The development of a nucleus pulposus-derived cartilage analog scaffold for chondral repair and regeneration

Abstract

Focal chondral defects (FCDs) significantly impede quality of life for patients and impose severe economic costs on society. One of the most promising treatment options-autologous matrix-induced chondrogenesis (AMIC)-could benefit from a scaffold that contains both of the primary cartilage matrix components-sulfated glycosaminoglycans (sGAGs) and collagen type II. Here, 17 different protocols were evaluated to determine the most optimum strategy for decellularizing (decelling) the bovine nucleus pulposus (bNP) to yield a natural biomaterial with a cartilaginous constituency. The resulting scaffold was then characterized with respect to its biochemistry, biomechanics and cytocompatibility. Results indicated that the optimal decell protocol involved pre-crosslinking the tissue prior to undergoing decell with trypsin and Triton X-100. The residual DNA content of the scaffold was found to be 32.64 ± 9.26 ng/mg dry wt. of tissue with sGAG and hydroxyproline (HYP) contents of 72.53 ± 16.43. and 78.38 ± 8.46 μg/mg dry wt. respectively. The dynamic viscoelastic properties were found to be preserved (complex modulus: 17.92-16.62 kPa across a range of frequencies) while the equilibrium properties were found to have significantly decreased (aggregate modulus: 11.51 ± 9.19 kPa) compared to the non-decelled fresh bNP tissue. Furthermore, the construct was also found to be cytocompatible with bone marrow stem cells (BMSCs). While it was not permissive of cellular infiltration, the BMSCs were still found to have lined the laser drilled channels in the scaffold. Taken together, the biomaterial developed herein could be a valuable addition to the AMIC family of scaffolds or serve as an off-the-shelf standalone option for cartilage repair.

Keywords: autologous matrix-induced chondrogenesis; bovine nucleus pulposus; cartilage; decellularization; focal chondral defect.

Fig. 1.

(A) Graphical Representation of Decellularization. (1) Extraction of fbNP from bovine IVD using 8 mm biopsy punch; (2) Physical compaction by insertion and squeezing in a compression chamber followed by lyophilization; (3) Crosslinking of compacted fbNP using EDC-NHS for 15 minutes; (4) Rinsing of crosslinked fbNP through a series of washes (dH2O + EtOH) leading to partial loss of compaction; (5) Cryo-trimming of swollen fbNP down to a height of 2–2.5 mm; (6) Creation of shallow grid pattern on fbNP surface using Dremel laser; (7) Biopsy punching of fbNP down to 5 mm followed by sonication in trypsin-EDTA solution for 10 minutes and then shaking in the same solution for 2 hours at 37°C; (8) Placement of fbNP in Triton-X100 solution and shaking for 24 hours followed by sonication, washing and incubation in DNase/RNase solution for 48 hours; red dots and green squiggles indicate cells and sGAGs in the matrix. (B) (Top) Compaction chamber and rod used in Step 2 of Fig. 1(A); (Middle) Compaction of fbNPs within the chamber by pressing down on them with the rod; (Bottom) Maintenance of compacted position by wrapping the system in parafilm.

Fig. 2.. Qualitative Confirmation of CA Decellularization.

(A, B) H&E staining showing that CAs that were (B) decelled for 2 hours in trypsin exhibited a distinct lack of nuclei compared to (A) non-decellularized fbNP (extracellular matrix = pink; nuclei = purplish blue); (C, D) AB-NFR staining yielded results similar to H&E with (D) CA showing empty lacunae and presence of sGAG and (C) fbNP samples showing an abundance of nuclei (GAG = blue; nuclei = red); (E, F) DAPI staining corroborated these findings with (F) almost no DNA in CA samples compared to (E) fbNP (DNA = fluorescent blue). Total Magnification: 200X.

Fig. 3.. Quantitative Confirmation of CA Decellularization.

(A) DNA quantification using the PicoGreen assay showed significant reductions in CA compared to fbNP samples; (B) Agarose Gel Electrophoresis indicated a lack of DNA in the CA (absence of bands) while the fbNP showed presence of large fragments of intact DNA (indicated by bands in white box). The DNA values for CA in both (A) and (B) fell within acceptable decellularization parameters. **** indicates significant differences (p≤0.0001); Dashed line in (A) indicates acceptable decellularization threshold (50 ng/mg dry wt.).

Fig. 4.. Gross view of the CA.

Representative images displaying the (A) diameter, (B) thickness and (C) CA surface profile with laser drilled channels and dimensions of the resulting grid architecture.

Fig. 5.. Biochemical Characterization of CA.

(A) sGAG content in CA was significantly reduced from fresh samples but still fell within range of human cartilage. (B) HYP content was unaffected by the decell; (C) sGAG to HYP ratio also fell within range of human hyaline cartilage. Crosslinking was found to significantly improve sGAG content. **** indicates significant differences (p≤0.0001). ns is not significant (p>0.05). Orange dashed lines show dry mass quantifications of respective proteins in human hyaline cartilage; For (A) orange dashed lines indicate range of sGAG found in human cartilage (i.e., between 50 – 200 ug/mg dry wt.) Continuous magenta line indicates quantifications of respective proteins in non-crosslinked CA samples.

Fig. 6.. Dynamic Mechanical Analysis of CA (Unconfined Compression).

(A) Complex Modulus; (B) Storage Modulus; (C) Loss Modulus; (D) Phase Angle; (E) Hysteresis at strains of 16%, 13%, 10% and 9% with ascendant order of frequency. No significant differences were found between fbNP and CA samples at any frequency. Dashed orange lines indicate values of respective parameters of human hyaline cartilage at 1 and 4 Hz.

Fig. 7.. Dynamic Mechanical Analysis of CA (Confined Compression).

(A) Aggregate Modulus; (B) Hydraulic Permeability; Significant differences were found between fbNP and CA samples for both parameters. ** indicates significant differences (p≤0.01). For (A) orange dashed lines indicate range of aggregate moduli found in human hyaline cartilage (i.e., between 500–900 kPa); For (B) orange dashed lines indicate range of hydraulic permeability found in human hyaline cartilage i.e., between (1–200) × 10⁻¹⁶ m⁴/Ns.

Fig. 8.. Cytotoxicity Assessment of CA.

hBMSCs were seeded on CAs and compared with non-seeded controls. Representative live/dead images of seeded CA scaffolds in the transverse plane at (A) 7 and (B, C) 14 days indicating that a majority of cells were alive on the surface and within the laser drilled channels of the CA (viable cells = green; dead cells = red) (D) alamarBlue assay performed on days 1, 7 and 14 illustrating consistent cell metabolic activity throughout the culture period. H&E staining showed no cellular infiltration into the scaffold matrix either (E) on the CA surface or (F) within the laser drilled channels on Day 1. Similarly, no infiltration was observed on Day 14 on (G) the scaffold surface or (H) in channels. Cells were seen lining the channels but not aggregating in them. ns indicates no significant differences. (A, B) have total magnification of 100x. (C, F, H) have total magnification of 200x. (E and G) have total magnification of 400x. Nuclei = dark purple; ECM = light purple.