Raman Spectroscopic Analyses of Jaw Periosteal Cell Mineralization

Abstract

To achieve safer patient treatments, serum-free cell culture conditions have to be established for cell therapies. In previous studies, we demonstrated that serum-free culture favored the proliferation of MSCA-1⁺ osteoprogenitors derived from the jaw periosteum. In this study, the in vitro formation of bone-specific matrix by MSCA-1⁺ jaw periosteal cells (JPCs, 3 donors) was assessed and compared under serum-free and serum-containing media conditions using the marker-free Raman spectroscopy. Based on a standard fluorescence assay, JPCs from one patient were not able to mineralize under serum-containing culture conditions, whereas the other cells showed similar mineralization levels under both conditions. Raman spectra from mineralizing MSCA-1⁺ JPCs revealed higher levels of hydroxyapatite formation and higher mineral to matrix ratios under serum-free culture conditions. Higher carbonate to phosphate ratios and higher crystallinity in JPCs cultured under serum-containing conditions indicated immature bone formation. Due to reduced collagen production under serum-free conditions, we obtained significant differences in collagen maturity and proline to hydroxyproline ratios compared to serum-free conditions. We conclude that Raman spectroscopy is a useful tool for the assessment and noninvasive monitoring of in vitro mineralization of osteoprogenitor cells. Further studies should extend this knowledge and improve JPC mineralization by optimizing culture conditions.

Figure 1

JPCs were seeded into uncoated/coated glass bottom dishes under untreated and osteogenic conditions. MSCA-1⁺ JPCs were seeded into uncoated/coated glass bottom dishes under untreated and osteogenic conditions. Plates without attachment were unproblematic for untreated JPCs. On the basis of increased detachment of osteogenically induced monolayers, glass bottom dishes coated with MesenCult-XF attachment substrate were used for all Raman measurements.

Figure 2

Mean Raman spectra collected from MSCA-1⁺ JPCs after 20 days of osteogenic differentiation in DMEM and MC medium. (a) Spectra from patient #1: averages were generated from 79 (MC) and 48 (DMEM) single spectra; (b) 128 (MC) and 37 (DMEM) single spectra were averaged for mean spectra from patient #2; (c) mean spectra from patient #3 were generated from 46 (MC) and 100 (DMEM) single spectra. Bone-specific peaks were assigned based on Mandair and Morris (2015) [13].

Figure 3

Fluorescence staining of hydroxyapatite and PCA of Raman spectra from differentiated MSCA-1⁺ JPCs. (a, b, d, e, g, h) OsteoImage fluorescence staining and (c, f, i) PCA scores plots of JPCs isolated from (a–c) patient #1, (d–f) patient #2, and (g–i) patient #3 after differentiation in (a, d, g) serum-containing DMEM and under (b, e, h) serum-free conditions (MC) for 20 days. Scale bar equals 200 µm. (c, f, i) PCA plots of Raman spectra using PCs shown in Figure 4. The PCAs reflect a more mature bone formation (green ellipse) under MC media conditions compared to DMEM conditions (yellow ellipse) after 20 days of differentiation for all patients.

Figure 4

Loadings of PCA describe the mineralization for each patient. (a) PC 1, PC 2, and PC 3 loadings demonstrate strong influences of hydroxyapatite peaks (v₂PO₄³⁻, v₁PO₄³⁻) at 961 and 430 cm⁻¹ in the PCA of Raman spectra from patient #1 (Figure 3(a)). (b) PC 1, PC 2, and PC 4 loadings referring to the PCA of patient #2 (Figure 3(b)) exhibit the four major hydroxyapatite vibrations. (c) Loadings of PC 1, PC 2, and PC 3 validate that the Raman spectra of patient #3 exhibit signals from hydroxyapatite.

Figure 5

Spectral ratios to assess the mineral compositions of DMEM- and MC-cultured MSCA-1⁺ JPCs in osteogenic differentiation. Mineral to matrix ratio as indicated by (a) ratio of hydroxyapatite (HA) to phenylalanine signal and (b) ratio of HA to amide III. Both ratios are significantly increased under MC conditions after 20 days of differentiation. (c) Carbonate to HA (1070/961 cm⁻¹) ratios were significantly higher in JPCs differentiated under DMEM conditions. (d) The crystallinity is indicated by the inverse of full width at half maximum (FWHM) of the HA peak and was significantly decreased under serum-free MC culture conditions compared to DMEM conditions indicating more mature differentiation. Asterisks indicate statistical significances: ^∗∗∗p < 0.001.

Figure 6

Spectral ratios to assess the maturity degree of collagen formed at mineralized sites in DMEM and serum-free cultured MSCA-1⁺ JPCs. (a) Proline to hydroxyproline ratio (853/872 cm⁻¹) indicating significant differences in maturation of collagen fibrils due to media conditions after 20 days of differentiation (statistical significance is indicated by asterisk: ^∗∗∗p < 0.001). (b) Collagen cross-linking calculated from the 1658/1682 cm⁻¹ ratio (n.s., not statistically significant).

Figure 7

Quantitative PCR measurements of type I collagen (α1 and α2 chain) gene expression levels in MSCA-1⁺ JPCs cultured under DMEM (serum-containing) and MC conditions (serum-free). (a) mRNA levels of type I collagen (α1 chain) and (b) (α2 chain) in relation to the housekeeping gene GAPDH are illustrated in untreated (co) and osteogenically induced (ob, for 10 days) MSCA-1⁺ cells. Statistical significances are indicated by asterisks: ^∗p < 0.05.