Skip to main content
NCBI home page
NIHPA Author Manuscripts logoLink to NIHPA Author Manuscripts
. Author manuscript; available in PMC: 2026 Jun 5.
Published in final edited form as: Connect Tissue Res. 2025 Dec 3;67(2):157–173. doi: 10.1080/03008207.2025.2593877

Flow cytometric identification of dynamic immune cell populations in a rat model of post-traumatic elbow contracture

Rebecca F Reals 1,*, J Caleb Snider 2,*, Ryan M Castile 2, Chelsey L Dunham 3, Necat Havlioglu 4, Paul C Cannon 5, Benjamin M Zmistowski 6, Aaron M Chamberlain 7, Matthew R Bersi 1,2,#, Spencer P Lake 1,2,6,#
PMCID: PMC13233149  NIHMSID: NIHMS2176039  PMID: 41334767

Abstract

Purpose/Aim of Study:

Post-traumatic joint contracture (PTJC) commonly occurs after elbow injury. Previous findings implied immune system activation in capsules of contracted joints; however, quantifying immune cell populations in rodent joint tissues is challenging due to small size and low cellularity. Here, we used flow cytometry to investigate the temporal immune response and enumerate cell populations in the rat elbow capsule after traumatic injury.

Materials and Methods:

After inducing PTJC, capsules were harvested from injured, sham, and control rat elbows at multiple time points, stained with surface markers for immune cells, and quantified via flow cytometry. Results were compared to previously published mechanics and histology data from the same model. Another injured group was treated with celecoxib to determine if changes due to anti-inflammatory treatment could be detected.

Results:

Compared to control, injured animals displayed elevated leukocytes, T cells, and natural killer cells. CD45+ cells exhibited similar temporal changes as mechanics, which increased and then decreased after injury. CD3+, CD4+, and CD8a+ T cells followed a similar pattern as histology scores, which increased and remained elevated. Treating injured animals with celecoxib increased leukocytes but decreased several immune subpopulations.

Conclusions:

A method for flow cytometry on rat elbow capsule was established and used to quantify immune cell populations, which changed in response to injury and anti-inflammatory treatment. Comparisons between flow cytometry and previously published mechanics and histology revealed additional insights about temporal patterns in cell-, tissue-, and joint-level changes. Future work will investigate whether changes in immune cells attenuate PTJC symptoms.

Keywords: elbow joint, post-traumatic joint contracture, immune response, flow cytometry, T cells, anti-inflammatory, celecoxib

Introduction

Post-traumatic joint contracture (PTJC) affects up to 50% of patients after a traumatic elbow injury1 and may result in debilitating pain, joint stiffness, and reduced range of motion2. Current treatments – such as physical therapy, capsular release surgery, or total elbow arthroplasty – fail to restore the full elbow range of motion in up to 12% of patients3. Furthermore, none of the current treatment strategies address the underlying biological mechanisms that cause the disease. Elbow PTJC is closely associated with thickening, adhesions, and fibrosis of the joint capsule, which was previously established via histology in our preclinical model of elbow PTJC in the rat4. Previous work also demonstrated that the anterior capsule was a major contributor to contracture after injury5.

In general, inflammation and activation of the immune system is an essential part of joint contracture development after traumatic injury6,7. Although the initial inflammatory response is necessary for the healing and regeneration of injured tissue8, abnormal or excessive inflammation may result in fibrosis which can lead to contracture6. For example, inflammatory cells such as monocytes and mast cells have been implicated in knee contracture9 and inflammatory gene expression was increased in the capsule of immobilized rat10 and rabbit11 knee contracture models. Moreover, previous work with our elbow PTJC model identified elevated circulating levels of the immune cell recruitment ligand monocyte chemoattractant protein-1 (MCP-1) after injury12. Immune cell recruitment, persistence, and biological activity following traumatic injury orchestrates healing and can thereby influence patient outcomes. A persistently increased inflammatory environment that does not resolve may cause chronic musculoskeletal soft tissue injury. Indeed, pro-inflammatory “M1” macrophages, natural killer cells, neutrophils, and TH1-polarized T cells are associated with increased severity of osteoarthritis13,14, rheumatoid arthritis15, and tendon disorders16.

While specific immune cell contributions to elbow PTJC have not yet been elucidated, several studies have shown that T cells contribute to fibrosis and joint contracture. For example, one study of fibrotic human knee capsules concluded that a T cell-mediated immune response initiated capsulitis and scar tissue formation leading to the development of arthrofibrosis17. Another study demonstrated that T cells mediated a positive feedback loop with fibroblasts leading to collagen deposition, inflammatory cytokine production, and further immune recruitment to joint capsules from patients with frozen shoulder, a fibrotic condition with similarities to PTJC18. Finally, tendon samples taken from patients with rotator cuff tears confirmed an increased presence of T cells and revealed a self-amplificatory loop between T cells and tenocytes, resulting in upregulation of inflammatory cytokines and adhesion molecules16.

T cell activity has been shown to increase expression of cyclooxygenase 2 (COX-2) , an enzyme that contributes to inflammation and pain19, in the joint capsule. Selective inhibition of COX-2 signaling by non-steroidal anti-inflammatory drugs (NSAIDs; e.g., celecoxib), which are commonly prescribed for the treatment of musculoskeletal pain and stiffness, can lead to decreased inflammation and reduced expression of effector molecules such as IL-17A, a pro-inflammatory cytokine produced by a subset of helper T cells19,20. Celecoxib treatment of primary synovial fibroblasts from patients with rheumatoid arthritis similarly resulted in a dose-dependent decrease in pro-inflammatory mediators19. Celecoxib also decreased mRNA expression of genes associated with fibrogenesis and may therefore promote improved collagen remodeling and decreased joint stiffness21,22. Inhibiting COX-2 has also shown promise in preclinical models of knee contracture, resulting in decreased joint adhesions and increased range of motion23,24.

Overall, these studies provide a link between inflammation, immune cell activity, and the development of fibrosis and joint contracture that warrants further investigation. However, elbow injuries are understudied compared to other joints25 and thus the temporal nature of the immune response and relative proportion of immune cells present during elbow contracture are poorly understood. Therefore, the purpose of this study was to establish a flow cytometry-based approach to investigate the time-dependent expression of immune cells in rat elbow joint capsule following traumatic injury. Our flow cytometry technique quantified several immune cell populations following capsule injury and even captured subtle increases in a sham group compared to unoperated controls. Additionally, some immune cell populations in the injured capsule were selectively decreased after broad-spectrum anti-inflammatory treatment (i.e., celecoxib), further demonstrating the sensitivity of the proposed methodology and ultimately showcasing the development of a platform to test immunomodulatory therapies for elbow PTJC using our preclinical rat model.

Materials and Methods

Injury Model and Experimental Design

In this IACUC-approved study, injuries mimicking an elbow dislocation with PTJC were surgically induced in the left elbows of male Long-Evans rats (200-250 g; 5-7 weeks old; Charles River Laboratories International, Wilmington, MA) according to an established protocol (Figure 1A,B)26. Briefly, the soft tissue damage that occurs during a simple elbow dislocation injury in humans was replicated with an anterior capsulotomy and lateral collateral ligament (LCL) transection (Figure 1C). The injured forelimb was then immobilized with bandages until sacrifice26. Injured animals received either no treatment (INJ; n = 3-6 per time point at 3, 7, 14, or 21 days post-surgery) or oral celecoxib treatment (CEL; n = 4 per time point at 3 or 7 days post-surgery). Celecoxib was administered orally (50 mg/kg in a liquid sweetener) via daily syringe feeding beginning on the day of surgery. Sham animals (SHAM; n = 3-6 per time point at 3, 7, 14, or 21 days post-surgery) received an incision through the skin and muscle of the left elbow, without anterior capsulotomy or transection of the LCL, followed by immobilization with bandages26. Age-matched control animals (CTRL; n = 10) received neither surgery nor immobilization. Separate groups of CTRL animals were not added for each time point due to ethical considerations to minimize the number of animals included in the study. CTRL animals were also not expected to exhibit significant differences in immune cell populations over 21 days in the absence of surgery and immobilization. In a previous study27, animals subjected to the same injury model were sacrificed at 3, 7, or 21 days post-injury (n = 8 per time point) and evaluated via post-mortem joint mechanical testing and histological analysis. Prior biomechanical and microstructural findings were correlated to a subset of data from the current study to provide additional functional insights into the flow cytometry results.

Figure 1:

Figure 1:

Open in a new tab

(A) A representative hematoxylin and eosin-stained sagittal histological section of the elbow of an uninjured control CTRL animal, with inset showing the normal appearance of the anterior capsule tissue. (B) A representative sagittal section from an injured INJ animal which was immobilized for 42 days, then sacrificed. Inset shows abnormal anterior capsule tissue exhibiting thickening, fibrosis, and adhesions to the bone. (C) Schematic representation of a sagittal view of the elbow joint showing the approximate locations of the anterior capsule and lateral collateral ligament (LCL) relative to other anatomical structures (not drawn to scale). The injury model replicates soft tissue injuries observed in a simple elbow dislocation using a LCL transection and anterior capsule puncture. (D) Schematic representation of the rat forelimb with skin and muscle removed for visualization of the anterior capsule, alongside a photo taken during dissection. (E) Schematic representation of the rat forelimb with skin, muscle, and anterior capsule removed for visualization of the underlying articulating surfaces, alongside a photo taken during dissection.

Flow Cytometry

Animals were sacrificed at either 3, 7, 14, or 21 days post-surgery via CO2 inhalation; anterior capsules (Figure 1D,E) and spleens were removed and immediately dissociated into single cell suspensions. To assess local immune responses to injury, anterior capsules were harvested, finely minced, and incubated in type 2 filtered collagenase (750 μL, Worthington Biochemical, LS004204) at 37°C. After 7 minutes, the supernatant containing liberated cells was collected and fresh collagenase was replaced on the remaining tissue for another 7-minute incubation at 37°C; this process was repeated eight times28 or until the tissue was fully digested. Collected digestion buffer was further diluted with FACS buffer (PBS + 3% FBS) to reduce enzymatic activity. To assess systemic T cell immune responses associated with joint injury, spleens were placed in FACS buffer prior to being finely minced, pressed between glass slides, filtered through a 100 μm filter, and centrifuged at 300 g for 5 min. The dissociated splenocyte solution was then incubated with red blood cell lysis buffer for 10 minutes at room temperature (ThermoFisher, 00-4300-54). For each tissue preparation, cell suspensions were filtered through a 40 μm filter, centrifuged, and incubated with Fc block (5 μg/mL CD32, BD Biosciences, 550270) prior to antibody staining. Cells were labeled with DAPI (ThermoFisher D1306, 1:500,000) to identify non-viable cells, then stained with conjugated antibodies in two different staining panels as shown in Table 1. Panel 1 was comprised of FITC His48 (Clone His48, eBioscience 11-0570-82, 1:100), PE CD161 (Clone 3.2.3, BioLegend 205604, 1:100), APC CD43 (Clone W3/13, BioLegend 202814, 1:100), and APC-Cy7 CD45 (Clone OX-1, BioLegend 202216, 1:100). Panel 2 was comprised of PE CD3 (Clone 1F4, BioLegend 201411, 1:100), PerCP CD8a (Clone OX-8, BioLegend 201712, 1:100), APC CD4 (Clone W3/25, BioLegend 201509, 1:100), and APC-Cy7 CD45 (Clone OX-1, BioLegend 202216, 1:100). Fluorescence minus one (FMO) controls for each surface marker were completed separately for spleen (Supplemental Figure 1) and capsule (Supplemental Figure 2) tissue. Compensation of each fluorescent channel was conducted using UltraComp compensation beads (Invitrogen 01-3333-42). Cells were analyzed with an Agilent Novocyte 3000, and gating was performed in FlowJo (Tree Star Inc, USA). Using this combination of surface stains adapted from a previous study29, we were able to identify the following immune cell populations: CD45+ hematopoietic cells, CD161+ natural killer (NK) cells, granulocytes (His48+ with high side scatter granularity), CD3+CD4+ helper T cells, CD3+CD8a+ cytotoxic T cells, and two monocyte subsets based on CD43 expression where His48HiCD43Lo monocytes represent alternatively activated-monocytes and His48Lo-IntCD43Hi represents classically-activated monocytes.

Table 1:

Antibodies used in two different flow cytometry panels, where panel 1 is focused on myeloid cell types (e.g., natural killer cells, granulocytes, and monocytes) and panel 2 is focused on lymphoid cell types (e.g., T cell subsets).

Panel Antibody Fluorochrome Supplier Clone Dilution
1 DAPI Pacific Blue ThermoFisher N/A 1:500,000
CD43 APC Biolegend W3/13 1:100
CD45 APC-Cy7 Biolegend OX-1 1:100
CD161 PE Biolegend 3.2.3 1:100
His48 FITC eBioscience His48 1:100
2 DAPI Pacific Blue ThermoFisher N/A 1:500,000
CD3 PE Biolegend 1F4 1:100
CD4 APC Biolegend W3/25 1:100
CD8a PerCP Biolegend OX-8 1:100
CD45 APC-Cy7 Biolegend OX-1 1:100
Open in a new tab

To analyze the composition of immune cells after injury and minimize spectral overlap of fluorochromes, samples were divided in half and subjected to two separate panels of antibody staining. Gating for panel 1 (Figure 2) began by separating cells from debris (Figure 2A,H), live cells from dead cells (Figure 2B,I), and single cells from doublets (Figure 2C,J). Single, viable cells were then separated into CD45+ and CD45− populations (Figure 2D,K) to delineate CD45+ leukocytes, which were further separated into CD161+ NK cells and CD161− cells (Figure 2E,L). From the CD161− population, neutrophil-like granulocytes were identified by His48+ and high granularity (Figure 2F,M). Finally, non-granulocytes were separated into two different types of monocytes: one His48Lo-IntCD43Hi population and one His48HiCD43Lo population (Figure 2G,N)29. These two populations are reported to be analogs to non-classical Ly6CLo and classical Ly6CHi murine monocytes, respectively30,31. Gating for panel 2 (Figure 3) also began by identifying single, viable cells (Figure 3A-C,G-I), and separating them into CD45+ and CD45− populations (Figure 3D,J). From the CD45+ population we then identified the CD3+ total T cell population (Figure 3E,K), which was further separated into CD4+ cells and CD8a+ cells (Figure 3F,L). Gate placement was determined using fluorescence minus one (FMO) controls (Supplemental Figures 1-2) applied first in spleen samples due to higher cellularity. Gate locations were then copied to the capsule samples prior to verification with capsule-specific FMO controls. Despite low cellularity, we were able to successfully label immune cell populations in rat elbow capsule tissue.

Figure 2:

Figure 2:

Open in a new tab

(A-G) Gating strategy for panel 1, shown with representative images from the spleen (A-G) and anterior capsule (H-N) of a control CTRL animal. Cells were separated from debris (A,H), live cells from dead cells (B,I), and single cells from doublets (C,J), before isolating CD45+ cells (D,K) which were split into CD161+ and CD161− populations (E,L). CD161− cells were separated into granulocytes and non-granulocytes (F,M), the latter of which were separated into two types of monocytes (G,N). (H-N) SSC-A = side scatter area, FSC-A = forward scatter area, FSC-H = forward scatter height.

Figure 3:

Figure 3:

Open in a new tab

Gating strategy for panel 2, shown with representative images from the spleen (A-F) and anterior capsule (G-L) of a control CTRL animal. Cells were separated from debris (A,G), live cells from dead cells (B,H), and single cells from doublets (C,I), before isolating CD45+ cells (D,J). From the CD45+ population, CD3+ cells were identified (E,K) and separated into CD4+ and CD8a+ populations (F,L). SSC-A = side scatter area, FSC-A = forward scatter area, FSC-H = forward scatter height.

Joint Mechanical Testing and Histology

In a previously published study27, the biomechanical properties of bilateral elbows from injured animals were measured via flexion-extension joint testing at 3, 7, and 21 days post-injury. Age-matched uninjured control animals were also subjected to joint mechanical testing. Measured biomechanical parameters include range of motion (ROM) difference between injured and contralateral elbows, neutral zone (NZ) length (which represents the functional ROM or the amount of motion possible before a larger force is required to move the joint further), and neutral zone (NZ) stiffness. Following mechanical testing, a subset of limbs (n = 3 per time point) was prepared for histological analysis as described previously26. Three separate 5 μm sagittal sections per elbow were cut, stained with hematoxylin and eosin, and blinded before being scored by a musculoskeletal pathologist using an established semi-quantitative scoring method32. Measured pathological scoring parameters included anterior capsule fibrosis, anterior capsule adhesions, and anterior capsule thickness, among others. The mechanical testing and histology data were each plotted against the flow cytometry data to evaluate potential relationships. Thus, these are groupwise, unpaired correlations because they are separate groups of animals.

Statistical Analysis

Flow cytometric data were processed and analyzed using FlowJo. Results are presented as mean ± standard error of mean (SEM). Statistical comparisons between injury groups (SHAM vs. INJ) and treatment groups (INJ vs. CEL) were evaluated using two-way ANOVAs with Bonferroni corrections. Because CTRL was only assessed at a single time point, statistical comparisons between CTRL and other groups (i.e., SHAM, INJ, CEL) were done using a separate one-way ANOVA with Dunnett correction at each time point. For mechanical testing data, statistical comparisons between INJ and CTRL were evaluated using a one-way ANOVA with Dunnett correction. For histological data, statistical comparisons were not performed due to the small sample size and semi-quantitative nature of the scoring method. Statistical significance for all comparisons was set as p < 0.05.

Results

Flow cytometry for control, sham, and injured animals

Overall, the proportion of immune cell populations in the elbow capsule was elevated in SHAM and INJ compared to CTRL (Figure 4; Supplemental Table 1). At day 3, INJ capsules showed an early inflammatory response characterized by elevated CD3+, CD4+, CD8a+, and CD161+ cells compared to CTRL capsules (Figure 4B-E). SHAM capsules showed some degree of inflammation over CTRL with increased CD161+ cells and decreased granulocytes (Figure 4E,F), albeit not to the extent of INJ. At day 7, CD45+, CD4+, and CD8a+ populations were elevated in INJ over CTRL (Figure 4A,C,D), while only His48HiCD43Lo monocytes were altered in SHAM relative to CTRL (Figure 4H). At day 14, both monocyte populations were elevated in INJ over CTRL (Figure 4G,H), while only CD4+ cells were increased in SHAM over CTRL (Figure 4C). At day 21, persistent elevation of CD3+ and CD4+ cells were observed in INJ compared to CTRL capsules (Figure 4B,C), whereas only His48HiCD43Lo monocytes were elevated in SHAM over CTRL (Figure 4H). Thus, the immune cell landscape of SHAM and INJ capsules were altered compared to CTRL; but, as expected due to the varying degrees of injury severity, SHAM capsules exhibited less pronounced differences than INJ.

Figure 4:

Figure 4:

Open in a new tab

(A-H) Flow cytometric analysis of anterior capsules from control (CTRL; black dotted line in gray shaded region), sham (SHAM; blue), and injured (INJ; purple) animals. Cell populations identified were (A) CD45+ immune cells, (B) CD3+ T cells, (C) CD4+ helper T cells, (D) CD8a+ cytotoxic T cells, (E) CD161+ natural killer cells, (F) granulocytes, (G) His48Lo-IntCD43Hi monocytes, and (H) His48HiCD43Lo monocytes. CTRL, SHAM, and INJ are shown as mean ± standard error of the mean. Statistical significance was set at p < 0.05. * = significant difference between CTRL and SHAM, # = significant difference between CTRL and INJ, $ = significant difference between SHAM and INJ. Results of two-way ANOVA that were significant are shown on the plots as either time (TIME), injury (INJ), or interaction (INT). Sample size for CTRL is n = 10, SHAM is n = 3 at Day 3, Day 7, and Day 21 and n = 6 at Day 14, and INJ is n = 3 at Day 3, Day 7, and Day 21 and n = 6 at Day 14.

Within each immune cell population, two-way ANOVAs were used to identify any overall group- (i.e., INJ vs. SHAM) or time-dependent effects (Supplemental Table 1). The effects of both time and injury were significant for CD8a+ cells, and the effects of time, injury, and their interaction were significant for CD161+ cells. Notably, the CD161+ percentage in INJ capsules increased compared to SHAM at 3 days but quicky returned to SHAM levels by day 7 (Figure 4E). Analysis of other cell types demonstrated that on average, INJ capsules tended to show slightly elevated immune cell populations compared to SHAM even though these differences did not reach statistical significance. INJ capsules displayed elevated trends of leukocyte infiltration (CD45+) for 14 days before returning to SHAM levels by 21 days (Figure 4A). We also observed a persistently elevated pattern for CD3+ and CD4+ T cells, but not CD8a+ T cells (Figure 4D), at 21 days after injury (Figure 4B-C). Interestingly, granulocytes were generally downregulated immediately after surgery and later returned to CTRL levels in SHAM but not INJ groups (Figure 4F). Monocytes (Figure 4G-H) in both SHAM and INJ were overall slightly elevated compared to CTRL.

Splenic tissue was also analyzed as an indicator of systemic immune response to elbow injury (Figure 5; Supplemental Table 2). In general, the spleen displayed fewer injury-induced differences from CTRL than capsule tissue. Namely, at days 3 and 21, neither INJ nor SHAM had any significantly different populations from CTRL; at days 7 and 14 only CD45+ cells were elevated over CTRL (Figure 5A). This suggests the surgical procedure, independent of capsule and ligament injury, induced a delayed splenic immune response. Two-way ANOVAs to identify group- (i.e., INJ vs. SHAM) or time-dependent effects showed no significant group or interaction terms; however, CD45+, His48Lo-IntCD43Hi, His48HiCD43Lo, and granulocyte populations changed significantly over time (Figure 5A,F-H).

Figure 5:

Figure 5:

Open in a new tab

Flow cytometric analysis of spleens from control (CTRL; black dotted line in gray shaded region), sham (SHAM; blue), and injured (INJ; purple) animals. Cell populations identified were (A) CD45+ immune cells, (B) CD3+ T cells, (C) CD4+ helper T cells, (D) CD8a+ cytotoxic T cells, (E) CD161+ natural killer cells, (F) granulocytes, (G) His48Lo-IntCD43Hi monocytes, and (H) His48HiCD43Lo monocytes. CTRL, SHAM, and INJ are shown as mean ± standard error of the mean. Statistical significance was set at p < 0.05. * = significant difference between CTRL and SHAM, # = significant difference between CTRL and INJ, $ = significant difference between SHAM and INJ. Results of two-way ANOVA that were significant are shown on the plots as either time (TIME), injury (INJ), or interaction (INT). Sample size is for CTRL is n = 10, SHAM is n = 3 at Day 3, Day 7, and Day 21 and n = 6 at Day 14, and INJ is n = 3 at Day 3, Day 7, and Day 21 and n = 6 at Day 14.

Mechanical and histological data

Previously collected post-mortem mechanical testing data following elbow injury showed several significant differences. Namely, INJ elbows showed a decreased ROM compared to CTRL at day 21 (Figure 6A). NZ length increased compared to CTRL at day 3 followed by a subsequent significant decrease at day 21 (Figure 6B). NZ stiffness was slighly decreased in INJ tissues at day 3 and significantly increased in INJ tissues compared to CTRL at day 21 (Figure 6C). This initial increase in joint motion and decrease in joint stiffness at day 3, followed by decreased joint motion and increased stiffness at day 21, mimics the longitudinal patterns of CD45+ cell expression in that CD45+ cells initially increased at Day 3 and then gradually decreased over time. Correlation analysis resulted in U-shaped (Figure 6D and E) and V-shaped (Figure 6F) relationships between mechanical parameters and CD45+ cell populations. In contrast, the persistent increase in CD3+, CD4+, and CD8a+ T cells resulted in plateaued correlation analysis plots (Figure 6G-O) when compared to mechanical data.

Figure 6:

Figure 6:

Open in a new tab

Joint biomechanical metrics27 and groupwise correlations with flow cytometry results. (A) Range of motion (ROM) difference between injured INJ and contralateral CTRL elbows, (B) neutral zone (NZ) length, and (C) neutral zone (NZ) stiffness show dynamic changes over post-injury time. Statistical significance (*) indicated for p < 0.05. Correlations show (D-F) an initial increase in CD45+ cells at 3 days post-injury followed by a delayed shift in joint mechanics as CD45+ cells return towards control levels at later time points. T cell markers CD3+ (G-I), CD4+ (J-L), and CD8a+ (M-O) showed increased quantities at 3 days that remained generally elevated as ROM difference, NZ length, and NZ stiffness became increasingly altered by 21 days post-injury. (Dashed lines and arrows indicate time progression of data post-injury).

Semi-quantitative histological scoring was conducted on INJ samples and correlated with different immune populations present in capsule tissue. Anterior capsule fibrosis (Figure 7A) and adhesion (Figure 7B) scores were increased at all time points in INJ samples compared to CTRL (which received a score of zero for both criteria) but did not change over time after injury. Anterior capsule thickness scoring was also sharply increased at all time points in INJ samples compared to CTRL but gradually decreased over time after injury (Figure 7C). An initial increase and gradual decrease in CD45+ cells, combined with a sustained increase in histological scores for (Figure 7D) fibrosis and (Figure 7E) adhesions resulted in U-shaped correlation plots that plateau on one side. Persistently increased capsule fibrosis and adhesion scores in INJ samples compared to CTRL began at day 3 and mirrored the temporal expression pattern of all three T cell subpopulations (i.e., CD3+, CD4+ and CD8a+). These similarities were visualized via INJ data clustering (Figure 7G, H, J, K, M, and N). The decrease in capsule thickness score from day 3 to day 21 mimics the CD45+ leukocyte expression pattern thereby resulting in U-shaped (Figure 7F) and inverted U-shaped plateaued correlation plots (Figure 7I, L, O).

Figure 7:

Figure 7:

Open in a new tab

Histological scores27 for anterior capsule and groupwise correlations with flow cytometry results. Scores for anterior capsule (A) fibrosis and (B) adhesion were zero for control CTRL animals but elevated at all post-injury time points. (C) Anterior capsule thickness increased sharply from control levels at 3-days post-injury then decreased at later time points. Correlations showed an initial increase in CD45+ cells and corresponding increase of histological scores that remain elevated for (D) fibrosis and (E) adhesions but gradually decreased for (F) capsule thickness as the number of CD45+ cells returned towards control levels at later time points. Correlations of histology metrics with (G-I) CD3+, (J-L) CD4+, and (M-O) CD8a+ cells showed increases post-injury that remained consistently elevated for (G, J, M) fibrosis and (H, K, N) adhesions but that decreased at later times for (I, L, O) capsule thickness, even though T-cell populations remained high. (Dashed lines and arrows indicate time progression of data post-injury).

Effect of drug treatment on injured animals

Having established the ability to detect immune cell populations from capsule tissue, we next attempted to modulate inflammation using treatment with the NSAID celecoxib (CEL; 50 mg/kg/day). Early changes in immune cell populations were measured at 3 and 7 days post-surgery to determine whether early treatment can effectively at modulate immune cell populations in the elbow capsule. In this experimental design, CEL samples were compared to untreated INJ and CTRL samples (Supplemental Table 3). Interestingly, we observed an increase in the percentage of CD45+ cells in CEL capsules relative to CTRL at both days 3 and 7 (Figure 8A). At day 7, CD8a+ and His48HiCD43Lo monocytes in CEL-treated capsules were significantly decreased and increased over CTRL, respectively (Figure 8D,H). A two-way ANOVA was used to evaluate treatment (i.e., INJ vs. CEL) and time effects in injured capsules. CEL decreased several immune populations relative to INJ including CD3+, CD4+, CD8a+, CD161+, and His48Lo-IntCD43Hi cells (Figure 8B-E,G); however, the effect of time was only significant for CD8a+ and CD161+ cells (Figure 8D,E; Supplemental Table 3). Post-hoc tests further identified CEL-induced decreases in CD4+, CD8a+, CD161+, and His48Lo-IntCD43Hi cells at day 3 (Figure 8C-E,G; Supplemental Table 3) and decreases in CD4+ and CD8a+ T cells at day 7 compared to INJ.

Figure 8:

Figure 8:

Open in a new tab

Flow cytometric analysis of anterior capsules from control (CTRL; black dotted line in gray shaded region) and injured animals (INJ; purple) with and without celecoxib treatment (CEL; green). Cell populations identified were (A) CD45+ immune cells, (B) CD3+ T cells, (C) CD4+ helper T cells, (D) CD8a+ cytotoxic T cells, (E) CD161+ natural killer cells, (F) granulocytes, (G) His48Lo-IntCD43Hi monocytes, and (H) His48HiCD43Lo monocytes. CTRL, SHAM, and CEL are shown as mean ± standard error of the mean. Statistical significance was set at p < 0.05. ^ = significant difference between INJ and CEL, & = significant difference between CTRL and CEL. Results of two-way ANOVA that were significant are shown on the plots as either time (TIME), treatment (TRT), or interaction (INT). Sample size is for CTRL is n = 10, INJ is n = 3 at Day 3 and Day 7 and CEL is n = 4 at Day 3 and Day 7.

Celecoxib treatment also induced changes in splenic immune cells compared to untreated samples (Figure 9; Supplemental Table 4). CEL-treated spleens showed a decrease in CD161+ cells relative to CTRL at day 3 and INJ at days 3 and 7 (Figure 9E). There was also an increase in CD8a+ cells after CEL treatment at day 7 relative to both CTRL and INJ (Figure 9D). While the CD45+ population was similar between CEL and INJ at day 3, the percentage of leukocytes was reduced in CEL spleens by day 7 (Figure 9A). CEL also reduced both CD45+ and CD161+ cells in the spleen. While the effect of time was significant for monocytes and granulocytes, the interaction between treatment and time was only significant for CD45+ leukocyte expression (Figure 9A,E-H; Supplemental Table 4).

Figure 9:

Figure 9:

Open in a new tab

Flow cytometric analysis of spleens from control animals (CTRL; black dotted line in gray shaded region) and injured animals (INJ; purple) with and without celecoxib treatment (CEL; green). Cell populations identified were (A) CD45+ immune cells, (B) CD3+ T cells, (C) CD4+ helper T cells, (D) CD8a+ cytotoxic T cells, (E) CD161+ natural killer cells, (F) granulocytes, (G) His48Lo-IntCD43Hi monocytes, and (H) His48HiCD43Lo monocytes. CTRL, SHAM, and CEL are shown as mean ± standard error of the mean. Statistical significance was set at p < 0.05. ^ = significant difference between INJ and CEL. Results of two-way ANOVA that were significant are shown on the plots as either time (TIME), treatment (TRT), or interaction (INT). Sample size is for CTRL is n = 10, INJ is n = 3 at Day 3 and Day 7 and CEL is n = 4 at Day 3 and Day 7.

Discussion

Inflammation has previously been shown to be present in our rat model of elbow PTJC, as well as other models of joint contracture, and thus may represent a potential target for treatment and mitigation of fibrosis. A previous study using our PTJC model found that the INJ anterior capsule exhibited increased inflammation, fibrosis, and adhesion to osseous surfaces compared to CTRL27. Similarly, a murine model of elbow PTJC reported capsule thickening, fibrosis, and chronic inflammation33. A study utilizing a rabbit model of knee contracture proposed that inflammation plays a key role in capsule fibrosis after treatment with the mast cell stabilizer ketotifen, which was found to improve cellular and biomechanical manifestations of fibrosis34. These observations are supported by multiple studies that have identified inflammatory cell infiltration and upregulation of inflammatory mediator genes in the injured joint capsule of a rat model of knee contracture; treatment with the anti-inflammatory drug dexamethasone inhibited the increase of inflammatory markers in this model35-37. The current standard of care for elbow contracture is physical therapy or surgery38,39, but these treatments are sometimes ineffective6 and do not address underlying biological mechanisms. Targeting the inflammatory response as an adjunctive therapy may limit the development of fibrosis in the joint, attenuate contracture progression, and result in a better outcome independently or when paired with physical therapy.

In this study, we developed a flow cytometry approach for the first characterization of immune cell populations present in the rat elbow capsule following traumatic joint injury. With two different panels, we identified the presence of CD45+ leukocytes, CD3+ T cells, CD4+ helper T cells, CD8a+ cytotoxic T cells, CD161+ natural killer cells, neutrophil-like granulocytes, and two subpopulations of monocytes analogous to classically- and alternatively-activated monocytes (as defined in mice and humans). Despite working with a small tissue of low cellularity with high levels of extracellular matrix deposition, we were able to successfully isolate single cell suspensions from elbow capsules, thus establishing a technique which could be adapted to investigate the effect of various interventions on inflammation in traumatic elbow injury.

Comparison of flow cytometric profiles between groups demonstrated altered immune cell populations in INJ compared to SHAM. CD45+ leukocytes showed a sustained, although not statistically significant, increase at multiple time points over 14 days in INJ compared to SHAM with resolution of inflammation present by day 21 (Figure 4A). The only statistically significant difference between INJ and SHAM occurred in CD161+ natural killer cells at day 3 (Supplemental Table 1). Indeed, the two-way ANOVA between INJ and SHAM showed that the effect of injury was only significant in CD8a+ and CD161+ cells, which may be attributed to either small sample sizes (n = 3-6 animals/group) or the relatively small number of cells isolated from each capsule (approximately 50,000-100,000 cells on average). Functionally, NK cells respond to stressed or damaged tissue by coordinating inflammation through macrophage, T cell, or dendritic cell interactions40; these cells are present abundantly and early in injured capsules. Interestingly, the monocyte populations detected after injury were similar between SHAM and INJ capsules, yet both were increased relative to CTRL (Figure 4G,H). Most of the capsule monocyte infiltration is likely due to surgical exposure of the joint and subsequent limb immobilization rather than the tissue-level damage that occurs in INJ only. It is important to note that the sham procedure implemented in the current study is relatively severe, with incisions made through the entire depth of the skin and muscle down to the joint tissues, and includes the same type of immobilization as INJ. Thus, the only difference between SHAM and INJ was the LCL transection and anterior capsule puncture. Previous studies using this model also reported that SHAM displayed differences compared to CTRL26,41. Mechanically, SHAM exhibited increased NZ stiffness and decreased ROM compared to CTRL, while histologically, SHAM capsule exhibited similar histological characteristics as INJ, such as increased adhesions, cellularity, and thickness relative to CTRL. Thus, it was unsurprising that the current study found few statistical differences between SHAM and INJ. However, the significant changes that do arise may reveal a cell-specific target for addressing joint contracture that would previously have been masked by general inflammation.

Comparisons between flow cytometry data and previously published joint mechanics and histology data27 provided additional context and revealed important insights for comparing INJ to CTRL. At 3 days post-injury, all T cell populations (CD3+, CD4+, and CD8a+) and NK cells were increased, and corresponding histological measures of corresponding capsular properties (i.e., fibrosis, adhesions, thickness) were also increased (Figure 7). Thus, cell- and tissue-level changes were both observed at the earliest time point, even though joint-level mechanical properties were unchanged (or slightly increased) at 3 days and only became impaired after 21 days (Figure 6). These observations suggest early joint instability induced by the injury (i.e., LCL transection and anterior capsule puncture) and a delay between the onset of cell/tissue disruption and subsequent disruption of joint mechanics. By 14 days post-injury, CD8+ T cells (Figure 4D) and NK cells (Figure 4E) returned to baseline or CTRL levels. Interestingly, CD3+ and CD4+ T cell populations remained elevated throughout the 21-day study period, similar to histological scores of capsule fibrosis and adhesions, while capsule thickness and CD45+ leukocyte expression both decreased over time (Figure 7). Decreasing capsule thickness over time may demonstrate extracellular matrix compaction wherein the tissue becomes more dense and stiffer, thereby restricting motion over time. While overall immune cell populations may have decreased in the capsule, the persistently elevated T cells (specifically helper T cells) after injury may indicate a pivotal cell population governing the healing response. If so, this observation suggests that T cells are a potential target for immunomodulatory treatment, in agreement with previous studies in similar soft tissue fibrosis injury models14,16,18,42-45. Taken together, these results suggest cell- and tissue-level changes occur first after traumatic injury and are followed by joint-level changes. However, it is important to note that mechanical testing captures overall joint-level changes that include contributions from other peri-articular joint tissues (e.g., posterior capsule, ligaments, tendons, etc.), while the flow cytometry and histological data focused only on the anterior capsule.

As T cells in the capsule showed notable differences between groups, we also investigated whether the composition of the spleen was altered following injury. The spleen acts as a T cell reservoir and site for T cell specialization that has been shown to respond to other joint injuries (e.g., increasing proportions of splenic T cells in osteoarthritis46. Furthermore, protein kinase inhibitors blocking immune cell signaling by the spleen have been a therapeutic target for rheumatoid arthritis47. However, flow cytometric analysis of spleens in our model of PTJC showed no significant differences between SHAM and INJ up to 21 days after injury (Figure 5, Supplemental Table 2). Since capsule injury did not result in chronic systemic inflammation, as shown by decreasing CD45+ leukocyte expression over time (Figure 4A), it is possible that this acute injury was not sufficient to promote measurable changes in the splenic immune landscape. One difference of note is the significant increase in CD45+ leukocytes in SHAM and INJ spleens compared to CTRL (Figure 5A). This suggests the sham procedure induced a generalized immune response in the spleen while not altering the specific makeup of leukocyte populations.

In addition to baseline immunological characterization of the rat elbow capsule, a subset of animals was treated with celecoxib (CEL) to diminish the early immune response to injury and evaluate our approach as a method to quantify changes associated with anti-inflammatory or immunomodulatory treatment strategies targeting elbow PTJC. An unexpected observation in CEL capsules was an increased fraction of CD45+ cells relative to untreated INJ (Figure 8A). A potential explanation for this could be altered crosstalk between immune cells and stromal cells that promotes fibroblast activity and deposition of extracellular matrix16. By attenuating inflammation with celecoxib, capsule fibroblasts may be prevented from expanding to a greater proportion of the cells present after injury. In the current study, our primary focus was on immune cell enumeration and thus we did not evaluate extracellular matrix remodeling of the capsule after injury or drug treatment. However, examining subpopulations of interest revealed a decrease in nearly every other type of immune cell in CEL-treated capsules relative to INJ including CD3+, CD4+, CD8a+, CD161+, and His48Lo-IntCD43Hi cells (Figure 8B-G). Because the overall proportion of leukocytes increased in CEL capsules, our observations of decreased immune subsets suggest another population not directly measured here, such as B cells or mast cells, must have increased to account for the increasing number of leukocytes. B cells are known to have pathogenic effects in rheumatoid arthritis48 and osteoarthritis49, which may be the subject of future studies. This is further supported by analysis of spleens in CEL rats. We observed minimal alterations in the spleen at 3 days after injury, but the increase in CD45+ cells observed in INJ capsules at day 7 was prevented by CEL treatment (Figure 9A). While we saw marginal decreases in CD4+ and CD161+ cells at day 7 with CEL (Figure 9C,E), it is possible that B cell expansion was also mitigated and contributed to the decrease in the leukocyte fraction. Overall, these data suggest that our method of flow cytometry could be implemented to identify a candidate cell population which may be associated with maladaptive remodeling and further expand our understanding of inflammation-driven wound healing in traumatic elbow injury.

The current study is not without limitations. First, the small size and low cellularity of the rat elbow capsule limited the number of detectable single viable cells by flow cytometry and may have contributed to sample-to-sample variability. Second, only male rats were included in the study to eliminate cyclical variations in hormone levels from consideration. While it is possible that female rats have a different immune response to injury, a previous study50 using this model demonstrated similar functional, mechanical, and morphological properties for males and females. In addition, the use of only male rats in this study was sufficient to demonstrate the ability to detect alterations in immune populations following elbow injury. Third, celecoxib is a broad-spectrum anti-inflammatory drug that does not selectively block the differentiation, recruitment, or proliferation of specific cell populations measured in this study. As a COX-2 inhibitor, celecoxib decreases prostaglandin production, but future studies may selectively inhibit specific cell types – such as T cells, NK cells, or monocytes – for a more targeted treatment approach. Fourth, there are immune cell populations that have not been evaluated here, such as B cells, mast cells, that have been shown to contribute to other joint diseases and should be characterized in future work with this PTJC model. Fifth, we performed flow cytometric analysis of the elbow capsule and spleen but did not investigate other sources of immune cells – such as the blood, bone marrow, lymph nodes, or thymus – which may better characterize systemic immune responses and provide insight into whether the observed immune cell alterations stem from increased production or recruitment at sites of injury. Lastly, mechanics and histology data are from different rats than those used for flow cytometry because flow cytometry requires the capsule to be removed from the elbow immediately after sacrifice. This limits the ability to perform full correlations, but the groupwise comparisons performed in the current study still provide valuable insights.

Although flow cytometry of the rat elbow capsule is novel, this technique has been used extensively to evaluate immune cell populations in other musculoskeletal joint tissues. For example, flow cytometry has been used to investigate immune cell populations in various rodent and human hip, knee, ankle, and shoulder tissues13-16,18,42,43,51-61. Due to the ability to multiplex multiple surface stains for the same cells, flow cytometry has proven, in this study and others, to be a useful technique for probing the relative proportions of immune cell populations in normal and diseased tissues. Our development of an approach to analyze the rat elbow capsule by flow cytometry allows for comparisons to be made with other studies that investigate the immune response after joint injury and clarify the functional impact of immunomodulatory treatments.

PTJC occurs as a consequence of immune-mediated fibrosis of the joint tissues including the anterior capsule. Findings from this study show that flow cytometry can be used to quantify several immune populations in the rat elbow capsule and spleen and that these populations were altered in response to sham surgery, injury, or anti-inflammatory drug treatment. Ongoing work will investigate how alterations in the proportion of immune cell populations after injury impacts mechanical or functional improvements, such as increased joint range of motion or changes in gait patterns.

Supplementary Material

Supplemental Figure 2

Supplemental Figure 2: Comparison between fluorescence minus one (FMO) controls (top row) and fully stained samples (bottom row) for panel 1 (A) and panel 2 (B). Graphs are representative images from the anterior capsule of a control CTRL animal. FSC-A = forward scatter area.

Supplemental Figure 1

Supplemental Figure 1: Comparison between fluorescence minus one (FMO) controls (top row) and fully stained samples (bottom row) for panel 1 (A) and panel 2 (B). Graphs are representative images from the spleen of a control CTRL animal. FSC-A = forward scatter area.

3

Acknowledgements:

The authors thank Dr. Jai Rudra for allowing us to collect data using his laboratory’s flow cytometer. This research was supported in part by the Intramural Research Program of the National Institutes of Health (NIH). The contributions of the NIH author were made as part of their official duties as an NIH federal employee, are in compliance with agency policy requirements, and are considered works of the United States Government. However, the findings and conclusions presented in this paper are those of the author and do not necessarily reflect the views of the NIH or the U.S. Department of Health and Human Services.

Funding:

This work was supported by the NIH under Grant R01 AR071444.

This work was supported by the NIH under Grant T32AR060719.

This work was supported by the Monticello College Foundation under the Spencer T. and Ann W. Olin Fellowship for Women in Graduate Study at Washington University in St. Louis.

Footnotes

Declaration of Interests:

Dr. Aaron Chamberlain: DePuy - consulting, royalties; Zimmer-Biomet - research support; Tigon Medical – royalties.

Dr. Benjamin Zmistowski: Zimmer-Biomet – consulting.

References

  • 1.Anakwe RE, Middleton SD, Jenkins PJ, McQueen MM, Court-Brown CM. 2011. Patient-Reported Outcomes After Simple Dislocation of the Elbow. JBJS. 93(13):1220. doi: 10.2106/JBJS.J.00860 [DOI] [PubMed] [Google Scholar]
  • 2.Magermans DJ, Chadwick EKJ, Veeger HEJ, van der Helm FCT. 2005. Requirements for upper extremity motions during activities of daily living. Clin Biomech. 20(6):591–599. doi: 10.1016/j.clinbiomech.2005.02.006 [DOI] [PubMed] [Google Scholar]
  • 3.Myden C, Hildebrand K. 2011. Elbow joint contracture after traumatic injury. J Shoulder Elbow Surg. 20(1):39–44. doi: 10.1016/j.jse.2010.07.013 [DOI] [PubMed] [Google Scholar]
  • 4.David MA, Reiter AJ, Dunham CL, Castile RM, Abraham JA, Iannucci LE, Shah ID, Havlioglu N, Chamberlain AM, Lake SP. 2022. Pleiotropic Effects of Simvastatin and Losartan in Preclinical Models of Post-Traumatic Elbow Contracture. Front Bioeng Biotechnol. 10:803403. doi: 10.3389/fbioe.2022.803403 [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 5.Dunham CL, Castile RM, Chamberlain AM, Lake SP. 2019. The Role of Periarticular Soft Tissues in Persistent Motion Loss in a Rat Model of Posttraumatic Elbow Contracture. J Bone Joint Surg Am. 101(5):e17. doi: 10.2106/JBJS.18.00246 [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 6.Hildebrand KA, Ademola A, Hart DA. 2021. Nonsurgical treatments for post-traumatic elbow contractures: approaches for the prevention of their development and progression. Ann Jt 6(0). doi: 10.21037/aoj-20-62 [DOI] [Google Scholar]
  • 7.KANEGUCHI A, OZAWA J. 2022. Inflammation and Fibrosis Induced by Joint Remobilization, and Relevance to Progression of Arthrogenic Joint Contracture: A Narrative Review. Physiol Res. 71(4):447–455. doi: 10.33549/physiolres.934876 [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 8.Arvind V, Huang AH. 2021. Reparative and Maladaptive Inflammation in Tendon Healing. Front Bioeng Biotechnol 9. https://www.frontiersin.org/articles/10.3389/fbioe.2021.719047 [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 9.Monument MJ, Hart DA, Salo PT, Befus AD, Hildebrand KA. 2013. Posttraumatic elbow contractures: targeting neuroinflammatory fibrogenic mechanisms. J Ortho Sci. 18(6):869–877. doi: 10.1007/s00776-013-0447-5 [DOI] [PubMed] [Google Scholar]
  • 10.Yabe Y, Hagiwara Y, Suda H, Ando A, Onoda Y, Tsuchiya M, Hatori K, Itoi E. 2013. Joint immobilization induced hypoxic and inflammatory conditions in rat knee joints. Connect Tissue Res. 54(3):210–217. doi: 10.3109/03008207.2013.786056 [DOI] [PubMed] [Google Scholar]
  • 11.Morrey ME, Abdel MP, Riester SM, Dudakovic A, van Wijnen AJ, Morrey BF, Sanchez-Sotelo J. 2017. Molecular landscape of arthrofibrosis: Microarray and bioinformatic analysis of the temporal expression of 380 genes during contracture genesis. Gene. 610:15–23. doi: 10.1016/j.gene.2017.01.025 [DOI] [PubMed] [Google Scholar]
  • 12.Reiter A. 2021. Post-traumatic Elbow Contracture Characterization and Physical Therapy-based Treatment Strategies in a Preclinical Model. McKelvey Sch Eng Theses Diss. doi: 10.7936/amn9-vj29 [DOI] [Google Scholar]
  • 13.Benigni G, Dimitrova P, Antonangeli F, Sanseviero E, Milanova V, Blom A, van Lent P, Morrone S, Santoni A, Bernardini G. 2017. CXCR3/CXCL10 Axis Regulates Neutrophil-NK Cell Cross-Talk Determining the Severity of Experimental Osteoarthritis. J Immunol Baltim Md 1950. 198(5):2115–2124. doi: 10.4049/jimmunol.1601359 [DOI] [PubMed] [Google Scholar]
  • 14.Wang W, Lin X, Xu H, Sun W, Bouta EM, Zuscik MJ, Chen D, Schwarz EM, Xing L. 2019. Attenuated Joint Tissue Damage Associated With Improved Synovial Lymphatic Function Following Treatment With Bortezomib in a Mouse Model of Experimental Posttraumatic Osteoarthritis. Arthritis Rheumatol Hoboken NJ. 71(2):244–257. doi: 10.1002/art.40696 [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 15.Yudoh K, Matsuno H, Nakazawa F, Yonezawa T, Kimura T. 2000. Reduced expression of the regulatory CD4+ T cell subset is related to Th1/Th2 balance and disease severity in rheumatoid arthritis. Arthritis Rheum. 43(3):617–627. doi: 10.1002/1529-0131(200003)43:3<617::AID-ANR19>3.0.CO;2-B [DOI] [PubMed] [Google Scholar]
  • 16.Garcia-Melchor E, Cafaro G, MacDonald L, Crowe LAN, Sood S, McLean M, Fazzi UG, McInnes IB, Akbar M, Millar NL. 2021. Novel self-amplificatory loop between T cells and tenocytes as a driver of chronicity in tendon disease. Ann Rheum Dis. 80(8):1075–1085. doi: 10.1136/annrheumdis-2020-219335 [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 17.Bosch U, Zeichen J, Skutek M, Haeder L, van Griensven M. 2001. Arthrofibrosis is the result of a T cell mediated immune response. Knee Surg Sports Traumatol Arthrosc Off J ESSKA. 9(5):282–289. doi: 10.1007/s001670100218 [DOI] [PubMed] [Google Scholar]
  • 18.Akbar M, Crowe LAN, McLean M, Garcia-Melchor E, MacDonald L, Carter K, Fazzi UG, Martin D, Arthur A, Reilly JH, et al. 2021. Translational targeting of inflammation and fibrosis in frozen shoulder: Molecular dissection of the T cell/IL-17A axis. Proc Natl Acad Sci U S A. 118(39):e2102715118. doi: 10.1073/pnas.2102715118 [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 19.Paulissen SMJ, van Hamburg JP, Davelaar N, Asmawidjaja PS, Hazes JMW, Lubberts E. 2013. Synovial fibroblasts directly induce Th17 pathogenicity via the cyclooxygenase/prostaglandin E2 pathway, independent of IL-23. J Immunol Baltim Md 1950. 191(3):1364–1372. doi: 10.4049/jimmunol.1300274 [DOI] [PubMed] [Google Scholar]
  • 20.van den Berg WB, McInnes IB. 2013. Th17 cells and IL-17 a--focus on immunopathogenesis and immunotherapeutics. Semin Arthritis Rheum. 43(2):158–170. doi: 10.1016/j.semarthrit.2013.04.006 [DOI] [PubMed] [Google Scholar]
  • 21.Limberg AK, Tibbo ME, Salib CG, McLaury AR, Turner TW, Berry CE, Jay AG, Carter JM, Bolon B, Berry DJ, et al. 2020. Reduction of arthrofibrosis utilizing a collagen membrane drug-eluting scaffold with celecoxib and subcutaneous injections with ketotifen. J Orthop Res Off Publ Orthop Res Soc. 38(11):2474–2483. doi: 10.1002/jor.24647 [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 22.Salib CG, Reina N, Trousdale WH, Limberg AK, Tibbo ME, Jay AG, Robin JX, Turner TW, Jones CR, Paradise CR, et al. 2019. Inhibition of COX-2 pathway as a potential prophylaxis against arthrofibrogenesis in a rabbit model of joint contracture. J Ortho Res. 37(12):2609–2620. doi: 10.1002/jor.24441 [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 23.LI F, HE B, LIU S, FAN C. 2013. Celecoxib effectively inhibits the formation of joint adhesions. Exp Ther Med. 6(6):1507–1511. doi: 10.3892/etm.2013.1336 [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 24.Trousdale WH, Limberg AK, Reina N, Salib CG, Thaler R, Dudakovic A, Berry DJ, Morrey ME, Sanchez-Sotelo J, van Wijnen A, Abdel MP. 2022. Intra-articular celecoxib improves knee extension regardless of surgical release in a rabbit model of arthrofibrosis. Bone Jt Res. 11(1):32–39. doi: 10.1302/2046-3758.111.BJR-2021-0546.R1 [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 25.Lake SP. 2021. Emerging trends in elbow injury, pathology and treatment. Ann Jt. 6(0). doi: 10.21037/aoj-2019-ei-09 [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 26.Lake SP, Castile RM, Borinsky S, Dunham CL, Havlioglu N, Galatz LM. 2016. Development and use of an animal model to study post-traumatic stiffness and contracture of the elbow. J Ortho Res. 34(2):354–364. doi: 10.1002/jor.22981 [DOI] [PubMed] [Google Scholar]
  • 27.Dunham CL, Castile RM, Havlioglu N, Chamberlain AM, Lake SP. 2018. Temporal Patterns of Motion in Flexion-extension and Pronation-supination in a Rat Model of Posttraumatic Elbow Contracture. Clin Orthop. 476(9):1878–1889. doi: 10.1097/CORR.0000000000000388 [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 28.Miller LJ, Lincoln J. 2014. Isolation of murine valve endothelial cells. J Vis Exp JoVE.(90):51860. doi: 10.3791/51860 [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 29.Barnett-Vanes A, Sharrock A, Birrell MA, Rankin S. 2016. A Single 9-Colour Flow Cytometric Method to Characterise Major Leukocyte Populations in the Rat: Validation in a Model of LPS-Induced Pulmonary Inflammation. PLoS ONE. 11(1):e0142520. doi: 10.1371/journal.pone.0142520 [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 30.Strauss-Ayali D, Conrad SM, Mosser DM. 2007. Monocyte subpopulations and their differentiation patterns during infection. J Leukoc Biol. 82(2):244–252. doi: 10.1189/jlb.0307191 [DOI] [PubMed] [Google Scholar]
  • 31.Sunderkötter C, Nikolic T, Dillon MJ, Van Rooijen N, Stehling M, Drevets DA, Leenen PJM. 2004. Subpopulations of mouse blood monocytes differ in maturation stage and inflammatory response. J Immunol Baltim Md 1950. 172(7):4410–4417. doi: 10.4049/jimmunol.172.7.4410 [DOI] [PubMed] [Google Scholar]
  • 32.Reiter AJ, Schott HR, Castile RM, Cannon PC, Havlioglu N, Chamberlain AM, Lake SP. 2023. Early Joint Use Following Elbow Dislocation Limits Range-of-Motion Loss and Tissue Pathology in Posttraumatic Joint Contracture. J Bone Joint Surg Am. 105(3):223–230. doi: 10.2106/JBJS.22.00064 [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 33.Moore-Lotridge SN, Oelsner WK, Ihejirika Y, Desai MJ, Gebhart SS, Schoenecker JG. 2018. Novel preclinical murine model of trauma-induced elbow stiffness. J Exp Orthop. 5:36. doi: 10.1186/s40634-018-0155-3 [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 34.Monument MJ, Hart DA, Befus AD, Salo PT, Zhang M, Hildebrand KA. 2010. The mast cell stabilizer ketotifen fumarate lessens contracture severity and myofibroblast hyperplasia: a study of a rabbit model of posttraumatic joint contractures. J Bone Joint Surg Am. 92(6):1468–1477. doi: 10.2106/JBJS.I.00684 [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 35.Kaneguchi A, Ozawa J, Kawamata S, Yamaoka K. 2017. Development of arthrogenic joint contracture as a result of pathological changes in remobilized rat knees. J Orthop Res. 35(7):1414–1423. doi: 10.1002/jor.23419 [DOI] [PubMed] [Google Scholar]
  • 36.Kaneguchi A, Ozawa J, Yamaoka K. 2018. Anti-inflammatory Drug Dexamethasone Treatment During the Remobilization Period Improves Range of Motion in a Rat Knee Model of Joint Contracture. Inflammation. 41(4):1409–1423. doi: 10.1007/s10753-018-0788-5 [DOI] [PubMed] [Google Scholar]
  • 37.Kaneguchi A, Ozawa J, Minamimoto K, Yamaoka K. 2020. Nitric oxide synthase inhibitor L-NG-nitroarginine methyl ester (L-NAME) attenuates remobilization-induced joint inflammation. Nitric Oxide. 96:13–19. doi: 10.1016/j.niox.2020.01.003 [DOI] [PubMed] [Google Scholar]
  • 38.Papadopoulos D, Papatheodorou LK, Sotereanos DG. 2021. Elbow contracture release. Ann Jt 6(0). doi: 10.21037/aoj-19-180 [DOI] [Google Scholar]
  • 39.Reiter AJ, Castile RM, Schott HR, Kivitz GJ, Chamberlain AM, Lake SP. 2021. Investigating the Effects of Physical Therapy Timing, Intensity and Duration on Post-Traumatic Joint Contracture in a Rat Elbow Model. Muscles Ligaments Tendons J. 11(3):547–553. doi: 10.32098/mltj.03.2021.20 [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 40.Vivier E, Tomasello E, Baratin M, Walzer T, Ugolini S. 2008. Functions of natural killer cells. Nat Immunol. 9(5):503–510. doi: 10.1038/ni1582 [DOI] [PubMed] [Google Scholar]
  • 41.Dunham CL, Castile RM, Havlioglu N, Chamberlain AM, Galatz LM, Lake SP. 2017. Persistent motion loss after free joint mobilization in a rat model of post-traumatic elbow contracture. J Shoulder Elbow Surg. 26(4):611–618. doi: 10.1016/j.jse.2016.09.059 [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 42.Blomgran P, Blomgran R, Ernerudh J, Aspenberg P. 2016. A possible link between loading, inflammation and healing: Immune cell populations during tendon healing in the rat. Sci Rep. 6:29824. doi: 10.1038/srep29824 [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 43.Blomgran P, Hammerman M, Aspenberg P. 2017. Systemic corticosteroids improve tendon healing when given after the early inflammatory phase. Sci Rep. 7(1):12468. doi: 10.1038/s41598-017-12657-0 [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 44.Howell KL, Kaji DA, Li TM, Montero A, Yeoh K, Nasser P, Huang AH. 2021. Macrophage depletion impairs neonatal tendon regeneration. FASEB J Off Publ Fed Am Soc Exp Biol. 35(6):e21618. doi: 10.1096/fj.202100049R [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 45.Arvind V, Howell K, Huang AH. 2021. Reprogramming adult tendon healing using regenerative neonatal regulatory T cells. BioRxiv: 2021.05.12.443424. doi: 10.1101/2021.05.12.443424 [DOI] [Google Scholar]
  • 46.Boneva B, Ralchev N, Ganova P, Tchorbanov A, Mihaylova N. 2022. Collagenase-Induced Mouse Model of Osteoarthritis-A Thorough Flow Cytometry Analysis. Life Basel Switz. 12(11):1938. doi: 10.3390/life12111938 [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 47.Palominos PE, Lineburger IB, Xavier RM. 2021. Emerging protein kinase inhibitors for the treatment of rheumatoid arthritis. Expert Opin Emerg Drugs. 26(3):303–321. doi: 10.1080/14728214.2021.1964472 [DOI] [PubMed] [Google Scholar]
  • 48.Wu F, Gao J, Kang J, Wang X, Niu Q, Liu J, Zhang L. 2021. B Cells in Rheumatoid Arthritis : Pathogenic Mechanisms and Treatment Prospects. Front Immunol. 12:750753. doi: 10.3389/fimmu.2021.750753 [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 49.Burt KG, Scanzello CR. 2023. B cells in osteoarthritis: simply a sign or a target for therapy? Osteoarthritis Cartilage. 31(9):1148–1151. doi: 10.1016/j.joca.2023.06.002 [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 50.Reiter AJ, Schott HR, Castile RM, Cannon PC, Havlioglu N, Chamberlain AM, Lake SP. 2021. Females and males exhibit similar functional, mechanical, and morphological outcomes in a rat model of posttraumatic elbow contracture. J Orthop Res. 39(9):2062–2072. doi: 10.1002/jor.24918 [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 51.Blomgran P, Blomgran R, Ernerudh J, Aspenberg P. 2017. Cox-2 inhibition and the composition of inflammatory cell populations during early and mid-time tendon healing. Muscles Ligaments Tendons J. 7(2):223–229. doi: 10.11138/mltj/2017.7.2.223 [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 52.Collins FL, Roelofs AJ, Symons RA, Kania K, Campbell E, Collie-duguid ESR, Riemen AHK, Clark SM, Bari CD. 2023. Taxonomy of fibroblasts and progenitors in the synovial joint at single-cell resolution. Ann Rheum Dis. 82(3):428–437. doi: 10.1136/ard-2021-221682 [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 53.Faryniarz DA, Bhargava M, Lajam C, Attia ET, Hannafin JA. 2006. Quantitation of estrogen receptors and relaxin binding in human anterior cruciate ligament fibroblasts. In Vitro Cell Dev Biol Anim. 42(7):176–181. doi: 10.1290/0512089.1 [DOI] [PubMed] [Google Scholar]
  • 54.Jiang N, Liu H-X, Liang H-Y, Feng X-H, Liu B-Y, Zhou Y-Y. 2021. Osteogenic differentiation characteristics of hip joint capsule fibroblasts obtained from patients with ankylosing spondylitis. Ann Transl Med. 9(4):331. doi: 10.21037/atm-20-7817 [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 55.Jones E, Churchman SM, English A, Buch MH, Horner EA, Burgoyne CH, Reece R, Kinsey S, Emery P, McGonagle D, Ponchel F. 2010. Mesenchymal stem cells in rheumatoid synovium: enumeration and functional assessment in relation to synovial inflammation level. Ann Rheum Dis. 69(2):450–457. doi: 10.1136/ard.2008.106435 [DOI] [PubMed] [Google Scholar]
  • 56.Katschke KJ, Rottman JB, Ruth JH, Qin S, Wu L, LaRosa G, Ponath P, Park CC, Pope RM, Koch AE. 2001. Differential expression of chemokine receptors on peripheral blood, synovial fluid, and synovial tissue monocytes/macrophages in rheumatoid arthritis. Arthritis Rheum. 44(5):1022–1032. doi: 10.1002/1529-0131(200105)44:5<1022::AID-ANR181>3.0.CO;2-N [DOI] [PubMed] [Google Scholar]
  • 57.Li C, Peng Z, Zhou Y, Su Y, Bu P, Meng X, Li B, Xu Y. 2021. Comprehensive analysis of pathological changes in hip joint capsule of patients with developmental dysplasia of the hip. Bone Jt Res. 10(9):558–570. doi: 10.1302/2046-3758.109.BJR-2020-0421.R2 [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 58.Manferdini C, Paolella F, Gabusi E, Silvestri Y, Gambari L, Cattini L, Filardo G, Fleury-Cappellesso S, Lisignoli G. 2016. From osteoarthritic synovium to synovial-derived cells characterization: synovial macrophages are key effector cells. Arthritis Res Ther. 18(1):83. doi: 10.1186/s13075-016-0983-4 [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 59.Nishio H, Saita Y, Kobayashi Y, Takaku T, Fukusato S, Uchino S, Wakayama T, Ikeda H, Kaneko K. 2020. Platelet-rich plasma promotes recruitment of macrophages in the process of tendon healing. Regen Ther. 14:262–270. doi: 10.1016/j.reth.2020.03.009 [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 60.Van Landuyt KB, Jones EA, McGonagle D, Luyten FP, Lories RJ. 2010. Flow cytometric characterization of freshly isolated and culture expanded human synovial cell populations in patients with chronic arthritis. Arthritis Res Ther. 12(1):R15. doi: 10.1186/ar2916 [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 61.Zhang F, Wei K, Slowikowski K, Fonseka CY, Rao DA, Kelly S, Goodman SM, Tabechian D, Hughes LB, Salomon-Escoto K, et al. 2019. Defining inflammatory cell states in rheumatoid arthritis joint synovial tissues by integrating single-cell transcriptomics and mass cytometry. Nat Immunol. 20(7):928–942. doi: 10.1038/s41590-019-0378-1 [DOI] [PMC free article] [PubMed] [Google Scholar]

Associated Data

This section collects any data citations, data availability statements, or supplementary materials included in this article.

Supplementary Materials

Supplemental Figure 2

Supplemental Figure 2: Comparison between fluorescence minus one (FMO) controls (top row) and fully stained samples (bottom row) for panel 1 (A) and panel 2 (B). Graphs are representative images from the anterior capsule of a control CTRL animal. FSC-A = forward scatter area.

NIHMS2176039-supplement-Supplemental_Figure_2.jpg (149.8KB, jpg)
Supplemental Figure 1

Supplemental Figure 1: Comparison between fluorescence minus one (FMO) controls (top row) and fully stained samples (bottom row) for panel 1 (A) and panel 2 (B). Graphs are representative images from the spleen of a control CTRL animal. FSC-A = forward scatter area.

NIHMS2176039-supplement-Supplemental_Figure_1.jpg (171.8KB, jpg)
3
NIHMS2176039-supplement-3.pdf (1.1MB, pdf)

RESOURCES