3D collagen-based scaffolds used in this study have been produced by Fin-Ceramica Faenza SpA (Faenza, Italy). They have a cylindrical shape, with an 8 mm diameter and 5 mm height, consisting of equine type I Collagen gel (1 wt%) supplied in aqueous acetic buffer solution (pH = 3.5) (Opocrin SpA, Modena, Italy). The process of fabrication, as well as the chemical and physical characterization and tolerability have been described previously (Calabrese et al., 2017a). Briefly, collagen gel was softly dilute in sterilized water and precipitated in fibers by drop-wise addition of 0.1 M NaOH solution up to the isoelectric point (pH = 5.5). A crosslinking reaction was performed by 48 h-long immersion of the agglomerated fibers in NaHCO3/Na2CO3 (Sigma Aldrich and Merck Millipore) aqueous solution with a 1,4-butanediol diglycidyl ether (BDDGE) solution at 37°C to maintain scaffold structure. Then, agglomerated fibers were freeze-dried for 25 h under vacuum conditions (P = 0.29 mbar) to obtain a porous 3D structure. Finally, scaffolds were gamma-sterilized at 25 kGy. The microstructural and morphological characterization of scaffold was assessed by Scanning Electron Microscopy (SEM) by using a SEM-LEO 438 VP (Carl Zeiss AG, Oberkochen, Germany). The samples were sputter-coated with gold before analysis.

Female mice (n = 44) (BALB/cOlaHsd, 6 weeks aged, weight: 17–22 g; Harlan Laboratories) were used. Animal care and handling were performed according to EU Directive 2010/63/EU and the Italian law (D.Lgs. 26/2014). All experiments involving animals have been approved by the Italian Ministry of Health. Animals were housed in groups of four in independently ventilated cages (15 changes/hour of filtered air), with ad libitum access to water and food (Teklad rodent diet, Harlan Laboratories, San Pietro al Natisone, Italy), with standard conditions of temperature (22 ± 2°C) and relative humidity (50 ± 5%) and a light/dark cycle of 12/12 h.

Surgery was performed under aseptic conditions, maintaining mice under gas anesthesia (isoflurane). All efforts were made to minimize the number of animals used and their suffering. Pre- and post-grafting procedures were performed as explained previously (Calabrese et al., 2016b, 2017b). Briefly, surgical procedures were performed under aseptic conditions, with the animals under gas anesthesia (isoflurane). One collagen type-I scaffold/animal was implanted into a subcutaneous pocket in the dorsum of mouse. The transplanted mice were randomly divided in five groups: 1 week (n = 8), 2 week (n = 8), 4 week (n = 8), 8 week (n = 8), 16 week (n = 8), and finally sacrificed by intracardiac injection of Tanax (MSD Animal Health Srl, Segrate, Italy) under deep anesthesia (isoflurane). Four untreated animals were used as negative controls for in vivo imaging analysis. The scaffolds were explanted to perform ex-vivo analyses.

In order of assessing the occurrence of angiogenesis within the scaffold in a time-course manner, animals were evaluated by in vivo FMT (FMT 2500, Perkin Elmer, Monza, Italy). Specifically, all animals received an injection of 100 μl of AngioSense 680EX (Perkin Elmer, Monza, Italy) into the tail vein. This fluorescent probe specifically binds to endothelial cells. Twenty-four hours after the probe injection, FMT images were acquired. During the imaging, mice were maintained under isoflurane anesthesia. Acquisition and analysis of FMT images were assessed by using the TrueQuant software (Perkin Elmer, Monza, Italy). For quantification, the region of interest (ROI) was selected and the extent of angiogenesis was analyzed by measuring the amount of fluorescence probe (in pmol) into the ROI after choosing a concentration threshold. This threshold has been determined by keeping the volume of ROI constant (50 mm³). Animals were sacrificed by decapitation under deep anesthesia immediately after imaging.

Immediately after animal sacrifice, scaffolds were collected and fixed for 2 h in 4% paraformaldehyde, dehydrated, embedded in paraffin and cut into 3 μm-thick sections. Sections were mounted on slides and treated for immunohistochemical analysis, as described previously (Calabrese et al., 2016a,b, 2017a,b). For immunofluorescence analysis the following rabbit polyclonal primary antibodies (LSBio, Seattle, WA, USA) were used: anti-Aggrecan (dilution: 1:150), anti-Matrilin-1 (1:200), anti-Sox9 (1:200), anti-type II Collagen (1:200). AlexaFluor anti-rabbit 568 was used as secondary antibody (dilution: 1:2000, Life Technologies) and nuclei were counterstained with DAPI (dilution: 1:10.000). Control of immunostaining specificity was assessed without the primary antibody. For immunohistochemistry analysis the mouse anti-CD15 (dilution: 1:100) monoclonal antibody (Abcam, Italy) was used. Washed sections were treated with biotinylated anti-mouse and anti-rabbit immunoglobulins and then incubated in streptavidin conjugated with peroxidase (Dako, Carpintera, Calif., USA). Staining was detected using diaminobenzidine (DAB). Sections were counterstained with hematoxylin, then dehydrated and mounted in permanent non-aqueous mounting medium.

Cell counts have been performed by using ImageJ software (NIH, USA). Data were analyzed as percentage of positive cells on the total number of DAPI stained cells. Three sections/animal have been used for counting.

For histological analysis alternate sections were also stained with Hematoxylin and Eosin (H&E), Alcian Blue (Panreac, Castellar del Valles, Barcelona, Spain) and Safranin O/Fast Green.

H&E staining was performed as previously reported (Calabrese et al., 2016a,b, 2017b). Alcian Blue staining was assessed according to the manufacturer protocol. The stained slides were examined by using a Leica DMI 4000B fluorescence microscope. A semi-quantitative evaluation of the formation of cartilage matrix has been performed by the measurement of optical density (OD) in gray scale images of the Alcian Blue stained specimens. Three images/animal have been randomly selected for OD quantification and measurements have been performed by using ImageJ.

For the staining a Safranin O solution was prepared according to manufacturer protocol. Sections were previously stained with Hematoxilin and Fast green, washed and incubated in 0.1% Safranin O solution for 5 min. Successively, slides were washed several times to get rid of the excess of the staining solution and, finally, mounted.

Total RNA was isolated from scaffolds by using TRIzol reagent (Invitrogen Inc.). Reverse transcription of total RNA was carried out using High-Capacity cDNA Reverse Transcription Kit (Applied Biosystems). The transcriptional analysis of chondral genes, including COL1A1, COL2A1 e Aggrecan, was performed using SYBR Green method on a StepOne System Real Time PCR (Applied Biosystems). Primers for selected murine mRNA were designed using primer blast39 using exon-exon junctions on messengers as target region. The mRNA level of glyceraldehyde-3-phosphate dehydrogenase (GAPDH) and beta-tubulin (TuBB4a) were used as endogenous controls.

Primers sequences:

COL1A1_Forward: CCGGAAACAGACAAGCAACCCAAA;

Data were analyzed either as raw data or as mean ± SEM, as appropriate. Differences between groups were evaluated by using one-way ANOVA, followed by Tukey's HSD post hoc test. For all experiments, a P < 0.05 was considered to be significant. All analyses were assessed by means of Systat 11 (Systat Software, USA).

The morphological and microstructural features of scaffold were assessed by SEM analysis. Figure 1 reports SEM images of the scaffold at two different magnifications, 50x (Figure 1a) and 150x (Figure 1b). SEM images reveal an elevated porosity with three-dimensionally intersected pores without any precise alignment of the collagen fibers. This specific structure can be useful both for functions and for guiding cells during migration and proliferation.

The results of FMT analysis have demonstrated the presence of new vascularization within the implanted scaffolds, in vivo. Specifically, after intravenous injection of AngioSense 680, a strong fluorescent signal appeared in the anatomical area where the scaffold implant was placed. This signal was higher at the earlier time-point (1–4 weeks; Figures 2b–d), with a decreasing trend along the longer time-points (8–16 weeks; Figures 2e,f), although this decrease was not statistically significant (P > 0.05; Figure 2a).

To assess the biocompatibility and capability of the new biomimetic 3D scaffold to promote augmentation of the cartilage-like tissue after implantation in vivo, we performed several histological analyses on the explanted scaffolds at 1, 2, 4, 8, and 16 weeks after implantation.

H&E staining confirmed the tolerability of the scaffold, already shown previously (Calabrese et al., 2016b), as revealed by the lack of necrotic areas and the slight and transitory inflammatory reaction seen sometimes in the explanted scaffolds (Figure 3). In fact, only at the first 2 weeks after implantation it was possible to see a small number of granulocytes, which are specific elements of inflammation (Figures 3a–f) that disappeared over time (Figures 3g–o).

H&E staining also showed the capacity of the scaffold to promote the formation of new chondral tissue, as demonstrated by the capacity of the biomaterial to recruit host cells that infiltrate, adhere and grow into the scaffold. Although host cells were confined in the surface during the first 2 weeks (Figures 3b,e), they started to infiltrate also the inner part of the scaffold at week 4 (Figures 3g–o). Furthermore, H&E staining showed a gradual increase of cartilage-like tissue within the pores of the scaffold starting at 4 weeks after implantation, as demonstrated by the presence of chondromyxoid-like matrix (Figures 3h,k,n). Interestingly, chondrocyte-like cells and pericellular lacunae, which are typical cartilage structures, were visible at 8 and 16 weeks after implantation (Figures 3k,l,n, arrows).

H&E staining also confirmed the FMT findings concerning the new vascularization of the scaffold by displaying the presence of structures resembling blood vessels (insert in Figure 3o).

These data were further sustained by the Alcian Blue staining (Figure 4), which revealed the presence of spindled and stellate mesenchymal cells as early as 2 weeks after implantation (Figure 4b). Then, the amount of extracellular matrix increases gradually (Figures 4c–d) and became strong at 16 weeks (Figure 4e), confirming the cartilage-like feature of the newly formed tissue. OD quantification confirms the gradual increase of cartilage-like matrix (ANOVA: P < 0.05; Figure 4a) that became significantly higher at 16 weeks, compared to average levels at 1 week (Tukey's post hoc test: P < 0.05, Figure 4a). To further reveal the cartilaginous nature of the freshly produced tissue within the scaffold, we assessed a Safranin O/ Fast green staining (Figures 5a–f). The red staining showed the GAGs of newly formed cartilage, while the green staining displayed collagens. In accordance with Alcian Blue, Safranin O/Fast green staining displayed a gradual increase of cartilage-like matrix overtime that became markedly more evident at 8 and 16 weeks (Figures 5e,f).

Moreover, to further confirm both the biocompatibility and the capacity of biomimetic 3D scaffold to promote chondrogenesis, we performed immunohistochemical analysis using specific markers of inflammation and cartilage, such as CD15 and Sox9, type-II Collagen, Aggrecan and Matrilin-1 respectively, on histological sections obtained at 1, 2, 4, 8, and 16 weeks post-grafting (Figures 6, 7).

In accordance with H&E staining, immunohistochemical analysis of CD15, a granulocytes marker, showed a signal only during the first 2 weeks, (Figures 6a,b enlarged in Figures 6f,g) after implantation, confirming the presence of a temporary inflammatory process that missing over time (Figures 6 c–e enlarged in Figures 6h–j).

A weak expression of Sox9 was seen during the first 2 weeks (Figures 7Aa,b,B) followed by a significant increase at weeks 4 and 8 (Figures 7Ac,d,B; ANOVA: P < 0.05; Tukey's HSD post hoc test: P < 0.05) and an evident reduction at week 16 (Figures 7Ae,B; Tukey's HSD post hoc test: P < 0.05). Type-II Collagen displayed a very weak staining during the first 2 weeks (Figures 7Af,g,B), but a more robust signal appeared after 4 weeks (Figures 7Ah–j,B; ANOVA: P < 0.05; Tukey's HSD post hoc test: P < 0.05) and increased further at 8 and 16 weeks (Tukey's HSD post hoc test: P < 0.05). The expression of Aggrecan was very low during the first 2 weeks (Figures 7Ak,l,B), and then it reaches a peak at week 4 (Figure 7B; ANOVA: P < 0.05; Tukey's HSD post hoc test: P < 0.05) and then significantly and progressively decreased at weeks 8 and 16 (Figures 5b, 7Am–o; Tukey's HSD post hoc test: P < 0.05). As evident in Figures 7Ap–t,B, the Matrilin-1 expression pattern was similar to that of Aggrecan.

RT-qPCR analysis has been also assessed on total RNA isolated from explanted scaffolds to quantify the expression of target genes related to chondral phenotype. The expression levels of chondrogenic genes, such as Type I Collagen (COL1A1), Type II Collagen (COL2A1), Aggrecan (ACAN) and Sox9, at 2, 4, 8, and 16 weeks have been compared to mRNA levels at week 1. Logarithmic RQ values are reported in Figure 8. COL1A1 shows an increased expression of almost three orders of magnitude during the second week that increased up to seven to 16 weeks. COL2A1 shows a distinctive peak of expression at week 4 (14 order of magnitude) with a subsequent decrease and reduced modulation (6 order of magnitude) until week 16. Sox9 mRNA concentration increases during the whole period with a massive regulation up to 3,5 order of positive modulation during the 16th week. ACAN shows a smallest expression during week 2 with peaks of positive regulation during the 4th and 8th week and a subsequent decrease modulation to week 16.

The authors declare that the research was conducted in the absence of any commercial or financial relationships that could be construed as a potential conflict of interest.