ARTHRITIS & RHEUMATISM Vol. 56, No. 5, May 2007, pp 1440–1445 DOI 10.1002/art.22643 © 2007, American College of Rheumatology
The Antiinflammatory Mechanism of Methotrexate Depends on Extracellular Conversion of Adenine Nucleotides to Adenosine by Ecto-5⬘-Nucleotidase Findings in a Study of Ecto-5⬘-Nucleotidase Gene–Deficient Mice M. Carmen Montesinos,1 Masahide Takedachi,2 Linda F. Thompson,2 Tuere F. Wilder,3 Patricia Ferna´ndez,3 and Bruce N. Cronstein3 Objective. Evidence from in vitro, in vivo, and clinical studies indicates that adenosine mediates, at least in part, the antiinflammatory effects of methotrexate (MTX), although the biochemical events involved have not been fully elucidated. This study was under-
taken to investigate whether MTX exerts antiinflammatory effects in mice that lack ecto-5ⴕ-nucleotidase (ecto5ⴕ-NT) (CD73) and are unable to convert AMP to adenosine extracellularly, in order to determine whether adenosine is generated intracellularly and transported into the extracellular space or is generated from the extracellular dephosphorylation of AMP to adenosine. Methods. Male CD73 gene–deficient mice and age-matched wild-type mice received intraperitoneal injections of saline or MTX (1 mg/kg/week) for 5 weeks. Air pouches were induced on the back by subcutaneous injection of air; 6 days later, inflammation was induced by injection of carrageenan. Results. Fewer leukocytes, but higher levels of tumor necrosis factor ␣ (TNF␣), accumulated in the air pouches of vehicle-treated CD73-deficient mice compared with those of wild-type mice. As expected, MTX treatment reduced the number of leukocytes and TNF␣ levels in the exudates and increased exudate adenosine concentrations in wild-type mice. In contrast, MTX did not reduce exudate leukocyte counts or TNF␣ levels or increase exudate adenosine levels in CD73-deficient mice. Conclusion. These results demonstrate that the antiinflammatory actions of MTX are mediated, at least in part, by increased release of adenine nucleotides that are hydrolyzed extracellularly to adenosine via an ecto5ⴕ-NT–dependent pathway.
Dr. Montesinos’ work was supported by the Ramo ´n y Cajal Program of the Spanish Ministry of Education and Science and grants from the Valencian Government (Conselleria d’Empresa, Universitat i Cie`ncia) (GV05/031) and the Instituto de Salud Carlos III (FIS 05/1659). Dr. Thompson’s work was supported by a grant from the NIH (AI-18220); she holds the Putnam City Schools Distinguished Chair in Cancer Research. Dr. Ferna´ndez is recipient of a postdoctoral fellowship from the Spanish Ministry of Education and Science. Dr. Cronstein’s work was supported by grants from the NIH (AR41911, GM-56268, AA-13336, and General Clinical Research Center grant M01-RR-00096), King Pharmaceuticals, and the Kaplan Cancer Center. 1 M. Carmen Montesinos, PhD: University of Valencia, Valencia, Spain, and New York University School of Medicine, New York, New York; 2Masahide Takedachi, DDS, PhD, Linda F. Thompson, PhD: Oklahoma Medical Research Foundation, Oklahoma City; 3 Tuere F. Wilder, MS, Patricia Ferna´ndez, PhD, Bruce N. Cronstein, MD: New York University School of Medicine, New York, New York. Dr. Cronstein has received consulting fees (less than $10,000 each) from King Pharmaceuticals, CanFite Biopharmaceuticals, Bristol-Myers Squibb, Cellzome, Tap Pharmaceuticals, Prometheus Laboratories, Regeneron (data safety and monitoring [Westat]), Sepracor, Amgen, Endocyte, and Protalex and has received honoraria or speaking fees (less than $10,000 each) from Tap Pharmaceuticals and Amgen. He owns stock in CanFite Biopharmaceuticals. He holds patents on use of adenosine A2A receptor antagonists to promote wound healing and to inhibit fibrosis, on testing for single-nucleotide polymorphisms in the adenosine A1 receptor to treat fibromyalgia, and on use of the adenosine A1 receptor to treat osteoporosis and other diseases of bone. Address correspondence and reprint requests to M. Carmen Montesinos, PhD, Department of Pharmacology, University of Valencia, Avenue Vicent Andre`s Estelle`s s/n, 46100 Burjassot, Valencia, Spain. E-mail:
[email protected]. Submitted for publication November 28, 2006; accepted in revised form February 16, 2007.
Low-dose weekly methotrexate (MTX) is a mainstay in the treatment of rheumatoid arthritis and other inflammatory diseases, with a relatively safe profile compared with other therapies (1). Although there is 1440
ROLE OF ECTO-5⬘-NT IN ANTIINFLAMMATORY ACTION OF MTX
still controversy about the mechanism of action of MTX (2), increasing evidence indicates that it promotes extracellular adenosine accumulation at sites of inflammation (3). The biochemical mechanism by which MTX promotes accumulation of extracellular adenosine remains unclear. In vivo treatment with MTX increases intracellular concentrations of 5-aminoimidazole-4-carboxamide ribonucleotide (AICAR), an intermediate in de novo purine biosynthesis (4). There are 2 potential mechanisms to explain previous observations that increased intracellular AICAR levels lead to increased extracellular adenosine (5). First, AICAR is a competitive inhibitor of AMP deaminase, and this inhibition may lead to accumulation and release of intracellular adenine nucleotides which are converted extracellularly to adenosine. Alternatively, AICA ribonucleoside, the dephosphorylated metabolite of AICAR, is a competitive inhibitor of adenosine deaminase, and this inhibition may lead to the direct accumulation and release of adenosine into the extracellular space. CD73 (ecto-5⬘-nucleotidase [ecto-5⬘-NT]) is a glycosyl phosphatidylinositol–anchored membranebound glycoprotein that catalyzes the hydrolysis of extracellular nucleoside monophosphates into bioactive nucleoside intermediates (6). Although previous studies showed that injection of a specific ecto-5⬘-NT inhibitor, ␣,-methylene ADP, suppressed the antiinflammatory properties of MTX in the murine air pouch model of inflammation (7), ␣,-methylene ADP may have other effects, including binding to adenine nucleotide receptors. In the present study we sought to selectively address the involvement of extracellular conversion of adenine nucleotides to adenosine by CD73 in the antiinflammatory effect of MTX, by studying mice with a targeted disruption of the gene encoding CD73. We report here that MTX treatment increases exudate adenosine concentrations, diminishes leukocyte accumulation, and reduces exudate tumor necrosis factor ␣ (TNF␣) levels in wild-type mice, but not in CD73deficient mice. MATERIALS AND METHODS Materials. Carrageenan (type I) was obtained from Sigma (St. Louis, MO). MTX was purchased from Immunex (San Juan, PR). All materials were the highest quality available. Animals. Mice with a targeted disruption of the gene for ecto-5⬘-NT were generated by homologous recombination in which the third exon of CD73 was replaced by a neomycin resistance cassette. They were healthy, bred normally, and
1441
appeared to have normal immune system development (6). Mice used for these experiments were the product of 6–8 backcrossings to C57BL/6 males. Genotyping was performed by polymerase chain reaction (PCR) on tail DNA, using the primers 5⬘-AATCCAGGGACAAATTTAGTC-3⬘ (forward) and 5⬘-AGAAAGGTGTTGGAGTGTCCT-3⬘ (reverse), which detect the wild-type CD73 allele, and 5⬘-CTTGGGTGGAGAGGCTATTC-3⬘ (forward) and 5⬘-AGGTGAGATGACAGGAGATC-3⬘ (reverse), which detect the mutated CD73 allele (neomycin resistance cassette). To perform the PCR, 0.3 g of genomic DNA was used in 30 l of final reaction mixture. PCR was performed with Mastercycler gradient (Eppendorf, Hamburg, Germany) under the following conditions: 94°C for 2 minutes, followed by 40 cycles of 94°C for 15 seconds, 57°C for 15 seconds, and 72°C for 15 seconds, and a final extension at 72°C for 10 minutes. Mice were housed in the Oklahoma Medical Research Foundation or New York University animal facilities, fed regular mouse chow, and given access to drinking water ad libitum. All procedures described below were reviewed and approved by the Institutional Animal Care and Use Committees of Oklahoma Medical Research Foundation and New York University Medical Center and carried out under the supervision of the facilities’ veterinary staff. Induction of air pouches and carrageenan-induced inflammation. Male mice (10–15 weeks old) were given weekly intraperitoneal (IP) injections of either MTX (1 mg/kg; freshly reconstituted lyophilized powder) or vehicle (0.9% saline) for 5 weeks. Air pouches were generated by subcutaneous injection of 3 ml of sterile air and reinflated (with 1.5 ml of sterile air) 2 days later. On day 6, inflammation was induced, within 3 days of the last administration of MTX, by injection of 1 ml of 2% carrageenan suspension. Four hours later, mice were killed by CO2 narcosis, and exudates harvested with 2 ml phosphate buffered saline (PBS) (4). Leukocytes were counted in a hemocytometer chamber, and concentrations of adenosine and cytokines (TNF␣ and keratinocyte chemoattractant) were quantified by high-performance liquid chromatography and enzyme-linked immunosorbent assay, respectively (8). Flow cytometry. Leukocytes in inflammatory exudates were stained with fluorescein isothiocyanate–conjugated anti– Mac-1 and phycoerythrin-conjugated anti–L-selectin (BD PharMingen, San Diego, CA), using standard methods. The mean fluorescence intensity (MFI) of Mac-1 and L-selectin staining on gated neutrophils was determined using CellQuest software and a FACSCalibur (Becton Dickinson Immunocytometry Systems, Mountain View, CA). Data were collected on 10,000 cells/air pouch. Statistical analysis. All statistical analyses were performed using GraphPad Prism version 4.00 for Windows (GraphPad Software, San Diego, CA) (www.graphpad.com). Differences between groups were analyzed by Student’s unpaired 2-tailed t-test. P values less than 0.05 were considered significant.
RESULTS To evaluate whether adenosine is generated intracellularly and transported into the extracellular space or, rather, is generated from the extracellular dephosphor-
1442
ylation of AMP to adenosine by ecto-5⬘-NT, we investigated the antiinflammatory effects of MTX in mice lacking CD73, using the air pouch model of inflammation. Carrageenan-induced inflammation led to accumulation of low levels of adenosine in the inflammatory exudates from wild-type mice treated with weekly IP injections of vehicle (mean ⫾ SEM 44 ⫾ 8 nM; n ⫽ 9), whereas levels of adenosine in the inflammatory exudates from vehicle-treated CD73-deficient mice were undetectable or very low (6 ⫾ 3 nM; n ⫽ 8) (P ⬍ 0.0005 versus vehicle-treated wild-type mice) (Figure 1A). Treatment of wild-type mice with MTX (1 mg/kg IP weekly for 5 weeks) increased exudate adenosine concentrations (114 ⫾ 12 nM; n ⫽ 9) (P ⬍ 0.0005 versus vehicle-treated wild-type mice), as we have previously shown (4,7,8). In contrast, MTX treatment did not affect adenosine levels in mice that were deficient in CD73 (6 ⫾ 3 nM; n ⫽ 8) (P not significant [NS] versus vehicle-treated CD73-deficient mice; P ⬍ 0.0005 versus vehicle-treated wild-type mice) (Figure 1A). The MTX dose used in our experiments (1 mg/kg) was much higher than is used in clinical practice. It is likely that MTX is metabolized differently in mice than in humans, and we found that higher doses were necessary to consistently suppress leukocyte recruitment in this model of inflammation. The concentrations of leukocytes in the inflammatory exudates from vehicle-treated CD73 gene– deficient mice were significantly lower than those in exudates from vehicle-treated wild-type mice (mean ⫾ SEM 2.6 ⫾ 0.2 versus 4.1 ⫾ 0.2 million cells/ml, n ⫽ 19 and n ⫽ 18 respectively) (P ⬍ 0.0001) (Figure 1B), similar to findings in adenosine A2A receptor–knockout mice studied using the same model of inflammation (9). To confirm that diminished leukocyte concentrations in CD73 gene–targeted mice were not due to neutrophil malfunction, we measured keratinocyte chemoattractant levels in the inflammatory exudates from wild-type and CD73-deficient mice and found no significant difference between the groups (mean ⫾ SEM 41.7 ⫾ 5.4 and 46. 8 ⫾ 5.0 ng/ml; n ⫽ 15 and n ⫽ 18, respectively). We also found that neutrophils from inflammatory exudates from both wild-type and CD73-deficient mice appeared to be activated to a similar degree, based on their levels of expression of Mac-1 (mean ⫾ SEM MFI 7,248 ⫾ 186 and 7,243 ⫾ 155; n ⫽ 10 and n ⫽ 13, respectively) and L-selectin (MFI 104 ⫾ 9 and 93 ⫾ 4; n ⫽ 10 and n ⫽ 13, respectively) (P NS for both), as measured by flow cytometry. These values represent a 2–3-fold increase in Mac-1 expression and a 50–70% decrease in L-selectin expression compared with those in
MONTESINOS ET AL
Figure 1. Effect of methotrexate (MTX) treatment on A, adenosine concentrations, B, leukocyte accumulation, and C, tumor necrosis factor ␣ (TNF␣) concentrations in air pouch exudates from CD73knockout (CD73-KO) mice and wild-type mice. Mice were treated with weekly intraperitoneal injections of MTX (1 mg/kg) or saline (control) for 5 weeks prior to induction of inflammation and subsequent collection of inflammatory exudates. Values are the mean and SEM. ⴱ ⫽ P ⬍ 0.05; ⴱⴱ ⫽ P ⬍ 0.0005; ⴱⴱⴱ ⫽ P ⬍ 0.0001 versus saline-treated wild-type mice, by Student’s t-test. NS ⫽ not significant.
neutrophils isolated from air pouches injected with PBS rather than carrageenan. Treatment with MTX reduced leukocyte accumu-
ROLE OF ECTO-5⬘-NT IN ANTIINFLAMMATORY ACTION OF MTX
lation in exudates from air pouches of wild-type mice by 49% (2.1 ⫾ 0.2 million cells/ml; n ⫽ 13) (P ⬍ 0.0001 versus vehicle-treated wild-type mice), again similar to previous results (4,7,8), but did not affect leukocyte concentrations in the exudates from CD73-deficient mice (2.3 ⫾ 0.3 million cells/ml; n ⫽ 9) (P NS versus vehicle-treated CD73-knockout mice; P ⬍ 0.0001 versus vehicle-treated wild-type mice) (Figure 1B). Under the conditions studied, there was no difference in the type of white cells that accumulated in the air pouch exudates from either MTX-treated or saline-treated wild-type or CD73-deficient mice (⬎90% polymorphonuclear cells). In contrast to the findings with regard to leukocyte accumulation, we found significantly increased accumulation of the inflammatory cytokine TNF␣ in exudates from vehicle-treated CD73 gene–deficient mice compared with their wild-type littermates (mean ⫾ SEM 2.1 ⫾ 0.33 versus 1.4 ⫾ 0.20 ng/ml; n ⫽ 23 and n ⫽ 20, respectively) (P ⬍ 0.05) (Figure 1C). MTX inhibited TNF␣ accumulation in inflammatory exudates from wild-type mice (0.98 ⫾ 0.07 ng/ml; n ⫽ 15) (P ⬍ 0.05 versus vehicle-treated wild-type mice), but had no effect on TNF␣ accumulation in exudates from CD73deficient mice (2.1 ⫾ 0.25 ng/ml; n ⫽ 15) (P NS versus vehicle-treated CD73-knockout mice; P ⬍ 0.05 versus vehicle-treated wild-type mice) (Figure 1C). DISCUSSION Adenosine is a ubiquitous purine nucleoside present in all tissues and body fluids, which modulates cellular and organ function via occupancy of specific cell surface receptors (A1, A2A, A2B, and A3). Adenosine receptors are all members of the large family of 7-transmembrane–spanning, heterotrimeric G protein– associated receptors. Extracellular adenosine concentrations tend to remain constant under resting conditions (30–300 nM) and are held in tight check by the equilibrium between adenosine production/release into the extracellular space and adenosine uptake by cells or catabolism to inosine and other purines not active at adenosine receptors. In contrast, under conditions of cellular or tissue necrosis or stress, adenosine levels may increase to the micromolar or even higher range, as a result of ATP catabolism (10). Extracellular ATP, like other nucleotides (ADP, UTP, and UDP), exerts many biologic effects through direct activation of cell surface receptors for adenine nucleotides. Adenine nucleotides are sequentially degraded to adenosine by the coordinated action of ectoapyrase (CD39), which dephosphorylates ATP and ADP
1443
to AMP, and finally ecto-5⬘-NT (CD73), which catalyzes the dephosphorylation of AMP to adenosine. Consistent with the known pathways of extracellular adenine nucleotide metabolism, we found that CD73-deficient mice, which are unable to catabolize the hydrolysis of AMP to adenosine, accumulate less adenosine in exudates after an inflammatory insult, in comparison with their wildtype littermates. Adenosine has been described as an endogenous antiinflammatory agent (11). Accordingly, in CD73 gene–deficient mice, vascular cell adhesion molecule 1 expression on vascular endothelial cells is up-regulated, and an increased number of monocytes adhere to the endothelial cell walls (12). One might therefore expect the number of neutrophils in exudates from genetargeted mice to be increased. However the concentrations of leukocytes in the inflammatory exudates from CD73 gene–deficient mice were significantly lower than those in exudates from wild-type mice. This finding is reminiscent of the 29% reduction in leukocyte concentrations observed in exudates from carrageenan-treated air pouches of adenosine A2A receptor–deficient mice compared with wild-type mice (9). Adenosine A2A receptor–deficient mice also showed a 27% decrease in the extravasation of Evans blue dye into the air pouch lumen and a reduction in the density of microvessels in the air pouch wall (9). These results suggest that the A2A receptor plays an important role in the known neovascularizationpromoting activity of adenosine. It is probable that the lower leukocyte concentrations observed in carrageenan-treated air pouches of vehicle-treated CD73-deficient mice can be explained by a similar phenomenon. It therefore appears that in our experimental system, the likely decreased angiogenesis in the air pouch wall in the CD73 gene–targeted mice has a greater effect on leukocyte migration into the air pouches than does the increased endothelial cell activation that is also occurring. It is unlikely that the reduced leukocyte concentrations in the air pouches of CD73-deficient mice are a consequence of abnormal neutrophil numbers or function. The numbers of neutrophils in the blood of these animals are normal (Thompson LF: unpublished observations), as are the cellularity and composition of all of the lymphoid organs (6). In addition, the concentration of keratinocyte chemoattractant, the main chemokine responsible for neutrophil migration into sites of inflammation, was similar in the exudates from gene-targeted and wild-type mice. Furthermore, neutrophil activation after carrageenan injection, as assessed by down-
1444
MONTESINOS ET AL
Figure 2. Proposed mechanism by which methotrexate (MTX) increases extracellular adenosine concentrations. RFC-1 ⫽ reduced folate carrier 1; FPGS ⫽ folylpolyglutamate synthetase; MTXglu ⫽ MTX polyglutamates; DHFR ⫽ dihydrofolate reductase; MTHFR ⫽ methylene tetrahydrofolate reductase; AICAR ⫽ 5-aminoimidazole-4-carboxamide ribonucleotide; AICART ⫽ AICAR transformylase; Pi ⫽ inorganic phosphate; AICAr ⫽ AICA ribonucleoside; ADA ⫽ adenosine deaminase; FAICAR ⫽ formyl AICAR; PPi ⫽ inorganic pyrophosphate.
regulation of L-selection expression and up-regulation of Mac-1 expression, was also equivalent in the 2 strains. We conclude that neutrophil activation and migration are normal in CD73-deficient mice, making abnormal neutrophil function an unlikely explanation for the lack of response to MTX treatment observed in these mice. We demonstrated increased levels of TNF␣ in the air pouch exudates from CD73-deficient mice. TNF␣ is not produced exclusively by neutrophils present in the inflammatory exudates. In fact, Garcia-Ramallo et al (13) observed a time-dependent increase in levels of TNF␣ in exudates preceding cell recruitment. Air pouches generated for 2, 6, or 9 days before inflammation was initiated showed a proportional increase in the number of cells lining the cavities. Two hours after carrageenan stimulation, the synthesis of TNF␣ increased in proportion to the lining cellularity (13). Since no differences in the numbers of infiltrating leukocytes were found, these data suggest that the early source of
TNF␣ is resident cells, such as tissue macrophages. Adenosine, through activation of membrane receptors, has been shown to inhibit the secretion of inflammatory cytokines by macrophages (11). Therefore, TNF␣ concentrations are higher in the gene-deficient mice even though the numbers of cells in their exudates are lower than in exudates from saline-treated wild-type mice. We have previously demonstrated, pharmacologically and in studies using appropriate adenosine receptor gene–targeted mice, that adenosine, acting at one or more of its receptors, mediates the antiinflammatory effect of MTX in both acute and chronic inflammation (4,8,14,15). Moreover, it has been shown that genetic factors contribute to the antiinflammatory efficacy of MTX, both in humans and in mice, and polymorphisms in enzymes involved in MTX-induced adenosine upregulation are likely responsible for either good response or resistance to MTX (3,16). However, the mechanism by which MTX increases adenosine levels at
ROLE OF ECTO-5⬘-NT IN ANTIINFLAMMATORY ACTION OF MTX
the site of inflammation has not been fully explained. We found in the present study that CD73-deficient mice are resistant to the antiinflammatory action of MTX. Our results strengthen the hypothesis that adenosine participates in the antiinflammatory effect of MTX in the carrageenan air pouch model of inflammation, and are most consistent with the following model of MTX’s antiinflammatory mechanism (Figure 2): MTX is first taken up by cells, where it undergoes polyglutamation; MTX polyglutamates then accumulate within cells or tissues. MTX polyglutamates inhibit AICAR transformylase, leading to intracellular accumulation of AICAR. Increased AICAR levels lead to increased intracellular AMP via competitive inhibition of AMP deaminase. The increased AMP in the cells is more likely to be exported to the extracellular space, where it is converted to adenosine. Alternatively, elevated levels of AMP could lead to increased levels of ADP and/or ATP, which can be exported and dephosphorylated to adenosine extracellularly by the combined action of CD39 and CD73. Finally, adenosine interacts with its receptors on stimulated inflammatory cells to inhibit cytokine production and diminish inflammation (7). Loss of ecto-5⬘-NT activity decreases the generation of adenosine and thus abrogates the capacity of MTX to diminish inflammation. It will be important to determine whether polymorphisms in the cd73 gene can account for resistance to MTX in some rheumatoid arthritis patients whose disease fails to respond to this therapy.
ACKNOWLEDGMENTS The authors are grateful to Aletha Laurent, James Vaughn, Stephanie McGee, and Scott Hooker for technical assistance.
AUTHOR CONTRIBUTIONS Dr. Montesinos 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. Montesinos, Takedachi, Thompson, Cronstein. Acquisition of data. Montesinos, Takedachi, Thompson, Wilder, Ferna´ndez. Analysis and interpretation of data. Montesinos, Takedachi, Thompson, Cronstein. Manuscript preparation. Montesinos, Takedachi, Thompson, Cronstein. Statistical analysis. Montesinos, Takedachi, Thompson, Wilder, Ferna´ndez.
1445
REFERENCES 1. Borchers AT, Keen CL, Cheema GS, Gershwin ME. The use of methotrexate in rheumatoid arthritis. Semin Arthritis Rheum 2004;34:465–83. 2. Benito-Garcia E, Heller JE, Chibnik LB, Maher NE, Matthews HM, Bilics JA, et al. Dietary caffeine intake does not affect methotrexate efficacy in patients with rheumatoid arthritis. J Rheumatol 2006;33:1275–81. 3. Wessels JA, Kooloos WM, de Jonge R, de Vries-Bouwstra JK, Allaart CF, Linssen A, et al. Relationship between genetic variants in the adenosine pathway and outcome of methotrexate treatment in patients with recent-onset rheumatoid arthritis. Arthritis Rheum 2006;54:2830–9. 4. Cronstein BN, Naime D, Ostad E. The antiinflammatory mechanism of methotrexate: increased adenosine release at inflamed sites diminishes leukocyte accumulation in an in vivo model of inflammation. J Clin Invest 1993;92:2675–82. 5. Gruber HE, Hoffer ME, McAllister DR, Laikind PK, Lane TA, Schmid-Schoenbein GW, et al. Increased adenosine concentration in blood from ischemic myocardium by AICA riboside: effects on flow, granulocytes and injury. Circulation 1989;80:1400–11. 6. Thompson LF, Eltzschig HK, Ibla JC, van de Wiele CJ, Resta R, Morote-Garcia JC, et al. Crucial role for ecto-5⬘-nucleotidase (CD73) in vascular leakage during hypoxia. J Exp Med 2004;200: 1395–405. 7. Morabito L, Montesinos MC, Schreibman DM, Balter L, Thompson LF, Resta R, et al. Methotrexate and sulfasalazine promote adenosine release by a mechanism that requires ecto-5⬘-nucleotidase-mediated conversion of adenine nucleotides. J Clin Invest 1998;101:295–300. 8. Montesinos MC, Desai A, Delano D, Chen JF, Fink JS, Jacobson MA, et al. Adenosine A2A or A3 receptors are required for inhibition of inflammation by methotrexate and its analog MX-68. Arthritis Rheum 2003;48:240–7. 9. Montesinos MC, Desai A, Chen JF, Yee H, Schwarzschild MA, Fink JS, et al. Adenosine promotes wound healing and mediates angiogenesis in response to tissue injury via occupancy of A2A receptors. Am J Pathol 2002;160:2009–18. 10. Schulte G, Fredholm BB. Signalling from adenosine receptors to mitogen-activated protein kinases. Cell Signal 2003;15:813–27. 11. Hasko G, Cronstein BN. Adenosine: an endogenous regulator of innate immunity. Trends Immunol 2004;25:33–9. 12. Zernecke A, Bidzhekov K, Ozuyaman B, Fraemohs L, Liehn EA, Luscher-Firzlaff JM, et al. CD73/ecto-5⬘-nucleotidase protects against vascular inflammation and neointima formation. Circulation 2006;113:2120–7. 13. Garcia-Ramallo E, Marques T, Prats N, Beleta J, Kunkel SL, Godessart N. Resident cell chemokine expression serves as the major mechanism for leukocyte recruitment during local inflammation. J Immunol 2002;169:6467–73. 14. Montesinos MC, Yap JS, Desai A, Posadas I, McCrary CT, Cronstein BN. Reversal of the antiinflammatory effects of methotrexate by the nonselective adenosine receptor antagonists theophylline and caffeine: evidence that the antiinflammatory effects of methotrexate are mediated via multiple adenosine receptors in rat adjuvant arthritis. Arthritis Rheum 2000;43:656–63. 15. Montesinos MC, Desai A, Cronstein BN. Suppression of inflammation by low-dose methotrexate is mediated by adenosine A2A receptor but not A3 receptor activation in thioglycollate-induced peritonitis. Arthritis Res Ther 2006;8:R53. 16. Delano DL, Montesinos MC, Desai A, Wilder T, Fernandez P, D’Eustachio P, et al. Genetically based resistance to the antiinflammatory effects of methotrexate in the air-pouch model of acute inflammation. Arthritis Rheum 2005;52:2567–75.
