Angiotensin II Type 2 Receptor Modulates Synovial Macrophage Polarization by Inhibiting GRK2 Membrane Translocation in a Rat Model of Collagen-Induced Arthritis

Rheumatoid arthritis (RA) is a severe, inflammatory autoimmune disease that is associated with substantial loss of quality of life, high mortality, and affects ∼1% of the population worldwide (1). One of the hallmarks of RA is an ongoing overactivation of cells of the innate and adaptive immune system that sustains an influx of immune cells into the joints from the periphery. In this pathology, activated effector cells are recruited into the synovium and joint cavity, thereby establishing a complex network of interactions that promote the production of proinflammatory cytokines, trigger significant proliferation of fibroblast-like synoviocytes (FLS), and contribute to cartilage and bone damage (2).

The mediators of the renin–angiotensin system have been described as regulators of the cardiovascular system, with a primary role in controlling blood pressure and maintaining body fluid homeostasis. However, there is increased evidence that documents a concurrent role of angiotensin II (Ang II) as a significant mediator of inflammation through its cognate, G protein–coupled angiotensin II type 1 (AT1R) and angiotensin II type 2 receptor (AT2R) (3). Generally, Ang II is involved in the generation of oxidative stress, the contraction of blood vessels, and the promotion of inflammation and fibrosis through AT1R. Opposite physiological effects, such as a relaxation of blood vessels and anti-inflammatory, anti-fibrotic and antioxidative responses are mediated through AT2R (4).

AT2R consists of 363 aa and its gene is located at xq22-23, containing three exons and two introns (5). In an atherosclerosis disease model, the additional absence of AT2R in apolipoprotein E-null mice leads to the disease progression. On the contrary, AT2R transgenic Agtr2^(+/+) mice shows an effective remission of atherosclerosis, suggesting that AT2R may plays a protective role in this disease (6, 7). In addition AT2R receptor agonists has the pharmacological effects of relaxing vascular endothelium, inhibiting inflammation, promoting the repair and regeneration of neurons (8–10).

FLS and synovial macrophages (SMs) are the two main cellular components of the synovium (11). It has been reported widely that FLS and SMs play essential roles in the joint pathology of RA. The lining of the synovium inside the fibrous outer layer (subintima) consists of two to three layers of cells. This intimal synovial lining (intima) is composed predominantly of two cell types: SMs and FLS. It has been shown that SMs contribute to RA progression by secretion of various factors, including reactive oxygen species, NO intermediates and matrix-degrading enzymes. In addition, SMs produce different kinds of cytokines in the rheumatoid synovium that can accelerate inflammation by recruiting other immune cells and activating FLS (12). SMs exhibits a range of phenotypes and functions when encountering different factors that generally are categorized as M1 and M2 macrophages and constitute different polarization stages of macrophages. In the active phase of RA, proinflammatory M1 macrophages are increased, whereas M2 macrophages that promote inflammation resolution and tissue repair are decreased. This imbalance of M1 and M2 macrophages exacerbates the inflammatory response (13). Macrophages express receptors of the renin–angiotensin system. As a consequence, angiotensin receptors could play an important role in maintaining the autoimmune inflammatory immune response.

The G protein–coupled receptors (GPCRs) comprise a huge receptor family that includes angiotensin receptors, PG receptors, chemokine receptors, and protease-activated receptors. The physiological function of GPCRs is regulated by G protein–coupled receptor kinases (GRK). Concurrently, the affinity of GPCR and β-arrestin is significantly increased in the cytoplasm, leading to the formation of a complex and thereby terminating the signal transduction of cells (14). The affinity between the GPCRs and the G protein is reduced, resulting in an uncoupling. Among the GRK subtypes, GRK2 is emerging as the pivotal integrative scaffolding for cell motility (15). Accumulating data indicate the dysfunction and overexpression of the cell membrane GRK2 in RA patients and animal models, suggesting that GRK2 plays a critical role in inflammation and that duration of mechanical hyperalgesia could be a promising target of RA treatment (16).

Our previous results have shown that treatment with AT2R agonist ameliorated the arthritis index and histological signs in the rat model of adjuvant-induced arthritis (AIA). Furthermore, AT2R agonist inhibited the chemotaxis of AIA monocytes in vitro, possibly because of the upregulation of AT2R expression (17). As a kind of GPCRs, the function of AT2R may be regulated by GRK2. Nevertheless, the underlying mechanism remains elusive. The peripheral blood monocytes migrate into synovial tissue, differentiating into SMs. The large number of SMs in the synovial tissue of RA patients contributes to the abnormal hyperactivity of the immune system and promotes joint inflammation. Whether AT2R has an anti-inflammatory effect on SMs is still unknown. Using a rat model of collagen-induced arthritis (CIA), this study investigates how AT2R activation exerts anti-inflammatory effects by regulating GRK2 membrane translocation and macrophage polarization.

Synovial tissue samples were obtained from eight RA patients. Furthermore, peripheral blood was collected from 32 RA patients who fulfilled the criteria for RA by the American College of Rheumatology and were undergoing total knee arthroplasty (18). As controls, synovial tissue from three sex- and age-matched patients with noninflammatory knee joint diseases and peripheral blood from 15 healthy volunteers were used. All patients gave their informed consent, and the study protocol was approved by Biomedical Ethics Committee of Anhui Medical University (20160094).

Female Sprague Dawley rats aged 6–7 wk with a weight of 160 ± 10 g, were purchased from the Experimental Animal Center of Anhui Medical University. The rats were kept in a specific pathogen–free environment. All procedures were performed according to the guidelines of the Animal Care and Use Committee of Anhui Medical University and were approved by the Animal Ethics Committee of the Clinical Pharmacology Institute at Anhui Medical University (no. 20160121).

Chicken type II collagen (CCII) was obtained from Chondrex (Redmond, WA). AT2R agonist CGP42112 (catalog no. HY-12405), AT2R antagonists PD123319 (catalog no. HY-10259) and etanercept (HY-108847) were obtained from MedChemExpress. Mouse anti-human CD68 Ab (clone Y1/82A; catalog no.333804; biotin) and recombinant rat M-CSF (carrier-free, catalog no.556904) were purchased from BioLegend; Cy5-labeled donkey anti-rabbit Abs and Alexa Fluor 488–labeled streptavidin were obtained from Jackson ImmunoResearch Laboratories (catalog no.711-175-152, catalog no.016-540-084; West Grove, PA). Rabbit anti-rat AT1R was obtained from Abcam (catalog no. ab124505; Cambridge, U.K.). Rabbit anti-rat AT2R was obtained from Novus Biologicals (catalog no. NBP1-77368; Toronto, ON, Canada). Alexa Fluor 594–conjugated goat anti-mouse IgG (H+L), Alexa Fluor 594–conjugated goat anti-rabbit IgG (H+L), and polyclonal iNOS Abs were obtained from Protientech (SA00013-3/4, 18985-1-AP; Wuhan, China). Mouse monoclonal TGase2 Ab (catalog no.sc-48387) was obtained from Santa Cruz Biotechnology. The rabbit mAbs anti–phospho-p38 MAPK (Thr¹⁸⁰/Tyr¹⁸²), anti-p44/42 MAPK (Erk1/2), anti–phospho-SAPK/JNK (Thr¹⁸³/Tyr¹⁸⁵), anti-SAPK/JNK, and anti–arginase-1 were ordered from Cell Signaling Technology. The mAbs anti–phospho-p44/42 MAPK(Erk1/2) and anti-GRK2 Ab were obtained from Affinity Biosciences. FITC-mouse anti-rat CD11b (clone WT.5), PE-mouse anti-rat CD86 (clone 24F) and mouse anti-rat CD3 (FITC; clone G4.18) were products of BD Biosciences (Franklin Lakes, NJ). Alexa Fluor 647–mouse anti-rat CD163 (clone ED2) were obtained from Bio-Rad Laboratories (Hercules, CA). ELISA kits for AT1R and AT2R were obtained from RY Biological (Shanghai, China) (AT1R kits catalog no. RY-02430 and AT2R kits catalog no. RY-02432). Rat Magnetic Luminex Assay was purchased from R&D Systems for cytokine screening (catalog no. LXSARM-08).

Human synovial tissues and rat knee joints were fixed in 4% formaldehyde for 48 h. The rat knee joints were decalcified subsequently. The samples were imbedded in paraffin, cut into 4-mm-thick sections, deparaffinized, and rehydrated for H&E staining, Masson staining, and immunohistochemistry (IHC) staining. Paraffin sections were stained with H&E. For Masson staining, a commercially modified Masson’s Trichrome Stain Kit was used to determine the proliferation of FLS in synovial tissue according to the manufacturer’s protocol (catalog no. G1345; Beijing Solarbio Science & Technology). For IHC staining, the sections were incubated for 20 min in 3% H₂O₂ to block endogenous peroxidases. After Ag retrieval and quenching of endogenous peroxidase, sections were incubated with CD68 mouse mAb (1:100 dilution) at 4°C overnight, followed by 30-min incubation with a secondary anti-mouse Ab at 37°C. The staining was developed with a HRP-linked polymer detection system and was revealed using 3,3-diaminobenzidine and hematoxylin.

Human tissue samples were harvested after surgical dissection, rapidly frozen in tissue freezing medium and stored at −80°C. Sections (4 μm) were cut using a cryotome (Leica, Wetzlar, Germany), air-dried for 10 min, and fixed in acetone at −20°C. Prior to staining, sections were rehydrated in PBS/1% BSA for 60 min. Sections were incubated for 2 h with CD68 biotinylated Abs, washed three times with PBS/BSA, incubated for 1 h with streptavidin–Alexa Fluor 488, and washed three times with PBS/BSA. Subsequently, the sections were incubated with purified rabbit anti-human AT1R, AT2R, iNOS, and mouse anti-human Ab GRK2 and TGase2 for 2 h; washed three times with PBS/BSA; and finally labeled for 1 h with Alexa Fluor 594–labeled secondary Ab.

PBMCs were isolated using density gradient centrifugation. PBS (100 μl) was added for three freeze and thaw cycles that released cellular proteins. The protein concentrations were determined with a BCA protein assay. Commercial ELISA kits were used to determine AT1R and AT2R levels in PBMCs lysis according to the manufacturer’s protocol.

The rat CIA model was prepared following a previously described protocol (19). Briefly, CCII was dissolved in 0.01 M filtered acetic acid at a concentration of 4 mg/ml. CFA was prepared by suspending heat-killed Mycobacterium butyricum in liquid paraffin at 6 mg/ml. CCII was emulsified with the same volume of CFA. CIA was induced by a single intradermal injection of 0.1 ml of emulsion into the right hind metatarsal rat footpad. Another 0.1-ml emulsion was injected intradermally into the base of the tail or back at multiple sites. One week later, the rats were given a booster injection. The day of the first immunization was defined as day 0. The clinical symptoms of the rats were evaluated every 3 d, and the volume of the hind paws was determined with a paw volume meter (YLS-7A; The Academy of Medical Sciences, Shandong, China). After the onset of clinical symptoms of CIA (around day 14 after immunization), model rats were randomly divided into a control group, a vehicle group, an AT2R agonist CGP42112 group, and an etanercept group as positive control. The groups can be characterized as follows: the control group was nonarthritic, healthy control rats; the vehicle group included untreated CIA rats receiving an intra-articular injection of normal saline twice a week at 50 μl vol; CGP42112 was diluted with normal saline into 4 μg/50 μl per 200 g rat (20 μg/kg). Before use, CGP42112 was diluted further into 2 μg/50 μl per 200 g rat (10 μg/kg) and 1 μg/50 μl per 200 g rat (5 μg/kg). Each rat was injected through this route and with this volume twice a week. Etanercept was diluted into 600 μg/200 μl per 200 g rat (3 mg/kg). Each rat was treated with around 200 μl per 200 g rat of etanercept i.p. twice per week. Body weight and the secondary paw swelling were determined by paw volume meter and recorded every 4 d to evaluate treatment of arthritis.

The suspension of splenic lymphocytes was adjusted to 2 × 10⁷ cells/ml in PBS. After incubation of FITC-labeled CD3 Ab for 30 min in the dark at 4°C for 30 min, T cells were isolated by a cell sorter (BD FACSAria).

Synovial tissues of CIA rats were washed with D-HBSS and cut into 1–2-mm² pieces, then incubated with 4 ml of collagenase type IV, which was dissolved in serum-free RPMI 1640 for 1 h at 37°C. Cells were incubated with 0.25% pancreatin for 30 min at 37°C. The cell suspension was filtered by 200-μm nylon mesh and stained before analyzing it in a flow cytometer.

Each rat synovial tissue single-cell suspension was divided into two tubes for AT1R and AT2R staining (1.0 × 10⁶ cells per tube). Cells were blocked by whole rat serum for 15 min. Cells were then incubated at 4°C for 30 min with AT1R (1:100) or AT2R (1:100). Subsequently, cells were washed and incubated with Cy5–goat anti-rabbit IgG (1:200) or PE-labeled secondary Ab (1:400) in the dark at 4°C. Finally, cells were washed and blocked by whole rat serum for 15 min, centrifuged, and the serum discarded. Last, cells were incubated FITC-CD11b (1:200). For SM polarization analysis, synovial tissue single-cell suspensions (1 × 10⁶) were incubated at 4°C for 30 min with FITC-CD11b (1:200), PE-CD86 (1:400) and Alexa Fluor 647–CD163 (1:10). The samples were analyzed by flow cytometry (CytoFLEX; Beckman Coulter). Data analysis was performed using CytExpert analysis software (Beckman Coulter). The sort gating strategy incorporated live cell and singlet gates prior to gating on CD11b, then in CD11b⁺ to detect other individual markers. The staining and fluorescent compensation strategy for flow cytometry are shown in Supplemental Figs. 1 and 2.

Femurs and tibia were put in 75% ethanol for 4 min and washed with normal saline. The bones were opened with sterile scissors, and the bone marrow was flushed out with icy normal saline containing 0.1% FBS and filtered through a 200-μm nylon mesh. Bone marrow cells were cultured at a density of 5 × 10⁵ cells per cm² in RPMI 1640 with 10% FBS and 10 ng/ml M-CSF at 37°C for 4 d.

CGP42112 and PD123319 were diluted in normal saline to prepare a 0.1 M stock solution before use. Bone marrow–derived macrophages (BMDMs) were pretreated agonist with or without an antagonist in different concentrations for AT2R before being stimulated with LPS (1 μg/ml, 24 h). Free NO was detected in cell culture supernatant by Griess reagent according to the manufacturers protocol (catalog no. G2930; Promega).

T cell culture supernatant and the serum of different experimental groups of rats were analyzed using bead-based, multiplex assays (Luminex Assay Panel, catalog no. LXSARM-08; R&D Systems) according to the manufacturers protocol.

Nuclear proteins were extracted from cells as previously described (20). In short, after lysis and centrifugation, cell pellets were resuspended in 400 μl of ice-cold PBS containing 0.1% Nonidet P-40 and supplemented with a mixture of protease inhibitors (Roche, Basel, Switzerland). Lysed cells were centrifuged for 20 s at 10,000 rpm. The nuclear pellet was resuspended in 60 μl of cell lysis buffer with 1 mM PMSF and 5 mM phosphatase inhibitor, incubated on ice for 30 min, and centrifuged for 10 min at 14,000 × g. The supernatant was removed. Protein concentrations were determined with a BCA Protein Assay (Pierce, Rockford, IL).

Western blotting was carried out as described in detail previously (21). The same amount of protein (10–15 μg) per lane was separated electrophoretically by SDS-PAGE and transferred to a PVDF membrane. The membranes were incubated with a primary Ab, p-ERK1/2 (1:1000), ERK1/2 (1:1000), p-P38 (1:1000), P38 (1:1000), p-JNK (1:1000), JNK (1:1000), or β-actin (1:5000) overnight in 4°C. For nuclear protein, Lamin A/C (1:2000) was used as a loading control. After washing four times for 10 minutes in PBS containing 0.1% Tween 20, the membranes were incubated with goat anti-rabbit IgG or goat anti-mouse conjugated with HRP Ab (1:10,000) for 2 h at room temperature. Membranes were washed, incubated with Western ECL Solution for 1 min, and exposed to ImageQuant LAS 4000 Mini (GE Healthcare Life Sciences, Parramatta, NSW, Australia) for 5 s–10 min. The density of a specific band was quantified using ImageJ software.

Cytoplasm and membrane protein were extracted from BMDMs as previously described (22, 23). In short, BMDMs were lysed in cell lysis buffer and centrifuged at 14,000 rpm for 15 min at 4°C. The supernatant was total protein. The supernatant was then centrifuged at 100,000 × g for 60 min at 4°C; the supernatant was cytoplasm protein. The precipitates containing membrane proteins were resuspended in cell lysis buffer with 1 mM PMSF. The total cytoplasm and membrane protein of BMDMs was then used to detect GRK2 expression by Western blot. The flow cytometry analysis was also used to detect GRK2 membrane expression. Cells were taken by scraper, and BMDM suspensions (1.0 × 10⁶) were incubated at 4°C for 30 min with Abs for GRK2 (1:50). Subsequently, cells were washed and incubated with allophycocyanin–goat anti-mouse IgG (1:200) in the dark at 4°C. The cells incubated with allophycocyanin-labeled secondary Ab but without GRK2 primary Ab was used as blank control for gating. The allophycocyanin-positive cells were analyzed on a CytoFLEX flow cytometer (Beckman Coulter CytoFLEX). Data analysis was performed using CytExpert analysis software (Beckman Coulter). GRK2 translocation was also confirmed by immunofluorescence (IF). BMDMs were cultured in 35-mm glass-bottom dishes and fixed with 4% paraformaldehyde for 30 min at 4°C. The cells were permeabilized with 0.5% Triton X-100 for 10 min, blocked with 5% BSA for 30 min at room temperature, then incubated overnight at 4°C with monoclonal anti-GRK2 Ab in a wet chamber. After washing with PBS, the cells were incubated for 1 h at room temperature with Alexa Fluor 594–labeled secondary Abs. Images of the fluorescent signal were acquired under a TCS SP8 confocal microscopy. (Leica Microsystems, Wetzlar, Germany).

Whole-cell lysate immunoprecipitation was performed by lysing cells in the lysis buffer (25 mM HEPES, 150 mM NaCl, 5 mM EDTA, 1% Triton X-100, 10% [v/v] glycerin with protease inhibitors and phosphatase inhibitors). The lysate was kept on ice for 30 min, and insoluble material was removed by centrifugation at 12,000 × g for 15 min at 4°C. The supernatant was centrifuged at 100,000 × g for 60 min at 4°C. Cytoplasmic proteins were precleared using normal mouse IgG and Protein A/G Plus Agarose for 2 h at 4°C. For the specific staining, anti-GRK2 or normal IgG Ab were added to Protein A/G Plus Agarose for 2 h at 4°C, then precleared lysates were associated with Protein A/G Plus Agarose and anti-GRK2 or normal IgG Ab and incubated at 4°C overnight.

All statistical analyses were undertaken using SPSS 16.0. All quantitative data were expressed as mean ± SEM. Sample sizes are presented in the figure legends. The multiple comparison of ANOVA was used to analyze the data from multiple groups. An independent samples t test was applied for comparisons between two groups. Correlation analyses use Pearson correlation test. Significance was established at p < 0.05.

Intraoperative observation showed that there is no synovium attachment on the tibial plateau and the surface of articular cartilage is smooth in the control group (Fig. 1). In RA patients, there is serious synovial proliferation accompanied with cartilage malformation and erosion (Fig. 1, arrow a–b). H&E staining indicated that the synovium of the healthy control group comprised two to three layers of resting cells. In contrast, RA synovia showed an excessive and disorganized proliferation of synoviocytes with large amounts of lymphocyte infiltration, abundant pannus formation, and obvious cartilage erosion (Fig. 1, arrow c–e). Masson staining demonstrated a lot of collagenous fiber formation (Fig. 1, arrow f). An evaluation of the presence of macrophages in the synovium by IHC showed that a large number of CD68⁺ macrophages had been recruited into the synovial tissue of RA patients when compared with the controls (Fig. 1, arrow g).

We performed secondary Ab staining to exclude unspecific staining (Supplemental Fig. 3A, 3B). As demonstrated above, one of the hallmarks of RA is significant hyperplasia of the synovium accompanied by strong cell proliferation (Fig. 2A, 2B, DAPI). An evaluation of the presence of SMs in the synovium by IF microscopy showed that a large number of cells had been recruited into the synovial tissue of RA patients when compared with healthy controls. Both iNOS⁺ M1 SM and TGM2⁺ M2 SM are activated and can be shown to have infiltrated into synovial tissue (Fig. 2A). Additionally, an examination of AT1R and AT2R expression on SMs showed that both receptors were upregulated significantly on SM (Fig. 2B).

The protein expression of AT1R and AT2R in PBMCs was quantified by ELISA after cell lysis. We found that both the AT1R and AT2R expression and the ratio of AT1R/AT2R were upregulated (Fig. 2C). Employing a Pearson correlation test, we showed that C-reactive protein (CRP) and erythrocyte sedimentation rate (ESR) were positively correlated with AT1R expression (r = 0.559, p < 0.001; r = 0.491, p < 0.001) and the ratio of AT1R/AT2R (r = 0.541, p < 0.001; r = 0.569, p < 0.001). What is more, CRP and ESR were negatively correlated with AT2R expression (r = −0.606, p < 0.001; r = −0.466, p < 0.001) (Fig. 2D).

We used the well-established in vivo model of an autoimmune inflammatory joint disease, CIA, and the AT2R agonist CGP42112 to evaluate the role of AT2R in vivo. To this end, we inoculated rats with an emulsion containing collagen and CFA in the right hind paw, causing a strong, primary inflammation within 24–48 h after the first immunization. Body weight was determined every 4 d after the booster injection. A secondary inflammation caused by a systemic autoimmune response developed around day 14. The left paw of CIA rats also displayed deformations, and movement was limited because of the inflammation in their joints (Fig. 3A). Additionally, CIA rats also showed a severe body weight loss and substantial secondary paw swelling (Fig. 3B, 3C). As treatment, CGP42112 was injected intra-articularly at a concentration of 5, 10, and 20 μg/kg. Etanercept, one of the biologics for RA treatment and already demonstrated to inhibit macrophage-abnormal activation and macrophage M1 polarization, was used as a positive control treatment. As a consequence of the injections of increasing amounts of CGP42112, the body weight loss after immunization improved and the secondary paw swelling was reduced in correlation with the amount of CGP42112 transferred (Fig. 3). The positive control injection with entanercept resulted in less weight loss and joint swelling than in those of CGP42112 treatment groups (Fig. 3).

We investigated the pathological changes of several rat joints in CIA and the AT2R activation on the pathology using H&E staining and radiographic analysis. H&E staining indicated that the control group synovium had one to three layers of synoviocytes and a smooth cartilage surface (Fig. 4A). The x-ray image of the control group showed no swelling, with both toe and ankle joints displaying a clear joint space (Fig. 4B). In contrast, CIA joints showed an excessive and disorganized proliferation of synoviocytes, with large amounts of lymphocyte infiltration, abundant pannus formation, obvious cartilage erosion, and distinct local inflammation (Fig. 4A). Radiographic analysis showed that toe and ankle joint spaces gradually disappeared or became indistinct (Fig. 4B, arrow a), and the bones exhibited monosaccate or multisaccate destruction (Fig. 4B, arrow b). After treatment with a low dose of CGP42112 (5 μg/kg), rat joints had all the hallmarks of an inflammatory response, resulting in obvious cartilage erosion and less joint spaces. After an intermediate dose of CGP42112 (10 μg/kg), joints showed excessive proliferation of synoviocytes combined with strong lymphocyte infiltration, but the joint spaces were preserved. Finally, a treatment with a high dose of CGP42112 (20 μg/kg) protected the rats from synovial hyperplasia, pannus formation, and saccate destruction (Fig. 4A, 4B). To gain further insights into the mechanistic aspects of the anti-inflammatory effect of AT2R activation, we examined the serum concentrations of cytokines in rats with CIA (Fig. 4C). The proinflammatory cytokines, such as MIP-2, IL-6, IL-1β, TNF-α, and IFN-γ, were expressed abundantly in the serum of CIA rats. In our analysis, we observed that neither of the three doses of CGP42112 downregulated the serum titer of MIP-2 and IFN-γ, whereas the intermediate dose (10 μg/kg) and the high dose (20 μg/kg) downregulated serum TNF-α and IL-6. Only the highest dose of 20 μg/kg downregulated the IL-1β level and upregulated the IL-10 level in serum (Fig. 4C). Based on these results, we chose the dose of CGP42112 (20 μg/kg) that showed a maximal effect for further experiments.

We used FACS to isolate rat CD3⁺ T cells from spleen to measure the T cell–derived cytokines IL-4, IFN-γ, and IL-17 (Fig. 4D). In the early stage of CIA model (d14), the IFN-γ and IL-17 levels were essentially unchanged in the different experimental groups, whereas IL-4 was significantly upregulated in the CIA group as compared with the control group. In the peak stage (day 28) of our model, additional CGP42112 treatment upregulated IL-4 level significantly, whereas the IFN-γ and IL-17 levels were upregulated only in CIA group. At the later relief stage (day 42) a similar picture emerged. The addition of CGP42112 and etanercept did not modulate the increased IFN-γ production of T cells in both the peak and relief stage. In contrast, CGP42112 and etanercept showed a strong effect in the suppression of T cell–produced IL-17. Meanwhile, a lower ratio of IFN-γ/IL-17 was observed in the CIA group during the peak and relief stage, indicating a more severe disease in the untreated CIA group, whereas after receiving CGP42112 or etanercept treatment, the ratio of IFN-γ/IL-17 was higher, indicating that these medical treatments have a therapeutic effect on CIA.

AT2R activation significantly downregulated the high level of SM infiltration in CIA rats (Fig. 5A). The level of AT1R and AT2R expression was significantly increased on SMs in the CIA group compared with the control group, and the ratio of AT1R/AT2R in SMs was elevated. These results suggested that SMs expressing proinflammatory AT1R predominate in CIA mice. As a specific AT2R agonist, treatment with 20 μg/kg CGP42112 significantly downregulated AT1R expression and upregulated AT2R expression on SMs and normalized the ratio of AT1R/AT2R (Fig. 5B, 5C).

We found a large proportion of CD11b⁺CD86⁺ M1 macrophages in the CIA group, and the ratio of CD86/CD163 in SMs was increased compared with the vehicle group, implying that M1 macrophages played a central role in CIA pathology. We investigated the characteristics of SM phenotypes after AT2R activation in CIA rats. As a specific AT2R agonist, treatment with CGP42112 (20 μg/kg) downregulated CD86 expression and upregulated CD163 expression. After AT2R activation, the CD11b⁺CD163⁺ M2 SMs may play a central role in SMs (Fig. 5D, 5E).

The proportions of induced BMDMs is more than 90% (Supplemental Fig. 4). Based on the BMDMs, we used NO expression in M1 macrophages to evaluate the optimal doses of AT2R agonist CGP42112 and AT2R antagonist PD123319 for further experiments. To polarize BMDMs into M1 macrophages, we used LPS (1 μg/ml, 24 h). The BMDMs were pretreated with different concentrations of PD123319 for 30 min before CGP42112 treatment for 2 h. LPS, as a strong inducer of M1 polarization, upregulated NO significantly. With the increase of CGP42112 (from 10⁻⁹ to 10⁻⁶ M), NO was decreased. Interestingly, the content of NO was increased when the CGP42112 concentration is above 10⁻⁶ M (Fig. 6A). The concentration of PD123319 above 10⁻¹⁰ M can block the pharmacological action of AT2R activation with regard to NO. The PD123319 (10⁻⁷ M) shows excellent antagonistic activity compared with 10⁻⁸ M PD123319 (Fig. 6B). Therefore, we choose the optimal concentrations of CGP42112 (10⁻⁶ M) and PD123319 (10⁻⁷ M) for our further investigation.

Both AT1R/AT2R expression, the ratio of AT1R with AT2R, and M1 macrophage marker iNOS were significantly increased in BMDMs after LPS treatment, but the M2 macrophage marker Arg1 was not changed. Pretreatment with of these cells with CGP42112 (10⁻⁶ M) downregulated AT1R expression and upregulated AT2R and Arg1 expression. Furthermore, giving AT2R antagonist PD123319 before CGP42112 blocked AT2R effects (Fig. 7).

Three branches of JNK, p38 MAPK and ERK1/2, are well-known downstream effectors of the MAPK cell signaling pathway. We determined the amount of phosphorylated and nonphosphorylated forms of each signaling molecule and found that the three branches were activated after LPS stimulation for 4 h in total cell protein. In NF-κB cell signaling, the levels of nuclear NF-κB p50/p65 subunits and phosphorylated IκB were elevated. Interestingly, AT2R activation only inhibited ERK1/2 phosphorylation in MAPK signaling (Fig. 8A). Moreover, CGP42112 downregulated IκB phosphorylation and p50/p65 nuclear translocation, indicating that AT2R activation inhibited NF-κB activation. This effect of CGP42112 was blocked by using AT2R antagonist PD123319 (Fig. 8B).

In our study, GRK2 was highly expressed in total cell protein, especially on the cell membrane, after LPS activation. Addition of AT2R activation did not influence GRK2 expression in total cell protein. Nevertheless, AT2R activation-downregulated GRK2 membrane expression and -upregulated GRK2 cytoplasmic expression were confirmed by three kinds of methods, such as Western blot, flow cytometry, and IF. Additionally, we showed that AT2R antagonists could block this effect. Therefore, AT2R activation can cause GRK2 membrane translocation (Fig. 9A–C).

After AT2R activation with CGP42112, we found GRK2 was translocated into cytoplasm and p-ERK1/2 was inhibited. To explore the relationship between cytoplasmic GRK2 and ERK1/2 phosphorylation, we used coimmunoprecipitation to detect the interaction between GRK2 and p-ERK1/2. The coexpression of p-ERK1/2 and GRK2 in BMDMs increased in cell cytoplasm after AT2R activation (Fig. 9D). Furthermore, giving AT2R antagonist PD123319 before could block its pharmacologic action. Taken together, AT2R activation causes cytoplasmic GRK2 to bind to p-ERK1/2 and may inhibit ERK1/2 activation (Fig. 10).

RA is a systemic autoimmune disease that is characterized by the infiltration of inflammatory immune cells and the uncontrolled hyperplasia of synoviocytes, ultimately leading to major pathological features, including synovitis and cartilage destruction (24). There are numerous treatment options for RA, such as nonsteroidal anti-inflammatory drugs, disease-modifying antirheumatic drugs, and biological agents, including TNF, IL-6, and IL-1 inhibitors. Unfortunately, these drugs are often associated with inadequate responses and can cause severe adverse reactions (25). The large transmembrane receptor family of GPCRs, which includes the AT2Rs, transfer signals across the cell membrane, involving them in many pathophysiological processes, including pathways relevant in RA (26). Two thirds of all currently available drugs target GPCRs directly or indirectly. However, the detailed mechanism of GPCRs signaling is still elusive. Selective modification of GPCR-dependent signaling cascades to inhibit disease progression in rheumatic diseases needs to be investigated, and there is an urgency to find safer and more effective therapeutic targets that help RA patients.

Articular damage is mainly driven by peripheral immune cells and synovial lining fibroblasts, also called FLS. Consequently, an important characteristic of the rheumatoid synovium is the strong hyperplasia of the lining layer, which is caused by more of both FLS and SMs. The stimulation of AT2R with the specific agonist CGP42112 significantly reduced gene expression of IL, IL-1β and IL-6, and activation of NF-κB in RA FLS, whereas opposite effects were elicited by AT2R small interfering RNA. Moreover, AT2R agonism efficiently decreased proliferation and migration of unstimulated RA FLS or after stimulation with proinflammatory cytokines. The activation of AT2R with a specific agonist may effectively dampen the proinflammatory behavior of RA FLS and might represent a novel therapeutic strategy for patients with RA (27). It could be of interest to further investigate whether AT2R agonism exhibits similar anti-inflammatory effects on other synovial cell types, especially SMs.

Indeed, in a combination of in vitro and vivo results, it has been demonstrated in a rat model of AIA that intra-articular injection of the AT2R agonist CGP42112A effectively decreased the severity of arthritis. This view is further supported by the in vitro evidence that the activation of AT2R with a specific agonist effectively counteracts the proinflammatory and aggressive behavior of RA FLS (17).

In our results, AT2R activation could also change the balance of T effector subset cells by inhibiting IL-17 expression and upregulating the ratio of IFN-γ/IL-17. This indicates that AT2R agonist could interfere with disease progression. In fact, it has been reported that AT2R stimulation is anti-inflammatory in LPS-activated THP-1 macrophage cell lines through downregulation of TNF-α and IL-6 and increased production of IL-10 (28), supporting our conclusion that AT2R agonism might also have an anti-inflammatory effect on SMs.

Macrophages have emerged as a key component of the innate immune system in the adult organism. Their complex pathway to maturity, their unique plasticity, and their various roles as effector and regulatory cells during an immune response have been the focus of intense research (29). A class of surface molecules, AT2R, plays important roles in many immune processes, and various members of this family of receptors have become a research focus for therapeutic targets that can modulate the response of immune cells in pathologies such as diabetes, atherosclerosis, and chronic inflammatory diseases (30).

Our results demonstrated both that AT1R expression is upregulated and the AT1R/AT2R ratio in RA patients PBMCs was significantly increased. The AT2R expression in RA patients PBMCs is slightly upregulated. The inflammatory biomarkers CRP and ESR of RA patients were positively correlated with AT1R expression and the ratio of AT1R/AT2R. Thus, the imbalance between AT1R and AT2R could be involved in advancing RA progression.

In our experiments, we used the AT2R agonist CGP42112 as a tool drug to study the physiological function of AT2R. We found that AT2R on RA patients PBMCs and SMs is upregulated and AT2R activation could ameliorate the symptoms of arthritis and macrophage infiltration in the CIA model.

Additionally, we found that treatment with CGP42112 not only suppressed AT1R expression but also increased AT2R expression on SMs, causing a beneficial change in the AT1R/AT2R ratio. Furthermore, we could show that SMs in the CIA model tended to be of the M1 phenotype, with higher CD86 and lower CD163 expression. After AT2R activation, the predominant SM phenotype was reversed and showed now a strong M2 phenotype with higher CD163 and lower CD86 expression. Therefore, activation of AT2R changed the AT1R/AT2R ratio and supported M2 polarization in SMs in the rat CIA model.

In CIA, the IL-4 level was significantly upregulated only in the early stage of disease. It has been demonstrated previously that IL-4 is upregulated in the synovial fluid of early RA but not established RA, suggesting that this cytokine is part of an early modulatory response that is lost when patients progress to fully established disease (31). Activation of AT2R induced T cells to express a significantly elevated level of IL-4 in peak stage. Functionally, IL-4 is potent antiarthritic cytokine that also inhibits cartilage damage and osteoclastogenesis. Thereby, it has combined anti-inflammatory and antiosteoclastogenic properties in models of inflammatory arthritis (32). Furthermore, IL-4 binds to macrophages and induces them to differentiate into an M2 phenotype via activation of STAT6, thus shifting the balance of macrophages from proinflammatory M1 macrophages to regulatory M2 macrophages. This increases the release of anti-inflammatory effector cytokines, such as IL-10 and TGF-β, and reduces the production of proinflammatory cytokines, such as TNF and IL-1β, by M1 macrophages (1). This is consistent with our result that after intra-articular injection of CGP42112 SM M2 polarization is promoted and TNF and IL-1β secretion is downregulated.

We showed that SM has two origins: embryonic SM and bone marrow SM (BMSM). Embryonic SM is seeded at joint synovium at a prenatal stage, whereas the classic BMSM infiltrated the synovium at the perinatal stage. In RA, large numbers of SMs are present in synovial tissues, with BMSMs accounting for nearly 90% of the total number of SMs (33). Although the number of SMs is upregulated in inflammation, the absolute number of SMs is still limited. Therefore, we based our further studies to define phenotype and function of SMs on BMDMs, which could be generated in large numbers. When pretreated with CGP42112 before activation with LPS, AT2R expression was upregulated significantly, changing the AT1R/AT2R ratio and inhibiting BMDM M1 polarization. We concluded that AT2R plays an anti-inflammatory role in preventing the progression of RA by inhibiting the two main synovial tissue cell-type FLS and SM abnormal activation (27).

The functions of AT2R are regulated by GRKs. In a previous study, we found that in chronic inflammation in AIA or CIA models, out of the four subtypes of the GRKs (GRK2, GRK3, GRK5, GRK6) only GRK2 showed a significant change (34). The function of GRK2 in the regulation of cell signaling depends not only on specific stimuli, cell type, and/or physiological context but also subcellular localization (35). We observed that total GRK2 protein and GRK2 localized on the membrane of inflammatory macrophages was increased significantly. After AT2R activation, the GRK2 membrane localization decreased, whereas GRK2 cytoplasm expression increased without a change in the total GRK2 expression. We already demonstrated that the abnormal activation of the MAPK–NF-κB signaling pathway plays an important role in CIA progression. The phosphorylation levels of the three branches of MAPK (JNK, P38, ERK) are different in different immune cells, and any of the three pathways can activate the NF-κB signal to promote an inflammatory immune response in CIA rats. AT1R blocker could downregulate both ERK1/2 and P38 phosphorylation but leave JNK phosphorylation in T cells and B cells in CIA unaffected (19). The underlying mechanisms that modulate the phosphorylation levels of the MAPK pathway remain unclear.

At this time, only ERK1/2 phosphorylation was reduced after AT2R activation. Our previous studies revealed that ERK1/2 activation is related to translocation of GRK2. When GRK2 translocated from membrane to cytoplasm, the increasing amount of GRK2 in cytoplasm is combined with p-ERK1/2, which causes a reduction of p-ERK1/2 entering the nucleus (36). Translocation of p-ERK1/2 to the nucleus from the cytoplasm can activate the inflammatory immune response when activated under inflammatory conditions (37), whereas cytoplasmic GRK2 could inhibit chemokine-mediated induction of ERK1/2 activity (22). ERK1/2 is a key signaling molecule that regulates NF-κB activation. Phosphorylated ERK promotes the phosphorylation of IκB, dissociates the NF-κB complex, and induces the translocation of the p65/p50 subunit from the cytoplasm to the nucleus (38). When the coexpression between p-ERK1/2 and GRK2 in cytoplasm is increased after AT2R activation, the decreased dissocation of p-ERK1/2 fails to promote IκB phosphorylation. Thus, the activation of NF-κB is downregulated because of low phosphorylation of IκB, reduced nuclear translocation, and a low level of inflammatory cytokines. Therefore, we hypothesize that AT2R activation-induced GRK2 cytoplasm translocation can downregulate the dissociative p-ERK1/2 and NF-κB activation. The possible underlying mechanism of AT2R anti-inflammatory role on SMs was shown in Fig. 10. What is more, the lower dissociative p-ERK1/2 causes less p-ERK1/2 nuclear translocation, as demonstrated before (36).

In summary, the current study shows that the AT1R and AT2R expression is directly correlated with RA severity. AT2R activation inhibited M1 macrophage polarization and reduced RA pathology. Further studies showed that AT2R activation inhibited activation of macrophages by promoting GRK2 cytoplasmic translocation. Therefore, attempts to change the AT1R/AT2R ratio might be a new mechanism in RA therapy, and AT2R agonists may be regarded as a novel class of drugs that could be used in the treatment of inflammatory and immune diseases in the future.

The authors have no financial conflicts of interest.