Introduction

Top Abstract Introduction Experimental Procedures Results Discussion References

Acute activation of G_i/o-coupled receptors inhibits drug-stimulated cyclic AMP accumulation. In contrast, prolonged activation of G_i/o-coupled receptors results in an enhanced responsiveness of forskolin- and G_s-stimulated adenylate cyclase [E.C. 4.6.1.1] activity through an unknown mechanism (Thomas and Hoffman, 1987; Ammer and Schulz, 1996; Watts and Neve, 1996; Rhee et al., 2000). This agonist-induced heterologous sensitization of drug-stimulated cyclic AMP accumulation has been proposed to be a mechanism by which cells adapt to prolonged inhibition of cyclic AMP synthesis and may be a cellular model of drug tolerance and dependence (Sharma et al., 1975; Ammer and Schulz, 1996; Watts and Neve, 1996; Nestler and Aghajanian, 1997). Heterologous sensitization of adenylate cyclase is caused by the persistent activation of a number of G_i/o-coupled receptors, including µ-opioid, D₂ and D₄ dopamine, A₃ adenosine, ₂-adrenergic, M₄ muscarinic, and CB₁ cannabinoid receptors (Sharma et al., 1975; Jones and Bylund, 1988; Ammer and Schulz, 1996; Thomas and Hoffman, 1996; Watts and Neve, 1996; Avidor-Reiss et al., 1997; Palmer et al., 1997; Schoffelmeer et al., 1997; Nevo et al., 1998; Watts et al., 1999; Rhee et al., 2000). In addition, studies have revealed that heterologous sensitization occurs in a number of cellular systems, including transient expression systems, stably transfected cultured cells, and primary cultured neurons (Watts and Neve, 1996; Schoffelmeer et al., 1997; Watts et al., 1998; Rhee et al., 2000; V. J. Watts, J. G. Lisinicchia, and M. A. Beazely, unpublished observations). Together these observations suggest that heterologous sensitization is a common phenomenon, and that understanding the mechanisms may reveal information about G protein-coupled receptor signaling in a variety of cell types.

Although the precise molecular mechanisms for heterologous sensitization remain unknown, recent studies have provided evidence that sensitization is influenced by the signal transduction machinery within the cellular environment (Thomas and Hoffman, 1996; Watts and Neve, 1996; Watts et al., 1998; Rhee et al., 2000). Specifically, evidence suggests that agonist-induced sensitization may be influenced by the complement of adenylate cyclase isoforms present within the cell (McDermott and Sharp, 1995; Watts and Neve, 1996; Avidor-Reiss et al., 1997; Varga et al., 1998). There are at least nine distinct molecular forms of adenylate cyclase (ACI-ACIX) that have been partially characterized and each isoform appears to have distinct regulatory requirements with the exception that each isoform is stimulated by G_s, (for review, see Taussig and Zimmerman, 1998). Types I, III, and VIII are stimulated by calcium/calmodulin and have differential sensitivity to modulation by _i/o (Nevo et al., 1998, V. J. Watts and M. G. Cumbay, unpublished observations). Additionally, ACI shows synergistic activation by G_s and Ca²⁺, whereas ACVIII does not (Cali et al., 1994; Wayman et al., 1994; Nielsen et al., 1996). Types V and VI are robustly stimulated by G_s and forskolin, whereas their activity is inhibited by calcium and G_i (Taussig et al., 1994; Chabardes et al., 1999). The family of adenylate cyclases with the most members is represented by ACII, ACIV, and ACVII, which are characterized by their ability to be conditionally stimulated by G protein subunits as well as their apparent lack of direct modulation by G_i/o (Taussig and Zimmerman, 1998). In addition, ACII and ACVII are stimulated by activators of protein kinase C such as PMA (Watts and Neve, 1997; Taussig and Zimmerman, 1998). Finally, ACIX is stimulated by G_s but is not potently stimulated by forskolin (Antoni et al., 1998; Hacker et al., 1998).

The regulatory properties described above and recent observations that the recombinant isoforms of adenylate cyclase show differential sensitization (Thomas and Hoffman, 1996; Watts and Neve, 1996; Avidor-Reiss et al., 1997; Nevo et al., 1998; Rhee et al., 2000) provide the basis for present studies. We have constructed HEK293 cells stably expressing D_2L dopamine receptors in combination with ACI, ACII, ACV, ACVIII, or ACIX in an effort to identify potential mechanisms for heterologous sensitization of adenylate cyclase. In an effort to control for sensitization of endogenous adenylate cyclase activity in HEK293 cells, we have characterized and used conditions that show selective activation of recombinant adenylate cyclase isoforms. We now report that both short- and long-term activation of D_2L dopamine receptors resulted in a marked degree of sensitization of ACI, ACII, ACV, and ACIX, but not ACVIII. This time-dependent heterologous sensitization appeared to be influenced by the regulatory properties of each adenylate cyclase isoform. In particular, sensitization appeared to be related to the ability of adenylate cyclase isoforms to be activated by G_s, supporting a hypothesis that heterologous sensitization involves G_s.

	Experimental Procedures

Top Abstract Introduction Experimental Procedures Results Discussion References

Materials. [³H]Cyclic AMP was purchased from PerkinElmer Life Science Products (Boston, MA). Forskolin, ()-quinpirole, and spiperone were purchased from RBI/Sigma (Natick, MA). PMA and A23187 were purchased from Calbiochem (La Jolla, CA). Fetal clone serum and bovine calf serum were purchased from Hyclone (Logan, UT). Most other reagents were purchased from Sigma (St. Louis, MO).

Cell Culture. HEK293 cells expressing ACI, ACII, ACV, and ACVIII were obtained from Dr. Daniel Storm (University of Washington, Seattle, WA). Each cell line was transfected with pcDNA1-D_2L as described previously (Watts and Neve, 1996), creating the cell lines ACI/D_2L, ACII/D_2L, ACV/D_2L, and ACVIII/D_2L. HEK293 cells were transfected with pcDNA3-ACIX using lipofectAMINE (Life Technologies, Grand Island, NY) according to the manufacture's instructions. Clones were isolated by selection with G418 (600 µg/ml) and were screened for isoproterenol-stimulated cyclic AMP accumulation as described below. The ACIX cell line was then transfected with pcDNA1-D_2L (10 µg) and pBabe Puro (1 µg) using lipofectAMINE and clones were isolated following selection with puromycin (2 µg/ml) and G418 (600 µg/ml). All cells were maintained in culture medium consisting of Dulbecco's modified Eagle's medium supplemented with 5% fetal clone serum, 5% bovine calf serum, 0.05 U/ml penicillin, and 50 µg/ml streptomycin. The medium used to maintain HEK-ACI/D_2L, ACII/D_2L, ACV/D_2L, and ACVIII/D_2L cells was supplemented with hygromycin (460 U/ml) and puromycin (2 µg/ml). The medium used for the ACIX/D_2L cell line was supplemented with G418 (300 µg/ml) and puromycin (2 µg/ml). Cells were propagated in a humidified incubator in the presence of 10% CO₂.

Cyclic AMP Accumulation Assay. Cells were seeded at densities between 100,000 and 150,000 cells/well in 48-well cluster plates and grown to confluence. Cells were pretreated for 2 or 18 h in culture medium, in the presence or absence of the D₂ agonist quinpirole (1 µM). Where indicated, some pretreatments were performed in the presence of the D₂ antagonist spiperone (1 µM). Following pretreatment, the cells were washed three times for 3 to 4 min with 200 µl/well of assay buffer (Earle's balanced salt solution containing 0.02% ascorbic acid and 2% bovine calf serum). The cells were then placed on ice and the indicated drugs were added in the presence of spiperone (1 µM) to prevent activation of D_2L receptors by residual quinpirole (Watts and Neve, 1996). The cells were then incubated in a 37°C water bath for 15 min. Following the incubation, the stimulation media was decanted and the reaction was terminated with 100 µl/well of ice-cold 3% trichloroacetic acid. The 48-well cluster plates were stored at 4°C for up to 1 week prior to analysis.

Quantification of Cyclic AMP. Cyclic AMP accumulation was quantified using a competitive binding assay adapted from Nordstedt and Fredholm (1990) with minor modifications (Watts and Neve, 1996). Duplicate samples of the cell lysate (15 µl) were added to reaction tubes. [³H]Cyclic AMP (~1 nM final concentration) and cyclic AMP binding protein (ca. 150 µg) were diluted in cyclic AMP assay buffer (100 mM Tris/HCl, pH, 7.4, 100 mM NaCl, 5 mM EDTA) and then added to each well for a total volume of 550 µl. The tubes were incubated on ice for 2 h and were harvested by filtration (Packard Unifilter GF/C) using a 96-well Packard Filtermate Cell harvester (Meriden, CT). The filters were allowed to dry and Microscint O scintillation fluid was added. Radioactivity on the filters was determined using a Packard TopCount scintillation/luminescence detector. Cyclic AMP concentrations in each sample were estimated in duplicate from a standard curve ranging from 0.1 to 300 pmol of cyclic AMP per assay.

	Results

Top Abstract Introduction Experimental Procedures Results Discussion References

Regulation of Endogenous Adenylate Cyclase in HEK-D_2L Cells. Acutely, D₂ agonists inhibit cyclic AMP accumulation in HEK293 cells expressing D_2L receptors, whereas short-term (2 h) pretreatment with the same drugs results in a heterologous sensitization of forskolin (10 µM)-stimulated cyclic AMP accumulation (Fig. 1; Watts and Neve, 1996). Long-term (18 h) activation of D₂ dopamine receptors also sensitized forskolin (10 µM)-stimulated cyclic AMP accumulation (Fig. 1). The apparent magnitude of agonist-induced heterologous sensitization in HEK-D_2L cells was dependent upon the concentration of forskolin that was used to subsequently stimulate cyclic AMP accumulation (Table 1; Watts and Neve, 1996). For example, agonist treatment of HEK-D_2L cells followed by stimulation with 10 µM forskolin caused a marked increase in cyclic AMP accumulation that was 7-fold greater than that observed in vehicle-treated cells (Table 1; Watts and Neve, 1996). On the other hand, sensitization studies using forskolin at concentrations of 100 nM revealed no increase in cyclic AMP accumulation compared with basal levels and no enhanced response in agonist-treated cells (Table 1). Agonist treatment also failed to induce significant sensitization of adenylate cyclase activity in HEK-D_2L cells following activation of the endogenous -adrenergic receptor by isoproterenol (Table 1). Additionally, agonist treatment did not alter basal levels of cyclic AMP accumulation in HEK-D_2L cells (Fig. 1).

View larger version (12K): [in this window] [in a new window]   Fig. 1.   Effect of agonist pretreatment on cyclic AMP accumulation in HEK-D2L cells. Cells were pretreated for 2 h (A) or 18 h (B) with vehicle or quinpirole (1 µM). Following agonist pretreatment, cells were washed extensively and cyclic AMP accumulation was determined in the absence (basal) or presence of 10 µM forskolin (FSK) for 15 min. Data shown are mean ± S.E. of seven to nine independent experiments. *p < 0.05 compared with vehicle-treated cells (Student's paired t test).

Effects of Short- and Long-Term D₂ Receptor Activation on ACI and ACVIII. Type I or VIII adenylate cyclase was stably transfected in combination with the dopamine D_2L receptor in HEK cells. Stimulation of either ACI/D_2L cells and ACVIII/D_2L cells with the calcium ionophore A23187 markedly increased cyclic AMP accumulation (Figs. 2 and 3). Exposure of ACI/D_2L cells to quinpirole enhanced A23187-stimulated cyclic AMP accumulation compared with vehicle-treated cells by 3- and 5-fold following short- and long-term agonist exposure, respectively (Fig. 2). Heterologous sensitization of ACI was prevented when quinpirole treatments were carried out in the presence of the D₂ antagonist spiperone (Fig. 2). In contrast to the robust effects of agonist treatment on ACI activity, short-term agonist exposure on ACVIII/D_2L cells did not result in significant heterologous sensitization of A23187-stimulated cyclic AMP accumulation (Fig. 3A). Furthermore, 18 h of agonist exposure produced only a modest degree of heterologous sensitization of ACVIII (1.3-fold) compared with vehicle-treated cells (Fig. 3B).

View larger version (13K): [in this window] [in a new window]   Fig. 2.   Agonist-induced heterologous sensitization of A23187-stimulated cyclic AMP accumulation in ACI/D2L cells. Cells were pretreated for 2 h (A) or 18 h (B) with vehicle or quinpirole (1 µM). Following agonist pretreatment, cells were washed extensively and cyclic AMP accumulation was determined in the absence (basal) or presence of 10 µM A23187 for 15 min. Where indicated some pretreatments were completed in the presence of 1 µM spiperone (+Spip). Data shown are mean ± S.E. of seven to eight independent experiments. *p < 0.05 compared with vehicle-treated cells (Student's paired t test).

View larger version (14K): [in this window] [in a new window]   Fig. 3.   Agonist-induced heterologous sensitization of A23187-stimulated cyclic AMP accumulation in ACVIII/D2L cells. Cells were pretreated for 2 h (A) or 18 h (B) with vehicle or quinpirole (1 µM). Following agonist pretreatment, cells were washed extensively and cyclic AMP accumulation was determined in the absence (basal) or presence of 10 µM A23187 for 15 min. Where indicated some pretreatments were completed in the presence of 1 µM spiperone (+Spip). Data shown are mean ± S.E. of seven to eight independent experiments. *p < 0.05 compared with vehicle-treated cells (Student's paired t test).

View larger version (11K): [in this window] [in a new window]   Fig. 4.   Regulation of cyclic AMP accumulation in ACI/D2L and ACVIII/D2L cells in the presence of 500 µM IBMX. ACI/D2L, and ACVIII/D2L cells were pretreated for 2 h with vehicle or quinpirole (1 µM) (A). Following agonist pretreatment, cells were washed extensively and cyclic AMP accumulation was determined in the presence of 10 µM A23187 for 15 min. B and C, basal and A23187-stimulated cyclic AMP accumulation in cells following 2 h of pretreatment in culture medium in the absence (control) or presence of 200 ng/ml cholera toxin (+CTX). Data shown are mean ± S.E. of three to seven independent experiments. *p < 0.05 compared with vehicle-treated cells (Student's t test).

Effects of Short- and Long-Term D₂ Receptor Activation on ACII. HEK ACII/D_2L cells were used to examine drug-stimulated cyclic AMP accumulation following pretreatment with quinpirole for 2 or 18 h. Quinpirole pretreatment resulted in marked sensitization of PMA-stimulated cyclic AMP accumulation compared with cells that were vehicle-treated (Fig. 5). The magnitude of sensitization was time-dependent and was greater following long-term agonist exposure. Both short- and long-term quinpirole-induced sensitization were blocked when pretreatments were carried out in the presence of spiperone (Fig. 5). In contrast to effects on PMA-stimulated cyclic AMP accumulation, agonist treatment of ACII/D_2L cells did not alter cyclic AMP accumulation under basal, forskolin-stimulated, isoproterenol-stimulated, or A23187-stimulated conditions (Table 2).

View larger version (14K): [in this window] [in a new window]   Fig. 5.   Agonist-induced heterologous sensitization of PMA-stimulated cyclic AMP accumulation in ACII/D2L cells. Cells were pretreated for 2 h (A) or 18 h (B) with vehicle or quinpirole (1 µM). Following agonist pretreatment, cells were washed extensively and cyclic AMP accumulation was determined in the absence (basal) or presence of 100 nM PMA for 15 min. Where indicated some pretreatments were completed in the presence of 1 µM spiperone (+Spip). Data shown are mean ± S.E. of seven to eight independent experiments. *p < 0.05 compared with vehicle-treated cells (Student's paired t test).

Effects of Short- and Long-Term D₂ Receptor Activation on ACV. ACV/D_2L cells were exposed to quinpirole for 2 or 18 h and basal and drug-stimulated cyclic AMP levels were examined. The basal activity of ACV was markedly enhanced (ca. 4-fold) following 2 h of quinpirole pretreatment compared with vehicle-pretreated cells (Fig. 6A). Extending agonist incubations to 18 h resulted in an even greater increase in ACV basal activity that was approximately 18-fold greater than vehicle-treated cells (Fig. 6B). Short- and long-term heterologous sensitization of basal levels of ACV-stimulated cyclic AMP accumulation was prevented when quinpirole treatments were carried out in the presence of spiperone (Fig. 6, A and B). The observed agonist-induced increases in basal cyclic AMP levels were specific to ACV/D_2L cells and were not observed with the other cell types examined in the present study (Figs. 1-7).

View larger version (24K): [in this window] [in a new window]   Fig. 6.   Agonist-induced heterologous sensitization of basal and drug-stimulated ACV activity. ACV/D2L cells were pretreated for 2 or 18 h with vehicle or quinpirole (1 µM). Following agonist pretreatment, cells were washed extensively. Basal cyclic AMP accumulation was examined following 2 h (A) and 18 h (B) of agonist treatment. Where indicated some agonist pretreatments were completed in the presence of 1 µM spiperone (+Spip). Isoproterenol- (Iso, 1 µM) and forskolin (FSK, 100 nM)-stimulated cyclic AMP accumulation was also measured following 2 h (C) and 18 h (D) of agonist pretreatment. Data shown are mean ± S.E. of five to nine independent experiments. *p < 0.05 compared with vehicle-treated cells (Student's paired t test).

Effects of Short- and Long-Term D₂ Receptor Activation on ACIX. Agents that selectively activate ACIX are not available, but HEK cells expressing ACIX show marked isoproterenol-stimulated cyclic AMP accumulation compared with HEK-D_2L cells (Fig. 7). Because HEK-D_2L cells showed a trend toward developing heterologous sensitization to isoproterenol (Table 1), sensitization experiments with ACIX/D_2L cells were completed simultaneously with HEK-D_2L cells for direct comparison. Both ACIX/D_2L and HEK-D_2L cells were pretreated with quinpirole for 2 or 18 h and cyclic AMP levels were examined. Pretreatment with quinpirole resulted in sensitization of isoproterenol-stimulated ACIX activity following both 2 and 18 h of treatment. Both short- and long-term agonist-induced sensitization of ACIX was prevented by coincubation with spiperone (Fig. 7). In contrast to effects of agonist treatment in ACIX/D_2L cells, quinpirole pretreatment failed to produce a significant degree (<2-fold) of sensitization to isoproterenol in HEK-D_2L cells. Furthermore, basal levels of cyclic AMP accumulation in ACIX/D_2L and HEK-D_2L cells were not altered following short- or long-term agonist treatment (Table 1; data not shown for ACIX).

View larger version (14K): [in this window] [in a new window]   Fig. 7.   Agonist-induced heterologous sensitization of isoproterenol-stimulated cyclic AMP accumulation in ACIX/D2L cells. Cells were pretreated for 2 h (A) or 18 h (B) with vehicle or quinpirole (1 µM). Following pretreatment, cells were washed extensively and cyclic AMP accumulation was stimulated with isoproterenol (1 µM) for 15 min. Where indicated, some pretreatments were completed in the presence of 1 µM spiperone (+Spip). Basal levels of cyclic AMP accumulation in vehicle-treated cells (18 h) were 3.0 ± 0.9 pmol/well for HEK-D2L cells and 2.1 ± 0.4 pmol/well for ACIX/D2L cells (n = 6 for both cell lines). Data shown are mean ± S.E. of four to six independent experiments. *p < 0.05 compared with vehicle-treated cells (Student's paired t test).

	Discussion

Top Abstract Introduction Experimental Procedures Results Discussion References

Sensitization of the cyclic AMP signaling pathway is a neuroadaptive response that may be involved in drug dependence, antidepressant drug therapy, and schizophrenia (Memo et al., 1983; Chen and Rasenick, 1995; Nestler and Aghajanian, 1997; Duman, 1998; Bohn et al., 2000). Mechanistic studies of sensitization in native tissues are difficult because individual brain regions routinely express multiple isoforms of adenylate cyclase. In the present study, we examined heterologous sensitization of select neuronal isoforms of adenylate cyclase by stably transfecting the D₂ dopamine receptor with individual isoforms of adenylate cyclase in HEK293 cells. We used protocols or conditions that would allow us to selectively activate ACI, ACII, ACV, ACVIII, or ACIX to determine the isoform specificity and the potential mechanisms for agonist-induced heterologous sensitization of adenylate cyclase.

The present study suggests that isoform specificity for short- and long-term sensitization is similar and that agonist-induced heterologous sensitization is time-dependent for the isoforms of adenylate cyclase examined here. We found that a significant degree of sensitization occurred rapidly (within 2 h) and extending the agonist pretreatment to 18 h appeared to enhance the observed magnitude of sensitization for ACI, ACII, ACV, and ACIX. These observations are consistent with previous studies examining ACI and ACV (Watts and Neve, 1996; Avidor-Reiss et al., 1997; Nevo et al., 1998; Rhee et al., 2000). A study completed in HEK293 cells reported that short-term activation (30 min) of M₂ muscarinic receptors sensitizes drug-stimulated ACVI activity, but not ACI or ACII activity (Thomas and Hoffman, 1996). These results may be explained by the short agonist treatment time and the fact that ACVI (and ACV) may show more rapid and robust sensitization compared with other adenylate cyclase isoforms.

Previous studies of recombinant isoforms of adenylate cyclase have suggested that sensitization is also influenced by the choice of agent or condition used to examine cyclic AMP accumulation (Thomas and Hoffman, 1996; Watts and Neve, 1996). Similar to our previous work, agonist pretreatment of ACII/D_2L cells resulted in a marked degree of sensitization to PMA-stimulated cyclic AMP accumulation (Watts and Neve, 1996). We also found that agonist treatment failed to alter forskolin- and isoproterenol-stimulated cyclic AMP accumulation in ACII/D_2L cells. The lack of ACII sensitization to these two agents appears to be consistent with other reports using nonselective activators such as forskolin, G_s-coupled receptor agonists, or constitutively active G_s (Thomas and Hoffman, 1996; Avidor-Reiss et al., 1997; Nevo et al., 1998; Rhee et al., 2000). In fact, agonist treatment of cells expressing ACII results in a significant decrease in stimulated cyclic AMP accumulation (Nevo et al., 1998; Rhee et al., 2000). The observed sensitization of ACII to protein kinase C activation is in contrast to one study that found that chronic µ-opioid receptor did not induce sensitization of phorbol ester-stimulated ACII activity (Avidor-Reiss et al., 1997). The reason for this discrepancy is unclear and may be due to differences in the G_i/o-coupled receptors (D₂ versus µ), the choice of cell lines (HEK293 versus COS 7), transfection methodology (stable versus transient), or other laboratory procedures (Watts and Neve, 1996; Avidor-Reiss et al., 1997). Specificity for sensitization of drug-stimulated cyclic AMP accumulation was also observed for other isoforms of adenylate cyclase. For example, long-term agonist treatment in ACI/D_2L cells enhanced the cyclic AMP response to A23187 by 5-fold, whereas the response to 10 µM forskolin was enhanced only 2.3-fold (data not shown) compared with the 7-fold sensitization observed in HEK-D_2L cells. Sensitization specificity was also observed for isoproterenol-stimulated ACIX activity compared with basal levels of cyclic AMP accumulation in ACIX/D_2L cells.

In addition to the specificity described above, unique sensitization characteristics were observed upon examining sensitization of ACV-modulated cyclic AMP accumulation. Initial studies in ACV/D_2L and HEK-D_2L cells revealed sensitization of ACV when 1 µM isoproterenol or 100 nM forskolin was used to stimulate cyclic AMP accumulation in both cell lines. Moreover, ACV was the only isoform that exhibited sensitization of basal cyclic AMP accumulation. Opioid receptor stimulation also caused sensitization of the basal activity of ACV and the closely related ACVI (Avidor-Reiss et al., 1997). These observations may indicate that agonist pretreatment increases basal activity through a mechanism that is distinct to ACV and ACVI. Recently, a mutant ACV (F400Y) was shown to exhibit enhanced basal activity in the absence of drug stimulation (Zimmermann et al., 1999). Furthermore, the F400Y ACV mutant was also sensitized to activation by forskolin or G_s (Zimmermann et al., 1999). Thus, persistent agonist treatment appears to result in a sensitized phenotype of ACV that shares a number of biochemical properties with the F400Y ACV mutant (Zimmermann et al., 1999).

Short- and long-term heterologous sensitization are blocked by pretreatment with pertussis toxin, implicating G_i/o in heterologous sensitization (Watts and Neve, 1996; Watts et al., 1998, 1999; Rhee et al., 2000). There is also evidence that a decrease in the expression of G_i may be involved in heterologous sensitization (Van Vliet et al., 1993; Reithmann and Werdan, 1995; Watts et al., 1999). More recently, a change in the detergent solubility of G_i was correlated with the onset of heterologous sensitization (Bayewitch et al., 2000). The present results showing that ACVIII is weakly sensitized, combined with observations that ACVIII is only weakly inhibited by G_i/o proteins, also support a strong role for G_i/o proteins in sensitization (Nevo et al., 1998; V. J. Watts and M. G. Cumbay, unpublished observations). However, we also found marked sensitization of adenylate cyclases that are not directly modulated by G_i/o, ACII and ACIX (Taussig and Zimmerman, 1998). Together, these observations suggest that there is an important role of G_i/o proteins in sensitization, but this role appears to differ among recombinant isoforms of adenylate cyclase.

Receptor activation of G_i/o proteins also promotes the release of subunits, and studies have shown that sequestration of subunits (via the coexpression of proteins that bind subunits) prevents the development of heterologous sensitization (Avidor-Reiss et al., 1996; Thomas and Hoffman, 1996; Rhee et al., 2000). Because the individual isoforms are differentially regulated by subunits, the mechanism for -mediated sensitization would be predicted to differ among isoforms. For example, G_s- and protein kinase C-stimulated ACII activity is potentiated by subunits, whereas, the activity of ACI is inhibited by subunits (Tang and Gilman, 1991; Watts and Neve, 1997; Taussig and Zimmerman, 1998), and several isoforms of adenylate cyclase are not modulated by subunits (for review, see Taussig and Zimmerman, 1998). The complexity of -effects are also evident from studies that have shown that sequestration of subunits can enhance the activity of ACV and also block agonist-induced sensitization of ACV (Bayewitch et al., 1998; Rhee et al., 2000). The observations described above suggest that unambiguously identifying the role of subunits in sensitization will be difficult. Furthermore, the effects of subunits will be dependent upon the recombinant isoforms under investigation.

The results discussed above suggest that, similar to their acute regulatory properties, the precise mechanisms of heterologous sensitization of adenylate cyclase may differ between individual isoforms (Thomas and Hoffman, 1996; Watts and Neve, 1996; Avidor-Reiss et al., 1997; Taussig and Zimmerman, 1998; Rhee et al., 2000). In spite of their differential regulation, all recombinant isoforms of adenylate cyclase are activated by G_s and we have proposed that one mechanism of sensitization is enhanced G_s activity or enhanced G_s/adenylate cyclase interactions (Watts and Neve, 1996). We have shown that select recombinant isoforms of adenylate cyclase as well as the endogenous adenylate cyclases expressed in C6 glioma, HEK293, and NS20Y cells are sensitized following persistent D₂ receptor activation (Watts and Neve, 1996; Watts et al., 1998). In addition, several other reports have provided support for a role of G_s in heterologous sensitization of adenylate cyclase (Jones and Bylund, 1990; Chen and Rasenick, 1995; Ammer and Schulz, 1996; Watts and Neve, 1996; Ammer and Schulz, 1997). We now report that persistent activation of the D₂ dopamine receptor results in marked sensitization of ACI, ACII, ACV, ACIX, but not ACVIII. Our present observations may be explained by the ability of these isoforms to be modulated by G_s. Persistent agonist treatment induced a marked degree of heterologous sensitization of A23187-stimulated ACI activity and PMA-stimulated ACII activity. Although the activation of ACI and ACII differs in many ways, both are synergistically activated by G_s under our sensitization conditions, suggesting that they may share an overlapping mechanism of heterologous sensitization (Jacobowitz and Iyengar, 1994; Wayman et al., 1994; Watts and Neve, 1997; Taussig and Zimmerman, 1998). In contrast, ACVIII is not synergistically activated by G_s in presence of A23187 and also shows very weak heterologous sensitization (Cali et al., 1994; Nielsen et al., 1996; present study). Thus, heterologous sensitization of ACI, ACII, and ACVIII may be dependent upon synergistic response to selective activators and G_s. Sensitization studies of ACV also provided additional support for a role of G_s in heterologous sensitization. Finally, ACIX, for which the only direct modulator appears to be G_s, showed robust agonist-induced heterologous sensitization in ACIX/D_2L cells (Hacker et al., 1998; Taussig and Zimmerman, 1998). The observations described above provide strong evidence that G_s is required for sensitization of adenylate cyclase and that the magnitude of sensitization is dependent on the ability of G_s to modulate each isoform.

The precise molecular mechanisms for heterologous sensitization of recombinant adenylate cyclase are unknown; however, the present study suggests a primary role for G_s. We have also demonstrated that the specificity for sensitization is dependent upon the specific modulators or conditions that are used to examine cyclic AMP accumulation following agonist treatment.