, Muraguchi A, 2009. A rapid and efficient single-cell manipulation method for screening antigen-specific antibody-secreting cells from human peripheral blood. Nat. Med 15(9), 1088–1092. [DOI] [PubMed] [Google Scholar]
  23. Jung T, Schauer U, Heusser C, Neumann C, Rieger C, 1993. Detection of intracellular cytokines by flow cytometry. J. Immunol. Methods 159(1–2), 197–207. [DOI] [PubMed] [Google Scholar]
  24. Junkin M, Tay S, 2014. Microfluidic single-cell analysis for systems immunology. Lab Chip 14(7), 1246–1260. [DOI] [PubMed] [Google Scholar]
  25. Kaveh DA, Whelan AO, Hogarth PJ, 2012. The duration of antigen-stimulation significantly alters the diversity of multifunctional CD4 T cells measured by intracellular cytokine staining. PloS one 7(6), e38926. [DOI] [PMC free article] [PubMed] [Google Scholar]
  26. Konry T, Dominguez-Villar M, Baecher-Allan C, Hafler DA, Yarmush ML, 2011. Droplet-based microfluidic platforms for single T cell secretion analysis of IL-10 cytokine. Biosens. Bioelectron 26(5), 2707–2710. [DOI] [PMC free article] [PubMed] [Google Scholar]
  27. Li L, Su H, Ji Y, Zhu F, Deng J, Bai X, Li H, Liu X, Luo Y, Lin B, Liu T, Lu Y, 2023. Deciphering Cell-Cell Interactions with Integrative Single-Cell Secretion Profiling. Adv. Sci 10(19), e2301018. [DOI] [PMC free article] [PubMed] [Google Scholar]
  28. Love JC, Ronan JL, Grotenbreg GM, van der Veen AG, Ploegh HL, 2006. A microengraving method for rapid selection of single cells producing antigen-specific antibodies. Nat. Biotechnol 24(6), 703–707. [DOI] [PubMed] [Google Scholar]
  29. Lovelace P, Maecker HT, 2018. Multiparameter Intracellular Cytokine Staining. Methods Mol. Biol 1678, 151–166. [DOI] [PubMed] [Google Scholar]
  30. Lu Y, Xue Q, Eisele MR, Sulistijo ES, Brower K, Han L, Amir el AD, Pe’er D, Miller-Jensen K, Fan R, 2015. Highly multiplexed profiling of single-cell effector functions reveals deep functional heterogeneity in response to pathogenic ligands. Proc. Natl. Acad. Sci. U. S. A 112(7), E607–615. [DOI] [PMC free article] [PubMed] [Google Scholar]
  31. Ma C, Fan R, Ahmad H, Shi Q, Comin-Anduix B, Chodon T, Koya RC, Liu CC, Kwong GA, Radu CG, Ribas A, Heath JR, 2011. A clinical microchip for evaluation of single immune cells reveals high functional heterogeneity in phenotypically similar T cells. Nat. Med 17(6), 738–743. [DOI] [PMC free article] [PubMed] [Google Scholar]
  32. Manz R, Assenmacher M, Pfluger E, Miltenyi S, Radbruch A, 1995. Analysis and sorting of live cells according to secreted molecules, relocated to a cell-surface affinity matrix. Proc. Natl. Acad. Sci. U. S. A 92(6), 1921–1925. [DOI] [PMC free article] [PubMed] [Google Scholar]
  33. O’Neil-Andersen NJ, Lawrence DA, 2002. Differential modulation of surface and intracellular protein expression by T cells after stimulation in the presence of monensin or brefeldin A. Clin. Diagn. Lab Immunol 9(2), 243–250. [DOI] [PMC free article] [PubMed] [Google Scholar]
  34. Pala P, Verhoef A, Lamb JR, Openshaw PJ, 2000. Single cell analysis of cytokine expression kinetics by human CD4+ T-cell clones during activation or tolerance induction. Immunology 100(2), 209–216. [DOI] [PMC free article] [PubMed] [Google Scholar]
  35. Peterson VM, Zhang KX, Kumar N, Wong J, Li L, Wilson DC, Moore R, McClanahan TK, Sadekova S, Klappenbach JA, 2017. Multiplexed quantification of proteins and transcripts in single cells. Nat. Biotechnol 35(10), 936–939. [DOI] [PubMed] [Google Scholar]
  36. Rudloff MW, Zumbo P, Favret NR, Roetman JJ, Detres Roman CR, Erwin MM, Murray KA, Jonnakuti ST, Dundar F, Betel D, Philip M, 2023. Hallmarks of CD8(+) T cell dysfunction are established within hours of tumor antigen encounter before cell division. Nat. Immunol 24(9), 1527–1539. [DOI] [PMC free article] [PubMed] [Google Scholar]
  37. Shao N, Qin L, 2018. Biochips—New Platforms for Cell‐Based Immunological Assays. Small Methods 2(1), 1700254. [Google Scholar]
  38. Shevyrev DV, Tereshchenko VP, Sennikov SV, 2022. The Enigmatic Nature of the TCR-pMHC Interaction: Implications for CAR-T and TCR-T Engineering. Int. J. Mol. Sci 23(23). [DOI] [PMC free article] [PubMed] [Google Scholar]
  39. Shi Q, Qin L, Wei W, Geng F, Fan R, Shin YS, Guo D, Hood L, Mischel PS, Heath JR, 2012. Single-cell proteomic chip for profiling intracellular signaling pathways in single tumor cells. Proc. Natl. Acad. Sci. U. S. A 109(2), 419–424. [DOI] [PMC free article] [PubMed] [Google Scholar]
  40. Stanley AC, Lacy P, 2010. Pathways for cytokine secretion. Physiology 25(4), 218–229. [DOI] [PubMed] [Google Scholar]
  41. Torres AJ, Hill AS, Love JC, 2014. Nanowell-based immunoassays for measuring single-cell secretion: characterization of transport and surface binding. Anal. Chem 86(23), 11562–11569. [DOI] [PMC free article] [PubMed] [Google Scholar]
  42. Wang RF, Wang HY, 2017. Immune targets and neoantigens for cancer immunotherapy and precision medicine. Cell Res 27(1), 11–37. [DOI] [PMC free article] [PubMed] [Google Scholar]
  43. Wei SC, Hsu MN, Chen CH, 2019. Plasmonic droplet screen for single-cell secretion analysis. Biosens. Bioelectron 144, 111639. [DOI] [PubMed] [Google Scholar]
  44. Wu T, Womersley HJ, Wang JR, Scolnick J, Cheow LF, 2023. Time-resolved assessment of single-cell protein secretion by sequencing. Nat. Methods 20(5), 723–734. [DOI] [PubMed] [Google Scholar]
  45. Xiong Y, Leng Y, Li X, Huang X, Xiong Y, 2020. Emerging strategies to enhance the sensitivity of competitive ELISA for detection of chemical contaminants in food samples. Trends Anal. Chem 126, 115861. [Google Scholar]
  46. Xue Q, Bettini E, Paczkowski P, Ng C, Kaiser A, McConnell T, Kodrasi O, Quigley MF, Heath J, Fan R, Mackay S, Dudley ME, Kassim SH, Zhou J, 2017. Single-cell multiplexed cytokine profiling of CD19 CAR-T cells reveals a diverse landscape of polyfunctional antigen-specific response. J. Immunother. Cancer 5(1), 85. [DOI] [PMC free article] [PubMed] [Google Scholar]
  47. Zhang S, Wang Y, Onck P, & den Toonder J, 2020. A concise review of microfluidic particle manipulation methods. Microfluidics and Nanofluidics, 24(4), 24. [Google Scholar]
  48. Zhao P, Bhowmick S, Yu J, Wang J, 2018. Highly Multiplexed Single-Cell Protein Profiling with Large-Scale Convertible DNA-Antibody Barcoded Arrays. Adv. Sci 5(9), 1800672. [DOI] [PMC free article] [PubMed] [Google Scholar]
  49. Zhou Y, Shao N, Bessa de Castro R, Zhang P, Ma Y, Liu X, Huang F, Wang RF, Qin L, 2020. Evaluation of Single-Cell Cytokine Secretion and Cell-Cell Interactions with a Hierarchical Loading Microwell Chip. Cell Rep 31(4), 107574. [DOI] [PMC free article] [PubMed] [Google Scholar]

Associated Data

This section collects any data citations, data availability statements, or supplementary materials included in this article.

Supplementary Materials

1
NIHMS2091697-supplement-1.docx (7.8MB, docx)
2
NIHMS2091697-supplement-2.xlsx (13.7KB, xlsx)
3
NIHMS2091697-supplement-3.xlsx (12KB, xlsx)
4
NIHMS2091697-supplement-4.xlsx (14.1KB, xlsx)

RESOURCES