A standardized, open-source, portable model for noninvasive joint injury in mice

Abstract

Objective: Preclinical models of osteoarthritis (OA) are crucial for the study of disease mechanisms and development of disease-modifying therapeutics. While surgical OA models, such as destabilization of the medial meniscus (DMM), have been the gold standard in the field for decades, noninvasive joint loading-based models have increased in popularity and utility. To facilitate standardization of the noninvasive anterior cruciate ligament rupture (ACLR) model in mice, we present the Mobile Joint-Injury Operator (MoJO) - an open-source protocol with accompanying fixtures and data, designed for a low-cost, commercially-available, portable, small-footprint uniaxial testing system.

Methods: We provide 3d-printable fixture designs and a comprehensive description of the loading protocol, describe the expected mechanical output, and offer various troubleshooting strategies. We validate the mechanical accuracy and inter-operator reproducibility of the procedure and characterize the resultant post-traumatic OA phenotype by knee hyperalgesia testing, flow cytometry, μCT imaging, and histological assessment.

Results: Across n ​= ​952 procedures, the MoJO protocol was highly accurate and repeatable, with a >99 ​% rate of successful ACLR and high repeatability between operators and institutions. ACLR-mediated joint injury resulted in the expected post-traumatic OA phenotype in male and female C57Bl/6 mice, including progressive hyperalgesia, histological and μCT evidence of cartilage damage, synovitis, and osteophyte formation, and increased expansion of fibroblasts, endothelial cells, and myeloid cells by flow cytometry.

Conclusions: Increased standardization of joint injury is a critical aspect of the overall refinement of animal models of OA. The MoJO represents an affordable and highly reproducible option for implementing the mouse ACLR model of OA.

Fig. 1

A). Schematic of mouse positioning and protocol overview. B). Isometric and side view of the Univert S2 and assembled ACLR fixtures. C). Isometric view of ACLR fixture components. D). Load vs time (top) and displacement vs time (bottom) plots of full loading protocol. E). Isolated view of load vs time (top) and displacement vs time (bottom) plots of preload and preconditioning segments to demonstrate expected creep. F). Isolated view of load vs time (top) and displacement vs time (bottom) plots of the final compressive loading segment to induce ACLR. The point of ACL failure is noted. G). Two representative examples of load vs time plots of successful and unsuccessful ACL ruptures, tibial fractures, and tibial physis ruptures. Note the characteristic secondary “catch” that should be observed following ACL rupture. H). Quantitative comparison of mechanical data, age at injury, and mass at injury between mice in successful ACLR, unsuccessful ACLR, and physis ruptures.

Fig. 2

A). Accuracy and repeatability data of the ACLR protocol with the Univert S2. B). Descriptive ACLR outcomes derived from analysis of mechanical data, stratified by sex across n ​= ​509 male and n ​= ​443 female mice. C). Linear correlations between body mass at the time of injury and failure displacement (left) and failure load (right). D). Linear correlations between age at the time of injury and failure displacement (left) and failure load (right). P and r² values were calculated via Pearson bivariate correlation analysis.

Fig. 3

A). Failure displacement (left) and failure load (right) in a cohort of matched ACLR procedures in which the ElectroForce protocol was employed on one limb and the MoJO protocol on the other (N ​= ​9 mice). B). Linear correlations between ElectroForce- and MoJO-derived failure displacement (left) and failure load (right). C). Flow cytometric assessment of synovial cells in naïve/uninjured mice, ElectroForce-injured mice, and MoJO-injured mice. n ​= ​3 samples per condition, each made up of two pooled synovia. D). Synovial cell proportions derived from flow cytometry between uninjured, ElectroForce-injured, and MoJO-injured mice. E). Knee withdrawal thresholds from ACLR joints (left) and contralateral joints (right), derived from blinded longitudinal knee hyperalgesia testing in ElectroForce-injured and MoJO-injured mice. n ​= ​6 mice per group assessed longitudinally.

Fig. 4

A). Safranin-O/Fast-green staining of sagittal sections from the medial and lateral compartments of Sham and ACLR 28d mice. B). 3D, sagittal, and coronal views of whole joints, derived from μCT imaging. Expected phenotypic characteristics are labeled with arrowheads.