Non-invasive detection of regulatory T cells with Raman spectroscopy

Area of Science:

• Immunology
• Spectroscopy
• Machine Learning

Background:

• Regulatory T cells (Tregs) are crucial for immune self-tolerance and hold therapeutic potential.
• Current Treg identification relies on invasive intracellular markers or less specific surface markers, compromising cell viability and purity.
• Existing methods face challenges in accurately isolating live, unaltered Tregs for research and clinical use.

Purpose of the Study:

• To develop and validate a non-invasive method for identifying live, unaltered regulatory T cells (Tregs).
• To overcome the limitations of invasive markers and surrogate surface markers in Treg purification and analysis.
• To establish a reliable method for Treg detection applicable to both murine and human cells.

Main Methods:

• Utilized Raman spectroscopy for optical detection of live, unaltered Tregs.
• Integrated machine learning, specifically regularized logistic regression, for data analysis and Treg identification.
• Validated the method on murine cells with a Foxp3 reporter and subsequently on human peripheral blood T cells.
• Incorporated strategies to manage sample purity and ensure model robustness.

Main Results:

• Achieved an accuracy higher than 80% in identifying Tregs, comparable to standard sorting purities.
• Demonstrated the ability to reliably detect Tregs in independent donors not included in the model training set.
• Successfully applied the non-invasive method to both murine and human T cell samples.

Conclusions:

• Raman spectroscopy coupled with machine learning offers a viable non-invasive approach for identifying live, unaltered Tregs.
• This method surpasses the limitations of traditional Treg identification techniques, enabling purer cell isolation.
• The demonstrated reliability in independent donors represents a significant advancement towards practical therapeutic applications of Tregs.