TL;DR
This paper introduces the SwissSPAD2, a large-area, time-gated SPAD camera, for near-infrared (NIR) fluorescence lifetime imaging (FLI). The authors demonstrate its capability for both in vitro and in vivo preclinical applications, showing it can accurately measure fluorescence lifetimes and Förster Resonance Energy Transfer (FRET) in complex biological systems, rivaling traditional gated-ICCD cameras.
What Problem Does This Paper Address?
Preclinical molecular imaging is crucial for drug development and understanding disease. Near-infrared (NIR) fluorescence lifetime imaging (FLI) offers a powerful way to monitor biological processes in vivo due to its unique contrast mechanism and ability to probe molecular environments. However, existing technologies for macroscopic FLI (MFLI) in deep tissues, primarily time-gated ICCD cameras, are expensive, prone to degradation, and have limited room for improvement. Furthermore, detecting short-lifetime NIR dyes, essential for in vivo imaging, presents a challenge for current detectors. This paper addresses the need for a more advanced, cost-effective, and high-performance solution for NIR MFLI by evaluating the SwissSPAD2, a novel time-gated SPAD camera.
Methods
The study employs two primary methods for analyzing fluorescence lifetime data:
- Nonlinear Least Square Fit (NLSF): A standard technique that fits the measured decay curve to a model, requiring an instrument response function (IRF) and good signal-to-noise ratio. The paper uses both single-exponential and bi-exponential NLSF to analyze different types of decays.
- Phasor Analysis: A user-friendly, fit-free approach that analyzes the Fourier transform of the decay curve. Phasors are plotted in a complex plane, and their position reveals information about the fluorescence lifetime(s). This method is particularly useful for analyzing complex mixtures and FRET.
The researchers benchmarked the SwissSPAD2 against a state-of-the-art gated-ICCD camera using these methods. They performed in vitro experiments to assess environmental effects on dye lifetimes and quantify FRET between dye mixtures. Subsequently, they applied these techniques to in vivo studies of drug-target engagement in live mice using FRET.
Datasets
The paper does not introduce new benchmark datasets. Instead, it utilizes well-characterized fluorescent dyes and probes for its experiments:
- IRDye 800CW-2DG: Used to study environmental effects (pH, solvent) on fluorescence lifetime.
- Alexa Fluor 750: Used for initial validation of the SPAD camera's performance.
- AF700/AF750 labeled IgG: Used to create FRET pairs for in vitro validation of FRET measurements.
- AF700/AF750 labeled Trastuzumab (TZM) and Cetuximab (CTM): These are clinically relevant antibodies used in in vivo studies to monitor drug-target engagement (TZM with HER2, CTM with EGFR) in tumor xenografts.
Key Insights and Ideas
- SPADs are ready for NIR MFLI: The SwissSPAD2, despite being a newer technology, demonstrates comparable or superior performance to established gated-ICCDs for NIR FLI, even with challenging short-lifetime dyes.
- Phasor analysis is robust: The paper highlights the power and simplicity of phasor analysis, showing its effectiveness in quantifying FRET and analyzing complex decays, even with imperfect IRFs.
- FRET as a molecular ruler: The study effectively uses FRET to quantify drug-target engagement in vivo, providing a direct measure of antibody binding to its receptor. This is a significant advancement for preclinical drug development.
- Distinguishing tumor models: The FRET measurements allowed the researchers to differentiate between tumor models with varying levels of receptor expression and drug binding, even when both models overexpressed the target. This has implications for understanding drug delivery and efficacy.
- Environmental sensitivity of NIR dyes: The in vitro experiments clearly showed how the fluorescence lifetime of NIR dyes can be influenced by their microenvironment, underscoring the importance of lifetime measurements for accurate biological interpretation.
Key Results
- Quantitative agreement: The SwissSPAD2 and the gated-ICCD camera showed excellent agreement in both in vitro and in vivo FRET measurements, with differences typically within a few percent.
- Environmental effects: IRDye 800CW-2DG's lifetime was found to increase significantly in DMSO compared to aqueous buffers, and a slight increase was observed at lower pH.
- FRET quantification: Bi-exponential NLSF and phasor ratio analysis provided consistent FRET fraction values for both detectors across various acceptor-to-donor ratios in in vitro FRET experiments.
- In vivo drug-target engagement:
- TZM-HER2: The study successfully quantified TZM-HER2 engagement in AU565 and SK-OV-3 tumor xenografts, showing significantly higher FRET in AU565 tumors, consistent with biological expectations.
- CTM-EGFR: Similar FRET measurements were performed for CTM-EGFR engagement, again demonstrating the ability to differentiate between tumor models and identify FRET in organs like the liver.
- Sub-nanosecond lifetime resolution: The SwissSPAD2 demonstrated the ability to resolve lifetimes of only a few hundreds of picoseconds and lifetime differences of tens of picoseconds.
Limitations and Future Work
- IRF influence: The paper notes a small, consistent bias between NLSF and phasor lifetimes, attributed to potential inadequacies in the IRF data used for convolution and calibration. This bias was corrected for post-analysis.
- In vivo IRF approximation: For in vivo experiments, the IRF was acquired using the mouse itself. The authors acknowledge that this is an approximation and may not fully capture the complex scattering and propagation effects of light deep within tissues. However, they argue it's a better approximation than a simple paper IRF.
- Future improvements for SPADs: The authors highlight several areas for future development in SPAD technology, including micro lenses for increased collection efficiency, shorter gate durations for better resolution, dual-gate architectures, rolling-shutter recording, and higher recording rates for faster imaging. They also suggest leveraging the built-in FPGA for advanced data pre-processing.
- Broader applications: The paper concludes by suggesting that the SwissSPAD2 and similar SPAD technologies are poised to become the technology of choice for a wide range of macroscopic FLI applications, from preclinical studies to optical-guided surgery.
Summary updated