|Home | About | Journals | Submit | Contact Us | Français|
Cyclic adenosine monophosphate (cAMP) was the original “second messenger” to be discovered. Its formation is promoted by adenylyl cyclase activation after ligation of G protein–coupled receptors by ligands including hormones, autocoids, prostaglandins, and pharmacologic agents. Increases in intracellular cAMP generally suppress innate immune functions, including inflammatory mediator generation and the phagocytosis and killing of microbes. The importance of the host cAMP axis in regulating antimicrobial defense is underscored by the fact that microbes have evolved virulence-enhancing strategies that exploit it. Many clinical situations that predispose to infection are associated with increases in cAMP, and therapeutic strategies to interrupt cAMP generation or actions have immunostimulatory potential. This article reviews the anatomy of the cAMP axis, the mechanisms by which it controls phagocyte immune function, microbial strategies to dysregulate it, and its clinical relevance.
This review will provide clinicians with an overview of the cyclic AMP axis, its role as a down-regulator of host antimicrobial defense functions, and the clinical and translational relevance of such actions.
The cyclic nucleotide cyclic adenosine monophosphate (cAMP)—the original member of the family of second messengers—was discovered by Dr. Earl W. Sutherland during his studies of the mechanisms of hormone action. Sutherland was awarded the 1971 Nobel Prize for this work, which would prove to be the first of five Nobel Prizes recognizing research on this molecule. cAMP is now recognized as a universal regulator of cellular function in organisms including amoebas, plants, and humans (1). Biological processes mediated by this second messenger include memory, metabolism, gene regulation, and immune function (2). This review will address cAMP regulation of innate immunity, with an emphasis on the lung. Although many contemporary reviews on innate immunity focus on pathogen recognition receptors and signaling pathways (3), the importance of cAMP as a controller derives from its ability to exert broad modulatory effects that are independent of the pathogen, the recognition receptor, or the signaling pathway in question.
The generation of cAMP is initiated when an extracellular first messenger (neurotransmitter, hormone, chemokine, lipid mediator, or drug) binds to a seven transmembrane–spanning G protein–coupled receptor (GPCR) that is coupled to a stimulatory G protein α subunit (Gαs) (Figure 1). Ligand binding results in the exchange of GDP for GTP on the Gαs protein and its subsequent dissociation from the βγ subunit complex (4). The free Gαs subunit stimulates the enzyme adenylyl cyclase (AC) to catalyze the cyclization of ATP to generate cAMP and pyrophosphate (5, 6). Some of the well-known Gαs-coupled GPCR ligands include epinephrine and norepinephrine, histamine, serotonin, and certain cycloxygenase (COX)-derived prostaglandins (particularly prostaglandin [PG] E2 and I2 [also known as prostacyclin]) (4). By contrast, Gαi subunits inhibit AC and the production of cAMP; some well-known Gαi-coupled GPCR ligands include chemokines CCR1–10 and CXCR1–6 as well as leukotrienes B4, C4, and D4 (4).
To date there are 10 known AC isoforms that are differentially expressed in various cell types (6). Intracellular levels of cAMP are tightly regulated, not only by AC but also by the enzyme phosphodiesterase (PDE) (Figure 1). PDEs oppose cAMP signaling by degrading intracellular cAMP (7). There are 11 distinct PDE gene families whose expression is likewise tissue-specific. Moreover, because both ACs and PDEs can be localized to different spatial compartments within the cell, the evanescent presence of cAMP can be further regulated at a subcellular level (see below) (8).
Diverse cAMP effector molecules also contribute to the complexity of this pathway. Binding of cAMP to the regulatory subunit of protein kinase A (PKA) leads to dissociation of its catalytic subunit, which is then free to phosphorylate specific Ser and Thr residues on numerous target proteins, including the transcription factor cAMP response element–binding protein (9). A-kinase–anchoring proteins serve to localize PKA as well as PDEs within specific cellular microdomains, thereby creating discrete subcellular pools of intracellular cAMP and its effector within a cell. Recently, alternative (PKA-independent) intracellular targets of cAMP, including cyclic nucleotide-gated channels and two isoforms of exchange proteins directly activated by cAMP (Epac), have been identified and their roles in mediating a variety of cellular functions are emerging (10). Previous work has shown that PKA and Epac may have redundant, independent, or even opposing effects within the same cell (11, 12) (discussed more below). cAMP-binding domains of PKA or Epac have been employed as biosensors to measure the spatially discrete pools of intracellular cAMP in live cells (13) using fluorescent resonance energy transfer (14). Cellular functions are ultimately influenced by the ability of these cAMP effectors to regulate activation of transcription factors as well as signaling molecules such as protein kinases, calcium, and small GTPases (4, 15).
Elevations in intracellular cAMP suppress innate immune functions of monocytes, macrophages (ms), and neutrophils (PMNs) (collectively referred to as phagocytes) predominantly through the modulation of three key effector functions of these cells: generation of inflammatory mediators (e.g., cytokine, chemokine, and lipids), phagocytosis, and intracellular killing of ingested pathogens. The host defense function of dendritic cells, important cellular adaptors of innate and acquired immunity, are also highly susceptible to regulation by cAMP but are not discussed here (16). While many mediators of innate immune defense (e.g., chemokines and leukotrienes) act through Gαi-coupled receptors, the extent to which their immunostimulatory effects depend on reductions in intracellular cAMP and the impairment of cAMP effector pathways is uncertain.
Leukocytes respond to microbial invasion by producing a balance of pro-inflammatory and anti-inflammatory mediators. Increased intracellular cAMP suppresses the expression of pro-inflammatory cytokines such as tumor necrosis factor-α (16) and interleukin-12 (17); chemokines, including macrophage inflammatory protein-1α and -1β (16); and the pro-inflammatory lipid mediator leukotriene B4 (18). In contrast, cAMP enhances the production of the anti-inflammatory cytokine interleukin-10 (16) and stimulates the expression of the suppressor of cytokine signaling-3 protein in peripheral blood mononuclear cells and PMNs (19). While the effects of cAMP on m inflammatory mediator generation were originally reported to be mediated by PKA rather than Epac-1 (16), Epac-1 has been implicated in the suppression of endotoxin-induced interferon-β production in a m cell line (20). It is likely that the roles of these two cAMP effectors vary depending on the cell type and the mediator under investigation.
Phagocytosis involves a highly regulated sequence of signal transduction events that lead to cytoskeletal and membrane rearrangements and eventually particle engulfment (21). A variety of specific cell surface receptors recognize microbial pathogens. These include opsonin-dependent complement receptors (CR) and Fcγ receptors (FcR) as well as the opsonin-independent class A scavenger receptors, mannose receptors, and dectin receptors. Elevations in intracellular cAMP suppress CR-, FcR-, and scavenger receptor–mediated phagocytosis (22–24), while it remains unknown whether mannose and dectin receptor–mediated phagocytosis are influenced by cAMP.
The molecular pathways involved in the inhibition of phagocytosis by cAMP are not completely defined. cAMP has been shown to both inhibit (25) and promote (26) F-actin polymerization during FcR-mediated phagocytosis. cAMP may also down-regulate phagocytosis through the modulation of phagocytic receptor expression. For example, Nambu and coworkers reported that cAMP decreased expression of the stimulatory Fcγ receptor I and increased expression of the inhibitory Fcγ receptor IIb in the monocytic cell line U937 (22). In contrast, stimulatory Fcγ receptor IIa expression was increased as a result of PKA-CREB activation in the m-like cell PL8 (27).
It was recently demonstrated that Epac-1, but not PKA, was involved in the cAMP-dependent reduction of FcR phagocytosis by alveolar macrophages (AMs) (11). However, others have reported that both PKA and Epac-1 can counterregulate phagocytosis in different cell types. Makranz and colleagues (24) showed that both PKA and Epac-1 suppress myelin phagocytosis by microglia and peritoneal ms. Bryn and coworkers (28) reported that PKA regulated FcR phagocytosis in circulating monocytes, while both Epac-1 and PKA inhibited this function of monocyte-derived ms. Canetti and colleagues recently reported that Epac-1 effects on FcR phagocytosis in AMs are mediated by the activation of the tyrosine phosphatase SHIP-1, resulting in increased activity of the phosphatase and tensin homolog on chromosome 10 (29).
cAMP also suppresses the microbicidal capacity of leukocytes toward bacteria (11, 30, 31), viruses (32), fungi (33), and eukaryotic parasites (34). The mechanisms by which it does so are not well understood. Potential targets of cAMP regulation include the production and release of reactive oxygen intermediates (ROIs), reactive nitrogen intermediates, phagosomal acidification, and lysosomal enzyme release.
cAMP down-regulates ROI release in phagocytes stimulated with a myriad of agonists (31, 35, 36). This effect is associated with, and presumably involves, inhibition of two pivotal steps in NADPH oxidase activation, namely, the phosphorylation as well as the translocation of the cytosolic p47phox subunit to the cell membrane (31, 35–37). However, the molecular mechanisms underlying these effects are poorly understood. While it has been shown that cAMP-activated PKA inhibits ROI release in phagocytes stimulated with formyl peptides, phorbol esters, and IgG-coated targets (31, 35–37), our group has shown that activation of both PKA and Epac-1 can inhibit ROI generation in Klebsiella pneumoniae–infected AMs (11).
Data on the regulation of reactive nitrogen intermediate generation by cAMP are conflicting. On the one hand, cAMP was found to be pivotal in the enhancement of inducible nitric oxide (NO) synthase expression and NO production by a mechanism dependent on PKA activity (38, 39). In addition, cAMP increased the stability of inducible NO synthase protein in a m cell line (40). On the other hand, elevation of cellular cAMP levels inhibited inducible NO synthase activation in hepatic ms (41).
Contradictory findings also apply to the regulatory role of cAMP on phagolysosome maturation. Kalamidas and coworkers (42) showed that increased cAMP levels decreased phagolysosome formation and acidification by a mechanism dependent on PKA activity, while Di and colleagues (43) demonstrated that cAMP and PKA activity were required for phagosomal acidification. It is of interest that machinery involved in cAMP formation and signaling has been localized to the phagosome. PDE4 was targeted to the forming phagosome in PMNs, where it was colocalized with the catalytic subunit of PKA (44). In addition, Brock and coworkers (45) found that both Epac-1 and its downstream effector, the small GTPase Rap1, associated with phagosomes containing IgG-opsonized targets in AMs. Epac-1, but not Rap1, appeared to accumulate on maturing phagosomes after cAMP stimulation.
Underscoring the importance of cAMP as a negative regulator of antimicrobial responses is the finding that several pathogenic microorganisms have evolved mechanisms to exploit host cell cAMP signaling as a virulence factor. Microbial pathogens can increase the intracellular cAMP production of host cells directly (reviewed below) or indirectly, through the elicitation of host autocrine and paracrine mediators that secondarily enhance cAMP generation (e.g., adenosine, PGE2, histamine, etc.). Only the former, direct actions are discussed here. By using cAMP to disable host cell phagocytosis, intracellular killing, and inflammatory mediator generation, pathogens gain an upper hand in establishing infection.
The respiratory pathogen Bordetella pertussis (the cause of whooping cough) provides a fascinating example of how a bacterium can derail innate immune defenses simply by overwhelming the cAMP regulatory system of host cells. B. pertussis produces several toxins, including the well-described pertussis toxin (PT) and the less well known AC toxin, CyaA (46). Both PT and CyaA toxins increase host cell cAMP, but by distinct mechanisms. B. pertussis PT catalyzes the ADP ribosylation of the inhibitory Gαi subunit, increasing the intracellular cAMP levels in target cells (47). The actions of PT impair m phagocytosis and ROI generation (48, 49) and suppress PMN recruitment to the lungs during infection (50). CyaA is a pore-forming toxin that, in addition to penetrating host cells, possesses an AC motif that is activated by eukaryotic calmodulin and catalyzes the unregulated conversion of cellular ATP to cAMP (reviewed extensively in Ref. 46). In vivo studies indicate that CyaA primarily inhibits the host defense functions of myeloid phagocytes, such as AMs and PMNs (46). CyaA impairs superoxide production, chemotaxis, cytokine production, and phagocytosis (46, 51, 52). In addition to CyaA, three other calmodulin-dependent AC toxins have been identified, including the edema factor of Bacillus anthracis, the ExoY of Pseudomonas aeruginosa, and the AC of Yersinia pestis (reviewed extensively in Ref. 53).
Although not a respiratory pathogen, Vibrio cholerae, the causative agent of cholera, is perhaps the best known microbial pathogen that uses cAMP amplification as a virulence factor. The cholera toxin (CT) is a member of the superfamily of heterodimeric AB toxins that bind to cell surfaces with their B subunit to facilitate the translocation of the A subunit into the cell. Within the cytoplasm, the A subunit of CT catalyzes the ADP-ribosylation of the stimulatory Gαs subunits of G proteins. After ADP-ribosylation, Gαs binds to AC and constitutively activates it, leading to a sustained increase in intracellular cAMP (54). A structurally related AB toxin that works through a similar mechanism is the heat labile toxin (LT) of Escherichia coli. Although the primary target for LT and CT is the intestinal epithelial cell (the poisoning of which causes diarrhea), data suggest that these toxins also increase cAMP production in phagocytes, resulting in the suppression of innate immune defenses. Thus, LT and/or CT have been shown to inhibit phagocytosis (55), intracellular bacterial killing (30), cytokine production (56), and chemotaxis (57) by monocytes, ms, and PMNs. A growing list of other pathogens are known to hijack host cAMP signaling to impair innate immunity and the reader is referred to other resources for more information (58–61).
In view of the multiple inhibitory effects of increased cAMP on innate immune function (discussed above), it is of substantial interest that a number of conditions predisposing to respiratory infection are characterized by increased levels of cAMP (Table 1). In vitro infection with HIV increases cAMP levels in T cells and accounts for impaired proliferative responses (62). HIV infection exerts a similar effect in ms, which was associated with impaired phagocytosis via both CR (63) and FcR (64). Although HIV infection has been shown to increase m PGE2 synthetic capacity (65, 66), the increased cAMP was not explained by this mechanism (63). It has recently been reported (67) that in vitro exposure of normal AMs to cigarette smoke impaired phagocytic capacity and increased cAMP levels, a finding that might explain reduced phagocytic function in AMs from subjects with chronic obstructive pulmonary disease (COPD) as well as from healthy smokers as compared with healthy control subjects.
Among substances that ligate Gαs-coupled receptors and elevate intracellular cAMP, PGE2 stands out because it is the major PG product of most organs and its synthesis is universally up-regulated during host responses to infection. As will be seen, conditions characterized by overproduction of PGE2 provide clinically relevant examples in which cAMP is, or is expected to be, increased. Increased plasma PGE2 levels have been reported in patients after bone marrow transplantation (68). In a murine bone marrow transplantation model, we have recently observed high PGE2 levels in both lung and peritoneal lavage fluid, as well as overproduction of PGE2 by multiple cell types including AMs, PMNs, and alveolar epithelial cells (69). Importantly, abrogation of PGE2 synthesis by the COX inhibitor indomethacin reversed both the in vitro phagocytic defects in AMs and PMNs as well as the in vivo defect in pulmonary bacterial clearance observed in these transplanted mice. Similarly, a bactericidal defect in PMNs from guinea pigs after thermal burn injury has been linked to increased intracellular cAMP and overproduction of PGE2 (70), and was completely overcome by treatment with COX inhibitors (71). Interestingly, overproduction of PGE2 has also been reported in a number of other conditions associated with increased susceptibility to infection, including protein-calorie malnutrition (72, 73), cancer (74), infancy (75), aging (76), and cystic fibrosis (77, 78). Several studies have demonstrated that high dose ibuprofen is able to blunt the decline in lung function in patients with cystic fibrosis (79, 80). While this is conventionally regarded as an “anti-inflammatory” strategy, an alternative possibility is that ibuprofen prevents overproduction of immunosuppressive PGE2 and instead represents an “immunostimulatory” strategy.
Short- and long-acting β2-adrenergic agonists are mainstays of therapy in obstructive lung diseases such as asthma and COPD. These agents are employed primarily for their bronchodilator actions, which are clearly attributed to cAMP accumulation in airway smooth muscle cells. A recent study showed that inhaled β-2 agonists salbutamol and salmeterol impaired the clearance of nontypeable Haemophilus influenzae from the murine respiratory tract, an effect which was prevented by the β receptor antagonist propranolol (81). Although the possibility that β-2 agonists might impair antimicrobial responses in patients has received little attention, recent data showing that salmeterol reduced the number of exacerbations in patients with COPD (82) suggests that this may not be a clinically relevant action. PDE inhibitors are currently under development for obstructive lung diseases, and these were shown to enhance the suppressive effect of PGE2 on AM phagocytosis (83).
Ligands that act via Gαs-coupled receptors and increase cAMP are used therapeutically in a variety of conditions. For example, analogs of PGI2 are commonly used for the treatment of pulmonary arterial hypertension. Interestingly, it was recently reported that patients receiving the PGI2 analog treprostinil had a higher rate of bloodstream infections than did patients receiving the alternative analog epoprostenol (84). Treprostinil inhibited phagocytosis, bacterial killing, and cytokine generation in AMs to a much greater degree than other PGI2 analogs, and this was due in part to the previously unknown ability of treprostinil to bind and activate the E prostanoid 2 receptor for PGE2 (85), whose expression in AMs exceeds that of the IP receptor for PGI2.
Nonsteroidal anti-inflammatory drugs (NSAIDs) are widely used in clinical practice for their analgesic and anti-pyretic actions, which depend primarily on inhibition of COX and resultant inhibition of PGE2 and other PGs. Beneficial effects of NSAIDs on bacterial clearance were mentioned above in animal models of bone marrow transplantation and burn injury, as well as in patients with cystic fibrosis. Additional studies demonstrate that administration of these agents enhanced microbial clearance and/or survival in murine models of infection with Mycobacteria tuberculosis (86), Leishmania amazonensis (87), and Taenia cysticercosis (88), and likewise enhanced AM phagocytosis of Klebsiella pneumoniae (83). Although not tested explicitly, we speculate that PG inhibition by NSAIDs in these studies leads to reductions in intracellular cAMP, which may account for the immunostimulatory effects of NSAIDs in these models.
Alterations in cAMP levels can profoundly influence the innate immune functions of phagocytes, with increased cAMP generally down-regulating inflammatory mediator generation, phagocytosis, and microbial killing. The immunosuppressive actions of cAMP are broad, applying to a variety of innate immune cell types and their interactions with the entire gamut of microbial organisms. The circumstances resulting in cAMP perturbations are myriad and common, reflecting the actions of host-derived molecules, pathogen-derived molecules, and pharmacologic agents. The molecular mechanisms and the clinical impact of such changes in cAMP on innate immune function remain incompletely defined. However, a better understanding of the cAMP axis is likely to provide new insights into the regulation of innate immunity that can be translated into therapeutic benefit.
This work was supported by National Institutes of Health grants HL078727 and HL058897 to M.P.-G, a Hartwell Foundation fellowship (M.N.B.), and a Doris Duke Clinical Scholar Award (D.M.A.).
Originally Published in Press as DOI: 10.1165/rcmb.2008-0091TR on March 6, 2008
Conflict of Interest Statement: None of the authors has a financial relationship with a commercial entity that has an interest in the subject of this manuscript.