Immune regulatory and anti-resorptive activities of tanshinone IIA sulfonate attenuates rheumatoid arthritis in mice

Abstract

Background and Purpose

Rheumatoid arthritis (RA) is an autoimmune disease characterized by chronic inflammation and painful joint destruction. Current treatments are helpful in RA remission, but strong immunosuppressive activity and patient resistance are clinical issues. This study explores a dual-action inhibitor, possessing both anti-inflammatory and anti-resorptive properties, as a novel treatment for RA.

Experimental Approach

Therapeutic efficacy and mechanisms of ectosteric (tanshinone IIA sulfonate [T06]) and active site-directed (odanacatib [ODN]) inhibitors of cathepsin K (CatK) were evaluated in RA mouse models. Pathology was assessed through biochemical analyses and histopathological examination. Flow cytometry analysis was performed to characterize immune cells. Anti-inflammatory effects of T06 on nuclear factor kappa beta (NF-κB) pathway were studied in macrophages.

Key Results

T06 effectively lowered the number of joint-resident immune cells, accompanied by significantly reduced production of inflammatory cytokines and collagenolytic proteases. This also included the suppression of Th17 cells and IL-17, resulting in the reduction of osteoclasts in arthritic joints and amplification of the overall anti-resorptive effect of T06, which has been attributed to its selective inhibition of the collagenolytic activity of CatK by preventing its oligomerization. The anti-inflammatory mechanism of T06 was based on blocking the phosphorylation of IκBα in the NF-κB pathway, resulting in reduced activation and expression of inflammatory cytokines. In contrast, ODN had no effect on inflammation and disease progression and was limited to the inhibition of CatK.

Conclusions

The combined anti-resorptive and anti-inflammatory activities characterize T06 as a novel therapeutic agent for RA.

Graphical Abstract

Abbreviations

BMD

bone mineral density

BV

bone volume

CIA

collagen-induced arthritis

CTX II

C-terminal cross-linked telopeptides of type II collagen

IKK

inhibitor of nuclear factor kappa B kinase

IκB

inhibitor of nuclear factor kappa B

MMP

matrix metalloproteinase

NF-κB

nuclear factor kappa B

ODN

odanacatib

PDPN

podoplanin

RA

rheumatoid arthritis

RANKL

receptor activator of nuclear factor kappa B

REL

NF-κB subunit

T06

tanshinone IIA sulfonate

TAK1

transforming growth factor-β-activated kinase 1

Tb.N

trabecular number

Tb.Sp

trabecular space

Tb.T

trabecular thickness

TRAP

tartrate-resistant acid phosphatase

What is already known

• Inflammation and tissue destruction are the central features in rheumatoid arthritis (RA) progression.
• We recently observed the cathepsin K-directed ectosteric mechanism of T06 in bone resorption inhibition.

What does this study add

• T06 reduces immune regulatory and anti-resorptive activities in RA joints without causing systemic immunosuppression.
• T06 attenuates inflammatory signalling by downregulating NF-κB pathway and suppressing the number of immune cells.

What is the clinical significance

• Dual-action inhibitor holding both anti-inflammatory and anti-resorptive properties represents a novel therapeutic candidate for RA.

1 INTRODUCTION

Rheumatoid arthritis (RA) is a chronic autoimmune disease characterized by joint inflammation, bone/cartilage destruction, pain and ultimately disability (Firestein, 2003). This disease affects up to 0.5% to 1% of the worldwide population, with a complex pathogenesis, where both innate and adaptive immune responses play important roles (Alamanos & Drosos, 2005; Müller-Ladner et al., 2005). Joint inflammation is induced by synovial hyperplasia and the infiltration of inflammatory cells such as macrophages, T-cells and B-cells into the synovium, and triggered and exacerbated by an elevated production of inflammatory cytokines and growth factors (Cope et al., 2007; Komatsu & Takayanagi, 2022; McInnes et al., 2016). T-cell-mediated immune abnormalities are the main pathogenetic pathway of RA, where T-cells are responsible for osteoclastogenesis and activation of macrophages and fibroblasts to produce cytokines and chemokines (Ponchel et al., 2020). T-cells differentiate into Th1 and Th17 cells and perpetuate the inflammatory environment in the synovium by releasing receptor activator of nuclear factor kappaB ligand (RANKL) and various inflammatory molecules, which subsequently results in inflammation and tissue destruction (Kim et al., 2015; Sato et al., 2006). Once the synovium becomes a highly immune-responsive environment, mast cells, neutrophils, osteoclasts, synovial fibroblasts and chondrocytes are activated to further amplify the production of cytokines, reactive oxygen species (ROS), neutrophil extracellular traps (NETs) and tissue-degrading proteases such as cathepsins, matrix metalloproteinases (MMPs) and aggrecanases (ADAMTS) (McInnes et al., 2016; Murphy et al., 2002; Nagase & Kashiwagi, 2003; Rengel et al., 2007; Wright et al., 2020). These proteases are the main culprits involved in tissue destruction.

Several signalling pathways are associated with the initiation and progression of RA. Dysfunctional regulation of the Janus kinase / signal transducers and activators of transcription (JAK/STAT) pathways results into rapid cytokine signalling (Malemud, 2018). The mitogen-activated protein kinase (MAPK) pathway directly regulates the vital cellular activities (migration, apoptosis and differentiation) during RA (Inoue et al., 2006), and wingless/integrated (Wnt) signalling plays a crucial role in inflammation and tissue destruction (Rabelo et al., 2010). Of particular importance is the nuclear factor-kappaB (NF-κB) pathway that regulates both innate and adaptive immune responses by steering the activation and differentiation of immune cells and production of pro-inflammatory factors (e.g., TNF-α, IL-1β, IL-6 and IL-17) (Simmonds & Foxwell, 2008). The NF-κB pathway is regulated by various kinases (Solt & May, 2008). Therefore, a better understanding of the mechanism and putative therapeutic targets of this pathway is essential for developing more potent inhibitors to attenuate both inflammation and joint destruction.

Current anti-rheumatic drugs are primarily anti-inflammatory and include non-steroidal anti-inflammatory drugs (NSAIDs) and immunosuppressant disease-modifying anti-rheumatic drugs (DMARDs) (Bullock et al., 2018). NSAIDs efficiently inhibit cyclooxygenase activity to reduce inflammation, but have various side effects (Crofford, 2013; Harirforoosh et al., 2013). DMARDs work to suppress the overactive immune/inflammatory systems, constituting either small molecule drugs such as methotrexate or biologicals such as TNF-α inhibitors (etanercept) and IL-1 receptor antagonists (anakinra), or antibodies against TNF-α (infliximab, adalimumab and humicade) (Aletaha & Smolen, 2018). These biologicals are effective and can lead to temporary remission of RA at the very least. However, besides being highly expensive, they are prone to trigger infections due to their strong immunosuppressive activity (Bitoun et al., 2023). Therefore, newer therapies such as targeted small molecules are in development to overcome these limitations (Harrington et al., 2020). Another drug target is cathepsin K (CatK), a cysteine protease upregulated in inflamed synovium, that contributes to bone erosion and immune response (Asagiri et al., 2008; Dai et al., 2020; Svelander et al., 2009). Active site-directed CatK inhibitors such as odanacatib (ODN), balicatib, relacatib and ONO-5334 demonstrated significant efficacies in osteoporosis trials by reducing bone degradation markers, increasing bone mineral density and reducing bone fracture rates (Duong et al., 2016). However, none of these inhibitors have achieved FDA approval due to adverse effects in skin and increased risk of cardiovascular events (Dai et al., 2020; Rünger et al., 2012). ODN and balicatib have been evaluated in human osteoarthritis trials, but these trials were terminated for undisclosed reasons in their phase II stages (clinical trials ID: NCT00397683; clinical trials ID: NCT00371670). ONO-5334 was evaluated in a collagen-induced monkey arthritis model and showed efficacy in reducing joint destruction but no effect on inflammation (Yamada et al., 2019). This body of work indicates that active site-directed CatK inhibitors have severe side effects and their therapeutic efficacy is limited to the inhibition of the degradation of the collagenous bone and joint matrix. We recently delineated the ectosteric mechanism of a medicinal plant-derived tanshinone (tanshinone IIA sulfonate [T06]) to inhibit CatK, by preventing the formation of collagenolytically active oligomers and thus inhibiting bone and cartilage resorption (Panwar et al., 2017). Inspired by our previous findings and the documented anti-inflammatory role of tanshinones (Ma et al., 2016), we hypothesized that selectively blocking both the inflammatory cascade and bone/cartilage erosion with a dual-activity inhibitor may efficiently reduce synovial inflammation and joint destruction in advanced RA.

In this study, we observed that T06 attenuates both inflammatory signalling and destruction of joint tissue by suppressing the number of immune cells and osteoclasts in the collagen-induced arthritis (CIA) mouse model. Histological, immunochemical, cell sorting, biochemical and micro-computed tomography (μCT) approaches were used to demonstrate the dual anti-inflammatory and anti-resorptive activity of T06. We delineated the pathways of T06 in IκBα/NF-κB signalling and the suppression of cytokine release and inflammatory cell distribution.

2 METHODS

2.1 CIA

CIA was induced in female DBA/1 mice (10-week-old) purchased from Charles River Laboratories, Canada. Each mouse was injected 200 μg of bovine type II collagen (Chondrex, Woodinville, WA, USA) emulsified with complete Freund adjuvant (CFA, Chondrex) containing a final concentration of 0.5 mg·mL⁻¹ of Mycobacterium tuberculosis into the tail base. Immunization was repeated 21 days after primary immunization with 200 μg of bovine type II collagen and incomplete Freund adjuvant (IFA, Chondrex) at a different site than the first injection. Lipopolysaccharide (LPS) from Escherichia coli 0111: B4 (LPS, Chondrex) (40 μg per mice; constituted in saline) was injected intraperitoneally on Day 24 to synchronize the onset of arthritis. Arthritis was developed within 48 h in >95% of mice. Treatment was started on Day 27 (3 days after of LPS injection). Inhibitors were mixed with food powder (LabDiet, USA) and given as T06 (40 mg·kg⁻¹·day⁻¹) and ODN (10 mg·kg⁻¹·day⁻¹) doses every day in the morning (3 g chow per mouse based on body weight). During food/inhibitor uptake, mice were housed in single cage to make sure that each mouse obtained the weight-based drug amount. Cages were thoroughly inspected every noon to make sure the drug consumption was complete, and mice were given subsequently 2 g of Dietgel 76A (ClearH₂O, Westbrook, ME, USA), a fortified dietary supplement for hydration and nutrition. We planned a restricted diet schedule for mice to leave them hungry overnight for the complete uptake of the food/inhibitor mix next morning within 1 h. Mice were trained for 2 weeks on a restricted diet prior to start of the inhibitor treatment, and mice weight, behaviour and food consumption were monitored every day. We used three DBA/1 mice in each group to determine the mouse plasma concentrations of T06 (40 mg·kg⁻¹·day⁻¹) and ODN (10 mg·kg⁻¹·day⁻¹) at 0.5, 2, 4, 6, 8, 12 and 24 h after complete uptake of the food/inhibitor. Samples were analysed by ultrahigh performance liquid chromatography–tandem mass spectrometry (UHPLC/MS/MS) (Agilent, Mississauga, Ontario, Canada), and pharmacokinetic analysis was performed using a non-compartmental analysis (PK Solutions 2.0, Montrose, CO, USA) as described previously (Panwar et al., 2017). Mice were placed in a 12 light/12 dark cycle in a pathogen free facility at temperatures of 21°C with 50% humidity. Animals were randomly assigned and allocated to each group (n = 11) before the arthritis induction and prior to experiments by simple randomization.

2.2 Evaluation of clinical disease

Disease severity was assessed by visual scoring from 0 to 4. Each paw was assigned a score from 0 to 4 (0, no detectable arthritis; 1, mild swelling; 2, moderate swelling or redness with one or two inflamed digits; 3, swelling and redness involving the entire paw, foot pad and joint with three or more inflamed digits; 4, severe swelling and redness involving the entire paw, foot pad and joint with all inflamed digits with visible joint deformity). The clinical scores for each mouse are presented as the sum of the scores for the four limbs, and the maximal score for each mouse was 16. Joint swelling was quantified every third day by measuring fore and hind paw thickness and the ankle and knee width with a calliper. In addition, the grip strength of each paw was determined using a wire (2.5 mm in diameter) for 30 s using a score system between 0 and 4 (0, normal; 1, mildly reduced; 2, moderately reduced; 3, severely reduced; 4, unstable or no grip). Clinical scores (swelling, redness, inflamed digits, paw thickness, knee width and disease severity) were analysed by an investigator blinded to the mouse treatment during the treatment period.

2.3 Histological and Immunohistochemical staining

Joints and paws were removed, fixed in 10% formalin, decalcified in 12% ethylenediaminetetraacetic acid (EDTA) and processed for paraffin embedding. Afterwards, the joint sections (4 μm) were mounted on glass slides and stained with haematoxylin and eosin (H&E), toluidine blue and tartrate-resistant acid phosphatase (TRAP; MilliporeSigma, MA, USA). Each of the joints was classified for the severity of joint lesions. The following histologic parameters were assessed: inflammation, synovitis, pannus formation, oedema, cell infiltration and erosion of cartilage/bone, and classified into four grades: 0, normal; 1, slight; 2, moderate; 3, maximum; 4, severe. Area of bone erosion and inflammation, and number of osteoclasts per bone surface, were assessed. For immunolocalization, formalin-fixed, paraffin-embedded rheumatoid tissues were cut into consecutive 4-μm-thick sections. After the sections were dewaxed in xylene, dehydrated via graded ethanol solutions, antigen retrieval was performed by heating the sections for 20 min at 100°C in citrate solution in a steamer. Then, the sections were cooled and rinsed in PBS for 5 min and subsequently blocked using 3% bovine serum albumin. The sections were then incubated with primary antibody against mouse TNF-α (Abcam, Cat# ab66579, RRID:AB_1310759), IL-1β (R and D Systems, Cat# AF-401-NA, RRID:AB_416684) and CatK (MS4) at 4°C overnight. Sections incubated with rabbit IgG in combination with mouse IgG served as a negative control. Subsequently, sections were washed and stained with biotinylated anti-rabbit IgG using Vectastain Elite ABC Kit (Thermo Fisher Scientific, Newark, NJ, USA) at 1:100 dilution for 1 h at room temperature and developed with the chromogenic substrate 3,3′-diaminobenzidine (DAB). Images were captured using a Nikon Eclipse Ci microscope with 20× and 40× objectives and analysed using NIS elements D 4.5 software. Control experiments involved incubating tissue sections with the antibody diluent but without primary antibody, followed by incubation with secondary antibodies and detected as described above. Quantification of histological and immunohistochemical analysis was performed in a blinded manner (investigator blinded to the treatment).

2.4 Inflammatory marker analysis

Blood was taken via cardiac puncture and collected in two separate tubes for (i) flow cytometry (1:1 in 3.9% sodium citrate), and (ii) ELISA where blood was allowed to clot for 10 min at room temperature to obtain serum. Subsequently, blood was spun for 10 min at 10,000 rpm, and serum was removed and immediately frozen at −80°C until used. Mice knee joints (including synovium, adjacent tissue and bone) were pulverized using a mortar and pestle filled with liquid nitrogen and suspended in 1 ml PBS/200 mg of tissue containing Complete Protease Inhibitor Cocktail (Thermo Fisher Scientific) and homogenized for 30 s using a tissue homogenizer. Mouse knee joint homogenates were centrifuged for 10 min at 10,000 rpm at 4°C, and supernatants were collected for cytokine analysis using ELISA according to the manufacturer's instructions for TNF-α (Abcam, Cat# ab208348,), IL-1β (Abcam, Cat# ab197742), IL-6 (Abcam, Cat# ab222503), IL-17 (MyBioSource, San Diego, CA, USA, Cat# MBS2508197), RANKL (MyBioSource, Cat# MBS3805457) and IFN-γ (Abcam, Cat# ab282874). Bone and cartilage turnover markers were detected in blood using C-terminal cross-linked telopeptides of type I collagen (CTX I) (MyBioSource, Cat# MBS703094) and C-terminal cross-linked telopeptides of type II collagen (CTX II) ELISA kit (MyBioSource, Cat# MBS706197).

2.5 Quantitative real-time PCR (qRT-PCR)

RNA was isolated from individual homogenized ankle joints using RNeasy® Mini-kits (Qiagen, Hilden, Germany). For cDNA synthesis, 1 μg of total RNA was reverse transcribed using qScript cDNA SuperMix (Quanta Biosciences, Beverly, MA, USA) as per kit instructions. Primers for the genes of interest were designed with primer design software Primer Express™ (Table S3). Each of forward and reverse primers was used at 10 μM concentration. qRT-PCR was performed using PerfeCTa SYBR Green qPCR SuperMix (Quanta Biosciences) with initial denaturation at 95°C for 1 min (one cycle) followed by 40 cycles of (denaturation 95°C for 10 s and annealing/extension at 60°C for 30 s) and finally kept at 4°C using qPCR thermocycler (Applied Biosystems). β-Actin and GAPDH were used as a control for adjusting the relative of total RNA. Relative gene expression changes were analysed by 2^−ΔCt method.

2.6 Protein extraction and western blotting

Mice knee joints were homogenized as described above and supernatants were used to determine the protein concentration using the Quick start Bradford reagent (Bio-Rad, Hercules, CA, USA). A total of 10 μg of extracted protein was used for immunoblotting. Samples were separated by sodium dodecyl sulphate (SDS) polyacrylamide gel electrophoresis and transferred to nitrocellulose membrane. After incubation with 5% skim milk in Tris-buffered saline (TBST) for 3 h, the membrane was incubated overnight at 4°C with primary antibodies against the following proteins: CatK (MS4, in-house), MMP13 (Thermo Fisher Scientific, Cat# PA5-27242, RRID:AB_2544718), MMP3 (Abcam, Cat# ab52915, RRID:AB_881243) and transforming growth factor beta-1 (TGFβ1) (Abcam, Cat# ab215715, RRID:AB_2893156). Membranes were washed four times for 5 min and incubated for 2 h with a 1:5000 dilution of horseradish peroxidase-conjugated anti-mouse antibody (W4021; Promega, Madison, WI, USA) or anti-rabbit antibody (Promega, Cat# W4011, RRID:AB_430833). The membrane was washed five times with TBST for 10 min and developed with the enhanced chemiluminescence system (GE Healthcare). β-Actin (Biolegend, San Diego, CA, USA, Cat# 622102, RRID:AB_315946) was used as control to normalize the levels of protein.

2.7 Three-dimensional μCT

At the end of treatment, mice were anesthetized with isoflurane and euthanized by cervical dislocation, and one pair of fore/hind legs were removed and fixed in 10% neutral buffered formalin. Ex vivo μCT scanning of the knee joints and paws was carried out using a Scanco μCT 100 (Scanco Medical, Bruttisellen, Switzerland). The bones were scanned with a 0.5 mm Al filter in batches of three at a nominal resolution of 7.4 μm. A hundred projections were acquired at an 360° angular range and the X-ray source was set at 70 kVp and 200 μA at 200 ms integration time. Three-dimensional render images of paws and knee joints were generated through original volumetric reconstructed images by Scanco μCT software (Scanco Medical). Trabecular and cortical regions were selected by drawing ellipsoid contours, and 150 slices were analysed in the same region of control, CIA and treated groups. Bone mineral density (BMD = BM/BV mg·cm⁻³) and other parameters (trabecular number [Tb.N], trabecular thickness [Tb.Th] and trabecular space [Tb.Sp]) were quantified from μCT scans using Scanco μCT software.

2.8 Flow cytometry and cell sorting

Hindlimb (knee joints) were digested with 10 mg·ml⁻¹ collagenase from Clostridium histolyticum type IV (MilliporeSigma) for 45 min at 37°C. Cells were resuspended in eBioscience™ flow cytometry staining buffer (Thermo Fisher Scientific), and cell numbers were quantified using an automated cell analyser and stained with trypan blue (MilliporeSigma) for cell viability. Cells were stained with a panel of antibodies (see Table S1). All antibodies were utilized at 1:200 dilution. Blood was collected 1:1 into 3.9% sodium citrate from either control or arthritic mice. Cells were stained with a panel of antibodies for 25 min, and red blood cell (RBC) lysis was performed using RBC lysis buffer (Thermo Fisher Scientific). Cells were isolated by mechanical dissociation of the spleen and were stained with a panel of antibodies, as described above. Debris (SSC-A vs. FSC-A) and doublets (FSC-H vs. FSC-A) were excluded, and live/dead discrimination was determined using viability dye iFluor 860 maleimide (FSC-H vs. Live/Dead). Cytoflex LX Analyser (Backman Coulter) was used for cell sorting, and data were analysed using FlowJo software.

2.9 NF-κB activation study

Bone marrow-derived macrophages were obtained in vitro by culturing bone marrow cells of DBA mice in Complete Macrophage Medium (Cell Biologics, Chicago, IL, USA) supplemented with granulocyte-macrophage colony-stimulating factor (GM-CSF; 25 ng·ml⁻¹; R&D Systems). Mature macrophages were plated on glass coverslips in 24-well plates at the density of 10 × 10⁴ per well and incubated in a complete media overnight to 80% confluence. Subsequently, macrophages were stimulated with LPS (50 ng·ml⁻¹, Thermo Fisher Scientific) in the presence and absence of T06 (10 μM) and ODN (1 μM) and incubated for 4 h. After incubation, conditional media were removed from each well, and cells were fixed with 4% paraformaldehyde followed by permeabilization with 0.3% Triton X-100 (MilliporeSigma) for 15 min and then blocked in 3% BSA for 2 h at room temperature. For nuclear translocation of the NF-κB p65 subunit, we used a NF-κB p65 primary antibody (Abcam, Cat# ab16502, RRID:AB_443394) at 1:200 dilution and anti-rabbit Alexa fluor 627-conjugated secondary antibody at 1:500 dilution. Slides were mounted in DAPI mounting medium (ProLong Gold, Thermo Fisher Scientific), imaged using Zeiss Axioplan II fluorescence microscope, and fluorescent areas were quantified using ImageJ software (NIH). Cell lysates were prepared with RIPA lysis buffer (Thermo Fisher Scientific), and nuclear lysates were prepared following manufacturer's instruction with nuclear extraction Kit (ab113474, Abcam). Changes in the protein expression and phosphorylation of NF-κB subunits were detected by standard western blot methods and using the following primary antibodies: p65 (Abcam, Cat# ab16502, RRID:AB_443394), p-p65 (Abcam, Cat# ab86299, RRID:AB_1925243), IκBα (Abcam, Cat# ab32518, RRID:AB_733068), p-IκBα (Abcam, Cat# ab133462, RRID:AB_2801653) and p-IKKα/β (Cell Signaling Technology, Cat# 9241, RRID:AB_2566820) at a dilution of 1:1000. Relevant horseradish peroxidase (HRP)-conjugated secondary antibodies, either anti-mouse (Promega, Cat# W4021, RRID:AB_430834) or anti-rabbit (Promega, Cat# W4011, RRID:AB_430833), at 1:5000 dilution were used and bands developed with the enhanced chemiluminescence system (GE Healthcare). β-Actin (BioLegend, Cat# 622102, RRID:AB_315946) at 1:1000 dilution was used as a control to normalize the levels of protein. NF-κB Transcription Factor Assay Kit (Abcam, Cat# ab207216) was used to quantify activation of NF-κB family members (p50, p52, p65, c-Rel and RelB) by ELISA in one assay. We used kinase enzyme system kits to determine the inhibitory effect of T06 on the kinases involved in the activation of NF-κB pathway via interfering with IKKα (Promega, Cat# V4068), IKKβ (Promega, Cat# V4502) and transforming growth factor-β-activated kinase 1 (TAK1) kinase activity (Promega, Cat# V4088).

2.10 Statistical analysis

A total of 11 mice for each condition were used in this study. Animals were randomly assigned prior to experiments by simple randomization. Evaluation of clinical disease, ELISA and μCT was done on all mice. All biochemical experiments were performed in three to five independent experiments. Mouse cell-based experiments including flow cytometry were performed in four independent experiments. Details of the number of experimental repeats are provided in the figure captions. GraphPad Prism (version 9; GraphPad Software, San Diego, CA, USA) used for analysis and statistical tests. Kruskal–Wallis, two-tailed, and Dunn's multiple comparisons tests were used in most of the analyses. Data are presented as mean ± SD (ns: not significant; significance *p < 0.05).

2.11 Animal study approval

All animal studies were conducted according to ethical guidelines for animals of the Canadian Council on Animal Care (CCAC) and approved by the University of British Columbia animal ethical committee (approval number: A16-0085). Animal studies are reported in compliance with the ARRIVE guidelines (Percie du Sert et al., 2020) and with the recommendations made by the British Journal of Pharmacology (Lilley et al., 2020).

2.12 Materials

T06 was purchased from Wuhan ChemFaces Biochemical Co., Ltd/China and ODN from Selleckchem.com, Houston, TX. Details of other materials and suppliers are provided in specific subsections in Methods.

2.13 Nomenclature of targets and ligands

Key protein targets and ligands in this article are hyperlinked to corresponding entries in https://www.guidetopharmacology.org and are permanently archived in the Concise Guide to PHARMACOLOGY 2021/2022 (Alexander et al., 2021).

3 RESULTS

3.1 Inhibition of clinical parameters of arthritis

Female DBA/1 mice immunized with type II collagen in CFA on Days 0 and 21, followed by LPS administration on Day 24, developed strong inflammatory responses in their joints. These mice were subsequently treated with either T06 or ODN for 21 days starting on Day 27 to allow an initiation of the disease (Figure 1a). There was no significant difference in the weight of mice in each group throughout the experiment (Figure 1b). Pharmacokinetic analysis of the plasma T06 concentration revealed: C_max: 76.5 ng·ml⁻¹, terminal elimination half-life (t_1/2): 1.95 h, AUC_INF: 415 (ng·h)·ml⁻¹, apparent volume of distribution (V/F): 287.2 ml·g⁻¹ and apparent oral clearance (CL/F): 83.6 ml·h⁻¹·g⁻¹. T06 concentration–time profile was characterized by peak plasma concentrations at 2 h followed by a continuous decrease and remained at 6.8 ng·ml⁻¹ at 12 h which further dropped below the detection limit at 24 h. These results are in accordance with our previous studies (Panwar et al., 2017). The ODN peak plasma concentration was 3940 ng·ml⁻¹ with a terminal elimination half-life (t_1/2) of 2.7 h, AUC_INF: 31,345 ng·h⁻¹·ml⁻¹, apparent volume of distribution (V/F): 4.9 ml·g⁻¹ and apparent oral clearance (CL/F): 1.3 ml·h⁻¹·g⁻¹. Within 1 week of T06 treatment, paws, knees and ankles of T06-treated mice showed a robust reduction in clinical arthritis symptoms when compared with untreated and ODN-treated CIA mice. At the end of the treatment (Day 48), T06-treated mice revealed a significant reduction in swelling, redness, inflamed digits and an about 60% decrease in the severity and incidence of arthritis when compared with CIA-untreated controls. In contrast to T06, ODN treatment did not produce a statistically relevant reduction in clinical scores when compared with untreated CIA mice (Figure 1d,e). However, we observed a more than 40% reduction in bone and cartilage degradation serum markers (CTX I and CTX II) in both T06- and ODN-treated mice when compared with untreated control animals (Figure 1f). This work supports the previously reported anti-resorptive CatK-inhibitory activities of both compounds, but additional CatK-independent anti-inflammatory effects of T06 make it a more promising drug candidate for RA.

FIGURE 1

(a) Schematic overview of the induction in collagen-induced arthritis (CIA) in DBA/1 mice and treatment protocol. Mice were orally administered with T06 (40 mg·kg⁻¹), ODN (10 mg·kg⁻¹) or vehicle for 21 days. (b) After the onset of clinical arthritis, no significant differences were observed in body weight among the groups. (c) Representative images of the hind limb from mice showing obvious remission of arthritis in T06-treated CIA mice. Severity of the clinical signs in untreated CIA and inhibitor-treated mice is represented in terms of (d) clinical score and (e) ankle thickness. (f) Serum CTX I and CTX II, assessed by ELISA, confirmed the anti-resorptive potential of the tested inhibitors. T06 treatment showed anti-inflammatory effect by reducing clinical score. Data are represented as the mean ± SD (n = 11 in each condition). ns, not significant; *P < 0.05 in comparison with the CIA-untreated control mice (Kruskal–Wallis test, two-tailed, and Dunn's multiple comparisons test).

3.2 Effect of T06 and ODN on bone/cartilage erosion and joint inflammation

Histological analysis of the paws, ankles and knees supported the observational clinical findings, demonstrating more than a 55% reduction in synovial inflammation and joint destruction in T06-treated CIA mice when compared with ODN-treated and CIA controls (Figures 2a,b and S1). Interestingly, we observed significantly lower bone erosion which was corroborated by lower osteoclast numbers in T06-treated CIA mice (Figures 2c,d and S2) and reduced expression of CatK in osteoclasts (Figures 2e,f and S2). μCT of the hind paws, knee and reconstituted trabecular bone in the distal femur indicated minor or no disease progression in T06-treated group compared with CIA control mice (Figure 2g,i). T06-treated mice revealed comparable parameters in BV/TV, Tb.N, Tb.Th, BMD and Tb.Sp relative to non-CIA controls whereas the ODN effect was weaker (Figure 2j). These results demonstrated that mice treated with T06 show a significant reduction in arthritic symptoms when compared with ODN.

FIGURE 2

Histological analysis of tissue sections prepared from hind paws of naive, CIA-, CIA-ODN- and CIA-T06-treated mice. (a) Toluidine blue staining showed extensive proteoglycan loss (visualized by fainted and lack of blue colour) and the destruction of cartilage during the progression of arthritis which was reduced with T06 treatment. (b) Quantification of histological arthritis score, in paw (both hind and front), ankle and knee of different mice groups confirmed the T06 inhibitory effect on the arthritis progression. (c,d) TRAP-positive osteoclasts (red) and their quantification, and (e,f) CatK expression in osteoclasts (brown) showed a reduction in T06-treated mice in femur head. All sections have the same magnification and are counterstained with haematoxylin. Data are presented mean ± SD (n = 4 in each group). Kruskal–Wallis test, two-tailed; ns, not significant, *P < 0.05 when compared with the CIA-untreated mice. μCT analysis confirmed the efficacy of T06 in CIA mice. Representative 3D model of (g) hind paw, (h) knee joints and (i) reconstituted bone trabecular in the distal femur of naive, CIA-, CIA-ODN- and CIA-T06-treated mice. (j) 3D parameters including bone volume/total volume (BV/TV), bone mineral density (BMD), trabecular number (Tb.N), trabecular thickness (Tb.Th) and trabecular spacing (Tb.Sp) of control, CIA- and inhibitor-treated mice. Values are the mean ± SD (n = 10 in each group). Kruskal–Wallis test, two-tailed, and Dunn's multiple comparisons test were used in most of the analyses. ns, not significant; *P < 0.05 when compared with the non-CIA mice.

A more detailed histological analysis of knee joints (safranin O / H&E staining) showed mostly intact cartilage surfaces with less cell infiltration, pannus formation and synovitis in the T06-treated group (more than 60% reduction). In contrast, there was a complete destruction of cartilage and trabecular bone observed in CIA mice with complete infiltration of inflammatory cells (Figures 3a,b and S3). ODN treatment showed no effect on inflammation parameters and arthritis progression but comparatively less eroded cartilage and bone surfaces when compared with CIA mice. Immunohistochemical analysis of the synovium revealed reduced TNF-α and CatK expression in T06-treated mice confirming the reduction of immune cell and synovial fibroblast activation in knee joints (Figure 3c,d). In contrast, no effect by ODN was observed. CatK, responsible for bone and cartilage destruction, was less expressed by synovial fibroblasts, macrophages, osteoclasts and other cells in the synovium of T06-treated mice. Similarly, in cartilage, expression of TNF-α, IL-1β and CatK was decreased in T06-treated mice when compared with the CIA control group (Figure 3e,f). Similar to the synovium, ODN had no significant effect on the expression of these proteins.

FIGURE 3

(a) Knee joint sections were stained with safranin O, a proteoglycan red marker showing severe cartilage damage in CIA- and CIA-ODN-treated mice and preservation of cartilage in T06-treated mice. (b) Quantification of inflammation, cellular infiltration, pannus tissue and cartilage/bone destruction was performed by ImageJ software in different conditions. Asterisks (*) highlights the synovium and arrow (↓) highlights cartilage in knee joint. (c) Immunohistochemical analysis revealing overexpression of TNF-α and CatK in synovium of the inflammatory pannus tissue. (d) Quantification of CatK and TNF-α expression in synovium showed their downregulation in T06-treated mice. (e) Chondrocytes of articular cartilage and osteoclast-like multinucleated cells showed massive overexpression of CatK, TNF-α and IL1β. However, reduction in the expression of these cytokines and CatK was observed in the T06-treated group. (f) Quantification of CatK, TNF-α and IL-1β expression in articular cartilage. Data represent mean ± SD (n = 4 in each group). Kruskal–Wallis test, two-tailed, and Dunn's multiple comparisons test; ns, not significant; *P < 0.05 versus CIA-untreated mice.

Furthermore, to determine the effect of T06 on the in situ joint production of pro-inflammatory cytokines (TNF-α, IL-6, IL-1β, IL-17, IFN-γ and RANKL) and ECM-degrading proteases (CatK and matrix metalloproteinases [MMPs]) both at mRNA, and protein levels were assessed in ankle and knee joints at the end of the experiment (Day 48). As shown in Figure 4a,b, expression of key inflammatory cytokines mediating arthritis progression was reduced in T06-treated mice by 50% or more when compared with ODN-treated and CIA control mice. In addition, serum cytokines were significantly reduced in T06-treated mice when compared with CIA controls (Table S2). The expression of CatK and MMPs which have been shown to participate in matrix destruction and processing of regulatory proteins was selectively decreased, based on their mRNA and protein levels in T06-treated but not in ODN-treated mice (Figure 4c–e).

FIGURE 4

Reduction of pro-inflammatory cytokines in the joints of T06-treated CIA mice. (a) Cytokine (TNF-α, IL-6, IL-1β, IL-17, RANKL and IFN-γ) levels were measured by ELISA in knee lysates. Cytokine concentrations in synovium was standardized to total protein levels (n = 4). (b) Cytokines (TNF-α, IL-6, IL-1β, IL-17, RANKL and IFN-γ) mRNA levels were analysed by real-time PCR in ankle joints (hindlimb). β-Actin was used as an internal control for normalization for real-time analysis. Data represent mean ± SD (n = 4 in each condition). *P < 0.05 versus CIA-untreated mice (Kruskal–Wallis test, two-tailed). (c,d) Protein levels of CatK, MMP13, MMP3 and TGFβ1 in knee lysate were identified by western blot analysis. Band intensities were quantified and normalized relative to the quantity of their respective actin bands and expressed as fold changes of the values in naive group (n = 3 in each condition). (e) Regulatory effect of T06 on gene expression levels of CatK, MMP13 and MMP3 were analysed by RT-qPCR in the knee of naive, CIA-, CIA-ODN- and CIA-T06-treated mice (n = 4). Significance was measured using Kruskal–Wallis test, two-tailed, and Dunn's multiple comparisons test. *P < 0.05 compared with untreated CIA. ODN-treated mice showed no significance difference compared with CIA-untreated mice at both protein and mRNA levels.

3.3 Effect of T06 on immune cells

Potential anti-inflammatory and anti-resorptive pathways contributing to joint damage in RA are summarized in Scheme S1. Effects of T06 and ODN treatment on immune and non-immune cells participate in these pathways were analysed. To compare the inflammatory cell response to the drug treatment in CIA mice with control mice (CIA inhibitor untreated, and non-CIA) in knee joints, we analysed local immune cells by flow cytometry (Figure 5). Podoplanin (PDPN), which has been implicated in chronic inflammatory processes, was used as marker to analyse synovial fibroblast populations (PDPN⁺/CD45⁻). PDPN⁺ fibroblasts showed a three-fold increase in CIA mice but were reduced by 55% in T06-treated mice (Figure 5a). CIA mice also exhibited myeloid cell (CD11b⁺) expansion, consistent with previous studies of autoimmune arthritis models (Hernandez et al., 2020) (Figure 5b). In addition, neutrophil (CD11b⁺Ly6G⁺) recruitment was reduced in inflamed limbs during T06 treatment by more than 50%. There was less than a one-fold increase in macrophage counts in the joints of T06-treated mice, compared with an approximately three-fold increase in the inflamed joints of CIA mice (Figure 5b). Flow-cytometric analysis revealed that dendritic cell (CD11c⁺) populations were reduced in the joints of T06-treated mice. Furthermore, decreased proportion of CD11c⁺CD11b⁺ DC subsets were observed in T06 treatment (Figure 5b). Quantification of CD3⁺ cells (a pan-T-cell marker) in the joints of arthritic mice revealed significantly lower numbers in T06-treated mice when compared with CIA control mice (Figure 5c). Analysis using a panel of T-cell-specific markers revealed an increase in CD4⁺ T-cells and a decrease in CD8⁺ T-cells in CIA joints. T06-treated mice showed a significant reduction of the CD4⁺ cell account but no significant effect on the CD8⁺ cell population. Of the CD4⁺ populations analysed, Th17 cells, a subset of T helper cells that not only strongly contributes to joint inflammation but also participates to accelerate osteoclast differentiation and bone destruction, were reduced by 60% to 70% after T06 treatment (Figure 5c). B-cell (B220⁺) numbers were decreased by almost half in T06-treated mice when compared with the CIA control group (Figure 5d). In contrast to T06, the selective CatK inhibitor, ODN, showed no effect on the distribution of the analysed immune cells when compared with CIA controls (Figure 5a–d).

FIGURE 5

Flow-cytometric gating strategy for immune cells from mice limbs. (a) Podoplanin positive cells (fibroblasts) were analysed by flow cytometry, and T06 treatment revealed obvious reduction in synovium fibroblasts. (b) Flow-cytometric gating strategy and quantifications of myeloid cells (CD45⁺CD11b⁺), neutrophils (CD45⁺CD11b⁺Ly6G⁺), dendritic cells (CD45⁺CD11c⁺) and subgroup of DCs (CD11⁺CD11c⁺) in cell suspensions from hind paws. (c) Flow-cytometric gating strategy of T-cell populations within inflamed joints. CD3 T-cells (CD45⁺CD3⁺) in limbs harvested and subsets of T-cells (CD4⁺ T-cells and CD8⁺ T-cells) were quantified. Subsets of CD4⁺ T-cells (Th17) in limbs also were analysed. (d) Flow-cytometric gating strategy of B-cells (CD45⁺B220⁺) and their quantification in naive, CIA- and CIA-treated mice. (e) Table of antibodies used in this experiment with specific fluorochrome. Graph shows individual data points with mean values (naive n = 5; CIA n = 6; CIA-ODN n = 5; CIA-T06 n = 5 mice) for all experiments. Data presented in fold change is normalized to cell numbers in control limbs. Quantification graph shows individual data points and mean values. T06-treated mice showed significant reduction in the number of myeloid cells, neutrophils, dendritic cells, T-cells, Th17 cells and B-cells. ODN treatment showed non-significant change in the cell population in CIA mice. Statistics: Kruskal–Wallis test, two-tailed, and Dunn's multiple comparisons test compared with untreated CIA. ns, not significant; *P < 0.05.

To further determine the impact of T06 and ODN on systemic immune cell populations in CIA mice, we performed flow cytometry on blood resident cells. The gating strategy and used antibodies are illustrated in Figure 6a,b. Blood parameters showed a 40% decrease in dendritic cells (CD11c⁺), a 55% decrease in myeloid cells (CD11b⁺) and an ~70% reduction in neutrophils (Figure 6c). Concerning the T-cell populations, we observed a significant decrease in the percentage of CD4⁺ cells, especially in the Th17 subset of CD4⁺ T-cells which showed a reduction of over 60% in T06-treated mice relative to untreated CIA mice (Figure 6d–f). A decreasing trend in CD8⁺ T-cells in CIA mice was observed, which seems to be recovered by T06 treatment. Similar to knee joints, a 40% decrease in B-cells (B220⁺) was observed in T06-treated animals when compared with CIA mice (Figure 6f).

FIGURE 6

(a) Flow-cytometric gating strategy and (b) Table of antibodies used in this experiment with specific fluorochrome. (c) quantifications of myeloid cells (CD45⁺CD11b⁺), neutrophils (CD45⁺CD11b⁺Ly6G⁺) and dendritic cells (CD45⁺CD11c⁺) in peripheral blood of naive, CIA- and CIA-treated mice. (d,e) Flow-cytometric gating strategy of T-cell populations within peripheral blood. CD3⁺ T-cells (CD45⁺CD3⁺) in blood harvested and subsets of T-cells (CD4⁺ T-cells and CD8⁺ T-cells) were quantified. Subsets of CD4⁺ T-cells (Th17) in peripheral blood also were analysed. Representative flow cytometry plots showing significant reduction in the fraction of Th17 T-cells in the blood of CIA-T06-treated mice. (f) Quantification of different subsets of T-cells (CD3⁺, CD4⁺ T-cells, CD8⁺ T-cells and Th17) in the blood. Graphs show individual data points with mean values (naive n = 5; CIA n = 5; CIA-ODN n = 5; CIA-T06 n = 5 mice) for all experiments. Data presented in fold change which is normalized to cell numbers in control samples. T06-treated mice showed significant reduction in the number of myeloid cells, neutrophils, dendritic cells (DCs), T-cells, Th17 cells and B-cells. No significant change in the blood cell population was observed with ODN treatment in CIA mice. Statistical difference was determined by Kruskal–Wallis test, two-tailed, and Dunn's multiple comparisons test compared with untreated CIA. *P < 0.05.

Flow cytometry analysis of spleen-resident cells in CIA mice showed no significant change in B-cells (B220⁺) and only slight decreases in CD4⁺ and CD8⁺ T-cell levels (Figure 7a–c) when compared with non-CIA mice. In contrast, Th17 cells strongly increased in CIA mice about four-fold. As shown for joints and blood cell levels, ODN had no effect on the distribution of these cells whereas T06 slightly increased CD4⁺ and CD8⁺ populations and decreased almost two-fold Th17 cells in the spleen. B-cells (B220⁺) were not affected by T06 (Figure 7c). Splenic CD11b⁺ cells showed more than a two-fold increase in CIA which was reduced by ~35% in T06-treated mice but not by ODN (Figure 7d,e).

FIGURE 7

(a) Flow-cytometric gating strategy to analyse the splenic cell populations. (b) Representative flow cytometry plots showing significant reduction in Th17 subset of T-cells in T06-treated CIA mice. (c) Percentage of main lymphoid populations in spleen: B-cells (CD45⁺B220⁺) and T-cells (CD4⁺, CD8⁺ and Th17) from control, CIA-, CIA-ODN- and CIA-T06-treated mice. (d,e) Flow-cytometric analysis of CD11b⁺ cells in the mice spleen showed reduction in the percentage of CD11b⁺ cells in T06-treated mice. (f) Table of antibodies used in this experiment with specific fluorochrome (naive n = 5; CIA n = 5; CIA-ODN n = 5; CIA-T06 n = 5 mice) for all experiments. Data presented in fold change and quantification graph show individual data points and mean values. Statistics: Kruskal–Wallis test, two-tailed, and Dunn's multiple comparisons test compared with untreated CIA. ns, not significant; *P < 0.05.

3.4 T06 effectively suppresses NF-κB activation

The nuclear translocation of p65 NF-κB was significantly reduced in T06-treated macrophages as shown by immunostaining (Figure 8a). T06 treatment slightly affected the relative levels of phosphorylated IKKα/β expression but significantly reduced the relative levels of phosphorylated IκBα and NF-κB p65 (Figure 8b,c). This effect results in a reduced release of NF-κB (p65) and its translocation into the nucleus. Because the phosphorylation of IκBα/NF-κB depends on IKKα and IKKβ activities, we tested the efficacy of T06 to inhibit either IKKα or IKKβ. In vitro assays demonstrated that T06 inhibits IKKβ (IC₅₀ 11.9 ± 3.8 μM) comparatively better than IKKα (IC₅₀ 28.7 ± 4.4 μM) (Figure S4A,B). We also analysed if upstream phosphorylation of IKKα or IKKβ by TAK1 required for their activation was interfered by T06 (Scheme 1) We only found a weak inhibitory effect of T06 on TAK1 activity (IC₅₀ 47.8 ± 5.3 μM) suggesting that it is likely a more direct effect on IKKβ preventing phosphorylation of IκBα (Figure S4C). After identifying the effect of T06 on the phosphorylation and translocation of p65 NF-κB at the protein level (Figure 8b,c), we determined the role of T06 in the regulation of nuclear transcription by assessing the binding of nuclear NF-κB (p50, p52, p65, c-Rel and RelB) subunits to promoter consensus sequences on immobilized DNA oligonucleotides. The decrease in binding of NF-κB subunits to consensus sequences was only observed in T06-treated but not in ODN-treated macrophages (Figure 8d). A similar decrease in the binding of NF-κB subunits was observed in the nuclear extracts from ankle joints of T06-treated mice compared with CIA mice (Figure 8e). These results support the reduced expression and secretion of TNF-α and other cytokines in T06-treated mice when compared with CIA mice as shown in Figure 4a,b. A significant reduction of secreted TNF-α protein as observed by ELISA was seen in T06-treated macrophages (2.72 ± 0.32 ng·ml⁻¹) compared with LPS-stimulated macrophages (5.02 ± 0.40 ng·ml⁻¹). In contrast, ODN did not show any anti-inflammatory effects on NF-κB activation in macrophages and CIA mice.

FIGURE 8

(a) Effect of T06 on nuclear translocation of NF-κB p65 induced by LPS. Macrophages were treated by LPS (50 ng·ml⁻¹) in presence and absence of inhibitors (T06: 10 μM and ODN: 1 μM) for 4 h; then, nuclear translocation of p65 was determined by immunofluorescence analysis. LPS activated the NF-κB signalling pathway and induced p65 (red) translocation into the nucleus (blue) in macrophages. However, T06-treated cells showed less p65 translocation into the nucleus confirmed the role of T06 in interfering the NF-κB signalling pathway (n = 4 in each condition). Scale bar: 10 μm. (b,c) Phosphorylation of IKKα/β, IκBα and p65 during translocation of NF-κB p65 was analysed by western blot analysis in total cell extracts (n = 3 in each condition). Quantification of the activation of NF-κB family members (p50, p52, p65, c-Rel and RelB) in (d) macrophages and (e) ankle joint nuclear extracts (hindlimb) of control, CIA-, CIA-ODN- and CIA-T06-treated mice (n = 4). Nuclear lysates were assayed for transcriptionally active, nuclear NF-κB using NF-κB transcription factor ELISA assay. Statistics: Kruskal–Wallis test, two-tailed, compared with LPS-treated condition or untreated CIA mice. ns, not significant; *P < 0.05.

SCHEME 1

NF-κB activation pathways: In the canonical pathway, activation of TLRs and TNFR receptors causes phosphorylation of the inhibitory IκBα by the IKK complex, leading to their phosphorylation and ubiquitin-dependent proteasomal degradation. NF-κB subunits RelA (p65/p50) and c-Rel (c-Rel/p50) free form inhibitory interaction with the IκBs translocating to the nucleus to activate the transcription of targeted NF-κB target genes. Here, we observed that T06 prevents the necessary IKK-mediated phosphorylation of inhibitory IκBα and resulted in less nuclear translocation of the NF-κB dimers (p65/p50), downregulating the gene expression of inflammatory cytokines (TNF-α) and proving its anti-inflammatory property. T06 inhibits IKKβ at (IC₅₀ 11.9 ± 3.8 μM) and IKKα at (IC₅₀ 28.7 ± 4.4 μM). Minor inhibition of TAK1 by T06 confirmed that TAK1-mediated phosphorylation of IKK subunits was not affected by T06. An additional step of phosphorylation of NF-κB (p65) in their nuclear translocation also was reduced with T06 treatment. In noncanonical NF-κB signalling pathways, activation of receptors (BAFF and lymphotoxin β) causes phosphorylation of NIK that participate in activation of IKKα, subsequently leading to the phosphorylation and proteasome-dependent processing of p100 into p52, ultimately resulting into nuclear translocation of RelB-p52 heterodimers and activation of the transcription of targeted NF-κB genes. T06 reduced the overall activation of NF-κB family members (p50, p52, p65, c-Rel and RelB) confirmed by the less active subunits present in nuclear lysates of T06-treated condition.

4 DISCUSSION

Here, we report the dual anti-resorptive and anti-inflammatory effects of an ectosteric CatK inhibitor (T06) in a mouse model of RA and compared its efficacy with a highly selective CatK inhibitor (ODN) (Anderson et al., 2014). CatK is a well-proven pharmaceutical target to inhibit osteoclast-mediated bone and cartilage degradation (Bone et al., 2015; Hou et al., 2001; Morko et al., 2005; Salminen-Mankonen et al., 2007) and, as expected, its inhibition significantly reduced type I and II collagen degradation markers in blood and improved various bone histomorphometric markers. The overall anti-resorptive efficacies of T06 and ODN were comparable, and these effects were likely of systemic nature and included the reduction of the overall bone turnover. In contrast, only T06 could suppress both autoimmune inflammation and tissue degradation in the arthritic joints. None of the mice in the T06-treated group advanced to severe disease as demonstrated in clinical scores, histology and μCT analysis which confirmed the absence or progression of severe joint inflammation. Also, the downregulation of CatK and MMPs expression was noted in the synovial tissue of T06-treated CIA mice which further corroborated with the reduction in bone/cartilage destruction within joints. This is highly relevant because overexpression of CatK has been previously reported at the site of cartilage/bone erosion in synovial tissue (Zhuo et al., 2014). More recently, it has been observed that CatK participates in inflammation in an osteoclast-independent manner in autoimmune arthritis (Asagiri et al., 2008). We hypothesize that the mechanisms of action of T06 involve anti-resorptive and anti-inflammatory effects through the blockade of NF-κB signalling pathways, the reduction of immune cells in inflamed joints and a drop in the expression of matrix-degrading proteases and pro-inflammatory cytokines.

Abnormal activation and infiltration of immune cells and the overexpression of pro-inflammatory factors within joints contribute to perpetuating inflammation in RA (Komatsu & Takayanagi, 2022; McInnes et al., 2016), and also increases the number of CatK-expressing cells such as osteoclasts and synovial fibroblasts. Signalling pathways triggered by the cellular exposure to growth factors, chemokines and cytokines subsequently lead to unregulated expression of genes involved in inflammation (Komatsu & Takayanagi, 2022; Liu et al., 2017; McInnes et al., 2016; Müller-Ladner et al., 2005). Our data suggest that T06-mediated reduction in the severity of arthritis is based upon the early restriction of the influx of pro-inflammatory cells and the suppression of pro-inflammatory cytokines such as TNF-α, IL-6, IL-1β, IL-17 and RANKL. The expression of these cytokines, which are involved in the activation of immune cells, and the differentiation of monocytes into synovial fibroblasts and osteoclasts was arrested by T06 treatment. The suppression of the expression of IL-17 and the Th17 cell population by T06 also explains the additional reduction in bone damage associated with IL-17-regulated osteoclast formation (Sato et al., 2006). Interestingly, we observed ~40% less osteoclasts present in the joints when compared with ODN treatment and thus a significant reduction of CatK expression in affected joints.

Of interest is the differential effect of T06 on immune cells in the inflamed joint tissues, blood and spleen. Whereas the reduction of myeloid cells, neutrophils, dendritic cells, CD4⁺, Th17 and B-cells in the presence of T06 was comparable in joints and blood samples, the effect on splenic cell counts was negligible with the exception of Th17 cells. This outcome indicates that the anti-inflammatory effect of T06 is not systemic but rather targets inflamed sites. Interestingly, T06 did not alter the CD8⁺ population size in any of the organs investigated and thus may not interfere with killer T-cell activities in pathogen defence or tumour suppression. As expected, ODN did not affect any of the immune cell populations. It should be emphasized that cytokine expression was only partially blocked by T06 (40% to 60%) at the dose tested, indicating that there remains space for anti-infectious activities when needed. This is comparable with results obtained for the reduction of inflammatory cytokines in mice treated with anti-TNF-α and anti-IL-1β (Williams et al., 2000). Treatment of CIA mice with etanercept, a TNF-α receptor antagonist, yielded similar outcomes as T06 in terms of μCT-based histomorphometric parameters and in the reduction in arthritis scores when given three times per week at 100 μg per mouse. Interestingly, at a four-fold dose, the TNF-α receptor antagonist completely lost its efficacy after 3 weeks due to the formation of anti-etanercept antibodies (Yi et al., 2014). This may explain why etanercept and other biologicals against RA are only effective in a subpopulation of patients (Atzeni et al., 2008; Finckh et al., 2006). This increasing refractory effect to biologicals would not be an issue for low molecular drugs such as T06.

Based on these observations, we elucidated pathways of T06 activity on cytokine expression which is ultimately connected with the recruitment of immune cells. Here, we observed the interference of T06 in the phosphorylation of IκBα/NF-κB during NF-κB signalling, which resulted in reduced activation and expression of TNF-α. This indicates that T06 interferes with upstream kinases, such as the IκB kinase (IKK) complex consisting of two kinases (IKKα and IKKβ). We observed that T06 inhibits IKKβ more than two-fold better than IKKα. IKKβ is considered to represent the more crucial activity in the canonical NF-κB pathway (Israël, 2010). TAK1, an upstream kinase of the NF-κB pathway, is only very poorly inhibited by T06 and thus can be excluded as a substantial T06 drug target. At this point, we cannot exclude additional T06 effects on other upstream components of the NF-κB pathway such as the regulatory subunit NF-κB essential modulator (NEMO) or with the dimerization of IKKα and IKKβ. Further investigation will shed light on the exact interference sites of T06 in the mechanism of NF-κB subunit phosphorylation. An additional interaction site for T06 might represent the inhibition of MSK1 because we observed a significant reduction of the phosphorylation of p65. However, this reduction also could be a consequence of the inhibition of IκBα and thus the release of the p65/p50 complex.

In conclusion, RA pathogenesis represents a network of interactions between mutually affecting tissues and cells (bone, cartilage, synovium and immune system) which amplifies inflammatory and tissue-destructive processes. Members of the tanshinone family, such as T06, have the unique property of controlling both pro-inflammatory and tissue-erosive pathways in RA by targeting the disease dominant collagenase, CatK and the pro-inflammatory NF-κB pathway. Moreover, T06 may even exert an anti-pain activity as recently demonstrated in an antibody-induced arthritis mouse model (Jurczak et al., 2022). A small molecule such as T06 that has dual or even pleiotropic activities targeting multiple disease-associated interactions would represent a potentially highly effective drug when compared with treatments designed for single selected targets with a high potency as currently employed using various biologicals.

AUTHOR CONTRIBUTIONS

P. Panwar and D. Brömme designed the study, validated results and were responsible for administration of project. P. Panwar performed the experiments and analysed the data. Data were validated by P. Panwar and discussed by all authors. D. Brömme participated in NF-κB activation study. D. Brömme acquired funding for the study. P. Panwar wrote the original draft of the manuscript. All authors reviewed and revised the final manuscript.

ACKNOWLEDGEMENTS

This work was supported by Canadian Institutes of Health Research (Grants PJT-391020 and CPG-396661) and Collaborative Health Research Projects (Grant CHRP 523434). D.B. was supported by Canada Research Chairs award funding. We are thankful to the CFI-supported UBC Centre for High-Throughput Phenogenomics for the use of the fluorescent microscope and micro-computed tomography (μCT). We thank the flow cytometry facility (ubcFLOW) for their support in conducting cell sorting experiments. We are thankful to Nooshin S. Safikhan for her help during animal tissue collection. We are grateful to Angela Tether, Research Grant Facilitator, University of British Columbia, for critical English-language proofreading.

CONFLICT OF INTEREST STATEMENT

No disclosures and all authors have no conflicts of interest.

DECLARATION OF TRANSPARENCY AND SCIENTIFIC RIGOUR

This Declaration acknowledges that this paper adheres to the principles for transparent reporting and scientific rigour of preclinical research as stated in the BJP guidelines for Natural Products Research, Design and Analysis, Immunoblotting and Immunochemistry and Animal Experimentation and as recommended by funding agencies, publishers and other organizations engaged with supporting research.

Open Research

DATA AVAILABILITY STATEMENT

All data generated or analysed during this study are included in this article (and its Supporting Information). Sufficient information regarding the methods and analysis is given in the article to reproduce reported results, and no additional data have been shared.