Extensive annulus fibrosus (AF) radial tears lead to intervertebral disc (IVD) herniation. While unrepaired defects in the AF are associated with postoperative reherniation and high IVD degeneration prevalence, current surgical strategies are limited to symptomatic treatment of pain and disregard the structural integrity of the AF. For all these reasons, this study is focused on i) designing polycaprolactone (PCL) electrospun implants that mimic the multi-lamellar fibrous structure of the native tissue and ii) assessing their ability to properly close and repair an AF defect in a sheep in vivo model. Oriented PCL mats were produced by electrospinning with average fiber diameters of 1.3µm and a tensile modulus (55±1MPa) matching the one of a native human AF lamella (∼47MPa). In vitro experiments demonstrated a spontaneous colonization of PCL mats by human and ovine AF cells. In vivo study was carried out on 6 sheep in which 5 lumbar discs were exposed using a left retroperitoneal approach. Defects (2×5mm, 2mm depth) were created in the outer annulus, with randomized distribution of conditions including 10-layer oriented or non-oriented mats, untreated and healthy groups. X-ray and MRI examinations were performed every month until explantations at 1, 3 and 6 months, followed by immuno-histological analysis. Data showed no dislocation of the implants, cell infiltration between the PCL mats and within the mats, and a continuous type I collagen tissue formation between the implants and the surrounding AF tissue. These results highlight that multi-layer PCL electrospun mat is a promising biomaterial for AF repair.

The recent description of progenitor/stem cells in degenerated intervertebral discs (IVDs) raised the possibility of harnessing their regenerative capacity for endogenous repair. The aim of this work is to develop an intradiscal polysaccharide microbead-based delivery system for the sequential release of chemokines and nucleopulpogenic factors. This delivery system would sequentially contribute to 1) the recruitment of resident progenitors (CXCL12 or CCL5), 2) the differentiation of the mobilized progenitors (TGF-β1 and GDF5), and 3) the subsequent regeneration of NP. To determine the effects of chemokines on in vitro cell recruitment, human mesenchymal stem cells (MSC) were cultured in Transwells for 4h, with or without CXCL12 or CCL5. In parallel, pullulan microbeads (PMBs) (100µm) were prepared by a simultaneous crosslinking protocol coupled to a water-in-oil emulsification process. Freeze-dried PMBs were loaded with biological factors then release assays were performed at 37°C for 21 days and supernatant concentrations were measured by ELISA. As compared to untreated MSC, MSC migration was improved with a 3.9 (CXCL12) and 7.5 (CCL5) fold increase, respectively. All factors were successfully adsorbed on PMBs and a burst release within the 1^st day was observed. At day 7, 27.5% and 83% of CXCL12 and CCL5 were released, respectively and at day 21, 20% and 100% of TGF-β1 and GDF5 were released, respectively. Currently, released cytokine bioactivity is being analysed and an ex vivo ovine IVD model is developed to determine the repair potential of this controlled release approach.