Chitosan powder (ChitoClear™ TM4030, DD of 75%; MW 50 kDa, Primex Ehf, Iceland) was dispersed at the concentration of 6% w/v into a 0.02% w/v aqueous solution of sodium hyaluronate (HA) (Halien, High MW, ACME s.r.l., Cavriago, Italy, batch Ha16003), magnetically stirred until homogenization (5 s) and made up to volume dropwise with glacial acetic acid (Sigma-Aldrich) until a final concentration of 2% v/v. Complete dissolution of chitosan was achieved after overnight magnetic stirring at 24°C. D-(+)-raffinose pentahydrate (Sigma-Aldrich, USA) was added at a concentration of 290 mM as viscosity agent (47) and dissolved by magnetic stirring. The same procedure was adopted to prepare the chitosan-based solution free of hyaluronic acid: in this case, dissolution of chitosan took place in a 2% v/v acetic acid solution in UltraPure water (Purelab Flex 1 system, ELGA Veolia) to which D-(+)-raffinose was added at 290 mM. Polymeric blends were stored at 4°C until use.

The CAD model for 3D printing of scaffolds was designed by means of Solidworks™ software (Dassault Systems, USA) by drawing a grid composed of overlapping parallel filaments set at a nominal distance of 200 μm. Total sizes of the object, formed by five layers, were also set, 1.5 × 1.5 × 0.1 cm (width/length/thickness). The file produced was then converted through the software Slic3r™ (RepRap) from the generated. stl format (STereo Lithography interface format) to the machine code (gcode format) readable by the in-house-built 3D printer. Briefly, the platform physically supporting the 3D construction consisted of a 0.6-mm-thick stainless-steel plate coated with a 0.6-mm chitosan or chitosan–HA film casted prior to printing, which was fixed on the Peltier cells of the 3D machinery and instantly frozen at temperature of −18°C: this film constituted the base for the scaffold. Polymeric blends were loaded into 5-ml syringes equipped with a blunt 26 G needle (internal ø: 0.260 mm), which were fixed on the machinery's robotic arms and connected to the pressure-assisted mechanical piston (extruder). Solutions were deposited by the extruding apparatus moving along x-, y-, and z-axis on the top of the base film, and instantly frozen at a constant temperature of −18°C exploiting freeze deposition layer by layer to give shape to the polymeric blend to obtain the 3D object designed, according to a method recently described (48). At the end of each manufacturing process, frozen scaffolds underwent ionotropic gelation by exposure to saturated ammonia vapors for 1 h at ambient temperature, following the technique reported in a previous work dealing with the settlement of suitable gelation processes for 3D chitosan hydrogels (45). Scaffolds were then washed by immersion in 80 ml of UltraPure water until neutral pH (four times for 10 min) to eliminate ammonia residues and raffinose (47). Finally, scaffolds were punched round to fit into 24-well plates, sterilized by immersion in 70% w/v ethanol, transferred to sterile phosphate buffered saline (PBS) and preserved at 4°C until use.

The mechanical resistance of scaffolds obtained by different gelation media was compared on 20-layer scaffolds of 5 × 1.5 cm. Hydrogels at the maximum swollen state were tested. Thickness was determined as a mean of six measurements of the scaffold performed with a digital micrometer (Mitutoyo, Japan). Each scaffold was fixed on a tensile tester (AG M1 Acquati, Italy) loaded with a 5-daN cell. Force and time signals were digitalized by means of a PowerLab 400 board and recorded with Scope 3.5 software. Elongation at break (% strain) and Young's modulus were determined from the relevant stress–strain curves, taking into consideration the linear portion and normalized by thickness (45).

The donor horses were referred to the Veterinary Teaching Hospital of the University of Parma and were examined by a veterinarian to assess their clinical history and to exclude systemic and hematological diseases. Blood was collected and placed into tubes supplemented with sodium citrate (0.38% final concentration). The citrated blood was centrifuged in a standard laboratory centrifuge (swinging rotor) for 20 min at 180 × g. Afterwards, the portion of plasma enriched with platelets was transferred to another tube and platelets were pelleted in a second centrifugation step for 15 min at 900 × g. The pellet was resuspended so as to obtain a concentration of 1 × 10⁹ platelets per 1 ml. The PRP was subjected to a double freezing cycle at −80°C and thawing at −37°C. To remove platelet remnants derived from freezing–thawing cycles, a further centrifugation was performed (Microfuge 22R Centrifuge) for 20 min at 11,000 × g. PL supernatant was collected and stored at −80°C until use (39).

Horses ranging from 5 to 8 years old and farmed for human consumption were slaughtered and the metacarpophalangeal and metatarsophalangeal joints were immediately delivered to our laboratory for collection of cartilage. Before cartilage harvesting, joints were carefully examined and those with macroscopic lesions related to overt OA or with evidence of synovitis were excluded from the study. The articular cartilage was obtained from three joints/experiment; chondrocytes were isolated separately from each cartilage and then checked for viability and the samples with viability <95% were excluded. Then, three samples were pooled for each experiment and three experiments were done, with three replications/experiment.

Chondrocytes were isolated from tissue following a protocol previously described (39). Briefly, cartilage was finely diced under sterile conditions and then washed several times in PBS. After pre-incubation in 0.1% pronase (Sigma) solution for 1 h at 37°C, the tissue was treated with 0.2% collagenase type IA (Sigma) in D-MEM (4.5 g/L glucose; 25 mM Hepes) for 2 h at 37°C. The digested material was filtered through 100-μm and 20-μm nylon filters and the cellular suspension was centrifuged at 1,800 rpm for 10 min. The supernatant was discarded, and the pellet was washed several times with D-MEM containing 10% fetal calf serum (FBS), 100 U/ml penicillin and 0.1 mg/ml streptomycin. The number of chondrocytes was determined using a hemocytometer and the cell viability (never <95%) was assessed by trypan blue (0.1%) exclusion. The chondrocytes were seeded into 3D chitosan scaffolds (CH) with or without hyaluronic acid (CH+HA) at a density of 8 × 10⁵ cells/scaffold.

The culturing medium contained 10% FBS or 5% PL or nothing (–FBS/–PL). All cultures were incubated at 37°C in a humidified atmosphere of 5% CO₂ in air for the time specified for each single experiment and the medium was exchanged biweekly.

The adhesion, colonization, and morphological characteristics of the cells in 3D systems was histologically evaluated after 2 weeks and 4 weeks of culture.

At the end of experimental time, scaffolds were fixed with 4% paraformaldehyde (PFA) in PBS for 40 min at room temperature. After fixation, the scaffolds were washed to eliminate any residues of PFA and then pre-embedded in 100 μl of 2% agarose solution in distilled water to preserve the 3D culture and facilitate scaffold handling. The samples were then dehydrated in increasing alcohol concentrations, clarified in xylene, and paraffin embedded. Histological sections 5 μm thick were obtained using a rotary microtome (Slee Cut 6062, Slee Medical, Mainz, Germany). Sections were stained with hematoxylin–eosin (H&E) and Masson's trichrome (with Aniline-Blue). Microphotographs were acquired with Nikon Digital Sight System.

Immunohistochemical analysis was performed at 2 and 4 weeks of culture in order to determine the positivity of ECM to collagen type II. Sections were treated with 2% Hyaluronidase (Sigma) in PBS for 30 min at 37°C. Endogenous peroxidases were quenched with 3% H₂O₂ for 12 min at room temperature and then aspecific antigens were blocked with 1% bovine serum albumin (BSA) (Sigma) in PBS for 1 h at room temperature. The slides were then incubated overnight at a 4°C with primary monoclonal antibody anti-collagen II (1:100) (orb 156420; Byorbit). Then, secondary biotinylated anti-rabbit antibody (BA-100, Vector Laboratories) was applied for 1 h at room temperature. Slides were incubated with an ABC Kit (PK7800, Vector laboratories), and then with 3,3′-diaminobenzidine (DAB) for 3 min, counterstained with Mayer's hematoxylin and then dehydrated in ascending scale of alcohols and permanently mounted.

Slides were then digitalized and collagen type II was quantified using NIS-elements AR (Nikon, Japan) on three (per case) digital images at 10 × magnification. The total number of cells was obtained identifying the nuclei through a nuclear threshold-based mask; this mask was converted into regions of interest (ROIs) (each nucleus represents a single counting ROI). In order to include cytoplasm and pericellular space, the mask was expanded for three layers through a mathematical morphology-thickening function (NIS–element AR, Nikon, Japan). The number of positive cells [(N of collagen Type II positive cells/total number of cells) * 100] was obtained overlaying the ROIs with a threshold based on collagen type II signal and considering only ROIs with positive signaling.

The scanning electron microscopy (SEM) of cells grown on chitosan scaffolds was performed after 2 weeks and 4 weeks of culture (49, 50). Scaffolds seeded with cells were fixed at each selected time point with 2.5% glutaraldehyde in 0.1 M sodium cacodylate buffer (pH 7.3) for 2 h at 4°C. They were then dehydrated through a series of increasing grades of ethanol (from 50% v until absolute) and then critical point dried with liquid carbon dioxide (CPD 030 Baltec, Wallruf, Germany). Specimens were then sputter-coated (Balzers device) with gold-palladium (Plano, Germany) using a SCD 040 coating device (Balzer Union, Wallruf, Germany). The samples were observed using a Philips 501 SEM scanning electron microscope at an accelerating voltage of 5 kV. The chitosan scaffolds without cells were used as control samples (blank).

The experimental results were presented as the mean ± SD. In each experiment, each treatment was performed with three replicate culture wells. Statistical differences were calculated with multifactorial ANOVA using the SPSS 21.0 for Window (IBM® SPSS® Statistics, IBM, New York, USA). When significant differences were found, means were compared by Scheffe's F test.

The structure of the 3D CH scaffold consisted of an upper part of a regular network of CH fibers forming pores of approximately 200 μm (Figure 1A) and a basal layer of CH film (Figure 1D) to hold cells and prevent their adhesion to the plastic plate.

Mechanical resistance was tested for both types of scaffold, resulting in 128 kPa ± 21 kPa (% strain at rupture 52.3 ± 5.3%) for the chitosan scaffold and 156 kPa ± 25 kPa (% strain at rupture 50.6 ± 10.9%) for the chitosan + hyaluronic acid scaffold, without any statistically significant difference. Furthermore, the scaffolds retained their structure for up to 6 months when stored in water or saline solution: no statistically significant alteration of the structure was observed. When immersed in culture media, all scaffolds retained their original dimensions for up to 21 days.

After 2 weeks of culture in both types of scaffolds (Figures 1B,C), chondrocytes colonized the scaffold reticulate. A few cells settled in the spaces within the reticulate, whereas others adhered to the surface of fibers of scaffold.

The chondrocytes penetrated inside the 3D structure and colonized the spaces between the fibers (Figure 1E). The chondrocytes used the spaces as niches for proliferation and growth by organizing in clusters, which became larger during culture (Figure 1F, black arrowhead) or remained as single cells adhering to the surface of the scaffold fibers (Figure 1F, white arrowhead).

After 2 weeks, the chondrocytes cultured with FBS colonized the pores of the scaffold's reticulate texture. However, these appeared with an evident double cell morphology, both roundish cells and fibroblast-like cells (Figure 2A, white arrowheads and white arrows). When the scaffold was prepared with HA, roundish chondrocytes became predominant and a little production of filamentous matrix became evident (Figure 2B, white stars). In the presence of PL, the chondrocytes had a roundish morphology, and the presence of filaments compatible with matrix production was apparent in both types of scaffolds (Figures 2C,D, white stars).

After 4 weeks of culture with FBS, the chondrocytes colonized the pores of the scaffold's reticulate texture. Moreover, these appeared to be more numerous in CH scaffolds both with and without HA enrichment, compared with that for PL treatment. In the FBS condition, the chondrocytes maintained the double morphological features, and a few fibroblast-like cells became evident in the scaffold prepared with HA (Figure 3B, white arrows). Furthermore, the cells began to produce a few matrix filaments (Figures 3A,B, white stars).

After 4 weeks of PL supplementation, the cell population increased and the chondrocytes that covered the scaffold texture were more numerous, compared with the scenario after 2 weeks. In this culture condition, the chondrocytes showed a round morphology and a significant amount of filamentous matrix production (Figures 3C,D, white stars) in both types of scaffolds.

The production of ECM by the chondrocytes was evaluated using Masson's trichrome stain. It highlighted the collagen-based ECM in blue, whereas the scaffold structure and cells were stained in dark red.

After 2 weeks of culture, matrix production was detected under all the conditions (Figure 4). However, this was more evident in the CH scaffolds with HA than in those without HA, even in the absence of FBS/PL supplementation (Figure 4B). Furthermore, the presence of PL increased matrix production in both types of scaffolds (Figures 4E,F).

After 4 weeks of culture, the cells seeded on both types of scaffolds (Figure 4, right panel) continued to present a roundish and regular morphology comparable to that observed at 2 weeks. However, higher cell density was observed in certain areas of the scaffolds in the presence of FBS and PL (Figures 4I,L–N).

The matrix production was increased compared with that at 2 weeks under all the conditions. This was particularly so in the CH scaffolds with HA without FBS/PL supplementation (Figure 4H).

Collagen type II was used as a marker of chondrocyte ECM production. The expression of collagen type II after 2 weeks of culture (Figure 5, left panel) was low in the scaffolds without HA under all the tested culturing conditions (Figures 5A,C,E,O). A marginal more evident collagen II positivity was observed only in the PL supplementation condition. Rather, CH scaffolds with HA showed ECM production characterized by expression of collagen type II, particularly in FBS and PL supplementation conditions (Figures 5D,F,O).

After 4 weeks (Figure 5, right panel), collagen type II positivity was also evident in the ECM of the CH scaffolds without HA (Figures 5I,M,O) and remained higher in the CH scaffold with HA (Figures 5H,I,N,O). Without FBS/PL supplementation, the collagen type II expression remained at baseline (Figures 5G,H,O).