Erythrocyte removal from bone marrow aspirate concentrate improves efficacy as intra-articular cellular therapy in a rodent osteoarthritis model

Introduction

Osteoarthritis (OA) is a multifaceted disease process impacting more than 10% of the US adult population and is highly associated with pain, disability, and economic burden (1,2). Current options to treat OA include nonsteroidal anti-inflammatories, intra-articular injections of corticosteroids or biological therapies, and arthroscopic debridement with chondral resurfacing techniques. Despite the high prevalence of OA, there remains a lack of effective treatment options that offer pain relief and improve quality of life without the risk of significant adverse effects (e.g., gastrointestinal side effects associated with nonsteroidal anti-inflammatory use). Cellular therapies, in many cases touted as “stem cell therapy”, for treatment of OA and other chronic orthopedic conditions have emerged as an attractive new option in both human and veterinary surgery (3-16). While cellular therapy is increasingly popular, efficacy remains unclear due to variability in composition of cellular products and lack of consistency among study designs supporting different therapies.

Increased Food and Drug Administration (FDA) regulation of cellular therapies has prompted evaluation of minimally manipulated treatments. Currently, biological therapies employed for human OA include autologous bone marrow aspirate concentrate (BMAC) and enriched autologous platelet products (3-5,8,17,18). Bone marrow aspirate concentrate has demonstrated benefit in healing full thickness cartilage defects in both animal models (19,20) and human clinical patients with naturally occurring disease (21-27). In equine models of induced full-thickness chondral defects, BMAC treatment resulted in improved macroscopic and histologic scoring of cartilage tissue quality based on magnetic resonance imaging compared to microfracture alone (19) and reduced surgical trauma to subchondral bone (20). It was further suggested by the authors of that paper that “given the few mesenchymal stem cells in minimally manipulated BMAC, other mechanisms such as paracrine, anti- inflammatory, or immunomodulatory effects may have been responsible for the tissue regeneration observed” (20). In clinical patients, BMAC has demonstrated therapeutic potential in several orthopedic conditions following intra-articular injection, including knee and spinal OA (28-32), with significant improvement in pain and functionality scores, in uncontrolled, open-label trials (32).

BMAC consists of a mixture of multiple different cell types, including red blood cells (RBCs), stromal cells, immune cells (macrophages, neutrophils, B and T lymphocytes), and true hematopoietic stem cells (HSCs), all of which are concentrated by the centrifugation and washing steps required to prepare BMAC (8). Although BMAC is used as a cellular therapy in the management of acute and chronic OA in humans, the individual contribution of the diverse cell populations in BMAC in suppressing inflammation and stimulating cartilage repair has not been established (3,33,34). Of note, BMAC injections are reportedly highly inflammatory and anecdotally result in significant morbidity following injection in many recipients (3). Indeed, numerous previous studies have highlighted the role that hemarthrosis plays in joint degeneration, demonstrating blood-induced joint damage with increased levels of pro-inflammatory cytokines and eventual cartilage degradation (35-56). Thus, we hypothesized that the RBC component within BMAC might interfere with the potential benefits of the overall cellular therapeutic effects of BMAC. If it was shown that RBC elimination improved the efficacy of BMAC in rodent models, it might be possible to remove RBCs from BMAC to improve activity in clinical patients.

Therefore, the overall objectives of this work were: (I) to compare the relative inflammatory properties of RBC-lysed and non-lysed BMAC using in vitro assays and (II) to determine whether RBC removal improved the efficacy of BMAC in a rodent destabilization of the medial meniscus (DMM) model of OA. The results of these studies could improve cellular therapies for OA by providing mechanistic understanding of optimal BMAC cellular therapies. We present this article in accordance with the ARRIVE reporting checklist (available at https://atm.amegroups.com/article/view/10.21037/atm-22-4256/rc).

Methods

This study was approved by the Institutional Animal Care and Use Committee at Colorado State University (IACUC protocol #1017) and conducted according to the national guidelines under which the institution operates, and NIH Guidelines for the Care of Use of Laboratory Animals (8^th edition).

In vitro assays for BMAC-induced inflammatory responses

In vitro assays using macrophages as the indicator cells assessed the relative immunogenicity (cytokine release) of BMAC and RBC-depleted BMAC. BMAC was prepared from mouse bone marrow by extrusion from the tibia and femur with a #27 needle from two 12-week-old male C57BL/6Nci mice (Charles River Laboratories, Wilmington, MA, USA), as described for preparation of human BMAC (3,8). Briefly, the tibia and femur were severed using a #10 scalpel blade at both proximal and distal aspects of the bone and bone marrow lavaged from the medullary cavity using a #27 needle with sterile phosphate buffered saline (PBS). The bone marrow was suspended in PBS and RBCs were removed by Ammonium-Chloride-Potassium (ACK)-lysis for 10 minutes followed by centrifugation 1,200 g for 5 minutes and three washing steps in PBS. BMAC and RBC-depleted BMAC preparations were then added at a 1:3 ratio to mouse AMJ (alveolar macrophage with the J2 retrovirus carrying the v-raf and v-myc oncogenes) murine alveolar macrophages [American Type Culture Collection (ATCC), Manassas, VA, USA] cultured in Roswell Park Memorial Institute (RPMI) medium (Gibco^TM, Thermo Fischer Scientific, Waltham, MA, USA) with 10% fetal bovine serum (FBS) + glutamine (57,58), and plated at a density of 2×10⁵ macrophage cells per triplicate wells of 96-well plates in complete Dulbucco’s Modified Eagle Medium (DMEM) (Gibco^TM, Thermo Fischer Scientific) culture, as described recently (59-62). Controls included macrophages cultured in complete DMEM media alone. After 48 h in culture, macrophage supernatants were assayed for tumor necrosis factor-α (TNF-α) concentrations by enzyme-linked immunosorbent assay (ELISA) according to manufacturer’s instructions (Mouse TNF-α DuoSet ELISA, R&D Systems, Minneapolis, MN, USA).

Animals

Studies were done using 12-week-old male and female C57BL/6Nci mice (Charles River Laboratories, Wilmington, MA, USA). Thus, this study considered sex as a biological variable, as male mice develop post-traumatic osteoarthritis (PTOA) more severely in the DMM model than female mice (63). Animals (n=10 per group; 5 males, 5 females) were randomly assigned to receive either surgery followed by no injection (untreated), complete BMAC, or RBC-depleted BMAC, for a total of thirty mice used. All animals were included in the final analyses. Mice were housed together by sex in groups of five in solid bottom cages with corncob bedding and allowed ad libitum water and standard rodent chow. Mice were maintained in the university’s Laboratory Animal Resources building and were assessed by a veterinarian daily with body weights monitored weekly. At the conclusion of the study, mice were euthanized via CO₂ inhalation with confirmation by cervical dislocation.

Bone marrow collection and preparation of BMAC

Bone marrow was collected from the tibia and femur of 12 healthy mice (two mice per treatment group and per sex) by extrusion with a #27 needle, as per above. The cell suspension was processed to produce BMAC using two centrifugation steps, as described previously and above (8). A portion of the BMAC was reserved for cytological analyses and flow cytometric evaluation.

Destabilized medial meniscus model of OA

Osteoarthritis was created in mice using a well-established method by DMM (64). To perform DMM, the right stifle (knee) of all mice was injured as previously described (64). Briefly, an approximately 3-mm longitudinal incision was made using a #15 blade overlying the distal patella of the right hindlimb to the proximal tibial plateau (64). Following medial parapatellar arthrotomy, the infrapatellar fat pad (IFP) was temporarily repositioned laterally, allowing access to the anterior medial meniscotibial ligament, which anchors the medial meniscus to the tibial plateau. This ligament was severed using a #11 scalpel blade. The IFP was repositioned, and the incision closed using 6-0 monofilament absorbable suture in simple interrupted pattern. Contralateral limbs served as naïve controls.

Gait and activity monitoring to assess OA severity

Mice were monitored prior to and for 3 weeks following intra- articular injection by two methods: individual cage monitoring and Digigait treadmill-based analysis. To determine general animal behavior and mobility, each mouse underwent individual cage monitoring for 10 minutes weekly during the experimental time-course. In all cases, animals were placed in their resident cage with their environmental enrichment hut for the duration of the assessment. Animals were acclimated to the system over 1 week, after which one baseline measurement was collected immediately before the start of the study. Animals received testing on days 7, 14, and 21 following intra-articular injection. Training and data collection occurred randomly among animals during the same time of day (8 am–12 pm) and involved the same handlers. The video analysis software (ANY-maze^TM, Wood Dale, IL, USA) automatically collected mobility parameters including total distance traveled, mean speed, time in hut, and entries to the top of hut. Parameters of interest were accessed both (I) cross-sectionally among groups normalized to pre-DMM surgery baseline and (II) longitudinally overtime.

Animals were monitored for gait abnormalities using the Digigait analysis system (Mouse Specifics, Inc., Framingham, MA, USA) (15). Briefly, animals were placed on a treadmill at 35 cm/s (flat or 10° inclined; three runs per time point for each grade) and video recorded for analysis. Animals were acclimated to the system over 1 week, during which baseline was collected in triplicate at one timepoint immediately prior to starting the study. Subsequent DigiGait data collection occurred on days 7, 14, and 21 following intra-articular injection. Animal testing order was randomly selected at each timepoint for data acquisition and testing occurred during the same time of day (1 pm–4 pm) and involved the same handlers. Data processing was performed using Digigait software. Parameters evaluated included % swing stride, % stance stride, % brake stride, and % propel stride. Parameters of interest were accessed both (I) cross-sectionally among groups normalized to pre-DMM surgery baseline and (II) longitudinally overtime.

Joint histology and immunohistochemistry

Twenty-eight days after surgery, knee/stifle joints were fixed in 10% neutral buffered formalin for 24 h, followed by decalcification in ethylenediamine tetra-acetic acid (EDTA) and paraffin embedding. Sections (5 µm) were taken from the center of the medial tibial plateau and stained with toluidine blue. Histological grading of joint tissues was conducted using established criteria for OA and were completed by two blinded, independent individuals trained in using this scoring method (65). Briefly, joint tissues were semi-quantitatively graded for osteoarthritic damage including cartilage fibrillation, and cartilage loss including clefts/erosions and calcification, in addition to synovitis and proteoglycan content for the whole joint and medial and lateral joint compartments (65).

NanoString analysis of immune transcriptomes of stifle tissues

RNA was harvested from formalin fixed, paraffin embedded articular cartilage tissues from both the medial and lateral tibia and femurs of study animals using Qiagen RNeasy formalin fixed paraffin embedded (FFPE) kit following manufacturer’s instructions. Samples from tissues of n=4 male mice per group (no treatment, BMAC, and RBC-depleted BMAC) were randomly selected for evaluation by NanoString analysis using a mouse immune panel (NanoString nCounter Murine Immunology Panel (catalog number XT-CSO-MIM1-12). Prior to analysis by NanoString, RNA quality and quantity were assessed by NanoDrop spectrophotometry and integrity evaluated by Agilent TapeStation. All RNA samples analyzed contained RNA molecules in size >200 nucleotides or larger exceeding 50% of total RNA (66). Analysis of transcript abundance and statistical significance was performed with the nCounter Analysis Flex software and with Partek^® Flow^® Genomics Analysis software v10.0 (Partek Inc., Chesterfield, MO, USA).

Statistical analysis

The experimental sample size for the in vivo study was calculated using GPower Version 3.1.1 (1). An a-priori power analysis was conducted from pilot gait data (stride length) obtained using this injury model in mice. To account for biological variability related to sex, both male and female mice were analyzed. Originally, statistics were run to stratify sex; however, no significant differences were seen (P>0.05). Therefore, the sexes were combined (10 animals per time-point), leaving time as the main factor of interest. Cross-sectional data was normalized to baseline values and were compared using repeated-measures analysis of variance (ANOVA) with Tukey’s correction. Longitudinal data were compared using repeated-measures ANOVA with Dunnett’s correction; each timepoint was compared to baseline values. Histological scoring at end-term was evaluated using a one-way non-parametric ANOVA (Kruskal-Wallis test) with Dunn’s multiple comparisons test. Statistical analysis was performed using GraphPad Prism v9.3.1 (GraphPad Software Inc., La Jolla, CA, USA). Significance for overhead enclosure monitoring was assessed at P<0.1. The outcomes utilized P<0.05 (67).

Gene expression count data and fold changes were analyzed via nSolver software, using housekeeping gene normalization without background thresholding or subtraction. Figures were generated using Partek Flow v10 (Partek Inc.). Differentially expressed genes (DEGs) were generated using ANOVA with multiple test correction. Panther database (http://pantherdb.org) was used to categorize genes and define molecular functions and biological processes (68,69).

Results

Macrophage responses to BMAC and impact of RBC depletion on inflammatory cytokine production

Initial studies were done using in vitro assays to assess the relative inflammation inducing properties of BMAC, and to determine whether RBC depletion modulated any inflammatory effects of BMAC. These studies were conducted using AMJ macrophages as the readout for inflammatory responses, as we have previously used this cell line as a sensitive indicator of immune responses in vitro (70). To assess inflammatory responses, macrophages were incubated with BMAC cells or RBC-depleted BMAC cells (adjusted to add equivalent numbers of nucleated cells for each cell population) and incubated for 48 h. At this time, the release of pro-inflammatory cytokine TNF-α was assessed by ELISA, using cell culture supernatants. We observed that BMAC stimulated significant production of TNF-α compared to resting AMJ macrophages (P=0.0001), and that RBC-depleted BMAC stimulated significantly less TNF-α than non-depleted BMAC (P<0.0001) (Figure 1). These results indicated that RBC depletion had the potential to significantly reduce the inherent inflammatory properties of BMAC in vivo, and to improve overall treatment outcomes for BMAC treatment of established OA.

Figure 1 Suppression of macrophage cytokine release by BMAC cells. Purified BMAC cell subpopulations (ACK-lysed or non-lysed) were co-cultured 48 h at 1:3 ratio to differentiated mouse AMJ macrophages. Controls included AMJ murine macrophages cultured in complete DMEM media alone, and BMAC and unlysed BMAC. After 48 h in culture, macrophage supernatants were assayed for TNF-α concentrations by ELISA (lower limit of detection 31.2 pg/mL). Significant differences noted with P values <0.05. ACK, Ammonium-Chloride-Potassium; AMJ, alveolar macrophage with the J2 retrovirus carrying the v-raf and v-myc oncogenes; BMAC, bone marrow aspirate concentrate; DMEM, Dulbucco’s Modified Eagle Medium; ELISA, enzyme-linked immunosorbent assay; RBC, red blood cell; TNF-α, tumor necrosis factor α.

Comparison of functional impact of treatment with BMAC or RBC-depleted BMAC in mice with induced OA

The in vitro findings (above) indicated that RBC depletion might reduce BMAC inflammation, but the critical question was whether this effect would also translate to greater improvement in overall joint function in the case of established OA. To assess functional improvement, both activity monitoring and gait analysis were employed.

Voluntary activity was monitored in animals with DMM-induced OA (no-injection control, BMAC-treated, and RBC-depleted BMAC-treated) using the ANY-maze^TM overhead cage monitoring software (Figure 2). When cross-sectional data was normalized to pre-DMM surgery baseline values, mice treated with RBC-depleted BMAC exhibited increased distance traveled (Figure 2A, P=0.05) and mean speed (Figure 2C, P=0.04) compared to no treatment control animals at day 27 post-DMM surgery. Additionally, RBC-depleted BMAC animals spent less time in their security huts at day 27 post-surgery relative to no-injection controls (Figure 2E, P=0.09). Both BMAC (Figure 2G, P=0.03) and RBC-depleted BMAC animals (Figure 2G, P=0.05) showed an increase in their top of hut entries at day 27 compared to no-injection control animals, indicating greater willingness to propel off their operated hindlimbs. This effect could reflect several processes, including reduction of intra-articular inflammation and/or improved cartilage integrity. Longitudinal data is provided (Figure 2B,2D,2F,2H); within group differences out to day 28 were not a significant finding.

Figure 2 ANY-maze^TM cage monitoring parameters following DMM on the right hind limb of mice and treatment dependent on group. Parameters of interest statistical significance is shown overtime and between groups for distance traveled (A,B), mean speed (C,D), time in hut (E,F), and entries to the top of hut zone (G,H). Filled in symbols indicate male animals and open symbols indicate female animals. Significant differences noted, with P value noted <0.1 and significant differences over time labeled (*, P<0.05). BMAC, bone marrow aspirate concentrate; DMM, destabilization of the medial meniscus; RBC, red blood cell.

We next assessed the impact of BMAC treatment and RBC depletion on gait in animals with OA. These analyses were done using the Digigait^® controlled treadmill walking system, which focused on parameters of interest at 35 cm/s at both a 10-degree incline (Figure 3) and on a flat surface (Figure S1). Results in the manuscript proper are depicted from animals evaluated at a 10-degree incline both (I) cross-sectionally among groups, normalized to pre-DMM surgery baseline Digigait values (Figure 3A,3C,3E,3G) and (II) longitudinally overtime (Figure 3B,3D,3F,3H).

Figure 3 Digigait controlled treadmill walking at 35 cm/s at a 10-degree incline. Parameters of interest are shown both overtime and between groups for %swing stride (A,B), %stance stride (C,D), %brake stride (E,F), and %propel stride (G,H). Filled in symbols indicate male animals and open symbols indicate female animals. Significant differences noted, with P value noted <0.1 and significant differences over time labeled (*, P<0.05; **, P<0.005). BMAC, bone marrow aspirate concentrate; RBC, red blood cell.

The first parameter evaluated was %swing stride. Mice treated with either BMAC or RBC-lysed BMAC trended toward having a lower %swing stride compared to non-injected controls at day 27 (Figure 2A; P=0.07 and P=0.09, respectively). When longitudinal data were assessed, only non-injected controls demonstrated an increased %swing stride from baseline by day 27 (Figure 3B, P=0.04).

When %stance stride was evaluated, BMAC treated animals trended towards a cross-sectional increase at day 27 compared to the non-injected group (Figure 2C; P=0.07). Longitudinal data confirmed that no-injection control animals showed a decreased %stance stride compared to baseline by day 27 (Figure 3D; P=0.04).

The %stance stride was further broken down into %brake stride and %propel stride. %Brake stride refers to the portion of the step from initial touchdown onto the belt until the maximum amount of the paws area is touching the belt. %Propel stride refers to the portion of the step beginning when the maximum amount of the paw is touching the belt until the paw is lifted off the belt. No differences were noted for %brake stride between or among groups at day 27 (Figure 3E,3F). In contrast, no-injection animals showed a longitudinal decrease in %propel stride at day 27 post-injury (Figure 2G; P=0.01); this group also trended toward decreased %propel stride from both BMAC and RBC-lysed BMAC animals at day 27 (Figure 2H; P=0.05 and P=0.08, respectively).

Impact of BMAC treatment on cartilage integrity

We next evaluated the impact of BMAC treatment on joint pathology, with a particular focus on articular cartilage integrity. Joint tissues were collected from the three treatment groups at day 28 after OA was induced in the DMM model. Images from these studies are depicted in Figure 4. In untreated animals with DMM-induced OA there was: (I) loss of proteoglycan, synovitis, and fibrillations and/or clefts on the cartilage surface in the medial compartments (Figure 4D); and (II) small fibrillations and loss of proteoglycan on the lateral surfaces (Figure 4G). In contrast, BMAC treated animals showed fewer fibrillations and relatively maintained proteoglycan within both the medial and lateral compartments (Figure 4E,4H). In animals treated with RBC-depleted BMAC, there was improved proteoglycan content and decreased OA progression in both the medial and lateral compartments compared to the other groups (Figure 4F,4I).

Figure 4 Toluidine blue photomicrographs from no-injection control (left), unlysed BMAC (middle), and lysed BMAC (right) limbs. Low magnification image of whole knee joints from no-injection control (A), unlysed BMAC (B), and lysed BMAC (C), post-DMM and applicable treatment. Higher magnification image for evaluation of the medial compartment no-injection control (D), unlysed BMAC (E), and lysed BMAC (F). Additional higher magnification image for evaluation of the lateral compartment for no-injection control (G), unlysed BMAC (H), and lysed BMAC (I). (A-C) 4×, scale bar =200 μm; (D-I) 20×, scale bar =20 μm. BMAC, bone marrow aspirate concentrate; RBC, red blood cell.

These histologic changes were quantified using the recommended Osteoarthritis Research Society International (OARSI) scoring system. RBC-lysed BMAC animals demonstrated a decreased whole joint OARSI score compared to no-injection control animals and trended towards a decreased score relative to BMAC treated animals (Figure 5A; P=0.0115 and P=0.0693, respectively). Additionally, decreased synovitis was seen in the RBC-lysed BMAC animals compared to no-injection control animals or BMAC treatment (Figure 5B; P=0.0007 and P=0.0336, respectively). As expected with the DMM mouse PTOA model, the majority of OA changes were noted within the medial compartment of the joint rather than the lateral portion. As such, the OARSI scores for the medial compartment were improved for the RBC-lysed BMAC animals compared to no-injection control or BMAC animals (Figure 5C; P=0.0003 and P=0.0086, respectively), with similar changes noted in the medial compartment proteoglycan scores (Figure 5D; P=0.0003 and P=0.0185, respectively). No differences among groups for the lateral compartment were seen for the OARSI (Figure 5E) or proteoglycan scores (Figure 5F). Additional scored and individual OARSI parameters are listed in Figure S2.

Figure 5 OARSI histopathology scores following DMM surgery and treatment with no-injection control (blue), unlysed BMAC (green), or lysed BMAC (purple). Parameters listed include whole joint OARSI score (A), synovitis score (B), medial compartment OARSI score (C), medial compartment proteoglycan score (D), lateral compartment OARSI score (E), and lateral compartment proteoglycan score (F). Filled in symbols indicate male animals and open symbols indicate female animals. Significant differences noted with P values noted <0.1. BMAC, bone marrow aspirate concentrate; DMM, destabilization of the medial meniscus; OARSI, Osteoarthritis Research Society International.

Impact of RBC depletion on joint immune transcriptome responses to BMAC injection

To better understand how BMAC injections affected the overall immune transcriptome of the joint, RNA was extracted from FFPE joint tissues, and analyzed using a NanoString panel (Murine Immune panel, catalog #XT-CSO-MIM1-12) to broadly examine multiple immune pathways. Using these data, we first assessed the degree of relatedness of three study populations (complete BMAC, RBC-depleted BMAC, and no-injection groups). Principal component analysis (PCA) of the three study populations showed distinct clustering patterns (Figure 6A).

Figure 6 NanoString gene expression results. (A) PCA plot of n=4 per group, legend shown in top left. BMAC in blue, RBC-lysed BMAC samples in yellow, and no-injection controls in red. (B) Volcano plot of ANOVA results from comparison of DEGs between RBC-lysed BMAC vs. no-injection controls. X-axis shows fold change and y-axis shows P values. Red and blue dots indicate significantly upregulated and downregulated respectively, with significance defined as P value ≤0.05 (unadjusted) and fold change ≥2. Number of significant genes indicated in red for upregulated and blue for downregulated. (C) Table of 20 genes containing top 10 most upregulated and top 10 most downregulated comparing RBC-lysed BMAC vs. no-injection controls, with P values, full gene name and description listed. (D) Volcano plot of BMAC vs. no-injection controls with axis and significance parameters as defined above. (E) Table containing 20 genes from RBC-lysed BMAC vs. no-injection comparison including all 4 upregulated and 16 most downregulated genes. ANOVA, analysis of variance; DEGs, differentially expressed genes; BMAC, bone marrow aspirate concentrate; PC, principal component; PCA, principal component analysis; RBC, red blood cell.

Comparison of RBC-depleted BMAC to no-injection control animals demonstrated significant upregulation of 15 genes and downregulation of 18 genes (significance defined as fold change ≥2 or P value ≤0.05) (Figure 6B). The 10 most upregulated and downregulated genes are shown in Figure 6C with full expression shown in online tables (available at: https://cdn.amegroups.cn/static/public/atm-22-4256-1.xls).

Comparison of complete BMAC injection to no-injection control animals yielded fewer significantly upregulated (n=3) genes and downregulated (n=25) genes (Figure 6D). The most up- and downregulated genes are listed in Figure 6E with full expression provided in online tables (available at: https://cdn.amegroups.cn/static/public/atm-22-4256-1.xls).

Differential gene expression analysis [gene set enrichment analysis (GSEA)] was done to compare the genes up-and down-regulated BMAC treatment. These analyses identified 33 DEGs (upregulation of n=15, downregulation of n=18 genes) in animals treated with RBC-lysed BMAC compared to untreated animals (Figure 6B) and 28 DEGs in animals treated with BMAC compared to non-treated animals (Figure 6C). In contrast, when RBC-depleted BMAC treated animals were compared to BMAC treated animals, we found 50 DEGs (upregulation of n=44, downregulation of n=6 genes) [Figure 7A, full expression provided in online tables (available at: https://cdn.amegroups.cn/static/public/atm-22-4256-2.xlsx)]. A heat map was generated of the significantly expressed DEGs in BMAC, RBC-depleted BMAC and untreated animals (Figure 7B).

Figure 7 NanoString analysis of specific genes related to eliminating RBC from treatment. (A) Volcano plot of RBC-lysed BMAC compared to BMAC, with x-axis as fold change and y-axis as unadjusted P value. Red shows significantly upregulated, while blue represents significantly downregulated genes. (B) 50 significant genes defined as P value ≤0.05 and fold change >2 in aforementioned comparison were used to construct a heat map, with expression levels represented on a scale of red (highest) to blue (lowest). Gene names are listed on the left rows, with groups on top of columns. Yellow (RBC-lysed BMAC), red (no-injection control), and blue (BMAC). Dendrograms in black show clustering by samples (top) and also by genes (right). (C) Venn diagram constructed using all differentially expresses genes with P value ≤0.05 and fold change >1. Each circle as labeled including 65 DEGs from RBC-lysed BMAC vs. BMAC (green), 41 DEGs from BMAC vs. no-injection control (blue) and 45 DEGs from RBC-lysed BMAC vs. no-injection control (red). BMAC, bone marrow aspirate concentrate; DEGs, differentially expressed genes; RBC, red blood cell.

To help visualize exactly how treatment with RBC-depleted BMAC impacted immune transcriptomes in the joint compared to BMAC treatment or untreated animals, a Venn diagram was generated (Figure 7C). After eliminating overlapping DEGs that were altered in BMAC group, 48 genes remained; 19/48 genes also demonstrated differential expression when comparing RBC-depleted to untreated joint groups. Full differential gene expression lists are provided in online tables (available at: https://cdn.amegroups.cn/static/public/atm-22-4256-2.xlsx). Discovery of genes that were uniquely altered in the RBC-depleted BMAC treated joints, compared to either BMAC treatment or no treatment, indicating the RBC removal uncovers other immune modulatory properties of BMAC that are obscured by RBC-induced inflammation.

Discussion

Bone marrow aspirate concentrate products have demonstrated therapeutic potential in preclinical models of OA and clinically in various musculoskeletal disorders (27,31). The most notable finding of the present work was that removal of erythrocytes from BMAC improved treatment of OA in a murine model. This was illustrated via significant increases in cage monitoring parameters (distance traveled, mean speed, time in hut, entries to top of hut zone) and gait parameters (lower %swing stride, higher %propel stride) at day 27 in RBC-lysed BMAC treated mice compared to no-injection control mice. Importantly, treatment with RBC-depleted BMAC also significantly improved cartilage integrity compared to control animals. These findings add to the body of literature regarding the use of regenerative therapies in the treatment of OA as well as the deleterious effects of hemarthrosis, specifically the negative influence of erythrocytes on synovial tissues, supporting previous work demonstrating blood-induced joint damage was associated with increased levels of pro-inflammatory cytokines and cartilage degradation (35-56). Thus, our study findings indicate that the overall effectiveness of BMAC therapy for OA can be significantly improved by RBC depletion prior to intra-articular injection.

While it has been previously noted that BMAC preparations can induce inflammatory responses in patients (3), this is the first study to our knowledge to demonstrate that the presence of RBCs in BMAC may have deleterious effects on joint healing and functional improvement. Given that it is nearly impossible to avoid blood contamination of BMAC using conventional aspiration techniques, our findings indicate that direct RBC elimination (by lysing) is necessary to fully remove RBCs. It is also possible to remove RBCs from bone marrow aspirates by Ficoll density separation, but this technique requires biosafety cabinets, considerable manual dexterity, and laboratory centrifuges to perform. Thus, a closed system using RBC lysis with agents such as ACK would be preferable for the direct to bedside use of autologous BMAC for treatment of orthopedic injuries and OA.

Pain and loss of mobility are the primary reasons people seek treatment for OA (71). As pain is a chief complaint for OA, tracking behavior and mobility changes are important aspects of post-traumatic OA treatment models to assess. Open-field testing, or cage monitoring can be an effective method to obtain objective research data on behavior and mobility measures (72). Cage monitoring systems, such as ANY-maze^TM, can track rodents within a designated area to view post-injury behavior and mobility differences compared to baseline. By 27-days post-DMM surgery, animals injected with RBC-lysed BMAC demonstrated increased activity, as seen by a greater distance traveled and overall mean speed, compared to untreated mice. Additionally, RBC-lysed BMAC treated animals spent less time in their security hut. This altered mobility may imply a potential difference in pain or clinical responses among groups. Of note, both RBC-lysed BMAC and complete BMAC treated animals showed increased entries to the top of hut zone by day 27 post-DMM surgery. Climbing requires use of the hind limbs, and as such, a willingness to enter the top of hut, indicating a maintained ability to utilize the hind limb despite injury. Importantly, similar to our results, intra-articular autologous BMAC injections have been shown clinically in humans to decrease pain, increase functionality and knee related quality of life, along with increasing patient physical function (4). These improvements have been seen in multiple studies out to 12 months post-injection (73,74).

Gait analysis systems can be an effective way to assess mobility and function (75). As rodents tend to walk with balanced symmetrical gait, alterations in gait can be indicative of pain and/or an attempt to protect the injured limb (76). As such, these deviations from a symmetrical gait can be used as an indicator of pain (77). To the authors’ knowledge, there have been no other studies that have utilized treadmill walking in the evaluation of BMAC injections in rodents. Inclined treadmill running was prioritized over flat running to pose a greater challenge to the mice at this relatively early timepoint. Animals running on an inclined surface are required to work against gravity, increase muscle work and combat a “toppling moment” or the forces that may cause an animal to fall down the surface (78). Notably, both the complete and RBC-lysed BMAC animals maintained gait parameters similar to baseline by day 27 post-DMM surgery. In contrast, by day 27 post-DMM surgery, gait parameters in the no-injection control animals varied the most from their preoperative baseline. Specifically, no-injection animals were less inclined to put their foot down on the belt, and in the short period their foot was on the belt, they were less likely to propel off that foot into their next step. Overall, no-injection animals demonstrated a modulated gait during controlled treadmill movement by day 27 post-DMM surgery compared to pre-surgery baseline levels.

Histopathology findings demonstrated significant improvement in mice treated with RBC-depleted BMAC compared to untreated animals or those treated with BMAC. By 28 days post-DMM surgery, no-injection animals showed expected PTOA progression based on previously published work (64). DMM surgery demonstrated increased OA progression on the medial aspect of the joint compared to the lateral portion of both the tibia and femur (64). For this reason, improvements in OA scores are most expected in the medial aspects, as seen in this study. Importantly, RBC-lysed BMAC demonstrated decreased whole joint OARSI scores compared to both the no-injection control and complete BMAC treated animals. Similar to reported DMM models, osteophytes and patellar dislocation were not present in any of our mice (64). Proteoglycan is one of the major components of the extracellular cartilage matrix. It is important for binding water to increase compressive load, but also to store a variety of factors including chemokines, cytokines and growth factors. Importantly, our RBC-lysed BMAC treated animals demonstrated a decreased proteoglycan score, implying improved proteoglycan health and integrity, compared to either the complete BMAC treated animals or no-injection controls. Notably, animals treated with RBC-lysed BMAC demonstrated decreased synovitis scores compared to either no-injection control animals or complete BMAC treated animals. Of note, in humans treated with bone marrow mononuclear cells, magnetic resonance imaging (MRI) revealed 63% of patients had improvements in synovitis and 57% showed improvements in bone marrow edema (79). Importantly, future use of RBC-lysed BMAC, rather than complete BMAC, may improve prognosis and therapeutic potential in both human and animal clinical OA patients.

Comparison of DEGs between treatment groups demonstrated that 19 genes were found to be directly related to the removal of RBCs from BMAC. Several gene signatures emerged as being particularly important to resolution of inflammation in the joint environment. For example, IL1RN [interleukin-1 receptor antagonist (IRAP)] was upregulated in the RBC-depleted BMAC treated samples. IL1RN has long been used to treat OA (80) and slows the rate of disease progression by blocking interleukin (IL)-1 induced inflammation and loss of synovial structural integrity (81,82) The most highly upregulated gene within specific to removal of RBCs was IL-19, which exhibits anti-inflammatory characteristics through polarization of local t-cells and macrophages, as well as decreasing inflammatory monocyte infiltration through decreased CCR2 expression (83). The HSC marker CD34 was also upregulated following RBC removal, indicative of likely increased survival of bone marrow stem cells (HSC) within the injected joints. It is not clear at this time the role that HSC may play in recovery of joint health, but their role bears investigation.

Finally, upregulation of musculoaponeurotic fibrosarcoma (MAF) and PD-L2 genes were noted in mice receiving RBC-depleted BMAC (but not in BMAC-treated animals), both of which are important in suppressing tissue inflammation. Molecular patterns in the joint after the removal of erythrocytes also indicated decreased expression of several genes, specifically S100a8, TNFRSF1B and CCR2, which have been linked to the degradation levels of cartilage and bone. S100a8 has been associated to bone erosion and decreased collagen deposition (84-86). TNFRSF1B is a transmembrane receptor in the tumor necrosis factor receptor superfamily that has been associated with inflammation in OA joints and reduced cartilage repair processes (87). CCR2, a receptor for monocyte chemotaxis, limits the inflammatory monocyte influx when reduced; ablation of CCR2 has been previously shown to decrease joint pain in mouse OA models (88). In summary, the transcriptome analysis revealed that removal of RBCs from BMAC significantly reduced joint inflammatory responses, including upregulation of genes associated with reduction of joint inflammation. Treatment with BMAC by contrast induced upregulated expression of a number of pro-inflammatory genes, and reduced expression of genes associated with cartilage healing.

Caveats to this study are worthy of mentioning, including the use of only one BMAC cell concentration for injection and the relatively short 28-day study duration. While significant differences in gait parameters were seen between treatment groups by day 27, additional information may have been gained by allowing OA to progress for a more prolonged period. In. addition, both male and female animals were pooled for analysis in this study after no changes were seen with preliminary analysis utilizing sex as a factor. However, previous studies have shown that gender can be an important influencer of disease progression in rodent models of OA (63). The study also employed a single BMAC injection whereas repeated injections may have demonstrated greater improvement (83,84). Examination of the major cell fractions present in BMAC may also yield further insights into the mechanism(s) of action and lead to further optimization of the BMAC approach.

Conclusions

In summary, findings of this study demonstrated that removal of erythrocytes from BMAC improved outcomes in a mouse model of OA, as reflected by improvements in mobility, gait analysis, and histologic parameters compared to treatment with BMAC or no treatment. Indeed, a consistent treatment benefit from complete BMAC in this animal model was not seen, particularly compared to findings demonstrated by RBC-depleted BMAC. Mechanistic studies suggested that RBC-depleted BMAC may have exhibited greater disease modification through reduced induction of pro-inflammatory cytokine expression from synovial macrophages and downregulation of inflammatory gene expression from joint tissues. Consideration therefore should be given to routine RBC removal when using BMAC, to improve efficacy and reduce inflammation and adverse events associated with injection. Further investigation of fractionated BMAC for improved treatment of OA is warranted.

Acknowledgments

The authors gratefully acknowledge the assistance of staff of the Laboratory of Animal Resources at Colorado State University and undergraduate and veterinary students who assisted with data collection.

Funding: Support for this work was provided by Animal Health and Disease (No. 19HMFPXXXXG039150001) from the USDA National Institute of Food and Agriculture, Carolyn Quan and Porter Bennett Foundation, NIH/NCATS CTSA 5TL1TR002533-02, NIH 5T32OD010437-19, and Colorado State University Veterinary Summer Scholars Program USDA Animal Health and Disease Fellowship.

Footnote

Reporting Checklist: The authors have completed the ARRIVE reporting checklist. Available at https://atm.amegroups.com/article/view/10.21037/atm-22-4256/rc

Data Sharing Statement: Available at https://atm.amegroups.com/article/view/10.21037/atm-22-4256/dss

Conflicts of Interest: All authors have completed the ICMJE uniform disclosure form (available at https://atm.amegroups.com/article/view/10.21037/atm-22-4256/coif). LMP reports that support for this work was provided by Animal Health and Disease (No. 19HMFPXXXXG039150001) from the USDA National Institute of Food and Agriculture, Carolyn Quan and Porter Bennett, NIH/NCATS CTSA 5TL1TR002533-02, NIH 5T32OD010437-19, and Colorado State University Veterinary Summer Scholars Program USDA Animal Health and Disease Fellowship. LMP reports that a provisional patent has been filed covering the fractionated bone marrow aspirate product described herein. LMP serves as a member of the advisory team for EqCell Inc., and has stock options in EqCell Inc. LC reports that a provisional patent has been filed covering the fractionated bone marrow aspirate product described herein. SD reports that a provisional patent has been filed covering the fractionated bone marrow aspirate product described herein and that he is a cofounder (without revenue) of a virtual startup company developing an unrelated cellular therapeutic product for treatment of chronic wound infections. The other authors have no conflicts of interest to declare.

Ethical Statement: The authors are accountable for all aspects of the work in ensuring that questions related to the accuracy or integrity of any part of the work are appropriately investigated and resolved. This study was approved by the Institutional Animal Care and Use Committee at Colorado State University (IACUC protocol #1017) and conducted according to the national guidelines under which the institution operates, and NIH Guidelines for the Care of Use of Laboratory Animals (8^th edition).

Open Access Statement: This is an Open Access article distributed in accordance with the Creative Commons Attribution-NonCommercial-NoDerivs 4.0 International License (CC BY-NC-ND 4.0), which permits the non-commercial replication and distribution of the article with the strict proviso that no changes or edits are made and the original work is properly cited (including links to both the formal publication through the relevant DOI and the license). See: https://creativecommons.org/licenses/by-nc-nd/4.0/.

References

1. Liu Y, Ding W, Wang HL, et al. Gut microbiota and obesity-associated osteoarthritis. Osteoarthritis Cartilage 2019;27:1257-65. [Crossref] [PubMed]
2. Zhao X, Shah D, Gandhi K, et al. Clinical, humanistic, and economic burden of osteoarthritis among noninstitutionalized adults in the United States. Osteoarthritis Cartilage 2019;27:1618-26. [Crossref] [PubMed]
3. Chahla J, Dean CS, Moatshe G, et al. Concentrated Bone Marrow Aspirate for the Treatment of Chondral Injuries and Osteoarthritis of the Knee: A Systematic Review of Outcomes. Orthop J Sports Med 2016;4:2325967115625481. [Crossref] [PubMed]
4. Kraeutler MJ, Chahla J, LaPrade RF, et al. Biologic Options for Articular Cartilage Wear (Platelet-Rich Plasma, Stem Cells, Bone Marrow Aspirate Concentrate). Clin Sports Med 2017;36:457-68. [Crossref] [PubMed]
5. Borakati A, Mafi R, Mafi P, et al. A Systematic Review And Meta-Analysis of Clinical Trials of Mesenchymal Stem Cell Therapy for Cartilage Repair. Curr Stem Cell Res Ther 2018;13:215-25. [Crossref] [PubMed]
6. Andia I, Maffulli N. New biotechnologies for musculoskeletal injuries. Surgeon 2019;17:244-55. [Crossref] [PubMed]
7. Iturriaga L, Hernáez-Moya R, Erezuma I, et al. Advances in stem cell therapy for cartilage regeneration in osteoarthritis. Expert Opin Biol Ther 2018;18:883-96. [Crossref] [PubMed]
8. Chahla J, Alland JA, Verma NN. Bone Marrow Aspirate Concentrate for Orthopaedic Use. Orthop Nurs 2018;37:379-81. [Crossref] [PubMed]
9. Harrell CR, Markovic BS, Fellabaum C, et al. Mesenchymal stem cell-based therapy of osteoarthritis: Current knowledge and future perspectives. Biomed Pharmacother 2019;109:2318-26. [Crossref] [PubMed]
10. Iijima H, Isho T, Kuroki H, et al. Effectiveness of mesenchymal stem cells for treating patients with knee osteoarthritis: a meta-analysis toward the establishment of effective regenerative rehabilitation. NPJ Regen Med 2018;3:15. [Crossref] [PubMed]
11. Li X, Wang M, Jing X, et al. Bone Marrow- and Adipose Tissue-Derived Mesenchymal Stem Cells: Characterization, Differentiation, and Applications in Cartilage Tissue Engineering. Crit Rev Eukaryot Gene Expr 2018;28:285-310. [Crossref] [PubMed]
12. Piuzzi NS, Khlopas A, Newman JM, et al. Bone Marrow Cellular Therapies: Novel Therapy for Knee Osteoarthritis. J Knee Surg 2018;31:22-6. [Crossref] [PubMed]
13. Şahin Ş, Tuncel SA, Salimi K, et al. Advanced Injectable Alternatives for Osteoarthritis. Adv Exp Med Biol 2018;1077:183-96. [Crossref] [PubMed]
14. Shah K, Zhao AG, Sumer H. New Approaches to Treat Osteoarthritis with Mesenchymal Stem Cells. Stem Cells Int 2018;2018:5373294. [Crossref] [PubMed]
15. Southworth TM, Naveen NB, Nwachukwu BU, et al. Orthobiologics for Focal Articular Cartilage Defects. Clin Sports Med 2019;38:109-22. [Crossref] [PubMed]
16. Xing D, Wang Q, Yang Z, et al. Mesenchymal stem cells injections for knee osteoarthritis: a systematic overview. Rheumatol Int 2018;38:1399-411. [Crossref] [PubMed]
17. Gaul F, Bugbee WD, Hoenecke HR Jr, et al. A Review of Commercially Available Point-of-Care Devices to Concentrate Bone Marrow for the Treatment of Osteoarthritis and Focal Cartilage Lesions. Cartilage 2019;10:387-94. [Crossref] [PubMed]
18. Fotouhi A, Maleki A, Dolati S, et al. Platelet rich plasma, stromal vascular fraction and autologous conditioned serum in treatment of knee osteoarthritis. Biomed Pharmacother 2018;104:652-60. [Crossref] [PubMed]
19. Fortier LA, Potter HG, Rickey EJ, et al. Concentrated bone marrow aspirate improves full-thickness cartilage repair compared with microfracture in the equine model. J Bone Joint Surg Am 2010;92:1927-37. [Crossref] [PubMed]
20. Chu CR, Fortier LA, Williams A, et al. Minimally Manipulated Bone Marrow Concentrate Compared with Microfracture Treatment of Full-Thickness Chondral Defects: A One-Year Study in an Equine Model. J Bone Joint Surg Am 2018;100:138-46. [Crossref] [PubMed]
21. Gobbi A, Karnatzikos G, Scotti C, et al. One-Step Cartilage Repair with Bone Marrow Aspirate Concentrated Cells and Collagen Matrix in Full-Thickness Knee Cartilage Lesions: Results at 2-Year Follow-up. Cartilage 2011;2:286-99. [Crossref] [PubMed]
22. Gobbi A, Karnatzikos G, Sankineani SR. One-step surgery with multipotent stem cells for the treatment of large full-thickness chondral defects of the knee. Am J Sports Med 2014;42:648-57. [Crossref] [PubMed]
23. Gobbi A, Scotti C, Karnatzikos G, et al. One-step surgery with multipotent stem cells and Hyaluronan-based scaffold for the treatment of full-thickness chondral defects of the knee in patients older than 45 years. Knee Surg Sports Traumatol Arthrosc 2017;25:2494-501. [Crossref] [PubMed]
24. Enea D, Cecconi S, Calcagno S, et al. One-step cartilage repair in the knee: collagen-covered microfracture and autologous bone marrow concentrate. A pilot study. Knee 2015;22:30-5. [Crossref] [PubMed]
25. Gobbi A, Whyte GP. One-Stage Cartilage Repair Using a Hyaluronic Acid-Based Scaffold With Activated Bone Marrow-Derived Mesenchymal Stem Cells Compared With Microfracture: Five-Year Follow-up. Am J Sports Med 2016;44:2846-54. [Crossref] [PubMed]
26. Krych AJ, Nawabi DH, Farshad-Amacker NA, et al. Bone Marrow Concentrate Improves Early Cartilage Phase Maturation of a Scaffold Plug in the Knee: A Comparative Magnetic Resonance Imaging Analysis to Platelet-Rich Plasma and Control. Am J Sports Med 2016;44:91-8. [Crossref] [PubMed]
27. Cavallo C, Boffa A, Andriolo L, et al. Bone marrow concentrate injections for the treatment of osteoarthritis: evidence from preclinical findings to the clinical application. Int Orthop 2021;45:525-38. [Crossref] [PubMed]
28. Jo CH, Lee YG, Shin WH, et al. Intra-articular injection of mesenchymal stem cells for the treatment of osteoarthritis of the knee: a proof-of-concept clinical trial. Stem Cells 2014;32:1254-66. [Crossref] [PubMed]
29. Kouroupis D, Ahari AF, Correa D, et al. Intralesional Injection of Bone Marrow Aspirate Concentrate for the Treatment of Osteonecrosis of the Knee Secondary to Systemic Lupus Erythematosus: A Case Report. Front Bioeng Biotechnol 2020;8:202. [Crossref] [PubMed]
30. Eder C, Schmidt-Bleek K, Geissler S, et al. Mesenchymal stromal cell and bone marrow concentrate therapies for musculoskeletal indications: a concise review of current literature. Mol Biol Rep 2020;47:4789-814. [Crossref] [PubMed]
31. El-Kadiry AE, Lumbao C, Rafei M, et al. Autologous BMAC Therapy Improves Spinal Degenerative Joint Disease in Lower Back Pain Patients. Front Med (Lausanne) 2021;8:622573. [Crossref] [PubMed]
32. Kim JD, Lee GW, Jung GH, et al. Clinical outcome of autologous bone marrow aspirates concentrate (BMAC) injection in degenerative arthritis of the knee. Eur J Orthop Surg Traumatol 2014;24:1505-11. [Crossref] [PubMed]
33. Schäfer R, DeBaun MR, Fleck E, et al. Quantitation of progenitor cell populations and growth factors after bone marrow aspirate concentration. J Transl Med 2019;17:115. [Crossref] [PubMed]
34. Sugaya H, Yoshioka T, Kato T, et al. Comparative Analysis of Cellular and Growth Factor Composition in Bone Marrow Aspirate Concentrate and Platelet-Rich Plasma. Bone Marrow Res 2018;2018:1549826. [Crossref] [PubMed]
35. Lyons LP, Weinberg JB, Wittstein JR, et al. Blood in the joint: effects of hemarthrosis on meniscus health and repair techniques. Osteoarthritis Cartilage 2021;29:471-9. [Crossref] [PubMed]
36. SWANTON MC. Hemophilic arthropathy in dogs. Lab Invest 1959;8:1269-77. [PubMed]
37. Madhok R, York J, Sturrock RD. Haemophilic arthritis. Ann Rheum Dis 1991;50:588-91. [Crossref] [PubMed]
38. Valentino LA, Hakobyan N, Enockson C. Blood-induced joint disease: the confluence of dysregulated oncogenes, inflammatory signals, and angiogenic cues. Semin Hematol 2008;45:S50-7. [Crossref] [PubMed]
39. Melchiorre D, Manetti M, Matucci-Cerinic M. Pathophysiology of Hemophilic Arthropathy. J Clin Med 2017;6:63. [Crossref] [PubMed]
40. Roosendaal G, Lafeber FP. Pathogenesis of haemophilic arthropathy. Haemophilia 2006;12:117-21. [Crossref] [PubMed]
41. van Meegeren ME, Roosendaal G, Jansen NW, et al. Blood-Induced Joint Damage: The Devastating Effects of Acute Joint Bleeds versus Micro-Bleeds. Cartilage 2013;4:313-20. [Crossref] [PubMed]
42. Roosendaal G, Vianen ME, van den Berg HM, et al. Cartilage damage as a result of hemarthrosis in a human in vitro model. J Rheumatol 1997;24:1350-4. [PubMed]
43. Hooiveld MJ, Roosendaal G, van den Berg HM, et al. Haemoglobin-derived iron-dependent hydroxyl radical formation in blood-induced joint damage: an in vitro study. Rheumatology (Oxford) 2003;42:784-90. [Crossref] [PubMed]
44. Roosendaal G, Vianen ME, Marx JJ, et al. Blood-induced joint damage: a human in vitro study. Arthritis Rheum 1999;42:1025-32. [Crossref] [PubMed]
45. Hooiveld M, Roosendaal G, Wenting M, et al. Short-term exposure of cartilage to blood results in chondrocyte apoptosis. Am J Pathol 2003;162:943-51. [Crossref] [PubMed]
46. Jansen NW, Roosendaal G, Bijlsma JW, et al. Exposure of human cartilage tissue to low concentrations of blood for a short period of time leads to prolonged cartilage damage: an in vitro study. Arthritis Rheum 2007;56:199-207. [Crossref] [PubMed]
47. Roosendaal G, Vianen ME, Wenting MJ, et al. Iron deposits and catabolic properties of synovial tissue from patients with haemophilia. J Bone Joint Surg Br 1998;80:540-5. [Crossref] [PubMed]
48. Morris CJ, Wainwright AC, Steven MM, et al. The nature of iron deposits in haemophilic synovitis. An immunohistochemical, ultrastructural and X-ray microanalytical study. Virchows Arch A Pathol Anat Histopathol 1984;404:75-85. [Crossref] [PubMed]
49. Hakobyan N, Kazarian T, Jabbar AA, et al. Pathobiology of hemophilic synovitis I: overexpression of mdm2 oncogene. Blood 2004;104:2060-4. [Crossref] [PubMed]
50. Wen FQ, Jabbar AA, Chen YX, et al. c-myc proto-oncogene expression in hemophilic synovitis: in vitro studies of the effects of iron and ceramide. Blood 2002;100:912-6. [Crossref] [PubMed]
51. Roivainen A, Söderström KO, Pirilä L, et al. Oncoprotein expression in human synovial tissue: an immunohistochemical study of different types of arthritis. Br J Rheumatol 1996;35:933-42. [Crossref] [PubMed]
52. Øvlisen K, Kristensen AT, Jensen AL, et al. IL-1 beta, IL-6, KC and MCP-1 are elevated in synovial fluid from haemophilic mice with experimentally induced haemarthrosis. Haemophilia 2009;15:802-10. [Crossref] [PubMed]
53. Acharya SS, Kaplan RN, Macdonald D, et al. Neoangiogenesis contributes to the development of hemophilic synovitis. Blood 2011;117:2484-93. [Crossref] [PubMed]
54. McNulty AL, Rothfusz NE, Leddy HA, et al. Synovial fluid concentrations and relative potency of interleukin-1 alpha and beta in cartilage and meniscus degradation. J Orthop Res 2013;31:1039-45. [Crossref] [PubMed]
55. Swärd P, Frobell R, Englund M, et al. Cartilage and bone markers and inflammatory cytokines are increased in synovial fluid in the acute phase of knee injury (hemarthrosis)--a cross-sectional analysis. Osteoarthritis Cartilage 2012;20:1302-8. [Crossref] [PubMed]
56. Tajima T, Sekimoto T, Yamaguchi N, et al. Hemoglobin stimulates the expression of ADAMTS-5 and ADAMTS-9 by synovial cells: a possible cause of articular cartilage damage after intra-articular hemorrhage. BMC Musculoskelet Disord 2017;18:449. [Crossref] [PubMed]
57. Soontararak S, Chow L, Johnson V, et al. Mesenchymal Stem Cells (MSC) Derived from Induced Pluripotent Stem Cells (iPSC) Equivalent to Adipose-Derived MSC in Promoting Intestinal Healing and Microbiome Normalization in Mouse Inflammatory Bowel Disease Model. Stem Cells Transl Med 2018;7:456-67. [Crossref] [PubMed]
58. Soontararak S, Chow L, Johnson V, et al. Humoral immune responses against gut bacteria in dogs with inflammatory bowel disease. PLoS One 2019;14:e0220522. [Crossref] [PubMed]
59. Chow L, Johnson V, Coy J, et al. Mechanisms of Immune Suppression Utilized by Canine Adipose and Bone Marrow-Derived Mesenchymal Stem Cells. Stem Cells Dev 2017;26:374-89. [Crossref] [PubMed]
60. Chow L, Johnson V, Regan D, et al. Safety and immune regulatory properties of canine induced pluripotent stem cell-derived mesenchymal stem cells. Stem Cell Res 2017;25:221-32. [Crossref] [PubMed]
61. Colbath AC, Dow SW, Phillips JN, et al. Autologous and Allogeneic Equine Mesenchymal Stem Cells Exhibit Equivalent Immunomodulatory Properties In Vitro. Stem Cells Dev 2017;26:503-11. [Crossref] [PubMed]
62. Colbath AC, Dow SW, Hopkins LS, et al. Allogeneic vs. autologous intra-articular mesenchymal stem cell injection within normal horses: Clinical and cytological comparisons suggest safety. Equine Vet J 2020;52:144-51. [Crossref] [PubMed]
63. Ma HL, Blanchet TJ, Peluso D, et al. Osteoarthritis severity is sex dependent in a surgical mouse model. Osteoarthritis Cartilage 2007;15:695-700. [Crossref] [PubMed]
64. Glasson SS, Blanchet TJ, Morris EA. The surgical destabilization of the medial meniscus (DMM) model of osteoarthritis in the 129/SvEv mouse. Osteoarthritis Cartilage 2007;15:1061-9. [Crossref] [PubMed]
65. Glasson SS, Chambers MG, Van Den Berg WB, et al. The OARSI histopathology initiative - recommendations for histological assessments of osteoarthritis in the mouse. Osteoarthritis Cartilage 2010;18:S17-23. [Crossref] [PubMed]
66. Boss MK, Ke Y, Bian L, et al. Therapeutic Intervention Using a Smad7-Based Tat Protein to Treat Radiation-Induced Oral Mucositis. Int J Radiat Oncol Biol Phys 2022;112:759-70. [Crossref] [PubMed]
67. Jennions MD, Moller AP. A survey of the statistical power of research in behavioral ecology and animal behavior. Behavioral Ecology 2003;14:438-45. [Crossref]
68. Mi H, Thomas P. PANTHER pathway: an ontology-based pathway database coupled with data analysis tools. Methods Mol Biol 2009;563:123-40. [Crossref] [PubMed]
69. Mi H, Ebert D, Muruganujan A, et al. PANTHER version 16: a revised family classification, tree-based classification tool, enhancer regions and extensive API. Nucleic Acids Res 2021;49:D394-403. [Crossref] [PubMed]
70. Bosio CM, Dow SW. Francisella tularensis induces aberrant activation of pulmonary dendritic cells. J Immunol 2005;175:6792-801. [Crossref] [PubMed]
71. Creamer P. Osteoarthritis pain and its treatment. Curr Opin Rheumatol 2000;12:450-5. [Crossref] [PubMed]
72. Molstad DHH, Bradley EW. Pain and Activity Measurements. Methods Mol Biol 2021;2221:291-9. [Crossref] [PubMed]
73. El-Kadiry AE, Lumbao C, Salame N, et al. Bone marrow aspirate concentrate versus platelet-rich plasma for treating knee osteoarthritis: a one-year non-randomized retrospective comparative study. BMC Musculoskelet Disord 2022;23:23. [Crossref] [PubMed]
74. Anz AW, Hubbard R, Rendos NK, et al. Bone Marrow Aspirate Concentrate Is Equivalent to Platelet-Rich Plasma for the Treatment of Knee Osteoarthritis at 1 Year: A Prospective, Randomized Trial. Orthop J Sports Med 2020;8:2325967119900958. [Crossref] [PubMed]
75. Lakes EH, Allen KD. Gait analysis methods for rodent models of arthritic disorders: reviews and recommendations. Osteoarthritis Cartilage 2016;24:1837-49. [Crossref] [PubMed]
76. Jacobs BY, Kloefkorn HE, Allen KD. Gait analysis methods for rodent models of osteoarthritis. Curr Pain Headache Rep 2014;18:456. [Crossref] [PubMed]
77. Allen KD, Adams SB, Setton LA. Evaluating intra-articular drug delivery for the treatment of osteoarthritis in a rat model. Tissue Eng Part B Rev 2010;16:81-92. [Crossref] [PubMed]
78. Birn-Jeffery AV, Higham TE. The scaling of uphill and downhill locomotion in legged animals. Integr Comp Biol 2014;54:1159-72. [Crossref] [PubMed]
79. Goncars V, Kalnberzs K, Jakobsons E, et al. Treatment of Knee Osteoarthritis with Bone Marrow-Derived Mononuclear Cell Injection: 12-Month Follow-up. Cartilage 2019;10:26-35. [Crossref] [PubMed]
80. Baltzer AW, Ostapczuk MS, Stosch D, et al. A new treatment for hip osteoarthritis: clinical evidence for the efficacy of autologous conditioned serum. Orthop Rev (Pavia) 2013;5:59-64. [PubMed]
81. Mehta S, Akhtar S, Porter RM, et al. Interleukin-1 receptor antagonist (IL-1Ra) is more effective in suppressing cytokine-induced catabolism in cartilage-synovium co-culture than in cartilage monoculture. Arthritis Res Ther 2019;21:238. [Crossref] [PubMed]
82. Nixon AJ, Grol MW, Lang HM, et al. Disease-Modifying Osteoarthritis Treatment With Interleukin-1 Receptor Antagonist Gene Therapy in Small and Large Animal Models. Arthritis Rheumatol 2018;70:1757-68. [Crossref] [PubMed]
83. Autieri MV. IL-19 and Other IL-20 Family Member Cytokines in Vascular Inflammatory Diseases. Front Immunol 2018;9:700. [Crossref] [PubMed]
84. Zreiqat H, Belluoccio D, Smith MM, et al. S100A8 and S100A9 in experimental osteoarthritis. Arthritis Res Ther 2010;12:R16. [Crossref] [PubMed]
85. Schelbergen RF, Blom AB, van den Bosch MH, et al. Alarmins S100A8 and S100A9 elicit a catabolic effect in human osteoarthritic chondrocytes that is dependent on Toll-like receptor 4. Arthritis Rheum 2012;64:1477-87. [Crossref] [PubMed]
86. Ruan G, Xu J, Wang K, et al. Associations between serum S100A8/S100A9 and knee symptoms, joint structures and cartilage enzymes in patients with knee osteoarthritis. Osteoarthritis Cartilage 2019;27:99-105. [Crossref] [PubMed]
87. Chisari E, Yaghmour KM, Khan WS. The effects of TNF-alpha inhibition on cartilage: a systematic review of preclinical studies. Osteoarthritis Cartilage 2020;28:708-18. [Crossref] [PubMed]
88. Ishihara S, Obeidat AM, Wokosin DL, et al. The role of intra-articular neuronal CCR2 receptors in knee joint pain associated with experimental osteoarthritis in mice. Arthritis Res Ther 2021;23:103. [Crossref] [PubMed]