The biological basis of degenerative disc disease: proteomic and biomechanical analysis of the canine intervertebral disc

Abstract

Introduction: In the present study, we sought to quantify and contrast the secretome and biomechanical properties of the non-chondrodystrophic (NCD) and chondrodystrophic (CD) canine intervertebral disc (IVD) nucleus pulposus (NP).

Methods: We used iTRAQ proteomic methods to quantify the secretome of both CD and NCD NP. Differential levels of proteins detected were further verified using immunohistochemistry, Western blotting, and proteoglycan extraction in order to evaluate the integrity of the small leucine-rich proteoglycans (SLRPs) decorin and biglycan. Additionally, we used robotic biomechanical testing to evaluate the biomechanical properties of spinal motion segments from both CD and NCD canines.

Results: We detected differential levels of decorin, biglycan, and fibronectin, as well as of other important extracellular matrix (ECM)-related proteins, such as fibromodulin and HAPLN1 in the IVD NP obtained from CD canines compared with NCD canines. The core proteins of the vital SLRPs decorin and biglycan were fragmented in CD NP but were intact in the NP of the NCD animals. CD and NCD vertebral motion segments demonstrated significant differences, with the CD segments having less stiffness and a more varied range of motion.

Conclusions: The CD NP recapitulates key elements of human degenerative disc disease. Our data suggest that at least some of the compromised biomechanical properties of the degenerative disc arise from fibrocartilaginous metaplasia of the NP secondary to fragmentation of SLRP core proteins and associated degenerative changes affecting the ECM. This study demonstrates that the degenerative changes that naturally occur within the CD NP make this animal a valuable animal model with which to study IVD degeneration and potential biological therapeutics.

Fig. 1

Robotic biomechanical setup for intervertebral disc (IVD) motion segments. a and b Potted IVD motion segment mounted upon the robotic platform. c The rotational axis x, y and z reflect the three dimensions of movement afforded by biomechanical testing

Fig. 2

Western blots for fibromodulin, biglycan, and hyaluronan and proteoglycan link protein 1 (HAPLN1) expression in non-chondrodystrophic (NCD) and chondrodystrophic (CD) canine nucleus pulposus (NP) homogenates. Lane (i) depicts the expression of these proteins in articular cartilage used as a control. Lanes (ii) and (iii) represent the expression of these proteins in intervertebral disc NP homogenates for CD (beagle) and NCD (mongrel) samples. It is clear that there is strong expression in both articular cartilage and CD samples for all three proteins. However, in the NCD samples, fibromodulin and HAPLN1 are undetectable, and the expression of biglycan is markedly reduced

Fig. 3

Comparative immunohistochemical analysis of non-chondrodystrophic (NCD) and chondrodystrophic (CD) canine nucleus pulposus (NP) expression and distribution of decorin, biglycan, fibromodulin, hyaluronan and proteoglycan link protein 1 (HAPLN1), and aggrecan. For all extracellular matrix (ECM) proteins, the NCD canine intervertebral disc NP reveals a cobweb appearance, demonstrating intense staining for all proteins located in the areas tightly between the cells. Immunostaining with decorin reveals diffuse intercellular staining with negative immunostaining within the large, physaliferous-appearing notochordal cells. This presentation is in contrast to the abundant staining of these proteins in every CD NP sample. Decorin, biglycan, and HAPLN1 reveal intense staining within the ECM, with abundant clusters of small numbers of cells present within the NP. Although the ECM staining is less intense than that for the other three proteins, fibromodulin and aggrecan immunostaining is present. Furthermore, the CD NP stained for aggrecan reveals intense pericellular immunostaining diffusely throughout the ECM that is much less cellular than the NCD NP staining. Safranin-O staining shows quite intense ECM staining in the CD NP, whereas the NCD sample demonstrates intense intercellular staining without large, acellular ECM areas rich in proteoglycan staining. The overall appearance of the CD NP bears a strong resemblance to a fibrocartilaginous phenotype that is distinctly different from the NCD canine NP

Fig. 4

Western blots depicting the expression of decorin and biglycan after proteoglycan extraction from non-chondrodystrophic (NCD) and chondrodystrophic (CD) canine intervertebral disc (IVD) nucleus pulposus (NP). For each Western blot, lane (a) is undigested (U) proteoglycan extract from NCD (mongrel) and lane (c) is from CD (beagle) canine NP. Lanes (b) and (d) depict the detection of the specific small leucine-rich proteoglycan (SLRP) indicated above after digestion (D) with chondroitinase ABC and keratinase in NCD canine (lane b) and CD canine (lane d). a The decorin core protein is visualized appropriately at the 43 kDa molecular weight after digestion (lane b), with a clear, single band found from NCD NP extracts. NP extracts from CD canines (lane d) reveal fragmentation of decorin with multiple lower molecular weight bands visualized at 30, 25–20, and 17–15 kDa. b The biglycan core protein provides a clear, single band at the 45 kDa molecular weight in the NCD NP extract; however, the CD NP extract reveals multiple fragments of the core protein visualized at 37–35, 28, and 25 kDa

Fig. 5

Averaged moment–rotation curves during (a) flexion and extension and (b) axial rotation movements for non-chondrodystrophic (NCD; red) and chondrodystrophic (CD; white) specimens. Differences observed between moment–rotation curves are indicative of greater stiffness during flexion and extension and axial rotation loading in the NCD motion segments compared with CD motion segments