Joint injury-induced angiogenesis and thrombospondin 2

Abstract

Purpose: It is controversial whether inhibition of angiogenesis slows the progression of post-traumatic osteoarthritis (PTOA). The matricellular protein thrombospondin-2 (TSP2) is a candidate for inhibiting intraarticular angiogenesis after joint injury as it is a well-characterized endogenous anti-angiogenic protein. In addition, prior literature demonstrates the inhibition of matrix metalloproteinase (MMP) bioavailability by TSP2. However, little is known of the involvement of TSP2 in joint trauma, injury-induced angiogenesis and MMP activity, and PTOA progression. The purpose of this study was to characterize the endogenous TSP2 response following traumatic joint injury in mice. Methods: With IACUC approval, noninvasive ACLR was induced in adult TSP2-GFP reporter mice and C57Bl/6 mice via tibial compression. Limbs were processed for histology at 3d, 14d, and 28d post-ACLR (n=3-4/timepoint). The proportion of TSP2+ cells in synovium was quantified in injured and contralateral limbs of TSP2-GFP mice using k-means clustering of reporter intensity. ACLR-induced angiogenesis was assessed in C57Bl/6 mice via CD31 immunostaining and in vivo near-infrared imaging (NIR) utilizing the AngioSense-agent (Perkin Elmer), which enables in vivo assessment of vascular changes. Synovia from C57Bl/6 ACLR and contralateral knees were dissected, homogenized in Trizol, and analyzed for TSP2, MMP2, MMP3, MMP13 gene expression via rtPCR at 3d and 14d post-ACLR (n=5-8/timepoint), representing timepoints of early and late inflammation. To determine which cytokines induce TSP2 in fibroblast-like synoviocytes (FLS) and whether ACLR-associated in vivo inflammation affects the in vitro FLS cytokine response, primary mouse FLS were isolated from knee synovium of 7d post-ACLR and contralateral knees (n=3/group). FLS were induced with IL-1β and TNFα (10 ng/ml) for 48 hours and analyzed for TSP2 and MMP13 gene expression. Gene expression was normalized to ATP5b and paired t-tests performed. Results: ACLR induced intraarticular angiogenesis at 14d and 28d, assessed by influx of CD31+ cells. Abundant CD31+ staining was observed in synovium, menisci, and osteophytes of injured knees up to 28d post-ACLR (Figure. 1 A-C). An observed increase in in vivo AngioSense signal of the injured knee 7d post-ACLR (P=0.021) corroborates this finding and indicates acute vascular changes (Figure. 1D-E). ACLR also induced significant upregulation of MMP-2, -3, and -13 mRNA by synovium at 3d and 14d post-ACLR (Figure. 2K). In TSP2-GFP reporter mice, injured knees exhibited abundant synovial TSP2 expression, as early as 3d and up to 28d (at which point mice exhibit moderate PTOA severity; data not shown). At 3d and 14d, TSP2 expression is confined to the synovium, and by 28d, developing osteophytes are also TSP2(+). At all timepoints, synovial TSP2 expression is localized to the subintima. The intima of injured joints, although hyperplastic, remains TSP2(-). The cruciate ligaments, menisci, articular cartilage, and subchondral bone remain TSP2(-) (Figure. 2A-H). Quantitative image analysis indicates ACLR-induced TSP2-GFP expression in ∼27% of synovial cells at 14d (P=0.017) and ∼24% at 28d (P=0.018) post-ACLR, compared to near-zero expression in contralateral synovia (Figure. 2J). rtPCR confirms strong TSP2 upregulation in injured C57Bl/6 synovium at 3d (P=.006) and 14d (P=.007) (Figure. 2I). Surprisingly, both IL-1β and TNFα downregulated TSP2 in FLS from healthy (IL-1β - P=.009; TNFα - P=.001) and injured (IL-1β - P=.001; TNFα - P=.10) joints. TNFα-treated FLS from injured joints exhibited significantly less downregulation compared to healthy FLS (P=0.025) (Figure. 3B). Both IL-1β and TNFα induced a strong MMP13 response in healthy (P=0.002; P=0.004) and injured (P=0.001; P=0.006) FLS (Figure 3A), and there was no difference in the magnitude of MMP13 expression between healthy and injured FLS. Conclusions: TSP2 is strongly upregulated by the synovium following ACLR. This coincides with a synovial MMP-2, -3, and -13 response, increased intraarticular angiogenesis, and the time course of post-injury inflammation. Interestingly, TSP2 expression is confined to the subintimal region of the synovium. Our in vitro data shows that primary FLS do not upregulate TSP2 in response to IL-1β or TNFα, and interestingly, FLS from injured joints do not exhibit differentially-elevated TSP2 compared to FLS from contralateral joints, despite abundant in vivo synovial TSP2 expression following ACLR. Thus, ongoing work is identifying local TSP2+ cell populations and ligands that induce TSP2 expression in FLS. Given evidence on the anti-angiogenic capacity of TSP2, ongoing work is assessing angiogenesis, inflammation, catabolism, and PTOA severity in global TSP2-KO mice. Our ultimate objective is to mitigate injury-induced angiogenesis with exogenous TSP2 treatment.View Large Image Figure ViewerDownload Hi-res image Download (PPT)View Large Image Figure ViewerDownload Hi-res image Download (PPT)