Introduction

Top Summary Introduction Procedures Results & Discussion References

Adenosine, released under conditions of oxidative stress, confers cytoprotection in the cardiovascular and central nervous systems by activating cell surface adenosine receptors. This nucleoside provides rapid recovery of ventricular functioning induced after ischemia and reperfusion (Babbit et al., 1989) and can effect a degree of protection similar to that observed after preconditioning (Liu et al., 1991). In the central nervous system, adenosine and adenosine analogs protect against transient ischemia (Rudolphi et al., 1992). Rats treated with caffeine to increase A₁AR expression in the brain were more resistant to ischemia, underscoring a cytoprotective role of this receptor subtype (Rudolphi et al., 1989). In contrast, down-regulation of the A₁AR by prolonged agonist treatment exacerbated the damage created by a subsequent ischemic episode (Von Lubitz et al., 1994). Several mechanisms contributing to the cytoprotective role of adenosine have been proposed. Adenosine hyperpolarizes neurons by activating K⁺ conductance and inhibits Ca²⁺ influx into the nerve terminal (Rudolphi et al., 1992). These two functions probably account for its ability to reduce the release of excitatory amino acids. Activation of A₁AR depresses neuronal firing rate and decreases neuronal metabolism, thereby promoting the preservation of ATP during ischemia (Rudolphi et al., 1992).

Chemotherapeutic agents such as doxorubicin, daunomycin, mitoxantrone, bleomycin, and cisplatin promote oxidative stress by generating ROS (Powis, 1991). Anthracyclines, such as doxorubicin and daunorubicin, produce superoxide anions through redox cycling of these compounds in the mitochondria and hydroxyl radicals after reaction of superoxide anions with iron. The generation of ROS likely contributes to both the antitumor effects and side effects associated with these agents by potentiating DNA alkylation and strand breaks (Berlin and Haseltine, 1981). In fact, the generation of ROS may account for cisplatin-induced ototoxicity and nephrotoxicity and the cardiotoxicity produced by doxorubicin and daunorubicin (Myers et al., 1977; Powis, 1991).

In the present study, we provide evidence that chemotherapeutic agents that generate ROS can directly influence A₁AR expression through activation of NFB. Such a pathway provides an efficient mechanism by which the cell could counter the detrimental effects of ROS generated by chemotherapeutic agents or during ischemia.

	Experimental Procedures

Top Summary Introduction Procedures Results & Discussion References

Cell cultures and radioligand binding. DDT₁MF-2 cells were cultured in Dulbecco's modified Eagle's medium supplemented with 10% fetal bovine serum and 5% calf serum. Cells were maintained as a monolayer and detached in ice-cold phosphate-buffered saline containing 5 mM EDTA. The cells were then lysed in 10 mM Tris·HCl buffer, pH 7.4, containing 5 mM EDTA, 10 µg/ml soybean trypsin inhibitor, 10 µg/ml benzamidine, and 2 µg/ml pepstatin and homogenized briefly by Polytron homogenizer. Membranes were obtained by differential centrifugation, which involved centrifugation of the homogenates at 1,000 × g for 10 min followed by centrifugation of the supernatant at 40,000 × g for 15 min. The final pellet was resuspended in 50 mM Tris·HCl buffer, pH 7.4, 10 mM MgCl₂, and 1 mM EDTA, containing protease inhibitors (described above) and pretreated with adenosine deaminase (5 units/ml) to eliminate endogenous adenosine. Quantification of A₁AR was performed using the antagonist [³H]DPCPX or the agonist radioligand ¹²⁵I-APNEA. ¹²⁵I-APNEA was diluted 10-fold with [I]APNEA to enable use of higher effective concentrations of the radioligand. Membrane preparations (approximately 80 µg of protein) were incubated with the radioligands in the absence or presence of 1 mM theophylline (to define nonspecific binding) for 1 hr at 37°. Samples were then filtered through polyethyleneimine-treated Whatman GF/B glass fiber filters, washed with 10 ml of ice-cold Tris buffer, and radioactivity determined by gamma or  counting. Saturation curves were analyzed by a computer-assisted curve-fitting program (Hancock et al., 1979; DeLean et al., 1982).

Western blotting. For quantifying the A₁AR, cell membranes were solubilized in 1% CHAPS buffer, desalted into 0.05% CHAPS, and 100-200 µg of protein samples were used for performing sodium dodecyl sulfate polyacrylamide gel electrophoresis. NFB and c-Rel were quantified in nuclear extracts. Proteins were transferred to nitrocellulose membranes, blocked in Blotto buffer (130 mM NaCl, 2.7 mM KCl, 1.8 mM Na₂HPO₄, 1.5 mM KH₂PO₄, 0.1% NaN₃, 0.1% Triton-X 100, and 5% low-fat skim milk) for 2 hr and then incubated with the primary antibody at 4° overnight. Polyclonal antibody for the A₁AR was obtained from Alpha Diagnostic International (San Antonio, TX), and polyclonal antibodies for NFB and c-Rel were obtained from Santa Cruz Biotechnology (Santa Cruz, CA). After five washes in blocking solution, blots were incubated with ¹²⁵I-labeled goat antirabbit IgG for 2 hr at room temperature, washed five times, and exposed to Kodak XAR film at 70° for about 48 hr. Quantification of the gels (¹²⁵I counts) was performed using a PhosphorImager (GS-250 Molecular Imager; BioRad, Hercules, CA) and by cutting and counting the labeled protein bands.

Electrophoretic mobility shift assay. Nuclear extracts were incubated with double stranded specific B oligonucleotide (5'-ATGTGAGGGGACTTTCCCAGGC-3') (Sen and Baltimore, 1986). Similar electrophoretic mobility shift assays were performed using a labeled oligonucleotide probe (5'-CGCTTGATGAGTCAGCCGGAA-3') for the AP-1 transcription factor binding sequence. Incubations were performed at room temperature for 30 min in a total volume of 15 µl of buffer containing 12% glycerol, 12 mM HEPES-NaOH (pH 7.9), 60 mM KCl, 1 mM EDTA, and 1 mM dithiothreitol, 1.0 µg of poly(dI-dC), and 10,000 cpm of the labeled probe. The DNA-protein complexes were resolved on nondenaturing 5% polyacrylamide gels, performed with 0.5 × Tris/borate/EDTA buffer (4.5 mM Tris, 4.5 mM boric acid, 0.1 mM EDTA, pH 8.0). The fold increase in NFB expression was determined with a PhosphorImager using background substract.

Luciferase assay. Cells were grown to a density of about 20-40% confluency and transfected with a mixture containing 100-250 ng of plasmid DNA, 500-650 ng of carrier DNA, 3 µl/g DNA of N-[1-(2,3-dioleoyloxy)propyl]-N,N,N-trimethylammonium methylsulfate (lipofectin) in a volume of 50 µl Opti-MEM (GIBCO BRL, Gaithersburg, MD). The mixtures were allowed to stand for 45-60 min at room temperature before being added to the culture plate. After ~6 hr, regular media (Dulbecco's modified Eagle's medium, 10% fetal bovine serum) were added to the plate and it was returned to the incubator for 24 hr. For luciferase assays, cells were then lysed by using 50 µl of Reporter lysis buffer (Promega, Madison, WI) and centrifuged at 4° in a microcentrifuge at 12,000 × g. The extract was used immediately or stored at 70°. Twenty microliters of cell extract (supernatant) was mixed with 100 µl of Luciferase assay reagent (Promega) at room temperature and the chemiluminescent signal determined in a luminometer (1 min counts).

Protein determination and statistical analysis. The level of protein in samples was determined by the established method (Bradford, 1976), using bovine serum albumin to prepare standard curves. Statistical differences were determined by the analysis of variance, followed by post hoc analysis using the Tukey's test.

	Results and Discussion

Top Summary Introduction Procedures Results & Discussion References

DDT₁MF-2 smooth muscle cells have been used previously as a model system for studying A₁AR (Ramkumar et al., 1991). Treatment of DDT₁MF-2 cells with cisplatin (2.5 µM) produced an approximately 2-fold increase in A₁AR expression, determined by the antagonist radioligand [³H]DPCPX. Levels of these receptors were increased from 389 ± 79 to 844 ± 250 fmol/mg protein by cisplatin (Fig. 1a). The equilibrium dissociation constants (K_d values) were 3.3 ± 1.2 nM and 5.1 ± 1.0 nM, respectively, for control and cisplatin-treated cells. A similar increase in A₁AR was obtained using ¹²⁵I-APNEA, an agonist radioligand for the A₁AR (Fig. 1, c-e). Western blotting studies performed on these membranes using a selective antibody for the A₁AR demonstrated increased levels of immunoreactivity (36-38 kDa band) in preparations derived from cisplatin-treated cells (Fig. 1b), which coincide with the size of the A₁AR by photoaffinity labeling. These data suggest that the increase in ¹²⁵I-APNEA binding probably represents an increase in total receptor number and not enhanced coupling to G proteins. In support of this, no change in the expression of G_i protein subunits was detected in cells treated with cisplatin for 24 hr (data not shown). The dose of cisplatin producing half maximal increase in radioligand binding (EC₅₀) was approximately 300 nM, whereas maximal increase in receptor levels was obtained with 2.5 µM cisplatin (Fig. 1c). With use of the latter concentration of cisplatin, optimal increases in receptors were obtained after 24 hr of treatment. Other chemotherapeutic agents known to generate free radicals were similarly effective in increasing the levels of A₁AR in these cells (Fig. 1d). The exception was transplatin, a platinum analog devoid of antitumor activity. Cells incubated with hydrogen peroxide (10 µM) for 24 hr also demonstrated increases in A₁AR levels, albeit to a lesser degree, suggesting a direct role of ROS in this process (Fig. 1d). Coincubation of cells with H₂O₂ and the antioxidant enzyme catalase blocked the increase in receptor expression. Administration of catalase alone led to reduction in receptor expression, supporting a role of ROS in the up-regulation of A₁AR. The increase in receptor level produced by cisplatin was attenuated after pretreatment with actinomycin D (4 µM) or cycloheximide (5 µg/ml) before administering cisplatin, suggesting a requirement for de novo synthesis of A₁AR in this process. Interestingly, although treatment with actinomycin D for 24 hr did not affect the basal expression of these receptors (110 ± 4% of control), administration of cycloheximide for the same period reduced receptor expression to 38 ± 1% of control, suggesting a fairly rapid rate of turnover of the A₁AR protein.

View larger version (26K): [in this window] [in a new window]   Fig. 1.   Cisplatin for 24 hr induced up-regulation of A1AR in DDT1MF-2 cells. Cells were treated with cisplatin (2.5 µM) for 24 hr, after which crude plasma membranes were prepared for determining A1AR expression using [3H]DPCPX or 125I-APNEA. a, Saturation curves showing increase in A1AR expression in DDT1MF-2 cells after cisplatin treatment. Saturation curves and receptor levels were analyzed by computer as described previously (Hancock et al., 1979; De Leon et al., 1982). b, Western blotting for the A1AR using a polyclonal antisera at a titer of 1:1000. Molecular weight markers are indicated on the left. c, Dose-response effect of cisplatin obtained with cells treated with cisplatin for 24 hr. In this and subsequent experiments (d, e), a single, nonsaturating concentration of 125I-APNEA (1 nM) was used. d, Up-regulation of A1AR by different chemotherapeutic agents and H2O2. Cells were incubated with 1 µM of the chemotherapeutic agents or 10 µM H2O2 or 200 units/ml catalase for 24 hr. e, Inhibition of NFB by PDTC, dexamethasone, or genistein. This treatment abrogated the stimulatory effect of cisplatin and H2O2 on A1AR expression. Cells were treated with 200 µM PDTC, 100 nM dexamethasone, or 1 µM genistein. *, statistically significant difference from control; **, statistically significant difference from the cisplatin-treated group, using analysis of variance followed by post hoc analysis using Tukey's test. Experiments presented were repeated at least three times.

Chemotherapeutic agents such as cisplatin (Ravi et al., 1995; Rybak et al., 1995; Sasada et al., 1996) and daunorubicin (Wang et al., 1997) mediate their toxicities, in part, by generating ROS (Rybak et al., 1995). In the case of daunorubucin, the generation of ROS presumably leads to activation of transcription factor NFB (Beg and Baltimore, 1997). To test whether the increase in A₁AR induced by cisplatin was mediated through NFB, the response to cisplatin was tested in cells pretreated with PDTC, an inhibitor of NFB (Sun et al., 1995). PDTC abrogated the increase in A₁AR expression induced by cisplatin (Fig. 1e). Similarly, inhibition of NFB by dexamethasone (Mukaida et al., 1994) also attenuated the response to cisplatin (Fig. 1e). PDTC was also effective in blocking the increase in A₁AR elicited by H₂O₂ (Fig. 1e), supporting the role of NFB in the latter response. Additional experiments were performed by incubating cells with genistein, a different inhibitor of NFB. This agent significantly reduced cisplatin-induced increase in A₁AR expression from 168 ± 15% to 128 ± 9% of control (three experiments). Administration of genistein alone had little effect on the expression of the A₁AR. The level of binding was 101 ± 3% of control. These experiments clearly underscore a role for NFB in the regulation of A₁AR expression.

To determine whether exposure to cisplatin promotes activation of NFB in DDT₁MF-2 cells, nuclear translocation of the p65 DNA binding subunit of NFB was determined in Western blotting studies. Cisplatin increased the levels of p65 in the nucleus by 2- to 3-fold (Fig. 2a). Similar increases in the levels of c-Rel p75, a member of the NFB family (Baeuerle, 1991; Baeuerle and Baltimore, 1996), were observed after cisplatin treatment (Fig. 2b). To test whether the increase in NFB in the nucleus was associated with increased DNA binding activity, electrophoretic mobility shift assays were performed using an end-labeled NFB DNA binding oligonucleotide (5'-ATGTGAGGGGACTTTCCCAGGC-3') (Sen and Baltimore, 1986). Retardation of two bands was observed (Fig. 2c), the appearance of which was blocked by the addition of 30- and 10-fold excess (0.6 and 0.2 pmol, respectively) of cold B probe (Fig. 2c, lanes 8 and 9). The addition of p65 antibody (1.0 and 0.5 µg/ml) to the reaction mixture (Fig. 2c, lanes 10 and 11, respectively) resulted in supershift of the upper band (Fig. 2c, lane 11), suggesting the presence of p65 in this complex. It is not clear why no supershifted band was detected in lane 10. Antibody against c-Rel (1 µg/ml) failed to significantly reduce DNA binding activity, possibly implying lower levels of this protein in the NFB complexes. Administration of cisplatin to the DDT₁MF-2 cells increased DNA binding activities of both by approximately 4-fold (Fig. 2c, compare lanes 2-4 with lanes 5-7). Similar electrophoretic mobility shift assays performed using a labeled oligonucleotide probe for the AP-1 transcription factor binding sequence indicate no change in the levels of these AP-1 factors after cisplatin treatment (Fig. 3). The intensities of the AP-1 labeled complexes were similar in the control versus cisplatin-treated cells (Fig. 3, compare lanes 2-4 with lanes 8-10). Inhibition of the labeled probe could be detected with increasing concentrations of cold AP-1 oligonucleotide (Fig. 3, lanes 5-7). This suggests that cisplatin induction of A₁AR expression is mediated via NFB and not by AP-1 transcription factors.

View larger version (53K): [in this window] [in a new window]   Fig. 2.   Nuclear translocation of NFB by cisplatin in DDT1MF-2 cells. Cells were exposed to cisplatin (2.5 µM) for 2 hr and the nuclear translocation of NFB complex was determined. Western blotting of p65 (a) and p75 (b) in the nucleus after exposure to cisplatin. Increase in NFB (c) as detected in electrophoretic mobility shift assays. Lanes 2-4 represent nuclear extracts from control cells, whereas lanes 5-12 are extracts from cisplatin-treated cells. Two complexes, denoted NFB, were detected. Note the increase in the binding of the B probe in the cisplatin-treated cells (compare lanes 2-4 with lanes 5-7). Binding of the labeled B probe was inhibited by a 30- and 10-fold molar excess of cold B probe (lanes 8 and 9). Samples were incubated with 1.0 µg/ml (lane 10) or 0.5 µg/ml (lane 11) of p65 antibody for supershift experiments. , position of the supershifted band (lane 11).

View larger version (54K): [in this window] [in a new window]   Fig. 3.   Cisplatin treatment did not increase AP-1 binding activity in the nucleus. Cells were treated with cisplatin (2.5 µM) and nuclear extracts were prepared as described in Results and Discussion. Incubation of preparations from control and cisplatin-treated cells with labeled AP-1 probe indicates no difference in binding (compare lanes 2-4 with lanes 8-10). Labeling of the labeled complex is inhibited in the presence of 30-, 15-, and 5-fold molar excess of cold AP-1 probe (lanes 5-7).

Analysis of the human A₁AR promoter sequence (Ren and Stiles, 1995) using Transfac (Quandt et al., 1995; Wingender et al., 1997) indicates the presence of consensus sequences for both NFB and AP-1 transcription factors. A consensus sequence for NFB (5'-GGAAGTCCC-3') was detected 623 base pairs upstream of promoter A transcription start site in plasmid construct pBLPnif/PmtA (Ren and Stiles, 1995). An oligonucleotide probe constructed incorporating the sequence 5'-AGCAAGGGACTTCCGGAGGT-3' was used as competitor for labeled B probe in electrophoretic mobility shift assays. Results presented in Fig. 4a indicate inhibition of binding of the labeled probe with increasing concentrations (0.3-1.2 pmol) of A₁AR oligonucleotide (15- to 60-fold molar excess). Labeling of both the upper and lower bands were inhibit (Fig 4a, compare lane 2 with lanes 3-5), whereas no change in intensity of the nonspecific band was detected. As shown before, unlabeled B probe effectively competed with the labeled probe (Fig 4a, lanes 6 and 7) whereas no inhibition was observed using a 30-fold molar excess of unlabeled AP-1 oligonucleotide (Fig 4a, lane 8). To test whether the increase in nuclear translocation of NFB could increase activity of the A₁AR gene promoter, DDT₁MF-2, cells were transiently transfected with pBLPnif/PmtA, which contains the firefly luciferase reporter gene driven by the A₁AR promoter (Ren and Stiles, 1995). Transfectional efficiency in each well was determined by cotransfecting a pCMV-CAT construct and assaying for CAT activities. Fig. 4b indicates a 2- to 3-fold increase in luciferase activity when these cells were treated with cisplatin for 24 hr. Because these experiments were performed with additional stress to the cells due to transfection, it is likely that this stress contributes to higher basal luciferase activity. To test whether the high basal activity was caused by ROS, culture media from control and cisplatin-treated cells were supplemented with either superoxide dismutase or catalase. Significant inhibition of luciferase activity was obtained in the presence of catalase but not superoxide dismutase, suggesting a role of H₂O₂ and not O₂ in the regulation of A₁AR gene activity. Furthermore, inhibition of NFB by PDTC led to significant reductions in the activity of the A₁AR promoter, supporting a role for NFB in the maintenance of normal A₁AR expression under control conditions. When cisplatin-stimulated luciferase activity was compared with the basal activity defined by PDTC (200 µM), a 10- to 15-fold increase in luciferase activity was obtained (Fig. 4b). DDT₁MF-2 cells transfected with plasmid (pBLPnif/PmtB) containing promoter B and a putative NFB consensus sequence (5'-GGGGCTCCCC-3') located 306 base pairs upstream of the second transcription start site (Ren and Stiles, 1995) did not exhibit any cisplatin-stimulated luciferase activity. These data suggest that while NFB plays a significant role in regulating A₁AR gene activity, it is likely the consensus sequence upstream of promoter A that is important in this regulation.

View larger version (36K): [in this window] [in a new window]   Fig. 4.   The A1AR promoter (promoter A) contains an NFB regulatory site whose activity is induced by cisplatin. a, An A1AR-derived B oligonucleotide competing with the labeled probe for binding. Increasing concentrations of cold A1AR oligonucleotide (lanes 3-5) reduced binding of both upper and lower NFB bands. Inhibition of these two bands was also observed with cold B probe (lanes 6, 7), but not with AP-1 probe (lane 8). b, Cisplatin-induced activation of luciferase activity. DDT1MF-2 cells were transfected with plasmid (pBLPnif/PmtA) that contains the A1AR promoter (promoter A) coupled to the firefly luciferase reporter gene (Ren and Stiles, 1995). Luciferase activity was determined in controls or in cells exposed to cisplatin for 24 hr in the absence or presence of PDTC (200 µM) or catalase (200 units/ml).

If the increases in A₁AR obtained after cisplatin treatment were a compensatory response to increased oxidative stress, one would expect modulation of receptor activation to regulate the degree of oxidative stress the cell experiences. To test whether this increase in A₁AR confers protection in these cells, the levels of MDA, a marker of lipid peroxidation, were determined under different culture conditions (Ohkawa et al., 1978). As shown in Fig. 5, cisplatin treatment for 24 hr led to an approximately 2-fold increase in MDA levels, which was reduced after activation of the A₁AR by 1 µM (R)-PIA. This protective effect of (R)-PIA was blocked after the addition of 1 µM of DPCPX, suggesting a cytoprotective role of the A₁AR in these cells. The reduction in MDA was also blocked after incubation of the cells with 100 µM theophylline, a nonselective antagonist of the A₁AR.

View larger version (20K): [in this window] [in a new window]   Fig. 5.   Cytoprotective role of the increase in A1AR induced by cisplatin. DDT1MF-2 cells were treated with cisplatin (2.5 µM) for 24 hr, with or without the various drugs listed. The concentrations of (R)-PIA, DPCPX, and theophylline were 1 µM, 1 µM, and 100 µM, respectively. The results presented are the mean ± standard error of six independent experiments. Basal level of MDA in the control was 0.15 ± 0.05 nmole/mg protein. The data are presented as the mean ± standard error of three to five experiments. *, statistically significant difference (p < 0.05) from control. **, statistically significant difference from cisplatin-treated group.

Taken together, this study provides strong evidence for a direct role of NFB in the expression of the A₁AR. Up-regulation of the A₁AR by this mechanism provides a means by which ischemic tissues can optimize the cytoprotective role of adenosine. Furthermore, the increase in A₁AR expression during ischemia may compensate for any down-regulation of this receptor (Lee et al., 1986) resulting from a substantial rise in adenosine levels (Berne and Rubio, 1974). Down-regulation of the A₁AR during ischemia would limit the cytoprotective role of adenosine. These data suggest a paracrine role for ROS, regulating the expression of the A₁AR and providing cytoprotection to cells in the local environment. For the cancer cells, increase in the A₁AR may aid in the development of resistance to chemotherapeutic agents.

The present data support the notion that NFB acts as a sensor of the oxidative status of the cell and regulates the expression of the A₁AR accordingly. As such, the extracellular environment can directly regulate the levels of this G protein-coupled receptor. Accordingly, one could titrate the expression of the A₁AR by altering the level of ROS in the extracellular environment. Such a close regulation of receptor expression could aid in cell survival during prolonged periods of exposure to ROS. Our data indicate that ROS and NFB contribute significantly to A₁AR gene activity (at least for promoter A) because substantial decreases in activity were observed in the presence of catalase and after inhibition of this transcription factor.

Although cisplatin treatment led to significant changes in NFB activity, there was surprisingly little change in AP-1 as assessed by electrophoretic mobility shift assays. Several AP-1 consensus sequences have been identified in the A₁AR promoter region (Ren and Stiles, 1995). Thus, it is possible that these sites are not important in ROS-mediated increases in A₁AR expression but may regulate receptor expression induced by other stresses. Alternatively, AP-1 sites may be important in maintaining a baseline level of A₁AR expression.

The administration of cisplatin is associated with significant oto- and nephrotoxicity in animals and humans (Rybak et al., 1995). In rats and chinchillas, the administration of cisplatin resulted in significant elevation in A₁AR in the cochleas and kidneys (Ford et al., 1997). Because the A₁AR mediates renal vascular constriction (Macias et al., 1983), it is possible that the renal toxicity produced by cisplatin is mediated, in part, by up-regulation of this receptor subtype. Accordingly, blockade of these receptors with selective antagonists may alleviate some of the renal toxicity produced by this chemotherapeutic agent. Furthermore, activation of these A₁AR through round window application of selective agonists may alleviate the ototoxic effect of cisplatin in cancer patients.