Enthesis-related progenitors recruited from the subacromial bursa contribute to rotator cuff healing in rats

Understanding the origin and regulatory mechanisms of enthesis-related progenitor cells expressing scleraxis (Scx) and SRY-box-containing gene 9 (Sox9) may help develop therapeutic approaches to improve endogenous rotator cuff (RC) tendon-to-bone healing capabilities. This study explores the characteristics of Scx⁺/Sox9⁺ cells during RC healing. The effects of subacromial bursa (SAB) preservation on Scx⁺ and Scx⁺/Sox9⁺ progenitor cell recruitment and the histological and biomechanical maturation of postoperative RC tendon-to-bone healing are evaluated in rats.

Twelve-week-old wild-type Sprague-Dawley rats (n = 71) and ScxGFP transgenic rats (n = 6) underwent unilateral surgery for supraspinatus tendon repair immediately after transection. Four in vivo models were used: intact, SAB injury-only, SAB-preserved RC repair, and SAB-resected RC repair. Cells isolated from reparative entheses 1 week postoperative and intact entheses were analyzed via single-cell RNA sequencing. Progenitor marker (CD34, PDGFRα, Scx, and Sox9) expression was assessed by immunostaining. The SAB-preserved and SAB-resected RC repair groups were compared through histological and biomechanical analysis 3 and 6 weeks postoperatively.

Single-cell RNA sequencing identified 17 transcriptionally distinct cell clusters isolated from the intact and SAB-preserved groups. After reclustering the mesenchymal population, 4 distinct clusters were identified: Pdgfra^+/Cd34⁺ mesenchymal progenitors, tendon progenitors, Scx⁺ tendon fibroblasts, and Scx⁺/Sox9⁺ enthesis progenitors. Lineage analysis revealed that Scx⁺ and Scx⁺/Sox9⁺ cells differentiated from Pdgfra⁺/Cd34⁺ mesenchymal progenitors. In immunostained intact entheses, PDGFRα⁺ and CD34⁺ cells were localized on the surface of the supraspinatus tendon, and humeral greater tuberosity in the area assumed to be the SAB layer. In reparative entheses, Scx⁺ and Scx⁺/Sox9⁺ cells were within the reparative tissue between the bone and supraspinatus tendon, near the SAB, and sparser in the SAB-resected group. The second harmonic generation signal revealed that the SAB-resected group exhibited less collagen fiber orientation and inferior biomechanical properties. At 3 weeks postoperatively, the SAB-preserved RC repair group exhibited higher ultimate load to failure (SAB-preserved: 6.8 ± 1.9 N vs. SAB-resected: 3.3 ± 2.2 N, P = .0004) and ultimate stress to failure (SAB-preserved: 1.1 ± 0.4 N/mm² vs. SAB-resected: 0.5 ± 0.3 N/mm², P = .0001).

During the RC repair process, Scx⁺ and Scx⁺/Sox9⁺ enthesis-related progenitors may be associated with endogenous PDGFRα⁺ and CD34⁺ mesenchymal stem/progenitors in the SAB near the tendon-bone repair site. SAB preservation positively impacts early histological and mechanical maturation after RC tendon-bone repair.