Runx2-Modified Adipose-Derived Stem Cells Promote Tendon Graft Integration in Anterior Cruciate Ligament Reconstruction

Abstract

Runx2 is a powerful osteo-inductive factor and adipose-derived stem cells (ADSCs) are multipotent. However, it is unknown whether Runx2-overexpressing ADSCs (Runx2-ADSCs) could promote anterior cruciate ligament (ACL) reconstruction. We evaluated the effect of Runx2-ADSCs on ACL reconstruction in vitro and in vivo. mRNA expressions of osteocalcin (OCN), bone sialoprotein (BSP) and collagen I (COLI) increased over time in Runx2-ADSCs. Runx2 overexpression inhibited LPL and PPARγ mRNA expressions. Runx2 induced alkaline phosphatase activity markedly. In nude mice injected with Runx2-ADSCs, promoted bone formation was detected by X-rays 8 weeks after injection. The healing of tendon-to-bone in a rabbit model of ACL reconstruction treated with Runx2-ADSCs, fibrin glue only and an RNAi targeting Runx2, was evaluated with CT 3D reconstruction, histological analysis and biomechanical methods. CT showed a greater degree of new bone formation around the bone tunnel in the group treated with Runx2-ADSCs compared with the fibrin glue group and RNAi Runx2 group. Histology showed that treatment with Runx2-ADSCs led to a rapid and significant increase at the tendon-to-bone compared with the control groups. Biomechanical tests demonstrated higher tendon pullout strength in the Runx2-ADSCs group at early time points. The healing of the attachment in ACL reconstruction was enhanced by Runx2-ADSCs.

Figure 1. Identification of ADSCs by multi-lineage differentiation and cell surface markers.

(a) Passage 3 ADSCs show fibroblast-like morphology (original magnification 100×). (b) Oil red stain shows lipid droplets 1 week after the induction of adipogenic differentiation. (c) ADSCs pellets were stained positively with toluidine blue after induction of chondrogenic differentiation at 14 days. (d) Alkaline phosphatase was detected in the cytoplasm 14 days after induction of osteogenic differentiation. The scale bar is 100 μm. (e) Specific cell surface markers were detected by flow cytometry. The ADSCs were negative for the hematopoietic lineage markers CD34 and CD45. The fractions of CD44-, CD90-, and CD105-positive cells were 67.63%, 68.13% and 92.43%, respectively, indicating their mesenchymal origin. The experiments were performed three times, and representative images are shown. Values are the mean ± SEM.

Figure 2. Infection of Ad-Runx2 (co-expression with EGFP) and expression of Runx2 in ADSCs.

(a,b) Runx2 expression (green) in Ad-Runx2-infected ADSCs at 48 h post-infection was examined by fluorescence microscopy (a) and control cell morphology by light microscopy (b). (c,d) Ad-Runx2 transduction efficiency was measured by flow cytometry at 48 h post-infection (c) and using non-transduced ADSCs as controls for autofluorescence (d). (e–h) Immunofluorescence staining showed nuclear expression of Runx2 in Ad-Runx2-ADSCs at 48 h post-infection. EGFP (green) is expressed in the nucleus and cytoplasm (e), and Runx2 expression assayed by immuno-histology (red) is confined to the nucleus in infected cells (f). A merged image of (e,f) is shown in (g). Non-transduced cells were used as a negative control (h). The experiments were performed three times and representative images are shown. The scale bar is 20 μm.

Figure 3. Expression of osteogenic and adipogenic genes in ADSCs infected with Ad-Runx2.

(a) Runx2 mRNA was detected by real-time RT-PCR in ADSCs infected with Ad-EGFP or Ad-Runx2 at 1, 3, 7, 10 and 14 days post-infection. (b–d) The expression of specific osteogenic genes, including OCN (b), BSP (c) and COLI (d) in ADSCs infected with Ad-EGFP or Ad-Runx2 at 1, 3, 7, 10 and 14 days post-infection, as assayed by real-time PCR. (e,f) The expressions of the specific adipogenic markers LPL (e) and PPARγ (f) mRNA levels by real-time RT-PCR in ADSCs infected with Ad-EGFP or Ad-Runx2 at 1, 3, 7, 10 and 14 days post-infection. (g) Analysis of ALP activity in the lysates of ADSCs infected with Ad-Runx2 or Ad-EGFP in at 2 3, 7, 10 and 14 days post-infection. ALP activity in Runx2-infected cells was elevated at day 3, continued to increase with a peak at 10 days, and then gradually declined, but remained at a high level for the duration of the experiment. The intrinsic cellular ALP activity in Ad-EGFP-infected cells remained unchanged. The experiments were performed three times, and representative images are shown. Values are mean ± SEM, **P < 0.01, *P < 0.05 vs. Ad-EGFP-ADSCs.

Figure 4. Ad-Runx2-infected ADSCs promote ectopic bone formation in vivo.

(a) Bone formation was detected by X-rays in the muscle of right lower limb, 8 weeks after intramuscular injection of Ad-Runx2-modified ADSCs into the right lower limb of nude mice, but not in the left control limb injected with ADSCs infected with Ad-EGFP (scale bar is 10 mm). (b–e) Histological analyses of in vivo osteogenesis by Ad-Runx2-modified ADSCs. Note that significant amounts of cartilage and bone formation were observed with hematoxylin & eosin (b) and toluidine blue staining (c) in the Ad-Runx2-ADSCs group. The Ad-EGFP-ADSCs group contained muscle and fibrous tissue, with no evidence of bone formation (d,e). Representative images from six replications are shown.

Figure 5. Surgical procedures for ACL reconstruction in rabbit models treated with Ad-Runx2-ADSCs, fibrin glue or RNAi Runx2.

(a,b) The semitendinosus tendon was harvested. (c,d) Drilled tunnels (1.5 mm) were created through the femur (c) and tibia (d) at the insertion of the native ACL. (e) The graft was passed through the bone tunnels to replace the ACL. (f) Using a special needle equipped with fibrin glue, we mixed 0.2 ml of the fibrin glue with 1 × 10⁶ Runx2-ADSCs. (g,h) Injection around the periphery of the bone tunnel of the femur (g) and tibia (h) surrounding the tendon graft. Representative images from 10 replications are shown.

Figure 6. Evaluation of the width of bone tunnels by CT 3D reconstruction.

(a–d) 12 weeks after surgery, the CT images were acquired to measure the CSA (cross sectional area) of the bone tunnels. First, we located the bone tunnel in cross sections (a). Then, using CT 3D reconstruction, the corresponding sagittal (b) and coronal (c) images were obtained to ensure that the orientation was perpendicular to the long axis of the bone tunnel. Based on the orientation, a series of cross sections were created, resulting in 1.0 mm-thick slices with no interslice gap (d). (e,f) Average diameters of bone tunnels (e) and the ossification areas (f) were analyzed from the CT 3D reconstruction of the joint after ACL reconstruction surgery treated with Ad-Runx2-ADSCs, fibrin glue or RNAi Runx2. Values are mean ± SEM, n = 10, *P < 0.05, **P < 0.01 as indicated.

Figure 7. Gross observation and histological analysis of the joint after ACL reconstruction surgery on rabbits

(a) Gross observation demonstrated no significant adverse effect of the treatments with Ad-Runx2-ADSCs, fibrin glue or RNAi Runx2 on the joint tissue. Note the lack of ossification or scar abundance at any of the intra-articular tissues, such as ligaments or menisci. (b) Histological photomicrographs of specimens post-surgery. Specimens were harvested and observed 2 (i, ii, iii), 4 (iv, v, vi), 6 (vii, viii, ix), 12 (x, xi, xii), 26 (xiii, xiv, xv), and 52 weeks (xvi, xvii, xviii) after surgery. Specimens were divided into three groups, includingAd-Runx2-ADSCs treated (i, iv, vii, x, xiii, xvi); fibrin glue treated (ii, v, viii, xi, xiv, xvii); and RNAi Runx2-treated (iii, vi, ix, xii, xv, xviii). T, tendon; IF, interface tissue; FC, fibrous cartilage; B, bone; CC, calcified cartilage; S, Sharpey fiber. Representative images from four replications are shown.

Figure 8. Biomechanical testing of the joint after ACL reconstruction surgery.

(a) Images taken during the biomechanical analysis of the rabbits joints. The femur-ACL graft-tibia complexes were fixed in specially designed clamps, allowing tensile loading along the axis of the graft in a material testing machine. (b,c) The average ultimate load to failure (b) and failure location (c) of the graft in the Runx2-ADSCs, fibrin glue and RNAi Runx2 specimens at 2, 4, 6, 12 and 52weeks after surgery. Values are mean ± SEM, n = 6. *P < 0.05, **P < 0.01. ns, not significant.