1：Experimental Design and Method Selection:

The study used micro-Raman spectroscopy to analyze undifferentiated and differentiated hPDLSCs. The methodology included isolating cells, inducing differentiation, and performing Raman measurements with statistical analysis via principal component analysis (PCA).

2：Sample Selection and Data Sources:

hPDLSCs were isolated from periodontal ligament tissues of healthy patients aged 18-25 years. Cells from third to sixth passages were used, with differentiation induced into adipogenic, chondrogenic, and osteogenic lineages using specific media.

3：List of Experimental Equipment and Materials:

Equipment included a TriVista 557 Raman system, Coherent VerdiG laser (532 nm), microscope objective (×100, NA=0.80), glass coverslips, 24-well plates, DMEM medium, FBS, penicillin/streptomycin, differentiation inducers (e.g., ascorbic acid-2-phosphate, β-glycerophosphate, TGF-β1, dexamethasone, IBMX, insulin), and staining dyes (Oil Red O, Safranin O, Alizarin red). Materials were from suppliers like PAA Laboratories, Sigma-Aldrich, R&D Systems, and Merck Chemicals.

4：80), glass coverslips, 24-well plates, DMEM medium, FBS, penicillin/streptomycin, differentiation inducers (e.g., ascorbic acid-2-phosphate, β-glycerophosphate, TGF-β1, dexamethasone, IBMX, insulin), and staining dyes (Oil Red O, Safranin O, Alizarin red). Materials were from suppliers like PAA Laboratories, Sigma-Aldrich, R&D Systems, and Merck Chemicals. Experimental Procedures and Operational Workflow:

4. Experimental Procedures and Operational Workflow: Cells were seeded on glass coverslips, grown, and induced to differentiate. After fixation with methanol, Raman spectra were measured at three random positions per cell using a laser power of ≈1 mW and acquisition time of 900 s. Spectra were processed by subtracting substrate contributions and normalizing to the 1660 cm?1 peak. PCA was applied for statistical analysis.

5：Data Analysis Methods:

Raman spectra were analyzed for changes in biomolecular bands. PCA was used to reduce dimensionality and group spectra based on similarities, with data normalized by mean and standard deviation.