|Home | About | Journals | Submit | Contact Us | Français|
Dr. Reiss 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 design. Reiss, Chan.
Acquisition of data. Reiss, Edelman, Rao, Zhang, Anwar.
Analysis and interpretation of data. Reiss, Chan, Fernandez, Carsons.
Manuscript preparation. Reiss, Chan, Cronstein.
Statistical analysis. Reiss, Chan, Fernandez, Anwar.
To determine whether MTX can overcome the atherogenic effect of COX-2 inhibitors and IFN-γ, both of which suppress cholesterol efflux protein levels and promote foam cell transformation in THP-1 human monocytes/macrophages.
Message and protein level of the reverse cholesterol transport (RCT) proteins cholesterol 27-hydroxylase and ABCA1 in THP-1 cells were evaluated by real-time polymerase chain reaction and immunoblot, respectively. Expression was evaluated in cells incubated in the presence or absence of the COX-2 inhibitor NS398 or IFN-γ with/without MTX. Foam cell transformation of lipid-loaded THP-1 macrophages was detected with oil red O staining and light microscopy.
MTX increased 27-hydroxylase message and completely blocked NS398-induced downregulation of 27-hydroxylase (112.8±13.1% for NS398+MTX versus 71.1±4.3% for NS398 alone, with untreated as 100%, n=3, p<0.01). MTX also negated COX-2 inhibitor-mediated downregulation of ABCA1. Reversal of inhibitory effects on 27-hydroxylase and ABCA1 in the presence of MTX were blocked by the adenosine A2A receptor-specific antagonist ZM-241385. MTX also prevented NS398 and IFN-γ from increasing transformation of lipid-loaded THP-1 macrophages into foam cells.
This study provides evidence supporting the atheroprotective effect of MTX. Through adenosine A2A receptor activation, MTX promotes RCT and limits foam cell formation in THP-1 macrophages. This is the first evidence that any commonly used medication can increase expression of anti-atherogenic RCT proteins and can counteract the effects of COX-2 inhibition. Our results suggest that one mechanism by which MTX protects against cardiovascular mortality in RA patients is through facilitation of cholesterol outflow from cells of the artery wall.
Methotrexate (MTX) has a long history of use in the treatment of various immunologic diseases, and has been used to treat rheumatoid arthritis and psoriasis since the 1960s (1-3). MTX was originally developed as a folate antagonist that inhibits dihydrofolate reductase activity and is used in high doses for the treatment of malignancies such as leukemia. Because of its anti-inflammatory and immunosuppressive effects, the drug is also efficacious in the treatment of diseases such as asthma, systemic lupus erythematosus, Crohn's disease, myositis, and vasculitis (4-8).
A number of clinical studies show that MTX not only inhibits inflammatory cell proliferation through action on dihydrofolate reductase but also inhibits the conversion of 5-aminoimidazole-4-carboxamide ribonucleotide to 5-formyl-5-aminoimidazole-4-carboxamide ribonucleotide (AICAR), thus increasing intracellular and extracellular levels of adenosine-5′-phosphate and adenosine (9, 10). Adenosine is a nucleoside that acts as a signaling molecule by triggering activation of adenosine receptors. These receptors are expressed on the surface of a wide variety of cells and are implicated in cellular protection against ischemia-reperfusion injury (11) and anoxia (12).
The risk of cardiovascular disease (CVD) is increased in patients with rheumatoid arthritis (RA) (13-15) and this elevated risk is not explained by traditional risk factors alone (16). Suggested explanations involve the inflammatory response that characterizes active RA and adverse effects of glucocorticoid therapy or other medications. Previous studies suggest that MTX has a beneficial effect on cardiovascular mortality that is not observed with other antirheumatic drugs (17, 18).
We previously reported that immune reactants interfere with cellular defense against cholesterol overload by diminishing expression of two proteins responsible for reverse transport of cholesterol out of the cell to the circulation for ultimate excretion: cholesterol 27-hydroxylase and ATP binding cassette transporter A1 (ABCA1) (19, 20). The atherosclerosis-promoting cytokine IFN-γ decreased 27-hydroxylase and ABCA1 message and protein expression in THP-1 human monocytes/macrophages. Further, IFN-γ-treated THP-1 macrophages exposed to acLDL formed foam cells more rapidly and in greater proportion than untreated control cells. Most recently, we have shown that inhibition of cyclooxygenase (COX) in THP-1 monocytes/macrophages acts in a pro-atherogenic manner by dose-dependently decreasing 27-hydroxylase and ABCA1 (21). THP-1 macrophages show a significant increase in foam cell transformation in the presence of the COX-2 selective inhibitor NS398 compared to control (21). This work suggests that compromise of reverse cholesterol transport may contribute to the known increase in cardiovascular risk in patients treated with COX-2 inhibitors (21, 22).
Our laboratory also employed a number of approaches to enhance expression of 27-hydroxylase and ABCA1. We discovered that these proteins can be upregulated via activation of the adenosine A2A receptor with specific agonists including CGS-21680 and MRE-0094. Ligation of this receptor also inhibits macrophage foam cell transformation under cholesterol loading conditions (20). Since MTX is known to affect both adenosine release and cardiovascular risk, we investigated whether MTX modulates cholesterol metabolism and vulnerability to foam cell formation. We report here that MTX treatment counteracts propensity toward cholesterol overload in THP-1 monocytes/macrophages exposed to IFN-γ or selective COX-2 inhibition. This data supports the hypothesis that MTX provides protection from atherosclerotic cardiovascular disease (ASCVD) by increasing expression of anti-atherogenic molecules involved in cholesterol efflux, likely via a pathway involving adenosine release.
THP-1 monocytes were obtained from ATCC (Manassas, VA).
Oil red O and OptiPrep Density Gradient Media were purchased from Sigma-Aldrich (St. Louis. MO).
Trizol reagent was purchased from Invitrogen (Grand Island, NY).
All reverse transcription- Polymerase chain reaction (RT-PCR) reagents were purchased from Applied Biosystems (Chicago, IL).
Recombinant human IFN-γ was purchased from R&D Systems (Minneapolis, MN).
NS-398 was purchased from RBI-Sigma (Natick, MA).
MTX was purchased from Bedford Laboratories (Bedford Ohio).
Acetylated LDL was purchased from Intracel (Issaquah, Washington).
Anti-cholesterol 27-hydroxylase antibody is an affinity-purified rabbit polyclonal anti-peptide antibody raised against residues 15-28 of the cholesterol 27-hydoxylase protein (23).
THP-1 monocytes were grown at 37°C in a 5% CO2 atmosphere to a density of 106 cells/ml. Growth medium for THP-1 cells was RPMI 1640 (GIBCO BRL, Grand Island, NY) supplemented with 10% Fetal Bovine Serum (FBS) (GIBCO BRL), 50 units/ml penicillin, and 50 units/ml streptomycin. To facilitate differentiation into macrophages, THP-1 monocytes (106 cells/ml) in 12 well plates were treated with 100nM PMA (Sigma) for 4 days at 37°C.
Blood from healthy donors was collected in EDTA treated tubes, pooled and kept at 4°C. The pooled blood was adjusted to a density of 1.120 g/ml with the addition of OptiPrep Density Gradient Media (Sigma) according to the manufacturer's instructions. The blood was then overlaid with a 1.074 g/ml density solution composed of complete RPMI containing 10% FBS and OptiPrep media. A layer of complete RPMI containing 10% FBS was then overlaid on top to prevent monocytes from sticking to the plastic tube. The blood was centrifuged at 750g for 30 minutes at 4°C. After centrifugation, the monocyte interphase was collected from between the 1.074 g/ml and RPMI layer. The collected cells were diluted with 2 volumes of complete RPMI and harvested by centrifugation. The pellet was re-suspended in complete RPMI. The monocytes were counted by hemocytometer and plated at a density of 2×106 cells/well in a 6-well plate.
When THP-1 cells had reached 106 cells/ml, media was aspirated and cells were rinsed twice with Dulbecco's Phosphate Buffered Saline (DPBS) without calcium and magnesium. The monocytes were then incubated for 24-48 hr in six well plates, (37°C, 5% CO2) under the following conditions: a) RPMI control; b) RPMI containing 5μM MTX c) RPMI containing NS398 (50 μM); d) RPMI containing NS398 (50 μM) and MTX (increasing doses of 0.1μM, 0.5μM and 5μM); e) RPMI containing IFN-γ (500 U/ml); f) RPMI containing IFN-γ (500 U/ml) and 5μM MTX.
THP-1 macrophages were exposed to the following conditions: a) RPMI control; b) RPMI containing ZM-241385 (10μM); c) RPMI containing MTX (5μM); d) RPMI containing IFN-γ (500 U/ml); e) RPMI containing IFN-γ (500 U/ml) and 5μM MTX; f) RPMI containing ZM-241385 (10μM) and MTX (5μM); g) RPMI containing IFN-γ (500 U/ml), ZM-241385 (10μM) and MTX (5μM); h) RPMI containing NS398 (50μM), ZM-241385 (10μM) and MTX (5μM).
Immediately after the incubation period, the cells were collected and centrifuged at 1500 RPM at room temperature, media was aspirated and cell protein and RNA were isolated.
PBMC were incubated for 18 h in RPMI with 10% FBS with and without the addition of MTX at a concentration of 5μM. Cells were collected and RNA isolated.
RNA was isolated using 1ml Trizol reagent per 106 cells and dissolved in nuclease-free water. The quantity of total RNA from each condition was measured by absorption at 260 and 280 wavelengths using quartz cuvettes by ultraviolet spectrophotometry (Hitachi U2010 spectrophotometer).
27-hydroxylase and ABCA1 mRNA were quantitated by real-time PCR. cDNA was copied from 5 μg of total RNA using M-MLV reverse transcriptase primed with oligo dT. Equal amounts of cDNA were taken from each RT reaction mixture for PCR amplification using cholesterol 27-hydroxylase-specific primers or ABCA1 specific primers as well as glyceraldehyde-3-phosphate dehydrogenase (GAPDH) control primers. The cholesterol 27-hydroxylase-specific primers span a 311 base-pair sequence encompassing nucleotides 491-802 of the human cholesterol 27-hydroxylase cDNA (24). ABCA1 primers yield a 234 BP amplified fragment (21). Real-time PCR analysis was performed using the SYBR Green PCR Reagents Kit (Applied Biosystems) with a Stratagene MX3005P QPCR System.
PCR was performed using techniques standardized in our laboratory. Each PCR reaction contained 2.5 μl of the 10x fluorescent green buffer, 3 μl of 25 mM MgCl2, 2 μl dNTP mix (2500 μM dCTP, 2500 μM dGTP, 2500 μM dATP, and 5000 μM dUTP), 0.15 μl polymerase (5 U/μl; AmpliTaq Gold; Applied Biosystems), 0.25 μl uracil-N-glycosylase (1 U/μl UNG; AmpErase; Applied Biosystems), 0.5 μl of the forward and reverse primers (10μM concentration), 4 μl cDNA, and water to a final volume of 25 μl. The thermal cycling parameters were as follows: 5 min at 95°C to activate the polymerase (AmpliTaq Gold; Applied Biosystems), followed by 45 cycles of 30 seconds at 95°C and 45 seconds at 58°C then 45 seconds at 72°C. Each reaction was done in triplicate. The amounts of PCR products were estimated, using software provided by the manufacturer (Stratagene). After completion of PCR cycles, the reactions were heat denatured over a 35°C temperature gradient from 60°C to 95°C. To correct for differences in cDNA load among samples, the target PCRs were normalized to a reference PCR involving the endogenous housekeeping genes glyceraldehyde-3-phosphate dehydrogenase (GAPDH) and β-actin. Nontemplate controls were included for each primer pair to check for significant levels of any contaminants. Fluorescence emission spectra were monitored and analyzed. PCR products were measured by the threshold cycles (CT), at which specific fluorescence becomes detectable. The CT was used for kinetic analysis and was proportional to the initial number of target quantity copies in the sample. A melting curve analysis was performed to assess the specificity of the amplified PCR products. The quantity of the samples was calculated after the CTs of the serial dilutions were compared with a control. QRT-PCR standards were prepared by making 1:10 serial dilutions of a purified PCR product.
Total cell lysates were prepared for Western immunoblotting using RIPA lysis buffer (98% PBS, 1% Igepal CA-630, 0.5% sodium deoxycholate, 0.1% sodium dodecyl sulfate [SDS]). 100μl of RIPA lysis buffer and 10μl of protease inhibitor cocktail (Sigma) were added to the cell pellet from each condition and incubated on ice for 35 min with vortexing every 5min. Supernatants were collected after centrifuging at 10,000g at 4°C for 10 min using an Eppendorf 5415C centrifuge. The quantity of protein in each supernatant was measured by absorption at 560nm using a Hitachi U2010 spectrophotometer.
Total cell lysate was used for Western blots. Protein samples (20μg/lane) were boiled for 5 min, loaded onto a 10% polyacrylamide gel, electrophoresed for 1.5 hr at 100V then transferred to a nitrocellulose membrane in a semi-dry transblot apparatus for 1 hr at 100V. The nitrocellulose membrane was blocked for 4 hr at 4°C in blocking solution (3% nonfat dry milk dissolved in 1×Tween20-tris-buffered saline [TTBS]) then immersed in a 1:300 dilution of primary antibody (18.7 μg/ml) in blocking solution overnight at 4°C. The primary antibody is an affinity-purified rabbit polyclonal anti-peptide antibody raised against residues 15-28 of the cholesterol 27-hydroxylase protein (23). The following day, the membrane was washed 5 times in TTBS for 5mins per wash then incubated at room temperature in a 1:3000 dilution of ECL donkey anti-rabbit IgG Horseradish peroxidase-linked species-specific whole antibody (Amersham Biosciences, product Code NA934). The five washes in TTBS were repeated, then the immunoreactive protein was detected using ECL western blotting detection reagent (Amersham Biosciences, Cat No RPN2106) and film development in SRX-101A (Konica Minolta).
As control, on the same transferred membrane, beta-actin was detected using mouse anti-beta-actin (diluted in 1:1000, from abCam, product Code: ab6276) and ECL sheep anti-mouse-IgG Horseradish peroxidase-linked species-specific whole antibody (diluted in 1:2000, from Amersham Biosciences, product Code NA931) and all other similar steps as above.
For ABCA1 detection, macrophage cell lysates were electrophoresed for 1.5 hr at 100V (10% polyacrylamide gel), then transferred to a nitrocellulose membrane. The membrane was blocked for 4 hr at 4°C in blocking solution then incubated overnight at 4°C in a 1:200 dilution of rabbit anti- ABCA1antibody (Santa Cruz Biotechnology). The following day, the membrane was washed 5 times in TTBS for 5mins per wash then incubated at room temperature in a 1:5000 dilution of ECL donkey anti-rabbit IgG Horseradish peroxidase-linked species-specific whole antibody. Development proceeded as described above for the 27-hydroxylase antibody.
THP-1 differentiated macrophages were washed three times with phosphate-buffered saline (PBS) and further incubated in RPMI (37°C, 5% CO2) for 48 hr under the following five conditions: a) acetylated LDL (50 μg/ml); b) acetylated LDL (50μg/ml) and IFN-γ (500U/ml); c) acetylated LDL (50μg/ml) and IFN-γ neutralizing antibody (1.2μg/ml); d) acetylated LDL (50μg/ml), IFN-γ (500U/ml) and IFN-γ neutralizing antibody (1.2μg/ml); e) acetylated LDL (50μg/ml) and IFN-γ receptor antibody (125ng/ml); f) acetylated LDL (50μg/ml), IFN-γ receptor antibody (125ng/ml) and IFN-γ (500U/ml).
Immediately following incubation, media was aspirated and cells were fixed in the same 12 well plates used for incubation, with 4% paraformaldehyde in water, for 2-4 min. Cells were stained with 0.2% Oil-Red-O in methanol for 1-3 min. Cells were observed via light microscope (Axiovert 25-Zeiss) with 100X magnification and then photographed using a Kodak DC 290 Zoom Digital Camera. The number of foam cells formed in each condition were calculated manually and presented as percentage foam cell formation.
Statistical analysis were performed using GraphPad version 4.02 (GraphPad, San Diego, CA). All data were analyzed by one-way ANOVA and pairwise multiple comparisions were made between control and treatment conditions using Bonferroni's method.
MTX (5μM, 18 hr) increased 27-hydroxylase mRNA expression (113.9±6.4%) and completely blocked NS398-induced downregulation of 27-hydroxylase message (112.8±13.1% for NS398+MTX versus 71.1±4.3% for NS398 alone, with untreated as 100%, n=3, p<0.01) (Figure 1A). MTX (5μM, 18 hr) also abrogated IFN-γ-induced downregulation of 27-hydroxylase message (86± 9.6% of control for IFN-γ+MTX versus 45± 6.0% for IFN-γ alone, with untreated as 100%, n=3, p=0.02) (Figure 1B).
To make certain that our THP-1 model accurately represents primary human monocyte behavior, we isolated PBMC from healthy human donors and incubated them for 18h in RPMI 1640 media in the presence and absence of MTX (5μM). MTX exposure resulted in a nearly fourfold increase in 27-hydroxylase mRNA (3.63±0.46 for MTX with untreated as 1.00, n=3, p=0.004) as assessed by QRT-PCR (Figure 1C).
MTX was also effective in blocking COX-2 inhibitor-mediated downregulation of ABCA1 message in THP-1 monocytes. Adenosine A2A receptor blockade with ZM-241385 abolished the ability of MTX to counter COX-2 inhibitor effects on both 27-hydroxylase (Figure 3A).and ABCA1 (Figure 3B).
Similarly, downregulation of 27-hydroxylase and ABCA1 by IFN-γ, which we showed previously in THP-1 monocytes (19), was also demonstrated in THP-1 macrophages and was prevented by MTX and this effect of MTX was negated by ZM-241385 (Figure 4A-C). THP-1 macrophages exposed to MTX alone exhibited a substantial increase in expression of both 27-hydroxylase and ABCA1 relative to untreated THP-1 macrophages (Figure 4A-C).
Addition of the adenosine A2A receptor agonist CGS-21680 to THP-1 monocytes exposed to NS398 overcame the reduction in 27-hydroxylase expression. This is demonstrated by immunoblot (Figure 5A) and QRT-PCR (Figure 5B). Addition of CGS-21680 to NS398-treated THP-1 cells gave rise to a 184% increase in 27-hydroxylase mRNA (167.2±8.57% in CGS+NS398 vs. 58.9±2.3% in NS398 alone, n=3, p<0.001) (Figure 5B).
Acetylated LDL-treated THP-1 macrophages showed a significant decrease in foam cell transformation in the presence of MTX compared to control (29.7±2.0% vs. 39.3±5.0%, p<0.001) (Figure 6A). NS398 treatment resulted in 72.7±4.9% foam cells while combined NS398+MTX resulted in only 36.3±3.2% foam cells, (n=3, p<0.001) (Figure 6B). IFN-γ treatment prior to cholesterol loading with acetylated LDL resulted in 71.0±5.0% foam cells while IFN-γ+MTX resulted in only 46.0±7.2% foam cells,(n=3, p<0.001) (Figure 6C). Preincubation of THP-1 macrophages with the selective A2A receptor antagonist (ZM-241385) prior to MTX treatment ablated the anti-atherogenic effect of MTX and resulted in a significant increase in foam cells (62.1 ±1.5%) (Figure 6D).
MTX modulates the expression of numerous inflammatory cytokines. Methotrexate has been successfully used in the treatment of many immune- or inflammatory-mediated diseases (eg, rheumatoid arthritis) by causing generalized immunomodulation. Several studies have demonstrated that binding of adenosine to the A2 receptor inhibits lymphocyte proliferation and production of TNF, IL-8, and IL-12, as well as increases secretion of IL-10 (25). Given the important role of inflammatory mediators in the pathogenesis of ASCVD, therapeutic modulation targeting inflammatory mediators might be a new and promising strategy for treating ASCVD.
Prior studies demonstrate that adenosine A2A receptor ligation can help prevent atherosclerosis in an experimental intimal injury model (26). In this model the recruitment of inflammatory cells to the intima following endothelial injury was markedly reduced. Our results suggest that MTX, which increases adenosine levels, may have a similar effect. Moreover, our results further suggest that MTX, and adenosine, might reduce atherosclerosis not only by reducing inflammation in the vessel wall but also by stimulating reverse cholesterol transport.
The concentration of MTX used was in the range of prior in vitro cell culture studies (27). In our experiments, the concentration of MTX (5 mM) falls within the span of serum concentrations described in the literature (28-31). Although previously measured serum MTX levels extend through a wide range (0.5-1800 μM), the relevance of serum levels to MTX effects is unclear since antiinflammatory activity has been attributed to multiple metabolites of methotrexate rather than methotrexate itself, some of which have vastly extended tissue half-lives (32).
Recent studies by Ghosh and colleagues (33) suggest that another mechanism by which COX-2 promotes atherosclerotic cardiovascular disease is by diminishing COX-2-mediated metabolism of endocannabinoids and downstream activation of PPARδ increasing endothelial tissue factor expression. Tissue factor promotes intravascular thrombosis and the increased expression of tissue factor likely predisposes to the further development of atherosclerotic cardiovascular disease. MTX may also reduce tissue factor expression by increasing adenosine levels and adenosine A2A receptor activation since it has been known for some time that adenosine suppresses tissue factor production (34-36).
In studies published recently, 27-hydroxycholesterol was shown to antagonize estrogen at its receptors, an effect which may negatively impact the development of atherosclerosis (37, 38). Although this effect may be an important factor in preventing postmenopausal estrogen therapy from diminishing the risk of atherosclerosis, it may not be relevant to the development of atherosclerosis in males. Moreover, conjugated estrogens also reduce blood concentrations of such anticoagulants as plasminogen activator inhibitor-1 (PAI-1) (39) and rising 27-hydroxycholesterol levels may reverse these estrogen-mediated effects. Furthermore, anti-atherogenic rather than pro-atherogenic effects have been observed with selective estrogen receptor modulators (SERMs) such as raloxifene and tamoxifen in vivo. Potential explanations have included favourable alterations of anti-oxidant activity, cholesterol, fibrinogen and homocysteine levels as well as vascular tone (40). Thus, the overall impact of possible estrogen receptor blockade on atherogenicity by 27-hydroxycholesterol may be eclipsed by a multitude of other factors at play.
This is the first evidence that any widely used pharmacotherapy can increase the expression of the anti-atherogenic 27-hydroxylase or ABCA1 and can counteract the effects of COX-2 inhibition or IFN-γ exposure on gene expression. We have demonstrated that MTX inhibits foam cell formation under conditions of lipid overload. Our results suggest that the capacity of MTX to reduce the burden of ASCVD in patients with RA may be ascribed, in part, to favorable alterations in cholesterol homeostasis mediated via activation of the adenosine A2A receptor. Thus, adenosine receptor ligation may provide a suitable mechanism for the development of a promising treatment paradigm with long term-benefit in ASCVD.
This work was supported by a grant from the National Institutes of Health/National Heart, Lung and Blood Institute HL073814 (Reiss). Additional support was provided by the Arthritis National Research Foundation (Chan). National Institutes of Health (AA13336, AR41911 and GM56268), King Pharmaceuticals, the General Clinical Research Center (M01RR00096) and by the Kaplan Cancer Center of New York University School of Medicine (Cronstein).