Detection of Age-Related Changes in Tendon Molecular Composition by Raman Spectroscopy-Potential for Rapid, Non-Invasive Assessment of Susceptibility to Injury

Abstract

The lack of clinical detection tools at the molecular level hinders our progression in preventing age-related tendon pathologies. Raman spectroscopy can rapidly and non-invasively detect tissue molecular compositions and has great potential for in vivo applications. In biological tissues, a highly fluorescent background masks the Raman spectral features and is usually removed during data processing, but including this background could help age differentiation since fluorescence level in tendons increases with age. Therefore, we conducted a stepwise analysis of fluorescence and Raman combined spectra for better understanding of the chemical differences between young and old tendons. Spectra were collected from random locations of vacuum-dried young and old equine tendon samples (superficial digital flexor tendon (SDFT) and deep digital flexor tendon (DDFT), total n = 15) under identical instrumental settings. The fluorescence-Raman spectra showed an increase in old tendons as expected. Normalising the fluorescence-Raman spectra further indicated a potential change in intra-tendinous fluorophores as tendon ages. After fluorescence removal, the pure Raman spectra demonstrated between-group differences in CH₂ bending (1450 cm^-1) and various ring-structure and carbohydrate-associated bands (1000-1100 cm^-1), possibly relating to a decline in cellular numbers and an accumulation of advanced glycation end products in old tendons. These results demonstrated that Raman spectroscopy can successfully detect age-related tendon molecular differences.

Keywords: Raman spectroscopy; ageing; autofluorescence; collagen; glycation; tendinopathy.

Figure 1

Averaged charge-coupled device (CCD) counts of young (40 spectra) and old (30 spectra) superficial digital flexor tendons (SDFTs, Left) and deep digital flexor tendons (DDFTs, Right) (40 spectra in both groups).

Figure 2

Averaged spectra after min–max normalised charge-coupled device (CCD) counts of young and old group in SDFTs (Left, young:old = 40:30 spectra) and DDFTs (Right, 40 spectra in both groups).

Figure 3

Principal component analysis on min–max normalised spectra. (A) SDFT, scatter plot of PC1 and PC3 axis. Red squares: young tendons; black squares: old tendons. (B) SDFT, PC1 (black line) and PC3 (red line) vector loading plot. (C) DDFT, scatter plot of PC1 and PC3 axis. Red squares: young tendons; black squares: old tendons. (D) DDFT, PC1 (black line) and PC3 (red line) vector loading plot.

Figure 4

Averaged pooled Raman spectra of young (70 spectra) and old (80 spectra) tendons.

Figure 5

Principal component analysis of pooled Raman spectra. Left, scatter plot of PC1 and PC3 axis. PC1 separates tendon types and PC3 separates age groups. Right, PC1 and PC3 vector loadings.