Not All ACL Autografts Are Biologically Created Equal: An In Vitro Analysis of Graft Cellular Responses to Cyclic Strain

OBJECTIVES: Common autografts for ACL reconstruction include patellar (PT), hamstring (HT), and quadriceps tendons (QT). Various adverse outcomes, including osteoarthritis and revision risk, have been associated with different graft choices. While the time-zero material properties of autograft tissues are known, there is limited information on the biologic properties of autografts. In addition, the early differential responses of autograft fibroblasts to the biomechanical environment after surgery are not known. Our study objective was therefore to evaluate the in vitro responses of common ACL graft fibroblasts to clinically relevant strain levels. We hypothesized that fibroblasts from common autografts would exhibit differential gene expression and production of inflammatory biomarkers in response to cyclic strain. METHODS: 12 QT, 12 PT, and 9 HT were harvested from skeletally mature dogs (n=12) (ACUC #9163, 9164). Tissues were digested and ACL graft fibroblasts were seeded onto BioFlex® plates (1 × 105 cells/well), and cultured for 48h prior to application of strain. Cells were subjected to mechanical stimulation (2-s strain, 10-s relaxation, 0.5 Hz frequency) with a biaxial sinusoidal waveform with 3 strains (stress deprivation-0%, physiologic-4%, and high-10%) for 5 days using a Flexcell. On day 5, media was collected for inflammatory biomarker analysis (PGE2, IL-8, KC, MCP-1) and RNA was extracted from cells for gene expression analysis with RT-PCR (COL1, COL2, COL3, MMP2, MMP3, MMP13, ELASTIN, SCX). Significant differences among graft types were set at p<0.05. RESULTS: Gene expression: QT had higher COL1 expression at 0% strain (p<0.001) (Fig 1A). At 4% strain, COL1 expression was significantly higher in PT vs. HT (p<0.001) (Fig 1A). COL2 expression was significantly higher in PT vs. HT at 0% and 4% strain (p=0.044, p=0.012 respectively), and at 10% strain vs. QT (p=0.015) (Fig 1B). No significant differences with COL3 expression were seen between grafts for all strains (Fig. 1C). MMP2, (p=0.017) MMP13 (p=0.015), TIMP1 (p=0.001), and TIMP3 (p=0.035) expression were significantly higher in PT vs. HT at 4% strain. Elastin expression was significantly higher in PT vs. QT (p=0.034) and HT (p=0.019) at 4% strain. SCX expression was significantly higher in PT vs. HT at 4% strain (p<0.001). Inflammatory biomarkers in the culture media: QT had a higher production of IL8 and KC at 0% and 10% strain vs. PT and HT (P<0.01). There was no differences in IL8 or KC levels at 4%. PGE2 production increased for all grafts with increasing strain. At 4%, HT produced significantly higher levels of PGE2 vs. QT (p<0.001) and PT (p=0.005). However, at 10% strain, HT produced significantly lower PGE2 vs. QT (p=0.01) and PT (p<0.001). No significant differences in MCP1 levels were seen at 0 or 4% strain. CONCLUSION: Our study showed that fibroblasts from common ACL graft tissues exhibited different cellular responses to strain levels. At 4% strain, PT tended to have higher expression of extracellular matrix (ECM) genes and may have a more robust remodeling response under physiologic strain. Furthermore, the data from this study indicates that all autografts have a similar inflammatory metabolism at 4% strain, but the QT and PT may have a higher inflammatory response in stress deprivation (0%) and over strain (10%) conditions. Based on these data, we conclude that production of ECM and inflammatory response from ACL grafts are not only dependent on the mechanical environment, but also graft source.