Clinical samples

Forty-eight patients with clinically active RA were diagnosed through fulfillment of the 1987 American College of Rheumatology criteria [37] and were recruited from the outpatient clinic of the Rheumatology department of First Affiliated Hospital of Kunming Medical University of Yunnan Province. Written informed consent for the present study’s use of human peripheral blood was obtained from all forty-eight patients and thirty-seven healthy donors, and ethical approval was obtained from both the Ethics Committee of the First Affiliated Hospital of Kunming Medical University and the Committee on Human Subject Research and Ethics of Yunnan University (Approved number: 20190801). All experiments were carried out according to the relevant guidelines.

Cell sorting and cell culture

For the isolation of PBMCs, the heparinized whole blood samples (4 mL) from each RA patient and healthy donor were diluted with an equal volume of PBS (Biosharp Life Sciences, China), layered onto Ficoll-Hypaque (Solarbio Life Sciences, China) and centrifuged at 2,000 rpm/min, at 20 °C for 20 min by density gradient centrifugation. Mononuclear cells were recovered from the interface and washed twice with PBS. The CD3+ T cells were purified from about 2–10 × 108 PBMCs, which were passaged over a 30 μm nylon wool column, incubated using immunomagnetic microbeads (Miltenyi Biotec, Germany) for 30 min at 4 °C, and finally selected with QuadroMACS™ Separator (Miltenyi Biotec, Germany) via magnetic LS cell separation columns (Miltenyi Biotec, Germany). All assays were performed according to the protocols recommended by the manufacturer. Aliquots of 4–10 × 106 PBMCs and purified CD3+ T cells were optionally stimulated with anti-CD3 (clone UCHT1, 2.5 µg/ml) and anti-CD28 (clone CD28.2, 1.25 µg/ml) antibodies (BD Bioscience, USA) for 48 h in the presence of MTX does (0.001–1 µg/mL or 0.01 µg/mL) or AMD3100 (1 µg/mL) in vitro. Jurkat cells were offered by Professor Shenghong Li from the Insititute of Botany, the Chinese Academy of Sciences, Kunming, China. The PBMCs, purified CD3+ T cells and Jurkat cells were cultured in flat-bottom plates (Corning, USA) in RPMI-1640 medium (Gibco, Invitrogen Corporation, USA) supplemented with 10% fetal calf serum (Viva Cell Biosciences, China), 100 U/mL penicillin and 100 µg/mL streptomycin (BasalMedia Technologies, China) at 37 °C under 5% CO2.

Flow cytometry (FCM)

Expression of chemokine receptor CXCR4 on human T cells was analyzed using flow cytometry. 100 µL of human PBMCs (about 5 × 105 cells) were incubated with mouse anti-human antibodies, including anti-CD184-PE and anti-CD3-FITC (20 µL, BD Biosciences, USA), anti- CD4- and CD8-FITC (5 µL, BioLegend, USA), and their corresponding mouse isotypes as the negative controls. Then, the cells were incubated for 30 min at 4 °C in the dark. Subsequently, the cells were washed twice with 1 mL FACS buffer (1% BSA in PBS) by centrifugation (400 g/min, 5 minutes, 4 °C) to remove unbound antibodies. Cells were then resuspended in 500 µL PBS and sieved through a 40-µm pore size mesh to generate single-cell suspensions. Finally, labeled cells were measured on an Accuri C6 Flow Cytometer (Accuri Cytometers, BD Biosciences, C6, USA) and analyzed by FlowJo software (Treestar FlowJo v10.8.1, Ashland, USA). Negative fluorescence was gated with FITC-conjugated IgG or PE-conjugated IgG. At least 10,000 events collected within the lymphocyte gate. The single fluorochrome-stained cells were used for instrument compensation and the photomultiplier tube (PMT) setup. Analysis of CD3+, CD4+, CD8+ T cell subsets were performed on the total CD3+ T cells within the lymphocyte gate.

Quantitative real-time polymerase chain reaction (qRT-PCR)

Total messenger RNA (mRNA) was extracted using RNAiso Plus Reagent (Clontech Laboratories, Takara Bio Group, Japan) according to the manufacturer’s instructions. Potential genomic DNA contamination was removed from the samples and complement DNA (cDNA) was synthesized from 1 µg total RNA per sample using random primers with PrimeScript™ RT reagent Kit with gDNA Eraser (Clontech Laboratories, Takara Bio Group, Japan) according to the recommended protocol. The qRT-PCR with up to 100 ng cDNA was amplified using the SYBR® Premix Ex Taq™ (Tli RNaseH Plus, Clontech Laboratories, Takara Bio Group, Japan). The ABI Prism 7500 Sequence Detection System (Applied Biosystems, ABI 7500, USA) was used to measure the amplified PCR product. The qRT-PCR thermal cycle included 1 cycle of 95 °C for 10 s, followed by a two-step PCR program: 40 cycles of 95 °C for 5 s, and 60 °C for 34 s, followed by a melt curve analysis. Each sample was performed in triplicate (technical replicates) and repeated at least three times independently. The oligonucleotide primers used for qRT-PCR analyses were listed in Table S2. The relative expression of qRT-PCR data was determined using the 2−∆∆Ct method to compare mRNA expression levels of the target gene with the housekeeping gene (β-actin or 18 S).

CXCR4 methylation-specific PCR analysis based on nucleotide sequencing

The Jurkat cells, or the 0.01 µg/mL MTX treated CD3+ T cells, genomic DNA was extracted by DNA mini isolation kits (Qiagen, Germany) according to the manufacturer’s protocol. Genomic DNA was bisulfite converted with the Zymo EZ DNA methylation kits (Zymo Research, Irvine, CA, USA), and the CpG sites of the bisulfite modified DNA were amplified with bisulfite specific PCR consisted of an initial denaturation step at 95 °C for 5 min, followed by 38 cycles of denaturation at 95 °C for 30 s, annealing at 54 °C for 30 s, extension at 72 °C for 30 s, and final extension step at 72 °C for 7 min by the EpiArt HS Taq Master Mix (Vazyme Biotech, Nanjing, China). The PCR products were verified on 2% agarose gel, followed with sequencing by Sangon Biotech Co., Ltd. (Shanghai, China). For each CpG site, the methylation ratio was calculated as the number of cytosine reads divided by the sum of cytosine and thymine reads [38]. Bisulfite PCR primer sequences were designed by Methyl Primer Express Software v1.0 (Applied Biosystems, USA) and listed in Table S2.

In vitro transmigration

The interaction between CXCR4 and extracellular matrix (ECM, Sigma-Aldrich, USA) assays were conducted on sorting cells with 0.01 µg/mL MTX (Sigma-Aldrich, USA) and CXCR4-specific antagonist 1 µg/mL AMD3100 (Sigma-Aldrich, USA) treatment, using NeuroProbe 48-Well Micro Chemotaxis Chambers (Neuro Probe, Inc. Ap48, USA). Briefly, the Chemotaxis Chambers were separated using 5 μm pore polycarbonate transwell filter membranes (Neuro Probe, Inc. USA) and the lower side was coated with 0.01 mg/mL ECM diluted with RPMI-1640 culture media. In the lower wells, the 100 ng/mL human CXCL12 (SDF-1α) protein (Miltenyi Biotec, Germany) was supplemented as the chemoattractant. And then, 1 × 104 mock or pre-treated cells suspended in 50 µL RPMI-1640 culture media were added into each upper chamber. After 4 hours of incubation at 37 °C and 5% CO2, the membrane was carefully removed. Cells on the upper membrane surface were gently scraped with a cotton swab. The membranes were fixed in 4% (w/v) paraformaldehyde (PFA, Servicebio Technology, China) for 30 min and washed in PBS before staining with 0.1% Hoechst 33342 (Sigma-Aldrich, USA) for 5 min at room temperature, followed by rinsing three times with PBS. The membranes were mounted on glass slides with Fluoro-Gel (Electron Microscopy Sciences, USA) and cells per unit area in triplicate were determined by counting four representative microscopic fields under an epifluorescence microscope at a magnification of ×10 (Leica, DMi8, Germany).

Cell viability assay

To examine the toxicity of low-dose MTX on the activated T cells, PBMCs were seeded into 96-well flat bottom plates and stimulated with anti-CD3 (clone UCHT1, 2.5 µg/ml) and anti-CD28 (clone CD28.2, 1.25 µg/ml) antibodies (BD Bioscience, USA) for 0, 24 and 48 h in PBS or MTX (0.001–1 µg/mL) at the density of 5 × 104 cells per well (100 µL) in quadruplicate. The same volume of PBS was kept consistent with the 0.001–1 µg/mL MTX group added in the vehicle control group for 0, 24 and 48 h, respectively. The effects of MTX on the survival of cells was evaluated with CellTiter 96® Aqueous One Solution Reagent Solution Cell Proliferation Assay kit (Promega, USA), following the manufacturer instructions. Briefly, at 2 h before each of the desired time points, each well was mixed with 20 µL of MTS ((3-(4,5-dimethylthiazol-2-yl)-5-(3-carboxymethoxyphenyl)-2-(4-sulfophenyl)-2 H-tetrazolium, solution inner salt, and cells were incubated at 37 °C with 5% CO2 for 2 h. The absorbances of the samples were detected at a wavelength of 490 nm using a micro-plate reader (Molecular Devices, SpectraMax 190, USA).

Enzyme-linked immunosorbent assay (ELISA)

The clinical sample blood serums were collected, and enzyme-linked immunosorbent assay (ELISA) kits for detecting human cytokines IL-1, IL-2, IL-6, IL-8, TNF-α (Elsbio Biotec, China) and CXCL12 were obtained (Solarbio Life Sciences, China). ELISA experiments were conducted following the manufacturers’ instructions. Plates were read by a micro-plate reader (Molecular Devices, SpectraMax 190, USA) at a wavelength of 450 nm.

In vivo MTX and AMD3100 treatment

For in vivo assessment of the effect of MTX on the treatment of RA, we used adjuvant-induced arthritis (AA) mouse models that were widely used for drug validation, in particular for T cell-related pathways leading to arthritis [39]. Briefly, thirty-two 10-week-old female C57BL/6J mice were purchased from Vital River Laboratories (Beijing, China) to be used as the experimental model of arthritis. They were kept in healthy conditions with clean air and given plenty of food and water. To avoid superimposed side effects of MTX/AMD3100 treatment and adjuvant immunization on the growth and body weight of arthritis mice, a preventative model with MTX and AMD3100 treatments began on day 1 and lasted 15 days. Briefly, the mice were divided into four groups (eight mice per group). The non-arthritic healthy control mice and untreated arthritis mice received 150 µl normal saline (NS) every three days for 15 days (5 times total) and were then divided into two groups: namely non-arthritic healthy control mice were fed normally, untreated arthritis mice were induced by intradermal tail injection with 150 µl of 4 mg/mL heat-killed Mycobacterium tuberculosis suspended in complete Freund’s adjuvant (CFA) (day16, Sigma Aldrich, USA). The MTX or AMD3100 treatment of them established arthritis group was fed with 2.5 mg/kg MTX or 5 mg/kg AMD3100 from day 1 to day 15 every three days, and arthritis was subsequently induced with the same regimen. Mice were weighed and scored for clinical signs of arthritis once every two days from day 17 to day 48 post-immunization. On day 20, 27, and 38, blood was collected and analyzed for CXCR4+PBMCs. All animal studies and relevant procedures were reviewed and approved by the ethics committee of the institutional animal care and use committee, Yunnan University (IACUC, YNU) (Approved number: 20190801).

Collagen induced arthritis (CIA) mouse model

The type II collagen induced arthritis (CIA) mouse models, which were extensively used to identify potential pathogenic mechanisms of autoimmunity, including the role of individual cell types in disease onset and progression [40], were induced and the effect of CXCR4 on the occurrence and development of RA was assessed. Briefly, 10-week-old female C57BL/6J mice were injected intradermally at the base of the tail with 15 mg/kg chicken type II collagen (CII, Chondrex, USA) dissolved in 10 mM acetic acid and emulsified in an equal volume of CFA (Sigma Aldrich, USA) containing 4 mg/mL Mycobacterium tuberculosis to induce the arthritis mice model (day 0). Twenty-one days (day 21) after the first primary immunization, the mice were given a single booster injection in the same manner. The arthritis index assay, blood collection, and histological assessment were carried out. The expression of TNFα, IL-1β and IL-6 in lung tissue was detected by qRT-PCR and the respective primers used were listed in Table S2.

CXCR4 conditional knockout

Mx1-Cre−/CXCR4flox/flox mice (C57BL/6J) were purchased from the Jackson laboratory (The Jackson laboratory, USA). In these mice, two LoxP sites were genetically inserted in the side of the second exon of the Cxcr4 gene. LoxP site was genotyped using genotyping PCR primers MG-Loxp-F and MG-Loxp-R. A 577-bp and 481-bp DNA band was expected from either flox homozygous mice or wild type/flox heterozygous mice respectively. Meanwhile, Mx1-cre+/CXCR4+/+ mice (C57BL/6J) were purchased from Nanjing Biomedical Research Institute of Nanjing University (Nanjing, China). These mice overexpressed the Cre protein under control of the inducible Mx1 promoter. Mx1-Cre+ was genotyped using genotyping primers MG-Cre-F and MG-Cre-R, and a 408-bp PCR DNA band was expected as well. The Mx1-Cre−/CXCR4flox/flox mice and Mx1-Cre+/CXCR4+/+ mice were founders to be crossbred to obtain Mx1-Cre+/CXCRflox/flox mice. To obtain CXCR4 conditional knockout mice, Mx1-Cre+/CXCR4flox/flox mice were administrated 12.5 mg/kg polyinosinic-polycytidylic acid (PI: PC, Novus Biologicals, USA) by intraperitoneal injection. PI: PC injections were performed every two days, for a total of 3 times. Genetic deletion of Cxcr4 allele genes was verified using genotyping primers MG-CXCR4-F and MG-CXCR4-R. CXCR4+PBMCs and CXCR4+CD3+ T cells on mouse PBMCs were confirmed by flow cytometry. All the primers used were listed in Table S2. The CIA mice model of Cxcr4 conditional knockout was induced as above (Collagen induced arthritis mouse model) mentioned. In brief, 10-week-old female Mx1-Cre+/Cxcr4flox/flox mice were immunized intradermally at the base of the tail with 15 mg/kg chicken type II collagen (CII, Chondrex, USA) dissolved in 10 mM acetic acid and emulsified in an equal volume of CFA (Sigma Aldrich, USA) containing 4 mg/mL Mycobacterium tuberculosis (day 0) after the PI: PC was injected at day -2, day -4, and day -6 respectively. On day 21, the mice were given a single booster injection with the same manner. The arthritis index assay, blood collection, histological assessment, and the expression of TNFα, IL-1β and IL-6 in lung tissue were carried out.

Histological assessments of arthritis

At the end of the mice experimental procedure, the knee joints and paws were analyzed for histological assessment after the mice were sacrificed. Animals were perfused with 4% phosphate-buffered paraformaldehyde (PFA). Then, animal joints, paws and lungs were removed, dissected, and sliced into 5-µm cross-sections. Sections were stained by hematoxylin and eosin (H&E) (Beyotime Life Sciences, China) according to standard procedures. The severity of arthritic changes, such as articular cartilage destruction and bone erosion were histologically scored as described previously [41]. Briefly, we scored cartilage destruction from 0 to 3 and bone erosion from 0 to 5. The positive inflammatory areas of mouse lung sections were measured on each slide and quantified using ImageJ software [42].

Statistical analysis

All reported values were expressed as mean ± SEM (Standard Error of Mean), unless otherwise stated. The arthritis mice and MTX and AMD3100 treated experimental groups included at least five mice. Statistical analyses were done with SPSS 22.0 computer software (SPSS Inc, Chicago, USA). The statistical significance between healthy controls and patient, sham and MTX treated mice, sham and arthritis mice were analyzed using unpaired Student t-test. The statistical significance between two groups of qRT-PCR, in vitro transmigration and cell viability assay were analyzed with a paired Student’s t-tests. In all statistical analyses, the probability (P) values less than 0.05 were statistically significant.