Introduction One of the hallmarks of developing tendinopathy after injury is the accumulation of toluidine-blue-positive sulfated glycosaminoglycan, attributed to increased expression and synthesis of the large proteoglycans (PGs) aggrecan and versican. Indeed, there is evidence that, in many instances, these molecules are already chronically present in the tendon extracellular matrix (ECM) before an acute rupture. Our previous study in a surgically induced infraspinatus (IST) tendinopathy model in sheep confirmed the progressive increase of large PGs throughout the pathological tensile tendon1 with frank chondroid metaplasia evident by 52 weeks. As the IST experiences compressive loading during shoulder movement and is relatively short (∼4 cm), it was unclear how widespread the chondroid metaplasia is from the injury site and whether similar pathology occurs in tendons subjected largely to tensile loading.

To address this question we used a similar partial transection model in the equine superficial digital flexor tendon (SDFT). This 25 cm long tensile tendon is commonly injured in athletic horses with a very poor long-term prognosis. The aim of this study was to determine the significant factors affecting histopathological distribution of tendon PGs and their expression, along with other genes involved in their metabolism, using mixed regression models.

Methods PG content (toluidine blue staining) was scored in regional SDFT (12 regions) sections from formalin-fixed tissue harvested from animals (n=6 per group) 6 weeks after sham or mid-tendon lateral hemitransection surgery at the midmetacarpus. Aggrecan (ACAN), versican (VCAN), ADAMTS4, ADAMTS5 and TIMP1, 2 & 3 gene expression was measured by real time RT-PCR on samples from the same regions. Mixed regression models (clustered by animal and region) were fitted to log-transformed expression data to determine whether surgery or any of the positional (side or proximity to cut/carpus/MCP) covariates was a significant factor.

Results Both PG histological scoring and ACAN expression (p=0.006) were higher with distance from the midmetacarpus before surgery but not afterwards, when ACAN expression increased (p<0.001), particularly in the regions around the surgery site, and PG score increased in all but one region (figure 1). VCAN was evenly expressed in control SDFT but increased upon surgery (p<0.001) to be significantly higher on the lateral side (p=0.002) and towards the carpus (p=0.010). ADAMTS4 was expressed significantly more on the lateral side (p<0.001) and towards the carpus (p<0.001) both before and after surgery however decreased with surgery (p<0.001) significantly more nearer the lesion (p<0.001). ADAMTS5 expression was not significantly altered by transection (p=0.29). Surgery increased TIMP1 (p=0.019), decreased TIMP3 (p<0.001) and had no significant effect on TIMP2 (p=0.36) expression. Only TIMP1 varied by region, being higher on the lateral side after surgery (p<0.001).

• Download figure
• Open in new tab
• Download powerpoint

Figure 1.

Map of proteoglycan scores of regional histological tendon sections from horses with and without a lateral SDFT hemitransection 6 weeks earlier. Region is shaded if difference between sham (light bars) and cut (dark bars) is significant.

Discussion By 6 weeks, a midtendon injury caused pathology with increased PG throughout the full length of the tendon. Increased ACAN and VCAN expression did not account for all regional changes; decreased ADAMTS4 expression suggested turnover of existing PGs was also impaired. The widespread chondroid metaplasia following a focal injury will significantly affect the tendon material properties and may, in part, explain the poor long-term prognosis.