Detection and Characterization of Multidimensional Information of Adipocyte Model Based on AFM-Raman

Abstract

Excessive accumulation of white adipose tissue leads to metabolic disorders, and the excessive differentiation of preadipocytes into white adipocytes is one of the contributing factors to obesity. The browning of white adipocytes has been regarded as a promising therapeutic strategy. To analyze the origins and potential solutions for obesity from a fundamental perspective, we employed atomic force microscopy, and Raman confocal microscopy to investigate and characterize multidimensional information regarding the differentiation process of 3 T3-L1 preadipocyte models into white adipocytes and their subsequent browning into beige adipocytes. The results from atomic force microscopy indicated that during the differentiation of preadipocytes into mature white adipocytes, there was an increase in cell height, a decrease in length, and a transformation in shape from fibroblast-like morphology to spherical form. Additionally, Young's modulus, stiffness, and adhesion decreased throughout this process. Following browning, cells maintained their spherical shape but exhibited reduced height compared to white adipocytes; lipid droplet decomposition resulted in increased surface roughness. Raman spectroscopy studies revealed that preadipocytes lacked specific lipid peaks; however, as they differentiated into white adipocytes, peak Raman signals transitioned from weak to sharp. After browning occurred, lipid peak signals became sparse and dispersed. Furthermore, by calibrating temperature standard curves based on water molecule hydrogen-oxygen stretching bands, it was found that beige adipocytes possess thermogenic capabilities. Based on Segment Anything Model for lipid droplet segmentation and color clustering 3D K-Means point cloud analysis: White adipocyte lipid droplets aggregated with deeper coloration post-staining appearing duller; conversely, beige adipocyte coloration appeared lighter and brighter with more clusters present within the clustering point cloud. In summary, this study provides a novel method for multidimensional detection and characterization through an interdisciplinary approach combining cellular biology with physical chemistry.