Experimental Overview.
All procedures were approved by the Institutional Animal Care and Use Committee and performed in accordance with the NIH Guide for the Care of Use of Laboratory Animals. Animals (n = 18) were maintained at Colorado State University’s Laboratory Animal Resources housing facilities and were monitored daily by a veterinarian. Guinea pigs were singly housed in solid bottom cages and provided standard chow, hay cubes, and water ad libitum. Sixteen Hartley guinea pigs of the same age, from a coincident but unrelated project, were utilized as an untreated control group for body weight comparisons. This study focused on male animals. A complementary study was performed in female guinea pigs and will be the focus of an independent manuscript.
Two sequential cohorts were utilized, as separate assessments were required for molecular-based properties of knee joints (n = 10) versus biomechanical outcomes (n = 8). For the former goal, the following evaluations were performed: treadmill-based gait collection; histopathology; clinical microcomputed tomography (microCT) grading; and transcript and protein expression. The sample size for this cohort was determined from a pilot study focused on histologic assessment of OA as the primary outcome. Using a within group error of 0.5 and an effect size of 1.0 in a Wilcoxon signed-rank test (matched pairs) for t tests on G*Power (Publisher, Location), power associated with an alpha level of 0.05 was calculated as 0.90 with a sample size of 10.
Biomechanical methods included: whole knee anterior drawer testing; patellar tendon pull to failure; indentation of articular cartilage and menisci; treadmill-based and voluntary gait assessment; quantitative microCT assessments; and atomic absorption spectroscopy (AAS) of articular cartilage. Sample size for this cohort was determined from a pilot study focused on indentation of articular cartilage as the primary outcome. Using the analysis provided above, but with an effect size of 1.2, power associated with an alpha level of 0.05 was calculated as 0.90 with a sample size of 8.
Surgical removal of the IFP.
Resection of the IFP was performed on all animals (n = 18). After medial parapatellar arthrotomy of the right knee, the patella was temporarily displaced cranially with the knee in extension to permit access to the femoral groove. The patella was repositioned once the IFP was exposed medially for removal; the skin incision was closed following full dissection. An identical sham procedure, with minor manipulation but without removal of the IFP, was performed on the left knee and served as a matched internal control for each animal.
Gait assessments.
Obligatory treadmill-based gait analysis was performed on all animals (n = 18) using a DigiGait™ treadmill system (Mouse Specifics, Inc., Framingham, MA). Additionally, animals dedicated to biomechanical testing (n = 8) performed voluntary weight bearing assessment using a Rodent Walkway System (Tekscan, South Boston, MA). Animals were acclimated to both apparati over a 2-week period. Data collection was performed by the same handlers during the same time period (8:00AM to 12:00PM); the order in which animals were analyzed was randomly selected. Baseline gait and weight-bearing analyses were performed the day before IFP removal surgery. Subsequent data were collected every 4 weeks after surgery, with the final time point occurring the day before termination. Data is absent for 8 guinea pigs during week 12 due to Covid-19 pandemic restrictions.
Tissue collection.
All animals (n = 18) were harvested at 7 months of age. Body weights were recorded at the time of harvest; animals were transferred to a CO2 chamber for euthanasia. Hind limbs were removed at the coxofemoral joint. The lengths of the left and right femurs were measured using digital calipers. Animals in the first cohort (n = 10) had their left and right hind limbs removed and placed in 10% neutral buffered formalin for 48 hours and transferred to phosphate buffer saline (PBS) for microCT grading. After microCT, limbs were placed in 12.5% solution of ethylenediaminetetraacetic acid (EDTA) at pH 7 for decalcification. EDTA was replaced twice weekly for 6 weeks. Animals utilized for biomechanical outcomes (n = 8) had their left and right hind limbs immediately frozen and stored at -20°C for quantitative microCT followed immediately by biomechanical assessments.
MicroCT.
Knee joints from the first cohort (n = 10) were scanned in PBS using the Inveon microPET/CT system (Siemens Medical Solutions, Malvern PA), with a voxel size of 18 mm, a voltage of 100 kV, and an exposure time of 1356 ms. Clinical features of boney changes of OA were scored using a published whole joint grading scheme25. Features graded included: presence/size and location of osteophytes, subchondral bone cystic changes, subchondral bone sclerosis, articular bone lysis, and intraarticular soft tissue mineralization. Images were scored by a board-certified veterinary radiologist (AJM) blinded to limb identification.
Knees dedicated to biomechanical outcomes (n = 8) were scanned in PBS for microCT using a Bruker Skyscan 1276 (Bruker, Billerica, MA) with a voxel size of 20 µm. Osteophytes were identified by manually outlining new bone growth every 10 slides by a single operator (GEN). Osteophytic bone volume was identified by reduced bone mineral density and as being outside the original joint margins using CTAn software. Similar to previous studies, four regions of interest (ROI) were chosen to coincide with areas of mechanical testing on the articular cartilage of the lateral and medial compartments26–29. Trabecular volumes of interests (VOIs) ranged from just below the cortical bone to the growth plate. CTAn software (Bruker, Billerica, MA) was used to determine bone volume/ tissue volume (BV/TV), trabecular thickness (Tb.Th), trabecular separation (Tb.Sp), trabecular number (Tb.N), and volumetric bone mineral density (vBMD). Cortical bone VOIs were located between the joint margin and trabecular bone and were identified by a single operator (NV) using CTAn software. Cortical bone mineral density, cortical bone porosity, and cortical thickness were determined using CTAn.
Histologic Grading of OA.
Following decalcification and paraffin-embedding of knees from the first cohort (n = 10), three sagittal 5mm sections were made through each knee joint: (i) mid-sagittal slices were used for histologic evaluation of the IFP; and sagittal slices through both the (ii) medial and (iii) lateral compartments were utilized to assess OA changes in four sites (medial tibia, lateral tibia, medial femur, and lateral femur). The mid-sagittal sections were stained with hematoxylin and eosin (H&E) and Masson’s trichrome stain to confirm structural modifications. The medial and lateral compartments were stained with toluidine blue for OA grading using OARSI recommended criteria24. The OARSI published grading scheme is based upon species-specific features of OA, including: articular cartilage structure, proteoglycan content, cellularity, and tidemarks integrity. Two blinded veterinary pathologists (LBR and KSS) performed histological grading. Scores from each of the four anatomic locations were summed to obtain a total knee joint OA score for each right and left hind limb. An intraclass correlation coefficient of 0.9 for between-reviewer consistency was calculated; the few minor discrepancies identified were resolved prior to statistical analysis.
Biomechanical Analyses.
Methods relevant to cranial/anterior drawer testing and pull to failure are provided in Supplemental Tables 1. Cartilage and menisci from both native IFP and FCT knee joints were subjected to indentation relaxation tests in a phosphate buffered saline bath to determine instantaneous and equilibrium moduli. The bath was mounted to a two degree of freedom camera mount and an X-Y stage to allow for the indentation surface to be oriented normal to the indenter tip. Both the femoral and tibial cartilage were indented on the medial and lateral plateaus, which is a location where there would be cartilage on cartilage contact. To test the cartilage, the tibia was potted distally and adjusted to indent perpendicular to the surface. The cartilage was assumed to be 0.5mm thick on the femur and 0.75 mm on the tibia. The menisci were extracted off the tibial plateau at the anterior and posterior root attachments. They were then glued onto a plexiglass mounting plate to allow the proximal surface to be mounted perpendicular to the indenter tip. The medial and lateral menisci were indented at both the anterior and posterior locations. Menisci were assumed to be 1.2 mm thick and, following a preload of 50mN, menisci were indented to 10% strain (0.12 mm) in 1 second and held for 900 seconds to reach equilibrium. Hertzian contact between an elastic half space (meniscus) and a rigid sphere (indenter) was assumed30,31. A custom MATLAB algorithm (Mathworks, Natick, MA) was used to calculate the instantaneous and equilibrium moduli from the collected data.
Atomic Absorption Spectroscopy (AAS).
Phosphorus, magnesium, calcium, zinc and iron quantification were performed on samples of tissues from the second cohort (n = 8), including: the native IFP or replacement tissue that developed in its place; and articular cartilage from the tibia and femur (representing pooled medial and lateral compartments). Briefly, dried tissue was weighed, reduced to ash, sonicated in 3.6N nitric acid, and diluted 30-fold with deionized water32. Diluted samples were analyzed using a Model 240 AA flame atomic absorption spectrometer and SpectraAA software (Agilent Technologies, Santa Clara, CA)33. Mineral levels were reported as parts per million (ppm) dry weight (dw)34.
Gene Expression using Nanostring technology.
Total ribonucleic acid (RNA) was isolated from either the IFP or replacement tissue that remained in formalin-fixed paraffin-embedded blocks (n = 9) following acquisition of adequate sections for histopathology and immunohistochemistry (IHC) using a commercially available kit specific for such (Roche, Basel, Switzerland). A custom set of guinea pig-specific probes were designed and manufactured by NanoString Technologies (Seattle, WA) for the following genes: adiponectin (ADIPOQ), complement component 3 (C3), catalase (CAT), collagen type 1 alpha 2 (COL1A2), fatty acid synthetase (FASN), G protein pathway suppressor 2 (GPS2), leptin (LEP), monocyte chemoattractant protein-1 (MCP-1), and matrix metallopeptidase-2 (MMP-2), nuclear factor kappa-β transcription complex (NF-κ B p65 & NF-κB p50), and nuclear receptor subfamily 4 group A member 2 (NR4A2) (Supplemental Table 2). Per initial RNA quantification (Invitrogen Qubit 2.0 Fluorometer and RNA High Sensitivity Assay Kit, Thermo Fisher Scientific, Waltham, MA) and Fragment Analyzer quality control subsets (Fragment Analyzer Automated CE System and High Sensitivity RNA Assay Kit, Agilent Technologies), the optimal amount of total RNA (800.00ng) was hybridized with the custom code-set in an overnight incubation set to 65ºC, followed by processing on the NanoString nCounter FLEX Analysis system. Results were reported as absolute transcript counts normalized to positive controls and three housekeeping genes (β-actin, succinate dehydrogenase, and glyceraldehyde-3-phosphate dehydrogenase). Any potential sample input variance was corrected by use of the housekeeping genes and application of a sample-specific correction factor to all target probes. Data analysis was conducted using nSolver software (NanoString Technologies).
IHC and quantitative analysis.
IHC was performed on mid-sagittal sections containing medial tibial cartilage, the IFP, or replacement tissue using polyclonal rabbit antibodies to MCP-1 (Abcam ab9669) or NF-kB p65 (Abcam ab86299), each at a concentration of 2.5 mg/ml. Prior to staining, slides were incubated in citrate buffer for 5 hours at 55°C for antigen retrieval, as recommend for skeletal tissues35. 2.5% normal goat serum was used as a blocking reagent. Slides were incubated in primary antibody overnight in a humidified chamber at 4°C, followed by a 30-minute incubation with a biotinylated goat anti-rabbit secondary antibody. Bone marrow hematopoietic cells and macrophages served as internal positive controls for each section. Negative assay controls – rabbit immunoglobulin at 2.5 mg/ml and secondary antibody, alone – did not result in background immunostaining. Sections were counterstained with hematoxylin, cover slipped, and imaged by light microscopy. Data was quantified as integrated optical density using ImagePro-Plus 7 Software (Media Cybernetics, Rockville, MD). Four 1-mm-wide regions of interest of medial tibial cartilage and replacement tissue were analyzed for MCP-1 and NF-κB p65 expression; data for each tissue type was averaged prior to statistical analysis.
Statistical analyses.
Rationale for excluding individual values from data sets were determined prior to analysis and included whether an appropriate sample was unable to be collected, did not pass quality control parameters, or if integrity was compromised. Exclusion of animals resulted in: n = 12 animals for treadmill-based gait results; n = 8 for OARSI scoring; n = 9 for transcript expression; n = 8 for MCP-1 cartilage IHC; n = 6 for NF-κB p65 cartilage IHC; n = 9 for MCP-1 and NF-κB p65 IFP/FCT IHC; n = 7 for biomechanical analyses; and n = 5 for AAS of IFP/FCT. Complete data sets were available for all other outcomes.
Statistical analyses were performed with GraphPad Prism 9.1.1(La Jolla, CA) with significance set at p < 0.05. Data underwent normality and variance testing with the Shapiro-Wilk test. For mobility outcomes, longitudinal data was analyzed using two-way or mixed model ANOVA with a Tukey post hoc test analysis. For all other analyses, normally distributed data with similar variance were compared using paired t-test† for normally distributed data; Wilcoxon matched-pairs signed rank test ◊ was used for non-normally distributed data.