Single-cell metabolic analysis by stimulated Raman scattering cytometry

Abstract

Understanding cellular heterogeneity has been a challenge in biology. Current bio-analytical methods such as mass spectrometry or fluorescence-based detection would destruct the sample or perturb the functions of targeted molecules. In situ imaging of bio-molecules at single cell level resolves the phenotypes at metabolomics domain, which can address the challenges in studying cellular heterogeneity. Stimulated Raman scattering (SRS) microscopy provides a label-free approach to identify molecules based on the signature of molecular vibrations. However, there are several challenges to overcome in order to use SRS as a single-cell analysis platform with high throughput, high content, and high sensitivity. My thesis work aims to overcome above-mentioned difficulties. To fulfill the first unmet need of single cell metabolic analysis, we developed an SRS flow cytometry, and demonstrated the discrimination of particles at a throughput of up to 11,000 particles per second, which is a four orders of magnitude improvement compared to conventional spontaneous Raman flow cytometry. Next, we addressed the second unmet need of single-cell metabolic analysis through the development of SRS imaging cytometry. Using this platform, we studied the response of human pancreatic cancer to drug-induced and starvation-induced stress, and discovered lipid-facilitated protrusion as a metabolic biomarker for stress-resistant cancer cells. Lastly, to probe low-concentration bio-molecules using fingerprint Raman bands, we utilized pre-resonance enhancement to increase the SRS signal by two orders of magnitude, and demonstrated ultra-sensitive imaging of retinoids in cells. We demonstrated in situ imaging of retinoid level in cancer cells and during neuronal development. Collectively, these efforts demonstrate SRS cytometry as a high-throughput, high-content, and high-sensitivity single-cell analysis platform with broad applications.