Supercontinuum intrinsic fluorescence imaging heralds free view of living systems

Geng Wang, Lianhuang Li, Xiaoxia Liao, Shu Wang, Jennifer W. Mitchell, Chanaka Rabel, Shirui Luo, Jindou Shi, Janet E. Sorrells, Rishyashring R. Iyer, Edita Aksamitiene, Carlos A. Renteria, Eric J. Chaney, Derek J. Milner, Matthew B. Wheeler, Martha U. Gillette, Alexander G. Schwing, Jianxin Chen, Haohua TubioRxiv 2024 · DOI

TL;DR

This paper introduces "Supercontinuum Intrinsic Fluorescence Imaging" (SCIFI), a novel label-free microscopy technique that generates a broad spectrum of intrinsic fluorescence from live biological samples. SCIFI overcomes the limitations of traditional methods by minimizing photodamage, enabling high-content phenotyping and real-time monitoring of cellular processes across diverse biological systems, from microbes to plant leaves.

What Problem Does This Paper Address?

Current fluorescence microscopy techniques often struggle to balance the need for detailed molecular information with the imperative to minimize harm to fragile living systems. Traditional methods can induce photodamage, limiting observation time and potentially altering cellular behavior. Furthermore, achieving comprehensive imaging often requires specific fluorescent labels, which can be invasive and may not be available for all biological targets. The authors sought to develop an imaging strategy that could provide rich, multi-dimensional information from unlabeled live samples without causing significant damage, thus enabling a more "free view" of biological processes.

Methods

The core innovation is the generation of "supercontinuum" intrinsic fluorescence from biological samples. This is achieved by carefully controlling ultrafast laser pulses (picosecond-femtosecond) and their spatial separation to excite nonlinear optical processes within the sample itself. Unlike previous methods that often resulted in photodamage, this approach leverages a narrow window of illumination where samples can tolerate higher irradiance before damage occurs.

Key techniques include:

  • Supercontinuum Generation: Utilizing nonlinear optical effects in biological tissues to generate a broad spectrum of fluorescence (340-740 nm) from a single excitation wavelength (1110 nm).
  • Multiphoton Excitation: Employing simultaneous two-, three-, and four-photon absorption of intrinsic fluorophores (like tryptophan, NADH, FAD, and porphyrins) to generate distinct fluorescence signals.
  • Label-Free Imaging: Relying solely on the natural fluorescence of biomolecules, eliminating the need for exogenous labels.
  • Crosstalk Compensation: Developing sophisticated algorithms to correct for spectral overlap between different fluorescence signals, allowing for accurate quantification of individual fluorophores.
  • Quantitative Biomarkers: Deriving dimensionless metrics like Lipofuscin Fluorescence Independence (LFI) and Optical Redox Ratio (ORR) from the compensated fluorescence signals to represent cellular aging and metabolic state, respectively.
  • Kinetic and Endpoint Assays: Utilizing time-lapse imaging and snapshot analysis to monitor cellular responses to drugs and other stresses in real-time.

Datasets

The paper does not explicitly mention the use of pre-existing, labeled datasets for training or evaluation in the traditional machine learning sense. Instead, it presents results from a wide array of in vitro and in vivo biological samples, effectively creating its own benchmarks through the application of the SCIFI technique. These samples include:

  • Rabbit intestinal cells and microbes
  • Bovine embryos
  • Mouse skin flap
  • Developing mouse brain
  • Hamster kidney cells
  • C. elegans worms (young and mature)
  • Rat brain slices
  • Ivy and maple leaves

The paper implicitly introduces a new benchmark for label-free, high-content cellular phenotyping and dynamic monitoring across these diverse biological systems.

Key Insights and Ideas

  • "Free View" is Achievable: The central idea is that by carefully controlling laser parameters, we can achieve a "free view" of living systems, meaning detailed, multi-modal imaging without invasive labels or significant photodamage.
  • Intrinsic Fluorophores as Rich Biomarkers: The broad spectrum of intrinsic fluorescence from common biomolecules (tryptophan, NADH, FAD, porphyrins) can be resolved and quantified to reveal complex cellular states like aging, metabolic activity, and even early signs of drug-induced cytotoxicity.
  • Lipofuscin Fluorescence Independence (LFI) as an Aging Biomarker: LFI, derived from the relative contributions of different intrinsic fluorophores, shows a strong correlation with cellular aging across different species and cell types, offering a label-free alternative to traditional aging markers.
  • Optical Redox Ratio (ORR) for Metabolic Insight: ORR, based on the ratio of FAD to NADH fluorescence, provides a quantitative measure of cellular metabolic state, which can be crucial for understanding cell proliferation and disease.
  • Dynamic Monitoring of Cellular Stress: SCIFI can track subtle, real-time cellular responses to drugs or environmental stress, such as recovery from transient exposure, which might be missed by static endpoint assays.
  • Plant Photosynthesis Under a New Light: The technique reveals previously unobserved dynamics of chloroplasts during photosynthesis, suggesting new avenues for studying plant photobiology.

Key Results

  • High-Content Phenotyping: SCIFI successfully differentiated multiple cell types and states in various tissues (e.g., rabbit intestine, mouse lung, developing brain) based on their unique 6-color intrinsic fluorescence profiles.
  • Embryo Polarization Identification: The technique identified distinct cell lineages in bovine embryos at the 16-cell stage, correlating with trophectoderm and inner cell mass fates, without the need for genetic markers.
  • Aging Biomarker Validation: LFI demonstrated a consistent trend of increasing with age in C. elegans, rat brains, and mouse blood cells, suggesting its utility as a general aging biomarker.
  • Drug Cytotoxicity Sensing: SCIFI could detect drug-induced cytotoxicity in real-time, differentiating between modes of action (e.g., BAM15 vs. STS) based on fluorescence changes.
  • Plant Photobiology Insights: The imaging of ivy and maple leaves revealed dynamic processes in chloroplasts during photosynthesis and provided insights into the origins of autumnal leaf colors.
  • Microbial Imaging: The technique was applied to intestinal microbes, revealing their intrinsic fluorescence and potential for differentiation from non-living particles.

Limitations and Future Work

  • Photodamage Threshold: While significantly reduced, the risk of photodamage still exists at higher laser powers, necessitating careful optimization of illumination conditions. The paper notes that "living samples can tolerate rather high irradiance up to one-half of water optical breakdown before nonlinear photodamage onset."
  • Signal-to-Noise Ratio: The "gentle" imaging approach, while minimizing damage, can lead to a lower signal-to-noise ratio, which the authors address by employing deep learning-based video denoising techniques.
  • Complexity of Compensation: Accurate quantification relies on sophisticated crosstalk compensation, which, while effective, adds a layer of complexity to data processing.
  • Future Work:
    • 3D Disease Models: The authors suggest that realistic 3D disease models, such as organoids, should be used to better approximate in situ cell types and improve drug discovery.
  • Expanding Applications: Further optimization of illumination conditions and the development of more advanced detection modules (e.g., optical fiber-coupled spectroscopic detection) are proposed to expand the applicability of SCIFI.
  • Optogenetics Integration: Harnessing unused portions of the supercontinuum for optogenetic control is mentioned as a potential enhancement.
  • Machine Learning Integration: The paper hints at the potential for further integration with machine learning for tasks like automated cell tracking and classification.

Summary updated

Abstract

Optimal imaging strategies remain underdeveloped to maximize information for fluorescence microscopy while minimizing the harm to fragile living systems. Taking hint from the supercontinuum generation in ultrafast laser physics, we generated supercontinuum fluorescence from untreated unlabeled live samples before nonlinear photodamage onset. Our imaging achieved high-content cell phenotyping and tissue histology, identified bovine embryo polarization, quantified aging-related stress across cell types and species, demystified embryogenesis before and after implantation, sensed drug cytotoxicity in real-time, scanned brain area for targeted patching, optimized machine learning to track small moving organisms, induced two-photon phototropism of leaf chloroplasts under two-photon photosynthesis, unraveled microscopic origin of autumn colors, and interrogated intestinal microbiome. The results enable a facility-type microscope to freely explore vital molecular biology across life sciences.
BibTeX
@misc{wang2024supercontinuum,
  author = {Geng Wang and Lianhuang Li and Xiaoxia Liao and Shu Wang and Jennifer W. Mitchell and Chanaka Rabel and Shirui Luo and Jindou Shi and Janet E. Sorrells and Rishyashring R. Iyer and Edita Aksamitiene and Carlos A. Renteria and Eric J. Chaney and Derek J. Milner and Matthew B. Wheeler and Martha U. Gillette and Alexander G. Schwing and Jianxin Chen and Haohua Tu},
  title = {{Supercontinuum intrinsic fluorescence imaging heralds free view of living systems}},
  year = {2024},
  doi = {10.1101/2024.01.26.577383},
  howpublished = {bioRxiv},
}
Figure