Background:
Fabry disease (FD) is caused by a deficient activity of the lysosomal enzyme α-galactosidase A (a-gal A), which leads to a systemic accumulation of globotriaosylceramide (Gb3) and subsequent progressive and heterogeneous multiorgan damage. Its manifestation in the heart represents the leading cause of death in these patients. Disease-specific enzyme replacement and chaperone therapy are available. However, treatment success strongly depends on its initiation before organ damage develops. Especially patients with residual a-gal A activity FD (“late onset” FD) frequently receive a late diagnose, because the clinical manifestation often affects only a single organ as the heart. Accordingly, late onset FD is often accompanied by less symptoms compared to classical FD and usually is presented by an unexplained cardiac hypertrophy. Hence, the need of novel early diagnostic tools for this rare inherited lysosomal storage disorder let us to evaluate the potential of nonlinear vibrational spectroscopy for the detection of FD associated biomolecular alterations.
Aim:
The presented contribution aimed to evaluate the potential of nonlinear spectroscopic imaging tools for the differentiation between controlled biosamples and those affected by rare inherited lysosomal storage disorder, i.e. Morbus Fabry.
Methods:
Coherent anti-Stokes Raman scattering (CARS) microscopy was utilized for cardiac FD-biosamples of α-Gal A-knockout mice and wild-type littermates as controls at the age of 20 weeks, a time point before phenotypic abnormalities have been reported. Since CARS is based on molecular vibrations, it is label free and allows rapid imaging of tissue sections with a high spectral selectivity based on CH2 and CH3 stretching vibrations. The spectroscopic measurements were assisted by multivariate data analysis, started by image pre-processing followed by image clustering and ended by data-driven modelling (PCA-LDA). The evaluation of the classification performance was accomplished based on the leave-one-mouse-out cross-validation aiming to differentiate between knockout and wild type groups.
Results/ Conclusion:
It was demonstrated that biomedical application of CARS microscopy is well suited for the visualization and quantitative analyses of lipid accumulations. This label-free nonlinear vibrational microscopy was implemented for the first time to investigate the FD mouse model as a novel fast and precise pathological screening tool. The tissue sections of over 50 mice were studied, the cardiac cells and heart tissue sections were imaged applying non-linear microscopy. Thanks to the high sensitivity of spectral information and strong input of computer-aided diagnosis, the subtle changes in protein-lipids contest between FD-affected and control heart tissues were detected with a sensitivity of up to 96 %. These findings open new diagnostic prospects in FD investigation and its cardiac involvements as a good add-in to the gold-standard histology.