2024
Research News
Aluminum Microwell Substrates Enhance Single-cell Cancer Detection with Raman Spectroscopy
|
Print
Researchers from the Changchun Institute of Optics, Fine Mechanics and Physics of the Chinese Academy of Sciences and the University of Glasgow developed a platform for single-cell analysis. They introduced a microwell-assembled aluminum substrate that significantly advances Raman spectroscopy for cancer cell profiling. The study was published in Talanta.
Single-cell analysis is crucial for understanding cellular heterogeneity, which influences disease outcomes and therapy responses. Conventional methods like flow cytometry and single-cell RNA sequencing offer are often invasive or require complex preparations.
Raman spectroscopy, a label-free and non-invasive tool, provides a detailed molecular snapshot but struggles with low signal-to-noise ratios (SNR) and sample instability in liquid environments.
In this study, researchers designed an innovative platform using microwell-assembled aluminum substrates, which isolates individual cells in suspension, ensuring stability and enhancing Raman signal detection.
The platform consists of over 120,000 hexagonal microwells etched into an aluminum-coated glass substrate. Each microwell’s dimensions are optimized to capture single cells, minimizing aggregation and motion. Aluminum’s high reflectivity amplifies Raman signals while reducing background interference, enabling precise molecular analysis in liquid suspension.
Researchers analyzed lung cancer cells (PC-9) and normal bronchial epithelial cells (BEAS-2B) using the platform. The aluminum substrates demonstrated a 70% improvement in SNR compared with traditional glass substrates, ensuring more reliable data with shorter collection times.
Besides, researchers identified distinct biochemical differences between cancerous and normal cells using this platform. Cancer cells exhibited higher levels of nucleic acids, cytochromes, and unsaturated lipids, biomarkers of altered metabolism and rapid proliferation.
These findings were validated with machine learning models. Notably, the eXtreme Gradient Boosting (XGBoost) algorithm achieved a perfect classification accuracy of 100%, highlighting the robustness of the spectral data.
Contact
