Exploring the relationship between epigenetic DNA methylation and cardiac fibrosis through Raman microspectroscopy

Abstract

Cardiomyopathies are associated with fibrotic remodeling of the heart, which is characterized by the excessive accumulation of collagen type I (COL I) due to chronic inflammation and suspected epigenetic influences. Despite the severity and high mortality rate of cardiac fibrosis, current treatment options are often inadequate, underscoring the importance of gaining a deeper understanding of the disease's underlying molecular and cellular mechanisms. In this study, the extracellular matrix (ECM) and nuclei in fibrotic areas of different cardiomyopathies were molecularly characterized by Raman microspectroscopy and imaging and compared with the control myocardium. Patient samples were obtained from heart tissue affected by ischemia, hypertrophy, and dilated cardiomyopathy and analyzed for fibrosis through conventional histology and marker-independent Raman microspectroscopy (RMS). Prominent differences between control myocardium and cardiomyopathies were revealed by spectral deconvolution of COL I Raman spectra. Statistically significant differences were identified in the amide I region of spectral subpeak at 1,608 cm^-1, which is a representative endogenous marker for alterations in the structural conformation of COL I fibers. Moreover, epigenetic 5mC DNA modification was identified within cell nuclei by multivariate analysis. A statistically significant increase in signal intensities of spectral features indicative of DNA methylation was detected in cardiomyopathies in accordance with immunofluorescence 5mC staining. Overall, RMS is a versatile technology in the discrimination of cardiomyopathies based on molecular evaluation of COL I and nuclei while providing insights into the pathogenesis of the diseases.NEW & NOTEWORTHY Cardiomyopathies are associated with severe fibrotic remodeling of the heart, which is characterized by the excessive accumulation of collagen type I (COL I). In this study, we used marker-independent Raman microspectroscopy (RMS) to gain a deeper understanding of the disease's underlying molecular and cellular mechanisms.

Keywords: Raman spectroscopy; collagen; extracellular matrix; non-destructive imaging; pathological tissue remodeling.

Graphical abstract

Figure 1.

Histological staining allows the localization of fibrosis. A: Masson’s trichrome staining of control myocardium, ischemic heart disease (IHD), hypertrophic cardiomyopathy (HCM), and dilated cardiomyopathy (DCM) identify fibrosis by excessive collagen staining in blue. Scale bar = 500 µm. B: Picrosirius red (PSR) staining under brightfield illumination of control myocardium, IHD, HCM, and DCM. Scale bar = 100 µm. C: PSR staining under polarized light of control myocardium, IHD, HCM, and DCM. Scale bars = 100 µm. D: immunofluorescence (IF) images of control myocardium, IHD, HCM, and DCM show the complexity of the tissues. Colors in IF staining: Nuclei (blue), COL I (yellow), and αSMA (red). Scale bar = 100 µm. E: quantification of red collagen and green collagen fibers. F: coherency analysis shows parallel aligned fibers only in IHD. G: quantification of the amount of COL I based on IF images normalized by the whole tissue area. H: quantification of the amount of αSMA based on IF images normalized by the whole tissue area. Statistical analysis: t test, n = 5, *P < 0.05, ns: not significant.

Figure 2.

Raman imaging and spectral deconvolution (Sdc) allows marker-independent analysis of myocardium and different cardiomyopathies. A: True Component Analysis (TCA) images of control, IHD, HCM, and DCM. Scale bar = 20 µm. B: spectra identified by TCA: nuclei (blue), myosin (pink), COL III (turquoise), COL I (yellow), and αSMA (red). C: quantification of the amount of COL I based on Raman images normalized by the whole tissue area. D: quantification of the amount of αSMA based on Raman images normalized by the whole tissue area. E: score value analysis of PC-3 from COL I Raman spectra from control myocardium and cardiomyopathies. F: corresponding loading plot. G: peak area at 1,608 cm⁻¹ calculated based on Sdc of amide I area of averaged COL I spectra from control myocardium and different cardiomyopathies. H: peak width at 1,608 cm⁻¹ calculated based on Sdc of amide I area of averaged COL I from control myocardium and different cardiomyopathies. I: peak ratio of averaged COL I spectra at 1,608 cm⁻¹ normalized by amide I maximum. J: frequency of modes from filter image ratio at 1,608 cm⁻¹ normalized by amide I maximum. K: histogram of filter image ratio at 1,608 cm⁻¹ normalized by amide I maximum (1,667 cm⁻¹) of control and IHD. Lines in the histogram represent the 25, 50, and 75 percentiles while the widest position of the histograms is displaying the mode. L: histogram of filter image ratio at 1,608 cm⁻¹ normalized by amide I maximum (1,667 cm⁻¹) of control and HCM. M: histogram of filter image ratio at 1,608 cm⁻¹ normalized by amide I maximum (1,667 cm⁻¹) of control and DCM. Statistical analysis: t test, n = 5, *P < 0.05, **P < 0.01, ***P < 0.005, ns: not significant. COL I, collagen type I; COL III, collagen type III; DCM, dilated cardiomyopathy; HCM, hypertrophic cardiomyopathy; IHD, ischemic heart disease.

Figure 3.

Raman imaging identifies epigenetic alterations in cardiomyopathies. A: IF images of 5mC staining. Colors in IF staining: Nuclei (blue), COL I (yellow), and 5mC (turquoise). Scale bar = 100 µm. B: quantification of the amount of 5mC based on IF images normalized by nuclei. C: quantification of the fluorescence intensity of 5mC. D: score value analysis of PC-1 from nuclei Raman spectra from control myocardium and cardiomyopathies. E: corresponding loading plot. Statistical analysis: t test, n = 5, *P < 0.05, **P < 0.01, ***P < 0.005, ns: not significant. COL I, collagen type I; IF, immunofluorescence.