Folate receptor-mediated targeted therapy for rheumatoid arthritis by methotrexate-phospholipid complex nano-emulsions

Abstract Rheumatoid arthritis (RA) is a common autoimmune and inflammatory disease. Activated macrophages in arthritic joints play a prominent role in the initiation and persistence of RA. Despite great progress in the clinical treatment of RA, poor response and high discontinuation due to systemic toxicity remain unsolved issues, especially the well-known methotrexate (MTX). Therefore, active targeted delivery of therapeutic drugs to pathogenic cells in arthritic joints is essential to increase in situ activity and decrease systemic toxicity. Here, we developed an MTX-loaded macrophage-targeted nano-emulsion (NE) based on the overexpression of folate receptor (FR) on activated macrophages, the inherent high affinity of FR for folate (FA), as well as the property of MTX and phospholipids to form complexes (MTX@PC). Intravenous injection of DID-labelled MTX@PC-FA NEs into adjuvant-induced arthritis (AIA) mice, in vivo images and flow cytometry results revealed that the NEs were highly targeted to inflamed joints and macrophages, respectively. Therapeutic studies suggested that this strategy was conducive to achieve high efficacy and low toxicity of MTX in the treatment of RA. Our research highlights MTX@PC-FA NEs as a potential treatment option for RA targeting the FR-expressed activated macrophages.


Introduction
Rheumatoid arthritis (RA) is a degenerative, autoimmune disease that affects the hands and feet and is characterised by widespread, symmetrical and invasive joint inflammation [1]. Erythema, swelling, pain and stiffness of the joints are the prominent symptoms, which can result in progressive articular damage and lifelong impairment [1,2]. According to epidemiological statistics, the prevalence of RA ranges from 0.5% to 1% worldwide [3], and the average yearly treatment cost for patients with difficult-to-treat RA can reach 37,605 euros [4]. As a result, this illness has a significant impact on patients' quality of life and places a significant social and financial burden on society.
The classic disease-modifying antirheumatic drug (DMARD) methotrexate (MTX) is regarded as the gold standard for treating RA [5]. It is currently a first-line medication for RA, either alone or in combination with biological or conventional DMARDs. However, taking MTX orally for a prolonged period of time might have substantial side effects, such as bone marrow suppression, hepatotoxicity, nephrotoxicity, etc. forcing 10%-37% of patients to cease using the medication [6]. In addition, the wide distribution in vivo and inadequate concentration in the target organs brought on by systematic administration may be one of the potential causes for the low response rate of patients to MTX, which leads to about 30%-50% of patients need to adopt the co-administration of additional medications to achieve the therapeutic objective [7]. Therefore, it is urgent to improve the inflammatory targeting and reduce the incidence of adverse reactions of MTX.
Nano-delivery technology can endow drug targeting ability, improve drug concentration in diseased tissues, reduce side effects and increase bioavailability [8,9]. These benefits make MTX based on nanotechnology attractive in treating RA with great efficacy and few side effects. However, it is difficult to construct stable and secure nanocarriers for MTX delivery due to the drug's weak acidity and insolubility in water, oil and most organic solvents.
Macrophages are one of the most significant immune cells in the occurrence and progression of RA, and the abundance and activation of macrophages in inflamed synovial membrane are directly connected with the severity of RA [10]. Activated macrophages secrete inflammatory cytokines including tumor necrosis factor-alpha (TNF-α), interleukin-1beta (IL-1β), interleukin-6 (IL-6) and interact with other immune cells to maintain the inflammatory response and continuously damage the joints [11,12]. Therefore, RA therapy that targets macrophages has become a research hotspot [13,14]. Folate receptor (FR) is a specific class of cell-surface binding receptor glycoprotein. In the RA synovium, folate receptor beta (FR-β), a significant member of FR family, is overexpressed on the surface of activated macrophages, and has a strong affinity for folate (FA) [15,16]. However, while a variety of nanomedicines that target FR-β have shown good efficacy in the treatment of RA, the possible immunogenicity of the carriers and the complex production procedure have limited the relevant clinical advancement [17][18][19][20].
In this work, we constructed a nano-emulsion (NE) for the targeted delivery of MTX to macrophages in inflamed joints, to effectively combat RA inflammation ( Figure 1). On the basis of the property that phospholipid and compounds containing reactive hydrogen might produce drug-phospholipid complex (drug@ PC) [21][22][23], we prepared MTX-phospholipid complex (MTX@PC) as emulsifier. Owing to the better stability of NEs compared with liposomes in the suspension state, the oil phase was used as the core to increase the stability. Moreover, the preparation was endowed with a long-circulating and targeting unit, DSPE-mPEG 2000 -FA, to actively target macrophages and extend the time of systemic circulation. The chemical synthesis and NEs production processes are established and straightforward, and the excipients used in this delivery system are all pharmaceutical excipients that have received FDA approval. We then performed systematic preparation characterisation, targeting verification in vitro and in vivo, and in vivo efficacy and safety assessment. The findings showed that this approach has excellent anti-RA inflammatory effects, good biocompatibility and clinical translation potential.

Fabrication and characterization of MTX@PC-FA NEs
To synthesise MTX@PC, 40 mg soybean lecithin and 10 mg MTX powder were added into a 50-mL round-bottom flask containing 10 mL THF. The mixture was mixed, stirred and reacted in a water bath at 37 °C for 12 h. The organic solvent was removed via a rotary evaporator, and the products were collected and stored in darkness. Subsequently, IR, uV and 1 H-NMR were used to identify the products.
MTX@PC NEs and MTX@PC Liposomes were prepared by thin-film hydration. Briefly, 10 mg MTX@PC was placed in a 50-mL round-bottom flask and then dissolved completely in 5 mL of chloroform, followed by removal of the organic solvent via a rotary evaporator in a water bath at 37 °C. Then, 10 mL of sterile water injection or suspended aqueous solution containing soybean oil and MCT (0.25%, vol/vol) were added into above flask, and original liposomes or NEs were obtained by sonication. Finally, initial MTX@ PC Liposomes or MTX@PC NEs were prepared by a high-pressure homogeniser (AH-NANO, ATS, Canada). About 2 mL of initial MTX@PC NEs were well mixed with 10 mg of DSPE-mPEG 2000 -FA or DSPE-mPEG 2000 to prepare MTX@PC-FA NEs or MTX@PC NEs. Hydrodynamic diameter and zeta potential of NEs were determined at 25 °C by dynamic light scattering on a Malvern Zetasizer Nano-ZS (Nano ZS90, Malevern, uK). Particle morphology was further investigated via a transmission electron microscopy (H-600, Hitachi, Japan) after being stained with phospho-tungstic acid aqueous solution (2%, wt/vol). High-performance liquid chromatography (HPLC) was used to quantify encapsulation efficiency of MTX. MTX-loaded NEs were separated from Free MTX through sephadex G-75, and MTX@ PC-FA NEs solution was demulsificated in 1% Triton-THF-C 2 H 5 OH (2:5:5, vol/vol/vol) and sonicated for 10 min to extract MTX completely. Moreover, the chromatographic conditions of HPLC are as follows: Column: C 18 (5 μm, 250 × 4.6 mm, Agilent, uSA); Detector: uV 302 nm; Flow rate: 1.0 mL/min; Injection size: 10 μL; Mobile phase: Acetonitrile -7% citrate solution -2% anhydrous disodium hydrogen phosphate solution (8.5:10:80) (the pH was adjusted to 6.0 by 7% citrate solution or 2% anhydrous disodium hydrogen phosphate solution).

Cellular uptake and cytotoxicity in vitro
RAW264.7 cells were cultured in RPMI-1640 medium and supplemented with 10% foetal bovine serum, 100 u/mL penicillin and 100 mg/mL streptomycin at 37 °C in a humidified environment containing 5% CO 2 . RAW264.7 cells used in the research were first activated with 10.0 ng/mL LPS for 24 h, and then replaced with fresh medium and planted in culture plates in accordance with the prescribed number. RAW264.7 cells were planted in a 6-well plate with a density of 1 × 10 6 cells per well for 12 h. Cells were exposed to medium containing CY5-labelled MTX@PC-FA NEs, MTX@PC NEs or DSPE-mPEG 2000 for 1 h. Cells were collected, washed with PBS after incubation and then resuspended in PBS for flow cytometric analysis (FACS Celeata, BD Biosciences, uSA). Moreover, cellular uptake on RAW264.7 cells was also investigated via a laser scanning confocal microscopy (LSM 800, Zeiss, Germany) after being stained by DAPI. RAW264.7 cells were seeded in a 96-well plate with a density of 1 × 10 4 cells per well for 12 h. Different solutions were diluted by medium to prepare the MTX concentration of 4, 8, 12, 20, 32 and 40 µg/mL before being added 200 mL to each well. CCK-8 reagent was added to each well after 48 h. After 2 h of incubation, the absorbance was assessed at 450 nm with a microtiter-plate spectrophotometer (Spark 10 M, Tecan, Swiss). Cell viability was calculated by the following equation: cell viability

In vivo FR-mediated delivery of MTX@PC-FA NEs to inflamed joints
BALB/c mice were provided by Dashuo Biotechnology Company (Chengdu, China). Adjuvant-induced arthritis (AIA) mouse model was used in our study. Briefly, male BALB/c mice aged 6-8 weeks were subcutaneously injected with 0.1 mL of complete Freund's adjuvant (CFA) (7008, Chondrex, uSA), which contains 1 mg/mL of heatkilled Mycobacterium tuberculosis. There was severe, acute inflammation within 30 min of the injection, which peaked 14-16 days later and frequently continued for 30 days. The equal volume of Free DID in 1,2-propanediol aqueous solution (30%, vol/vol), DID-labelled MTX@PC-FA NEs or MTX@PC NEs were injected intravenously into AIA mice via the tail vein, with dosing 1 μg/mouse. Subsequently, the in vivo fluorescence imaging system (IVIS) was used to capture the biodistribution at 12 h after administration, and the region of interest (ROI) fluorescence analysis of joints was carried out by a living image program and displayed as mean intensity.
To prepare single cell suspension of inflamed joints, the mice were sacrificed at 13 h following injection, and the paws in each group were harvested and divided into pieces. About 1 mL collagenase type IV (1 mg/mL) was added into the above solution, followed by mixing evenly and incubating for 2 h at 37 °C on a shaker. The supernatant was discarded after centrifugation at 3000 g, and the cells were washed with 1 mL of PBS twice. Subsequently, cell surface molecular staining was used to label macrophages and flow cytometry was used to detect the frequency of DID + macrophages.

In Vivo therapeutic experiment
After 3 days of AIA induction, the animals were randomly divided into 4 groups (6 animals in each group). A 100 μL of sterile PBS (Control), MTX in PBS (5 mg/kg, MTX per body weight), MTX@PC NEs (5 mg/kg, MTX per body weight), MTX@PC-FA NEs (5 mg/kg, MTX per body weight) was administered through tail vein every 3 days for a total of six doses. Mouse paw thicknesses were measured with a digital calliper every 3 days. Clinical score was evaluated every 3 days by referring to the following standards (score: 0-4): 0, normal; 1, slight redness around the ankle or tarsal joints; 2, slight redness and swelling spreading to the tarsals; 3, significant redness and swelling reaching to the metatarsal joints; and 4, severe redness and edoema covering the ankle and foot. On day 21, pictures of paws were obtained. On day 22, the mice were sacrificed, and the blood, organs and paws were collected to study the relevant indications. To conduct the histopathological study, the collected paws were first fixed for 2 days in paraformaldehyde aqueous solution (4%, vol/vol), and then decalcified for 6 weeks in PBS (PH 7.4) with 15% EDTA. After decalcification, staining and image collection will be performed according to standard operating procedures (SOP). Furthermore, mouse serum samples were obtained at the end of the therapy experiment, and the concentration of cytokines such as TNF-α, IL-1β, IL-6 were measured by enzyme-linked immunosorbent assay (ELISA). In particular, mice's orbital entire blood was taken and allowed to coagulate at 4 °C for 4 h. To separate the serum from the supernatant, samples were centrifuged at 6000 g for 10 min. Subsequently, samples should be stored at -20 °C until use. Finally, multiple mouse ELISA kits were used to detect the concentration of various cytokines according to SOP.

Systemic toxicity evaluation
To conduct routine blood test, blood was collected in a blood vessel containing heparin and tested immediately with a haematology analyser (BC-2800Vet, Mindray, China). In addition, the blood sample without heparin was placed at 4 °C overnight and centrifuged at 1000 g for 20 min, and the supernatant was collected to analyse ALT according to the instruction in the kit.

Statistical analysis
All data were presented as mean ± standard deviation (SD), and the two-tailed unpaired Student's t test was used to determine the statistical significance of the differences between the groups with a level of *p < 0.05, **p < 0.01 and ***p < 0.001.

Fabrication and characterisation of MTX@PC-FA NEs
First, MTX@PC was synthesised in the nonprotic solvent tetrahydrofuran (THF). Compared with 0 h, the reaction solution after 12 h was transparent (Figure 2(A)), which might be due to the carboxyl of MTX formed MTX-phospholipid complex (MTX-PC) through hydrogen bonding with the carbonyl group on the phospholipid, thus hiding the polar group of MTX and dissolving in THF in complex form. The 1 H NMR, uV-vis and IR of the product all confirmed the successful synthesis of MTX@PC (Supplementary Figure S1(A,B), Figure 2(B)). Next, MTX@PC liposomes or MTX@PC nano-emulsions (NEs) were prepared by using MTX@PC as the emulsifier and sterile water injection or soybean oil and MCT as the core, respectively. The results showed that the NEs were much more stable than liposomes after 7 days of exposure to darkness at 4 °C, and MTX sedimentation in liposomes was visible ( Figure  2(C)), indicating that the oil-phase core was conducive to stability MTX@PC. Then, NEs was modified with DSPE-mPEG 2000 -FA to produce MTX@PC-FA NEs. The hydrodynamic size and potential were 118 nm and -13.5 mV, respectively (Figure 2(D)), and TEM revealed relatively uniform round-like particles (Figure 2(E)). In the stability experiment, the particle size of MTX@PC-FA NEs remained essentially unaltered after 7 days in the dark at 4 °C, showing good stabilisation (Figure 2(F)). Subsequently, the encapsulation efficiency of MTX@PC-FA NEs measured by HPLC was 84.7%.

Cellular uptake cytotoxicity in vitro
To examine the cellular uptake of the preparation on RAW264.7 cells, DSPE-mPEG 2000 and DSPE-mPEG 2000 -FA were tagged with fluorescent dye CY5. According to laser scanning confocal microscopy, the MTX@PC-FA NEs group had the best absorption efficiency (Figure 3(A)). In addition, outcomes from flow cytometry revealed that MTX@PC-FA NEs had an uptake efficiency of 43%, which was much greater than that of MTX@PC NEs (33%) and Free DSPE-mPEG 2000 (Figure 3(B)), demonstrating that FA modification significantly enhanced the macrophages targeting of NEs. DSPE-mPEG 2000 -FA also exhibited a significant difference compared with DSPE-mPEG 2000 , indicating the binding specificity between FA and FR (Supplementary Figure S2).
The effects of MTX, MTX@PC NEs and MTX@PC-FA NEs on cell viability in vitro were investigated. The results showed that the cytotoxicity of the Free MTX, MTX@PC NEs and MTX@PC-FA NEs increased proportionally to an increase in MTX concentration, with the preparation group exhibiting a more pronounced increase (Figure 3(C)). Moreover, the inhibitory efficiency of MTX@ PC-FA NEs was higher than that of MTX@PC NEs when the drug concentration was small, which may be related to the high uptake efficiency of MTX@PC-FA NEs mediated by FR receptors on RAW264.7 cells. As the concentration increased, receptor saturation resulted in comparable cytotoxicity of MTX@PC NEs and MTX@PC-FA NEs.

In vivo FR-Mediated delivery of MTX@PC-FA NEs to inflamed joints
NEs labelled with lipophilic fluorescent dye DID were employed to investigate the arthritic joint-targeting efficiency of the preparation. In vivo fluorescence imaging system (IVIS) showed that the fluorescence intensity of inflammatory joints in the MTX@PC-FA NEs group was significantly higher than that in the MTX@PC NEs and Free DID groups 12 h after intravenous administration (Figure 4(A)), and the two preparation groups are more potent than free group. The result was further confirmed by the semi-quantitative fluorescence efficiency values (Figure 4(B)). To further investigate the effect of NEs on intra-articular macrophages, the inflamed paws of Figure 3. Characterisation of MTX@PC-fa nes. a: Cellular uptake of CY5-labelled MTX@PC-fa nes in raW264.7 cells after incubation for 1 h, photos were taken by a laser confocal scanning microscope. Scale bar: 10 µm. B: uptake efficiency of CY5-labelled MTX@PC-fa nes in raW264.7 cells after incubation for 1 h, results were analysed by a flow cytometry. Data are presented as the mean ± SD, n = 3. *p < 0.05, **p < 0.01. C: Cytotoxicity of formulations with MTX at various concentrations. raW264.7 cells were treated with free MTX, MTX@PC nes and MTX@PC-fa nes for 48 h, and the inhibition rate was determined by CCK-8 assay.
AIA mice were harvested at 13 h and prepared into single-cell suspension. Macrophages were labelled with fluorescent antibody F4/80, and the proportion of DID + macrophages was detected by flow cytometry. The findings demonstrated that the ratio of DID + F4/80 + cells was approximately twice as high in the MTX@ PC-FA NEs group as it was in the MTX@PC NEs group (Figure 4(C)).
The aforementioned findings suggest that FA-modified MTX@ PC NEs can significantly enhance the inflammatory joint and cell targeting ability of the NEs via the FR over-expressed on activated macrophages, which will contribute to exert the therapeutic benefit of MTX and lessen the side effects.

In vivo therapeutic efficacy of MTX@PC-FA NEs
AIA mouse model has recently been widely used in the development of innovative therapies for RA due to its affordability and applicability [24][25][26]. The AIA mice were randomly divided into four groups on the 3rd day after the model establishment, with six animals in each group. The Control group was injected with PBS through the tail vein, and the other three groups were sequentially treated once every 3 days for a total of six times with Free MTX, MTX@PC NEs and MTX@PC-FA NEs. Each group received 5 mg/kg of MTX, while healthy mice served as the positive control ( Figure 5(A)). Importantly, paw thickness and clinical score was measured every 3 days during treatment.
The results showed that the Control group had the highest paw thickness and clinical score ( Figure 5(B,C)), as well as the most severe inflammation. The inflammation in each treatment group was suppressed to varying degrees, the effects of the two preparation groups were significantly superior to those of Free MTX, and the MTX@PC-FA NEs had the most prominent therapeutic effect. Photographs of paws at the end of the treatment more intuitively confirmed the aforementioned results ( Figure 5(D)).
Haematoxylin and eosin (H&E) staining of the ankle joint further demonstrated that MTX@PC-FA NEs therapy effectively reduced synovial inflammation, and the clarity and smoothness of the articular cavity were even comparable to that of healthy animals. The Control group, however, had the most severe joint injury compared to the other groups ( Figure 5(E)). In addition, the red cartilage content was also significantly larger in the MTX@PC-FA NEs group than in the Control, Free MTX and MTX@PC NEs groups according to safranin O-fast green pictures ( Figure 5(F)), with a continuous distribution similar to that in the healthy group.
Therapeutic experiment showed that MTX@PC-FA NEs with macrophage-targeting capabilities were demonstrated to be more effective at combating RA inflammation and cartilage degradation.

Evaluation of the therapeutic efficacy and toxicity of MTX@ PC-FA NEs
To further understand the mechanism of therapeutic effects, ELISA was used to measure the serum concentrations of various cytokines, including pro-inflammatory factor TNF-α, IL-1β, IL-6 and anti-inflammatory factor IL-10. The results demonstrated that the levels of pro-inflammatory factors in the Control group were the highest, and the levels in Free MTX group remained high, while the MTX@PC NEs and MTX@PC-FA NEs treatment effectively alleviated this phenomenon, especially in the MTX@PC-FA NEs group ( Figure 6(A-C)). However, the results of the anti-inflammatory factor IL-10 were just the opposite, with a significant increase in IL-10 concentration after MTX@PC-FA NEs treatment (Figure 6(D)). These experiments revealed that the alleviation of RA inflammation was compatible with the changes of cytokines in vivo. Targeted therapy reversed and rebalanced the secretion level of cytokines, and finally effectively curbed the development of RA inflammation.
The most serious side effect of methotrexate is bone marrow suppression, as well as liver damage. Therefore, we monitored blood routine and liver function at the end point of treatment. The outcomes demonstrated that MTX@PC-FA NEs treatment had no significant effect on white blood cell count (WBC), platelets (PLT) and alanine aminotransferase (ALT) in mice (Supplementary Figure S3(A-C)), and the values of MTX@PC-FA NEs group were all within the normal range. The Free MTX group, on the other hand, had elevated ALT and reduced WBC and PLT. These findings suggested that MTX@PC-FA NEs could effectively increase the enrichment of MTX in inflamed joints, avoiding the side effects of MTX extensively dispersed in normal tissues.

Conclusion
Despite being the gold standard for the treatment of RA, traditional MTX formulations have long been criticised for their poor targeting and significant side effects. In this study, we developed a macrophage-targeted NEs delivery system based on the overexpression of FR on activated macrophages in arthritic joints. The MTX@PC formed by MTX and phospholipid was used as the emulsifier, the oil phase composed of soybean oil and MCT served as the core to stabilise the preparation and FA was utilised to modify the surface of the NEs to target macrophages. In vitro and in vivo experiments showed that this strategy could effectively target Figure 5. Therapeutic experiment of MTX@PC-fa nes. a: The experimental outline of different formulations on aia mice. aia mice received treatments six times from 3rd to 21th day every 3 days. aia mice were sacrificed on day 22. groups included Healthy, Control, free MTX, MTX@PC nes and MTX@PC-fa nes, and the dose of MTX was 5 mg/kg. B: Changes of paw thickness during treatment. Data are showed as mean ± SeM, n = 6. ***p < 0.001, ****p < 0.0001. C: Changes of clinical score during treatment. Data are shown as mean ± SeM, n = 6. *p < 0.05, **p < 0.01. D: representative photos of arthritic paws from various treatment groups. Scale bar: 5 mm. e: representative photos of H&e staining from ankle joints. Scale bar: 200 µm. f: representative photos of Safranin-o staining from ankle joints. Scale bar: 200 µm. macrophages in arthritic joints. Moreover, the therapeutic experiments indicated that the MTX@PC-FA NEs improved the therapeutic index of MTX and generated minor systemic adverse effects compared to free MTX, due to the stronger accumulation of MTX@ PC-FA NEs in inflamed joints. Importantly, the components of MTX@PC-FA NEs are all pharmaceutical excipients, and the manufacturing process is straightforward and repeatable, which has obvious advantages compared with the intricate MTX delivery system commonly reported today. Finally, our research provides a reasonable and innovative approach for the nano-based therapy of RA, which has great potential for clinical translation. Figure 6. Serum concentrations of various cytokines. a: Serum concentration of pro-inflammatory cytokine Tnf-α. B: Serum concentration of pro-inflammatory cytokine il-1β. C: Serum concentration of pro-inflammatory cytokine il-6. D: Serum concentration of anti-inflammatory cytokine il-10. Data are shown as mean ± SD, n = 6. *p < 0.05, **p < 0.01, ***p < 0.001.