A fine balance in tendon healing

Tendons have limited regenerative potential, and injuries often cause scarring. A study now identifies a tendon stem cell population that contributes to regeneration and a tendon fibro–adipogenic progenitor population involved in fibrosis.

Tendons connect muscles to bones. They are essential for locomotion and composed of highly organized connective tissue capable of withstanding high mechanical loads. These intriguing mechanical properties are the result of a complex structure of parallel-aligned collagen I fibers, proteoglycans, glycoproteins, elastin, and tenocytes1,2. The majority of tendons are surrounded by a sheath (paratenon or synovial sheath) comprised of loose connective tissue to reduce friction3. Despite their strength, tendons have limited healing capacity and do not fully regain their initial structural integrity and strength after injury4. Even with surgical intervention, tendon injuries can result in complex adhesion and scar-tissue formation, ultimately leading to increased pain and a decreased range of motion5. The limitations of surgical repair have led scientists to research tendon regeneration, but a more comprehensive picture of the cellular and signalling mechanisms underlying tendon regeneration and fibrosis is just starting to emerge. Although stem/progenitor cells have been isolated from tendons and cultured in vitro6,7, their cellular origin remains a topic of debate, and surface markers have not been clearly defined. Moreover, tendon stem/progenitor cells have never been studied in vivo. In this issue of Nature Cell Biology, Harvey et al.8 utilized single-cell transcriptomics, transgenic mice, inducible lineage tracing, and ectopic drug delivery to identify and map a previously unappreciated stem cell population within the tendon sheath, capable of generating new tenocytes and self-renewing in response to injury. In addition, the authors provided evidence to support that platelet-derived growth factor receptor alpha (PDGFRα) signalling facilitates tendon regeneration and fibrosis.

In an effort to gain an improved understanding of the cellular composition of adult tendons, the authors first categorized all of the cells present in adult mouse patellar tendons using single-cell RNA sequencing. Eight clusters of cells were identified, one of which was confirmed as tenocytes, and an unknown cluster in close proximity to the tenocytes on the T-distributed stochastic neighbor embedding (t-SNE) plot was described as potential tendon stem cells. This population was enriched in tubulin polymerization-promoting protein family member 3 (Tppp3+), a marker reportedly expressed in developing tendon sheath tissues, epitenon, and paratenon9. A third cluster of cells resembled the gene expression profile of fibro–adipogenic progenitor (FAP) cells and was termed T-FAPs (tendon FAPs) by the authors.

To further explore the relationship between the cluster of Tppp3+ cells and tenocytes, Harvey et al.8 used a Cre reporter mouse to trace the fate of Tppp3+ cells and found that these cells are present within patellar tendon sheaths during mouse development and persist in the adult tendon sheath, although they are mostly quiescent. This finding raised the question of whether these cells are also involved in a response to injury. Using lineage tracing, the authors demonstrated that Tppp3+ cells are activated after a full thickness transection of the middle of the tendon and incorporated into the tendon during healing.

Next, the authors investigated how Tppp3+ cells are involved in tendon regeneration by creating a Tppp3 knock-in mouse model that allowed for simultaneous monitoring of Tppp3 and scleraxis (scx) expression. Three cell types were observed: Tppp3+/Scx+ (‘de novo tenocytes’), Tppp3/Scx+ (cells possibly derived from unmarked Tppp3+ cells or perhaps proliferating on their own), and Tppp3+/Scx (putative ‘renewed stem cells’). Sequencing of these cells demonstrated that the Tppp3 lineage cells proliferate, give rise to tenocytes in the mid-substance of the tendon, and self-renew within the tendon sheath.

Transcriptome analyses of Tppp3+/Scx and Tppp3/Scx+ cells further revealed the enrichment of PDGFRα in the putative stem cells. Pdgfa and the decoy PDGF receptor, pdgfrl, were enriched in tenocytes. The differential expression of several PDGF signalling components suggests that PDGF signalling is involved in the regulation of tendon biology. Using Monocle 2 unsupervised pseudotime analysis, the authors modelled the relationship between the two clusters from the initial t-SNE plot (a tenocyte cluster and a putative stem cell cluster) and detected five states. Three Tppp3+ states were linked to tenocyte fate, with one specific state representing the Tppp3+/pdgfra+ cells, which had higher levels of Tppp3+ expression than the other two states. Based on these results, Harvey et al.8 proposed that the Tppp3+/pdgfra+ subpopulation likely represents the tendon stem cell state.

Lineage tracing of Tppp3 and PDGFRα populations further helped to clarify the location in which the putative tendon stem cells reside in vivo. This strategy allowed the authors to evaluate not only the tendon stem cell (Tppp3+/pdgfra+), but also Tppp3+/pdgfra and Tppp3/pdgfra+ cell populations. Ultimately, they found that all three populations are present within the paratenon sheath in different proportions. In addition, Tppp3+/PDGFRα cells rarely gave rise to tenocytes or fibroblasts. Tppp3/pdgfra+ cells were identified as T-FAPs, a possible cellular source for fibrotic scar formation within tendons.

Harvey et al.8 then deciphered the dual role of PDGFRα signalling in tenocyte generation and fibrosis. Ectopic delivery of PDGF-AA, platelet-derived growth factor-AA, into patellar tendons increased the number of Tppp3+ lineage tenocytes. Elevated numbers of ER-TR7+ cells (reticular fibroblasts) were also evident. The authors then conditionally knocked out PDGFRα in Tppp3+ cells and tracked the cell fate to test whether PDGFRα signalling is required for tendon regeneration. Results indeed confirmed that PDGFRα signalling promotes the proliferative capacity of tendon stem cells and is important for tenocyte differentiation. An increased number of ER-TR7+ cells in injured knockout mice was also identified, suggesting that a substantial amount of fibrosis originates from the T-FAPs secondary to failed tendon regeneration (Fig. 1).

Fig. 1: A schematic of the two cell populations identified by Harvey et al. and their distinct roles in tendon healing.

One population, termed a tendon stem cell (TSC), is characterized by its expression of Tppp3+/pdgfra+ and involved in tendon regeneration. The second population is the tendon-fibro–adipogenic progenitor (T-FAP), with a role in tendon fibrosis. The authors observe that PDGFRα signalling plays a role in both tenocyte generation and fibrosis. Without PDGFRα signalling, TSC regeneration is impaired and T-FAPs dominate in this context with more extensive fibrosis.

Finally, the authors examined the differentiation capacities of three cell types. Not surprisingly, T-FAPs (Tppp3/pdgfra+) have the same potential as FAPs, which are mesenchymal stem-cell-like progenitors found in many tissue types, and differentiated into adipogenic, chondrogenic, and osteogenic cells7. Tendon stem cells (Tppp3+/pdgfra+) differentiated into chondrogenic and osteogenic cells, but not adipocytes, unlike tendon progenitor cells previously described7. In this aspect they resembled skeletal stem cells, which undergo differentiation into bone, cartilage and stroma, but not adipocytes10. Tppp3+/pdgfra cells only underwent chondrogenic differentiation.

In summary, this body of research sheds new light on our basic understanding of tendon regeneration and identifies a possible tendon stem cell population. Differentiation assays support the multipotency of these cells, and transcriptomic evidence supports their ability to self-renew. Further investigation is needed to address whether these cells also participate in wound healing when the tendon is completely disrupted and whether the same surface markers label human tendon stem cells, as markers can differ between species10,11. Furthermore, it would be interesting to evaluate the possibility of a crosstalk between tendon stem cells and skeletal stem cells at the tendon-to-bone interface in both homeostasis and repair.

The finding that PDGF signalling is involved in both regeneration and fibrosis is confirmatory to some extent. Previous literature has demonstrated that PDGF signalling can attract both tenocytes and fibroblasts, and in animal models ectopic platelet-derived growth factor-BB (PDGF-BB) delivery has yielded mixed results in tendon regeneration12.

The links between tendon stem-cell-mediated regeneration, (possibly T-FAP-mediated) fibrosis and PDGF signalling, as reported here, warrant further in-depth exploration of the mechanisms that regulate this process. With a deeper understanding of PDGF signalling responses, alternative therapeutic agents could be developed to enhance tendon regeneration and ideally prevent fibrosis by modifying PDGFRα signalling.

Additionally, it remains unknown whether the cellular makeup, specifically the ratio of tendon stem cells to T-FAPs, can alter regeneration after injury in the presence of PDGFRα signalling. Heterogeneous populations of fibroblasts adjacent to the tendon injury may likely also exert an influence during repair, similarly to what has been reported for healing skin wounds13,14.T-FAPs may give rise to fibroblasts with different roles in tendon fibrosis. A detailed characterization of fibroblast subtypes within tendons and their sheaths, in combination with an improved understanding of the possible contributions of these subtypes within tendon fibrosis and scarring, may offer efficient strategies to prevent tendon fibrosis and help mitigate the pain and loss of function.

In conclusion, even though the complex contribution of the microenvironment requires further elucidation, the latest work by Harvey et al.8 represents an important step forward in the study of tendon regeneration and the cellular and molecular mechanisms involved.


  1. 1.

    Titan, A. L., Foster, D. S., Chang, J. & Longaker, M. T. Plast. Reconstr. Surg. 144, 639e–647e (2019).

    CAS  Article  Google Scholar 

  2. 2.

    Juneja, S. C. & Veillette, C. A. Arthritis 2013, 154812 (2013).

    Article  Google Scholar 

  3. 3.

    Kannus, P. Scand. J. Med. Sci. Sports 10, 312–320 (2000).

    CAS  Article  Google Scholar 

  4. 4.

    Killian, M. L., Cavinatto, L., Galatz, L. M. & Thomopoulos, S. J. Shoulder Elbow Surg. 21, 228–237 (2012).

    Article  Google Scholar 

  5. 5.

    Snedeker, J. G. & Foolen, J. Acta Biomater. 63, 18–36 (2017).

    Article  Google Scholar 

  6. 6.

    Zhang, J. & Wang, J. H. Methods Mol. Biol. 1842, 217–228 (2018).

    CAS  Article  Google Scholar 

  7. 7.

    Bi, Y. et al. Nat. Med. 13, 1219–1227 (2007).

    CAS  Article  Google Scholar 

  8. 8.

    Harvey, T., Flamenco, S. & Fan, C.- M. Nat. Cell Biol. (2019).

  9. 9.

    Staverosky, J. A., Pryce, B. A., Watson, S. S. & Schweitzer, R. Dev. Dyn. 238, 685–692 (2009).

    CAS  Article  Google Scholar 

  10. 10.

    Chan, C. K. et al. Cell 160, 285–298 (2015).

    CAS  Article  Google Scholar 

  11. 11.

    Chan, C. K. F. et al. Cell 175, 43–56.e21 (2018).

    CAS  Article  Google Scholar 

  12. 12.

    Evrova, O. & Buschmann, J. Eur. Cell. Mater. 34, 15–39 (2017).

    CAS  Article  Google Scholar 

  13. 13.

    Hu, M. S., Moore, A. L. & Longaker, M. T. J. Invest. Dermatol. 138, 729–730 (2018).

    CAS  Article  Google Scholar 

  14. 14.

    Rinkevich, Y. et al. Science 348, aaa2151 (2015).

    Article  Google Scholar 

Download references

Author information



Corresponding author

Correspondence to Michael T. Longaker.

Ethics declarations

Competing interests

The authors declare no competing interests.

Rights and permissions

Reprints and Permissions

About this article

Verify currency and authenticity via CrossMark

Cite this article

Titan, A.L., Longaker, M.T. A fine balance in tendon healing. Nat Cell Biol 21, 1466–1467 (2019).

Download citation


Quick links

Nature Briefing

Sign up for the Nature Briefing newsletter — what matters in science, free to your inbox daily.

Get the most important science stories of the day, free in your inbox. Sign up for Nature Briefing