Autofluorescence lifetime imaging resolves cell heterogeneity within peripheral blood mononuclear cells

This study demonstrates that label-free optical metabolic imaging (OMI) using autofluorescence lifetime imaging can non-destructively resolve metabolic heterogeneity to accurately identify immune cell subtypes and activation states within heterogeneous peripheral blood mononuclear cell (PBMC) cultures, offering a promising tool for improved disease diagnostics and cell therapy development.

The Gist

The Big Picture: Taking a "Metabolic Snapshot" of Blood Cells

Imagine your blood is a bustling city filled with different types of workers: Monocytes (the security guards), T-Cells (the specialized detectives), B-Cells (the antibody factories), and NK Cells (the rapid-response strike teams).

Usually, when doctors want to check who is in the city and what they are doing, they have to stop the workers, put bright neon vests on them (staining with fluorescent dyes), and then scan them. This is like taking a photo of the city, but the process of putting on the vests might change how the workers behave, and it takes a long time. Also, this method mostly tells you what the workers look like, not how hard they are working.

This paper introduces a new, high-tech camera called "Optical Metabolic Imaging" (OMI). Instead of putting neon vests on the workers, this camera takes a "metabolic snapshot." It looks at the natural, internal glow of the cells to see how much energy they are burning. It's like being able to tell if a car is idling or racing by listening to the engine sound, without ever opening the hood or touching the car.

The Problem: We Need to Know Who Is Who, and Who Is "Waking Up"

When the body is sick or fighting an infection, these blood cells get "activated." They wake up, get angry, and start working overtime.

• The Challenge: In a mixed crowd of blood cells, it's hard to tell who is who just by looking at them.
• The Goal: The researchers wanted to see if they could use this new "engine sound" camera to identify the different types of cells and tell if they were sleeping (quiescent) or working hard (activated) without touching them or using dyes.

The Experiment: The "Wake-Up Call"

The researchers took blood from three healthy people and set up two groups:

1. The Nap Group: Cells left alone to rest.
2. The Wake-Up Group: Cells given a strong "wake-up call" (a chemical mix called PMA and ionomycin) to simulate an immune attack.

They waited just 2 hours—a very short time—and then used their special camera to take pictures of the cells. To make sure their camera was right, they also used a tiny bit of standard dye (the "neon vest") on a separate batch to act as a "ground truth" check.

The Results: The Camera Got It Right!

The results were impressive. The camera could "hear" the metabolic differences between the cells:

• Identifying the Security Guards (Monocytes): The camera was incredibly good at spotting Monocytes. It could tell them apart from the other cells with 96% accuracy when they were resting and 88% accuracy when they were working.
  • Analogy: It's like being able to spot a police car in a traffic jam just by the specific hum of its engine, even if you can't see the lights.
• Spotting the Strike Teams (NK Cells): It could also identify Natural Killer (NK) cells with about 74% accuracy.
• Detecting the "Wake-Up": Most importantly, the camera could tell the difference between the "Nap Group" and the "Wake-Up Group" with 93% accuracy after only 2 hours.
  • Why this matters: Traditional methods often take hours or days to show that cells are activated because they have to wait for new proteins to grow on the surface. This camera sees the energy change immediately.

The "Secret Sauce": How It Works

The camera looks at two tiny molecules inside every cell that act like fuel gauges: NAD(P)H and FAD.

• These molecules glow naturally when hit with a specific laser.
• The speed at which they stop glowing (called "lifetime") changes depending on how the cell is using energy.
• Resting cells are like a car idling at a stoplight (using a specific type of fuel).
• Activated cells are like a car speeding down the highway (switching to a different, faster fuel).

The camera measures these tiny changes in the "glow speed" to create a unique metabolic fingerprint for each cell.

Why This Changes Everything

1. No Touching, No Messing Up: Because it doesn't need to dye the cells, the cells stay alive and healthy. You can take a picture, wait an hour, and take another picture of the same cells to watch them change in real-time.
2. Super Fast: It detects activation in 2 hours, whereas traditional methods might wait much longer for the cells to show signs of work.
3. Better Medicine: This could help doctors diagnose diseases like sepsis or autoimmune disorders faster. It could also help scientists make better "cell therapies" (like CAR-T cancer treatments) by ensuring the cells they are giving to patients are healthy and ready to fight.

The Bottom Line

Think of this technology as a metabolic X-ray. Instead of just seeing the shape of the cells, it sees their "energy signature." The researchers proved that this method can instantly sort through a messy crowd of blood cells, identify the specific types, and tell you exactly who is sleeping and who is ready to fight, all without ever touching a single cell. It's a faster, cleaner, and smarter way to understand our immune system.

Technical Summary

1. Problem Statement

Peripheral Blood Mononuclear Cells (PBMCs) are a heterogeneous mix of immune cells (T cells, B cells, NK cells, and monocytes) critical for disease diagnostics, immune monitoring, and cell therapy development (e.g., CAR-T).

• Limitations of Current Methods: Standard characterization relies on flow cytometry using fluorescently labeled surface markers. These methods are often destructive, lack metabolic information, and require lengthy staining procedures that can alter cell features.
• The Gap: There is a need for non-destructive, label-free tools that can measure single-cell metabolism in real-time to characterize disease states and cell therapy potential without altering the sample. While metabolic heterogeneity is known to exist, it is difficult to resolve in mixed cultures using traditional surface markers alone.

2. Methodology

The study utilized Optical Metabolic Imaging (OMI), a label-free technique based on Fluorescence Lifetime Imaging Microscopy (FLIM), to image endogenous metabolic co-factors NAD(P)H and FAD.

• Sample Preparation:
  • PBMCs were isolated from three healthy human donors via density gradient centrifugation.
  • Cells were plated in two conditions: Quiescent (unstimulated) and Activated (stimulated with 40 nM PMA and 1 μM ionomycin).
  • To establish ground truth for cell subtypes, specific fluorescent antibodies (anti-CD4, CD8, CD14, CD19, CD56) were added to identify specific cell populations within the mixed culture.
• Imaging System:
  • A custom two-photon microscope was used with a 40x water-immersion objective.
  • Excitation: 750 nm for NAD(P)H and antibody detection; 890 nm for FAD.
  • Detection: Time-correlated single-photon counting (TCSPC) was used to capture fluorescence decays.
• Data Analysis:
  • Lifetime Fitting: Pixel-wise decay curves were fitted to a biexponential model to extract parameters: mean lifetime ($\tau_m$), short/long lifetimes ($\tau_1, \tau_2$), and fractional amplitudes ($\alpha_1, \alpha_2$).
  • Optical Redox Ratio (ORR): Calculated as $I_{NAD(P)H} / (I_{NAD(P)H} + I_{FAD})$.
  • Classification: Random Forest classifiers were trained on 10 OMI variables (including lifetimes, amplitudes, ORR, and cell area) to distinguish cell types and activation states.
  • Validation: Activation was confirmed at 24 hours via lactate assays and CD69 surface marker expression.

3. Key Contributions

• First Demonstration in Mixed Cultures: This is the first study to show that OMI can classify both activation states and immune cell subpopulations within complex, heterogeneous PBMC samples from primary human donors, rather than just in monocultures.
• Label-Free Metabolic Phenotyping: The study proves that intrinsic metabolic signatures (NAD(P)H/FAD lifetimes) can serve as a proxy for cell identity and functional state without the need for exogenous labels.
• Early Detection of Activation: The method successfully detected metabolic shifts associated with activation as early as 2 hours post-stimulation, significantly faster than traditional surface markers (CD69) which take longer to express.

4. Key Results

• Cell Subtype Classification (Quiescent State):
  • Monocytes: Identified with 96% accuracy (AUC = 0.98). They were distinguished by low NAD(P)H mean lifetime ($\tau_m$), high NAD(P)H bound fraction ($\alpha_1$), and larger cell area.
  • NK Cells: Identified with 74% accuracy (AUC = 0.91), primarily driven by ORR and FAD lifetime variables.
  • T/B Cells: Helper T, Cytotoxic T, and B cells were not strongly resolved from one another in the quiescent state using OMI alone.
• Cell Subtype Classification (Activated State):
  • Monocytes: Maintained high classification accuracy (88%, AUC = 0.94), though slightly lower than in the quiescent state due to increased metabolic homogeneity across cell types upon activation.
  • NK Cells: Retained moderate accuracy (74%, AUC = 0.91).
• Activation State Classification:
  • OMI distinguished between quiescent and activated PBMCs with 93% accuracy (AUC = 0.98) just 2 hours after stimulation.
  • The most predictive variables for activation were the Optical Redox Ratio (ORR) and NAD(P)H $\tau_1$.
• Metabolic Heterogeneity:
  • Activation increased metabolic heterogeneity (coefficient of variation) in most variables, particularly FAD lifetimes and cell area.
  • UMAP Analysis: Revealed distinct clustering of quiescent vs. activated cells. Notably, quiescent monocytes clustered closer to activated lymphocytes than to quiescent lymphocytes, reflecting their inherently higher metabolic activity.
  • CD56 Bright vs. Dim NK Cells: OMI successfully identified metabolic differences between CD56 bright (high glycolysis) and dim NK cells, showing distinct lifetime signatures.

5. Significance and Impact

• Clinical Diagnostics: OMI offers a rapid, non-destructive tool to monitor immune activation in diseases characterized by immune dysregulation (e.g., sepsis, SLE, rheumatoid arthritis) without altering the sample.
• Cell Therapy Manufacturing: The technique can be integrated into cell therapy workflows (e.g., CAR-T production) to screen starting materials and final products for metabolic fitness and heterogeneity before clinical use, potentially improving therapeutic efficacy.
• Workflow Efficiency: By eliminating the need for staining and allowing repeated imaging of the same cells, OMI enables longitudinal studies and high-throughput metabolic screening that is currently impossible with destructive methods like flow cytometry or RNA sequencing.
• Future Translation: The authors note that NAD(P)H lifetime measurements can be adapted to flow geometry, paving the way for clinical and biomanufacturing integration.