Second Messenger Systems

views updated


All cells of the body respond to their environment. For most cell types, the responses can be both general and be related to the cell's particular function. An example of a general response is the regulation of sugar utilization: When sugar is plentiful, glucose is polymerized to glycogen, a storage form. When sugar is scarce, the biochemistry of the cells will be adjusted so that glycogen is broken down to make sugar available. Stimulation of gland cells to release their secretion is a specialized response. In neurons the regulated specialized functions include properties of ion channels; availability of synaptic vesicles for exocytosis (a process called mobilization); and responsiveness of postsynaptic receptors. These functions can be altered transiently (in a time frame of minutes to hours) or for much longer periods, if not permanently. Short-term changes are produced by modifying existing proteins in the cell, modifications that are rapidly reversed when the second-messenger is removed. Long-term changes result when the second-messenger pathway changes transcription, ultimately producing a change in gene expression and the production of new proteins.

Many of these responses, both general and specialized, are produced by signal transduction, a process in which an extracellular stimulus activates a specific receptor on the surface of the responding cell. As a result, the receptor initiates changes in the biochemical state of the cell through the production of a substance that, in turn, alters the cell's responsiveness. This modulating substance is called a second messenger because it is evoked by an environmental cue (the "first messenger"). The process is called signal transduction because the external stimulus is recoded (transduced) into the change in biochemical state. Although ionized calcium (Ca2+) is often thought of as a second messenger, it may be more instructive to call it a primary regulator. Along with membrane potential, it is the ultimate governor of excitability and synaptic transmission in nerve cells.

Historically, formulation of the concepts and principles for understanding second messenger systems arose from studies of two physiological processes: regulation of sugar metabolism in muscle and fat cells (adipocytes), and the conversion of light energy into nerve impulses in the rod cells of the retina. In the 1950s, Sutherland and Rall (1957) discovered the first second messenger, cyclic adenosine monophosphate (cAMP), and Sutherland and his coworkers identified the enzymatic pathway by which cAMP is synthesized and showed how it brings about the changes in sugar metabolism. At about the same time, Wald (1959) investigated how rhodopsin converts the energy of a photon into a chemical signal; the enzymatic and electrophysiological events in phototransduction were thus identified. We now know that both of these processes are mediated by sets of proteins encoded by genes that are closely related phylogenetically and that operate by similar mechanisms.

Production of Second Messenger Molecules

Most second messenger systems follow a similar molecular strategy. The pathway is initiated when an external cue—most often a neurotransmitter, a hormone, an odorant, a physical stimulus such as light, or a mechanical force—activates a receptor. These receptors, which are single polypeptide chains, have seven membrane-spanning ("serpentine") domains. Although these molecules have several domains exposed on the external surface of the responding cell, the actual binding site for the first messenger is slightly buried within the membrane. Some of the receptor is also situated on the intracellular side of the membrane; a part of this intracellular domain associates with another protein, called a G-protein because it binds a guanosine nucleotide (GDP or GTP).

The receptors that usually mediate signal transduction are closely related members of a large gene family that, in addition to rhodopsin, contains adrenergic receptors, muscarinic acetylcholine receptors, and receptors for serotonin, dopamine, histamine, and all neuropeptides. The trimeric G-proteins belong to another large gene family. They are made up of three subunits (αβγ). The α subunit binds GDP when associated with an inactive receptor—a receptor that has not been activated by an external stimulus; it binds GTP when the receptor is activated. More than a dozen isoforms of α subunits have been identified, each of which interacts with different receptors. There are fewer types of the β and γ subunits.

When the α subunit binds GTP because of receptor activation, it dissociates from the βγ portion of the complex. Nevertheless, it retains its association with the inner surface of the cell's external membrane. Depending on the particular combination of receptor and G-protein, the dissociated α subunit either activates or inhibits the next component of the second messenger system, which can be called a primary effector. If the α subunit is stimulatory, its interaction with the primary effector results in synthesis of a second messenger.

The cAMP system is the best-understood intracellular signaling pathway. Binding of a neurotransmitter—such as norepinephrine to the β1-adrenergic receptor—activates a stimulatory G-protein (Gs), which promotes the synthesis of cAMP from adenosine triphosphate (ATP) by adenylyl cyclase. This primary effector enzyme, which is a twelve-membrane-spanning protein, operates only when it is associated with an αs subunit with bound GTP. The αs-cyclase complex also acts as a GTPase, an enzymatic activity that hydrolyzes bound GTP to GDP and Pi. When GTP is replaced by GDP, the αs subunit dissociates from the cyclase and reassociates with the receptor, to be activated again. Some receptors—for example, the muscarinic acetylcholine receptor—interact with an inhibitory G-protein (Gi). The αiα subunit, when activated by the muscarinic receptor, associates with adenylyl cyclase to block the synthesis of cAMP. The opposing actions of norepinephrine and acetylcholine on the cyclase through Gs and Gi represent one form of second messenger interaction called cross-talk.

An important functional aspect of signal transduction pathways, inherent in the arithmetic of the relationships among these components (receptor/G-protein/primary effector), is amplification of the external stimulus. A neuron has far fewer receptor molecules than G-proteins and primary effectors. Amplification occurs when the relatively few stimulated receptors activate many primary effectors. Moreover, since primary effectors are enzymes that produce the second messenger catalytically, this step amplifies the signal even further.

What Second Messengers Do

Second messengers activate secondary effectors, enzymes that are protein kinases in most instances. Typically these enzymes, which catalyze the transfer of the terminal (γ) phosphoryl group of ATP to hydroxyl groups of serine or threonine residues in proteins, are multifunctional: They can phosphorylate many different protein substrates. Secondary effector enzymes include the cAMP-dependent protein kinases (PKA), protein kinase C (PKC), and the Ca2+/calmodulin-dependent protein kinases. Phosphorylation of substrate proteins, which can be called secondary regulators, changes their properties either to stimulate or to inhibit their function. These changes in the activity of substrate proteins by phosphorylation are the means by which the responses to the environmental stimulus are produced. For example, phosphorylation of a channel protein can alter the flux of ions into the neuron to raise or lower the membrane potential. Another example is the phosphorylation of transcription activators, which promotes their binding to DNA. An example important to the neurobiology of memory is the cAMP response element binding protein (CREB). When phosphorylated by cAMP, CREB joins with other proteins to form a complex that binds to the promoter region of genes whose transcription is stimulated by cAMP.

Whether a substrate can be phosphorylated depends on the amino acid sequences around the serine and threonine residues in the substrate protein: Each type of kinase prefers special sequences. Often more than one type of protein kinase can phosphorylate the same protein. The phosphorylations then occur at different sites in the substrate molecule, however.

A serine/threonine protein kinase can exist in several isoforms in the same neuron. The isoforms of each type of kinase are closely related, some encoded by the same genes with diversity produced by alternative RNA splicing and others encoded by distinct but closely related genes. Most protein kinases are related phylogenetically, including tyrosine-specific kinases that phosphorylate proteins on the hydroxyl group of tyrosine residues. Protein kinases that can phosphorylate a variety of substrates are called multifunctional. There are also dedicated kinases that phosphorylate only a special protein substrate (for example, the β -adrenergic receptor kinase) or kinases that are present in only a few kinds of neurons (for example, the cGMP-dependent protein kinase).

PKA was the first protein kinase to be described. This enzyme consists of two regulatory (R) subunits and two catalytic (C) subunits. (There are two major types of R subunits, RI and RII, and several isoforms of C.) The PKAs illustrate how second messenger kinases are regulated: The common mechanism of activation is through binding of the second messenger to release the catalytic center from inhibition. With PKA, the inactive (inhibited) kinase is activated when the concentration of cAMP is elevated within the cell according to the reaction R2C2 + 4 cAMP + 2(R - 2cAMP) + 2C. The two R subunits each bind two molecules of cAMP and dissociate from the C subunits, which, when released, are then free to phosphorylate protein substrates. The free C subunits remain enzymatically active until the concentration of cAMP within the cell falls, which results in the dissociation of the cAMP bound to the R subunits. Lacking cAMP, R subunits reassociate with C subunits, thereby inhibiting the enzyme.

Inositol Polyphosphates, Diacylglycerol and Arachidonic acid

Many receptors activate phospholipase C (PLC) through other G-proteins. PLC catalyzes the hydrolysis of phospholipids in the external membrane of the responding cell. Phospholipids consist of a glycerol moiety esterified at the first (sn 1 position) and second (sn 2) hydroxyl groups to fatty acids, and, at its third, to a diester of phosphoric acid and one of four special alcohols (inositol in phosphatidylinositol, PI; choline in PC; ethanolamine in PE; and serine in PS). In the PI of nervous tissue, the fatty acid at the sn 1 position is usually stearic (an eighteen-carbon saturated fatty acid); the second hydroxyl is usually esterified to arachidonic acid, an unsaturated twenty-carbon fatty acid. The PLC activated in this second messenger system is a diesterase that hydrolyzes PI to an inositol phosphate and diacylglycerol (DAG). Inositol is an unsaturated six-membered cyclic polyalcohol that, in addition to the phosphoryl linkage, can be phosphorylated at hydroxyl groups on the other five carbons. Most often the fourth and fifth hydroxyl groups are phosphorylated in the inositol moiety of nerve cells (PIP2). When hydrolyzed by PLC, it is therefore called inositol 1,4,5-trisphosphate (IP3), a water-soluble second messenger. Other inositol polyphosphates exist, but it is not yet certain whether they, too, act as second messengers. Diacylglycerol, which is soluble only in lipid, remains within the membrane and also serves as a second messenger. PLA2, which hydrolyzes PI at the sn 2 position to release arachidonic acid, also is activated by a G-protein-coupled receptor for histamine and other neurotransmitters. The arachidonic acid produced by receptor-mediated activation of PLA2 can serve as a second messenger itself or it can be converted into many metabolites, some of which alter synaptic transmission and neuronal excitability.

PKC, another multifunctional enzyme, is activated by DAG. (Phorbol esters, tumor-promoting substances from plants, act as potent pharmacological analogues of DAG.) There now are fourteen isoforms of PKC known. All require the presence of membrane lipid, notably PS, to be active. The PKCs fall into four groups (classical, novel, atypical and μ-like), whose enzymatic properties reflect the presence or absence of various functional domains. For example classical PKCs which require CaM2+-ion for enzymatic activity, have a C2 consensus sequence that is responsible for binding Ca2+. Other PKCs lack the sequence and are active independently of Ca2+. The functional significance of this variety of isoforms is not yet known, but structural diversity is presumed to cause differences in substrate specificity and subcellular localization. Unlike PKA, the regulatory and catalytic regions of the PKCs are both parts of a single polypeptide chain. The catalytic part is masked by the regulatory domain. When a lipid activator and membrane (and Ca2+ for the classical forms) bind to the regulatory domain of the enzyme, its conformation changes, exposing the catalytic part of the kinase for action. The dependence of the PKC isoforms on Ca2+ is an important instance of the complexity of regulation by second messengers. In the pathway involving PLC, both DAG and IP3 are formed. The function of IP3 as a second messenger is to bind an intracellular receptor that is located on the cytoplasmic surface of the endoplasmic reticulum. When this receptor is activated, stored Ca2+ is released, thereby raising the intracellular concentration of the free ion.

Ca2+ is required for the action of many enzymes in neurons, either as the free ion or often complexed to calmodulin, a small protein that can bind four Ca2+ ions. In addition to the major forms of PKC, Ca2+ is required by some forms of adenylyl cyclase, guanylyl cyclase, PLC, phospholipase A2 (PLA2), 5-lipoxygenase, some protein phosphatases, and one other kinase, Ca2+/calmodulin-dependent protein kinase II. This multifunctional kinase, like the PKCs, is a single polypeptide chain containing both regulatory and catalytic domains. Unlike either of the other kinases, however, it exists in the cell as a complex of several individual kinase molecules, the exact number varying with the type of cell and with the location within the neuron. This enzyme is quite abundant but is highly concentrated in dendritic spines of neurons where it is the most abundant protein in the postsynaptic density.

MAP Kinases

There are three families of MAP (mitogen activated protein) kinases: kinases of the ERK type, p38 MAP kinases, and JUN kinases. These enzymes are all activated through cascades of kinase kinases, each one specific to the MAP kinase to be activated. The first kinase kinase in a cascade typically is activated by the receptor-mediated mobilization of a small (MW 20,000-40,000) G-protein (Ras, Raf, Rho, Ran, which are distantly related to the α-subunits of the trimeric G-proteins.) This mobilization, which initiates the second-messenger pathway, results in the phosphorylation of the first "upstream" kinase, which then goes on to phosphorylate the next. Finally the last upstream kinase (a MAP kinase kinase, or MAPKK) is a distinctive enzyme that phosphorylates the MAP kinase at two adjacent, amino acid residues, one a threonine, the other a tyrosine.

The activated MAP kinase is imported into the nucleus, where it phosphorylates (and activates) specific transcription factors. Thus, even though MAP kinases do phosphorylate proteins in the cytoplasm (for example, p38 kinase phosphorylates and activates PLA2), their most characteristic function appears to be the second-messenger mediation of changes in gene expression.

Degradation of Second Messengers

cAMP and cGMP are rapidly degraded by several different phosphodiesterases. There are many phosphodiesterase inhibitors that prolong the action of these cyclic nucleotides, including caffeine and theophylline, which occur naturally. Some of these degradative enzymes can also be secondary effectors—for example, in rod cells, where receptor-activated G-proteins activate a cGMP-phosphodiesterase. The second messengers derived from membrane phospholipids are inactivated by being reincorporated into the membrane. IP3 is dephosphorylated by several phosphatases, one of which is blocked by Li+ ion. (This inhibition may be important for the effectiveness of Li+ in the treatment of bipolar depression.)

Although individual protein substrates phosphorylated by the various protein kinases are discussed elsewhere, it is important to point out here that they exist only transiently because of the rapidity with which protein phosphatases remove phosphoryl groups from the proteins. Phosphoryl groups from proteins phosphorylated at serine/threonine residues are removed by one class of protein phosphatases, phosphate on tyrosine residues, by another. With MAP kinases, the phosphoryl groups on both serine/threonine and tyrosine are removed by a specific class of phosphatase.

Variations of Signal Transduction Pathways

Although the typical second messenger system consists of all of the components described, there are examples in which one or more of them is absent. In some instances the receptor-activated G-protein itself acts directly on a secondary regulator—for example, in heart muscle cells, where the G-protein modulates an ion channel without any intervening constituents. Some ion channels are gated directly by cyclic nucleosides, thereby circumventing the need for protein phosphorylation. Guanylyl cyclase can be activated by nitric oxide (NO), a short-lived gas that passes through membranes, bypassing a receptor/G-protein complex. NO is formed in neurons from the amino acid arginine as a consequence of the activation of the enzyme NO synthetase (which requires reduced nicotinamide adenine dinucleotide phosphate and Ca2+ ions) through a second messenger pathway. NO is unusual because it does not act as a second messenger in the neuron in which it is synthesized, but in a neighboring nerve cell. It is possible that arachidonic acid and its metabolites also may act transcellularly as second messengers; in this way they also would bypass receptor, G-protein, and primary effector enzyme. Finally, some secondary effector enzymes are linked directly to their own special receptor through the membrane; important examples of this kind of signaling are the so-called receptor tyrosine protein kinases.



Caron, M. G., and Lefkowitz, R. J. (1993). Catecholamine receptors: Structure, function, and regulation. Recent Progress in Hormone Research 48, 277-290.

Cobb, M. H. (1999). MAP kinase pathways. Progress in Biophysics and Molecular Biology 71, 479-500.

Dekker, L. V., Palmer, R. H., and Parker, P. J. (1995). The protein kinase C related gene families. Current Opinion in Structural Biology 5, 396-402.

Greengard, P. (2001). The neurobiology of slow synaptic transmission. Science 294, 1,024-1,030.

Kennedy, M. B. (2000). Signal-processing machines at the post-synaptic density. Science 290, 750-754.

Nishizuka, Y. (1995). Protein kinase C and lipid signaling for sustained cellular responses. FASEB Journal 9, 484-496.

Schulman, H., and Hyman, S. E. (1999). Intracellular signaling. In M. J. Zigmond, F. E. Bloom, S. C. Landis, and L. R. Squire, eds., Fundamental neuroscience. San Diego: Academic Press.

Schwartz, J. H. (2001). The many dimensions of cAMP signaling. Proceedings of the National Academy of Sciences of the United States of America 98, 13,483-13,484.

Siegelbaum, S. A., Schwartz, J. H., and Kandel, E. R. (2000). Modulation of synaptic transmission: Second messengers. In E. R. Kandel, J. H. Schwartz, and T. M. Jessell, eds., Principles of Neuroscience, 4th edition. New York: McGraw-Hill.

Soderling, T. R. (2000). CaM-kinase: Modulators of synaptic plasticity. Current Opinion in Neurobiology 10, 375-380.

Sutherland, E., and Rall, T. W. (1957). Isolation of cyclic AMP. Journal of the American Chemical Society 79, 3,608.

Takai, Y., Takuya, S., and Takahashi, M. (2001). Small GTP binding proteins. Physiology Review 81, 153-208.

Taylor, S. S., Buechler, J. A., and Yonemoto, W. (1990). cAMP- dependent protein kinase: Framework for a diverse family of regulatory enzymes. Annual Review of Biochemistry 59, 971-1,005.

Wald, G. (1959). Life and light. Scientific American 201, 92-108.

Zang, X., and Majerus, P. W. (1998). Phosphatidylinositol signaling reactions. Seminars in Cell Development and Biology 9, 153-160.

James H.Schwartz