Using FTIR and Raman spectroscopy, we investigated the applicability of these technique to detect changes in the secondary structure of proteins and lipids amount, which were correlated with cortisol concentration in plasma of volleyball players. In our research we studied the differences in the value of absorbance of individual functional groups and the frequency shift of the peaks. Moreover, to determine secondary structure of proteins, we calculated second derivative of obtained FTIR spectra and we made deconvolution of the amide I IR region. First we analyzed all IR region, where the measurements were performed. In the Fig. 1a, characteristic FTIR peaks for phospholipid, protein and lipid vibrations in plasma, were visible. Two peaks at 1079 cm− 1 and 1243 cm− 1 corresponded to symmetric and asymmetric stretching vibrations of PO2− from phospholipids, respectively. Moreover, three characteristic peaks originating from protein vibrations were noticed at 1265 cm− 1, 1525 cm− 1 and 1654 cm− 1 (amide III, amide II, amide I, respectively). Furthermore, in Fig. 1a peaks corresponded to symmetric as well as asymmetric vibrations of CH2 and CH3 groups at 2872 cm− 1, 2920 cm− 1 and 2957 cm− 1, respectively22–28. Moreover, also peaks characteristic for PO2− from phospholipids, amide bonds and CH2 and CH3 groups from lipids, were marked in Raman spectra in Fig. 1b. The Raman shifts at 1032 cm− 1 and 1240 cm− 1 originated from vibrations of phosphorate groups from phospholipids29. Amide III, amide II and amide I vibrations were placed at 1266 cm− 1, 1544 cm− 1 and 1655 cm− 1 30–33. Symmetric as well as asymmetric vibrations of CH2 and CH3 groups were noticed between 2800 cm− 1 and 2950 cm− 1 34.
When we compared FTIR spectra presented in Fig. 1a, we noticed, that along with a decrease of cortisol concentration, the absorbance of symmetric as well as asymmetric vibrations of CH2 and CH3 lipid groups decreased. We also observed increase of absorbance of amide bonds, when the cortisol concentration in plasma decrease. The smallest differences in the absorbance were visible in IR region corresponding to phospholipid vibrations. However, we observed, that the intensity of asymmetric stretching vibrations of PO2− was the highest in plasma of women with 12.7 µg/dL cortisol concentration (red spectrum). Furthermore, Raman spectra demonstrated, that the highest Raman intensity for lipid vibrations was observed for women with the highest cortisol concentration in plasma, (Fig. 1b, green spectrum). Moreover, the smallest Raman intensity for phospholipid vibrations, was visible for this spectrum. Raman spectra also presented differences in the Raman intensity and positions of peaks corresponding to amide vibrations. For women with the smallest cortisol concentration (blue and red spectra), similar Raman intensity for amide II and amide I bonds, were visible and for these vibrations the Raman shift was also similar. Interestingly, obtained spectra revealed, that along with the cortisol concentration, the shape of structures in the IR and Raman regions originating from analyzed vibrations, was changed. Consequently, to uncover more details in phospholipid, protein and lipid vibrations, second derivative of analyzed region (Figs. 2a-2d) was calculated. Moreover, to obtain information about secondary structure of proteins and calculated the percentage of alpha helix and beta sheet, second derivative of amide I bonds (Figs. 2e-2h) and deconvolution of amide I vibrations, was calculated (Figs. 2i-2l). From obtained spectra similarity in the all analyzed regions, in women with 15.2 µg/dL (blue spectrum) and 12.7 µg/dL (red spectrum), was found, (Figs. 2a-2d). In general, bands which were visible in Figs. 2e-2h corresponding to 1611–1630 cm− 1 were attributed as cross-β, at 1630–1644 cm− 1, and 1647–1662 cm− 1 as α-helix, and at 1662–1699 cm− 1 as anti-parallel β-sheet35. In the Figs. 2e-2h, the most visible differences in the structure of protein vibrations between samples with small (15.2 µg/dL, 12.7 µg/dL) and high (19.1 µg/dL, 27.91 µg/dL) cortisol concentration was obtained for IR region between 1665 cm− 1 and 1700 cm− 1, which corresponded to β-sheet secondary structure of protein.
The deconvolution of amide I region (Figs. 2i-2l) clearly presented differences in the number of fitting curves between analysed plasma samples. However, the most important differences were visible in the number of individual secondary structure of proteins. We calculated the area of peaks corresponding to α and β secondary structure, from which we obtained a ratio between them. For women with 27.91 µg/dL cortisol concentration the ratio between α and β was 1.214, for women with cortisol level 19.1 µg/dL – 1.227. In case of 15.2 µg/dL cortisol level, the α and β ratio was 1.261 and for 12.7 µg/dL the hormone stress concentration was 1.471.
To obtain information about differences in the number of phospholipid, protein and lipid functional groups, peak area corresponding to these vibrations was calculated and presented in Fig. 3.
Figure 3 showed differences in the peak area of peaks corresponding to phospholipid, protein and lipid vibrations in FTIR (Fig. 3a) and Raman (Fig. 3b) spectra collected from women with different cortisol concentration. Based on the FTIR data we observed, that along with the decrease of cortisol concentration the amount of lipids decreased, while the amount of proteins was increasing. Moreover, FTIR spectra presented small differences in the level of phospholipids in cortisol concentration function. While in the Raman spectra was noticed, that when cortisol concentration decreased, the area of peaks originating from phospholipids increased. Moreover, Raman spectra presented similar results when we compared cortisol concentration and protein functional groups, while in the case of peaks corresponding to lipid vibrations, we noticed amount increase along with an increase of cortisol concentration.
Table 1
Pearson correlation test between the concentration of cortisol in the plasma and the surface area of peaks originating from phospholipids, proteins and lipids, determined from FTIR and Raman spectra. p < 0,001.
FTIR spectroscopy – area of peaks corresponding to functional groups from:
|
|
phospholipids
|
proteins
|
lipids
|
cortisol concentration
|
phospholipids
|
1.00
|
0.95
|
-0.71
|
0.70
|
proteins
|
0.95
|
1.00
|
-0.89
|
0.10
|
lipids
|
-0.71
|
-0.89
|
1.00
|
0.87
|
cortisol concentration
|
0.70
|
0.10
|
0.87
|
1.00
|
Raman spectroscopy – area of peaks corresponding to functional groups from:
|
|
phospholipids
|
proteins
|
lipids
|
cortisol concentration
|
phospholipids
|
1.00
|
0.48
|
-
|
-0.57
|
proteins
|
0.48
|
1.00
|
-0.80
|
-0.90
|
lipids
|
-
|
-0.80
|
1.00
|
0.76
|
cortisol concentration
|
-0.57
|
-0.90
|
0.76
|
1.00
|
Pearson correlation test, Table 1 presented positive, as well negative correlations between properties of analyzed samples calculated from FTIR and Raman spectra. Positive correlation means, that when amount of measured factors increases, also second factor increases, while negative correlation means, that if we notice an increase of some factor, the amount of next factor decreases. From the FTIR spectra, positive correlations between phospholipids and proteins and cortisol concentration and phospholipids, proteins and lipids, respectively, were identified. Negative correlation between lipids and phospholipids, was visible from FTIR data. Moreover, a strong negative correlation between lipids and proteins was observed in FTIR spectra. In Raman data also positive and negative correlations were obtained. Positive correlation between phospholipids and proteins and between cortisol concentration and lipids, were visible, while for cortisol concentration and proteins, as well as phospholipids, negative correlations were found. Furthermore, lack of a correlation between proteins and lipids was noticed from Raman spectra.