T I B S - May 1985
Biological roles of cAMP: similarities and differences between organisms Juana M. Gancedo, Maria J. Maz6n and Pilar Eraso The mode of action of cAMP is completely different in eukaryotes and in prokaryotes, although the proteins which bind cAMP in both cases probably have a common evolutionary origin. Some patterns of regulation through cAMP have been conserved among very different organisms but there are cases where the role of cAMP changes widely from one type of cell to the other. In particular, cAMP cannot be considered a universal 'hunger' signal. Since the discovery of cAMP by Sutherland in the late 1950s, this metabolite has been implicated in many types of cellular processes in a variety of organisrns ranging from bacteria to mammals. However the studies on the physiological role of cAMP have not always proceeded in a systematic fashion and this has resulted in a large mass of data whose complexity makes general patterns of action difficult to recognize. In this article, we will attempt to organize disperse knowledge about cAMP action by considering the mechanisms which bring about changes in cAMP concentrations as well as the effect of these on the metabolism of prokaryotic and eukaryotic cells.
a carbohydrate-rich diet. In various fungi, no correlation has been found between the carbon source available and cAMP concentrationsL Changes in cAMP concentrations are most often produced by signals acting on the cell membrane. Depolarization of the plasma membrane increases cAMP in nerve cells6, as in Neurospora and possibly in other fun# 1,7. In higher eukaryotes, cAMP changes generally depend on the action of hormones. A set of hormones, often secreted as a response to stress (epinephrine, glucagon, histamine, vasopressin, etc.) increases cAMP concentrations, while other hormones like ct-adrenergic Catecholamines or insulin either decrease cAMP or antagonize the action of the Changes in cAMP concentrations first set. In lower eukaryotes, analogues A first step in assessing the regulatory to the hormonal signals are found. In role of. cAMP is the measurement of Dictyostelium discoideum, extracellular variations in the intraceUular con- cAMP binds to a specific receptor and centration of cAMP in different meta- causes an increase in intracellular bolic situations. Such measurements, cAMP (Ref. 8), while in yeast the however, are not completely straightfor- mating pheromone a-factor has been ward ~-3 and reliable data are scarce. reported to inhibit adenylate cyclase There seems to be no general rela- and is therefore thought to depress tionship between cAMP concentrations cAMP concentrations9. and the nutritional situation of cells. In E. coli, cAMP concentrations are low in Control of cAMP concentrations the presence of glucose and increase The intracellular cAMP concentration about 10-fold when other carbon sources depends on the balance between the are available3. In Pseudomonas, how- activities of adenylate cyclase and phosever, cAMP concentrations appear to be phodiesterase. In addition, changes in independent of the carbon source 2. In the amount and/or affinity of intracellueukaryotic cells, there is no uniform pat- lar proteins able to bind cAMP have tern either. In the presence of glucose, been postulated as a way of modulating cAMP is high both in Tetrahymena 4 and cAMP concentrations~. Although bacin yeasts5, while in liver cAMP con- teria as well as animal cells can excrete centrations fall when the animal is given cAMP, there is no evidence that this excretion regulates the intracellular cAMP concentration. J. M. Gancedo, M. J. Mazrn and P. Eraso are at In most organisms, changes in the the lnstituto de lnvestigaciones Biorn~dicas del CSIC and Departamento de Bioqulmica, Facultad activity of adenylate cyclase appear to de Medicina de la Universidad Autrnoma be the major point of control. Adenylate Arzobispo Morcillo, 4. 28029 Madrid, Spain. cyclase from higher eukaryotes is found ~) 1985. Elsevier Science Publishers B.V.. Amsterdam 0376 5067/85/$02.(~)
in the plasma membrane as part of a complex regulatory system which mediates the action of a variety of hormones. This system includes receptors for hormones and neurotransmitters and two different regulatory components (G, and G O which bind GTP and mediate the stimulation and the inhibition of the adenylate cyclase, respectively~°. Adenylate cyclase activity can also be modulated by calcium concentrationsit. In addition, it has recently been proposed that phosphorylation of the stimulatory Gs component by protein kinase C inhibits the cyclase from hepatocytes12. In lower eukaryotes, the adenylate cyclase system has not been studied in such detail. There is evidence, however, that in some species like
Saccharomyces cerevisiae, Neurospora crassa and Dictyostelium discoideum the system is similar to that found in higher eukaryotes, including a catalytic subunit and a regulatory component that binds GTP (Ref. 13). It is not yet clear if there is some equivalent to the receptor subunit that could also control adenylate cyclase activity. In bacteria, adenylate cyclase is a single protein. Its activity appears to be modulated by one of the components of the phosphoenolpyruvate-dependent transport system of carbohydrates, the so-called protein III°~. This protein is phosphorylated when no glucose is present in the medium and, in this form, it activates the cyclase3. It has also been proposed that the proton electrochemical gradient may directly regulate adenylate cyclase activity and that the cAMP receptor protein itself (CRP) could repress the synthesis of adenylate cyclase3. Changes in phosphodiesterase activity ~to not seem to play a regulatory role in prokaryotes 3 but can contribute to variation in cAMP concentrations in eukaryotes. In such organisms several phosphodiesterase isoenzymes which can be the target of different regulatory mechanisms have been identified. In higher eukaryotes phosphodiesterase can be activated by calcium--calmodulin, by an increase in the ratio NADH/NAD or by phosphorylation of membrane components~1. In lower eukaryotes, cAMP concentrations are sometimes inversely correlated with the activity of phosphodiesterase. cAMP on prokaryotes The mode of action of cAMP in E. coli has been studied in great detail with respect to effects on gene expression. Cyclic AMP forms a complex with a specific cAMP receptor protein (CRP)
T I B S - May 1985 and the cAMP--CRP complex interacts enzymes involved in glycogen synthesis with particular stretches of DNA to and degradation. exert a dual control on the initiation and A cAMP-dependent protein kinase termination of transcription of certain can also control a metabolic pathway operons3. The best studied systems a r e indirectly by changing the conoperons subject to catabolite repression centrations of a regulatory metabolite. by glucose, like those for the utilization For instance in liver, there is such a of lactose, arabinose or maltose. It 'protein kinase which modulates the should be stressed that although cAMP activity of the bifunctional enzyme is essential for relieving catabolite responsible for the synthesis and degrarepression, this phenomenon is not dation of fructose 2,6-bisphosphate, an modulated exclusively by changes in the important effector of glycolysis and concentrations of cAMP (Ref. 3). In gluconeogenesis15. addition to its well-known role in For a large group of cAMP actions derepression, cAMP can also repress the mediated by protein phosphorylation, synthesis of several E. coli proteins and the target proteins may not be enzymes this negative control depends also on a but proteins which control processes like functional CRP (Ref. 3). cell division, meiosis, transcription, No cAMP-dependent protein kinase translation, membrane permeability, has been identified up to now in any hormone secretion or muscle relaxation. bacteria. However, phosphorylation of Some representative examples are given proteins has been detected in bacteria below. and bacterial protein kinases have been Although the influence of cAMP on characterizedTM. cell proliferation has been the subject of considerable controversy, cAMP has Effects of cAMP in eukaryotes recently been shown to act as a Whereas the only known function of mitogenic signal both in animal cells16 cAMP in bacteria is to regulate gene and in yeast17. In this last case, there is expression, this metabolite produces a evidence that cAMP works via activagreat variety of effects in eukaryotic tion of a cAMP-dependent protein cells. In most cases where the mech- kinase. On the other hand, an elevated anism of cAMP action has been eluci- protein kinase activity blocks the initiadated, it has been shown to involve the tion of meiosis, as shown in Xenopus activation of a protein kinase through and Rana oocytesTM and in yeast17. It is dissociation of an inactive complex of suggested that meiosis is inhibited by a regulatory and catalytic subunits. phosphoprotein whose concentration Depending on the type of cell, the acti- will decrease when cAMP convated protein kinase can phosphorylate centrations are lowered. a variety of substrates, thus explaining In rat pituitary cells, an increase in the wide range of cAMP effects. The intracellular cAMP rapidly stimulates developmental cycle of Dictyostelium the transcription of the prolactin gene ~9. discoideum is a peculiar case because The effect of cAMP correlates well with cAMP plays a dual role: intracellular the phosphorylation of a chromatincAMP activates a protein kinase (as in associated protein. In rat cells too, the other eukaryotes) but extracellular genes coding for lactate dehydrogencAMP acts like a hormone, binding to a ase, tyrosine aminotransferase and specific membrane receptor and stimu- phosphoenolpyruvate carboxykinase lating adenylate cyclases. (PEPCK), show transcriptional activaThe phosphorylation of key regula- tion by cAMP2°. Although in these cases tory enzymes of different metabolic the mechanism of action has not yet pathways is perhaps the best studied been elucidated, it has recently been function of cAMP-dependent protein shown that a region at the 5' end of the kinase and the most easily interpreted as PEPCK gene is necessary for cAMP well. Phosphorylation can change the regulation of gene expression21. Cyclic activity of an enzyme not only by affect- AMP may act by changing the phosing its catalytic capacity but also by phorylation state of a nuclear protein, changing its sensitivity to effectors or which in turn affects the transcription modifying its susceptibility towards rate of the gene. Alternatively, as in proteolytic degradation. The effects of bacteria, cAMP could bind to a receptor phosphorylation can be more complex protein which would interact directly when the modified enzyme is a kinase or with the regulatory region of DNA. a phosphatase that in turn modifies a The ribosomal protein, $6, has been second enzyme. The amplifying power shown to be phosphorylated in vivo in of such a cascade is well exemplified in response to cAMP and to growth factors the coordinated regulation of the in all cell types examined. The pos-
211 sibility that differential phosphorylation by a variety of protein kinases could direct translation of different mRNAs has been proposed22, and is supported by studies/n vitro with synthetic messengers. Phosphorylation of membrane proteins by a cAMP-dependent protein kinase increases ion transport in higher eukaryotic cells23. In particular, it has been shown that in Aplysia neurons cAMP mediates the activation of K + conductance induced by serotonin while serotonin stimulation alters protein phosphorylation /n vivo24. In yeast, cAMP enhances membrane permeability towards a variety of compounds by a mechanism which could involve activation of the plasma membrane ATPase 25. The release of adrenocorticotropin (ACTH) from the anterior pituitary, in response to corticoprotein releasing factor, catecholamines or the vasointestinal peptide, appears to be mediated by an increase of intracellular cAMP 26. Cyclic AMP stimulates the phosphorylation of several distinct proteins and it is possible that these endogenous proteins induce ACTH synthesis or the release of granular ACI'H. The best-known examples of cAMPdependent phosphorylation in muscle are those of troponin I in cardiac muscle in response to epinephrine and of myosin (light-chain) kinase of smooth muscleH. In both cases, the phosphorylation is associated with muscle relaxation. The possibility of cAMPdependent phosphorylation of proteins present in sarcoplasmic reticulum causing changes in Ca 2÷ fluxes has been the subject of intense studiesn that show that phosphorylations triggered by either cAMP or Ca 2÷ are to some degree interdependent. To conclude this brief survey, it should be pointed out that although many proteins can be phosphorylated by cAMP-dependent protein kinases, in only a few cases has this phosphorylation been proved to play a physiological role. The metabolic significance of a cAMP-dependent phosphorylation must therefore be carefully established, case by case. General patterns of cAMP action The modes of action of cAMP in different organisms are summarized in Fig. 1. As shown, a sharp distinction can be drawn between prokaryotes and eukaryotes: in prokaryotes, cAMP controls transcription by binding to the CRP protein; in eukaryotes, cAMP controls a
.~ PROKARYOTES X
TRANSDUCTION MECHANISM cAMP-RECEPTOR PROTEIN
M a y 1985
PROCESSES AFFECTED ~ transcription / cellproliferation v
//meiosis membrane depolarization
cAMP-DEPENDENT . PROTEIN KINASE
~ membranepermeability degradationof reserves
stimulation of membranereceptors
\ "~,muscuar "~h°rm°lnesl re eclaxeaoitltni~:n
Fig. 1. Modes of action of cAMP
variety of processes by binding to the regulatory subunit of cAMP-dependent protein kinases. It should be noted however that the two types of protein which interact with c A M P show a significant homology, that points to a common evolutionary origin27. Some patterns of regulation also appear to have been conserved through evolution. One of them is the transient rise of c A M P concentration in the G1 phase of the growth cycle which occurs in prokaryotic aetinomycetes as well as in yeast or in rat fiver cells16. Thus, c A M P appears to allow progress along the cell division cycle. A t low c A M P concentrations, the ceils either remain quiescent or enter into meiosis. A n o t h e r conserved pattern is the increase in c A M P concentrations in response to stress or membrane depolarization. The final effect is often stimulation of the degradation of reserves leading to an increase in the availability of ATP. O n the other hand, many regulatory mechanisms do not remain uniform among organisms. For instance, although c A M P increases the degradation of carbohydrate reserves both in liver and in yeast, c A M P cannot be considered a universal 'hunger' signal since c A M P concentrations increase in the presence of a well utilized carbon source (such as glucose) in yeast s and in TetrahymenaL A n o t h e r example of divergent regulatory mechanisms is the well-known p h e n o m e n o n of catabolite repression; it occurs in bacteria like E. colP a n d B a c i l l u s subtilis 2s and in yeast 5 but only in E . coli is the repression mediated by a decrease in c A M P concentrations. Although regulation through changes
in c A M P concentrations is widespread, it is not a necessary feature in every kind of organism. Some bacteria appear to lack c A M P 2s. In plants, although c A M P has been reported 29, its role is probably ancillary, regulation through calcium being more important. In addition, some mutants of E . coli and yeast lack adenylate cyclase and are therefore devoid of cAMP. Mutant bacteria can thrive in the absence of added cAMP, providing the carbon source does not require induction of enzymes subject to catabolite repression a. In the case of yeast mutants, external c A M P is required unless there is a second mutation that makes the cAMP-dependent protein kinase active, even in the absence of c A M P 17. In $49 lymphoma ceils cAMP-dependent protein kinase can be entirely lacking with no measurable effect on cell growth 3°. It can therefore be concluded that, at least in certain types of cells, regulation through cAMP is dispensable.
Acknowledgement Critical reading of the manuscript by C. Gancedo, A. Pestafia and R. Serrano is gratefially acknowledged.
Rderences 1 Pall, M. L. (1981)Microbiol. Rev. 45, 462--480 2 Phillips, A.T. and Mulfinger, L. M. (1981) Z Bacteriol. 145, 1286-1292 3 Ullmann, A. and Danchin, A. (1983) Adv. Cyclic Nudeotide Res. 15, 1-53 4 Nandini-Kishore,S. G. and Thompson, G. A. (1979) Proc. Natl Acad. Sci. USA 76, 27082711 5 Eraso, P. and Gancedo, J. M. (1984) Eur. J. Biochem. 141, 195--198 6 Drummond, G.I. (1983) Adv. Cyclic Nucleotide Res. 15, 373--494 7 Maz6n, M. J., Gancedo, J. M. and Gancedo,
C. (1982) Eur. Z Biochem. 127, 605-608 8 Devreotes, P.N. (1983) Adv. Cyclic Nucleotide Res. 15, 55--96 9 Thorner, J. (1982) Cell 30, 5-6 10 Gilman, A. G. (1984) Cell 36, 577-579 ql Rasmussen, H. (1981) Calcium and cAMP as synarchic messengers, John Wiley & Sons 12 Heyworth, C. M., Whetton, A. D., Kinsella, A. R. and Houslay, M. D. (1984) FEBS Lett. 170, 38--42 13 Pall, M.L. (1984) Mol. Cell. Biochem. 58, 187-191 14 Cozzone, A. J. (1984) Trends Biochem. Sci. 9, 400--403 15 Hers, H.-G. and Van Schaftingen, E. (1982) Biochem. Z 206, 1-12 16 Boynton, A.L. and Whittield, J.F. (1983) Adv. Cyclic Nucleotide Res. 15, 193-294 17 Matsumoto, K., Uno, I. and Ishikawa, T. (1983) Cell 32, 417-423 18 Mailer, J. L. and Krebs, E. G. (1980) Curt. Top. Cell. Reg. 16, 272-311 19 Murdoch, G.H., Rosenfeld, M.G. and Evans, R. M. (1982) Science 218, 1315--1317 .20 Jungmann, R.A., Kelley, D.C., Miles, M. F. and Milkowski, D. M. (1983) J. BioL Chem. 258, 5312-5318 21 Wynshaw-Boris, A., Lugo, T.G., Short, J. M., Fournier, R. E. K. and Hanson, R. W. (1984) J. Biol. Chem. 259, 12161-12169 22 Burkhard, S. J. and Traugh, J. A. (1983) J. Biol. Chem. 258, 14003-14008 23 Greengard, P. (1978) Science 199, 146-152 24 Lemos, J. R., Novak-Hofer, I. and Levitan, I. B. (1984) Proc. Nag Acad. Sci. USA 81, 3233-3237 25 Foury, F. and Goffeau, A. (1975) J. Biol. Chem. 250, 2354--2362 26 Axelrod, J. and Reisine, T. D. (1984) Science 224, 452-459 27 Weber, I. T., Takio, K., Titani, K. and Steitz, T.A. (1982) Proc. Natl Acad. Sci. USA 79, 7679-7683 28 Price, V. L. and Gallant, J. A. (1983) Eur. J. Biochem. 134, 105-107 29 Brown, E.G. and Newton, R.P. (1981) Phytochemistry 20, 2453-2463 30 Gottesman, M. M. (1980) Cell 22, 329-330