振動分光法を用いた赤血球の分光化学的病期分類による次世代糖尿病診断

Type 2 diabetes mellitus (T2DM) induces progressive biochemical and morphological alterations in red blood cells (RBCs), driven by hyperglycemia-induced oxidative stress and disrupted lipid-protein homeostasis. These changes compromise membrane integrity and serve as sensitive molecular indicators of disease severity. Yet, a non-invasive strategy for capturing such alterations at the molecular level remains an unresolved challenge.

This study employs Fourier Transform Infrared Spectroscopy (FT-IR) and Raman Spectroscopy (RS) to uncover sex- and age-specific molecular alterations in RBCs induced by T2DM. By integrating classical and advanced analytical techniques, we identify novel spectrochemical markers of disease progression, offering a non-invasive, label-free framework for staging and monitoring diabetic pathology at the molecular level.

A total of 120 db/db and C57BL/6J mice were tested at 7, 12 and 24 weeks of age. Standard hematological and biochemical analyses were combined with vibrational spectroscopy (FT-IR, RS) and atomic force microscopy (AFM), to assess RBC morphology and molecular composition. Spectral data was analyzed using univariate and multivariate statistics to identify disease-specific spectral markers.

Partial Least Squares Discriminant Analysis reliably distinguished diabetic and control RBC profiles, confirming the presence of robust, disease-specific molecular patterns. Integration with hematological and biochemical parameters further validated the diagnostic relevance of this label-free, non-invasive approach. Structural protein analysis revealed a consistent decline in α-helical content and an increase in β-sheet and β-turn structures, reflecting protein misfolding and aggregation, particularly in older diabetic females. Alterations in disulfide bonding, hydrated β-sheets and H-bonded antiparallel β-sheets highlighted oxidative stress-mediated membrane destabilization. Additionally, elevated plasma triglyceride levels and increased saturation of RBC membrane lipids indicate impaired lipid handling.

Vibrational spectroscopy enables non-invasive, molecular-level profiling of diabetes-induced RBC alterations, revealing sex- and age-dependent signatures of disease progression. This approach offers a promising platform for precise staging and stratification in metabolic disorders.