Background
Rheumatoid arthritis (RA) is a chronic inflammatory disorder characterized by synovial hyperplasia, angiogenesis, inflammatory cell infiltration, pannus formation, cartilage destruction, and bone erosion. As RA progresses, it can lead to joint destruction and disability [
1].
The nuclear factor-κB (NF-κB) is a family of inducible transcription factors that regulates numerous genes involved in immune and inflammatory responses. Various factors, such as IL-1, IL-17, TNF-α, platelet-derived growth factor, lipopolysaccharide, oxidative stress, and viral products, can induce NF-κB activation. NF-κB, in turn, triggers the transcription of IL-1, TNF-α, IL-6, IL-8, IL-17, GM-CSF, and inducible nitric oxide synthase [
2]. NF-κB plays a significant role in RA pathology [
3], and novel therapeutic strategies aimed at specific inhibition of key elements in the NF-κB activation pathway have been under development in recent years [
4,
5].
Exosomes are small membrane-enclosed vesicles released by cells for intercellular communication. They are derived from the fusion of multivesicular bodies with the cell membrane [
6]. Exosomes, as biologically derived nanoparticles, offer efficient drug delivery and excellent biocompatibility with minimal side effects. They can elicit robust cellular responses both in vitro and in vivo, making them promising therapeutic agents [
7,
8]. NF-kB exists in an inactive state within the cytoplasm of nearly all mammalian cells, and it forms associations with inhibitory proteins collectively known as IκB (inhibitory κB proteins). Our group has previously reported that delivering exosomal super-repressor IκB (Exo-srIκB) using “EXPLOR” (exosomes for protein loading via optically reversible protein–protein interactions) technology [
9] can alleviate inflammation in disease models such as sepsis [
4] and acute kidney injury [
5]. However, the potential of delivering srIκB using exosomes in RA remains unexplored.
With this in mind, we employed the same technology to load srIkB into exosomes and deliver them systemically as Exo-srIkB to assess their impact on RA. This study aimed to investigate the efficacy of Exo-srIκB in alleviating arthritis, bone damage, and inflammation using human ex vivo samples and mouse models of RA.
Materials and methods
Human samples
All patients met the RA criteria established by the American College of Rheumatology/European League Against Rheumatism [
10]. Peripheral blood mononuclear cells (PBMCs) and synovial fluid mononuclear cells (SFMCs) were collected from active RA patients. The demographic characteristics of the patients are provided in Table
1. The study received approval from the Ethics Committee of Chonnam National University Hospital (CNUH). Written informed consent was obtained from all participants (institutional review board [IRB] no. CNUH-2011–199).
Table 1
Clinical characteristics and laboratory findings of patients with RA
Total number | 9 | 6 |
Age, mean ± SD (years) | 60.8 ± 19.2 | 52.5 ± 20.3 |
Male, n (%) | 3 (33.3) | 1 (16.6) |
RF positive, n (%) | 8 (88.8) | 5 (83.3) |
Anti-CCP Ab positive, n (%) | 8 (88.8) | 5 (83.3) |
Recent medications |
Steroid, n (%) | 7 (77.7) | 6 (100.0) |
Steroid dose, mean ± SD | 5.4 ± 3.4 | 6.6 ± 2.7 |
Methotrexate use, n (%) | 9 (100.0) | 6 (100.0) |
Leflunomide use, n (%) | 3 (33.3) | 1 (16.6) |
Tacrolimus use, n (%) | 3 (33.3) | 2 (33.3) |
TNF-α blocker use, n (%) | 2 (22.2) | 0 (0.0) |
Jak inhibitor use, n (%) | 0 (0.0) | 0 (0.0) |
Production of Exo-srIκB
The process of exosome production was previously described [
11,
12]. Briefly, Expi293F-producing cells were incubated in a wave culture system for 4 days and exposed to blue light for target protein loading and exosome production. Next, the harvested culture medium was centrifuged at 2000 g for 10 min to remove cells and debris and filtered with a 0.22-µm polyethersulfone filter to remove large particles. The exosome was subsequently purified using ultrafiltration and diafiltration for concentration, buffer exchange, and anionic and multimodal resin chromatography. Finally, a formulation and sterilization filter process was performed.
Characterization of Exo-srIκB
The morphology and lipid bilayer of extracellular vesicles (EVs) were absorbed on carbon-coated copper, stained with 2% uranyl acetate, and confirmed by transmission electron microscopy. Nanoparticle tracking analysis was used to measure EVs’ particle number and size distribution using NS300, and samples were diluted (1:100–1:10,000) in particle-free PBS to an acceptable concentration. Immunoblotting was performed by lysing cells in RIPA buffer or exosomes, followed by SDS/PAGE gel electrophoresis and transfer onto nitrocellulose membrane. Membranes were blocked with 5% skim milk in Tris-buffered saline containing 0.1% Tween-20 (TBS-T) and probed with primary antibodies against srIκB, CRY2 (customized antibody, AbClon, Seoul, Korea), CD9, CD81 (SBI, Tokyo, Japan), TSG101, Alix, GM130, calnexin (Abcam, Cambridge, UK), lamin B1, GAPDH (Santa Cruz Biotechnology, Dallas, TX, USA), and prohibitin (Novusbio, Centennial, CO, USA) at 4 °C overnight. After incubation with specific secondary antibodies, blots were developed using Clarity and Clarity Max ECL Western blotting substrates and imaged with the ChemiDoc imager.
Cell viability assay
Cell viability was assessed using the Cell Titer 96 AQueous One Solution Reagent (G3580, Promega, USA). Cells were seeded and treated with varying concentrations of Exo-srlκB for specified durations. Following the manufacturer’s instructions, 20 μl of MTS solution was added to 100 μl of cell culture medium and incubated at 37 °C for 2–4 h. Absorbance was then measured at 490 nm using a Molecular Devices Reader 96-well microplate reader (USA). For flow cytometry analysis, whole cells were surface stained with anti-Fixable Viability Dye-eFluor780 (65–0865-14, Invitrogen, USA).
Co-culture of Exo-srIκB with human inflammatory cells
PBMCs and SFMCs were isolated and cultured in RPMI1640 media (LM011-01, Welgene, Korea) supplemented with 10% fetal bovine serum (S001-01, Welgene, Korea) and 1% penicillin–streptomycin solutions (LS202-02, Welgene, Korea). Cells were seeded at a density of 5 × 105 cells/well in a 96-well plate. After a 3-h pretreatment with non-engineered cell-derived control exosomes (Exo-Naïve) or Exo-srIκB, the cells were stimulated with phorbol 12-myristate 13-acetate (PMA; P1585, Sigma, USA) at a concentration of 100 ng/ml, ionomycin (I9657, Sigma, USA) at a concentration of 1 μM, and brefeldin A (a Golgi plug protein transport inhibitor; 555,029, BD, USA). The cells were then incubated in CO2, 37 °C incubator for 4 h. Following stimulation, cells were stained with anti-Fixable Viability Dye-eFuor780 (65–0865-14, Invitrogen, USA). After washing, cells were fixed and permeabilized using perm/wash buffer and stained with APC-conjugated anti-IL-17A (512,334, BioLegend, USA) and PerCP-Cy5.5-conjugated anti-GM-CSF (502,312, BioLegend, USA) antibodies. FlowJo Software (BD, USA) was used for flow cytometry analysis. Additionally, supernatants from PBMCs and SFMCs were analyzed for human IL-17A, IL-6, and TNF-α levels using ELISA kits (88–7176, 88–7066, 88–7344, Invitrogen, Austria), and human GM-CSF levels were measured using an ELISA kit (K0331120, LABISKOMA, Korea). Optical density (OD) was measured at 450 nm using a SpectraMax® M2 microplate reader (Molecular Devices Corp., USA).
Experimental arthritis mouse model, intervention and scoring
Experiments were conducted with the approval of the Institutional Animal Care and Use Committee (animal experiment IRB no. CNU IACUC-H-2021–17). SKG mice on a BALB/c background were obtained from CLEA Japan (Tokyo, Japan) and housed in a specific pathogen-free (SPF) facility. The negative control mice were not injected with curdlan to assess the baseline response in the absence of the experimental intervention (n = 9). Eight-week-old female mice were treated with curdlan (3 mg/kg) by intraperitoneal injection. After the onset of symptoms following curdlan injection, the mice were randomly stratified into two groups (n = 9 per group): one receiving Exo-Naïve treatment and the other receiving Exo-srIκB treatment. Either Exo-Naïve (1 × 1010 pn/0.2 ml) or Exo-srIκB (1 × 1010 pn/0.2 ml) was repeatedly administered intraperitoneally three times a week until sacrifice.
Collagen-induced arthritis (CIA) model was prepared in 8–9-week-old male DBA/1 J mice and conducted in a SPF facility. The mice underwent primary immunization followed by a 21-day interval before secondary immunization. Random assignment of experimental groups was performed prior to the study (n = 5 mice for each group). Repeated treatments of all the test substances were initiated on day 23. The negative control mice were not immunized, serving as a baseline for assessing the response in the absence of the experimental intervention. The Exo-Naïve (1 × 1010 pn/0.2 ml) group received only exosome treatment, while the treatment group received 1 × 1010 pn/0.2 ml of Exo-SrlκB, which was administered three times a week until sacrifice. Methotrexate (MTX), as a positive control, was intraperitoneally administered at a dosage of 1 mg/kg per administration, given twice a week.
Clinical signs were monitored and scored twice a week by two independent observers using the following scale for affected joints: 0 = asymptomatic, 1 = slight swelling of ankles or toes, 2 = severe ankle swelling, 3 = severe ankle and toe swelling, and 4 = ankle and toe swelling with twisting [
13].
Tissue preparation and histological analysis
After completing the experiment, mice were sacrificed, and ankle samples were collected. Ankle specimens were fixed in 10% formalin for 1 week, decalcified in 10% formic acid at 37 °C for 1 week with shaking, and then embedded in paraffin. Paraffin blocks were sectioned at a thickness of 3.5 µm and deparaffinized using neo-clear (109,843, Merck, USA). Gradually graded ethanol was used for hydration, followed by staining with hematoxylin (105,174, Merck, USA) and eosin (HT110216, Sigma, USA). Additionally, safranin-O staining was performed on the joints to assess cartilage destruction. Two blinded readers independently scored the histological arthritis samples according to a previous report [
14].
Immunofluorescent staining
The section slides were deparaffinized in neo-clear and rehydrated in serial ethanol, followed by antigen retrieval with proteinase K (Abcam, ab64220) at RT for 30 min and blocking with BLOXALL (Vector, SP-6000) for 1 h. To observe co-localization of CD4, IL-17A, and TNF-α, the slides were incubated with primary mouse antibody for TNF-α (1:100, Santa, sc-52746), rabbit antibody for IL-17A (1:50; Abcam, ab79056), and rat antibody for CD4 (1:50; Santa, sc-19641) at RT for 1 h, followed by incubation with 488-conjugated anti-mouse antibody (1:100; Invitrogen, A11001), Cy3-Alexa-conjugated anti-rabbit antibody (1:100; Jackson ImmunoResearch, 111–165-144), and Cy5-conjugated anti-rat antibody (1:100; Jackson ImmunoResearch, 712–175-153) at RT for 1 h. To avoid nonspecific staining, the stained slides were treated with DAPI using the Autofluorescence Quenching Kit (Vector, SP-8500). Immunofluorescent images were collected by a confocal microscope (Leica Microsystem, Germany).
Micro-computed tomography analysis
The Quantum FX (μCT, Perkin Elmer) was utilized for imaging purposes. The scanning parameters were configured to 90 kV and 180 uA, with a scan duration of 2 min. The field of view (FOV) encompassed 20 mm, and the resolution achieved was 40 µm. Following the immobilization of the representative foot tissue that best reflected the clinical indicators of each group, a μCT scanner was employed for scanning. Radiographic images were acquired using the Quantum FX μCT imaging system (Perkin Elmer, MA, USA) and subsequently subjected to 3D rendering. Radiographic scoring was conducted, involving the independent evaluation of joint destruction by two researchers [
15]. The scoring value was determined by averaging the evaluations of the researchers, and the average scoring value of the foot tissue was computed as the score for each group.
Statistical analysis
Statistical analysis was conducted using Prism 9.0 Software (GraphPad Software, San Diego, CA, USA). Differences between means were evaluated for statistical significance using various tests, including Kruskal–Wallis test with Dunn’s multiple comparisons, T-test, Wilcoxon matched-pairs signed-rank test, two-way analysis of variance (ANOVA), and Mann–Whitney test. Significance levels were indicated on the graphs as follows: *P < 0.05, **P < 0.01, ***P < 0.001, and ****P < 0.0001. A P-value less than 0.05 was considered statistically significant.
Discussion
NF-κB is upregulated in RA synovium and contributes to RA pathology by enhancing the proliferation, invasion, and survival of fibroblast-like synoviocytes [
3]. The activation of NF-κB leads to dysregulation of osteoclasts and osteoblasts, resulting in increased bone resorption. Inhibiting NF-κB reduces the secretion of pro-inflammatory mediators and decreases both osteoclast differentiation and bone resorption [
16]. Although conventional anti-inflammatory and antirheumatic drugs are known to inhibit NF-κB activation, their potency as NF-κB inhibitors is limited, and they often lack specificity.
Studies using animals with genetically inactivated NF-κB signaling have shown promising results on specific NF-κB inhibition in RA treatment [
17]. These genetic engineering studies align well with experiments utilizing highly specific NF-κB inhibitors. Miagkov et al. demonstrated the efficacy of liposomal delivery of NF-κB decoys in preventing the recurrence of streptococcal cell wall arthritis in rats [
18]. Similarly, the administration of NF-κB decoys reduced the severity of CIA in rats and suppressed IL-1 and TNF-α production within the joints [
19]. Nevertheless, the safety of specific NF-κB inhibitors remains a concern, as NF-κB is crucial for normal development, including the protection of the liver against apoptosis and the immune response against pathogens. Thus, systemic inhibition of the NF-κB pathway could potentially lead to adverse effects [
20].
Exosomes are promising for drug delivery due to their safety and unique characteristics. These cell-derived vesicles have natural lipid bilayers, ensuring biocompatibility and minimize immune reactions or toxicity in vivo. Low levels of immunogenic surface proteins allow exosomes to evade immune recognition, thereby reducing associated adverse effects. They possess unique characteristics ideal for protein delivery, including biocompatibility, minimal toxicity, extended circulating half-life, stability, and customizable targeting efficiency [
21].
Many types of exosomes derived from body fluids have been found to be immunoregulatory in RA [
22,
23]. However, exosomes originating from blood plasma or serum have heterogeneous cellular origins and a poorly defined composition. Utilizing biomimetic exosomes loaded with dexamethasone sodium phosphate nanoparticles effectively enhances the therapeutic impact of glucocorticoids against RA [
24]. However, the absence of a separation mechanism between cargo proteins and lipid nanoparticles not only limits the efficiency of cytosolic delivery but also means that the preparation of these particles often involves complicated protein purification steps. Exosomes derived from IL-4 dendritic cells (DCs) exhibit the potential to reduce both the severity and incidence of established CIA [
25]. Moreover, administering a single systemic dose of exosomes sourced from IL-10 DCs after the onset of CIA has proven effective in ameliorating disease progression [
26]. Research findings indicate that systemic injection of DC/FasL exosomes is an effective treatment for established murine CIA [
27]. Furthermore, the injection of miR-150-5p-enriched exosomes derived from mesenchymal stem cells (MSCs) leads to reduce hind paw thickness and improved clinical arthritic scores in a CIA mouse model [
28]. However, the use of these genetically modified DC or MSC cells may face stringent regulatory scrutiny; manufacturing standards must all be satisfactorily addressed. In the present study, we used a protein carrier, EXPLOR, which has a higher loading capacity and delivery efficiency [
9]. In addition, we examined the efficacy in human samples from patients.
We observed that administering Exo-srIκB did not significantly impact cell viability. Flow cytometry analysis corroborated this, showing no decrease in cell viability for PBMCs and SFMCs. These outcomes suggest that Exo-srIκB does not compromise cell viability. Our findings reveal that Exo-srIκB effectively mitigates inflammatory cytokines in both PBMCs and SFMCs from RA patients. Furthermore, we observed significant improvements in clinical arthritis, inflammatory cytokine production, joint damages, and inflammatory cell infiltration in animal models of RA following the administration of Exo-srIκB. In the subset analysis of PBMCs, Exo-srIκB suppressed inflammation in human monocytes and CD4-positive cells, as indicated by our in vitro results (Suppl. Figure
3A, B). However, it did not produce the same effect on CD8 or mucosal-associated invariant T (MAIT) cells (Suppl. Figure
3C, D).
This study has limitations. We did not thoroughly investigate the effects of Exo-srIκB on specific immune cell subtypes in the ankle joint, apart from CD4 cells. Future investigations should explore the impact of Exo-srIκB on various cell types at inflammatory sites through expanded differentiation and analysis. We need to conduct further research on the specific mechanisms by which Exo-srIκB affects monocytes and CD4-positive T cells in RA. In this pilot study, Exo-srIkB exhibited effectiveness comparable to methotrexate. Additional experiments exploring drug dosage and intervals are necessary to establish a more optimized treatment regimen for RA.
Publisher’s Note
Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.