Fig. 3

Walking velocity and spatiotemporal gait pattern analysis before and after intra-articular injection of IDO-Gal3 or saline in MCLT + MMT operated knees. Prior to treatment (weeks 3–7), saline- and IDO-Gal3-treated rats used similar walking velocities (p ≥ 0.89, A). However, following treatment, rats treated with IDO-Gal3 used faster walking velocities than saline controls on week 10 and week 11 (p < 0.001, p = 0.033, respectively). Balanced gaits are defined as stance time being equal on the left and right limb, or a difference of stance time between limbs being near zero. Temporally symmetric gaits have foot strike sequences that are equally spaced in time, where a right foot strike occurs halfway between two left foot strikes (i.e., temporal symmetry ≈ 0.5). Prior to treatment, saline- and IDO-Gal3-treated rats used balanced, symmetric gaits. After treatment, the gait of saline-treated rats became temporally asymmetric at weeks 8–11 (^symmetry > 0.5, p < 0.05, Bonferroni-corrected t-test) and differed from IDO-Gal3 treatment at weeks 8–10 (p ≤ 0.035, Tukey’s HSD pairwise test, B). Neither saline- nor IDO-Gal3-treated animals had imbalanced stance times, though saline-treated animals tended to spend more time on the contralateral limb (imbalance > 0) while IDO-Gal3-treated animals tended to spend more time on the affected limb (imbalance < 0). Here, at weeks 8 and 9, stance time imbalance differed between saline- and IDO-Gal3-treated animals (p ≤ 0.025, Tukey’s HSD pairwise test, C). Spatial symmetry measures the symmetry of the foot placement, rather than the timing of the foot strike. Again, a spatially symmetric gait has a right foot placement about halfway between two left foot placements (spatial symmetry ≈ 0.5). Here, both IDO-Gal3- and saline-treated animals had spatially symmetric gait patterns throughout the experiment (D). This experiment includes over 1500 gait trials, with 65–131 trials collected at each treatment timepoint. As such, data are plotted using density plots, with bars indicating the 95% confidence interval of each treatment-timepoint mean as predicted by our linear mixed effects statistical model