Injectable perlecan domain 1-hyaluronan microgels potentiate the cartilage repair effect of BMP2 in a murine model of early osteoarthritis

Abstract

The goal of this study was to use bioengineered injectable microgels to enhance the action of bone morphogenetic protein 2 (BMP2) and stimulate cartilage matrix repair in a reversible animal model of osteoarthritis (OA). A module of perlecan (PlnD1) bearing heparan sulfate (HS) chains was covalently immobilized to hyaluronic acid (HA) microgels for the controlled release of BMP2 in vivo. Articular cartilage damage was induced in mice using a reversible model of experimental OA and was treated by intra-articular injection of PlnD1-HA particles with BMP2 bound to HS. Control injections consisted of BMP2-free PlnD1-HA particles, HA particles, free BMP2 or saline. Knees dissected following these injections were analyzed using histological, immunostaining and gene expression approaches. Our results show that knees treated with PlnD1-HA/BMP2 had lesser OA-like damage compared to control knees. In addition, the PlnD1-HA/BMP2-treated knees had higher mRNA levels encoding for type II collagen, proteoglycans and xylosyltransferase 1, a rate-limiting anabolic enzyme involved in the biosynthesis of glycosaminoglycan chains, relative to control knees (PlnD1-HA). This finding was paralleled by enhanced levels of aggrecan in the articular cartilage of PlnD1-HA/BMP2-treated knees. Additionally, decreases in the mRNA levels encoding for cartilage-degrading enzymes and type X collagen were seen relative to controls. In conclusion, PlnD1-HA microgels constitute a formulation improvement compared to HA for efficient in vivo delivery and stimulation of proteoglycan and cartilage matrix synthesis in mouse articular cartilage. Ultimately, PlnD1-HA/BMP2 may serve as an injectable therapeutic agent for slowing or inhibiting the onset of OA after knee injury.

Conflict of interest statement

Conflict of interest

One or more co-authors of the present manuscript have been named as inventors of one or more U.S. patents and/or applications relating to the technology described in the manuscript.

Figures

Fig. 1

Injectable PlnD1-HA microgels visualized following hydration in saline buffer..

Fig. 2

Histological sections of mouse knees processed 7 days post repair or control treatments and stained with Safranin-O and Fast Green. Knees treated with PlnD1-HA/BMP2 showed a normal smooth articular cartilage appearance (B) and showed no OA damage when compared with papain-damaged knees treated with either saline (C), PlnD1-HA (D), HA (E), or BMP2 (F). Proteoglycan depletion indicated by a loss of Safranin-O staining is marked by asterisks. An increase in chondrocyte clusters was observed in the proteoglycan-depleted region of articular cartilage of PlnD1-HA-treated knees (see arrows in D). Such clusters are absent in BMP-2 treated knees where articular cartilage fissures and delamination is observed (see arrowheads in F). No obvious difference is seen between control knees (A) injected twice with saline (in place of papain and the repair treatment) and the papain-injected knees treated with PlnD1- HA/BMP2 microgels (B). T, tibia; F, femur; m, meniscus. Scale bar represents 50 μm

Fig. 3

Box and whisker plot showing the median (central line), 25–75 percentile (boxes) and the entire range of scores obtained 7 days after treatment of saline or papain-damaged knees (n=5 for control injected twice with saline; n=9 for all the other groups). ** indicates p

Fig. 4

Effect of the combined administration of BMP2 and PlnD1-HA particles on mRNA levels of articular cartilage ECM components and ECM-modifying enzymes. Fold changes in mRNA levels are shown for knees treated with BMP2-loaded PlnD1-HA particles relative to knees treated with growth factor free particles (control PlnD1-HA) on days 1 and 7 following intra-articular injections. mRNA levels of the α1 chain of type II collagen (COL2a1), aggrecan (ACAN), perlecan (HSPG2) and xylosyltransferase 1 (XYLT1) were significantly increased in knees treated with PlnD1-HA/BMP2 compared with control (PlnD1-HA) knees whereas the opposite was seen with matrix degrading enzymes (MMP13, MMP3, and ADAMTS5) and the α1 chain of type X collagen (COL10a1). Each fold change value equal or higher to 2 (above the dashed line in A) and equal or lower to 0.5 (below the dashed line in B) is considered statistically significant, when PlnD1-HA/BMP2 is compared with control PlnD1HA. Error bars represent standard deviations.

Fig. 5

The extent of the type II collagen signal (green) is diminished in the articular cartilage of saline (D) versus PlnD1-HA/BMP2 (C) -treated knees after 7 days of treatment. For each group, consecutive coronal knee sections were stained with Safranin O/Fast Green (A, B) and immunostained with an antibody specifically directed against type II collagen (green in C, D). A and B are the adjacent histological stains for C and D, respectively. DRAQ5™ was used as nuclear stain (blue in C, D). Growth plate (GP) cartilage served as an internal positive control for type II collagen immunodetection. Asterisks indicate the presence of a healthy articular cartilage of high cellularity in the superficial layer of PlnD1-HA/BMP2-treated knees. F, Femur; T, Tibia; m, meniscus. Scale bars represent 100μm.

Fig. 6

Articular cartilage of knees treated with PlnD1-HA/BMP2 showed increased expression of aggrecan relative to control (PlnD1-HA or saline) knees after 7 days of treatment. For each group, consecutive coronal knee sections were stained with Safranin O/Fast Green (A-C) and immunolabeled with an antibody specifically directed against aggrecan (D–F). The aggrecan-specific immune signal (green) was noticeably increased in knees treated with PlnD1-HA/BMP2 (D and G) when compared with control knees treated with PlnD1-HA (E and H) or saline (F and I). A, B, C are the adjacent histological stains for D, E and F, respectively. DRAQ5™ was used as nuclear stain (blue in D-F). Growth plate (GP) cartilage served as an internal positive control for aggrecan immunodetection in osteoarthritic knees that displayed severe depletion of proteoglycans at their articular cartilage surface (see arrows in panel D-F). Scale bars represent 100μm for either A-F or G-I. F, Femur; T,Tibia, m, meniscus.

Fig. 7

Box and whisker plot showing the median (central line), 25 to 75 percentile (boxes) and the entire range of scores obtained 14 days after treatment of papain-damaged knees (A, n=9 for each group). ** indicates p