Carrier-Free Gambogic Acid Dimer Self-Assembly Nanomedicines for Rheumatoid Arthritis Treatment

Introduction

Rheumatoid arthritis (RA) is a chronic autoimmune inflammatory disease that can cause synovial inflammation, joint destruction, and disability, thereby severely reducing the quality of life of RA patients.¹ According to the statistics, RA affects approximately 1% of the global population and 0.28% of population in China.² In view of the limited clinical overcomes and severe side effects of these anti-RA agents in the market, anti-inflammatory and immunosuppressive natural compounds have attracted increasing attention. Gambogic acid (GBA) is a natural compound that exhibits significant therapeutic advantages in cancer and RA treatment.^3–6 To overcome its poor aqueous solubility and undesirable biodistribution, which lead to unavoidable systemic toxicity issues,⁷ a series of polymer matrix-based nano-delivery systems for GBA like micelles, nanoparticles, nanofibers, and nanorods have been developed for cancer treatment.^8,9 However, few studies have focused on targeted delivery of GBA for RA treatment. Currently, nanomedicines are thought to passively accumulate in chronically inflamed tissues via a similar extravasation through leaky vasculature and subsequent inflammatory cell-mediated sequestration (“ELVIS”) effects. Previous findings have demonstrated promising opportunities for the use of nanomedicines in the RA treatment.^10–12

However, these nanosystems based on nanocarrier materials have some common drawbacks, such as a complex drug encapsulation process, low drug loading efficiency, and poor batch-to-batch reproducibility in large-scale production.^13,14 Additionally, the repeated administration of large amounts of carrier materials in the body poses critical challenges for clinical translation owing to potential adverse side effects. Carrier-free nanomedicines generated by the self-assembly of small-molecule drugs have been proposed to effectively resolve these issues. Unfortunately, most chemotherapeutics cannot spontaneously self-assemble into nanoparticles (NPs). An approach relying on drug dimerization based on chemical linkers between small-molecule drugs has been employed to construct self-assembled nanoparticles. The strong intermolecular forces between homodimers, such as π-π stacking interactions, hydrogen bonding, and electrostatic forces, etc.¹⁵ drive the formation of nanoassembly. Given the active chemical groups in the GBA structure, we readily synthesized the gambogic acid dimers (GBA₂) based on an ester linkage via intermolecular interactions. Simple insertion of an ester linkage can transform the GBA dimers into a uniform nanosystem in water, with three advantages: high drug loading (~85.78%, w/w), high self-assembly stability, and low hemolysis when administered intravenously.

Both macrophages and osteoclasts are key players in the persistent progression of joint inflammation and destruction. Generally, the life spans of macrophages and osteoclasts are precisely regulated by apoptosis to maintain immune homeostasis and bone function balance, respectively.¹⁶ It was reported that targeted apoptosis of osteoclasts in arthritic joints is effective against inflammatory arthritis,¹⁶ hence, we investigated that the cell apoptosis of GBA₂/NPs assembled by GBA dimers against osteoclasts herein. In addition, regarding the advantages of the accumulation in inflamed joints and osteoclasts, as showed in Scheme 1, the carrier-free GBA₂/NPs were designed as a novel therapeutic approach for RA treatment.

Scheme 1 Schematic representation on the therapeutic benefits GBA2/NPs for rheumatoid arthritis treatment. (Created with BioRender.com).

Materials and Methods

Materials

Gambogic acid (purity >98%) was purchased from Chengdu Alfa Biotechnology Co. Ltd. (China). 1-ethyl-3-(3-dimethylaminopropyl) carbodiimide hydrochloride (EDC), 4-dimethylaminopyridine (DMAP), and dichloromethane (DCM) were obtained from Shanghai Aladdin Bio-Chem Technology Co., LTD (China). Near-infrared lipophilic carbocyanine dye 1, 10-dioctadecyltetramethyl indotricarbocyanine iodide (DiR) was purchased from Mellon Biological Technology Co., Ltd. (Dalian, China). Dimethyl sulfoxide (DMSO) and Lipopolysaccharide (LPS) were purchased from Sigma-Aldrich (St. Louis, MO, USA). Fetal bovine serum (FBS), phosphate-buffered saline (PBS), Hank’s balanced salt solution (HBSS), dulbecco’s modified Eagle’s medium (DMEM, Gibco), penicillin-streptomycin, and 0.25% (w/v) trypsin supplemented with 1 mM ethylene diamine tetraacetic acid (EDTA) were obtained from Invitrogen (Carlsbad, CA, USA). The cell counting kit-8 (CCK-8) assay kit and Annexin V-FITC/PI apoptosis kit were purchased from MultiSciences Lianke Biotech Co. Ltd. (Hangzhou, China). Commercial enzyme linked immunosorbent assays (ELISA) kits were obtained from Jingmei Biotech Co., Ltd. (Jiangsu, China). Anti-CD68 and Anti-CD51 antibodies were purchased from Abcam (Cambridge, UK).

The Cell Culture and Animal Studies

RAW 264.7 cells were obtained from ATCC (Manassas, VAUA) and were incubated in DMEM culture medium containing 10% FBS and 1% penicillin-streptomycin. The differentiation of RAW 264.7 cells into osteoclasts was induced by adding receptor activator of nuclear factor kappa-Β ligand (RANKL) (100 ng/mL) to the α-MEM culture medium for 6 days. Cells were cultured at 37 °C in a humidified atmosphere containing 5% CO₂. Male BALB/c mice (20 ± 2 g) and male Sprague Dawley rats (200 ± 20 g) were got from SiPeiFu (SPF, Beijing) biotechnology co., LTD (Beijing, China). All animals were allowed to acclimate for one week before the experiments and were provided with a fresh diet as well as free access to water. Ethic approval for animal experiments was obtained from the Chengdu University of Traditional Chinese Medicine (CDUTCM, permit CDU2022S156). Animal experiments were complied in accordance with the national standards of the People’s Republic of China (Laboratory animal: Guideline for ethical review of animal welfare [GB/T 35892–2018], Guideline of assessment for humane endpoints in animal experiment [RB/T 173–2018]). Animal welfare was ensured during whole animal experiments.

The Synthesis of Gambogic Acid Dimers

Gambogic acid dimers were synthesized by coupling two GBA molecules with EDC and DMAP as the catalysts (Figure 1A). GBA (1.0 eq), EDC (1.5 eq), and DMAP (1.5 eq) were dissolved in DCM. The reaction was performed in the dark at room temperature for 24 h, and the solvent was evaporated under reduced pressure on a rotary evaporator. The reaction mixture was purified using silica gel column chromatography to afford GBA₂ as a yellow solid. ¹H nuclear magnetic resonance (¹H NMR), ¹³C nuclear magnetic resonance (¹³C NMR) and MALDI-TOF/TOF-MS profiles confirmed the chemical structure of dimers. ¹H NMR and ¹³C NMR spectra of GBA₂ in CDCl₃-d₆ solution were recorded on a Bruker Ascend 600 spectrometer with ¹H resonance frequency of 600 MHz and ¹³C resonance frequency of 150 MHz, respectively. For MALDI-TOF-MS analysis, 1 μL of the sample was placed on MALDI target plates and immediately mixed with 1 μL 2, 5-dihydroxybenzoic acid in 50% acetonitrile/0.1% trifluoroacetic acid (TFA). Laser-desorbed positive and negative ions were analyzed after acceleration at 29 kV in linear mode on a Bruker ultrafleXtreme instrument (Bruker Daltonik, Bremen, Germany).

Figure 1 Identification of GBA2. (A) Synthesis route of GBA2. (B) 1H NMR spectrum of GBA2. (C) 13C NMR spectrum of GBA2. (D) MALDI-TOF/TOF-MS profile of GBA2 showing the expected molecular weight.

Preparation of Gambogic Acid Dimer Nanoparticles

Gambogic acid dimers were dissolved in a small amount of organic solvent and then added dropwise into deionized water under vigorous stirring to self-assemble the GBA₂/NPs using a one-step solvent evaporation method without using any surfactants or adjuvants. Briefly, 4 mg of GBA₂ was dissolved in 1 mL) acetone, which was then dropped into 10 mL of deionized water at 50 °C under continuous stirring. GBA₂ was successfully self-assembled within 2 h. The organic solvent was evaporated. Finally, the unloaded GBA₂ was removed a 0.45 μm filter (Millipore, USA). The obtained GBA₂/NPs were stored at 4°C.

Characterization of Gambogic Acid Dimer Nanoparticles

The size distribution, polydispersity index (PDI), and ζ-potential of GBA₂/NPs were determined by dynamic light scattering (DLS) using a LITESIZER 500 laser particle analyzer (Anton Paar, Austria). The morphology of the GBA₂/NPs was observed using transmission electron microscopy (TEM, Hitachi-7800, Japan) and scanning electron microscopy (SEM, Zeiss Sigma 300, Germany). The drug-loading efficiency (DLE) of the GBA₂/NPs was determined by high performance liquid chromatography (HPLC) using a standard curve method. Briefly, 1 mL GBA₂/NPs were mixed with methanol (1 mL). The mixture was then ultrasonicated for 30 min to extract GBA₂. After filtration through 0.22 μm filters, 10 μL of the suspension was injected into the HPLC system. The amount of GBA₂ was measured with a C18 reversed phase column (250 mm ×4.6 mm, 5 μm) at 30 °C at the maximum absorption wavelength of 360 nm. The mobile phase was a methanol–0.1% phosphoric acid–water solution (Table 1, v/v), and the flow rate was 1.0 mL/min.

Table 1 Information of Mobile Phase System

The storage stability of GBA₂/NPs at 4°C was evaluated by measuring the particle size and PDI using DLS during the storage period. Additionally, a serum stability test was performed. GBA₂/NPs were mixed with FBS at 1:1 (v/v) ratio, which was gently shaken at 37 °C. The supernatant of the mixture was collected at specific time points (0, 0.5, 1.0, 1.5, 2.0, 2.5, 3.0, 3.5, and 4.0 h), and the absorbance at 560 nm was measured using a microplate reader (SpectraMax iD5, USA).

In vitro hemolytic activity of GBA₂/NPs was evaluated. A 2% red blood cells (RBCs) suspension was obtained from rabbit heart blood as described in previous study.¹⁷ About 0.5 mL of the RBC suspension was incubated with 0.5 mL of free drug and nanoparticles at equivalent GBA concentrations (1.0, and 2.0 μg/mL) at 37 °C. After 4 h, the samples were centrifuged at 3000 rpm for 10 min. Absorbance (A) of the supernatant was measured using a microplate reader at 545 nm. RBC suspensions incubated with normal saline and distilled water were used as negative (0% lysis) and positive controls (100% lysis), respectively. The hemolytic rates of the samples were calculated using the following formula:

Where A_s represents the absorbance value of the samples and A_n and A_p are the negative and positive controls, respectively.

Molecular Dynamics Simulation

To gain further insight into the self-assembly mechanisms, the molecular dynamics simulations were conducted. CHARMM-GUI was used to construct the GBA₂ molecules. All MD simulations were performed using GROMACS software. The results are presented in VMD-1.9.2. The model system was composed of 25 GBA₂ molecules and 7000 water molecules. The side length of the water box was initially set to 10 nm. First, energy minimization was performed, and then a 1 ns simulation in the NPT ensemble. Finally, the size of the simulation box was 7*7*7 nm^3, followed by a 100 ns balance simulation (NPT,298 K, 1atm).

Biological Evaluation in vitro

The viability of GBA and GBA₂/NPs on Raw 264.7 cells was evaluated using the cell count kit-8 (CCK-8). RAW 264.7 cells were seeded in 96-well plates (8 × 10³ cells/well) for 24 h in advance. The cells were then cultured in 0.5% FBS medium containing free GBA and GBA₂/NPs at equivalent GBA concentrations (0–8 μM) for another 24 h. Based on kit instructions, 10 µL CCK-8 solution was added to measure the absorbance at 450 nm with a microplate reader (SpectraMax iD5, USA) to evaluate cell viability. Besides, we investigated the cellular uptake to observe the difference between GBA₂/NPs and free drug. RAW 264.7 cells were seeded at a density of 2.5 × 10⁵ cells/well in 6-well plates and incubated for 24 h to allow cell attachment. Following that, cells were treated with free GBA or GBA₂/NPs at equivalent GBA concentrations (100 μg/mL) for 4 h at 37°C. The cells were then washed thrice with cold Hank’s balanced salt solution (HBSS). Next, the cells were collected, freeze-thawed thrice, and ultrasonicated for 15 min. The lysate was extracted with methanol (0.5 mL), mixed by vibration, and centrifuged at 15,000 rpm for 5 min. Ten microliters of the resulting supernatant were used for the measurement of GBA₂ in the samples by HPLC analysis. Analytical conditions for gambogic acid dimers followed what we described above, while that of GBA was as follows: mobile phase, methanol−0.1% acetic acid water solution (93:7, v/v); detection wavelength, 360 nm; column, C18; column temperature, 25 °C; flow rate: 1.0 mL/min. In this study, the cytotoxicity and cellular uptake of GBA and GBA₂/NPs were also evaluated in osteoclasts. The inadequate apoptosis of osteoclasts in the RA joint contributes toward the persistent progression of joint inflammation and destruction.¹⁶ To investigate whether GBA₂/NPs could effectively induce the apoptosis of osteoclasts, the apoptotic profile was determined by flow cytometry assay.

Biodistribution of Gambogic Acid Dimer Nanoparticles in Adjuvant-Induced Arthritis Mice

To evaluate the biodistribution of GBA₂/NPs in vivo, adjuvant-induced arthritis (AIA) mice were intravenously injected with free DiR or DiR/GBA₂/NPs. BALB/c mice were injected with 50 μL Freund’s complete adjuvant (FCA, 1 mg/mL) into the left-hind paw to establish the AIA model. After 17 days, AIA mice were intravenously (i.v.) injected with free DiR or DiR/GBA₂/NPs containing 1 mg/kg DiR (3 mice/group). At selected time points post-injection (2, 4, 6, 8, 24, 30, and 48 h), the mice were anesthetized with isoflurane and observed using an in vivo imaging system (IVIS) (PerkinElmer, USA). At the end of the experiment, the mice were euthanized. The heart, liver, spleen, lungs, kidneys, and hind limbs were harvested for the ex vivo imaging of DiR fluorescence.

Anti-Rheumatoid Arthritis Efficacy of Gambogic Acid Dimer Nanoparticles in Adjuvant-Induced Arthritis Rats

The anti-RA effects of GBA₂/NPs were evaluated in AIA rats that were intracutaneously injected with 50 μL of FCA (10 mg/mL) into the left-hind footpads. After 17 days, the rats were randomly divided into three groups (n = 8), which were intravenously injected with saline, free GBA (4 mg/kg), or GBA₂/NPs (equivalent GBA dosage of 4 mg/kg) twice weekly. Healthy rats were treated with saline as a control group. The body weight, left-hind paw volume, and paw thickness were measured twice per week. Survival was monitored throughout the experimental period. On the 31st day, the rats were sacrificed and whole blood, major organs, and ankle joints from each group were collected. First, the ankle joints were fixed in 4% paraformaldehyde and scanned using ex vivo micro-computed tomography (micro-CT, PerkinElmer, 230 Inc. USA). The dataset was then reconstructed to obtain 3D images. Bone mineral density (BMD), trabecular number (Tb.N), trabecular bone thickness (Tb.Th), and trabecular separation (Tb.Sp) were quantitatively analyzed. Subsequently, the ankle joints were immersed in 15% tetrasodium ethylenediaminetetraacetic acid (EDTA) solution for 2 months for decalcification. The decalcified joints were embedded in paraffin and sectioned for H&E, safranin O, toluidine blue, tartrate-resistant acid phosphatase (TRAP), CD68, RANKL, TNF-α, and IL-1β staining. Additionally, the joints were sectioned for TUNEL, anti-CD68, and anti-CD51 costaining. Second, whole blood was prepared to determine red and white blood cell counts using an automated hematology analyzer (Tecom, China). Serum was extracted from the whole blood. Serum levels of TNF-α, IL-1β, RANKL, osteoprotegerin (OPG), total alanine aminotransferase (ALT), aspartate aminotransferase (AST), blood urea nitrogen (BUN), and creatinine (CRE) were measured using commercial ELISA kits. Finally, the main organs (heart, liver, spleen, lung, and kidney) were collected, fixed in 4% paraformaldehyde solution, embedded in paraffin, and stained with H&E and TUNEL.

Statistical Analysis

GraphPad Prism 8.0.1 (GraphPad Software, San Diego, USA) was employed for statistical analysis. All data were expressed as mean ± standard error of mean. One-way ANOVA with Tukey’s post-hoc test was performed for multi-group comparisons, and *P < 0.05 was considered statistically significant.

Results and Discussion

Identification of Gambogic Acid Dimers

Gambogic acid dimers were synthesized via esterification. ¹H NMR (Figure 1B), ¹³C NMR (Figure 1C), and MALDI-TOF/TOF-MS profiles (Figure 1D) indicated successful dimerization. Yield 58.8%. ¹H NMR (600MHz, CDCl₃-d₆): δ 12.77 (s, 1H), 7.55 (d, J = 6.9 Hz, 1H×2), 6.67 (d, J = 10.1 Hz, 1H×2), 6.64–6.60 (m, 1H×2), 5.44 (d, J = 10.1 Hz, 1H×2), 5.11 (t, J = 6.9 Hz, 1H×2), 5.06 (t, J = 6.9 Hz, 1H×2), 3.51 (dd, J = 7.0, 4.5 Hz, 1H×2), 3.25 (d, J = 7.1 Hz, 2H×2), 2.68–2.54 (m, 2H×2), 2.53 (d, J = 9.4 Hz, 1H×2), 2.34 (dd, J = 13.5, 4.8 Hz, 1H×2), 2.04 (q, J = 8.0 Hz, 2H×2), 1.82–1.74 (m, 2H×2), 1.72 (d, J = 8.4 Hz, 6H×2), 1.64 (d, J = 3.6 Hz, 6H×2), 1.55 (s, 3H×2), 1.40 (s, 3H×2), 1.38 (d, J = 5.0 Hz, 1H×2), 1.35 (s, 3H×2), 1.30 (s, 3H×2). ¹³C NMR (150MHz, CDCl₃-d₆): δ 203.15, 178.97, 171.42, 161.57, 157.76, 157.50, 137.07, 135.46, 133.51, 131.98, 131.92, 128.70, 124.93, 123.96, 122.32, 116.08, 108.07, 102.97, 100.55, 90.85, 83.89, 83.80, 81.48, 49.20, 47.02, 42.07, 30.09, 29.11, 27.65, 25.84, 25.78, 25.49, 22.91, 21.76, 18.24, 17.75, 11.56. MALDI-TOF/TOF-MS for GBA₂:1238, 1261.5846, [M + Na]⁺; 1237.5901, [M-H]⁻.

Characterization of Gambogic Acid Dimer Nanoparticles

According to dynamic light scattering (DLS) analysis, the hydrodynamic diameter and zeta potential of GBA₂/NPs was 178.21 ± 1.74 nm and −31.2 ± 0.3 mV, respectively (Figure 2A). It has been demonstrated that a diameter of 10 to 200 nm as well as a slightly negative charge could contribute to superior in vivo circulation time and inflamed joint targeting, known as extravasation through leaky vasculature and subsequent inflammatory cell-mediated sequestration (“ELVIS”) effect.¹⁸ The polydispersity index (PDI) was 0.13 ± 0.01, indicating a narrow size distribution. SEM and TEM were used to characterize the morphology of GBA₂/NPs. Both showed a uniformly spherical morphology with nanoscale sizes (Figure 2A). Notably, the drug loading efficiency of GBA₂/NPs was determined as 85.78 ± 3.67%. Taken together, the above results show that this carrier-free nanoplatform with a suitable size distribution, zeta potential, and ultra-high drug loading is suitable for GBA₂ delivery for RA therapy.

Figure 2 Characterization of GBA2/NPs. (A) Particle size distribution and zeta potential by DLS, and morphology image by TEM and SEM. (B) Storage stability of GBA2/NPs at 4 °C for 7 days. (C) The serum stability of GBA2/NPs during 4 h incubation with 10% FBS at 37 °C. (D) Hemolysis assay at different concentrations of GBA or GAB2/NPs. PBS and purified water were employed as negative and positive controls, respectively.

Subsequently, GBA₂/NPs were stored at 4 °C for one week. As shown in Figure 2B, the particle size and PDI values remained relatively constant, suggesting that GBA₂/NPs had excellent storage stability. Furthermore, no noticeable changes were observed after incubation in PBS containing 10% FBS at 37 °C for 4h (Figure 2C), indicating good serum stability of GBA₂/NPs. However, GBA is known to exhibit significant hemolytic toxicity.¹⁹ To ensure the biosafety of GBA₂/NPs after intravenous injection, a hemolysis analysis was conducted. 2.0 μg/mL GBA exhibited obvious red blood cell membrane damage comparable to that of the positive control, whereas GBA₂/NPs at the equivalent GBA concentration showed no hemolysis (Figure 2D). As expected, the hemolytic toxicity of GBA was significantly reduced by its dimerization and self-assembly. These findings indicate its potential application in vitro/vivo studies.

Molecular Dynamics Simulation of Gambogic Acid Dimer Nanoparticles

Configuration 3D diagram showed that GBA₂ self-assembled into a stable nanostructure for approximately 20 ns (Figure 3A), indicating the good self-assembly ability of the dimers. The sharp change in the solvent-accessible surface area (SASA) values also corroborated the increase in the compactness of the nanoclusters from 0 to 20 ns, which can reduce the surface area of the dimers exposed to water (Figure 3B). Analysis of the molecular interaction patterns further illustrated that hydrogen bonds (red dotted line) were the main driving force, as shown in (Figure 3C). Further calculations over a simulation time of 100 ns revealed that the number of GBA₂−water hydrogen bonds decreased, particularly from 0 ns to 20 ns, probably because of the increased aggregation of GBA₂ which kept the hydrogen bond acceptor and donor of GBA₂−water away and prevented hydrogen bond interactions (Figure 3D). According to thermodynamic laws, negative values of the free binding energy indicate favorable stability of the system.²⁰ As shown in Figure 3E, the absolute value of the potential energy between GBA₂ and water at 20 ns is smaller than that at 0 ns, indicating an unstable structure between them. GBA₂ preferentially forms intramolecular hydrogen bonds with water molecules. These findings demonstrate that the amphiphilic properties of GBA₂ contribute to the self-assembly process via hydrogen bonding and hydrophobic interactions.

Figure 3 Molecular dynamics simulation of GBA2/NPs. (A) Instantaneous simulation snapshots of the equilibrated system (GBA2 dispersed in water, upper panel; water was not shown for clarity, lower panel). (B) SASA calculation of the aggregates. (C) Hydrogen bond interaction of dimer nanoassemblies. (D) Distribution of the number of different types of hydrogen bonds during simulation time. (E) The binding energy between the GBA2 and water.

Biological Evaluation of Gambogic Acid Dimer Nanoparticles in vitro

Cytotoxicity of GBA₂/NPs on Raw 264.7 cells was assessed after 24 hours incubation. The IC₅₀ values of GBA and GBA₂/NPs were 1.476 μM and 2.531 μM, respectively. Compared to GBA, GBA₂/NPs showed lower dose-dependent cytotoxicity, which might be attributed to the enhanced cell uptake efficiency of the dimeric GBA nanoassemblies (Figure 4A). High intracellular uptake efficiency is necessary for therapeutic efficacy. The in vitro cellular uptake of GBA and GBA₂/NPs at equivalent concentrations was examined on Raw 264.7 cells by HPLC. As illustrated in Figure 4B, the cellular uptake of GBA₂/NPs was approximately 2-fold higher than that of GBA, possibly because of the nanoscale size. We presumed that the targeting ability of GBA₂/NPs contributes to their biosafety. Besides, the similar results of cell viability and cellular uptake of GBA₂/NPs were shown in osteoclasts (Figures S1 and S2). Inducing the apoptosis of osteoclasts in RA joints is a promising strategy for RA therapy. In this study, the results showed that GBA₂/NPs exhibit higher apoptosis rate in osteoclasts (Figure S3), which might effectively alleviate the process of joint inflammation.

Figure 4 Cytotoxicity and cell uptake profiles of Raw264.7 cells in vitro. (A) Cytotoxicity of GBA2/NPs or GBA against Raw264.7 cells with 24 h treatment. (B) Cell uptake of GBA or GBA2/NPs by Raw264.7 cells at 100 μg/mL for 2 h treatment. All doses were calculated as the free GBA. *p < 0.05 indicated the significant difference.

In vivo Biodistribution of Nanoparticles

The fluorescent signal in the inflamed paws of AIA mice treated with DiR/GBA₂/NPs was markedly stronger than that in the right non-inflamed paws, suggesting that DiR/GBA₂/NPs selectively accumulated in inflamed lesions. Notably, intense fluorescence was observed in inflamed paws as early as 2 h after injection of DiR/GBA₂/NPs, and the signal persisted for up to 48 h (Figures 5 and S4). In contrast, the fluorescence signal was barely detectable in the major tissues of the free DiR-treated AIA mice. Our data clearly demonstrated the rapid accumulation and prolonged circulation of GBA₂/NPs.

Figure 5 In vivo biodistribution of nanoparticles. Ex vivo fluorescence images showing the biodistribution of DiR and DiR/GBA2/NPs in the mice with a unilateral inflamed joint at predetermined time intervals, ie 2, 4, 6, 8, 24, 30, and 48 h after injection.

Therapeutic Effects of Gambogic Acid Dimer Nanoparticles in Adjuvant-Induced Arthritis Rats

Subsequently, a rat AIA model was established to investigate the therapeutic effects of GBA₂/NPs in vivo (Figure S5A). Saline, GBA, or GBA₂/NPs were intravenously injected into rats with AIA (4 mg/kg equivalent of GBA). Untreated healthy rats were used as controls. Compared with healthy rats, the thickness of the left-hind paw (Figure S5C) and volume (Figure S5B) were highest in rats treated with saline, followed by those treated with GBA or GBA₂/NPs.

Notably, GBA₂/NPs showed therapeutic effects comparable to those of GBA alone. Micro-computed tomography (micro-CT) and histological assays were used to further investigate joint destruction and synovial inflammation in different groups. Consistent with the above results, GBA and GBA₂/NPs showed similar efficacy, negligible swollen paws and ankle joints (Figure 6A), relatively smooth bone surfaces, and minor bone erosion (red arrows) (Figure 6B), together with high bone mineral density (BMD) (Figure 6F), which was close to that of healthy rats. The trabecular parameters also confirmed that GBA and GBA₂/NPs administration were equally effective in increasing the trabecular number (Tb.N) (Figure 6G) and trabecular bone thickness (Tb.Th) (Figure 6H) but decreased trabecular separation (Tb.Sp) (Figure 6I). Additionally, H&E staining revealed that both GBA and GBA₂/NPs reduced inflammatory cell infiltration (yellow arrows) and synovial hyperplasia (black arrows) (Figure 6C). Safranin-O and toluidine blue staining were used to label glycosaminoglycans, thus examining the cartilage integrity of the joint ankles.^21,22 Areas positive for safranin-O and toluidine blue staining of GBA and GBA₂/NPs were similar to those of the healthy group (Figure 6D and E), indicating that either GBA or GBA₂/NPs could recover cartilage destruction. However, our attention was drawn to the systemic toxicity of GBA, which was shown by sharp changes in body weight (Figure S5D) and low survival rate (Figure S5E) in the rats treated with GBA. In contrast, GBA₂/NPs showed good biocompatibility. Taken together, no difference in the anti-arthritic effects between GBA and GBA₂/NPs was found, whereas the toxic effects of GBA were remarkably attenuated by the self-assembly of dimers.

Figure 6 The administration of GBA2/NPs influenced histopathology of AIA rats. (A) Representative photographs of the hindlimbs at the endpoint of the experiment from different treatment groups. Histopathology evaluation of the ankle joints was identified using (B) micro-CT, (C) H&E, (D) safranin-O, and (E) toluidine blue staining. Quantitative micro-CT analysis of (F) BMD, (G) Tb.N, (H) Tb.Th, and (I) Tb.Sp of the ankle joints. Red arrows, bone erosion; Yellow arrows, inflammatory cell infiltration; Black arrows, synovial hyperplasia. *p < 0.05 indicated the significant difference. Abbreviations: CT, Computed tomography; H&E, Hematoxylin-eosin; BMD, Bone mineral density; Tb.N, Trabecular number; Tb.Th, Trabecular bone thickness; Tb.Sp, Trabecular separation.

Pro-Apoptotic Effects of Gambogic Acid Dimer Nanoparticles

Next, to investigate whether GBA₂/NPs could target osteoclasts and macrophages in arthritic joints and simultaneously induce apoptosis in both cell types, the level of cellular apoptosis and number of osteoclasts and macrophages in the inflamed joints were determined by TUNEL staining and immunofluorescence analysis, respectively. As shown in Figure 7A and B, GBA₂/NPs triggered a higher level of apoptosis in inflamed joints than free GBA. Furthermore, the apoptotic cells induced by GBA₂/NPs were mainly macrophages and osteoclasts, as evidenced by the overlap of green and red fluorescence. However, the GBA treatment resulted in more severe off-target apoptosis in the major organs (Figure S6). Together, these results indicate that GBA₂/NPs were preferentially enriched in inflamed joints, selectively targeted macrophages and osteoclasts, and efficiently induced apoptosis, thereby ameliorating adjuvant-induced arthritis without systemic toxicity. The reduction in osteoclasts and macrophages in inflamed lesions further confirmed the targeting activity of GBA₂/NPs. Immunohistochemical assays for CD68 expression showed that GBA₂/NPs remarkably decreased the number of macrophages (Figure 7C). Similar results can be seen in TRAP staining of osteoclasts (red area) (Figure 7D). Macrophages are the main producers of pro-inflammatory cytokines in RA, such as interleukin-1 beta (IL-1β) and tumor necrosis factor-alpha (TNF-α).²³ We observed that the GBA₂/NPs group had a lower secretion of IL-1β and TNF-α in the serum than the GBA group, which was similar to that of the healthy group (Figure 7H and I). The immunohistochemical analysis revealed that GBA₂/NPs potently decreased the expression of IL-1β and TNF-α in the ankle joint (Figure 7E and F). Furthermore, a balance between RANKL and osteoprotegerin (OPG) is essential for bone remodeling. Decreased RANKL levels and a lower RANKL/OPG ratio are helpful for RA remission.^24,25 Our study found that GBA₂/NPs exhibited a recovery trend of bone function, as confirmed by the lower expression of RANKL in arthritic joints (Figure 7G), together with a smaller RANKL/OPG ratio (Figure 7J) in the serum than in the GBA group. In brief, GBA₂/NPs targeting macrophages and osteoclasts promoted apoptosis, thereby reducing inflammation and bone destruction in arthritic joints.

Figure 7 Pro-apoptotic effects of GBA2/NPs in macrophages and osteoclasts in AIA rats. (A and B) Detection of apoptotic cells in ankle joints from AIA rats by TUNEL assay. Determination of macrophages and osteoclasts in ankle joints from AIA rats by immunofluorescence analysis of (A) CD68 and (B) CD51, respectively. Immunohistochemical analyses of (C) the CD68-stained synovial macrophages and (D) TRAP-stained osteoclasts in the joint tissues from rats subjected to the indicated treatment. Detection of (E) IL-1β, (F) TNF-α, and (G) RANKL expression levels in arthritic joints. The serum level of (H) IL-1β, (I) TNF-α, and (J) RANKL/OPG. *p < 0.05 indicated the significant difference. Abbreviations: TRAP, Tartrate resistant acid phosphatase; IL-1β, interleukin-1 β; TNF-α, tumor necrosis factor-α; RANKL, Receptor of activator of NF-kB ligand; OPG, Osteoprotegerin.

Considering the severe toxicity of GBA, we further evaluated the biosafety of GBA₂/NPs through histopathological analysis of major organs as well as hematological and serum biochemical examinations. As shown in Figure 8A, GBA administration resulted in histopathological abnormalities in major organs, mainly demonstrated by structural destruction (yellow arrows), necrosis (black arrows), and inflammatory cell infiltration (red arrows). ALT and aspartate aminotransferase (AST) were sensitive indicators of hepatocyte injury, whereas blood urea nitrogen (BUN) and creatinine (CRE) were those of nephrotoxicity.^26,27 The GBA group showed significant increases in BUN (Figure 8D), CRE (Figure 8E), AST (Figure 8F), and ALT (Figure 8G) levels, which indicated kidney and liver damage. White blood cells (WBC) and red blood cells (RBC) are closely associated with infection and hematological toxicity, respectively.²⁸ Similarly, abnormally high WBC (Figure 8B) and lower RBC counts (Figure 8C) suggest poor hemocompatibility of GBA. These adverse effects of GBA were largely abolished by the nanoassembly of GBA dimers. The structure of the cells remained intact and the nuclei were arranged in an orderly manner without necrosis or degeneration. WBC, RBC, BUN, CRE, AST, and ALT levels were similar to those in the healthy rats. Collectively, these results implied that GBA₂/NPs are the safe platform for RA therapy.

Figure 8 Reduced systemic toxicity of GBA2/NPs in AIA rats. (A) H&E staining of major organs (heart, liver, spleen, lung, and kidney). The amount of (B) WBC, (C) RBC, in whole blood and the serum level of (D) BUN, (E) CRE, (F) AST, and (G) ALT in rats receiving the indicated treatment. Yellow arrows, structural destruction; Black arrows, structural necrosis; Red arrows, inflammatory cell infiltration. *p < 0.05 indicated the significant difference. Abbreviations: WBC, White blood cells; RBC, Red blood cells; BUN, Blood urea nitrogen; CRE, Creatinine; AST, Aspartate aminotransferase; ALT, Alanine aminotransferase.

Conclusion

In conclusion, GBA dimers were successfully synthesized by coupling two GBA molecules via simple insertion of an ester linkage, and conjugates could subsequently self-assemble into GBA₂/NPs driven by hydrogen bonding and hydrophobic interactions. The homodimers could self-assemble into GBA₂/NPs with extraordinarily high drug loading capacity via the facile one-step nanoprecipitation. Mediated by the “ELVIS” effect, GBA₂/NPs showed high sensitivity to inflamed joints, enabling the targeted apoptosis of macrophages and osteoclasts. The anti-arthritic efficacy of GBA₂/NPs, including inflammatory remission and bone erosion repair, was comparable to that of GBA, without systemic toxic effects. The nanoassembly of homodimers is pretty eye-catching. This study highlights the potential of nanomedicines fabricated via dimerization and self-assembly for RA therapy, which might provide new insights into the design of nano-drug delivery system with high drug loading for the treatment.