Addressing complex autofluorescence signatures in solid tissue samples to enhance full spectrum flow cytometry of non-immune cells.

This paper presents an optimized full spectrum flow cytometry protocol and control strategy that overcomes complex tissue autofluorescence to enable high-dimensional proteomic phenotyping and isolation of rare endothelial cell subpopulations from solid murine tissues.

The Gist

Imagine you are trying to identify a specific group of people in a massive, crowded, and very noisy concert hall.

The Problem:
In the world of biology, scientists have been using a powerful tool called Single-Cell RNA Sequencing (scRNAseq) to read the "instruction manuals" (genes) inside individual cells. It's like reading a person's diary to understand who they are. However, a diary doesn't tell you what they are wearing right now, how they are acting, or who they are shaking hands with.

To see the "clothes and actions" (proteins on the cell surface), scientists use a technique called Flow Cytometry. Think of this as a high-tech security scanner that shines lights on cells as they pass by, reading the colors of their "clothes" to identify them.

The Challenge:
The researchers in this paper were trying to scan Endothelial Cells (ECs). These are the cells that line your blood vessels, acting like the wallpaper inside your body's pipes.

The problem? These cells come from solid organs like the liver and heart. Unlike blood cells (which are clean and easy to scan), cells from solid organs are messy. They are covered in "dirt" and "glow" from the tissue itself.

• The Analogy: Imagine trying to spot a specific person wearing a bright red shirt in a room where everyone is wearing a glowing, neon-green suit. The "neon green" is autofluorescence (natural glow from the tissue). It's so bright and chaotic that it drowns out the specific colors the scientists are trying to see. In the past, this made it impossible to tell the different types of blood vessel cells apart.

The Solution:
The team (Gkantsinikoudi et al.) developed a new, super-advanced way to do this scanning using a machine called the Cytek Aurora. They call it Full Spectrum Flow Cytometry (FSFC).

Here is how they solved the "neon green" problem, explained simply:

1. The "Spectral Unmixing" Magic Trick

Instead of just looking at one color at a time, this new machine captures the entire rainbow of light coming from every cell.

• The Analogy: Imagine the "neon green" glow isn't just one shade; it's a specific, messy pattern of light. The scientists took a "fingerprint" of this messy glow from healthy livers, sick livers, and hearts.
• The Fix: They taught the computer software to recognize that specific "messy pattern" and subtract it from the picture. It's like using noise-canceling headphones to block out the background chatter so you can hear the singer clearly.

2. The "Outfit" Selection (Panel Design)

They had to choose exactly which "colors" (fluorochromes) to tag the cells with.

• The Analogy: If you are trying to identify people in a crowd, you wouldn't give everyone a red hat and a red shirt. You need a mix of unique colors.
• The Strategy: They carefully picked 14 different colors that don't clash with each other. They also picked specific "badges" (antibodies) that only stick to blood vessel cells, ignoring the immune cells and dead cells. They even figured out that for liver cells, they needed a specific "dead cell detector" (a viability dye) that didn't get confused by the liver's natural glow.

3. The "Traffic Control" (Gating and Sorting)

Once they could see the cells clearly, they needed to sort them.

• The Analogy: Imagine a bouncer at a club. The machine can now say, "Okay, you are a blood vessel cell from the liver, and you are wearing a 'Thy1.2' badge. You go to the VIP room."
• The Result: They were able to isolate very rare, special groups of cells that were previously invisible.

Why Does This Matter?

The researchers used this new method to study what happens when the liver gets damaged (like in fatty liver disease or fibrosis).

• The Discovery: They found that blood vessel cells aren't all the same. Some are "peaceful," some are "angry" (inflamed), and some are "changing shape" to become scar tissue (a process called EndMT).
• The Impact: By being able to see and separate these specific groups, scientists can now find new drug targets. Instead of treating the whole liver, they might be able to target just the "angry" blood vessel cells to stop the disease.

Summary in a Nutshell

• Old Way: Trying to read a book in a room full of strobe lights. You can't see anything.
• New Way: They built a special camera that can filter out the strobe lights, identify the specific colors of the book pages, and even separate the pages you want to keep from the ones you don't.

This paper provides the "instruction manual" for other scientists to do the same thing with their own difficult tissue samples, opening the door to understanding diseases in the heart, liver, and beyond with much greater clarity.

Technical Summary

1. Problem Statement

While Single-cell RNA sequencing (scRNAseq) has revolutionized the understanding of cellular heterogeneity, it fails to capture protein expression, post-transcriptional modifications, and cell-surface interactions. Full Spectrum Flow Cytometry (FSFC) offers a solution by enabling high-dimensional proteomic analysis at single-cell resolution. However, the application of FSFC to non-immune cells derived from solid tissues (specifically Endothelial Cells or ECs) has been severely limited.

The primary barrier is complex autofluorescence (AF). Solid tissues, particularly those affected by steatosis (fatty liver) or fibrosis, contain lipids, extracellular matrix proteins, and metabolites that generate unique, high-intensity AF signatures. These signatures:

• Obscure specific fluorochrome signals.
• Complicate spectral unmixing (the mathematical separation of overlapping emission spectra).
• Lead to unmixing errors and inaccurate phenotyping.
• Are highly variable depending on the tissue type, sex, disease model, and time point.

2. Methodology

The authors developed and optimized a comprehensive 14-color FSFC protocol specifically for murine liver and heart endothelial cells. The methodology focuses on overcoming AF challenges through rigorous controls and panel design.

A. Panel Design and Optimization

• Platform: Cytek® Aurora (4 lasers: UV, Violet, Blue, Red; 54+3 detectors).
• Fluorochrome Selection: Used the Cytek® Spectrum Viewer to calculate similarity scores, complexity indices, and stain reduction matrices. The final panel included:
  • 6 EC Identity Markers: CD31, Thrombomodulin (TM), Endomucin (EMCN), LYVE1, Podoplanin (PDPN), MCAM.
  • 3 Activation Markers: ICAM1, LY6A, CD34.
  • 3 Mesenchymal Transition Markers: Transgelin (TAGLN), THY1.2, PDGFRα.
  • Exclusion/Controls: CD45 (to exclude immune cells) and a viability dye.
• Viability Dye Optimization: Tested DNA-binding vs. amine-reactive dyes. LIVE/DEAD Green was selected over NIR dyes because the specific chemical compounds used in their liver injury model induced high autofluorescence in the Near-Infrared (NIR) region, which would have obscured the signal.

B. Critical Controls for Autofluorescence

• Multi-AF Extraction: Instead of a single unstained control, the authors utilized the Cytek® Multi-AF extraction tool. They generated unstained controls for every specific variable (tissue type, sex, disease model, and time point) to capture the unique AF signatures of each condition.
• Reference Controls: Generated single-stain references using both compensation beads and cells to ensure accurate spectral signatures, particularly for tandem dyes.
• FMO (Fluorescence Minus One): Used to establish accurate gating thresholds for positive signals, accounting for spectral spreading.

C. Experimental Models

• Tissues: Murine liver and heart.
• Disease Models:
  • Fibrosis: Induced by Carbon Tetrachloride (CCl4) injections (4 and 8 weeks).
  • Steatosis/Fibrosis: Induced by a Western-type diet (high fat/cholesterol/fructose/sucrose) for up to 24 weeks.

D. Data Analysis Pipeline

• Pre-processing: Used FlowAI for automated quality control (removing anomalies due to flow rate fluctuations).
• Clustering: Applied FlowSOM (Self-Organizing Maps) for unsupervised clustering of EC subpopulations.
• Visualization: Used UMAP (Uniform Manifold Approximation and Projection) for dimensionality reduction.
• Trajectory Analysis: Applied Wishbone to model cellular plasticity and differentiation trajectories (e.g., Endothelial-to-Mesenchymal Transition or EndMT).

3. Key Results

A. Resolution of Endothelial Heterogeneity

• The optimized protocol successfully resolved 12 distinct EC subpopulations in the liver and 7 in the heart.
• Tissue Specificity: Liver ECs showed greater diversity (8 of 12 clusters expressed >4 identity markers) compared to heart ECs (dominated by a single large subpopulation).
• Zonation: The panel successfully recapitulated known zonation patterns (e.g., peri-portal vs. central vein in liver; endocardial vs. epicardial in heart) identified via immunohistochemistry.

B. Impact of Disease on Autofluorescence

• Healthy livers exhibited 3–4 distinct AF signatures.
• Fibrotic livers (8 weeks CCl4) exhibited up to 7 distinct AF signatures, significantly increasing panel complexity.
• The Multi-AF extraction strategy was essential to prevent unmixing errors in these complex models.

C. Discovery of Rare Subpopulations and Plasticity

• EndMT Identification: The protocol identified a pro-fibrotic EC phenotype expressing THY1.2 (a mesenchymal marker) preferentially on CD31^low cells.
• Inflammatory Phenotype: A pro-inflammatory LY6A⁺ phenotype was found preferentially on CD31^high cells.
• Trajectory Analysis: Wishbone analysis successfully mapped the trajectory of ECs transitioning from a quiescent state (ICAM1^- TAGLN^-) to activated/fibrotic states in response to Western diet and CCl4 injury.

D. Full Spectrum Sorting

• The authors demonstrated the ability to sort rare, complex EC subpopulations (e.g., CD31^low THY1.2⁺) using a streamlined sorting panel on the Cytek® Aurora CS sorter, validating the utility of the phenotyping data for downstream isolation.

4. Key Contributions

1. Protocol for Non-Immune Solid Tissues: Provides the first robust, optimized FSFC protocol specifically designed for endothelial cells in solid tissues, addressing the unique challenges of tissue-derived AF.
2. Multi-AF Extraction Strategy: Establishes a critical workflow using "Multi-AF extraction" controls to handle the dynamic and variable autofluorescence found in disease models (fibrosis/steatosis), a major bottleneck in solid tissue flow cytometry.
3. Viability Dye Selection Logic: Highlights the necessity of model-specific viability dye selection, demonstrating that standard "low AF" NIR dyes may fail if the disease model itself induces AF in that spectral region.
4. Integration of Trajectory Analysis: Successfully applies trajectory inference (Wishbone) to flow cytometry data to visualize endothelial plasticity and EndMT in real-time.
5. Rare Cell Isolation: Proves that high-dimensional FSFC data can be translated into effective sorting strategies to isolate rare, disease-relevant EC subpopulations for further study.

5. Significance

This work bridges a critical gap between transcriptomics and proteomics in vascular biology. By overcoming the autofluorescence barrier, the authors enable:

• High-Resolution Phenotyping: Moving beyond scRNAseq to validate protein-level expression and surface interactions in ECs.
• Therapeutic Target Discovery: The identification of specific, rare EC subpopulations (like the THY1.2⁺ pro-fibrotic cells) offers new potential therapeutic targets for liver and heart fibrosis.
• Standardization: The provided protocol and control strategies serve as a framework for applying FSFC to other non-immune cell types in complex solid tissues, accelerating research in inflammation, fibrosis, and cancer.

Limitations Noted: The study acknowledges that FSFC is constrained by the pre-selected antibody panel (unlike the unbiased discovery of scRNAseq) and that the requirement for model-specific AF controls increases the sample size needed for robust statistical power.