Dual Role for Adenine Nucleotides in the Regulation of the Atrial Natriuretic Peptide Receptor, Guanylyl Cyclase-A*

The ability to both sensitize and desensitize a guanylyl cyclase receptor has not been previously accomplished in a broken cell or membrane preparation. The guanylyl cyclase-A (GC-A) receptor is known to require both atrial natriuretic peptide (ANP) and an adenine nucleotide for maximal cyclase activation. When membranes from NIH 3T3 cells stably overexpressing GC-A were incubated with ATP, AMPPNP, or ATPγS, only ATPγS dramatically potentiated ANP-dependent cyclase activity. When the membranes were incubated with ATPγS and then washed, GC-A now became sensitive to ANP/AMPPNP stimulation, suggestive that thiophosphorylation had sensitized GC-A to ligand and adenine nucleotide binding. Consistent with this hypo- thesis, the ATPγS effects were both time- and concentration-dependent. Protein phosphatase stability of thiophosphorylation (ATPγS) relative to phosphorylation (ATP) appeared to explain the differential effects of the two nucleotides since microcystin, β-glycerol phosphate, or okadaic acid coincident with ATP or ATPγS effectively sensitized GC-A to ligand stimulation over prolonged periods of time in either case. GC-A was phosphorylated in the presence of [γ32P]ATP, and the magnitude of the phosphorylation was increased by the addition of microcystin. Thus, the phosphorylation of GC-A correlates with the acquisition of ligand sensitivity. The establishment of an in vitro system to sensitize GC-A demonstrates that adenine nucleotides have a daul function in the regulation of GC-A through both phosphorylation of and binding to regulatory sites.

Atrial natriuretic peptide (ANP) 1 is produced principally within the heart but is also synthesized in many other areas of the body (1). Major effects of ANP include the induction of natriuresis and diuresis in the kidney, the inhibition of adrenal gland aldosterone synthesis, and the relaxation of vascular smooth muscle (1). The two major binding proteins for ANP, the natriuretic peptide clearance receptor (NP-CR) and a guanylyl cyclase-linked receptor (GC-A), are found in the above target tissues as well as in many other regions of the body (1). Although NP-CR, a 65-kDa disulfide-linked dimer, has been suggested to function as a signaling molecule in addition to its role as a clearance receptor (2), GC-A appears to mediate a majority, if not all, of the known physiological effects of ANP. Mice lacking genes for either ANP or GC-A display an elevated form of blood pressure that has been suggested as salt-resistant (3,4) or salt-sensitive (5,6). In the GC-A deficient mouse infused ANP is unable to elicit a diuretic, natriuretic or smooth muscle relaxant effect (7).
Guanylyl cyclases, which catalyze the formation of cGMP, are divided into two classes: one is the family of soluble heterodimeric enzymes that are receptors for NO and possibly CO (8), and the other is the family of transmembrane receptorlinked enzymes, for which GC-A is a prototypical member (9). In mammals, 7 transmembrane forms of guanylyl cyclase (GC-A-G) are known to exist (10 -14), whereas more than 25 putative guanylyl cyclases have been identified in Caenorhabditis elegans (15). These transmembrane proteins share several conserved features, including an amino-terminal extracellular domain separated by a single transmembrane segment from intracellular protein kinase-like and cyclase catalytic domains (10).
GC-A appears to exist as a higher ordered oligomer in the absence of ligand (16,17), and it has been suggested that ANP binding induces a conformational change, perhaps resulting in relief of protein kinase homology domain inhibition of the cyclase catalytic domain (18). ANP binding to GC-A causes dramatic increases in intracellular cGMP levels, and the resultant homologous desensitization has been tightly correlated with GC-A dephosphorylation. Serine and threonine represent the phosphorylated amino acids in the basal state (19). Agents that activate protein kinase-C, such as certain pressor hormones and phorbol esters, also have been shown to desensitize GC-A (20 -24). This heterologous desensitization also correlates with GC-A dephosphorylation (25). The tryptic phosphopeptide maps of GC-A appear different, dependent on whether desensitization is homologous or heterologous. ANP results in maps indistinguishable from untreated cells, although dephosphorylation occurs, while phorbol esters result in the disappearance of one major tryptic phosphopeptide (25). ANP, therefore, may cause complete dephosphorylation of a population of GC-A while phorbol esters result in selective dephosphorylation (25). Possibly, ANP activates a protein phosphatase or induces a GC-A conformation conducive to dephosphorylation by a constitutively active protein phosphatase, whereas phorbol esters activate a protein phosphatase that selectively dephosphorylates due to the absence of ligand. Alternatively, the homolo-gous or heterologous desensitization could result from selective effects on protein kinases.
Although a modest ligand-dependent stimulation of purified GC-A occurs showing that other factors are not absolutely required (26), initial attempts to stimulate the enzyme in broken cell preparations were not successful. Subsequent studies demonstrated an absolute requirement of adenine nucleotides for facilitation of ANP signaling (27)(28)(29). Since AMPPNP is resistant to hydrolysis by ATPases and protein kinases, but supports some ligand-dependent stimulation of GC-A, it has been suggested that binding of adenine nucleotides is sufficient to generate ligand-dependent stimulation (27)(28)(29). The ATP binding site has not yet been identified, but since deletion of or point mutations within the protein kinase homology domain disrupt ANP/ATP-stimulated guanylyl cyclase activity, this domain may reflect the adenine nucleotide binding region (18,30,31).
We now demonstrate that adenine nucleotides play multiple roles in the activation of GC-A and for the first time demonstrate in vitro sensitization of GC-A to ligand. In a preparation of crude membranes, ATP␥S, which can serve as a substrate for protein kinases for thiophosphorylation of proteins, sensitizes GC-A to stimulation with ANP and adenine nucleotides. After ATP␥S treatment, even a nonhydrolyzable analog of ATP, AMPPNP, now facilitates stimulation of GC-A by ANP. However, the activity of the enzyme rapidly declines, presumably recapitulating desensitization. Consistent with dephosphorylation of a site-regulating desensitization, not only does the protein phosphatase inhibitor microcystin block ligand-induced inactivation while potentiating ANP/ATP␥S stimulation, but structurally different phosphatase inhibitors enhance ANP/ adenine nucleotide signaling. Furthermore, phosphorylation of GC-A is observed under sensitization conditions, and phosphorylation is increased when microcystin is included in the reaction mixture.

EXPERIMENTAL PROCEDURES
Materials-Rat ANP was from Peninsula Laboratories. Microcystin-LR and okadaic acid were from Life Technologies, Inc. Cyclosporin A was from Calbiochem. ATP␥S and AMPPNP were from Boehringer Mannheim or Sigma, and ATP was from Sigma.
Cell Culture and Preparation of Particulate Fraction-NIH 3T3 cells stably expressing GC-A were prepared as described (25). 3T3GC-A cells were maintained in Dulbecco's modified Eagle's medium supplemented with 10% heat-inactivated fetal bovine serum, 100 units/ml penicillin, 100 g/ml streptomycin, 0.25 g/ml amphotericin B, and 50 g/ml biologically active G418. Particulate fractions were prepared by washing confluent 10-or 15-cm plates with ice-cold phosphate-buffered saline and scraping cells in 0.75-1 ml of homogenization buffer (HB) (50 mM Hepes, pH 7.4, 10% glycerol, 100 mM NaCl, 10 g/ml aprotinin, 10 g/ml leupeptin, and 1 g/ml pepstatin A). Cells from multiple plates were homogenized by sonication and centrifuged at 16,000 ϫ g for 10 min at 4°C. The pellets were resuspended in HB by passing through a 22-gauge needle ten times, and the membranes were then aliquoted, frozen immediately in liquid nitrogen, and stored at -80°C. Protein concentrations were estimated with the BCA Protein Assay Kit (Pierce) using BSA as a standard and typically ranged between 1.5-2 mg of protein/ml. Membranes were diluted to 1 mg/ml in HB prior to guanylyl cyclase assays.
Guanylyl Cyclase Assay-Assays were performed in a reaction buffer containing 25 mM Hepes, pH 7.4, 50 mM NaCl, 0.1 mM GTP, 0.25 mM IBMX, 0.1% BSA, 5 mM MgCl 2 , 10 mM NaN 3 , 1 mM adenine nucleotide unless otherwise indicated, 1 M ANP, 0.5 Ci [␣ 32 P]GTP (NEN Life Science Products, 3000 Ci/mmol), and membranes in a total volume of 100 l for the indicated times. Phosphatase inhibitors also were included at the indicated concentrations. 3 mM MnCl 2 and 0.1% Triton X-100 were included to determine maximal guanylyl cyclase activity. Assays were initiated by the addition of 20 l (usually 20 g of membrane protein) of membranes to the above mixture that had been pre-warmed at 37°C for 20 s and terminated by the addition of zinc acetate/sodium carbonate. [␣ 32 P]cGMP was determined as described previously (32) and was produced in a linear manner as a function of protein concentration.
Primary Incubation and Washing Membranes-Membranes were incubated at 37°C with the indicated variables in an assay buffer containing 25 mM Hepes, pH 7.4, 50 mM NaCl, 0.1 mM GTP, 0.25 mM IBMX, 0.1% BSA, 5 mM MgCl 2 , 10 mM NaN 3 , and the indicated concentration of different adenine nucleotide. Reactions were terminated by addition were added to the assay mixture and incubated for another 10 min to determine ANP-stimulated guanylyl cyclase activity. cGMP formed during the initial incubation was below the detection limit of the assay. Data are from one representative experiment, and error bars show the range of duplicate determinations.
of ice-cold HB (described above) containing 0.2 M microcystin and centrifuged at 16,000 ϫ g for 10 min at 4°C. Pellets were washed with HB/microcystin and resuspended in HB with a 22-gauge needle. Membranes were used immediately for guanylyl cyclase assays as described above. The experiment in Fig. 2 contained 0.5 Ci [␣ 32 P]GTP (NEN Life Science Products, 3000 Ci/mmol) in the initial incubation, and the membranes were not washed before the addition of ligand(s). Membranes were also washed with HB or HB containing 1 M NaCl to determine the effect of washing membranes on either the ability of ATP␥S to sensitize GC-A, or microcystin to potentiate ANP/adenine nucleotide-dependent activity. Membranes were washed twice with HB or HB with 1 M NaCl and subsequently washed with HB and resuspended in HB by passing through a 22-gauge needle. When the ATP␥S effect was examined, membranes were again washed after the incubation with ATP␥S reaction mixture that included 25 mM Hepes, pH 7.4, 50 mM NaCl, 5 mM MgCl 2 , 10 mM NaN 3 , and 1 mM ATP␥S. Differently washed membranes were then subjected to the guanylyl cyclase assay as described.

FIG. 3. ATP␥S Sensitization of GC-A is stable.
A, 3T3GC-A membranes were initially incubated with 1 mM of the indicated adenine nucleotides for 10 min at 37°C and were subsequently washed as described under "Experimental Procedures." 1 mM ATP␥S allowed for maximal sensitization of GC-A (see Fig. 4B). Washed membranes were then assayed for guanylyl cyclase activity for the indicated times in the presence of 1 M ANP and 1 mM AMPPNP. ANP was absolutely required for GC-A stimulation after ATP␥S incubation (data not shown). Data are from one representative experiment, and error bars show the range of duplicate determinations. B, 3T3GC-A membranes were washed with the indicated buffers and incubated in the presence or absence of 1 mM ATP␥S to examine ATP␥S-mediated sensitization of GC-A. The membranes were washed again and assayed for guanylyl cyclase activity as described in panel A. Reversibility of Sensitization-Membranes were incubated with 1 mM ATP and 1 M okadaic acid for 15 min at 37°C followed by washing as described above. Membranes were then incubated at 37°C for increasing times after which they were placed on ice. Some membranes were subsequently incubated with ATP␥S reaction mixture that included 25 mM Hepes, pH 7.4, 50 mM NaCl, 10 mM NaN 3 , 5 mM MgCl 2 , and 1 mM ATP␥S for 15 min at 37°C followed by washing as described. Membranes were then subjected to the guanylyl cyclase assay as described.
Phosphorylation and Immunoprecipitation of GC-A-Membranes were incubated at 37°C for 20 min in an assay buffer containing 25 mM Hepes, pH 7.4, 50 mM NaCl, 0.1 mM GTP, 0.25 mM IBMX, 0.1% BSA, 5 mM MgCl 2 , 10 mM NaN 3 , 10 M ATP, and 200 Ci of [␥ 32 P]ATP (Amersham Pharmacia Biotech), in the presence or absence of 0.2 M microcystin. Reactions were terminated by the addition of immunoprecipitation buffer (HB buffer containing 1% Triton X-100, 10 mM NaPO 4 , pH 7.0, 0.1 M NaF, 1 mM Na 3 VO 4 , 0.1 M okadaic acid, 80 M ␤-glycerol phosphate, 1 M microcystin-LR, and 10 mM EDTA). After rocking at 4°C for 60 min, the detergent extract was cleared by centrifugation at 436,000 ϫ g for 20 min at 4°C. The extract was pre-cleared by initially incubating with a nonspecific antibody (Z660) raised against the carboxyl terminus of GC-C, the heat stable enterotoxin receptor, and protein A-agarose (Pierce). This antibody does not cross-react with GC-A. An antibody (A034) raised against the carboxyl-terminal 15 amino acids of GC-A was added to the cleared extract at a dilution of 1:200 and incubated at 4°C for 60 -120 min. Protein A-agarose was also included in this incubation to precipitate immune complexes. Immune complexes were washed with immunoprecipitation buffer supplemented with 0.1% SDS and 1% sodium deoxycholate, fractionated by SDS-polyacrylamide gel electrophoresis, blotted to Immobilon-P polyvinylidene fluoride membranes (Millipore), and GC-A visualized after autoradiography. 32 P was quantitated using the ImageQuant software on a Molecular Dynamics PhosphorImager.
Immunoblot Analysis-Polyvinylidene fluoride membranes to which immunoprecipitated GC-A had been blotted were blocked for 15 min at room temperature with 5% non-fat dry milk in TBST (Tris-HCl, pH 7.5, 150 mM NaCl, and 0.1% Tween-20). Membranes were washed with TBST and incubated for 60 min at room temperature in 1:2500 dilution of antibody (A034) raised against the carboxyl-terminal 15 amino acids of GC-A. Membranes were washed and incubated with a 1:30,000 dilution of a horseradish-peroxidase-conjugated secondary antibody in TBST. After washing the membrane with TBST, protein bands were detected using the enhanced chemiluminescence detection method (ECL) (Amersham Pharmacia Biotech).

Effect of Different Adenine Nucleotides on GC-A Activity-
Various groups have demonstrated that adenine nucleotides are required in addition to ANP for maximal GC-A activity; it has been suggested that adenine nucleotide binding and not phosphoryl transfer is the regulatory event since a nonhydro- lyzable ATP analog, AMPPNP, also facilitates slight ANP stimulation (27)(28)(29). The adenine nucleotide binding site has been suggested to be within the protein kinase homology domain of GC-A, but direct measurements of adenine nucleotide binding have not been reported. Although the current model of adenine nucleotide regulation of GC-A suggests that binding to a regulatory site is sufficient for transmission of ANP binding to activation of guanylyl cyclase, a positive correlation exists between the phosphorylation state of GC-A and its sensitivity to ligand (19,25). The resensitization of GC-A in a broken cell preparation has not been achieved, and it therefore has remained unclear whether phosphorylation merely correlates with the sensitization/desensitization of the receptor or whether it is a primary regulatory event.
Crude membranes were obtained from cells stably expressing GC-A and different adenine nucleotides were tested for their ability to stimulate the cyclase in the presence of ANP. The effects of the adenine nucleotides were different, in that ATP␥S was markedly more effective that ATP while AMPPNP was almost ineffective in yielding an ANP sensitive receptor (Fig. 1). These results suggested multiple adenine nucleotide requirements for ANP-dependent activation of GC-A, possibly as a substrate for transphosphorylation as well as for binding to a regulatory site.
Therefore, membranes were incubated with adenine nucleotides at concentrations that would not themselves yield an ANP-dependent activation of GC-A, and then ANP alone or ANP plus high concentrations of AMPPNP were added to determine whether such a prior incubation would sensitize the receptor. Low concentrations of ATP␥S markedly sensitized GC-A to subsequent ANP/AMPPNP stimulation (Fig. 2). AMP-PNP at millimolar concentrations in the absence of the first ATP␥S incubation failed to significantly stimulate cyclase activity in the presence or absence of ANP. ATP also failed to yield an ANP/AMPPNP sensitive enzyme, but given that thiophosphorylated proteins are known as particularly poor substrates for protein phosphatases (33), the sensitization of GC-A by prior incubation with low concentrations of ATP␥S but not ATP could be explained by the stability of thiophosphorylation. If a relatively stable thiophosphorylation explained the acquisition of a sensitive cyclase, then incubation with ATP␥S followed by washing of the membrane preparation should now yield an ANP/AMPPNP-sensitive cyclase.
ATP␥S Sensitization of GC-A Is Stable-Membranes were initially incubated with ATP␥S, no adenine nucleotide, or AMPPNP and then washed, and the resulting membranes were assayed for ANP/AMPPNP-stimulated cyclase activity (Fig.  3A). Clearly, ATP␥S but not AMPPNP sensitized GC-A to subsequent ANP/AMPPNP stimulation, consistent with thiophos-phorylation of a regulatory site. When membranes were isolated and subsequently washed with buffers containing 1 M NaCl, the ability of ATP␥S to sensitize GC-A was retained, suggestive that the proteins required for sensitization are tightly associated with the membrane (Fig. 3B).
Time-and Concentration-dependence of ATP␥S Sensitization-When membranes were incubated with 1 mM ATP␥S for increasing periods of time and washed, the resulting membranes displayed a time-dependent increase in ANP/AMPPNPsensitive cyclase activity (Fig. 4A). The membranes incubated without ATP␥S lost the slight ANP/AMPPNP responsiveness as a function of time. This was not due to a decreased capacity of the catalytic domain to synthesize cGMP since Mn 2ϩ /Triton X-100 activity (a treatment thought to maximally activate the cyclase independent of ligand) remained relatively constant throughout the incubation (Fig. 4A). The loss in ANP/AMPPNP responsiveness could be explained if a small amount of residual phosphoamino acid at the putative regulatory site was lost during the incubation. Increasing the concentration of ATP␥S in the initial incubation also increased ANP/AMPPNP-stimu- FIG. 8. In vitro sensitization of GC-A is reversible. ANP/AMP-PNP-stimulated guanylyl cyclase activity was determined in membranes that had been activated in the presence of 1 mM ATP and 1 M okadaic acid followed by washing as described under "Experimental Procedures." Membranes were then incubated at 37°C for the indicated times. To examine resensitization, membranes incubated for 15 min were again incubated with ATP␥S and washed. lated cyclase activity in a concentration-dependent manner (Fig. 4B). The activation by ATP␥S was maximal at approximately 500 M, and the apparent EC 50 was about 50 M.
The difference in concentration dependence for ATP␥S in the first incubation and adenine nucleotide/ANP in the second incubation, which is similar to the previously reported concentration-dependence for adenine nucleotides (data not shown) (28), by itself suggests different mechanisms of action, but it remained possible that ATP␥S became bound to a regulatory site in the first incubation as opposed to serving as a substrate for thiophosphorylation. Thus, AMPPNP was added during the first incubation to determine whether or not it would compete with ATP␥S. In fact, AMPPNP effectively blocked ATP␥S sensitization of the subsequently washed membranes (Fig. 5). A simple interpretation of these results is that AMPPNP effectively competes with ATP␥S to prevent thiophosphorylation.
Phosphatase Inhibitors and Adenine Nucleotides Potentiate Acquisition of GC-A Sensitivity-Given that phosphorylation by ATP or thiophosphorylation by ATP␥S could explain the acquired sensitivity to ligand, microcystin, a potent Ser/Thr protein phosphatase inhibitor, was tested for its ability to potentiate signaling and/or block desensitization of the ATP-or ATP␥S-sensitized receptor. Again, when AMPPNP was the lone adenine nucleotide added, minimal stimulation of GC-A occurred. The presence of microcystin yielded an initially ANP/ AMPPNP-sensitive enzyme, but the activity rapidly declined to near zero. As previously shown, ATP was not particularly effective as an activator, but the addition of microcystin now yielded a markedly ATP-sensitive enzyme. Microcystin also potentiated the response to ATP␥S (Fig. 6A). The putative protein phosphatase targeted by microcystin is likely associated tightly with the membrane given that microcystin potentiated an ANP/ATP stimulation of GC-A even after membranes were washed with buffers containing 1 M NaCl (Fig. 6B).
Although the effect of microcystin and ATP␥S on ANP-stimulated GC-A activity was consistent with inhibition of dephosphorylation of a regulatory site, it remained possible that microcystin acted in another manner. Thus, membranes were incubated with ANP and different adenine nucleotides in the presence of structurally different phosphatase inhibitors (Fig.  7). The general phosphatase inhibitor, ␤-glycerol phosphate, as well as the relatively selective PP1 and PP2A inhibitors, microcystin and okadaic acid, all potentiated GC-A signaling when AMPPNP and ATP were included in the incubation while microcystin and okadaic acid even potentiated the effect of ATP␥S. In contrast, cyclosporin A, which inhibits calcineurin when complexed with cyclophilin, had no effect on ANP/adenine nucleotide-stimulated cyclase activity.
In Vitro Sensitization of GC-A Is Reversible-Phosphorylation is a reversible means by which to regulate numerous physiological processes. The data presented thus far are consistent with a model that phosphorylation of GC-A regulatory sites modulates its sensitivity to stimulation with ANP/adenine nucleotide. Therefore, the reversibility of sensitization was examined. Membranes initially incubated with ATP and okadaic acid acquired the ability to be stimulated with ANP and AMP-PNP. This sensitization, however, declined as a function of incubation time at 37°C after the ATP and okadaic acid had been removed by washing the membranes (Fig. 8). Importantly, the membranes retained an ability to resensitize the now desensitized form of GC-A as demonstrated by the acquisition of ANP/adenine nucleotide sensitivity after a subsequent incubation of membranes with ATP␥S. Sensitization by ATP␥S was not notably reversible, consistent with the stability of thiophosphorylation relative to phosphorylation (data not shown). Thus, the sensitivity of GC-A to ANP is modulated in vitro by appar- ently altering the phosphorylation state of unidentified regulatory phosphorylation sites.
Phosphorylation of GC-A Correlates with Acquisition of Ligand Sensitivity-Previous reports demonstrated that dephosphorylation of GC-A correlated with its loss of sensitivity to ANP (19,25). However, a phosphorylation-induced sensitization of GC-A has not been accomplished. Given that conditions to sensitize GC-A have now been established here (Fig. 3) and phosphorylation of a regulatory site appears to represent a mechanism of sensitization, we examined phosphorylation of GC-A under sensitizing conditions. Phosphorylation of GC-A was observed, and clearly the GC-A-associated 32 P was increased when microcystin was included in the reaction mixture (Fig. 9, A and B). The increase in 32 P was not due to an increased amount of GC-A in the immunoprecipitate (Fig. 9A). Sensitization of GC-A by ATP and microcystin correlated well with GC-A phosphorylation (Fig. 9C), and thus for the first time, phosphorylation of GC-A is shown to correlate with GC-A sensitivity to stimulation by ANP/adenine nucleotides.
Model for Regulation of GC-A Activation-This is the first successful sensitization/desensitization of GC-A, a prototype of the other guanylyl cyclase receptors, in a broken cell preparation. Multiple roles for adenine nucleotides in regulation of the receptor are now clearly established based on this work. Phosphorylation of one or more regulatory sites as well as adenine nucleotide binding are required to generate a ligand-sensitive cyclase (Fig. 10). Although it is possible that phosphorylation of other sites participates in different aspects of GC-A regulation, such as ANP-independent activity, phosphorylation of these regulatory sites does not itself appear to alter the nonliganded activity of the receptor. The model now raises a number of important questions. The first is the site(s) of phosphorylation. These studies demonstrate that phosphorylation or thiophosphorylation generates an ANP/AMPPNP-sensitive receptor, however, they do not identify the localization of the regulatory sites. These could be within the cyclase itself or they could reside on associated proteins.
We have not established that adenine nucleotide binding is to the cyclase although the presence of the protein kinase homology domain is strongly suggestive that it represents the site of binding. Relatively high (millimolar) concentrations of ATP or AMPPNP are required to facilitate ANP signaling, and thus it remains possible that ATP or AMPPNP mimic the actions of a much more potent nucleotide regulator. The ability to obtain sensitized GC-A in vitro now allows searches for more potent regulators. Importantly, the ability to sensitize GC-A with ATP␥S and prevent desensitization with phosphatase inhibitors is retained even after washing membranes with buffers with high salt concentrations, demonstrating that the components required for sensitization and desensitization of GC-A likely reside in the membrane itself ( Fig. 3B and Fig. 6B).
The model also predicts several potential means by which to alter GC-A signaling. Other signaling systems could impinge on the sensitizing protein kinase or the putative desensitizing phosphatase, and given the dual role of adenine nucleotides in activation of GC-A, fluctuations in intracellular levels of ATP or of another regulatory nucleotide could also regulate its activity.