Methotrexate-Loaded Nanoparticles Ameliorate Autoimmune Arthritis by Regulating the Balance of Interleukin-17-Producing T Cells and Regulatory T and B Cells

Jin-Sil Park Catholic University of Korea Donghyun Lee Catholic University of Korea SeungCheon Yang Catholic University of Korea Hyun Sik Na Catholic University of Korea Keun-Hyung Cho Catholic University of Korea JeongWon Choi Catholic University of Korea Heebeom Koo Catholic University of Korea Mi-La Cho (  iammila@catholic.ac.kr ) Catholic University of Korea https://orcid.org/0000-0001-5715-3989 Sung-Hwan Park Catholic University of Korea


Background
Rheumatoid arthritis (RA) is a progressive autoimmune disease characterized by synovial in ammation, hyperplasia, and formation of rheumatoid pannus, resulting in destruction of the adjacent articular cartilage and bone structure. The development of RA involves a complex interplay between immune cells and in ammatory mediators including in ammatory cytokines, proteolytic enzymes, and prostanoids [1][2][3]. Although the causes of RA are unclear, in ltration of T and B cells into the joints leads to induction of in ammatory cytokines, such as tumor necrosis factor (TNF)-α, interleukin (IL)-6, IL-1β, and IL-17A and autoantibodies and drives the proliferation of broblast-like synoviocytes and destruction of bone [3,4].
Methotrexate (MTX) is a rst-line disease-modifying antirheumatic drug (DMARD) that alleviates articular damage in RA [14]. It is an antifolate metabolite that inhibits folate-dependent enzymes in the de novo synthesis of purines and pyrimidines [15,16]. MTX is used as monotherapy for patients with early RA [17] and as an anchor drug for combination therapy with other DMARDs or biologics in patients with established RA who are MTX-insu cient responders [18]. However, long-term use of MTX leads to drug resistance and causes severe side effects such as nausea, neutropenia, pulmonary brosis, and hepatitis [15,19].
Nanoparticles (NPs) are promising therapeutics due to their ability to deliver and release drugs [20,21].
Multiple NPs have been developed for drug delivery in various diseases. They inhibit loss due to fast excretion of drug from body by sustained release. NPs are typically administered by intravenous injection, oral feeding, or subcutaneous injection. Intravenously injected NPs are rapidly effective but also rapidly excreted and have a high risk of side effects. Oral feeding is the easiest and low risk, but the absorption rate is too low. Subcutaneous injection shows moderate effectiveness in relation to these methods, so has an advantage in occupying the middle ground. However, most NP studies for RA therapy have used intravenous injection and oral feeding.
MTX-loaded nanoparticles (MTX-NPs) ameliorated RA after subcutaneous injection. They were formulated with hydrophobic poly (D, L lactide-co-glycolide) (PLGA) and amphiphilic polyvinyl alcohol (PVA). MTX was stably loaded into the NPs. MTX-NPs attenuated the severity of autoimmune arthritis and reciprocally regulated Th17 and regulatory T and B cells in vivo.
Preparation and characterization of MTX loaded nanoparticles PLGA (50 mg) and methotrexate (5 mg) were dissolved in DMSO at 60℃. The solution was added dropwise to an aqueous solution of 1% PVA (w/v). The solution was homogenized at 7000 RPM for 2 min using a homogenizer (Ultra Turrax® T-25 homogenizer; IKA®-Werke, Staufen, Germany). After homogenization, non-encapsulated substances were removed by dialysis in distilled water for 1 h through a 14,000 molecular weight cut-off membrane. The size distribution of MTX-NP was measured in PBS using a Zetasizer Nano ZS90 (Malvern Instruments, Malvern, UK) at 25℃. Encapsulated MTX was quanti ed by assaying the absorbance of MTX at 370 nm and encapsulation e ciency was calculated by the formula: ((amount of encapsulated drug/amount of added drug) × 100%). MTX-NP was placed in a dialysis bag and immersed in a container containing 50 mL of PBS to analyze the drug release pro le. At predetermined time points, 200 μL of external PBS were removed and the absorbance at 370 nm of MTX was measured using Microplate Reader Synergy H1 (Bio-Tek, USA).

CIA induction and treatment with nanoparticles
Six-week-old male DBA/1J mice were purchased from Orient Bio Inc. (Seongnam, Korea). To induce CIA in mice, CII was dissolved overnight in 0.1 N acetic acid (4 mg/mL) with gentle rotation at 4°C. DBA/1J mice were injected intradermally at the base of the tail with 100 mg of CII emulsi ed in Freund's adjuvant (Chondrex). Two weeks later, 100 µg of type II collagen dissolved and emulsi ed 1:1 with incomplete Freund's adjuvant (Difco) was administered to the hind leg of mice as a booster injection. On day 24 after the rst immunization, mice were injected subcutaneously with 2.5 mg/kg MTX or MTX-NPs twice weekly.
Animals were maintained under speci c pathogen-free conditions at the Institute of Medical Science of the Catholic University of Korea and were fed standard mouse chow and water. All experimental procedures were examined and approved by the Animal Research Ethics Committee of the Catholic University of Korea; the procedure conformed to all National Institutes of Health of the United States guidelines (Permit number: 2020-0067-01).

Histology
Mouse joint tissues were xed in 10% neutral-buffered formalin, decalci ed in a decalcifying agent (National Diagnostics, Atlanta, GA, USA), embedded in para n, and sectioned. The sections (5 µm thick) were stained with hematoxylin and eosin (H&E) and scored as described previously [22]. Immunohistochemical analysis of IL-1β, TNF-α, and vascular endothelial growth factor (VEGF) (Santa Cruz Biotechnology, Dallas, TX, USA) was performed using the Vectastain ABC kit (Vector Laboratories, Burlingame, CA, USA). The sections were examined by light microscopy (Olympus, Tokyo, Japan). Positive cells were enumerated visually by four individuals, and the mean values were calculated.

Statistical analysis
All statistical analyses were performed using Prism (v. 8 for Windows; GraphPad Software). P-values were calculated by two-tailed paired t-test and two-way analysis of variance (grouped). P < 0.05 was considered indicative of statistical signi cance.

Preparation and characterization of MTX-NPs
We prepared PLGA NPs via a homogenization method using PVA as stabilizer. MTX was physically loaded into the NPs (Fig. 1a). MTX was stably encapsulated in a hydrophobic core (PLGA) with an encapsulation e ciency of 76.6 ± 4.9%. The size of MTX-NPs was 227.2 ± 1 nm and MTX-NPs were spherical by transmission electron microscopy (TEM) (Fig. 1b). The size is suitable for injection into body by syringes with small needle. The polydispersity index (PdI) and zeta potential of MTX-NP were 0.139 ± 0.041 and − 2.8 ± 0.02, respectively, showing a homogeneous formulation without aggregation and nearneutral surface, respectively. In PBS at pH 7.4, free MTX was slowly released from the PLGA core (Fig. 1c), indicating sustained release.

MTX-NPs attenuate the severity of autoimmune arthritis
To determine whether MTX-NPs could modulate the development of arthritis in vivo, free MTX or MTX-NPs were administered to mice with CIA at 3 weeks after CII immunization. Subcutaneous injection of MTX-NPs in arthritic mice signi cantly reduced the arthritis score and incidence compared with vehicletreated CIA mice. Injection of free MTX also reduced the arthritis score and incidence in CIA mice, but statistical signi cance was not consistently achieved (Fig. 2a). Histologic examination of joints stained with H&E showed that the ankles of MTX-NPs-treated mice exhibited less severe in ammation, bone damage, and cartilage damage compared with vehicle-treated mice. Application of MTX-NPs, in particular, exerted a more profound inhibitory effect on joint destruction compared with free MTX (Fig.  2b). Furthermore, the levels of in ammatory mediators-including IL-1β, TNF-α, and VEGF-were signi cantly lower in the joint sections from MTX-NPs-treated mice compared with vehicle-treated mice (Fig. 3).

MTX-NPs reciprocally regulate the Th17 cells and Treg cells in vivo
To evaluate whether MTX-NPs suppress Th17 cells in vivo, the number of CD4 + IL-17 + Th17 cells in the spleens from CIA mice injected with MTX-NPs was investigated by confocal microscopy. The number of Th17 cells was lower in MTX-NPs-or free MTX-treated CIA mice compared with vehicle-treated CIA mice (Fig. 4a). STAT3 phosphorylation in CD4 + cells decreased in MTX-NPs-or free MTX-treated CIA mice compared to vehicle-treated CIA mice, but there was no statistical signi cance (Fig. 4b). To investigate whether MTX-NPs reciprocally regulate the population of Th17 and Treg cells in vivo, the number of CD4 + CD25 + Foxp3 + Treg cells in the spleen was investigated. The number of Treg cells in the spleen of MTX-NPs-injected mice signi cantly increased compared to vehicle-injected mice, and the increase was also signi cant compared with free MTX-injected mice (Fig. 4c). In addition, the mRNA levels of IL-6 and IL-17A, in ammatory cytokines related to Th17 cell development, signi cantly decreased in ex vivo splenocytes of mice injected with MTX-NPs compared to mice injected with vehicle (Fig. 4d).

MTX-NPs increase regulatory B cells in vivo
To evaluate whether MTX-NPs act on B-cell responses in vivo, we examined the number of ex vivo CD19 + B220 + GL-7 + Fas + GC B cells and CD19 + CD25 + Foxp3 + regulatory B (Breg) cells in splenocytes from CIA mice injected with MTX-NPs by ow cytometry. The number of GC B cells was lower whereas the number of regulatory B cells was higher in ex vivo splenocytes from MTX-NPs-injected CIA mice compared with vehicle-injected mice, but there was no statistical signi cance (Fig. 5).

Discussion
We investigated the therapeutic potential of MTX-NPs administered subcutaneously in a murine model of autoimmune arthritis. MTX-NPs signi cantly reduced the clinical and histologic severity of arthritis. MTX-NPs reduced the number of TNF-α-and VEGF-positive cells and reciprocally regulated the number of Th17 and Treg cells in the spleen from MTX-NP-treated CIA mice. Moreover, injection of MTX-NPs showed a tendency to increase the number of regulatory B cells in ex vivo splenocytes of CIA mice.
Synovial in ammation in RA involves in ltration of in ammatory cells, including monocytes/macrophages, B cells, and T cells, which are sources of chemokines and in ammatory cytokines [23]. T cells are a key player in the in amed joints of patients with RA [24]. Although most subsets of CD4 + T cells are involved in RA pathogenesis, Th17 cells in particular positively correlate with disease activity in RA [25]. In CIA mice, de ciency of IL-17 suppressed the development of arthritis and reduced the production of type II collagen-speci c immunoglobulin [26]. Moreover, blockade of IL-17 ameliorated the severity of arthritis and prevented synovial in ammation and joint destruction in CIA mice [27]. MTX exerts a potent therapeutic effect by modulating humoral and cellular immune responses in RA management [14,28]. MTX increases the sensitivity of T cells to apoptosis [16] and inhibits NF-κB activation in T cells via tetrahydrobiopterin 4 depletion and JNK activation [29]. MTX suppresses the expression of IL-6 and IL-6-driven proliferation of broblast-like synoviocytes from patients with RA [30]. Thomas et al. demonstrated that MTX suppressed the JAK/STAT pathway in the Drosophila system and human macrophage lines [31], but the effect of MTX on JAK/STAT signaling and Th17 and Treg frequencies in the CIA model are unclear. The number of STAT3 phosphorylated CD4 + cells and of Th17 cells in the spleen of CIA mice were decreased by injection of free MTX, and the effect was increased by injection of MTX-NPs.
The importance of B cells in RA pathogenesis has been demonstrated. B-cell-de cient mice had impaired CIA development [32], and B-cell depletion by an anti-CD20 antibody delayed CIA development [33]. Immunization with type II collagen induced the formation of germinal centers in lymph nodes, which was necessary for CIA development [34]. Breg cells control autoimmune diseases by secreting antiin ammatory cytokines (e.g., IL-10 and IL-35) and transforming growth factor-β [35,36]. Moreover, Breg cells suppress the differentiation of proin ammatory lymphocytes (e.g., Th1 and Th17 cells) and dendritic cells [37]. The frequency of Breg cells increased and that of Th17 cells and germinal center B cells decreased in ex vivo splenocytes of MTX-NP-treated CIA mice.
The elimination half-life of free MTX in human is 0.16 h for subcutaneous injection [38]. Our NPs provided sustained release of MTX and increased the residence time. Although the release of MTX from NPs was faster than in prior reports, it affected the therapeutic results signi cantly. It is necessary to test and optimize further the release pattern of MTX by changing NP size or composition. In addition, MTX in NP could be taken up by T cells and B cells before release. This is because the NP surface was coated with PVA; we did not use polyethylene glycol groups to prevent fouling and unwanted uptake into immune cells. In addition, further modi cation with biological ligands will provide information for speci c targeting of NPs to particular cell types.

Conclusion
In summary, our study showed a therapeutic e cacy of MTX-NPs in mice with CIA. Subcutaneous injection of MTX-NPs effectively alleviated the development of arthritis and suppressed the in ltration of in ammatory factors-expressing cells in the joint from CIA mice. Moreover, the number of pathogenic Th17 cells decreased while the number of regulatory T and B cells increased in the spleen of the MTX-NPs-injected group. These ndings suggest that MTX-NPs have potential as a more advanced therapeutic strategy to overcome the limitations of MTX therapy.
JSP and DHL participated in the study design, data interpretation, and writing the manuscript. SCY, HSN, KHC and JWC carried out the animal experiments and analyzed data. CWY and SHP participated in the study design and data interpretation. HK, MLC and SHP conceived and designed the study, interpreted the data, and made critical revisions of the manuscript for important intellectual content. All authors read and approved the nal manuscript.   MTX-NPs suppressed the levels of in ammatory mediators in vivo. Beginning 3 weeks after the rst immunization with type II collagen (CII), mice were injected subcutaneously with vehicle, free MTX, or MTX-NPs twice per week for 7 weeks (n = 5/group). At 70 days after the rst immunization with CII, sections of joint tissues (n = 5/group) were stained with antibodies against interleukin (IL)-1β, tumor necrosis factor (TNF)-α, and VEGF. Graphs present numbers of antibody-positive cells for each cytokine.
Data are means ± SEM of two independent experiments. *P < 0.05 vs. control group.