Increased sensitivity to apoptosis induced by methotrexate is mediated by JNK




Low-dose methotrexate (MTX) is an effective therapy for rheumatoid arthritis (RA), yet its mechanism of action is incompletely understood. The aim of this study was to explore the induction of apoptosis by MTX.


Flow cytometry was performed to assess changes in the levels of intracellular proteins, reactive oxygen species (ROS), and apoptosis. Quantitative polymerase chain reaction was performed to assess changes in the transcript levels of select target genes in response to MTX.


MTX did not directly induce apoptosis but rather “primed” cells for markedly increased sensitivity to apoptosis via either mitochondrial or death receptor pathways, by a JNK-dependent mechanism. Increased sensitivity to apoptosis was mediated, at least in part, by MTX-dependent production of ROS, JNK activation, and JNK-dependent induction of genes whose protein products promote apoptosis. Supplementation with tetrahydrobiopterin blocked these MTX-induced effects. Patients with RA who were receiving low-dose MTX therapy expressed elevated levels of the JNK target gene, jun.


Our results support a model whereby MTX inhibits reduction of dihydrobiopterin to tetrahydrobiopterin, resulting in increased production of ROS, increased JNK activity, and increased sensitivity to apoptosis. The finding of increased jun levels in patients with RA receiving low-dose MTX supports the notion that this pathway is activated by MTX in vivo and may contribute to the efficacy of MTX in inflammatory disease.

The folic acid antagonist methotrexate (MTX) was initially developed during the 1940s to inhibit dihydrofolate reductase (DHFR) in the treatment of malignancies (1–3). The clinical potential of MTX in treating rheumatoid arthritis (RA) was initially suggested by Gubner et al in 1951 (4, 5) after studying the effects of MTX in 6 patients in whom RA was diagnosed and was confirmed by further studies conducted during the 1980s (6–8). These studies had durations of 12–18 weeks, and dosages varied from 7.5 mg/week to 25 mg/week. MTX possessed antiinflammatory effects in patients with RA, as evidenced by improved function, global assessments, joint scores, and marked decreases in pain. Use of weekly low doses of MTX is not limited to RA therapy. Over the years, this treatment option has expanded to include additional inflammatory and autoimmune diseases (9–12).

In contrast to the doses used for the treatment of malignancy, much lower and more infrequent doses of MTX are used to treat inflammatory disease, and basic mechanisms may differ greatly from those targeting cancer. The exact mechanisms underlying the antiinflammatory actions of MTX remain inconclusive in spite of its widespread application (13). MTX, like natural folates, is polyglutamated once taken up by cells. MTX polyglutamates, found in red blood cells, neutrophils, and mononuclear cells after oral administration, are believed to represent the active form of the drug, and levels of MTX polyglutamates correlate with clinical efficacy in patients with RA (14, 15). MTX-stimulated synthesis of adenosine and its release by cells and subsequent activation of adenosine receptors may be one mediating factor contributing to the antiinflammatory actions of MTX (13, 16, 17). MTX also inhibits T cell activation, induces T cell apoptosis, and alters expression of T cell cytokines and adhesion molecules (18–21). Such actions may be partly mediated by the synthesis and release of adenosine. The action of MTX may also be dependent, in part, on its ability to stimulate production of reactive oxygen species (ROS) (22).

From these studies, the underlying mechanisms behind the induction of apoptosis by low concentrations of MTX are not immediately apparent. In general, cells undergo apoptosis via their ability to activate intrinsic (mitochondrial) or extrinsic (death receptor) pathways (23–29). For example, the stress-inducible p53 protein plays a central role in transducing DNA damage signals to cause cellular apoptosis. Death receptors such as Fas or the family of tumor necrosis factor receptors (TNFRs) transmit signals through adapter molecules such as FADD, TRADD, or Daxx. These adaptor proteins activate death caspases, causing apoptotic cell death. Both intrinsic and extrinsic apoptosis pathways are thought to involve JNK signaling (29).

Here, we sought to investigate induction of apoptosis by low concentrations of MTX in a transformed human T cell line. We observed that low concentrations of MTX do not directly induce apoptosis. Rather, MTX markedly increases the sensitivity of cells to apoptosis mediated via either death receptor or mitochondrial pathways, in part by increasing the expression of genes whose protein products play key roles in the induction of apoptosis. This alteration in the transcription profile of cells treated with MTX is dependent on induction and activation of JNK by ROS. Apoptosis sensitivity and ROS production are prevented by supplementation with tetrahydrobiopterin (BH4), suggesting that inhibiting the reduction of dihydrobiopterin (BH2) by DHFR initiates this pathway. Studies of patients with RA who are currently receiving MTX therapy support the notion that these pathways may be activated in vivo by MTX and contribute to the therapeutic efficacy of MTX.


Drugs and reagents.

MTX, adenosine, caffeine, theophylline, N-acetyl-L-cysteine (NAC), H2O2, and BI-78D3 were obtained from Sigma-Aldrich, and pepJIP1 was obtained from Enzo Life Sciences. Anti-Fas antibody was obtained from Medical and Biological Laboratories. Caspase 3 activity was determined by enzymatic assay (CaspACE Colorimetric Assay System; Promega). TaqMan Low Density Array 384-well plates were obtained from Applied Biosystems. Plasmids containing JNK1 (MAPK8)– and JNK2 (MAPK9)–dominant-negative (DN) mutants were obtained from Addgene.

Cell preparation and culture.

The Jurkat human T cell leukemia line was obtained from American Type Culture Collection. Peripheral blood mononuclear cells (PBMCs) were purified by Ficoll-Hypaque density-gradient centrifugation. Cells were cultured in RPMI 1640 medium (1 μg/ml folic acid, with 10% fetal calf serum, penicillin/streptomycin, and L-glutamine) at 37°C in an atmosphere of 5% CO2 in air. Cells were plated at 0.8 × 106 cells/ml in 5-ml cultures. Cell viability and numbers were determined microscopically after staining with trypan blue.

The concentrations of MTX used in the different experiments ranged from 0.01 μM to 1 μM, and the culture periods ranged from 24 hours to 72 hours of continuous exposure to MTX. Pharmacokinetic analysis indicated that ingestion of a 20-mg tablet of MTX yields plasma MTX concentrations of ∼0.5 μM after 1 hour and ∼0.1 μM after 10 hours (30).

Cell transfection.

Cells were transfected using the Cell Line Nucleofector Kit V (Amaxa). Briefly, 1 × 106 Jurkat T cells were suspended in 100 μl of Cell Line Nucleofector Solution V containing 2 μg of plasmid DNA and 2 μg of pmaxGFP vector. The cell suspension was transferred to the provided cuvette and nucleofected using program X-001 on the Amaxa Nucleofector apparatus. Cells were incubated at room temperature for 10 minutes and then transferred to prewarmed culture medium in 6-well plates.

Quantitative reverse transcription–polymerase chain reaction (RT-PCR).

Total RNA was purified from blood collected in PAXgene tubes according to manufacturer's instructions (Qiagen) or from cell cultures using TRI Reagent and quantified using a NanoDrop 1000 spectrophotometer. Five micrograms of total RNA was used for complementary DNA (cDNA) synthesis (SuperScript III First-Strand Synthesis Kit; Invitrogen) with oligo(dT) as the primer. Duplicate PCRs were performed in volumes of 25 μl containing 50 ng of cDNA, TaqMan Gene Expression Master Mix, and the TaqMan probe. GAPDH was used as a housekeeping gene and control. RT-PCR was performed using the Applied Biosystems 7300 Real-Time PCR System.

Western blotting.

Whole cell lysates were prepared in phosphate buffered saline (PBS) containing 1% Nonidet P40 (NP40; Igepal CA-630), 50 mm Tris HCl, 150 mm NaCl, 2 mm EDTA, and 0.1% sodium dodecyl sulfate (SDS), plus a cocktail of protease inhibitors (Roche) and sodium orthovanadate. Equal amounts of protein were resolved by SDS–polyacrylamide gel electrophoresis (PAGE) and transferred to polyvinylidene difluoride membranes. Membranes were blocked in 5% nonfat milk, 0.1% Tween 20 in PBS. Rabbit polyclonal antibodies to JNK1 (ab10664) and phosphorylated JNK1 (cross-reactive with p-JNK2; ab4821) were obtained from Abcam. The ECL Plus Chemiluminescent Kit (Applied Biosystems) was used to visualize protein bands.

Flow cytometry.

For apoptosis determinations, cells were labeled using the PE Annexin V Apoptosis Detection Kit I (BD PharMingen). For intracellular protein determinations, cells were fixed (paraformaldehyde), permeabilized (Triton X-100 and NP40), and labeled with primary antibodies for 24 hours at 0–4°C, followed by incubation with fluorescence-labeled secondary antibodies for 1 hour at 0–4°C. The following antibodies were used: primary antibodies, rabbit anti-JNK (sc-571; Santa Cruz Biotechnology), polyclonal rabbit anti–p-JNK (pT183/pY185) (catalog no. 558268; BD PharMingen), rabbit monoclonal antibodies to p45 up-regulated modulator of apoptosis (PUMA) (ab33906; Abcam), c-Jun (ChIP Grade, ab31419; Abcam), TRAILR-1 (DR4) (ab18362; Abcam), and c-Fos antibody (ab7963; Abcam). Fluorescein isothiocyanate–conjugated goat anti-rabbit immunoglobulin (catalog no. 554020; BD PharMingen) was used as the secondary antibody. Cells were analyzed using the 3-laser BD LSR II flow cytometer (BD Biosciences).

Patient populations.

The study group was composed of 36 control subjects who had no current chronic or acute infections and no family history of autoimmune diseases and 50 patients meeting the American College of Rheumatology 1987 revised criteria for the classification of RA (31); 28 of these patients were currently receiving MTX treatment, and the other 22 patients were not. No other exclusion or inclusion criteria were used, except for the ability to provide informed consent. The Committees for the Protection of Human Subjects of Vanderbilt University and University of Texas–Southwestern Medical Center approved these studies. The approximate female-to-male ratio in all study groups was 3:1. The ages of the subjects (36–58 years) and the racial distributions in all groups were similar. Current therapies were determined by questionnaire and confirmed by chart review. Patients being treated with MTX were receiving dosages of 15–25 mg/week.

Statistical analysis.

Statistical significance was determined by the unpaired t-test with the Welch correction. P values less than 0.05 were considered significant.


Apoptosis “priming” by MTX.

Various studies have demonstrated the ability of MTX to induce apoptosis or alter cell viability. In the current study, we cultured Jurkat T cells with MTX and monitored apoptosis by measuring the activity of caspase 3. Jurkat cells cultured with MTX, low concentrations of either H2O2 or anti-Fas for 24 hours, or combinations of MTX and H2O2 or MTX and anti-Fas exhibited minimal activation of caspase 3 relative to cells cultured with high concentrations of H2O2 (Figure 1A). We next cultured cells for 48 hours with MTX and then exposed them to either anti-Fas antibody or H2O2 for an additional 24 hours to activate death receptor or mitochondrial apoptosis pathways, respectively. Pretreatment of Jurkat cells with MTX at concentrations of 0.1 μM resulted in a marked increase in activity of caspase 3 following culture with either H2O2 or anti-Fas (Figure 1B).

Figure 1.

Methotrexate (MTX) primes Jurkat cells for increased sensitivity to apoptosis. A and B, Jurkat cells were cultured for 48 hours with MTX, H2O2, or 50 ng/ml anti-Fas (A) or were cultured for 48 hours with the indicated concentrations of MTX, followed by culture with H2O2 or anti-Fas for 24 hours (B), and the levels of caspase 3 were determined. C, JNK1–dominant-negative (DN), JNK2-DN, or empty-vector plasmids with a green fluorescent protein (GFP) plasmid were introduced by transient transfection. Cells were cultured for 48 hours with or without MTX 10−7M and treated with anti-Fas for an additional 6 hours. The percentage of annexin V–positive cells was determined by flow cytometry after gating on GFP-positive and GFP-negative cells. D and E, Jurkat cells were cultured for 48 hours with MTX (10−7M) in the presence or absence of BI-78D3 or pepJIP1 (D) or in the presence or absence of N-acetyl-L-cysteine (NAC) (E), followed by culture with anti-Fas for 6 hours. The percentage of annexin V–positive cells was determined by flow cytometry. Bars show the mean ± SD. ∗ = P < 0.05 versus untreated cells.

As a second measure of apoptosis, we used flow cytometry to investigate changes in annexin V labeling. We used JNK1-DN and JNK2-DN mutants to assess the relative contributions of JNK1 and JNK2 to increased apoptosis sensitivity. Jurkat cells, either untreated or cultured with MTX for 48 hours, exhibited low percentages of annexin V–positive cells. Treatment with H2O2 or anti-Fas only slightly increased the percentages of annexin V–positive cells. However, treatment of MTX-cultured cells with H2O2 or anti-Fas resulted in a marked increase in the percentage of annexin V–positive cells (Figure 1C). This increase in apoptosis (annexin V–positive cells) was slightly abrogated by the presence of either JNK1-DN or JNK2-DN mutants but was substantially abrogated by the presence of both JNK1-DN and JNK2-DN mutants.

We also tested the ability of the specific JNK inhibitors BI-78D3 and pepJIP1, which target JNK–JNK-interacting protein 1 interaction sites, to prevent apoptosis (32–34). Both inhibitors blocked MTX-dependent increases in sensitivity to apoptosis (Figure 1D). NAC, a free radical scavenger and reducing agent, also blocked MTX-dependent apoptosis (Figure 1E). Based on these results, we concluded that MTX primes cells for markedly increased sensitivity to apoptosis via death receptor and mitochondrial pathways, which is dependent on JNK1 and JNK2 enzymes. Additionally, ROS may contribute to increased apoptosis sensitivity.

We next sought to determine whether MTX treatment of Jurkat cells altered the expression levels of a panel of genes whose protein products are known to have activating or inhibitory effects on apoptosis. Jurkat cells were cultured with MTX for 48 hours prior to RNA isolation, cDNA synthesis, and analysis by quantitative PCR. Although we did not observe reduced expression of genes such as BCL2, whose protein products protect cells from apoptosis (see Supplementary Figure 1, which is available at the Arthritis & Rheumatism Web site at, we observed increased expression of several genes, BBC3 (PUMA, Bcl-2 binding component 3), BCL3 (B cell CLL/lymphoma 3), HRK (harakiri), LTB (lymphotoxin β), TNFRSF10A (cytotoxic TRAIL receptor, TRAILR-1), TNFSF10 (TRAIL, APO2L), TNFRSF1B (TNFR2), and TNFRSF25 (TRAMP, APO-3), whose protein products are known to play critical roles in promoting sensitivity to apoptosis (Figure 2A). We also investigated changes in expression of the prototypical JNK target genes, jun and fos. These genes also had increased expression levels following MTX treatment.

Figure 2.

MTX induces expression of genes whose protein products promote apoptosis. A, Jurkat cells were cultured with MTX for 48 hours, and transcript levels of genes whose protein products are known to influence apoptosis were determined. Results are expressed as the fold increase compared with untreated cells after normalization to GAPDH. B, JNK1-DN, JNK2-DN, or empty-vector plasmids with a GFP expression plasmid were introduced into Jurkat cells by nucleofection. Left and middle, C-Jun levels in untreated cells (black lines), MTX-treated cells (blue lines), and MTX-treated cells transfected with JNK1-DN (red lines; left) or MTX-treated cells transfected with JNK2-DN (red lines; middle) are shown. Right, Changes in the expression levels of c-Jun, Fos, PUMA, and TRAILR-1 in MTX-treated Jurkat cells were determined by flow cytometry and are expressed as the fold change relative to untreated cells. C, MTX-mediated increases in c-Jun and Fos protein levels were inhibited by the specific JNK inhibitors BI-78D3 and pepJIP1. Bars show the mean ± SD. ∗ = P < 0.05 versus untreated cells. Max = maximum (see Figure 1 for other definitions).

To assess changes in the protein levels of target genes and to assess the relative contributions of JNK1 and JNK2 to these changes, we measured intracellular protein levels by flow cytometry with specific antibodies and used JNK1- or JNK2-DN mutants to specifically inhibit JNK1 or JNK2. MTX treatment resulted in increased protein expression of the known JNK target, c-Jun, and this increase was inhibited by the JNK1-DN mutant but not the JNK2-DN mutant (Figure 2B). We also determined changes in TRAILR-1, c-Fos, and PUMA protein levels after MTX treatment in the presence of either JNK1- or JNK2-DN mutants, using flow cytometry. Induction of TRAILR-1 by MTX was also abrogated by the JNK1-DN mutant, whereas the induction of c-Fos and PUMA was relatively unaffected by the JNK1-DN mutant but was almost completely abrogated by the JNK2-DN mutant (Figure 2B). The JNK-specific inhibitors BI-78D3 and pepJIP1 also inhibited induction of c-Jun and c-Fos expression by MTX (Figure 2C). Taken together, these results provided evidence that MTX increased sensitivity to apoptosis, at least in part, by inducing expression of genes whose protein products promote apoptosis. The results observed with the JNK-DN mutants implicated both JNK1 and JNK2 as regulators of this gene family.

Induction of jun and fos by MTX.

Jun and fos are prototypical JNK target genes (35, 36). Thus, we further explored induction of these genes by MTX. We also examined the effects of adenosine, because MTX is known to induce adenosine production (13). MTX at concentrations of 0.1–1.0 μM stimulated a gradual increase in jun levels over the course of 24–72 hours (Figure 3A). In contrast, adenosine at concentrations of 0.01–100 μM failed to induce changes in jun expression levels over this time period (Figure 3B). This range of adenosine concentrations should be sufficient to activate known human adenosine receptors, which have 50% maximum response concentrations for adenosine of ∼0.3 μM to 30 μM (37–-39). We also determined the expression levels of jun family members, junB and junD. Treatment of Jurkat cells with MTX or adenosine did not change expression levels of these jun family members. Culture with MTX or adenosine did not affect cell numbers or viability, as determined microscopically. Thus, exposure to MTX, but not adenosine, directly increased jun expression levels in a uniform T cell population but did not change expression levels of the Jun family members, junb and jund.

Figure 3.

Induction of jun and fos expression by methotrexate (MTX) in Jurkat T cells. A–D, Jurkat T cells were cultured with the indicated concentrations of MTX (A and C) or adenosine (B and D). At the indicated times, cell cultures were harvested, total RNA was purified, and cDNA was synthesized. The relative expression of jun and fos was normalized to GAPDH levels and expressed as the fold induction relative to untreated cultures. E and F, Jurkat cells were also cultured with MTX and the adenosine receptor antagonists caffeine (20 μM) and/or theophylline (20 μM) for 48 hours. The expression levels of jun and fos were determined by polymerase chain reaction. Bars show the mean ± SD. ∗ = P < 0.05 versus unstimulated cultures.

Because the transcription factor activator protein 1 (AP-1) is a heterodimer composed of a Jun family member and a Fos family member (40), we determined the impact of MTX treatment on the levels of Fos family member transcripts. Similar to what was observed for jun, fos messenger RNA (mRNA) expression levels increased in Jurkat cells after stimulation by MTX (Figure 3C). In contrast, stimulation by adenosine did not alter fos mRNA expression (Figure 3D). We examined changes in the mRNA levels of additional fos family members, fosl1 and fosB. We did not detect changes in transcript levels of these 2 genes following exposure to MTX or adenosine. The kinetics of fos induction were different from the kinetics of jun induction following MTX stimulation. Fos mRNA levels reached a peak after 24–48 hours, depending on the concentration of MTX, and declined by 72 hours, whereas jun mRNA levels reached a peak at 72 hours. Taken together, these results demonstrated that MTX induced increased expression of both the jun and fos components of the AP-1 transcriptional complex in Jurkat cells. They also showed that MTX specifically induced jun and fos mRNA but not mRNA encoding other Jun and Fos family members.

As an alternate approach, we sought to determine whether the adenosine receptor antagonists with broad specificity, caffeine and/or theophylline, interfered with the induction of jun and fos mRNA. Caffeine, theophylline, or their combination, at pharmacologically active concentrations, did not interfere with the induction of jun and fos mRNA by MTX (Figures 3E and F). Taken together, these results do not provide evidence against the notion that the antiinflammatory effects of MTX may be mediated, in part, via stimulation of adenosine release and activation of adenosine receptors. Rather, the results presented here clearly demonstrated that adenosine alone is not sufficient to increase the expression of fos and jun mRNA in T cells, and that adenosine receptor antagonists with broad specificity do not interfere with the induction of fos and jun mRNA by MTX.

Blockade of MTX-induced apoptosis by free radical scavengers and tetrahydrobiopterin (BH4).

MTX, at concentrations similar to those inducing jun and fos expression, leads to the production of ROS and priming for apoptosis (22). Furthermore, ROS is known to activate JNK, leading to increased jun expression (29, 41–43). We cultured Jurkat cells with MTX and the free radical scavenger NAC and examined changes in apoptosis sensitivity and jun expression levels. The addition of NAC inhibited MTX-mediated changes in apoptosis sensitivity (Figure 4A) and MTX-mediated increases in jun expression in Jurkat cells but did not alter baseline jun levels (Figure 4B). Therefore, we attempted to determine whether MTX increased the production of ROS, using flow cytometry after labeling cells with CM-H2DCFDA. MTX caused a marked increase in ROS production (Figure 4C) that was effectively reduced by NAC (Figure 4D). We concluded that the JNK-dependent increases in apoptosis sensitivity and jun transcript levels induced by MTX were mediated, at least in part, through ROS production.

Figure 4.

Free radical scavengers or tetrahydrobiopterin (BH4) supplementation blocks methotrexate (MTX)–mediated increases in apoptosis sensitivity. A and B, Jurkat cells were cultured for 48 hours with MTX and/or the free radical scavenger N-acetyl-L-cysteine (NAC), stimulated with anti-Fas for 5 hours, and the percentage of annexin V–positive cells was determined (A) or the cultures were harvested and processed, and the jun transcript levels were determined (B). C and D, Jurkat cells were treated for 48 hours with MTX alone (C) or with MTX in the presence or absence of NAC, and reactive oxygen species (ROS) production was determined by labeling cells with CM-H2DCFDA. E and F, Jurkat cells were cultured for 48 hours with MTX (0.1 μM) with or without BH4 supplementation (30 μM). The percentage of annexin V–positive cells after anti-Fas treatment was determined by flow cytometry (E), and ROS levels were determined after loading Jurkat cells with CM-H2DCFDA. Bars show the mean ± SD. ∗ = P < 0.05 versus unstimulated cultures; ∗∗ = P < 0.05 versus MTX- and NAC-treated cultures (A and D) or MTX- and BH4-treated cultures (E and F).

MTX inhibits the reduction of dihydrofolate to tetrahydrofolate and the reduction of dihydrobiopterin (BH2) to BH4 catalyzed by DHFR (44–47). BH4 is a necessary cofactor for nitric oxide synthases (NOS) including endothelial NOS (eNOS), which is expressed in T lymphocytes (48, 49). The loss of BH4 uncouples eNOS from NO synthesis, leading to excess production of ROS. Thus, we tested whether supplementation with BH4 could reverse MTX-mediated increased apoptosis and ROS production. We observed that BH4 supplementation effectively reduced MTX-dependent apoptosis (Figure 4E), and MTX-dependent increased ROS production (Figure 4F). These results are consistent with a model whereby inhibition of DHFR by MTX reduces intracellular levels of BH4, resulting in NOS-catalyzed ROS overproduction, leading to activation of JNK and increased apoptosis sensitivity.

Induction of JNK expression and activity by MTX.

Inhibition of MTX-mediated increased jun expression, priming for apoptosis, and expression of genes whose protein products are proapoptotic, by JNK inhibitors and by JNK1- and JNK2-DN mutants, suggest that MTX may increase JNK expression or activity. JNK is a member of the MAP kinase family and is activated by phosphorylation on threonine and tyrosine residues by JNK kinases, also termed MAP kinase kinases, in response to an array of stimuli (50, 51). There are 3 JNK genes that produce multiple mRNA and protein isoforms. We used Western blotting to determine whether stimulation by MTX changed JNK1 total protein levels and phosphorylated JNK1/2 levels. Jurkat cells were cultured with different concentrations of MTX for 2 days. Whole cell lysates were prepared and analyzed by SDS-PAGE and Western blotting with antibodies specific for JNK1 total protein or antibodies specific for the phosphorylated forms of JNK1 and JNK2. Culturing cells with MTX resulted in an increase in total JNK1 protein levels and levels of phosphorylated JNK1/2 (Figure 5A).

Figure 5.

Methotrexate (MTX) induces increased levels of JNK and p-JNK. A and B, Jurkat T cells (A) or human peripheral blood mononuclear cells (PBMCs) (B) were treated with the indicated concentrations of MTX. After 48 hours, cells were harvested, whole cell lysates were prepared, and JNK1 and p-JNK levels were determined by Western blotting with specific antibodies for either JNK1 or p-JNK1/2. C and D, Jurkat cells (C) and PBMCs (D) were treated with or without MTX, labeled with specific antibodies for p-JNK or JNK, and analyzed by flow cytometry. Bars show the mean ± SD mean fluorescence intensity (MFI). ∗ = P < 0.05 versus untreated cells.

We also cultured human PBMCs with MTX for 2 days and analyzed whole cell lysates by SDS-PAGE and Western blotting. Similar to what was observed in Jurkat cells, culturing PBMCs with MTX resulted in concentration-dependent increases in total JNK1 protein and phosphorylated JNK1/2 (Figure 5B). We also examined changes in total phosphorylated JNK and total JNK levels by flow cytometry. These experiments confirmed changes in phosphorylated JNK and total JNK levels in Jurkat cells (Figure 5C) and PBMCs (Figure 5D) following MTX treatment. We concluded that MTX induced an increase in the cellular levels of phosphorylated JNK protein and total JNK protein in both Jurkat cells and human PBMCs.

Jun expression in RA patients receiving MTX.

Given that MTX increased p-JNK levels in both Jurkat cells and PBMCs and expression of the known JNK target gene, jun, we sought evidence for similar effects of MTX in vivo. To test this hypothesis, we determined expression levels of jun in control subjects relative to patients with RA who were or were not currently receiving low-dose weekly MTX therapy. Blood samples were collected in PAXgene tubes. The expression levels of target genes were determined relative to GAPDH. We observed increased expression of jun in the blood of patients with RA who were currently receiving MTX therapy relative to patients who were not receiving MTX therapy and control subjects (Figure 6, far left). In contrast, expression levels of 3 other genes, GNB5, TXK, and NRAS, were reduced in patients with RA compared with control subjects, independent of MTX therapy (or other therapies). We concluded that low-dose MTX therapy increased expression levels of jun in whole blood, while the expression levels of other target genes that are reduced in patients with RA did not change in response to MTX therapy.

Figure 6.

Expression levels of the JNK target gene jun in vivo in patients with rheumatoid arthritis (RA) as a function of current methotrexate (MTX) therapy. Blood was obtained in PAXgene tubes from control subjects (CTRL; n = 36), patients with RA not receiving MTX (RA −MTX; n = 28), and patients with RA receiving MTX (RA +MTX; n = 22). Bars show the mean ± SD gene expression levels relative to GAPDH. P values are versus control. NS = not significant.


MTX was developed in the 1940s as an antagonist of DHFR, and since that time, the therapeutic efficacy of MTX has been attributed to inhibiting the reduction of dihydrofolate to tetrahydrofolate, resulting in inhibition of purine synthesis necessary for DNA and RNA synthesis (7.8). However, folate supplementation does not interfere with the antiinflammatory effects of MTX, suggesting additional mechanisms of action of low-dose MTX for the treatment of this class of disease (52). In the 1990s, it was proposed that production of adenosine, as a by-product of DHFR inhibition, simulates adenosine receptors, thus contributing to the antiinflammatory effects of MTX (13, 16). An additional possible mechanism by which MTX may exert its antiinflammatory effects is by induction of apoptosis of inflammatory cells, such as lymphocytes; however, it is not immediately apparent how the known biochemical pathways activated or inhibited by MTX may induce apoptosis.

Our new model suggests that MTX inhibits DHFR-catalyzed BH2 reduction to BH4. The downstream biochemical effector pathways stimulated by loss of BH4 lead to altered cell sensitivity to apoptosis, thus shifting the decades-long focus of MTX action away from inhibition of dihydrofolate reduction to inhibition of BH2 reduction to BH4 (see Supplementary Figure 2, which is available at the Arthritis & Rheumatism Web site at Our results support a model whereby submicromolar concentrations of MTX stimulate increased sensitivity to apoptosis via the death receptor and mitochondrial pathways. Increased apoptosis sensitivity depends on increased expression and activity of JNK and subsequent increased expression of JNK target genes. JNK target genes include those whose protein products increase the sensitivity of cells to apoptosis. JNK activation is mediated in part by ROS, which we propose is induced by MTX-dependent depletion of BH4 levels, uncoupling eNOS from NO synthesis, resulting in overproduction of ROS. Specific JNK inhibitors prevent both MTX-mediated effects on gene and protein induction and changes in sensitivity to apoptosis. Both JNK1-DN and JNK2-DN mutants interfere with these MTX-mediated effects. Thus, both JNK enzymes contribute to MTX-mediated changes in the transcriptional program in Jurkat cells and to alterations in sensitivity to apoptosis by regulating expression of different target genes. Furthermore, patients with RA receiving low-dose MTX weekly exhibit increased expression of jun, the prototypical JNK target gene, thus supporting the notion that this pathway is activated by MTX in vivo and may contribute to the efficacy of MTX in inflammatory disease.

It is well established that JNK enzymes play both positive and negative roles in both death receptor and mitochondrial pathways of apoptosis (35). These roles can be mediated via JNK translocation to the nucleus and phosphorylation of transcription factors. C-Jun is the prototypical transcription factor phosphorylated by JNK, but JNK also phosphorylates additional Jun family members, activating transcription factor (ATF) and Elk family members, p53, c-Myc, and others (53). JNK can also translocate to mitochondria and regulate the activity of proteins involved in apoptosis, such as Bad, Bim, and 14-3-3, via phosphorylation. In large part, these studies have examined the roles of JNK in stimulating or preventing apoptosis in response to either external or internal stimuli. Our results support a new role for JNK in regulating apoptosis.

The treatment of Jurkat T cells with MTX induces increased expression and activity of JNK, increased expression of jun mRNA, and increased expression of a number of genes whose protein products are proapoptotic. However, MTX does not directly stimulate apoptosis under these conditions. Rather, MTX-treated Jurkat cells are primed to exhibit markedly increased sensitivity to apoptosis when exposed to stimuli activating the death receptor or mitochondrial pathways via a JNK-mediated pathway. Further experimentation will be required to address whether JNK also primes Jurkat cells for apoptosis by modulating the activity of proapoptotic or antiapoptotic proteins via direct phosphorylation.

In many cell types, activation of JNK is mediated by its phosphorylation in response to stimuli, including cellular stress and inflammatory cytokines. In contrast, T lymphocytes exhibit a second level of regulation. Resting or naive T lymphocytes express low quantities of JNK enzymes (54). Activation via T cell receptor ligation leads to markedly increased expression of JNK genes and protein, whereas JNK phosphorylation requires CD28-mediated costimulatory signals. Our results demonstrated that MTX treatment of Jurkat cells or human PBMCs (70% T lymphocytes) leads to increased levels of p-JNK and total JNK protein. Whether MTX treatment of non-T cells or non-lymphoid cells induces increased levels of p-JNK and total JNK protein remains to be determined.

The potential of JNK inhibitors as therapeutic agents for inflammatory disease has also attracted considerable interest (35, 40, 55–59). Modulation of JNK enzymes, either by genetic means or with small molecule inhibitors, interferes with a number of processes linked to inflammatory disease. For example, many genes linked to inflammation, such as those encoding TNFα, matrix metalloproteinases, and adhesion molecules, possess AP-1 response elements in their promoters and are regulated by JNK enzymes through activation of AP-1 and ATF-2 transcription factors. Our results suggest a second paradigm shift. Our model suggests that correctly targeted activation of JNK, rather than its inhibition, represents a therapeutic goal for the treatment of inflammatory disease, especially RA. In lymphocytes, increased JNK activity and subsequent increased sensitivity to apoptosis may also have therapeutic benefits in inflammatory diseases that are dependent on T lymphocyte function by eliminating self-reactive T cells. For example, activated T cells undergo apoptosis if they are exposed to secondary T cell receptor ligation in a process termed activation-induced cell death (23).

We speculate that MTX therapy may prime self-reactive T cells to undergo increased activation-induced cell death in response to secondary exposure to self antigen and thus produce its therapeutic benefit. Future studies will be required to determine whether this novel MTX-induced pathway can be exploited for therapeutic benefit.


All authors were involved in drafting the article or revising it critically for important intellectual content, and all authors approved the final version to be published. Dr. T. M. Aune had full access to all of the data in the study and takes responsibility for the integrity of the data and the accuracy of the data analysis.

Study conception and design. Spurlock, Z. T. Aune, Tossberg, Crooke, Olsen, T. M. Aune.

Acquisition of data. Spurlock, Z. T. Aune, Tossberg, Collins, J. P. Aune, Huston.

Analysis and interpretation of data. Spurlock, Z. T. Aune, Tossberg, Collins, J. P. Aune, T. M. Aune.