Regulation of Rat Hepatocyte Function by P2Y Receptors: Focus on Control of Glycogen Phosphorylase and Cyclic AMP by 2-Methylthioadenosine 5′-Diphosphate

Glycogen phosphorylase, the rate-controlling enzyme in hepatic glycogenolysis, is activated by increases in both cyclic AMP and cytosolic free Ca²⁺ concentration ([Ca²⁺]_c), resulting in a net output of glucose to supply extrahepatic tissues, crucially the brain, when blood glucose levels drop. Extracellular nucleotides play a well established role in the regulation of this key function in rat hepatocytes through the activation of P2Y receptors (Okajima et al., 1987; Keppens et al., 1992, 1993; Keppens, 1993), a family of G-protein-coupled receptors responding to the native nucleotides ADP, ATP, UDP, UTP, and UDP-glucose (Boarder and Hourani, 1998; Abbracchio et al., 2003). Curiously, stimulation of P2Y receptors on rat hepatocytes leads to both increases in [Ca²⁺]_c and inhibition of adenylyl cyclase; these responses will stimulate and limit, respectively, activation of glycogen phosphorylase (Okajima et al., 1987; Dixon et al., 1990, 1995, 2000; Keppens et al., 1992; 1993; Keppens, 1993; Edgecombe et al., 1999). The mechanism underlying the regulation of glycogen phosphorylase by cyclic AMP is well established, with increased cyclic AMP levels leading to phosphorylation and activation of the enzyme, and with dephosphorylation by protein phosphatase 1 as cyclic AMP levels fall. Little is understood of P2Y receptor regulation of cyclic AMP in hepatocytes, despite its significance in the control of glycogen phosphorylase. An understanding of the nature of the P2Y receptors present on hepatocytes and their influence on cell function is of considerable physiological relevance since it is known that native nucleotides reach micromolar concentrations in the extracellular compartment of the liver (Chari et al., 1996). Furthermore, release of ATP by hepatocytes and activation of P2Y receptors on neighboring cells has been shown to play a role in intercellular communication in the liver (Schlosser et al., 1996; Wang et al., 1996).

In common with many other cell types, rat hepatocytes are known to express multiple P2Y subtypes, although these have not been conclusively identified (Dixon et al., 1990, 1995, 2000, 2003; Keppens 1993). Reverse transcriptase-polymerase chain reaction (RT-PCR) has demonstrated the presence in primary rat hepatocytes of mRNA transcripts encoding all four of the cloned rat P2Y subtypes investigated (P2Y₁, P2Y₂, P2Y₄, and P2Y₆; Dixon et al., 2000). These receptors are coupled to increases in [Ca²⁺]_c. We have previously argued that rat hepatocytes express functional P2Y₁ and P2Y₂ receptors mediating the effects of ADP (P2Y₁) and ATP and UTP (P2Y₂), with a possible contribution from P2Y₄ receptors (Dixon et al., 2000, 2003). The P2Y₆ receptor is apparently not expressed as a functional receptor, since UDP, which is the most potent agonist at this receptor (Nicholas et al., 1996), failed to elicit a [Ca²⁺]_c response in rat hepatocytes (Dixon et al., 2000).

The most recently characterized members of the P2Y receptor family are coupled to changes in cyclic AMP levels. Thus, the P2Y₁₁ receptor is coupled positively through G_s to adenylyl cyclase, in addition to the phosphoinositide pathway (Communi et al., 1999; Zambon et al., 2001). P2Y₁₂, P2Y₁₃, and P2Y₁₄ receptors are coupled negatively to adenylyl cyclase through G_i (Communi et al., 2001; Hollopeter et al., 2001; Zhang et al., 2001, 2002; Abbracchio et al., 2003).

2-MeSADP is a partially selective P2Y agonist, acting at P2Y₁, P2Y₁₂, and P2Y₁₃ receptors (Communi et al., 2001; Hollopeter et al., 2001; Zhang et al., 2002). Here we demonstrate that 2-MeSADP both stimulates glycogen phosphorylase and inhibits cyclic AMP in rat hepatocytes and have investigated the hypothesis that this occurs by action at distinct P2Y receptors.

Materials and Methods

Cell Preparation and Stimulation. Hepatocytes were isolated from fed, male Wistar-strain rats (150–250 g) by collagenase perfusion as described previously (Dixon et al., 1995). Briefly, the rat was killed by cervical dislocation in accordance with institutional guidelines, and the hepatic portal vein was cannulated. An initial Ca²⁺-free perfusion was followed by perfusion with collagenase (0.04%, w/v) and Ca²⁺ (3.8 mM) for 15 min. The perfusion rate was 30 ml min^-1 throughout. The cells were harvested and resuspended in Hepes-buffered medium at the appropriate density (116 mM NaCl, 5.6 mM KCl, 0.8 mM MgSO₄, 1.1 mM KH₂PO₄, 4.8 mM NaHCO₃, 11 mM glucose, 1.8 mM CaCl₂, 20 mM Hepes, 2% BSA w/v, pH 7.4).

Cell Culture. Isolated cells were cultured for 24 h at a density of 1 × 10⁵ cells well^-1 in 24-well plates in William's medium E with 10% FCS, 50 IU ml^-1 penicillin, 50 μg ml^-1 streptomycin, and 50 μg ml^-1 gentamicin, in a volume of 0.5 ml. Cells were then cultured for a further 24 h without FCS. For experiments investigating the effect of pertussis toxin (100 ng ml^-1), cells were cultured for a total of 24 h with the toxin added, where appropriate, at the same time that the cells were plated out.

Cyclic AMP Measurement. Cultured cells or freshly isolated cells resuspended at 2 × 10⁶ cells ml^-1 were treated with 300 μM IBMX and P2Y antagonists where applicable for 10 min before agonist addition. Aliquots of freshly isolated cells were dispensed into a 24-well plate (0.25 ml per well) and shaken at 37°C. Cells were stimulated by coaddition of nucleotide and glucagon (10 nM final concentration) for 3 min, and incubations were terminated by addition of trichloroacetic acid. Samples were left on ice for 60 min, extracted with a 1:1 mixture of tri-n-octylamine and 1,1,2-trichlorotrifluoroethane, neutralized with 250 mM NaHCO₃, and assayed for cyclic AMP content using the protein binding assay of Brown et al. (1971).

Glycogen Phosphorylase Assay. Cells were resuspended at 4 × 10⁶ cells ml^-1 and treated with MRS 2179 for 10 min where applicable. Aliquots of 0.25 ml were dispensed into each well of a 24-well plate and shaken at 37°C, and cells were stimulated with 2-MeSADP for 2 min. Incubations were terminated by removing 100 μl of cell suspension and adding to 100 μl of extraction buffer [400 mM NaF, 80 mM EDTA, 2% (w/v) glycogen, 200 mM glycyl glycine, pH 7.4) and freezing immediately in liquid nitrogen. Cell extracts were subjected to three cycles of freeze-thawing and centrifuged at 10,000g for 1 min at 4°C. A 100-μl portion of supernatant was incubated with 20 μl of assay buffer (600 mM glucose 1-phosphate, 1.8 M NaF, 6 mM caffeine, 12% (w/v) glycogen, 29.6 kBq ml^-1 [¹⁴C]glucose 1-phosphate, pH 6.1) for 40 min at 30°C. The reaction was terminated by spotting 100 μl of the assay mixture onto filter papers that were washed in 66% ethanol for 1.5 h, briefly soaked in acetone, and dried before being counted for radioactivity.

Calculation of Antagonist Potency. An estimate of the pA₂ value for MRS 2179 was calculated with a rearrangement of the Schild equation: pA₂ = -log₁₀ ([B]/CR - 1), where [B] is the concentration of agonist tested and CR is the concentration ratio of curve midpoint in the presence and absence of antagonist. In the absence of a full Schild analysis, this is referred to as an apparent pA₂.

High-Pressure Liquid Chromatography (HPLC) Analysis. HPLC analysis was used to determine whether 2-MeSADP and antagonists were broken down during the incubation period. An Alltech SAX column (Alltech Associates, Deerfield, IL) was eluted with a gradient running from A, 0.05 M, to B, 1 M K₂HPO₄ of 100% A for 10 min, to 20% B for 10 to 12.5 min, to 70% B for 12.5 to 25 min.

RT-PCR. RNA was prepared from freshly isolated hepatocytes (∼6 × 10⁶ cells) using TRIzol according to the manufacturer's instructions and was treated with RNase-free DNase I (200 U) for 30 min at 37°C, and then re-extracted with TRIzol. Isolated hepatocytes were also cultured (3 × 10⁶ cells/9-cm dish) as described above for 24 h and then for a further 24 h in the absence of FCS, and RNA was extracted as described above. First-strand cDNA was prepared from 2.5 μg of RNA using random hexamers and Superscript II reverse transcriptase in a 20-μl reaction volume, according to the manufacturer's instructions. PCR reactions for the rat P2Y₁, P2Y₂, P2Y₄, and P2Y₆ receptors used the primers described previously (Dixon et al., 2000). Reactions for the P2Y₁₂ and P2Y₁₃ receptors included the following primer pairs designed to amplify partial cDNAs from each sequence (P2Y₁₂: 5′-ATCTGGGCCTTCATGTTCCTGCTGTC-3′, 5′-CTTTCTTCTTATTTGCCCCGGATGTTGAG-3′, 532 bp; P2Y₁₃: 5′-TACCATACACCCACAGTCAAACCACC-3′, 5′-GGGCATAATCTTTTCCGTACACGAG-3′, 405 bp). The rat P2Y₁₂ primers were designed from the published sequence (Simon et al., 2002), and the rat P2Y₁₃ primers were designed from a rat expressed sequence tag (GenBank accession number BF390809) that when translated shared 80% amino acid identity with the corresponding region of the human protein. Amplifications of the housekeeping gene, glyceraldehyde-3-phosphate dehydrogenase (GAPDH) were also performed using the following primers (5′-ACCACAGTCCATGCCATCAC-3′, 5′-TCCACCACCCTGTTGCTGTA-3′, 415 bp).

PCRs were performed using 1.25% (v/v) of each first-strand cDNA reaction and 7.5 pmol of each subtype-specific forward and reverse primer, with 2.5 units of Biotaq in a 25-μl amplification volume containing 1.5 mM MgCl₂, according to the manufacturer's instructions. Amplification conditions were 30 s at 94°C, 30 s at 65°C, and 45 s at 72°C for 30 cycles. Reverse transcriptions were also performed in the absence of the enzyme as a control for contaminating DNA, and control amplifications were performed in the absence of added template. Amplicons (5 μl for GAPDH, P2Y₁, and P2Y₂; and 10 μl for P2Y₄, P2Y₆, P2Y₁₂, and P2Y₁₃) were subjected to gel electrophoresis on a 2% (w/v) agarose gel.

Statistical Analysis. Glycogen phosphorylase and cyclic AMP data presented in the figures are representative of three experiments from different hepatocyte preparations and are expressed as milli-units of enzyme activity or picomoles of cyclic AMP well^-1. All data presented in the text were pooled across three experiments, each in triplicate, and cyclic AMP data were normalized to the glucagon response after subtraction of basal values. Statistical analysis was by one-way ANOVA with Bonferroni's post-test, or Student's t test where indicated.

Materials. Cell culture medium was obtained from Invitrogen (Paisley, UK). Cyclic [³H]AMP and [¹⁴C]glucose-1-phosphate were purchased from Amersham Biosciences UK, Ltd. (Little Chalfont, Buckinghamshire, UK). Collagenase was obtained from Boehringer (Lewes, UK). 2-MeSADP, A3P5P, MRS 2179 (N⁶-methyl 2′-deoxyadenosine 3′,5′-bisphosphate), UTP, pertussis toxin, and all other chemicals were from Sigma Chemical (Poole, Dorset, UK). AR-C67085 (2-propylthio-β,γ-dichloromethylene-d-ATP) was a kind gift from AstraZeneca R and D Charnwood (Loughborough, UK). Biotaq was purchased from Bioline (Edgeware, UK), and all other molecular biology reagents were obtained from Invitrogen.

Results

Activation of Glycogen Phosphorylase by 2-MeSADP in Freshly Isolated Rat Hepatocytes.Figure 1 shows the concentration-response curve for activation of glycogen phosphorylase by 2-MeSADP in the presence and absence of the highly selective P2Y₁ antagonist, MRS 2179 (Boyer et al., 1998). 2-MeSADP potently stimulates glycogen phosphorylase with responses starting at 30 nM. In all experiments the curve was biphasic; an initial plateau (∼80% of maximal response) formed at 10 to 30 μM 2-MeSADP, followed by a small further rise at higher concentrations. The mean EC₅₀ for the major, high potency component of the response was 121 nM (pEC₅₀ = -6.92 ± 0.02: data pooled across three separate experiments, each in triplicate). In the presence of 100 μM MRS 2179 there was a substantial shift of the concentration-response curve to the right, with an EC₅₀ of 81 μM (pEC₅₀ = -4.17 ± 0.21 μM: data pooled across three experiments). The apparent pA₂ for MRS 2179, calculated only from the initial major phase of the concentration-response curve for 2-MeSADP (i.e., up to 100 μM 2-MeSADP), was 6.82 (data pooled from three experiments).

Inhibition of Glucagon-Stimulated Cyclic AMP Accumulation by 2-MeSADP and UTP in Freshly Isolated Rat Hepatocytes. As expected, glucagon (10 nM) in the presence of the phosphodiesterase inhibitor IBMX (300 μM), stimulated cyclic AMP accumulation from basal values of 7.8 ± 0.6 to 40.5 ± 0.8 pmol of cyclic AMP per well (n = three separate experiments). Stimulation of cells with 2-MeSADP led to a decrease in cyclic AMP levels stimulated by glucagon (Fig. 2a). The threshold concentration of 2-MeSADP for this inhibitory response was greater than 30 μM; application of 10 or 30 μM 2-MeSADP had no significant effect. Application of 100 μM 2-MeSADP reduced the glucagon-stimulated cyclic AMP level to 27.2 ± 1.1 pmol per well. Increasing the 2-MeSADP concentration to 300 μM led to a further decrease in cyclic AMP concentrations to 15.4 ± 0.4 pmol per well (Fig. 2a). This level of inhibition was found to be maximal; application of 1 mM 2-MeSADP had no additional effect (data not shown).

UTP was found to inhibit glucagon-stimulated cyclic AMP levels in freshly isolated hepatocytes to a similar extent (Fig. 2a). Thus, after subtraction of basal values, a maximal concentration of UTP (300 μM) reduced the glucagon-stimulated cyclic AMP level by 61.2 ± 1.6% compared with 61.9 ± 0.3% in response to 300 μM 2-MeSADP (n = three separate experiments). When applied alone to cells neither 2-MeSADP nor UTP (300 μM) had any significant effect on the basal level of cyclic AMP (data not shown).

Breakdown of UTP and 2-MeSADP. When cells were incubated with 300 μM UTP for 3 min and the medium analyzed by HPLC, only 95.0 ± 8.7 μM UTP was recovered. The majority of nucleotide (219.0 ± 6.0 μM) was recovered as UDP. Similarly, following incubation of cells for 3 min with 300 μM 2-MeSADP, 132.9 ± 9.5 μM was recovered. Two additional peaks were identified with retention times of 8.7 and 20.7 min compared with 27.9 min for 2-MeSADP, consistent with the 2-methylthio-substituted derivatives of adenosine and AMP, respectively. These two peaks accounted for 29.4 ± 2.4% and 26.7 ± 0.6% of the added nucleotide concentration (data not shown).

The products of 2-MeSADP degradation could conceivably act on P1 receptors present on rat hepatocytes and therefore influence the observed response. The P1 receptor present on these cells has been characterized as an A_2B receptor coupled to increases in cyclic AMP (Yasuda et al., 2003). Stimulation of this receptor would therefore counteract the inhibitory effect of 2-MeSADP on cyclic AMP generation. The inclusion of 300 μM IBMX should block P1 receptor activation as this compound is a general low potency P1 antagonist (Fredholm et al., 1994). However, to eliminate this possibility we stimulated cells with 2-MeSADP in the additional presence of the nonselective P1 antagonist CGS 15943 (Alexander et al., 1996). This treatment had no significant effect on the ability of 2-MeSADP to inhibit glucagon-stimulated cyclic AMP levels. Thus in response to 100 and 300 μM 2-MeSADP, glucagon-stimulated cyclic AMP levels were reduced by 28.0 ± 7.9% and 71.8 ± 1.7%, compared with 24.2 ± 9.9% and 70.8 ± 3.9% in the presence of 5 μM CGS 15943 (n = 3; data not shown). In addition, inclusion of CGS 15943 did not increase the potency of 2-MeSADP; 30 μM 2-MeSADP had no significant effect in the presence or absence of CGS 15943 (data not shown).

Effect of Culturing Rat Hepatocytes on the Response to 2-MeSADP and UTP. Hepatocytes were cultured for 48 h and the effects of 2-MeSADP and UTP examined and compared with the effects in freshly isolated cells from the same preparation. As illustrated in Fig. 2b, the inhibitory effect of 300 μM 2-MeSADP on glucagon-stimulated cyclic AMP levels was reduced significantly from 61.9 ± 0.3% in freshly isolated cells to 22.9 ± 6.2% following culture (P < 0.01, Student's t test). In contrast, the inhibitory effect of 300 μM UTP was significantly increased from 61.2 ± 1.6% to 88.7 ± 4.5% after 48 h in culture (P < 0.01, Student's t test).

To investigate the effects of pertussis toxin, the culture period was reduced to 24 h, at which time an inhibitory effect of 2-MeSADP was still apparent. Expressed as picomoles of cyclic AMP per well (mean ± S.E.), glucagon (10 nM) stimulation (control-deducted) with no pertussis toxin treatment was 9.7 ± 0.5, which was reduced by 2-MeSADP (300 μM) to 7.0 ± 1.2 and by UTP (300 μM) to 1.1 ± 0.1. After pertussis toxin, the equivalent figures were: glucagon, 11.3 ± 0.5; glucagon plus 2-MeSADP, 10.44 ± 0.31; glucagon plus UTP, 9.82 ± 1.34 pmol of cyclic AMP per well. These results show that pertussis toxin pretreatment removes the majority of the inhibitory cyclic AMP response to 2-MeSADP and to UTP. HPLC analysis revealed that during the 3 min incubation with cultured cells, 2-MeSADP and UTP were not degraded (data not shown).

Expression of mRNA for P2Y Receptor Subtypes. RT-PCR studies were undertaken to compare the expression of the P2Y₁, P2Y₂, P2Y₄, P2Y₆, P2Y₁₂, and P2Y₁₃ receptor transcripts in RNA isolated from freshly isolated hepatocytes and cells from the same preparations cultured for 48 h. Products for rat P2Y₁, P2Y₂, P2Y₄, P2Y₆, P2Y₁₂, and P2Y₁₃ receptor mRNAs were detected when cDNA derived from freshly isolated hepatocytes was subjected to 30 PCR cycles in reactions that included reverse transcriptase, but not when it was omitted (Fig. 3).

Amplifications using RNA isolated from cultured hepatocytes revealed an altered expression pattern for P2Y receptor transcripts, with a stronger signal from P2Y₄ and P2Y₆ amplicons when compared with freshly isolated samples. Conversely, transcripts for the P2Y₁₂ receptor were barely detectable, and P2Y₁₃ receptor transcripts were absent after 30 cycles of PCR (Fig. 3). However, the intensities of bands for P2Y₁, P2Y₂, and the housekeeping gene GAPDH were indistinguishable between freshly isolated and cultured preparations, indicating that the altered expression pattern is likely to be a reflection of true differences in P2Y receptor transcript levels rather than differences resulting from sample preparation. No PCR products were detected when reactions were performed in the absence of an added template (data not shown).

Effects of P2Y₁ Antagonists on 2-MeSADP-Stimulated Cyclic AMP Responses. In a series of experiments, we investigated the effect of two P2Y₁ antagonists, A3P5P (Boyer et al., 1996) and MRS 2179 (Boyer et al., 1998). Figure 4, a and b, show that MRS 2179, in the concentration range 1 to 300 μM, did not reverse the inhibition of glucagon-stimulated cyclic AMP levels induced by either 100 or 300 μM 2-MeSADP. Data pooled across three experiments with 300 μM 2-MeSADP showed no significant effect of 1 to 300 μM MRS 2179 (P > 0.05, one-way ANOVA). However, at the submaximal concentration of 2-MeSADP (100 μM) there was a tendency for MRS 2179 to further reduce cyclic AMP levels (Fig. 4a), which was significant when data were pooled from three experiments (P < 0.001, one-way ANOVA). This effect is seen in the absence of stimulation with nucleotide agonists. The application of 300 μM MRS 2179 to rat hepatocytes in the absence of 2-MeSADP reduced glucagon-stimulated cyclic AMP levels by 31.2 ± 11.4% (n = three separate experiments).

The less selective P2Y₁ antagonist A3P5P gave essentially the same results as MRS 2179. Data pooled from three experiments revealed that A3P5P (1–300 μM) had no significant effect on the inhibition of glucagon-stimulated cyclic AMP levels induced by 300 μM 2-MeSADP (P > 0.05, one-way ANOVA; data not shown). Again at the submaximal concentration of 2-MeSADP (100 μM), there was a significant effect of A3P5P to further reduce cyclic AMP levels (P > 0.05; one-way ANOVA). Application of 300 μM A3P5P alone led to a reduction of glucagon-stimulated cyclic AMP levels by 34.8 ± 3.0% (n = three separate experiments).

Effects of AR-C67085 on 2-MeSADP-Stimulated Cyclic AMP Responses. As 2-MeSADP is also a highly potent agonist at the rat P2Y₁₂ receptor, we investigated the effect of the P2Y₁₂ receptor antagonist, AR-C67085 (Humphries et al., 1995a,b). Figure 5, a and b, shows the results from a typical experiment in which increasing concentrations of AR-C67085 (10 nM–1 μM) did not affect the inhibition of glucagon-stimulated cyclic AMP levels by a submaximal (100 μM) or maximal (300 μM) dose of 2-MeSADP. Data pooled from three experiments revealed that AR-C67085 had no significant effect at either concentration of 2-MeSADP (P > 0.05, one-way ANOVA). Unlike MRS 2179 and A3P5P, application of AR-C67085 alone had no significant effect on glucagon-stimulated cyclic AMP levels; response to 10 nM glucagon in the presence of 1 μM AR-C67085 was 100.3 ± 13.3% of the response to glucagon alone.

AR-C67085 has recently been described as a weaker potency antagonist at the P2Y₁₃ receptor (Marteau et al., 2003). Increasing the AR-C67085 concentration to 300 μM did not lead to inhibition of the 2-MeSADP-mediated attenuation of cyclic AMP levels (data not shown).

AR-C67085 is also a potent agonist at the human P2Y₁₁ receptor, with an EC₅₀ for cyclic AMP accumulation in transfected cells of 1.5 μM (Communi et al., 1999). When applied to cells at concentrations up to 300 μM, AR-C67085 had no effect on cyclic AMP levels. Thus, in the presence of 300 μM AR-C67085, the cyclic AMP concentration was 15.0 ± 2.3 compared with basal values of 13.3 ± 2.3 pmol of cyclic AMP per 10⁶ cells (n = three separate experiments; P > 0.05, Student's t test). In a parallel series of experiments, AR-C67085 increased cyclic AMP levels in 1321N1 cells heterologously expressing P2Y₁₁ receptors (unpublished data).

HPLC analysis confirmed that there was no significant breakdown of any of the antagonists used during incubation (data not shown).

Discussion

Understanding the regulation of hepatocyte function by extracellular nucleotides is complicated by the presence of multiple P2Y receptor subtypes that are activated by a range of native agonists. Here we have examined regulation of rat hepatocytes by the partially selective synthetic ADP derivative 2-MeSADP, which acts at P2Y₁, P2Y₁₂, and P2Y₁₃ receptors. In a previous report, this nucleotide was shown to increase [Ca²⁺]_c (Dixon, 2000), and here we demonstrate that at similar concentrations it activates glycogen phosphorylase. However, we also report an apparently paradoxical attenuation of glucagon-stimulated cyclic AMP levels by 2-MeSADP, and have investigated this inhibitory effect to determine whether the same receptor mediates the conflicting influences on glycogen phosphorylase activity.

2-MeSADP is an agonist at the rat P2Y₁ receptor (Hechler et al., 1998) that is expressed by hepatocytes (Dixon et al., 2000). The potency of 2-MeSADP to stimulate glycogen phosphorylase (EC₅₀ 121 nM) is consistent with this response being mediated by P2Y₁ receptors, and this is confirmed by inhibition by the competitive and selective P2Y₁ antagonist, MRS 2179. We report a parallel shift to the right of the concentration-response curve for 2-MeSADP in the presence of this drug. The apparent pA₂ value of 6.82 compares well with the reported pA₂ of 6.99 for MRS 2179 at the turkey erythrocyte P2Y₁ receptor (Boyer et al., 1998).

Characteristically there was a small secondary rise in glycogen phosphorylase activity at the highest concentrations of 2-MeSADP tested. This may reflect the accumulation of the products of 2-MeSADP degradation. As described under Results, incubation of freshly isolated cells with 2-MeSADP led to accumulation of 2-methylthio-substituted derivatives of adenosine and AMP. These compounds could activate adenosine A_2B receptors present on rat hepatocytes that are coupled to increased cyclic AMP levels (Yasuda et al., 2003). This could conceivably stimulate glycogen phosphorylase further through the classical pathway via protein kinase A activation.

In single rat hepatocytes, 2-MeSADP evokes oscillations in [Ca²⁺]_c at concentrations as low as 100 nM, consistent with a role in the P2Y₁-mediated activation of glycogen phosphorylase as described above. Furthermore, these oscillations were blocked by the P2Y₁ receptor antagonist A3P5P (Dixon, 2000). The P2Y₁ receptor, however, does not appear to mediate the inhibition of glucagon-stimulated cyclic AMP accumulation by 2-MeSADP observed here, as this effect was not reversed by P2Y₁ receptor antagonists. Interestingly, both A3P5P and MRS 2179 had significant inhibitory effects on adenylyl cyclase in hepatocytes, which could potentially complicate the interpretation of experiments using these compounds. However, the maximal ability of these two drugs to directly attenuate cyclic AMP levels was substantially smaller than the inhibitory response to 2-MeSADP. It would therefore not completely mask an antagonist effect if 2-MeSADP was acting at P2Y₁ receptors. We conclude therefore that inhibition of adenylyl cyclase by 2-MeSADP is mediated by a receptor distinct from the P2Y₁ receptor, which is responsible for increased [Ca²⁺]_c and activation of glycogen phosphorylase.

To determine which receptors may be expressed in rat hepatocytes, we have extended our previously reported RT-PCR analysis (Dixon et al., 2000) to include recently identified members of the P2Y family. Consistent with the previous study (Dixon et al., 2000), we have shown that transcripts for P2Y₁, P2Y₂, P2Y₄ and P2Y₆ are expressed in freshly isolated hepatocytes, but in addition we report the expression of transcripts for P2Y₁₂ and P2Y₁₃ receptors. This is particularly significant as these receptors are negatively coupled to adenylyl cyclase through G_i (Communi et al., 2001; Zhang et al., 2002), and 2-MeSADP has been identified as an agonist at the rat P2Y₁₂ receptor (Hollopeter et al., 2001) and the mouse and human P2Y₁₃ receptors (Communi et al., 2001; Zhang et al., 2002). (The rat P2Y₁₃ receptor has yet to be cloned and its agonist profile is therefore unknown.)

RT-PCR results demonstrated that following culture for 48 h, transcripts for the G_i-coupled P2Y receptors were either barely detectable (P2Y₁₂) or absent (P2Y₁₃). Experiments conducted in parallel with RT-PCR revealed a concomitant large reduction in the ability of 2-MeSADP to inhibit glucagon-stimulated cyclic AMP levels. Culturing cells also led to an apparent increase in the effect of UTP, although this is likely to reflect the difference in stability of UTP under the two conditions. Significantly though the observation that UTP remains effective at inhibiting the cyclic AMP levels is consistent with the view that UTP and 2-MeSADP act at separate P2Y receptor subtypes and that the reduced response to 2-MeSADP reflects changes at the level of the receptor rather than an altered transduction mechanism.

These results suggested that the inhibitory response to 2-MeSADP may be mediated by P2Y₁₂ or P2Y₁₃ receptors. We therefore investigated the effect of AR-C67085, an antagonist with high potency (IC₅₀ = 0.6 nM; Humphries et al., 1995b) at the rat P2Y₁₂ receptor and low potency at the human P2Y₁₃ receptor (IC₅₀ = 0.63 μM; Marteau et al., 2003). AR-C67085, at concentrations up to 300 μM, did not antagonize the 2-MeSADP-evoked inhibition of glucagon-stimulated cyclic AMP levels in rat hepatocytes, arguing against the involvement of P2Y₁₂ receptors. Assuming the pharmacology of the rat P2Y₁₃ receptor reflects that of the human homolog, the lack of antagonism by AR-C67085 also argues against P2Y₁₃ receptor involvement. The accumulation during incubations of 2-MeSAMP, a known P2Y₁₂ and P2Y₁₃ antagonist (Hollopeter et al., 2001; Marteau et al., 2003), is unlikely to affect these conclusions since any antagonism would not be effective at the onset of the incubation. Furthermore, studies with adenosine 5′-O-(2-thiodiphosphate), a stable but low potency agonist of both P2Y₁₂ and P2Y₁₃ receptors (Zhang et al., 2001; Marteau et al., 2003), also showed no evidence of either of these receptors (unpublished data).

In comparison to the high potency of 2-MeSADP to evoke [Ca²⁺]_c oscillations in single rat hepatocytes (Dixon, 2000) or to activate glycogen phosphorylase as demonstrated here, the threshold concentration of 2-MeSADP required to elicit an inhibitory effect on cyclic AMP levels was high. It is, however, consistent with a previous report by Edgecombe et al. (1999) in which the inhibitory effects of ATP and Ap₄A in freshly prepared rat hepatocytes were apparent only at concentrations greater than 100 μM. We have demonstrated by HPLC that following a 3-min incubation of hepatocytes with 300 μM 2-MeSADP only 56% was recovered. It may therefore be assumed that in the absence of any degradation the threshold concentration of 2-MeSADP for inhibition of glucagon-stimulated cyclic AMP levels will be lower than that reported here. However, it remains apparent that the inhibition of cyclic AMP accumulation in response to P2Y receptor agonists will only be effective at concentrations that are high compared with those required to stimulate glycogen phosphorylase. The functional consequences may be to act to restrain the effects of the highest concentrations of extracellular nucleotides.

In conclusion, we have demonstrated that in freshly isolated rat hepatocytes, 2-MeSADP both stimulates glycogen phosphorylase and inhibits adenylyl cyclase and that these effects are the result of stimulation of distinct subtypes of P2Y receptor. The results of antagonist studies show that the activation of glycogen phosphorylase is mediated by the P2Y₁ receptor responsible for the previously reported elevation of [Ca²⁺]_c (Dixon, 2000) but that the inhibition of adenylyl cyclase is not mediated by P2Y₁ receptors or any other of the characterized receptors for which 2-MeSADP is a known agonist. It is therefore possible that 2-MeSADP acts at a hitherto uncloned receptor. It has recently been recognized that the P2Y receptor family is structurally diverse, the most recently identified receptor, P2Y₁₄, being a receptor for UDP-glucose (Abbracchio et al., 2003). Through phylogenetic analysis a number of orphan receptors have been identified that closely correlate with the P2Y family of G-protein-coupled receptors (Fredriksson et al., 2003), one of which may correspond to that on rat hepatocytes mediating the effect of 2-MeSADP. However, it remains a possibility that the rat P2Y₁₃ receptor displays an altered pharmacology with respect to antagonism by AR-C67085 compared with the human P2Y₁₃ receptor and that this receptor subtype mediates the inhibition of glucagon-stimulated cyclic AMP levels in rat hepatocytes. A further possibility is that receptor dimerization of known P2Y receptors results in a previously uncharacterized physiological response.