Fibres alignment has also a positive influence on collagen production. In fact, some studies showed that it is more abundant on aligned NFs-based scaffolds than on matrixes made of randomly assembled NFs [45, 46].

NFs parallelism can also produce a positive effect on cellular differentiation. It increases the expressions of the Scx gene and collagen XIV immediately after scaffold implantation [47].

The above results show the importance of surface nano-topography in terms of improved cellular behaviour.

One of the most critical weak points of current tissue engineering techniques applied to tendons is the regeneration of the tendon-bone transition zone [44, 48]. NMs could represent the major breakthrough towards this problem resolution. For example, Li et al. [49] developed a collector that imitated the collagen fibres organization that could be seen at the tendon-bone transition site. It comprised both aligned and randomly oriented NFs that imitated the structure of the collagen fibres in tendons and bones, respectively [50].

In order to repair a tendon, it is necessary to regenerate the sheath it glides into. It should be made of two walls, one external with anti-adhesion properties and one internal with a lubricated surface. For this reason, some researchers [45] developed a nanostructured membrane comprising a mixture of PCL and hydroxyapatite (HA) NFs internally and only PCL NFs externally (Fig. 2). They employed the electrospinning technique.

The HA on these NFs mimicked the one which was on tendons and stimulated the tissue regenerative process as well.

Some adherence areas were identified on the control group, while they were completely absent on the HA-based NFs.

In order to obtain an effective tendon repair, NFs could also be combined with some specific bio-molecules. For example, Sahoo et al. [51] developed a PLGA NFs-based scaffold onto which they applied a micrometre mesh reproducing the ECM structure. The researchers cultured rabbit-derived mesenchymal stem cells and showed that the combination of micro and nano features stimulated the collagen I and III production.

The tendon healing process is characterised by an increase in the beta-transforming growth factor (TGF-β) production. TGF-β₁ is one of its isoforms and causes the formation of fibrous tissues and adhesions areas [52]. In order to reduce the concentration of this protein, some specific antibodies could be employed. However, their half-life is very short and may cause a reduced clinical efficacy [53–55].

Alternatively, the concentration of these proteins could be varied by sending micro-RNA segments directly into the tendinous cells. Viruses are among the most largely experimented vectors [16, 17, 56, 57]. They showed high transfection values but were also characterised by a considerable toxicity. Since they may cause very dangerous immunologic and oncogenic responses [58, 59], they are not still ready for an in vivo clinical application.

This obviously stimulated the research on non-viral vectors. In this sense, nanoparticles (NPs) played a very important role [60, 61]. In fact, thanks to their tiny dimensions, they proved to be able to pass through the cellular membranes by the endocytosis mechanism [62, 63].

In order to further reduce the TGF-β₁ expression, other researchers [64] inserted micro-RNA filaments into plasmids and encapsulated this system into PLGA NPs. Plasmids release was dependent on pH and extended over a long time period. The researchers showed that the PLGA NPs–plasmids–micro-RNAs complex were able to penetrate the tendinous cellular membranes more deeply than the other non-viral vectors. However, this depended on the tendon area under study. In fact NPs penetration was very high in the injection site while their diffusion reduced as the thickness of the tissue increased.

Collagen NFs-based scaffolds for rotator cuff repair are partly already available on the market. Although they showed a satisfactory biological response on animal studies, their mechanical properties were insufficient to allow their clinical use [65, 66]. For example, Derwin et al. [67] compared four different types of scaffolds and showed that the porcine one was affected by a very rapid reabsorption.

Another important problem to deal with is the cellular penetration of the substrate. To resolve this issue, some researchers showed that soluble NFs could be employed to fix the poor infiltration of cells into the scaffold [68, 69]. In this case, the NFs arrangement did not influence the substrate permeability.

Other authors [70], by employing the electrospinning technology, developed a PCL NFs-based scaffold. They also produced a second hybrid scaffold made of poly-ε caprolactone (PCL) and polyethylene oxide (PEO) (Fig. 3). They aimed at finding out a more effective solution to repair rotator cuff tendons. The above NFs were combined and then studied on a rat model. Animals were sacrificed after 4 and 8 weeks. No postoperative problems were noticed. Histologic analyses revealed that PCL NFs-based scaffolds allowed a better cellular infiltration and colonisation. It is worth noting that these results were less significant if a percentage of the fibres were sacrificial. In fact, the fibres removal reduced the space available for the scaffold and therefore the cellular colonisation. However, this result was in contradiction with other studies [68], showing that the removal of part of the NFs produced an improved biological response in vitro and in sub-cutaneous implants.

Other researchers developed PLLA NFs-based scaffolds and cultured fibroblast obtained from the Long Head of the Biceps Tendon [71], onto their surface. Articular instability, tendon inflammation or its complete rupture were the main inclusion criteria of the study. The polymer showed hydrophobic character which reduced the cellular adhesion [72]. Same results were also obtained when mesenchymal stem cells were cultured on PLLA NFs combined with collagen and gelatin [73, 74].

The researchers paid a lot of attention to the ECM regeneration. They aimed at creating a substrate able to stimulate collage type I production [75–77].

Tendon-derived fibroblasts were cultured on the combination of PLLA and collagen type I NFs. The researchers observed that the genic expression and collagen type I production increased more on the combination of these NFs than on the PLLA NFs alone [71].

With respect to the fibroblasts that were cultured on PLLA/Collagen I NFs, researchers found out that the FAK, PYK and PI3K genes expressions were higher than the ones observed on glass and PLLA NFs. The results depended on the scaffold chemical composition and the physical properties were influenced by the nano-topography of the device’s surface.

The chemical interaction with integrins was the main responsible for the increased production of collagen type I. This result was also observed when osteoblasts and mesenchymal stem cells were cultured on different kinds of NFs [78]. The composite PLLA/Collagen I NFs were also able to stimulate the production of collagen type III. This protein along with collagen type X has a huge influence on the tendon healing process [79, 80].

Although the above studies look to be very encouraging, further investigations are still needed. In particular, researchers should check out whether the collagen type III and X influences the collagen type I production or not. Collagen type I is one of the most important constituents of the tendon as it provides the tissue with the mechanical resistance needed to resist to the external loads it is subjected to.