PERSPECTIVE
Plastic Adenylyl Cyclase

Stephen B. Liggett

Departments of Medicine and Pharmacology, University of Cincinnati College of Medicine, Cincinnati, Ohio

Signal transduction via receptors that couple to guanine nucleotide binding proteins (G proteins) is pervasive throughout eukaryotic and prokaryotic biology. In humans, there are thought to be thousands of different G protein-coupled receptors (GPCRs), which carry out an amazingly diverse array of signaling, such as odorant detection, phototransduction, central nervous system neurotransmission, and sympathetic and parasympathetic nervous system functions, to name only a few broad categories. All GPCRs have a similar structure: an extracellular amino-terminus, seven transmembrane-spanning domains, three extracellular and three intracellular loops, and an intracellular carboxy terminus. In simplistic terms, GPCR signaling appears to be like that of a switch: an agonist binds to a receptor, whose conformation then favors binding and activation of a G protein, which then alters the activity of an effector such as an enzyme or channel. In keeping with the analogy to electronic circuits, several aspects of this signaling have become clear (Figure 1). First, GPCR signaling is not like that of a simple switch but rather more like a group of amplifiers, attenuators, and timers. Second, signal splitting occurs immediately after agonist binding, activating a number of different pathways relevant to asthma. In addition, there are multiple levels of feedback inhibition and potentiation. Also, these circuits have both series and parallel components with little in the way of pure signal isolation. Each signal is multistaged so that modulation can occur at various levels.

View larger version (24K): [in this window] [in a new window]   Figure 1.   Schematic of G protein-coupled receptor signaling. Each module represents a receptor (A), G protein (B, D) or effector (C, E, F), which in a series circuit evokes a given cellular response. Within each module, the signaling has characteristics of an amplifier (a) with a variable output (b) and a timer (c). Signal splitting can result in parallel events, such as the coupling of the receptor to a second G protein (D), or when a protein activates two different effectors (E, F). Feedback regulation (d) can occur at points affecting amplitude of the response (such as by AR kinase phosphorylation) or the time course of the response (such as by RGS activity).

Many of the GPCRs couple to three G-protein families: G_s, G_i, and G_q. G_s-coupled receptors are typified by the ₂-adrenergic receptor (₂AR), which serves to stimulate adenylyl cyclase. Such stimulation acts to convert adenotriphosphate (ATP) to cyclic adenosine monophosphate (cAMP), which then activates protein kinase A (PKA). PKA acts to inhibit the production of inositol phosphates, and phosphorylates myosin light-chain kinase, cell surface K⁺ channels, a Na⁺/K⁺ adenosine triphosphase (ATPase), phospholamban, and one or more pumps that lead to sarcoplasmic reticulum uptake of Ca²⁺. The net effect is a decrease in intracellular Ca²⁺ and phosphorylation of contractile proteins, which results in smooth-muscle relaxation (1, 2). Receptors that couple to the inhibitory G protein G_i (such as the M2 muscarinic receptor) inhibit adenylyl cyclase activity. This is particularly obvious when adenylyl cyclase is being activated; but even in the absence of stimulation, basal cAMP levels can be lowered by activation of G_i-coupled receptors. Receptors coupling to the G_q class of G proteins lead to activation of phospholipase C, with generation of 1,2-diacylglycerol and inositol 1,4,5 triphosphate (IP₃). The former activates some protein kinase C (PKC) isoforms, and the latter increases intracellular calcium via IP₃ receptors. The M3-muscarinic receptor and the cysteinyl leukotriene-1 receptor couple through G_q. All G proteins consist of , , and  subunits, which are maintained in the heterotrimeric form in the absence of coupling to a receptor. Contact with the receptor in the agonist conformation results in the binding of the  subunit to the receptor and the dissociation of the subunits, which remain bound to each other. Traditionally, the G subunits have been considered the primary signaling element because they interact with effectors such as adenylyl cyclase. It is now appreciated that G also initiates other signaling pathways independent from the  subunit (3). The intrinsic guanosine triphosphatase (GTPase) activity of activated G, along with the action of a newly discovered group of proteins termed regulators of G protein signaling (RGS proteins), act to terminate the cycle, leading to dissociation of G from the receptor and reformation of the inactive heterotrimer.

The ₂AR of airway smooth muscle is the target of -agonists, used therapeutically for the treatment of bronchospasm in asthma and chronic obstructive pulmonary disease. As such, the mechanisms of ₂AR regulation have been a subject of intense interest. Three broad categories of ₂AR regulation have been identified: genetic (polymorphisms), agonist-promoted, and that due to crosstalk with other receptors (4). Rapid and long-term regulation of signaling by ₂AR and other GPCRs serves an important adaptive function in that it permits the cell to integrate the myriad of signals being received via these receptors. As such, the plasticity of GPCR signaling makes it difficult to know at any given instant which regulatory factors are at play. Nevertheless, substantial progress has been made regarding regulation of ₂AR function by agonist (5). Brief (seconds to minutes) agonist exposure results in phosphorylation of ₂AR by PKA and the AR kinase (ARK). Phosphorylation by the former kinase probably serves to directly depress coupling of the receptor to G_s, whereas ARK-mediated desensitization requires the binding of -arrestins to the phosphorylated receptor for uncoupling to occur. ₂ARs also undergo internalization with agonist exposure. This loss of cell-surface number can serve to further depress cellular responsiveness; internalization also serves to dephosphorylate receptors that have been phosphorylated by ARK. With prolonged agonist exposure, the total receptor complement becomes depressed, a phenomenon called downregulation. The mechanism involves decreases in transcription, messenger RNA stability, and increased receptor-protein degradation.

Compared to what we know about regulation of the receptor, little is known about how adenylyl cyclase itself is regulated, particularly in regard to ₂AR signaling in airway smooth muscle. A summary of events that could take place in smooth muscle and could serve to alter ₂AR signaling is shown in Figure 2. These include mechanisms of receptor crosstalk, which may be particularly relevant to the asthmatic milieu. The study by Billington and colleagues in this edition of the Journal (6) have begun to address issues related to regulation of the cyclase moiety itself within this complex series of events. There are nine cloned isoforms of adenylyl cyclase (7). Each isoform is stimulated by the  subunit of G_s. Only types V and VI are inhibited by the  subunit of G_i. Free subunits stimulate type II. Adenylyl cyclases are amenable to a number of regulatory factors relevant to airway smooth muscle, with some isoforms being substrates for phosphorylation by PKA and PKC. Physiologic levels of free Ca²⁺ regulate types V and VI and perhaps other isoforms. Ca²⁺/calmodulin and CAM kinase IV phosphorylate type I, and CAM kinase II phosphorylates type III adenylyl cyclase. Given some of these differences between adenylyl cyclase isoforms, it becomes imperative at some point to begin to understand how adenylyl-cyclase signaling is regulated in the cell type of interest since there may be a limited expression of different isoforms in such tissues. Using reverse transcription-polymerase chain reaction (RT-PCR) the authors have shown that human airway smooth muscle predominantly expresses type VI adenylyl cyclase, and to a lesser extent, type IX. There may also be some expression of types II and VII. Quantitative assessment of adenylyl cyclase isoform expression in human airway smooth muscle will be necessary to confirm these findings. Further studies by this group show that adenylyl cyclase responsiveness is regulated in a complex way by agonists of several different GPCRs relevant to lung biology. Brief incubation of cells with the muscarinic agonist carbacol resulted in desensitization of ₂AR responsiveness, as well as that of the cyclase as assessed by its responsiveness to forskolin. With 18 h of treatment, ₂AR desensitization was not apparent (indeed, it was slightly augmented), basal cAMP was increased, and the forskolin response was enhanced. M2 muscarinic function was maintained. This "sensitization" of adenylyl cyclase by carbacol was ablated with coincubation with pertussis toxin, implicating M2 muscarinic receptor-G_i interaction as necessary. Similar findings were observed with exposure to serotonin, histamine, and a thromboxane A2 receptor agonist. Since G released from G_i can stimulate phospholipase C and ultimately activate PKC, the effects of PKC inhibition were also examined. Adenylyl cyclase sensitization by long-term agonist was not affected by such inhibition. Since most of the PKC-sensitive adenylyl cyclase isoforms were found to be minimally, if at all, expressed, the data are consistent with the fact that these isoforms have relatively little impact on signaling in human airway smooth-muscle cells. So, whereas acute activation of M2 and other G protein-coupled receptors can desensitize ₂AR and adenylyl cyclase, adenylyl cyclase activity with prolonged activation is enhanced so that desensitization is no longer apparent. The mechanism of such compensation is not known and requires additional study. But does it make sense that the level of function of adenylyl cyclase is the "weak link" in the ₂AR signal transduction pathway? That is, which of the three major components of the ₂AR-G_s-AC pathway is the limiting factor? Billington and coworkers suggest that it is in fact adenylyl cyclase because by transfection of each component, an increase in ₂AR signaling was afforded only with increased adenylyl cyclase expression. This issue is complex and requires more investigation using multiple approaches. We have recently shown, for example, that in the cardiomyocyte, adenylyl cyclase is not the limiting factor in the AR signaling of the normal heart. With the desensitization of this pathway that occurs in a transgenic mouse model of hypertrophy, in vivo and in vitro AR function can be restored by transgenic expression of type V adenylyl cyclase (8), suggesting that in certain pathologic conditions, adenylyl cyclase may become the limiting factor. Interestingly, with marked ₂AR transgenic overexpression (~ 160-fold) we have shown that AR function can also be restored under these same conditions in the cardiomyocyte, but this may be due to coupling to a G_s in a compartment not normally available to the receptor (9).

View larger version (34K): [in this window] [in a new window]   Figure 2.   Complex interactions of the components of G protein-coupled receptor signaling in smooth-muscle cells.

The plasticity of GPCR regulation is once again demonstrated. Inflammatory/contractile agents associated with the asthma state initially result in ₂AR desensitization, but complete normalization (if not enhanced) function of the receptor and its effector occurs with prolonged exposure. It would seem then that the cell has a priority of maintaining ₂AR function, consistent with its beneficial effect on maintenance of airway caliber. As in the film The Graduate, the future still lies in plastics.

	Footnotes

Address correspondence to: Stephen B. Liggett, University of Cincinnati College of Medicine, 231 Bethesda Avenue, Cincinnati, OH 45267-0564. E-mail: Stephen.Ligget{at}UC.Edu

(Received in original form September 9, 1999).

	References

1. Paul, R. J., and P. de Lanerolle. 1996. Regulation of smooth muscle contractility. In Genetics of Asthma. S. Liggett and D. Meyers, editors. Marcel Dekker, Inc., New York. 91-117.

2. Hakonarson, H., and M. Grunstein. 1998. Regulation of second messengers associated with airway smooth muscle contraction and relaxation. Am. J. Respir. Crit. Care Med. 158(Suppl.): S115-S122 [Abstract/Free Full Text].

3. Muller, S., and M. J. Lohse. 1995. The role of G-protein subunits in signal transduction. Biochem. Soc. Trans. 23: 141-148 [Medline].

4. Green, S. A., and S. B. Liggett. 1996. G protein coupled receptor signaling in the lung. In The Genetics of Asthma. S. Liggett and D. Meyers, editors. Marcel Dekker, Inc., New York. 67-90.

5. Liggett, S. B., and R. J. Lefkowitz. 1993. Adrenergic receptor-coupled adenylyl cyclase systems: regulation of receptor function by phosphorylation, sequestration and downregulation. In Regulation of Cellular Signal Transduction Pathways by Desensitization and Amplification. D. Sibley and M. Houslay, editors. John Wiley & Sons, London. 71-97.

6. Billington, C. K., I. P. Hall, S. J. Mundell, J.-L. Parent, R. A. Panettieri Jr., J. L. Benovic, and R. B. Penn. 1999. Inflammatory and contractile agents sensitize specific adenylyl cyclase isoforms in human airway smooth muscle. Am. J. Respir. Cell Mol. Biol. 21: 597-606 [Abstract/Free Full Text].

7. Taussig, R., and A. G. Gilman. 1995. Mammalian membrane-bound adenylyl cyclases. J. Biol. Chem. 270: 1-4 [Free Full Text].

8. Tepe, N. M., and S. B. Liggett. 1999. Transgenic replacement of type V adenylyl cyclase identifies a critical mechanism of -adrenergic receptor dysfunction in the G_q overexpressing mouse. FEBS Lett. 458: 236-240 [Medline].

9. Dorn, G. W. II, N. M. Tepe, J. N. Lorenz, W. J. Koch, and S. B. Liggett. 1999. Low- and high-level transgenic expression of ₂-adrenergic receptors differentially affect cardiac hypertrophy and function in G_q-overexpressing mice. Proc. Natl. Acad. Sci. USA 96: 6400-6405 [Abstract/Free Full Text].