Characterization of nanoparticles and fluorescent recombinant extracellular vesicles with three high-sensitivity flow cytometers

This study comparatively evaluates the performance of three high-sensitivity flow cytometers (NanoFCM, BD Influx, and CytoFLEX LX) in characterizing silica nanoparticles and fluorescent recombinant extracellular vesicles to define the strengths and limitations of each platform for analyzing particles of varying sizes and concentrations.

The Gist

Imagine you are trying to count and identify tiny, invisible bubbles floating in a glass of water. Some bubbles are huge, some are tiny, and some are so small they are almost invisible. Now, imagine you have three different pairs of "super-eyes" (microscopes) to look at these bubbles. This paper is essentially a report card comparing how well three different high-tech flow cytometers (the "super-eyes") can spot these tiny bubbles, which in science are called nanoparticles and extracellular vesicles (tiny packages released by our cells).

Here is the breakdown of the study using simple analogies:

1. The Players: Three Different "Super-Eyes"

The researchers tested three different machines, each representing a different generation of technology:

• The Specialist (NanoFCM): Think of this as a specialized, high-powered microscope built specifically for looking at tiny dust motes. It's very sensitive but has a very narrow field of view.
• The Veteran (BD Influx): This is an older, classic machine that has been modified over the years. It's like a rugged, old-school camera that has been upgraded with new lenses. It's versatile but can be a bit "noisy."
• The Modern All-Rounder (CytoFLEX LX): This is a newer, commercial machine designed to do everything. It's like a modern smartphone camera: great for taking pictures of people, but maybe a bit confused when trying to photograph a single grain of sand.

2. The Test Subjects: The "Beads" and the "Bubbles"

To test these machines, the scientists used two types of targets:

• Silica Beads (The "Dust Motes"): These are tiny, non-glowing glass beads of different sizes (68nm to 155nm). They are like trying to spot different sizes of dust in a sunbeam. Because they don't glow, the machines have to rely entirely on how much light they bounce back (scatter).
• Recombinant Vesicles (The "Glowing Fireflies"): These are biological bubbles made by cells, but they have been engineered to glow green (like fireflies). This makes them easier to spot because you can look for the light they emit, not just the light they bounce.

3. The Challenges: Noise and Crowds

The study found that using these machines isn't as simple as just turning them on. Two big problems arose:

• The "Static" Problem (Background Noise):
  Imagine trying to hear a whisper in a quiet library versus a rock concert.

  • The Specialist (NanoFCM) was the best at hearing the whisper (seeing the smallest 68nm beads) because its "library" was very quiet.
  • The Modern All-Rounder (CytoFLEX) was like a rock concert; the background noise was so loud that it drowned out the smallest beads. It could only see the bigger ones.
  • The Veteran (BD Influx) was somewhere in the middle.

• The "Crowded Room" Problem (Swarm Detection):
  Imagine trying to count people in a room. If there are only 10 people, you can count them easily. But if you pack 1,000 people into a tiny hallway, they bump into each other, and you can't tell where one person ends and another begins.

  • When the scientists put a high concentration of particles into the Veteran and All-Rounder, the particles clumped together in the machine's "hallway." The machines thought one big clump was a single giant particle, leading to wrong counts.
  • The Specialist handled the crowd better because it moves samples very slowly, like a single-file line, preventing the clumping.

4. The Winning Strategy: The "Glowing Firefly" Trick

The most important discovery was about how you look for the particles.

• Looking for "Bounces" (Light Scatter): When the machines tried to find the particles just by how they bounced light, the results were messy and inconsistent. The machines disagreed on how many particles were there.
• Looking for "Glow" (Fluorescence): When the scientists told the machines to ignore everything that didn't glow green, the results changed dramatically.
  • It was like turning off the lights in the room and only looking for fireflies. Suddenly, the background noise disappeared, and all three machines agreed on exactly how many fireflies were in the room.

The Takeaway

This paper teaches us that there is no "one-size-fits-all" machine for studying tiny biological particles.

1. If you are looking for the tiniest, invisible specks: You need the most sensitive, specialized machine (like the NanoFCM), but you have to be careful not to make the sample too dilute, or the machine gets confused by background noise.
2. If you are counting biological particles: Don't just rely on how they look (size/shape). If you can make them glow (use fluorescence), you get much more accurate counts, regardless of which machine you use.
3. Dilution is key: You can't just dump a sample in. You have to find the "Goldilocks" concentration—not too crowded, not too empty—for each specific machine to get a true count.

In short, to see the invisible world of cell bubbles, you need the right tool, the right lighting (fluorescence), and the right amount of space to count them properly.

Technical Summary

1. Problem Statement

High-sensitivity flow cytometry (FC) is increasingly used for the single-particle analysis of nanoparticles (NPs) and extracellular vesicles (EVs). However, significant challenges remain in reliably detecting and quantifying these submicron particles due to:

• Instrument Heterogeneity: The field utilizes instruments from different generations (optimized first-gen, commercial second-gen, and dedicated third-gen), each with distinct optical configurations and detection limits.
• Detection Thresholds: EVs scatter very little light, often operating near the instrument's noise level. Traditional Forward Scatter (FSC) thresholds are often insufficient.
• Reference Material Limitations: Polystyrene (PS) beads are commonly used for calibration but have a higher refractive index (RI) than biological EVs, leading to overestimation of detection capabilities. Silica nanoparticles (SiNPs) with lower RI are better mimics but require rigorous cross-platform validation.
• Concentration Artifacts: High sample concentrations cause "swarm detection" (coincidence), while low concentrations increase the impact of background noise, making robust quantification difficult across different platforms.

There is a critical need for comparative studies using identical samples and reference materials to define the strengths, limitations, and optimal operating conditions for different high-sensitivity FC platforms.

2. Methodology

The study employed a controlled, cross-platform comparative approach using identical samples measured simultaneously by dedicated operators.

• Instruments Evaluated:

  1. NanoFCM (NF): Third-generation, lab-built design (commercialized), dedicated for nanoscale analysis. Uses a 532 nm laser for Side Scatter (SSC) and Single Photon Counting Modules (SPCM). Very low flow rate (0.005 µL/min).
  2. BD Influx (IF): First-generation, jet-in-air, optimized for small particles. Uses a 488 nm laser with reduced-wide angle FSC (rw-FSC) and PMT detectors. Moderate flow rate (6 µL/min).
  3. CytoFLEX LX (CF): Second-generation, cuvette-based. Uses a 405 nm laser for Violet Side Scatter (VSSC) and APD detectors. Higher flow rate (10 µL/min).

• Reference Materials:

  • Silica Nanoparticles (SiNPs): A non-fluorescent cocktail of 68, 91, 114, and 155 nm particles (low RI, mimicking EVs).
  • Polystyrene Nanoparticles (PSNPs): A cocktail of 64, 94, 127, and 156 nm particles (high RI).
  • Fluorescent Recombinant EVs (rEVs): Commercially available EVs expressing EGFP (approx. 125 nm), serving as a biological reference material.

• Experimental Design:

  • Thresholding Strategies: Evaluated SSC (488 nm), FSC (488 nm), VSSC (405 nm), and Fluorescence (FL) thresholds.
  • Concentration Series: Samples were measured across a wide range of dilutions (from $10^6$ to $10^9$ particles/mL) to determine the dynamic range and identify swarm detection vs. background noise limits.
  • Data Analysis: Events were gated and quantified using FlowJo, with concentrations corrected for flow rate, dilution factor, and background subtraction.

3. Key Contributions

• Cross-Platform Benchmarking: Provided the first direct comparison of three distinct generations of high-sensitivity FC instruments using identical SiNP and rEV samples.
• Validation of SiNPs: Demonstrated the utility of mixed SiNP populations (low RI) over PSNPs for determining the true Limit of Detection (LOD) for biological EVs.
• Optimization of Detection Strategies: Identified that a "one-size-fits-all" thresholding strategy does not exist; optimal triggering depends on the specific instrument architecture (e.g., VSSC for CF, FSC for IF, SSC for NF).
• Quantitative Guidelines: Established that robust quantification requires specific sample dilutions tailored to each instrument's flow rate and background noise profile to avoid swarm effects or background overestimation.

4. Key Results

A. Sensitivity to Non-Fluorescent Nanoparticles (SiNPs)

• NanoFCM (NF): Detected all four SiNP populations (down to 68 nm) using SSC (488 nm) thresholding. However, at low concentrations, the 68 nm and 91 nm populations overlapped with background noise.
• BD Influx (IF): Using SSC (488 nm), it failed to detect 68 nm particles. Switching to rw-FSC (488 nm) improved background rejection, allowing detection of 91 nm particles, but 68 nm remained undetected.
• CytoFLEX LX (CF): SSC (488 nm) failed to resolve any populations. Switching to VSSC (405 nm) allowed detection of three populations (91 nm, 114 nm, 155 nm), though 91 nm was not fully resolved from background.
• Conclusion: NF had the highest sensitivity for the smallest particles, but all platforms struggled with the 68 nm SiNPs at low concentrations due to background noise.

B. Impact of Sample Concentration

• NanoFCM: Required higher sample concentrations ($10^9$ particles/mL) to overcome background noise. It successfully avoided swarm detection at these levels due to its ultra-low flow rate.
• BD Influx & CytoFLEX LX: Required significantly lower concentrations ($10^8$ particles/mL) to prevent swarm detection (coincidence), which caused non-linear event counts and aberrant scattering signals at higher concentrations.
• Finding: Low concentrations challenge detection on NF (noise dominance), while high concentrations compromise single-particle analysis on IF and CF (swarm dominance).

C. Analysis of Fluorescent rEVs

• Qualitative Analysis:
  • NF: Fluorescent rEVs were clearly separated from background using SSC thresholding + fluorescence gating.
  • IF & CF: SSC thresholding alone resulted in high background noise. Fluorescence thresholding (FLt) was essential to reduce background events by 785-fold (IF) and 111-fold (CF), enabling clear visualization of the rEV population.
• Quantitative Analysis:
  • NF: SSC-only gating overestimated concentrations in dilute samples due to background inclusion. Adding fluorescence gating (SSCt + FLg) reduced variability from ±27% to ±4%, yielding ~$7 \times 10^{11}$ events/mL.
  • IF: Showed high consistency across strategies (SSCt, SSCt+FLg, FLt) when using optimal dilutions, yielding ~$3 \times 10^{12}$ events/mL.
  • CF: Showed the highest discrepancy. SSC-only gating led to underestimation due to swarm effects in concentrated samples. Fluorescence thresholding (FLt) provided the most reliable results, aligning best with IF and NF data (~$7.5 \times 10^{11}$ events/mL).

5. Significance and Conclusion

This study highlights that while high-sensitivity flow cytometry is a powerful tool for EV analysis, instrument choice and operational parameters are critical:

1. No Universal Strategy: Researchers must select detection thresholds (SSC, FSC, VSSC, or FL) based on the specific instrument's optical design.
2. Importance of Fluorescence: For biological EVs, fluorescence-based detection (either as a trigger or a gate) is superior to light-scatter-only methods for reducing background noise and ensuring accurate quantification.
3. Dilution is Key: There is no single "correct" concentration. Samples must be titrated to find the "sweet spot" where background noise is minimized without inducing swarm detection, which varies by instrument flow rate.
4. Reference Materials: The study advocates for the use of fluorescent rEVs and low-RI SiNPs as standard reference materials to validate instrument performance, moving beyond polystyrene beads which do not accurately mimic biological EV scattering properties.

The findings provide a roadmap for standardizing EV analysis across different generations of flow cytometers, facilitating more reliable data comparison in clinical and research settings.