Methotrexate upregulates circadian transcriptional factors PAR bZIP to induce apoptosis on rheumatoid arthritis synovial fibroblasts

Synovial tissues were obtained from RA patients (eight females and two males; aged 59.3 ± 3.4 years) during joint surgery. Eight of 10 patients took methotrexate and their average dosage was 7.4 ± 1.4 mg/week (Table 1) (see Additional file 2). This study has been approved by the ethics committee of Kobe University Graduate School of Health Sciences (#579–1) and Kobe Kaisei Hospital (#0072), in accordance with the Declaration of Helsinki. Written informed consent was obtained from each patient before study enrolment. Tissues were minced and treated with 2 mg/ml collagenase (Wako, Tokyo, Japan) in serum-free Dulbecco’s modified Eagle’s medium (DMEM; Nissui, Tokyo, Japan) at 37 °C for 1 h. Primary cultured synovial cell lines were established and maintained in DMEM including 10% heat-inactivated fetal bovine serum (FBS; Thermo, Waltham, MA, USA), 1% penicillin–streptomycin (100 U/ml; Life Technologies, Carlsbad, CA, USA), and 1% l-glutamine (2 mM; Life Technologies), in a humidified incubator at 37 °C in the presence of 5% CO₂. Harvested cells were continuously cultured to obtain synovial fibroblasts, and cells of passages 3–6 were used in the entire experiments.

Synovial fibroblasts were incubated in 50% horse serum and incubated for 2 h to synchronize the expression of clock genes [17, 31].

Synovial fibroblasts (3.0 × 10³ cells) were cultured in serum-free DMEM with or without MTX (1, 10, and 100 nM, 1 μM; Sigma Aldrich, St. Louis, MO, USA). After incubating for 24–72 h, cell viabilities were measured as the 450-nm absorbance of reduced WST-8 (2-(2-methoxy-4-nitrophenyl)-3-(4-nitriphenyl)–5-(2, 4-disulphonyl)-2H–tetrazolium, monosodium salt) using the Cell Counting Kit-8 (Dojindo, Kumamoto, Japan). The values were represented relative to MTX-untreated cells (control).

Synovial fibroblasts (8.0 × 10⁴ cells) were cultured in serum-free DMEM with or without MTX (10, 100 nM) for 24–32 h. After incubation, total RNA was extracted using the RNeasy Mini Kit (QIAGEN, Hilden, Germany). Then, reverse transcription was performed with the ReverTra Ace® qPCR RT Kit (Toyobo, Osaka, Japan) and analyzed on the StepOnePlus™ Real-Time PCR System (Applied Biosystems, Foster City, CA, USA). The TaqMan probes used were: Hs00154147_m1 for Bmal1, Hs00231857_m1 for Clock, Hs00256143_m1 for Per2, Hs00172734_m1 for Cry1, Hs00609747_m1 for Dbp, Hs00171406_m1 for Hlf, Hs01115720_m1 for Tef, Hs00993282_m1 for E4bp4, Hs00154189_m1 for Bik, and Hs00427621_m1 for TATA box binding protein (Tbp). Expression levels were normalized to Tbp.

Synovial fibroblasts (5.0 × 10⁵ cells) were cultured in serum-free DMEM with or without MTX (10, 100 nM) for 24–48 h, and lysed with RIPA buffer (Wako) to obtain cytoplasmic proteins. Samples were subjected to sodium dodecyl sulfate polyacrylamide gel electrophoresis (SDS-PAGE), transferred to polyvinylidene difluoride (PVDF) membrane (Millipore, Bedford, MA, USA), probed with antibodies, and developed by ImmunoStar® LD (Wako). Antibodies (Abs) used were: anti-PER2 Ab (sc-101,105; Santa Cruz, Dallas, TX, USA), anti-CYTOCHROME C Ab (ab13575; Abcam, Cambridge, UK), anti-BIK Ab (NB100–56109; Novus, Littleton, CO, USA), anti-Actin Ab (sc-1616; Santa Cruz), anti-mouse IgG Ab (NA9310; GE Healthcare, Chicago, IL, USA), and anti-rabbit IgG Ab (NA9340; GE Healthcare).

Synovial fibroblasts (3.0 × 10³ cells) were cultured in serum-free DMEM with or without MTX (10, 100 nM) for 32–48 h, and protein expression of PER2 and CYTOCHROME C was measured by the In Cell ELISA test kit (3440–02; Thermo). Antibodies were as already described.

Genomic DNA was obtained from RA synovial fibroblasts and amplified by PCR to generate the luciferase reporters. The primer pairs used for amplifying human Bik promoter containing D-box (−780 to +176) were 5′-TGGCCTAACTGGCCGTAAACAAGCTTTGCCGTGC-3′ (forward) and 5′-CGCCGAGGCCAGATCATGCTGGCAGCGTCTGTA-3′ (reverse). The primer pairs used for amplifying human Bik promoter without D-box (−260 to + 323) were 5′-TGGCCTAACTGGCCGCCTCTTGGAGCCTCGGTT-3′ (forward) and 5′-CGCCGAGGCCAGATCTTGCTGGAGCGGTAAAACC-3′ (reverse). The KpnI recognition sequence was added to the 5′ ends of forward primers and the BglII recognition sequence was added to the 5′ ends of reverse primers. The PCR products were cloned into the KpnI and BglII site of the pGL4.10 (luc2) vector (Promega, Madison, WI, USA) using In-Fusion® HD Cloning Plus (Takara, Shiga, Japan). The luciferase reporters containing Per2 promoters were generated with reference to Yoshida et al. [17]. D-box motifs of Per2 promoter were mutated from 5′-TTATGTAA-3′ to 5’-CGCCAGGC-3′ (−372 to −365) and from 5′-TTACGTAA-3′ to 5′-CAGCGTAA-3′ (−47 to −40) (see Additional file 3).

Synovial fibroblasts (4 × 10⁴ cells) were transfected with 500 ng of the pGL4.10 (luc2) vector containing various Per2 and Bik promoters using Lipofectamine 3000 Transfection Reagent (Thermo). As an internal control, 35 ng of pRL-TK (Promega) containing the herpes simplex virus thymidine kinase promoter driving Renilla luciferase was cotransfected. After 24 h of incubation for transfection, cells were treated with 10 and 100 nM MTX for 24 h, and analyzed for luciferase activity using the Dual-Luciferase Reporter Assay (Promega). Activities of both firefly and Renilla luciferases were measured, and the activity of firefly luciferase was normalized by Renilla luciferase. The values were shown as relative variations to MTX-untreated cells.

Synovial fibroblasts (6.0 × 10³ cells) were cultured in serum-free DMEM with or without MTX (10 nM) for 24 h. Then, cells were fixed with 4% formaldehyde, stained with anti-PER2 Ab, anti-BIK Ab, anti-CYTOCHROME C Ab, anti-mouse IgG (H + L), F(ab′)₂ Fragment (Alexa Fluor® 594 Conjugate) (#8890; Cell Signaling Technology, Danvers, MA, USA), anti-rabbit IgG (H + L), F(ab′)₂ Fragment (Alexa Fluor® 488 Conjugate) (#4412; Cell Signaling Technology), and DAPI (0.5 μg/ml; Sigma, St. Louis, MO, USA). Protein expression and morphological changes of the nucleus were examined under fluorescence microscopy.

Per2 small interfering (si) RNA (s16931; Life Technologies) and Bik siRNA (s1990; Life Technologies) were transfected into synovial fibroblasts (3.0 × 10³ cells) using Lipofectamine™ RNAiMAX (Life Technologies) for 48 h. Noncoding siRNA (4,390,843; Life Technologies) was also used as controls. After that, cells were cultured in serum-free DMEM with 10 nM of MTX for 24 h to measure cell viabilities using the Cell Counting Kit-8. The viabilities were represented relative to those of noncoding siRNA without MTX treatment.

In dynamite-plunger plots, values were expressed as the mean ± standard error of the mean (SEM). In boxplots, values were expressed with 10th, 25th, 50th (median), 75th, and 90th percentiles, and the single values were superimposed on the boxplots using black symbols.

For statistical analyses, a one-sample Kolmogorov–Smirnov test was used to test the normality, A one-sample t test was used to compare one single control value and others, a paired t test was used to compare differences between two experimental groups, the Tukey test was used to compare differences between more than three experimental groups, and Dunnett’s test was used to compare differences between the control and others. All statistical tests were two-sided and p < 0.05 was considered statistically significant.

The statistical analyses were performed by EZR, based on R and R commander [32].

We first examined the effect of MTX on viabilities of synovial fibroblasts. A lower concentration of MTX (1, 10 nM) significantly decreased cell viabilities, while cell viability was relatively weak with a higher concentration of MTX (100 nM, 1 μM) (Fig. 1). Since 10 or 100 nM was the optimum MTX concentration in human sera [33], experiments were subsequently conducted with MTX 10 nM treatment as an appropriate effective concentration and a concentration of 100 nM as a less-effective control.

We next determined the effect of MTX on expressions of circadian clock genes. Under treatment with MTX (10 nM for 24 h), the expression of Per2 messenger RNA (mRNA) was significantly increased, while expression of Bmal1, Clock, and Cry1 mRNA was not. In contrast, under treatment with MTX (100 nM for 24 h), the expression of Per2 did not show any significant differences, while the expression of Clock and Cry1 significantly decreased (Fig. 2a). In accordance with this, the expression of PER2 protein was significantly increased by 10 nM of MTX for 32 h, but not by 100 nM (Fig. 2b, c).

The circadian transcriptional factor PAR bZIP (Dbp, Tef, and Hlf) and E4bp4 genes regulate Per2 gene expression by binding to D-box on the promoter region [23‐28]. We therefore examined the effect of MTX on the mRNA expression of Dbp, Tef, Hlf, and E4bp4. Expression of Dbp, Tef, and Tef mRNA was significantly increased by 10 nM MTX treatment for 24 h, while E4bp4 mRNA expression was not. However, expression of Dbp and E4bp4 significantly decreased with 100 nM of MTX treatment (Fig. 3a).

Next, we examined the mRNA expression of the proapoptotic factor Bik, since the PAR bZIP-binding site also exists on the promoter region of Bik [29]. As shown in Fig. 3b, the expression of Bik mRNA was significantly increased by 10 nM of MTX treatment for 32 h.

Finally, we observed cytosolic release of CYTOCHROME C since Bik interacted with BCL-2 family proteins to induce mitochondria-related apoptosis. As a result, expression of CYTOCHROME C was increased by 10 nM of MTX treatment for 48 h, but not by 100 nM (Fig. 3c, d).

To examine the roles of PAR bZIP in the transcription of Per2 and Bik, we transfected luciferase reporters, constructed with or without D-box, to synovial fibroblasts. With 10 nM of MTX treatment for 24 h, D-box(+) Per2 promoter activity was significantly increased as compared to those of the control and D-box(−) Per2 promoter. By 100 nM of MTX treatment, D-box(−) Per2 promoter activity was significantly decreased as compared to that of the control. Bik promoter activity was significantly increased after 10 nM of MTX treatment, and was higher than that of D-box(−) Bik promoter (Fig. 4).

To certify the specific expression of Per2 and Bik on MTX-induced apoptotic synovial fibroblasts, we used fluorescent immunostaining to examine expressions of PER2, BIK, and CYTOCHROME C and, simultaneously, morphological changes of nucleus. As a result, nuclear blebbing, shrinkage, or fragmentation were observed indicating that MTX (10 nM for 24 h) could induce cell death. PER2 was overexpressed on apoptotic cells, but not on cells with normal appearance (Fig. 5a). BIK and CYTOCHROME C were also overexpressed specifically on the apoptotic cells (Fig. 5b).

To confirm whether PER2 and BIK were responsible for MTX-induced cell death, Per2 and Bik siRNA were transfected into synovial fibroblasts to examine cell viabilities. As shown in Fig. 6a, cytotoxicity of MTX was attenuated by silencing of Per2 and Bik, independently and synergistically.

When we examined the long-term expression of Per2 and Bik mRNA (Fig. 6b), they gradually increased from 8 h until 24 or 32 h after synchronized circadian oscillations by serum shock, consistent with our previous report that the expression of Per2 on synovial fibroblasts was increased after synchronization and peaked at 24 h [17]. Since these genes were responsible for MTX-induced cell death, synovial fibroblasts were separately treated by 10 nM of MTX at 0–8 h (the lower Per2/Bik expression period) and at 24–32 h (the higher Per2/Bik expression period), and viabilities of synovial fibroblasts were again measured. As shown in Fig. 6c, the reduction rate of cell viabilities was significantly increased by 24–32 h of MTX treatment compared with 0–8 h of treatment, depending on the expression levels of Per2 and Bik.