Adenosine A_2A Receptor Activation and Macrophage-mediated Experimental Glomerulonephritis

Abstract

In immune-induced inflammation, leukocytes are key mediators of tissue damage. Since A_2A adenosine receptors (A_2AR) are endogenous suppressors of inflammation, we examined cellular and molecular mechanisms of kidney damage to determine whether selective activation of A_2AR will suppress inflammation in a rat model of glomerulonephritis. Activation of A_2AR reduced the degree of kidney injury in both the acute inflammatory phase and the progressive phase of glomerulonephritis. This protection against acute and chronic inflammation was associated with suppression of the glomerular expression of the MDC/CCL22 chemokine and down-regulation of MIP-1α/CCL3, RANTES/CCL5, MIP-1β/CCL4, and MCP-1/CCL2 chemokines. The expression of anti-inflammatory cytokines, IL-4 and IL-10, also increased. The mechanism for these anti-inflammatory responses to the A_2AR agonist was suppression of macrophages function. A_2AR expression was increased in macrophages, macrophage-derived chemokines were reduced in response to the A_2AR agonist, and chemokines not expressed in macrophages did not respond to A_2AR activation. Thus, activation of the A_2AR on macrophages inhibits immune-associated inflammation. In glomerulonephritis, A_2AR activation modulates inflammation and tissue damage even in the progressive phase of glomerulonephritis. Accordingly, pharmacological activation of A_2AR could be developed into a novel treatment for glomerulonephritis and other macrophage-related inflammatory diseases.

INTRODUCTION

We still treat immunologically induced inflammatory diseases with nonspecific immunosuppressive drugs that can cause significant morbidity (1). For this reason defining the pathways of organ damage could lead to design of more specific therapies.

Crescentic glomerulonephritis is a rapidly progressive glomerular disease with a poor prognosis. Macrophages can play an important role in the induction and development of the disease and both the magnitude of proteinuria and the percentage of crescentic glomeruli are correlated with the number of macrophages that infiltrate the glomerulus; this is not the case with the number of CD8⁺ cells (2, 3). Macrophages are constituents of glomerular crescents in progressive crescentic glomerulonephritis and they probably play a major role in the irreversible scarring that leads to kidney failure (4–8). In Wistar-Kyoto rats (WKY), small doses of anti-glomerular basement membrane (GBM) antibody induce proliferative and necrotizing glomerulonephritis with crescent formation plus infiltration of CD8⁺ cells and macrophages into the glomeruli (9–13). In this model macrophages accumulate in glomeruli during later phases and contribute to the progressive decline in kidney function (3, 14).

Chemokines mediate selective attraction and activation of various subsets of leukocytes (6, 15–17). In the induction of anti-GBM glomerulonephritis in WKY rats CC chemokines play an important role. We found that CC chemokines, MCP-1/CCL2, MIP-1β/CCL4, RANTES/CCL5, and MDC/CCL22, and the CX3C chemokine, fractalkines/CX₃CL1, are induced during the disease and that the expression of multiple chemokines coincides with the influx of CD8⁺ cell and ED1⁺ macrophages into the glomeruli. In in vitro studies these chemokines induced strong migratory response in inflammatory cells prepared from the nephritic glomeruli (9–11).

The degree of inflammation can be tightly regulated by endogenous anti-inflammation pathways, including, PPARγ, adiponectin, NO, and adenosine receptors (18–24). Genetic and pharmacological evidence support a non-redundant role for both endogenous adenosine and A_2A receptors in protecting against acute inflammatory damage in models of inflammatory injury and systemic inflammation (25). The anti-inflammatory mechanism involves downregulation of activated immune cells in vivo (25). Thus, the A_2AR could modulate inflammatory processes because it is expressed on most cells involved in inflammation, including a variety of hematopoietic cells, endothelial cells, and smooth muscle cells (19, 20, 25, 26). We have found that quiescent peritoneal macrophages do not express A_2AR but following addition of LPS, these cells express abundant A_2AR. In vitro studies indicate that A_2AR are present on polymorphonuclear leukocytes (PMN) and can inhibit the expression of β2-integrins and adhesion, can suppress oxygen radical production, degranulation, and production of TNF-α (27–31). In addition, expression and activation of A_2A receptor on macrophages can inhibit the production of IL-12, TNF-α, and nitric oxide, and enhance the secretion of IL-10 in response to lipopolysaccharide (LPS) (32–35). Adenosine can inhibit proliferation, activation, and production of inflammatory cytokines in peripheral T-cells it also can enhance the production of anti-inflammatory cytokines (36, 37).

Because of the potential for beneficial effects of pharmacological activation of A_2AR as a therapeutic target, we tested the hypothesis that A_2AR activation would protect against immune-mediated inflammation in experimental glomerulonephritis. Activation of A_2AR significantly reduced leukocyte infiltration into the kidney and prevented kidney injury during the acute inflammatory phase and the progressive phase of glomerulonpehritis. Our results indicate that activation of A_2AR represents a potential therapeutic strategy for glomerulonephritis and possibly other macrophage-mediated diseases.

MATERIAL AND METHODS

Induction, treatment, and analysis of anti-GBM glomerulonephritis

Animal studies were approved by the IACUC at Baylor College of Medicine. Male WKY rats (Harlan Sprague Dawley Inc., Indianapolis, IN), weighing 180–200 grams received an intravenous injection of 25µl/100g body weight of anti-GBM Ab as described (9, 10, 12, 38). We activated A_2AR with a selective A_2A receptor agonist, CGS 21680, (Tocris Cookson Inc., Ellisville, MO, 1.5mg/Kg i.p. twice a day, n = 6) and inactivated A_2AR using a selective receptor antagonist, ZM241385 (Tocris, 2mg/Kg i.p. twice daily, n = 6) for a period of 5 days, as described (25). In initial experiments, we pharmacologically activated or inhibited A_2AR during the acute inflammatory phase of the disease by starting treatment 8 hours after the injection of anti-GBM antibody (maximum glomerular deposition of IgG occurs one hour after injection of anti-GBM antibody (39)). In the second set of experiments we began treatment with A_2AR agonist at day six after the injecting the anti-GBM antibody (the progressive phase of crescentic glomerulonephritis). Urine protein was assayed using sulfosalicylic acid and rats were euthanized at day six, or day 12 to collect blood and kidney tissues. During the first hour after injecting the A_2AR agonist or the A_2AR antagonist, tail-cuff blood pressure was measured in conscious rats (Visitech Systems, Apex NC) as described (40).

mRNA expression of A_2A R, chemokines, cytokines, and adhesion molecules

Rat A_2AR nucleotides 459 to 801 (GenBank sequence accession number NM_053294) were used to generate A_2AR probe from brain tissue by RT-PCR. MDC/CCL22 (400 bp), RANTES/CCL5 (246 bp), MIP-1a/CCL3 (284 bp), factalkine/CX3CL1 (420 bp), MIP-3β/CCL19 (380 bp), MIP-1β/CCL4 (210 bp), MCP-1/CCL2 (239 bp), VCAM-1 (371 bp), ICAM-1 (292 bp), and L-32 (92 bp) riboprobes were generated by PCR reaction using cDNA templates. rCK1 (BD Pharmingen, San Diego, CA) was used to investigate cytokine expression. Glomeruli were prepared by sequential sieving and total RNA was isolated from glomeruli (11, 41). Three µg of total RNA from each sample was used in an RNase protection assay using the Torrey Pines Biolabs kit (Houston, TX) as described (11, 16, 42, 43). Phosphoimage quantitiation was performed using the PhosphorImager SI scanning instrument and ImageQuaNT software (Molecular Dynamics, Sunnyvale, CA) (16, 44, 45).

Morphological analysis, immunohistochemical phenotyping, and quantitation of leukocytes

Kidney samples fixed in formalin or methanol-Carnoy fixative solution were embedded in paraffin. Two to three-µm sections were stained with periodic acid-Schiff reagent to assess glomerular hypercellularity, necrotizing lesions, and formation of glomerular crescents (crescentic glomeruli per 100 glomeruli was calculated and expressed as a percentage). Infiltrating leukocytes were immunohistochemically stained for CD8⁺ and ED1⁺, as described (9, 11, 12). Positively stained cells per 100 glomeruli were counted and expressed per glomerular section. All quantitative morphological analyses were performed in a blinded fashion.

Immunohistochemistry of A_2AR, ED1⁺ cells, CD8⁺ cells, mesangial cells, and podocytes

Paraffin sections of methanol-Carnoy fixed tissue were stained with goat polyclonal anti-A_2AR antibody (1:100, Santa Cruz Biotechnology, Santa Cruz, CA) and monoclonal antibody ED-1 against rat macrophages (Chemicon, Temecula, CA), or monoclonal OX-8 against rat CD8 (BD Biosciences Pharmingen). Two-color dual antigen immunostaining was obtained by serial avidin-biotin peroxidase and alkaline phosphatase staining reactions with final chromogenic substrates of diamibenzidine (brown color) and Fast Red (red color), respectively as described (12, 46, 47).

Kidney sections were also stained with α-SMA (Dako, Carpinteria, CA), a marker of activated mesangial cells or α-Actinin-4 (AXXORA, LLC, San Diego, CA), a marker for podocytes (48, 49).

RESULTS

A_2A R expression in nephritic glomeruli

Normal glomeruli do not express A_2A R. In response to the anti-GBM antibody, however, there was strong expression of A_2AR from day 3 onwards (Figure 1 A). The glomerular level of A_2AR mRNA was 6.2 fold higher in anti-GBM Ab-treated rats than in normal glomeruli on day 3. A_2AR expression peaked at day 5 (6.8-fold increase) and started to decrease by day 7 (5.2-fold increase). Double immunohistochemical staining of ED1⁺ cells and A_2AR demonstrated that A_2AR was virtually confined to macrophages. There was some staining in the location of glomerular endothelial cells (Figure 1 B) but A_2AR was not expressed in either CD8⁺ cells (Figure 1 C) or in monocytes located in extraglomerular vessels (Figure 1 D). These results suggest that increased expression of A_2AR in macrophages might yield an anti-inflammatory defense against kidney damage.

1.

Induction of A_2AR in nephritic glomeruli. (A) RNase protection analysis was performed to determine A_2AR mRNA expression in the glomeruli of anti-GBM glomerulonephritis in WKY rats. A_2AR was induced from day 3 onwards during anti-GBM glomerulonephritis. The expression of A_2AR peaked at day 5 and started to decrease by day 7. Rat ribosomal L-32 gene was used as a housekeeping gene. Probe contains polylinker regions and is longer than the protected bands. The expression levels of A_2AR mRNA are expressed relative to the mRNA levels of the housekeeping gene L-32. Data shown is a representative of three separate experiments. (B) Double immunohistochemical staining of ED1⁺ cells (brown color) overlap with A_2AR (red color), thus A_2AR was mainly in Mϕ. (C) Interestingly, A_2AR (red color) was not expressed in CD8+ cells (brown color) or (D) monocytes (red color) located in extraglomerular vessels. (B) and (C), right panel, show ED1⁺ or CD8⁺ cells staining after removal of red color (A_2AR staining) from (B) and (C) slides at left panel to demonstrate the overlap of A_2AR expression in ED1+ but not in CD8+ cells. Sections for staining were sampled on day 7 after anti-GBM antibody injection.

A_2AR activation prevents glomerulonephritis

In the acute (day 1 to day 5) phase of crescentic glomerulonephritis control rats exhibited severe glomerular hypercellularity, necrotizing lesions, and crescentic formation (Figure 2 and Figure 3). Increased expression of α-SMA, a marker for mesangial cells injury, was inversely correlated with reduced α-Actinin-4, a marker for podocytes, indicating the degree of glomerular injury (Figure 4). In rats treated with CGS21680, a specific A_2AR agonist, however, there was only minimal damage to the kidneys. Notably, crescent formation, a major characteristic of this model, was abolished by CGS21680 treatment and necrotizing lesions were not observed while glomerular hypercellularity was markedly attenuated by ~ 70% (Figure 2 and Figure 3 a). Treatment with the A_2AR agonist also decreased α-SMA expression and restored the expression of α-Actinin- 4 (Figure 4). There also was no proteinuria (Figure 5 a). In contrast, in rats with anti-GBM glomerulonephritis treated with a specific antagonist for A_2AR, ZM241385, there was more severe glomerular hypercellularity, necrotizing lesions, and crescentic formation compared to control rats (Figure 2 and Figure 3 a). A_2AR antagonist also enhanced the expression of α-SMA and increased the loss of α-Actinin- 4 (Figure 4). Finally, proteinuria was also significantly higher in ZM241385-treated vs. control rats with glomerulonephritis (Figure 5 a). Thus, selective activation of A_2AR at acute phase (day 1 to day 5) confers kidney protection from damage.

2.

A_2AR attenuates inflammatory infiltrates and renal injury during the acute inflammatory phase and progressive phase in anti-GBM glomerulonephritis. Immunohistochemistry stained for CD8⁺ cells or ED1⁺ monocytes/macrophages, and periodic acid-Schiff staining of kidney sections of control, CGS216809 (CGS), and ZM241385 (ZM)-treated rats with anti-GBM glomerulonephritis. In the acute phase of the disease (day 1 to day 5), the control group of glomerulonephritis rats display severe glomerular hypercellularity, necrotizing lesions, crescentic formation, and prominent accumulation of CD8⁺ and ED1⁺ macrophages infiltration. In CGS-treated rats, crescentic formation and necrotizing lesions were prevented, and CD8⁺ and ED1⁺ macrophages infiltrates attenuated. Worsening of glomerular injury was observed in rats treated with ZM. In the progressive phase of the disease (day 6 to day 11) CGS markedly reduced glomerular lesion and ED1⁺ cell infiltrate.

3.

Quantitation of CD8⁺ and ED1⁺ cell infiltration, glomerular hypercellularity, necrotizing lesions, and crescent formation in the glomeruli from WKY rats with anti-GBM glomerulonephritis that were treated with vehicle, CGS21680 (CGS) or ZM241385 (ZM) during day 1 to day 5 and day 6 to day 11. One hundred glomeruli per section were counted. Each data point represents sections sampled from six rats and is expressed as mean ± SEM. *p<0.0001 vs. control, **p<0.005 vs. control, ***p<0.05 vs. control, σp<0.0001 vs. CGS

4.

A_2AR activation decreases α-SMA expression and restores α-Actinin-4 expression in anti-GBM GN. Immunohistochemistry stained for α-SMA and α-Actinin-4 of kidney sections of normal, control, CGS216809 (CGS), and ZM241385 (ZM)-treated rats with anti-GBM glomerulonephritis. In the acute phase of the disease (day 1 to day 5), the control group of glomerulonephritis rats display increased expression of α-SMA and lost of α-Actinin-4. In CGS-treated rats, the expression of α-SMA was reduced and the expression of α-Actinin-4 restored. Increased expression of α-SMA and enhanced loss of α-Actinin-4 was observed in rats treated with ZM. In the progressive phase of the disease (day 6 to day 11) CGS markedly reduced α-SMA and restored α-Actinin-4.

5.

Determination of proteinuria (milligrams of urine protein per 24 h) in Wistar-Kyoto rats with anti-GBM glomerulonephritis that were treated with vehicle, CGS21680 (CGS), and ZM241385 (ZM). In the acute phase of the disease, CGS completely blocked proteinuria in anti-GBM antibody-injected rats. Proteinuria was significantly higher in ZM-treated rats than the control group. In the progressive phase of glomerulonephritis, CGS markedly reduced proteinuria. Treatment Samples were sampled from six rats per group and expressed as mean ± SEM. *p<0.0001 vs. control, **p<0.005 vs. control, ***p<0.05 vs. control, δp<0.005 vs. ZM, &p<0.0001 vs. ZM

Increased or normal blood pressure has been reported in mice lacking A_2AR. We determined if A_2AR antagonism could increase blood pressure and contribute to kidney damage (46, 50). There was no significant difference in systolic blood pressure among the groups at day 4 or day 6 after induction of anti-GBM GN (D4: 147.75 ± 18.7, 141.0 ± 14.2, 149.7 ± 14.1; D6: 134.6 ± 8.01, 131.0 ± 9.43, 137.0 ± 5.6 in control, A_2AR agonist, andA_2AR antagonist treated rats respectively).

Next, we treated rats in the progressive phase of glomerulonephritis (day 6 to day 11) using the A_2AR agonist, CGS21680. As shown in Figure 2 and Figure 3b, glomerular hypercellularity, necrotizing lesions, and crescentic formation were markedly reduced at day 12 compared to the results in the control group (88%, 90%, and 74%, respectively). A_2AR activation also decreased α-SMA expression and restored the expression of α-Actinin-4 vs. results in the control group (Figure 4). Finally, CGS21680 reduced proteinuria within 24 hours of administration; the inhibitory effect was sustained during the treatment period (Figure 5 b). Thus, selective activation of A_2AR also reduces kidney injury and improves proteinuria during the progressive phase of glomerulonephritis.

Prevention of glomerulonephritis by A_2AR activation reduces lymphocyte/macrophage infiltration

To understand the mechanism for the prevention of kidney-injury, we examined the infiltration of leukocytes into the kidney. As shown in Figure 2 and Figure 3a there was a prominent accumulation of CD8⁺cells and ED1⁺ macrophages in glomeruli of control rats during the acute (day 1 to 5) inflammatory phase of glomerulonephritis. Treatment with CGS21680 dramatically attenuated infiltration of CD8⁺ cells and ED1⁺ macrophages into glomeruli. Pharmacological activation of A_2AR in the progressive phase (from day 6 to day 11) blocked macrophage infiltration to levels only 26% of the level in glomeruli of control rats (Figure 2 and Figure 3 b). This model of glomerulonephritis is characterized by an early infiltration of CD8⁺ cells with a maximum increase on day 3 after the injection of anti-GBM Ab (13). Consequently, CD8⁺ cells were not observed in either control or CGS21680 treated rats.

A_2A R activation modulates the expression of chemokines and cytokines in anti-GBM glomerulonephritis

A potential mechanism by which activation of A_2AR protects from inflammatory damage and leukocyte infiltration could be via suppression of cytokines/chemokines. We found that pharmacological A_2AR activation in the acute inflammatory phase (day 1 to day 5) did not reduce glomerular expression of pro-inflammatory cytokines IL-1β or TNF-α mRNA. In contrast, it significantly increased the expression of anti-inflammatory cytokines, IL-4 and IL-10 (69% and 66% respectively), compared to levels in glomeruli of control rats (Figure 6 a and Figure 7a). Activation of A_2AR abolished the induction of MDC/CCL22 and significantly attenuated the expression of MIP-1a/CCL3, MIP-1β/CCL4, RANTES/CCL5, and MCP-1/CCL2 (Figure 6 b, c and 7 b, c) but did not affect expression of fractalkine/CX₃CL1 (Figure 6 c and Figure 7 c). Interestingly, inhibition of A_2AR by ZM241385 did not significantly change the expression of these chemokines. Expression of anti-inflammatory cytokines seemed to be attenuated by ZM241385 but the changes were not statistically significant compared to the control results. When compared with levels in CGS21680-treated rats, IL-10 and IL-4 were significantly reduced in rats treated with ZM241385 (64.2% and 61% respectively) (Figure 6 a and Figure 7 a).

6.

Selective A_2AR activation alters anti-GBM glomerulonephritis-induced cytokine expression. RNase protection assay of cytokines and chemokines expressed in the glomeruli of anti-GBM glomerulonephritis in WKY rats. CGS in the acute inflammatory phase of the disease (day 1 to day 5) increased the expression of IL-4 and IL-10 (a), abolished the increase of the expression of MDC/CCL22, and attenuated the expression of MIP-1α/CCL3, RANTES/CCL5, MIP-1β/CCL4, and MCP-1/CCL2 (b,c). ZM reduced IL-4 and IL-10 expression compared with CGS group (a). In the progressive phase of the disease (day 6 to day 11), CGS significantly enhanced the expression of IL-4 (d) and suppressed the increase of the expression of MDC/CCL22 (e). λ Refers to undigested probe. Each lane represents a single rat.

7.

Selective A_2AR activation alters anti-GBM glomerulonephritis-induced cytokine expression. Densitometric analysis of blots from RNase protection assay of cytokines and chemokines expressed in the glomeruli of anti-GBM glomerulonephritis in WKY rats. The data are presented as a ratio of the cpm for the specific mRNA/L-32 mRNA to ensure a constant quantity of RNA in each sample. Results were sampled from six rats per group and expressed as mean ± SD. *p<0.05 vs. nl, δp<0.005 vs. nl, γp<0.0005 vs. nl, **p<0.05 vs. ctrl, θp<0.005 vs. ctrl, λp<0.0005 vs. ctrl, κp<0.05 vs. CGS, &p<0.005 vs. CGS. nl refers to normal rat.

We also investigated the effects of the A_2AR agonist and antagonist on the expression of VCAM-I and ICAM-1 in anti-GBM glomerulonephritis. Increased expression of VCAM-1 and ICAM-1 was observed in anti-GBM glomerulonephritis and A_2AR activation or inactivation did not significantly modify their expression in the acute inflammatory phase (day 1 to day 5) compared to control rats (Figure 8).

8.

Expression of adhesion molecules is not modified by pharmacological activation of A_2AR. Densitometric analysis of blots from RNase protection assay of VCAM-1 and ICAM-1 expressed in the glomeruli in anti-GBM glomerulonephritis. The data are presented as a ratio of the cpm for the specific mRNA/L-32 mRNA to ensure a constant quantity of RNA in each sample. Results were sampled from six rats per group and expressed as mean ± SD. *p<0.05 vs. nl, δp<0.005 vs. nl, γp<0.0005 vs. nl,

Pharmacological activation of A_2AR in rats with established glomerulonephritis (day 6 to day 11), significantly enhanced the expression of the anti-inflammatory cytokine IL-4 (104%) but not of IL-10 (Figure 6 d and Figure 7 d). The induction of MDC/CCL22 chemokine was suppressed, but the expression of other chemokines and VCAM-1 or ICAM-1 did not change. Fractalkine/CX₃CL1 expression significantly increased (Figure 6 e, f;Figure 7 e, f, and Figure 8).

These results suggest that A_2AR pharmacological activation modulates renal injury by altering expression of inflammatory and anti-inflammatory cytokines in the kidney.

DISCUSSION

Our results indicate that activation of adenosine A_2A receptors prevents infiltration of leukocytes into the kidney, suppresses glomerular inflammation, and protects the kidney from inflammatory injuries induced by anti-GBM glomerulonephritis. These benefits were observed both in the acute inflammatory phase and during the progressive phase of glomerulonephritis. The protection can be linked to the ability of activated A_2AR to suppress infiltration of leukocytes, the key mediators of kidney damage.

We found that A_2AR activation alters the expression of chemokines. Specifically, pharmacological activation of A_2AR suppresses MDC/CCL22 expression and markedly reduces the expression of other chemokines during the acute inflammatory phase of glomerulonephritis (Figure 6). These changes in chemokine expression can explain the decreased infiltration of CD8⁺ and macrophages (6, 9–11). During the progressive phase of glomerulonephritis (day 6 to day 11), we found suppressed MDC/CCL22 induction while the expression of other chemokines (except for fractalkine/CX₃CL1) was not modified. These results are consistent with the report that MDC/CCL22 plays a critical role in stimulating the influx of macrophages into glomeruli in anti-GBM glomerulonephritis (11). Notably, MDC/CCL22 expression in anti-GBM glomerulonephritis is confined to macrophages (11). As shown in Figure 1, at day 7 after injecting the anti-GBM antibody, macrophages are the major cellular site of A_2AR expression in nephritic glomeruli. When MDC/CCL22 expression is inhibited by the A_2AR agonist, CGS21680, there also is a dramatic inhibition of ED1⁺ macrophage infiltration. These results raise the following possibilities: 1) Activation of macrophage-specific A_2AR suppresses MDC/CCL22 expression resulting in decreased macrophage infiltration in the acute phase of glomerulonephritis; and 2) macrophage infiltration in later stages of glomerulonephritis depends on MDC/CCL22 because blocking expression of this chemokine ameliorates the degree of functional damage. Interestingly, A_2AR activation in the progressive phase of glomerulonephritis did not alter the expression of other chemokines, except for fractalkine/CX₃CL1. Possibly increased fractalkine/CX₃CL1 expression occurs because of endothelial cell activation in the nephritic glomeruli. This interpretation is suggested because fractalkine/CX₃CL1 is mainly expressed by endothelial cells and we have shown that fractalkine/CX₃CL1 is induced in the endothelium of nephritic glomeruli (10). In contrast, we have not found A_2AR in cultured glomerular endothelial cells (quiescent or activated) or HUVEC (quiescent). Thus, a plausible explanation for the finding that A_2AR occupancy during the progressive phase of glomerulonephritis does not alter the expression of other chemokines, is that the chemokines at day 6 to day11 could be produced by A_2AR-negative kidney cells.

Our finding that activation of A_2AR did not suppress the expression of pro-inflammatory cytokines, IL-1β and TNF-α, (Figure 5 a and d) is consistent with the notion that macrophages are not the predominant source of IL-1β and TNF-α in this model of glomerulonephritis (51, 52). On the other hand, in vitro studies indicate that IL-1β and TNF-α can up-regulate A_2AR in THP-1 cells, human dermal microvascular endothelial cells, and PC12 cells (35, 53, 54). Our results, therefore, suggest that IL-1β and TNF-α could induce A_2AR expression in other cells and contribute to the beneficial effect of A_2AR agonists in preventing glomerular damage.

In addition, we found that the expression of the anti-inflammatory cytokines, IL-10 and/or IL-4, was enhanced by the A_2AR activation during anti-GBM glomerulonephritis. These cytokines might also contribute to the beneficial effects of A_2AR activation on kidney injury (55, 56).

Increasing evidence suggest that A_2AR tissue-protection requires activation of receptors expressed on bone marrow-derived cells (57, 58). Our data indicate that macrophages are the target cells mediating the benefits of A_2AR activation in both the acute and progressive phases of glomerulonephritis. First, A_2AR expression was increased in macrophages but not in CD8⁺ cells during anti-GBM GN (Figure 1B and 1C). Second, macrophage derived-chemokines such as MDC/CCL22, MCP-1/CCL2, RANTES/CCL5, and MIP-1α/CCL3 were decreased in response to A_2AR agonist. Third, chemokines not expressed in macrophages (e.g., fractalkine/CX3CL1) were unaffected by the A_2AR agonist. Finally, macrophages are present in the crescents and in progressive glomerulonephritis and are correlated with the degree of glomerular injury (5, 6, 8). It has been found that depletion of macrophages in this and other models of glomerulonephritis suppresses glomerular crescent formation and reduces proteinuria (3, 7, 59). Accordingly, suppression of the expression of MDC/CCL22 and reduced expression of other chemokines when macrophage-A_2AR activation occurs could explain in part the beneficial effect of CGS21680 in both acute and progressive phases of anti-GBM glomerulonephritis.

Notably, we found that the A_2AR agonist also was beneficial in the late phase of anti-GBM glomerulonephritis when macrophage influx and crescent formation were at their maximum (day 6) (13, 60). These responses were associated with a significant reduction in macrophage infiltration (74%) suggesting that activation of A_2AR on macrophages was responsible for the suppression of glomerular injury.

To examine mechanisms underlying the beneficial effects of A_2AR we determined the expression of α-SMA, a marker for mesangial cell injury and activation (49)and α-Actinin-4, a marker for podocytes, in glomeruli of rats with anti-GBM glomerulonephritis. Earlier reports have demonstrated that changes in podocytes contribute to the severity of anti-GBM glomerulonephritis (48, 61, 62). We found that the expression of α-SMA is increased and that α-Actinin- 4 expression is reduced in anti-GBM GN. Treatment with A_2AR agonist significantly decreased α-SMA expression and restored the expression of α-Actinin- 4. In contrast, A_2AR antagonist enhanced the expression of α-SMA and increased the loss of the podocyte marker, α-Actinin- 4, compared to the controls rats. These data suggest that A_2AR agonist may attenuate kidney injury through effects on mesangial cells and podocytes. Future experiments will determine if A_2AR activation prevents mesangial cells injury and restores α-Actinin-4 by reducing inflammation or through a direct effect on mesangial cells and podocytes.

A_2AR activation or inactivation did not modify blood pressure in rats with anti-GBM glomerulonephritis. Earlier reposts in mice lacking A_2AR provide inconsistent changes in blood pressure (46, 50). The inconsistent results could reflect differences in strains of mice, administration of A_2AR agonists, conscious vs. anesthetized animals, etc. Regardless, the effects of the A_2AR agonist and antagonist we found are not related to changes in blood pressure.

Prolonged exposure to adenosine agonists could produce desensitization of A_2AR (63–65). In certain reports, desensitization was dose-dependent and reversible; suggested mechanisms for desensitization have included down-regulation of the expression of A_2AR or of Gs proteins involved in the cellular signaling or phosphorylation of the A_2A, G protein-coupled receptor (GPCR) by GPCR kinases. There are also reports, however, providing evidence that prolonged stimulation of A_2AR does not lead to loss of functional response to adenosine agonists (66, 67). In our experiments, treatment with adenosine agonists increased expression of A_2AR in glomeruli of rats with anti-GBM glomerulonephritis, providing a mechanism that would prevent desensitization (Supplementary results). Interestingly, in rats treated with ZM241385, we found decreased expression of A_2AR. This finding could eliminate responses to endogenous adenosine-induced activation of A_2AR and lead to additional kidney damage.

In summary, we have demonstrated that pharmacological activation of A_2AR attenuates leukocyte infiltration into the kidney during both the acute and progressive phases of anti-GBM glomerulonephritis. Subsequent responses include decreased expression of chemokines and upregulation of anti-inflammatory cytokines with suppressed inflammation. These events are associated with marked protection against glomerular injury. Consequently, A_2AR activation could be a novel therapeutic strategy for modifying macrophage-mediated glomerulonephritis or other diseases dependent on macrophages.