Multiplexed, Sequential Secretion Analysis of the Same Single Cells Reveals Distinct Effector Response Dynamics Dependent on the Initial Basal State

Zhuo Chen, Yao Lu, Kerou Zhang, Yang Xiao, Jun Lü, Rong Fan

Abstract

The effector response of immune cells is a highly dynamic process governed by the secretion of various proteins. Understanding these dynamics requires sequential measurement of the same single cells over time. This study introduces a microchip-based, 10-plexed sequential secretion assay capable of monitoring approximately 5,000 single cells across multiple time points. By applying this platform to human macrophages stimulated with lipopolysaccharide (LPS), researchers identified distinct activation modes and discovered that a cell's initial basal state significantly influences its subsequent inflammatory response. These results demonstrate a cell-intrinsic heterogeneous response within a phenotypically homogeneous population.

Introduction: Beyond the Snapshot

Immune responses are inherently temporal. To understand how the body combats pathogens, it is essential to track the secretion of cytokines and chemokines in real-time. Traditional methods like flow cytometry or ELISpot provide only "snapshots"—single data points that often require the destruction of the cell. While single-cell RNA sequencing (scRNA-Seq) offers deep insights into gene expression, the correlation between mRNA levels and actual protein secretion is frequently poor. There is a critical need for longitudinal tracking to observe how the functional output of live cells evolves over hours or days.

A High-Throughput Platform for Longitudinal Secretomics

The researchers developed a novel microchip-based assay consisting of two primary components: a PDMS microtrough array and antibody barcode slides. The microtrough array features a "honeycomb" structure that captures and holds thousands of single cells. The antibody barcode slides are patterned with microscopic stripes of antibodies targeting specific proteins.

The innovation of this system lies in its ability to swap antibody slides sequentially. Every two hours, the researchers replaced the slide—which had captured proteins secreted during that interval—with a fresh one. This allowed for the measurement of 10 different proteins (including TNF, IL-6, and CCL2) from the same individual cells across four distinct time points. This method achieved a five-fold increase in throughput compared to previous microfluidic secretomic technologies, capturing data from over 5,000 cells per chip.

Results: The Predictive Power of the Basal State

The study focused on U937-derived human macrophages stimulated with Lipopolysaccharide (LPS), a bacterial component that triggers strong immune signaling. Analysis of 1,752 single cells with complete longitudinal data revealed several key insights into immune cell behavior.

The Basal State as a Predictor

One of the most striking findings was that a cell’s activity in its "resting" (basal) state strongly predicts its future response to stimulation. There was a near-perfect linear correlation ($R^2$ between 0.96 and 0.99) between a cell's basal polyfunctionality—the variety of proteins it secretes at rest—and its averaged polyfunctionality after activation. Cells that were more active at rest became the "super-responders" upon encountering the pathogen.

Divergent Activation Pathways

Upon activation, the macrophage population did not respond uniformly. Instead, the cells split into two distinct functional states, as confirmed by both protein data and scRNA-Seq:

  • Cluster 0: Primarily focused on protein translation, building the cellular machinery required for large-scale protein production.
  • Cluster 1: Primarily focused on the inflammatory response, actively secreting chemical signals to alert the immune system.

These states appeared to be mutually exclusive in the short term, suggesting a biological "fork in the road" where a cell prioritizes either internal preparation or external signaling.

Temporal Secretion Modes

The researchers identified four distinct modes of protein secretion:

  1. On-Off (Transient): Rapid secretion followed by a shutdown. Factors regulated by the transcription factor NF-κB, such as TNF and CCL2, predominantly followed this mode.
  2. Off-On (Delayed): A delayed start to secretion, common for proteins like IL-6.
  3. All-On (Continuous): Sustained secretion throughout the observation period.
  4. Others: Including oscillatory patterns or no detectable secretion.

Integrating Transcriptomics and Proteomics

To validate these findings, the team performed scRNA-Seq on matched samples at 2, 4, and 6 hours. The integrated analysis highlighted a significant "transcript-protein lag." For example, while IL-10 mRNA levels peaked at the 2-hour mark, the actual protein secretion did not reach its maximum until the 4–6 hour window. This discrepancy underscores the importance of measuring the "finished product" (secreted proteins) rather than relying solely on the "instruction manual" (mRNA).

Conclusion and Future Directions

This study demonstrates the power of longitudinal tracking in understanding immune cell heterogeneity. By monitoring thousands of cells over time, the researchers showed that immune responses are not random but are deeply influenced by a cell's intrinsic state.

While the platform currently faces some challenges—such as a 56% cell retention rate during slide swaps—it paves the way for "functional proteomics." Future applications could extend this live-tracking approach to cancer cells or neurons, providing a dynamic view of how individual cells react to drugs or environmental stressors, ultimately leading to a better understanding of how individual reactions constitute a collective population response.

Summary updated

Abstract

The effector response of immune cells dictated by an array of secreted proteins is a highly dynamic process, requiring sequential measurement of all relevant proteins from single cells. Herein, a microchip-based, 10-plexed, sequential secretion assay on the same single cells and at the scale of ≈5000 single cells measured simultaneously over 4 time points are shown. It is applied to investigating the time course of single human macrophage response to toll-like receptor 4 (TLR4) ligand lipopolysaccharide (LPS) and reveals four distinct activation modes for different proteins in single cells. Protein secretion dynamics classifies the cells into two major activation states dependent on the basal state of each cell. Single-cell RNA sequencing performed on the same samples at the matched time points further demonstrates the existence of two major activation states at the transcriptional level, which are enriched for translation versus inflammatory programs, respectively. These results show a cell-intrinsic heterogeneous response in a phenotypically homogeneous cell population. This work demonstrates the longitudinal tracking of protein secretion signature in thousands of single cells at multiple time points, providing dynamic information to better understand how individual immune cells react to pathogenic challenges over time and how they together constitute a population response.
BibTeX
@inproceedings{chen2019multiplexed,
  author = {Zhuo Chen and Yao Lu and Kerou Zhang and Yang Xiao and Jun Lü and Rong Fan},
  title = {{Multiplexed, Sequential Secretion Analysis of the Same Single Cells Reveals Distinct Effector Response Dynamics Dependent on the Initial Basal State}},
  booktitle = {Advanced},
  year = {2019},
  doi = {10.1002/advs.201801361},
}
Figure