Cellulose Nanofiber-Reinforced Chitosan Hydrogel Composites for Intervertebral Disc Tissue Repair

The development of non-cellularized composites of chitosan (CHI) hydrogels, filled with cellulose nanofibers (CNFs) of the type nanofibrillated cellulose, was proposed for the repair and regeneration of the intervertebral disc (IVD) annulus fibrosus (AF) tissue. With the achievement of CNF-filled CHI hydrogels, biomaterial-based implants were designed to restore damaged/degenerated discs. The structural, mechanical and biological properties of the developed hydrogel composites were investigated. The neutralization of weakly acidic aqueous CNF/CHI viscous suspensions in NaOH yielded composites of physical hydrogels in which the cellulose nanofibers reinforced the CHI matrix, as investigated by means of microtensile testing under controlled humidity. We assessed the suitability of the achieved biomaterials for intervertebral disc tissue engineering in ex vivo experiments using spine pig models. Cellulose nanofiber-filled chitosan hydrogels can be used as implants in AF tissue defects to restore IVD biomechanics and constitute contention patches against disc nucleus protrusion while serving as support for IVD regeneration.


Introduction
Fiber-reinforced hydrogels can be engineered to mimic many biological tissues as they present similar microstructural and mechanical properties, and functionality. Thus, bioinspired composite hydrogels have great potential for tissue engineering applications. The intervertebral disc (IVD), which is the largest avascular tissue of the human body [1,2] can constitute a source of bioinspiration for the engineering of hydrogel biomaterials. The IVD connects the vertebral bodies (VBs) while keeping dispersion of CNFs in CHI to produce cast films has been reported [31,36]. The question of CNF toxicity has been addressed and the low toxicity risk potential for CNFs has been reported according to ecotoxicological, cytotoxicity, and pro-inflammatory response studies [38,39]. In vitro tests carried out inhuman and murine cells have demonstrated the absence of cytotoxicity and genotoxicity for MFC suspensions [40]. In previous works, we demonstrated the good biocompatibility of CNFs in cytotoxicity evaluations in human dermal fibroblasts (HFIB-D) and human bone marrow stromal cells (HBMSCs) [41]. Cellulose nanofibers have been used as biomaterial reinforcement in tissue engineering [41][42][43]. The biocompatibility of CNFs and their non-biodegradability in the human body motivated the proposal of CNFs for the development of implants which require long-term stability to engineer tissues of high-mechanical performance (e.g., bone). Cellulose nanofiber scaffolds have been proposed for the long-term proliferation of bone-related cells [44]. In hydrogel biomaterials, the use of CNFs is promising. In addition to their mechanical performance, CNFs present high-water retention and can yield transparent biomaterials [42,45]. Moreover, CNFs can be oriented within hydrogels, by the uniaxial stretching of their composites under controlled environmental humidity. Osorio-Madrazo et al. [21] performed pioneering works where CNFs, specifically nanowhiskers, were oriented in bulk hydrogel matrices to produce reinforced anisotropic hydrogels. Thermosensitive injectable hydrogels of CHI containing TOCNFs at concentrations of 0.2%, 0.4%, 0.6%, and 0.8% (w/v) have been designed for biomedical applications [46]. These hydrogels undergo sol-gel transition at body temperature through reversible interactions between CHI and β-glycerophosphate. Both MC3T3-E1 pre-osteoblast cells and L929 fibroblast cells have shown biocompatibility towards CHI/TOCNF 0.4% (w/v) [46]. On one hand, the excellent biological properties of CHI, its ability to form hydrogels, the CNFs outstanding mechanical properties and demonstrated biocompatibility; and on the other hand, the fiber-reinforced hydrogel structure of the IVD motivated us to propose the development of CNF-filled CHI hydrogels for IVD tissue repair and regeneration. Different parameters, such as CHI concentration and CNF content, were considered in the processing of the composite hydrogels. The effect of these parameters in the structure and mechanical properties of the achieved hydrogel biomaterials was investigated. To assess the suitability of the obtained CNF/CHI composite hydrogels for IVD tissue repair, we performed studies in which these biomaterials were implanted in the AF tissue region of the disc to re-establish the biomechanics of lesioned discs, while serving as support for IVD regeneration and contention patch against NP protrusion.

Cellulose Nanofibers
Gel-like suspensions of CNF were obtained from bleached pine sulfite dissolving pulp by a mechanoenzymatic method adapted from Doench et al. [41] and Pääkkö et al. [47] at the Centre Technique du Papier (CTP, Grenoble, France). Pine sulfite pulp was incubated for 1 hat 50 • C with a solution of endoglucanase FiberCare R ® (Novozymes Biologicals, Paris, France) at pH 5.0. The pulp was refined at 4.5% consistency with a 30.5 cm single disk refiner for 25 min. The digested samples were further refined to obtain a pulp suspension of Schopper-Riegler (SR) number higher than 80 and mean fiber length smaller than 300 µm. A quantity of 2% (w/w) fiber suspensions was processed with an Ariete homogenizer (GEA, Düsseldorf, Germany), involving one pass at 1000 bar followed by three passes at 1500 bars. The obtained CNFs exhibited a surface charge density of 40-80 mmol/kg. Thus, they were weakly charged with carboxylate moieties. The morphology of the CNFs was characterized by atomic force microscopy (AFM)with a NanoScope V controller using the NanoScope 9.1 software (Bruker Corporation, Santa Barbara, CA, USA). A droplet of a 0.001% CNF suspension was placed on a freshly cleaved mica surface. After sample drying, observations were performed in tapping mode with the AFM equipped with a tube scanner from Veeco Digital Instruments (Santa Barbara, CA, USA), using silicon tips (PPP-NCH, Nanosensors, Sindelfingen, Germany) with resonance frequency and spring constant of 360 kHz and 50 N m −1 , respectively.

Preparation of the CNF-Filled Chitosan Composite Hydrogels
A fine powder of CHI was mixed at 2 or 3% (w/w) with CNFs in water at a given CNF content (0.2%; 0.4% (w/w)). The dispersions were sonicated with a SONOPULS Ultrasonic homogenizer (Bandelin electronic GmbH, Berlin, Germany) for 5 min at 40% amplitude. Then, acetic acid was added in stoichiometric ratio to the free amine groups of the glucosamine units of CHI, to solubilize the CHI (DA = 2.5%). The mixture was kept under mechanical stirring overnight. Afterwards, the CNF/CHI viscous suspensions were neutralized with 2 M NaOH for 1 h using Petri dishes as molds. After neutralization, the gelled material could be removed from the molds and washed with distilled water untilneutral and complete removal of the salts formed during the neutralization. Finally, composite hydrogels of different CNF/CHI concentrations, of around 1-2 mm thickness, were obtained depending on the amount (1 or 2 g, respectively) of CNF/CHI viscous suspension added to the 3 × 3 cm Petri dish molds to proceed to the neutralization step. Hydrogels of CNF:CHI weight ratios of 0.1:1 and 0.2:1 with a CHI concentration of 2% (w/w); and of 0.07:1 and 0.1:1 with a CHI concentration of 3% (w/w) were produced. To characterize the dispersion of the CNFs in the CHI matrix, the freeze-dried CNF/CHI scaffolds were carefully fractured and gold sputtered in a Polaron SC 7640 (VG Microtech, East Sussex, UK), and observed by scanning electron microscopy (SEM) (Amray Inc., Bedford, MA, USA) at an accelerating voltage of 15 kV. It is worth noticing that with this technique the analysis was performed in the dry state. Nevertheless, it allowed to get insight into the dispersion of the CNFs in the dry scaffold composites, which were originated from the hydrogel processing.

Microtensile Testing of CNF-Filled Chitosan Composite Hydrogels
For the tensile straining experiments, ≈2 mm wide and ≈7 mm long strips were cut out from the prepared flat nanocomposite hydrogels with a razorblade. Each sample was glued immediately onto a foliar frame with a test span of about 7 mm by using cyanoacrylate glue (Loctite ® 454, Germany). Afterwards, the frame with the sample was mounted onto a home-made microtensile tester surrounded by a sealed sample chamber, which allowed for controlling relative humidity (RH) during the experiment. The humidity generator (Wetsys, Setaram Instrumentation, Caluire, France) mixes dry air and water-saturated air in a controlled way to supply a gaseous flow with a stable RH that can be tuned between 5 and 95% (±0.3%) at a given temperature. At a RH of 45%, the hydrogel nanocomposite was uniaxially stretched at a constant speed of 0.2 µm/s. The RH was kept constant during the whole experiment (±2% RH). The applied force (F) was measured by a load cell with a maximum capacity of 50 N. The displacement of the motorized linear stage was recorded. Nominal stress σ = F/(w 0 t 0 ) and nominal strain ε = (l − l 0 )/l 0 values were estimated with w 0 being the initial width, t 0 the initial thickness, l the actual length, and l 0 the initial length of the cut hydrogel strip. At least three different strips of a given sample were tested under similar experimental conditions. It is worth noticing that the true stress of the sample should be different from the calculated one because of a reduction of the sample cross-sectional area (wt) caused by straining and/or partial drying. The design of a bioactive hydrogel to seal, reinforce, and repair the AF disc region would ensure appropriate long-term tissue regeneration while providing a mechanical support to avoid post-surgery hernia reoccurrence. To this end, the determination of spine mechanical loading and motion is challenging in biomechanics. To understand its functionality, the needed smallest unit (motion segment) consists of two vertebrae and the disc joining them together [6]. Around 30 kg growing-finishing pigs were used as models for compression testing of healthy and damaged discs, these latter repaired with the composite hydrogels. Discs localized between the lumbar L5 and thoracic T7 vertebrae were considered ( Figure 1). Motion segments were dissected, consisting of two adjacent VEPs and the in-between disc. A total of 9 motion segments were tested: T8-T9, T10-T11, T11-T12, T12-T13, T14-T15, L1-L2, L2-L3, L3-L4, and L4-L5 discs ( Figure 1). The design of a bioactive hydrogel to seal, reinforce, and repair the AF disc region would ensure appropriate long-term tissue regeneration while providing a mechanical support to avoid postsurgery hernia reoccurrence. To this end, the determination of spine mechanical loading and motion is challenging in biomechanics. To understand its functionality, the needed smallest unit (motion segment) consists of two vertebrae and the disc joining them together [6]. Around 30 kg growingfinishing pigs were used as models for compression testing of healthy and damaged discs, these latter repaired with the composite hydrogels. Discs localized between the lumbar L5 and thoracic T7 vertebrae were considered ( Figure 1). Motion segments were dissected, consisting of two adjacent VEPs and the in-between disc. A total of 9 motion segments were tested: T8-T9, T10-T11, T11-T12, T12-T13, T14-T15, L1-L2, L2-L3, L3-L4, and L4-L5 discs ( Figure 1). The compression tests have been performed on a Shimadzu Autograph AG-X plus (Kyoto, Japan) equipped with a 10 kN load cell. After a pre-loading of 200 N [7], compression was applied until failure. Concerning lesioned discs, a fenestration of length ≈10 mm, width ≈5 mm, and thickness ≈1.5 mm was performed in the ventral part of the AF. A composite hydrogel (3% CHI/0.4% CNF (w/w)) piece of similar dimensions as the performed fenestration was carefully implanted within the AF defect. Then, the disc was sutured at the implanted position and subjected to compression loading. Figure 2 shows the methodology followed to prepare and mechanically test a fenestrated disc that was implanted with the hydrogel. The compression tests have been performed on a Shimadzu Autograph AG-X plus (Kyoto, Japan) equipped with a 10 kN load cell. After a pre-loading of 200 N [7], compression was applied until failure. Concerning lesioned discs, a fenestration of length ≈10 mm, width ≈5 mm, and thickness ≈1.5 mm was performed in the ventral part of the AF. A composite hydrogel (3% CHI/0.4% CNF (w/w)) piece of similar dimensions as the performed fenestration was carefully implanted within the AF defect. Then, the disc was sutured at the implanted position and subjected to compression loading. Figure 2 shows the methodology followed to prepare and mechanically test a fenestrated disc that was implanted with the hydrogel.

Cell Culture of NIH/3T3 Fibroblasts on CNF/CHI Hydrogels
To evaluate the suitability of the CNF/CHI composite hydrogels for IVD tissue engineering, experiments were performed with fibroblast cells cultured on the hydrogels of different CHI concentrations and CNF contents. Cell cultures of 3T3 cells developed using the NIH Swiss mouse embryo fibroblasts were performed (murine fibroblast, strain: NIH/Swiss). NIH/3T3 fibroblast cells were grown in T75 (75 cm 2 ) cell culture flasks (Sarstedt, Nümbrecht, Germany). The cells were cultured in Dulbecco's modified Eagle medium (DMEM) supplemented with 2 mM L-glutamine and 10% fetal bovine serum (FBS) (Gibco, Thermo Fisher Scientific, Leicestershire, UK) at 37 °C in a humidified atmosphere of 5% CO2 for 1 week. Upon 90% confluence, cells were rinsed twice with phosphate-buffered saline (PBS) (Gibco, Thermo Fisher Scientific, Leicestershire, UK) followed by detachment with trypsin/ethylenediaminetetraacetic acid (EDTA) for 5 min and neutralization with the corresponding cell culture medium. After detachment, cells were spun down in a centrifuge for 5 min at 110 rcf (Rotor F-45-30-11, Eppendorf 5417R, Hamburg, Germany). The supernatant was discarded and cells were diluted into the culture medium. The CNF/CHI hydrogel pieces were put in 24-well cell culture plates, covering the whole well surface for subsequent use for cell growth. Firstly, the hydrogel pieces were incubated with 1 mL of culture medium for 30 min. Then, the medium was removed and cells were transferred to each well in wished dilution. A suspension of

Cell Culture of NIH/3T3 Fibroblasts on CNF/CHI Hydrogels
To evaluate the suitability of the CNF/CHI composite hydrogels for IVD tissue engineering, experiments were performed with fibroblast cells cultured on the hydrogels of different CHI concentrations and CNF contents. Cell cultures of 3T3 cells developed using the NIH Swiss mouse embryo fibroblasts were performed (murine fibroblast, strain: NIH/Swiss). NIH/3T3 fibroblast cells were grown in T75 (75 cm 2 ) cell culture flasks (Sarstedt, Nümbrecht, Germany). The cells were cultured in Dulbecco's modified Eagle medium (DMEM) supplemented with 2 mM L-glutamine and 10% fetal bovine serum (FBS) (Gibco, Thermo Fisher Scientific, Leicestershire, UK) at 37 • C in a humidified atmosphere of 5% CO 2 for 1 week. Upon 90% confluence, cells were rinsed twice with phosphate-buffered saline (PBS) (Gibco, Thermo Fisher Scientific, Leicestershire, UK) followed by detachment with trypsin/ethylenediaminetetraacetic acid (EDTA) for 5 min and neutralization with the corresponding cell culture medium. After detachment, cells were spun down in a centrifuge for 5 min at 110 rcf (Rotor F-45-30-11, Eppendorf 5417R, Hamburg, Germany). The supernatant was discarded and cells were diluted into the culture medium. The CNF/CHI hydrogel pieces were put in 24-well cell culture plates, covering the whole well surface for subsequent use for cell growth. Firstly, Biomimetics 2019, 4, 19 7 of 17 the hydrogel pieces were incubated with 1 mL of culture medium for 30 min. Then, the medium was removed and cells were transferred to each well in wished dilution. A suspension of the cells in culture medium (500 µL) was added on the hydrogel surface and NIH/3T3 cells were seeded at 200,000 cells per well on average in triplicates. Cells seeded in empty wells (i.e., without the hydrogel) were used as control.

Results and Discussion
3.1. Structure of the CNF/CHI Hydrogels Figure 3 shows atomic force microscopy (AFM) images of the CNFs used as nanoreinforcement of the CNF/CHI composite hydrogels. They reveal an entangled network of interconnected nanofibrils with average width of 35.2 ± 8.1 nm and bundles up to 100 nm width. The relative larger width of the fibrils observed by this method could be due to the drying of the fibrils inducing the formation of aggregates [47]. The mechanoenzymatic hydrolysis during the CNFs processing leaves long nanoscale fibrils, preserving the native cellulose I crystalline allomorph with partly amorphous regions. Thus, nanofibrils are able to inherently entangle [47]. Such long entangled nanofibers may be feasible for nanoreinforcement percolation in the composite hydrogels.
Biomimetics 2019, 4, 19 7 of 16 the cells in culture medium (500 µL) was added on the hydrogel surface and NIH/3T3 cells were seeded at 200,000 cells per well on average in triplicates. Cells seeded in empty wells (i.e., without the hydrogel) were used as control. Figure 3 shows atomic force microscopy (AFM) images of the CNFs used as nanoreinforcement of the CNF/CHI composite hydrogels. They reveal an entangled network of interconnected nanofibrils with average width of 35.2 ± 8.1 nm and bundles up to 100 nm width. The relative larger width of the fibrils observed by this method could be due to the drying of the fibrils inducing the formation of aggregates [47]. The mechanoenzymatic hydrolysis during the CNFs processing leaves long nanoscale fibrils, preserving the native cellulose I crystalline allomorph with partly amorphous regions. Thus, nanofibrils are able to inherently entangle [47]. Such long entangled nanofibers may be feasible for nanoreinforcement percolation in the composite hydrogels.      Figure 5 shows SEM micrographs of CNF/CHI hydrogels of different compositions which were freeze-dried for observation. The scaffolds revealed a sponge-like network microstructure with interconnected pores for all CNF/CHI concentrations considered in the hydrogel preparation. In all considered compositions, a regular network structure with indistinguishable cellulose nanofibers was observed, which can be due to the good interfacial compatibility of the polysaccharide nanofibers and matrix allowing for good dispersion of the CNFs in the composite. Scanning electron microscopy (SEM)images reveal that with the increase in CHI concentration the porosity decreases. For a given CHI concentration, the increase in CNF content yielded biomaterials of smaller pore sizes. In a given hydrogel sample, observations at the hydrogel outer top surface (Figure 5b,h) (which was directly in contact with the NaOH solution during neutralization) differed from those of surface fractures corresponding to the inner part of the hydrogel (Figure 5a,g). At the outer top surface, a smoother microstructure was observed with a pore size smaller (Figure 5b,h) than in the inner hydrogel ( Figure  5a,g). It could be explained due to the hydrogel processing with a directional diffusion of the base (NaOH) during the neutralization, in which the neutralization front moves from the top to the bottom of the Petri dish mold (Figure 4). Sereni et al. [48] also reported on porosity variations throughout the cross-section of CHI physical hydrogels due to the preparation procedure similar to that used in this work to prepare composite hydrogels.  Figure 5 shows SEM micrographs of CNF/CHI hydrogels of different compositions which were freeze-dried for observation. The scaffolds revealed a sponge-like network microstructure with interconnected pores for all CNF/CHI concentrations considered in the hydrogel preparation. In all considered compositions, a regular network structure with indistinguishable cellulose nanofibers was observed, which can be due to the good interfacial compatibility of the polysaccharide nanofibers and matrix allowing for good dispersion of the CNFs in the composite. Scanning electron microscopy (SEM)images reveal that with the increase in CHI concentration the porosity decreases. For a given CHI concentration, the increase in CNF content yielded biomaterials of smaller pore sizes. In a given hydrogel sample, observations at the hydrogel outer top surface (Figure 5b,h) (which was directly in contact with the NaOH solution during neutralization) differed from those of surface fractures corresponding to the inner part of the hydrogel (Figure 5a,g). At the outer top surface, a smoother microstructure was observed with a pore size smaller (Figure 5b,h) than in the inner hydrogel (Figure 5a,g). It could be explained due to the hydrogel processing with a directional diffusion of the base (NaOH) during the neutralization, in which the neutralization front moves from the top to the bottom of the Petri dish mold (Figure 4). Sereni et al. [48] also reported on porosity variations throughout the cross-section of CHI physical hydrogels due to the preparation procedure similar to that used in this work to prepare composite hydrogels.

Micromechanical Properties of the CNF/CHI Composite Hydrogels
Mechanical properties characterization of the composite hydrogels is shown in Figure 6. Micromechanical studies were performed under controlled RH. Figure 6 shows the impact of the increase of both CHI concentration and CNF content on the hydrogel mechanical properties. Table 1 gives the elastic modulus (E) values estimated from the slope of the tangent comprising the range between the origin and the lowest strain values of the stress-strain curves ( Figure 6). Strain at break values of 45-65% and E between 0.073 and 0.30 MPa were obtained depending on the hydrogel composition. As expected, the increase of the CNF content enhanced the hydrogel mechanical properties. The highest value of E was obtained for composites containing a CHI concentration of 3% (w/w) and a CNF content of 0.4% (w/w). Results were in the range of relatively good mechanical performance for polysaccharide hydrogel biomaterials [21,49].These results indicate that the nanofibers provided reinforcement to the hydrogel matrix of both 2 and 3% (w/w) CHI. The surface of CNFs used in this work was weakly charged with carboxylate moieties displaying a surface charge density of 40-80 mmol/kg [50,51]. During hydrogel processing, in the viscous suspensions electrostatic interactions could be established between CHI (a polycation) and CNFs (a polyanion). Hydrogen bond interactions also could be established between the -OH and -NH functionalities of CHI and the -OH groups of cellulose. Besides, hydrophobic interactions together with H-bonding contribute to the formation of the hydrogel network during the neutralization step. All mentioned interactions contribute to the good compatibility between the nanofibers and the CHI matrix and may allow for stress transfer from the CHI hydrogel matrix to the nanofibers, resulting in improved mechanical properties of the composites. Chitosan chains may absorb on the surface of CNF and play a role in the bridging of nanofibers [41,52,53].

Micromechanical Properties of the CNF/CHI Composite Hydrogels
Mechanical properties characterization of the composite hydrogels is shown in Figure 6. Micromechanical studies were performed under controlled RH. Figure 6 shows the impact of the increase of both CHI concentration and CNF content on the hydrogel mechanical properties. Table 1 gives the elastic modulus (E) values estimated from the slope of the tangent comprising the range between the origin and the lowest strain values of the stress-strain curves ( Figure 6). Strain at break values of 45-65% and E between 0.073 and 0.30 MPa were obtained depending on the hydrogel composition. As expected, the increase of the CNF content enhanced the hydrogel mechanical properties. The highest value of E was obtained for composites containing a CHI concentration of 3% (w/w) and a CNF content of 0.4% (w/w). Results were in the range of relatively good mechanical performance for polysaccharide hydrogel biomaterials [21,49]. These results indicate that the nanofibers provided reinforcement to the hydrogel matrix of both 2 and 3% (w/w) CHI. The surface of CNFs used in this work was weakly charged with carboxylate moieties displaying a surface charge density of 40-80 mmol/kg [50,51]. During hydrogel processing, in the viscous suspensions electrostatic interactions could be established between CHI (a polycation) and CNFs (a polyanion). Hydrogen bond interactions also could be established between the -OH and -NH functionalities of CHI and the -OH groups of cellulose. Besides, hydrophobic interactions together with H-bonding contribute to the formation of the hydrogel network during the neutralization step. All mentioned interactions contribute to the good compatibility between the nanofibers and the CHI matrix and may allow for stress transfer from the CHI hydrogel matrix to the nanofibers, resulting in improved mechanical properties of the composites. Chitosan chains may absorb on the surface of CNF and play a role in the bridging of nanofibers [41,52,53].

Suitability of the CNF/CHI Composite Hydrogels for IVD Repair and Biomechanics Re-Establishment
Examples of optical phase-contrast images of NIH/3T3 fibroblasts after six days of culture on a CNF/CHI hydrogel composite are shown in Figure 7. After 48 h (data not shown), confluent cell spreading was practically achieved on both the neat 2% (w/w) CHI hydrogels and the corresponding CNF/CHI composites. The differences in cellular behavior observed for different CHI concentrations were mainly explained by the increase of stiffness and density (porosity decrease) of the hydrogel when increasing CHI concentration. Higher CHI concentrations led to less accessibility for the cells, reduced adhesion, and longer time was needed for representative spreading. Ata given CHI concentration, for the CNF contents considered in this study (0.2-0.4% (w/w)), similar cell behaviors were observed when using CHI alone or CNF-filled CHI hydrogels. A remarkable aspect was that for CHI concentrations higher than 2% (w/w), as considered in this study, the fibroblasts mostly grew on the hydrogel surface, which hardly revealed a three-dimensional (3D) cell culture appearance (Figure 7a). When the top surface of the hydrogel was cut to produce a fresh surface revealing the inner microstructure of the flat hydrogel (i.e., farther from the original top surface), proliferation of the cells within the hydrogel structure was observed, as revealed in Figure 7b. This relatively low proliferation in the bulk of the hydrogels for CHI concentration higher than 2% could be also related to variations in the density and stiffness throughout the hydrogel cross-section (when moving from the top to the inner part). Actually, in the hydrogel processing, a directional neutralization occurred since the neutralization front moves from the top to the bottom of the Petri dish mold (Figure 4). Sereni et al. [48] reported on porosity variations in the cross-section of CHI hydrogels due to the preparation procedure, which was similar to that used in this study. This effect may also be present in CNF-filled CHI hydrogels, which indeed could influence the fibroblast behavior on the hydrogel.

Suitability of the CNF/CHI Composite Hydrogels for IVD Repair and Biomechanics Re-Establishment
Examples of optical phase-contrast images of NIH/3T3 fibroblasts after six days of culture on a CNF/CHI hydrogel composite are shown in Figure 7. After 48 h (data not shown), confluent cell spreading was practically achieved on both the neat 2% (w/w) CHI hydrogels and the corresponding CNF/CHI composites. The differences in cellular behavior observed for different CHI concentrations were mainly explained by the increase of stiffness and density (porosity decrease) of the hydrogel when increasing CHI concentration. Higher CHI concentrations led to less accessibility for the cells, reduced adhesion, and longer time was needed for representative spreading. Ata given CHI concentration, for the CNF contents considered in this study (0.2-0.4% (w/w)), similar cell behaviors were observed when using CHI alone or CNF-filled CHI hydrogels. A remarkable aspect was that for CHI concentrations higher than 2% (w/w), as considered in this study, the fibroblasts mostly grew on the hydrogel surface, which hardly revealed a three-dimensional (3D) cell culture appearance (Figure 7a). When the top surface of the hydrogel was cut to produce a fresh surface revealing the inner microstructure of the flat hydrogel (i.e., farther from the original top surface), proliferation of the cells within the hydrogel structure was observed, as revealed in Figure 7b. This relatively low proliferation in the bulk of the hydrogels for CHI concentration higher than 2% could be also related to variations in the density and stiffness throughout the hydrogel cross-section (when moving from the top to the inner part). Actually, in the hydrogel processing, a directional neutralization occurred since the neutralization front moves from the top to the bottom of the Petri dish mold (Figure 4). Sereni et al. [48] reported on porosity variations in the cross-section of CHI hydrogels due to the preparation procedure, which was similar to that used in this study. This effect may also be present in CNF-filled CHI hydrogels, which indeed could influence the fibroblast behavior on the hydrogel. Using CNF/CHI hydrogel patches for the repair of the AF disc region is justified by the reported modeling of IVD tissue, in which the fibrous network of AF plays a critical role in the IVD biomechanics [54]. A rigorous material definition maybe achieved to incorporate the role of fiber- Using CNF/CHI hydrogel patches for the repair of the AF disc region is justified by the reported modeling of IVD tissue, in which the fibrous network of AF plays a critical role in the IVD biomechanics [54]. A rigorous material definition maybe achieved to incorporate the role of fiber-matrix interactions in the complex IVD tissue. The design of hydrogel composite implants needs to reach adequate mechanics and relative long-term performance as well as fix onto the tissue defect (e.g., by adhesion or suture). The design of a bioactive hydrogel to seal, reinforce, and repair the AF, as proposed here, may ensure appropriate long-term tissue regeneration, while providing a mechanical support to avoid post-surgery hernia reoccurrence or compressive scar. We studied the input of the engineered CNF/CHI composite hydrogels on the biomechanics of the IVD. The hydrogel composition with the best mechanical properties achieved in the microtensile testing results (3% CHI/0.4% CNF) was selected for implantation studies. Figure 8 shows the resulting nonlinear stress-strain relationship of the motion segments' mechanical behavior. The curves of the healthy, damaged, and hydrogel-implanted discs showed an initial neutral zone (NZ), where the discs undergo relatively large deformation under low applied force. Then, the elastic zone (EZ) was reached. This linear EZ corresponds to the physiological loading range (Figure 8, grey shadow) with stresses reported to be 0.6-3.8 MPa, related to body weight or exerted muscle forces during bending and torsion (0.6-1.3 MPa) [5], and typical or extreme daily loadings like object lifting and vibrating vehicle driving (2.5-3.8 MPa) [7,55,56]. The EZ of the healthy disc revealed the physiological loading region and a pronounced slope. The positive input of the hydrogel implantation in AF defect on disc biomechanics is also shown. While both curves of the damaged and hydrogel-treated discs revealed an EZ within the physiological loading range, the linear EZ of the curve of the hydrogel-treated disc revealed a higher slope similar to the healthy disc. Moreover, the stress at failure of the healthy disc, hydrogel-implanted disc, and the untreated damaged disc (6.7, 5.9, and 5.0 MPa, respectively) demonstrates the beneficial role of hydrogel composite implantation to re-establish IVD biomechanics. matrix interactions in the complex IVD tissue. The design of hydrogel composite implants needs to reach adequate mechanics and relative long-term performance as well as fix onto the tissue defect (e.g., by adhesion or suture). The design of a bioactive hydrogel to seal, reinforce, and repair the AF, as proposed here, may ensure appropriate long-term tissue regeneration, while providing a mechanical support to avoid post-surgery hernia reoccurrence or compressive scar. We studied the input of the engineered CNF/CHI composite hydrogels on the biomechanics of the IVD. The hydrogel composition with the best mechanical properties achieved in the microtensile testing results (3% CHI/0.4% CNF) was selected for implantation studies. Figure 8 shows the resulting nonlinear stressstrain relationship of the motion segments' mechanical behavior. The curves of the healthy, damaged, and hydrogel-implanted discs showed an initial neutral zone (NZ), where the discs undergo relatively large deformation under low applied force. Then, the elastic zone (EZ) was reached. This linear EZ corresponds to the physiological loading range (Figure 8, grey shadow) with stresses reported to be 0.6-3.8 MPa, related to body weight or exerted muscle forces during bending and torsion (0.6-1.3 MPa) [5], and typical or extreme daily loadings like object lifting and vibrating vehicle driving (2.5-3.8 MPa) [7,55,56]. The EZ of the healthy disc revealed the physiological loading region and a pronounced slope. The positive input of the hydrogel implantation in AF defect on disc biomechanics is also shown. While both curves of the damaged and hydrogel-treated discs revealed an EZ within the physiological loading range, the linear EZ of the curve of the hydrogel-treated disc revealed a higher slope similar to the healthy disc. Moreover, the stress at failure of the healthy disc, hydrogel-implanted disc, and the untreated damaged disc (6.7, 5.9, and 5.0 MPa, respectively) demonstrates the beneficial role of hydrogel composite implantation to re-establish IVD biomechanics. Figure 8. (a) Examples of healthy, damaged, and hydrogel-implanted discs for which spinal motion segments, consisting of two vertebral endplates and the disc joining them together, biomechanical testing were performed. (b) Compression stress-strain curves of healthy, damaged, and hydrogelimplanted intervertebral disc. For the latter, the intervertebral disc annulus fibrosus was repaired with a 3% CHI/0.4% CNF composite hydrogel patch. Grey shadow: Physiological loading range. The range of motion (ROM)-the displacement at the largest applied load-of the hydrogel-implanted disc approached more that of the healthy disc, than did the damaged disc. In contrast to the damaged disc, the shape of the curve of the hydrogel-implanted disc was much closer to that of the healthy disc. Similarly to what is observed for the healthy disc, the hydrogel-implanted disc revealed a plastic deformation range after the EZ. The dotted curve in the Figure 8 corresponds to a hydrogel-implanted disc, which initially reached complete failure, but was retreated with a new hydrogel patch in the AF defect and injected with 120 µL of CNF/CHI viscous formulation (sol state) in the nucleus [41]. With this combination of hydrogel implant and injectable treatment, a stress of 2.2 MPa, still in the range of physiological loading, could be achieved. Pictures of discs subjected to this treatment combination after complete failure are shown in Figure 9. In a first disc, when the failure was achieved for the second time, the integrity of the disc was preserved at the implanted side (Figure 9b), revealing the ability of the hydrogel implant to withstand any fluid leak during nucleus prolapse. In a second disc, the implanted hydrogel could not withstand fluid leak during nucleus prolapse (Figure 9c). Noticeably, the images in Figure 9 correspond to the appearance after mechanical testing of motion segments which already achieved failure in a previous compression testing. The range of motion (ROM)-the displacement at the largest applied load-of the hydrogelimplanted disc approached more that of the healthy disc, than did the damaged disc. In contrast to the damaged disc, the shape of the curve of the hydrogel-implanted disc was much closer to that of the healthy disc. Similarly to what is observed for the healthy disc, the hydrogel-implanted disc revealed a plastic deformation range after the EZ. The dotted curve in the Figure 8 corresponds to a hydrogel-implanted disc, which initially reached complete failure, but was retreated with a new hydrogel patch in the AF defect and injected with 120 µL of CNF/CHI viscous formulation (sol state) in the nucleus [41]. With this combination of hydrogel implant and injectable treatment, a stress of 2.2 MPa, still in the range of physiological loading, could be achieved. Pictures of discs subjected to this treatment combination after complete failure are shown in Figure 9. In a first disc, when the failure was achieved for the second time, the integrity of the disc was preserved at the implanted side (Figure 9b), revealing the ability of the hydrogel implant to withstand any fluid leak during nucleus prolapse. In a second disc, the implanted hydrogel could not withstand fluid leak during nucleus prolapse (Figure 9c). Noticeably, the images in Figure 9 correspond to the appearance after mechanical testing of motion segments which already achieved failure in a previous compression testing. Figure 9. Appearance of discs after complete failure, which were subject to a treatment combination of CNF/CHI hydrogel ("gel state") implantation in the disc annulus fibrosus region ( Figure 2) and (a) injection of CNF/CHI viscous suspension ("sol state") in the nucleus pulposus region. (b) Appearance after failure in a second mechanical testing, after already achieved failure in a previous compression test ( Figure 8); the integrity of the disc was preserved at the hydrogel-implanted disc side. (c) The implanted hydrogel could not withstand a fluid leak during nucleus prolapse.
Research on hydrogel biomaterial therapies for IVD tissue engineering is facing many challenges. Our designed materials appear compatible with early mechanical stimulation, which also appears to affect the behavior of chondrocytes and thus the self-regeneration of cartilaginous tissues [54]. A proportional relationship was found between compression of a chondrocyte-loaded tissue engineered biomaterial and the culture time-dependent increase of stiffness of the biomaterial [54,57,58].

Conclusions
Cellulose nanofiber-filled chitosan composite hydrogels were developed and structurally and mechanically characterized. The addition of CNF contributes to the improvement of the mechanical properties of CHI hydrogel matrices. From ex vivo experiments performed in pig vertebral unit models, we showed that the implantation of CNF-reinforced CHI hydrogels in AF disc defects contributes to the restoration of the disc biomechanics by approaching the functionality of a healthy disc. Furthermore, these hydrogel composites can serve as contention patches against nucleus protrusion and also as hydrogel biomaterial to support cell growth in a 3D environment, and finally show characteristics and properties which are promising for IVD tissue engineering.  . Appearance of discs after complete failure, which were subject to a treatment combination of CNF/CHI hydrogel ("gel state") implantation in the disc annulus fibrosus region ( Figure 2) and (a) injection of CNF/CHI viscous suspension ("sol state") in the nucleus pulposus region. (b) Appearance after failure in a second mechanical testing, after already achieved failure in a previous compression test ( Figure 8); the integrity of the disc was preserved at the hydrogel-implanted disc side. (c) The implanted hydrogel could not withstand a fluid leak during nucleus prolapse.
Research on hydrogel biomaterial therapies for IVD tissue engineering is facing many challenges. Our designed materials appear compatible with early mechanical stimulation, which also appears to affect the behavior of chondrocytes and thus the self-regeneration of cartilaginous tissues [54]. A proportional relationship was found between compression of a chondrocyte-loaded tissue engineered biomaterial and the culture time-dependent increase of stiffness of the biomaterial [54,57,58].

Conclusions
Cellulose nanofiber-filled chitosan composite hydrogels were developed and structurally and mechanically characterized. The addition of CNF contributes to the improvement of the mechanical properties of CHI hydrogel matrices. From ex vivo experiments performed in pig vertebral unit models, we showed that the implantation of CNF-reinforced CHI hydrogels in AF disc defects contributes to the restoration of the disc biomechanics by approaching the functionality of a healthy disc. Furthermore, these hydrogel composites can serve as contention patches against nucleus protrusion and also as hydrogel biomaterial to support cell growth in a 3D environment, and finally show characteristics and properties which are promising for IVD tissue engineering.