Cartilage regeneration by chondrogenic induced adult stem cells in osteoarthritic sheep model

Abstract

Objectives: In this study, Adipose stem cells (ADSC) and bone marrow stem cells (BMSC), multipotent adult cells with the potentials for cartilage regenerations were induced to chondrogenic lineage and used for cartilage regenerations in surgically induced osteoarthritis in sheep model.

Methods: Osteoarthritis was induced at the right knee of sheep by complete resection of the anterior cruciate ligament and medial meniscus following a 3-weeks exercise regimen. Stem cells from experimental sheep were culture expanded and induced to chondrogenic lineage. Test sheep received a single dose of 2 × 10(7) autologous PKH26-labelled, chondrogenically induced ADSCs or BMSCs as 5 mls injection, while controls received 5 mls culture medium.

Results: The proliferation rate of ADSCs 34.4 ± 1.6 hr was significantly higher than that of the BMSCs 48.8 ± 5.3 hr (P = 0.008). Chondrogenic induced BMSCs had significantly higher expressions of chondrogenic specific genes (Collagen II, SOX9 and Aggrecan) compared to chondrogenic ADSCs (P = 0.031, 0.010 and 0.013). Grossly, the treated knee joints showed regenerated de novo cartilages within 6 weeks post-treatment. On the International Cartilage Repair Society grade scores, chondrogenically induced ADSCs and BMSCs groups had significantly lower scores than controls (P = 0.0001 and 0.0001). Fluorescence of the tracking dye (PKH26) in the injected cells showed that they had populated the damaged area of cartilage. Histological staining revealed loosely packed matrixes of de novo cartilages and immunostaining demonstrated the presence of cartilage specific proteins, Collagen II and SOX9.

Conclusion: Autologous chondrogenically induced ADSCs and BMSCs could be promising cell sources for cartilage regeneration in osteoarthritis.

Conflict of interest statement

Competing Interests: The authors have declared that no competing interests exist.

Figures

Figure 1. The arthroscopic and arthrotomy representation of the right knee joints before and after surgical inductions.

(a) The arthroscopic representation of the right knee joints before the surgical operations revealed no chondral lesion on medial femoral condyle (MFC) and medial tibial plateau (MTP). (b) The arthrotomy photo during the surgical operations revealed no previous lesions at Patella femoral groove (PFG), medial femoral condyle (MFC), medial tibial plateau (MTP). (c) The arthrotomy photo of patella (P) before OA induction revealed no chondral degenerations (black arrow). (d) The arthroscopic images after OA inductions at MFC and MTP showed slight indentations and moderate focal lesions (white arrows). (e) Arthroscopic images after OA inductions at the PFG revealed severe and prominent chondral erosion (white arrows). (f) The arthroscopic images after OA inductions at P, which is the opposite contact of PFG, depicted degenerations of cartilage caused by the frictional contacts. Arthroscopy scale  = 0.5 cm, Arthrotomy  = 1.5 cm.

Figure 2. Multipotency evaluations of ADSCs and BMSCs.

(a) Inverted phase contrast images of the Adipogenic inductions of ADSCs and BMSCs. The image depicted the collection of fat droplets in clusters (black arrows) from day 4 of the inductions process. The oil red staining showed that lipids formed on both induced cell samples, picked up red stain (white arrow) showing their positivity to adipogenic lineage. Scale  = 70 µm. (b) Inverted phase contrast images of the osteogenic inductions of ADSCs and BMSCs. The picture revealed similar clustered mineralization on both cell samples from day 4 of the inductions process. The evaluation of the induced ADSCs and BMSCs with alizarin red staining demonstrated some mineralization activities (white arrow) showing their commitment to osteogenic lineage. Scale  = 70 µm. (c) Inverted phase contrast images of the chondrogenic inductions of ADSCs and BMSCs. The picture revealed similar aggregation of the cells in condensations seen in early chondrogenesis from day 4 of the inductions process. The evaluation of the induced ADSCs and BMSCs with toluidine blue staining demonstrated that the condensed matrixes picked up the blue stain (red arrows), showing their positivity to chondrogenic lineage. Scale  = 70 µm.

Figure 3. Monolayer analysis of BMSCs and ADSCs.

(a) The morphological images of the BMSCs and ADSCs from P0 to P4. BMSC looked more spindle upon isolation and early attachment, while ADSCs were broader (P0-P1). Both cell samples attain the same size and shape as the passage progressed from P2–P4 showing more mesenchymal-like structure. Scale bar represents 35 µm. (b) The Morphological changes of the cells during chondrogenic differentiation. BMSCs formed the aggregates of cartilage in a film-like sheet, more readily than ADSCs. The aggregates and sheet clumped together (white arrow) to form a firm cartilage structure by 3rd week. ADSCs split its formed aggregates and dispersed on the medium. Scale represents 35 µm. (c) The Population doubling time (PDT) of the cell samples during culture. ADSCs were 34.4±1.6 hrs, while that of BMSCs were 48.8±5.3 hrs. ADSCs had a significantly higher growth rate compared to the BMSCs with P value of 0.008. (d) The viability of cells after trypsinization, at the end of each passage. From P0–P4, BMSCs had mean viabilities of (87, 88, 92, 86, and 90); while ADSCs had (93, 96, 95, 96, and 95) respectively. ADSCs had higher viabilities at each passage compare to BMSCs but were significantly higher at P1, P2 and P3; P<0.05.

Figure 4. Gene expression analysis of ADSCs and BMSCs.

(a) Comparison of the gene expressions of chondrogenically induced ADSCs to the uninduced. After 21 days of induction, the induced samples had significantly higher expressions of Col II, SOX9, Aggrecan and Col I genes in fold increases of (12.96, 23.44, 44.36 and 18.44) respectively compare to the uninduced (week 1 post-isolation) cells, with P values (∞) respectively (P<0.05). (b) Comparison of the gene expressions of chondrogenically induced BMSCs to the uninduced. After 21 days of induction, the induced samples had significantly higher expressions of Col II, SOX9, Aggrecan and Col I genes in fold increases of (16.21, 51.40, 120.33 and 19.17) respectively compare to the uninduced (week 1 post-isolation) cells, with P values (*) respectively (P<0.05). Comparing the Induced ADSCs and Induced BMSCs, BMSCs had significantly higher expressions of Col II, Aggrecan and SOX9 genes in fold increases of (1.85, 2.03 and 5.31) respectively with P values (**) (p<0.05). There was no significance difference on Col I expression.

Figure 5. The gross evaluations of the right knee joint samples.

(a) The gross images of knee representing the six regions. Patella femoral groove (PFG), medial femoral condyle (MFC), lateral femoral condyle (LFC), medial tibia plateau (MTP), lateral tibia plateau (LTP) and patella (P) were used as reference points. The treated knees showed regenerations. At PFG, the regenerated cartilages had unique morphological appearances different from the native (black ring); the control still retained severe cartilage degeneration (black ring). At MFC, the ADSCs still retain a slight focal defect (black arrow); BMSCs had no sign of defect, while the control retained severe cartilage degeneration (black ring). At LFC, there were no lesions on the treated knees, but the control had reduced cartilage thickness (black arrow). At MTP, both treated samples revealed structural appearances of a crescent regenerating meniscus-like cartilage (black arrow), but none at the control. There is no evidence of meniscus regeneration on the medial tibia plateau (yellow arrow). There is also a remnant of resected spur bone formed at the anterior region of the medial tibia plateau (white arrow), which is a supportive mechanism seen in severe OA. At the LTP, there were no conspicuous lesions at the treated samples but the control retained a severe degeneration (black ring). At P, the control presented worse degeneration compared to either of the ADSCs or BMSCs (black arrows). (M =  medial, and L =  lateral). Scale represents 1.5 cm. (b) Combined gross and histological ICRS grading of the right knee joints. The control sheep scored a mean grade of 3.33±0.2, while the test groups ADSCs and BMSCs scored 1.5±0.2 and 1.3±0.3 respectively. Both treated sheep samples had significantly higher grades (*) and (**) respectively compared to the control P<0.05. BMSCs treated sheep had better grade score than ADSCs, but was not significant P = 0.465.

Figure 6. Fluorescence evaluations of PKH26 dye on the resected regenerated cartilages.

Samples were taken from PFG and MFC of the BMSCs samples; then from MTP and P of the ADSCs samples. PKH26 fluorescence was shown in 2D, the composite images and 3D images. The composite and 3D confocal images revealed the integration and arrangements of the labeled chondrogenically induced cells. The fluorescence of the dye proved the participation of the injected cells in the cartilage regeneration. Scale: gross images  = 1.5 cm and microscopic images  = 35 µm.

Figure 7. Histological and immunohistochemical evaluations of the right knee joints.

(a) Haematoxylin and Eosin stain of the right knee joint PFG samples. Yellow arrow points at the exposed area of cartilage, leading to subchondral bone (green arrow) of the control knee joints. Treated samples (BMSCs and ADSCs) reflected regenerated engineered cartilages (white arrows) covering the subchondral bone, though not smoothly packed like the native cartilage (black arrow). Scale represents 70 µm. (b) Safranin O stains of the right knee joint (patella) samples. The regenerated cartilage on both treated knee (BMSCs and ADSCs) stained positive (white arrow). They were homogenous to the histochemical properties of accumulated proteoglycans revealed via Safranin O, but not smooth as the native cartilage (black arrow). Yellow arrow points at the degenerated area of cartilage on the control; green arrow, the subchondral bone. Scale represents 70 µm. (c) Immunohistochemistry images of the right knee joint (PFG) samples. Collagen type II protein staining for both treated samples (BMSCs and ADSCs) demonstrated positive staining within the ground tissues and were visible throughout the matrixes, evidenced by the green fluorescence tagged with the secondary antibody (black arrow). The native articular cartilage served as positive control (black arrow). The absence of green fluorescence on the ground tissue of the engineered cartilages without the primary antibody (yellow arrow) served as the negative control. Scale represents 35 µm. (d) Immunohistochemistry images of the right knee joint (PFG) samples. SOX9 staining for BMSCs and ADSCs demonstrated positivity within the ground tissues and were visible throughout the matrixes by the green fluorescence tagged with the secondary antibody (black arrow). The native articular cartilage served as positive control (black arrow), while the absence of green fluorescence on the ground tissue of the engineered cartilages without the primary antibody (yellow arrow) served as the negative control. Scale represents 35 µm. (e) Immunohistochemistry images of the right knee joint (PFG) samples. Collagen type I staining for BMSCs and ADSCs demonstrated diminished positivity within the ground tissues. This occurred mainly as scattered clusters of the green fluorescence tagged with the secondary antibody (black arrow). The fibrous cartilage served as positive control (black arrow), while the absence of green fluorescence on the ground tissue of the engineered cartilages without the primary antibody (yellow arrow) served as the negative control. Scale represents 35 µm. Note: The authors wish to state that a superimposed histological image of Safranin O and PKH26 dye or the immuno signals with PKH26 dye would have made better proves of the presence of the injected cells, but owing to technical challenges with these processes and the dye stability, it was not accomplished.