20:  20.7 Interaction and Regulation of Signaling Pathways

The Same RTK Can Be Linked to Different Signaling Pathways

We have seen that activated RTKs can initiate signaling via Ras and the downstream MAP kinase pathway (see Figure 20-6). These receptors also can trigger the inositol-lipid pathway by binding PI-3 kinase and PLC_γ, two of the enzymes needed to form the second messengers IP₃ and DAG (see Figure 20-38). The SH2 domains of PLC_γ, for example, bind to specific phosphotyrosines of certain RTKs, thus positioning the enzyme close to its membrane-bound substrate. In addition, the RTK phosphorylates tyrosine residues on the bound PLC_γ, enhancing its lipase activity. Thus activated RTKs promote PLC_γ activity in two ways: by localizing the enzyme to the membrane and by phosphorylating it.

The initiation of tissue-specific signaling pathways by stimulation of the same receptor in different cells is exemplified by EGF-stimulated signaling in C. elegans. The central importance of RTK-Ras-MAP kinase signaling, stimulated by EGF, in development of the vulva in C. elegans was demonstrated in studies analogous to those described earlier for development of R7 cells in Drosophila (see Figure 20-25). Other genetic studies, however, have shown that stimulation of the EGF receptor triggers a Ras-independent pathway in some tissues. For example, one of the many functions of EGF in C. elegans is to control contractility of smooth muscle, which in turn regulates the extrusion of oocytes from one compartment of the hermaphrodite gonad to another, where they are fertilized. Coupling of the EGF receptor to Ras is not required for the EGF-induced contractions of the gonad. Analysis of several different types of mutation led researchers to conclude that in C. elegans smooth muscle, the EGF receptor is linked to the inositol-lipid pathway. Ligand binding to the receptor leads to an increase in IP₃, which then promotes release of intracellular Ca²⁺ stores (see Figure 20-39). The increased cytosolic Ca²⁺ level then promotes muscle contraction.

Multiple G Proteins Transduce Signals to Different Effector Proteins

We noted previously that studies with bacterial toxins originally led to identification of several G proteins coupled to different effector proteins. More recently, molecular cloning has led to the isolation of a large number of proteins related to the α, β, and γ subunits of these previously characterized G proteins. In mammals, for instance, 16 distinct G_α subunits, 5 G_β subunits, and 12 G_γ subunits have been identified so far. Analysis of the C. elegans genome, whose entire sequence has recently been completed, revealed that this organism also encodes multiple G_α, G_β, G_γ subunits.

Clearly, eukaryotes generally produce multiple trimeric G proteins that link GPCRs to a variety of effector proteins, including ion channels, adenylyl cyclase, phospholipase C, and a cGMP-specific phosphodiesterase in photoreceptor cells (Table 20-5). Stimulation of the receptors coupled to these G proteins thus can modulate many cellular functions. The role of trimeric G proteins in these signaling pathways has been demonstrated by their sensitivity to nonhydrolyzable GTP analogs and, in some cases, by their sensitivity to cholera or pertussis toxin. Because the same cell may express diverse G proteins, it is often difficult to determine which specific G protein mediates the effect of a particular ligand.

The presence of multiple G_α subunits in a single cell raises the possibility that a single ligand could initiate signaling through more than one effector protein. Several examples of such multiplex signaling have been described, although the precise molecular details of which G proteins and which specific subunits mediate these effects are not yet known. In some cells, modulation of different effectors coupled to the same GPCR is observed at different ligand concentrations or when different concentrations of the receptor are expressed on the surface. The large number of possible combinations of different G protein subunits, coupled receptors, and effector proteins provide cells with the ability to respond in remarkably diverse ways to precisely control their development and function.

G_βγ Acts Directly on Some Effectors in Mammalian Cells

In the signaling pathway stimulated by binding of mating factors to haploid yeast cells, the signal is transduced by the G_βγ subunit complex, not by the G_α · GTP complex (see Figure 20-30). This activation of a downstream signaling pathway by G_βγ initially was thought to reflect an idiosyncrasy of yeast biology. More recent research has shown that in some mammalian cells G_βγ can directly regulate certain effector proteins.

For example, one isoform of adenylyl cyclase (ACI) present in the brain is stimulated by G_sα · GTP as described previously (see Figure 20-16); G_βγ inhibits the activation of ACI by G_sα. Another isoform (ACII), however, is stimulated by binding of G_βγ but only if free G_sα · GTP also is present. Yet other adenylyl cyclase isoforms (e.g., ACIII) are insensitive to G_βγ. Some brain cells contain two adenylyl cyclase isoforms that are regulated differently; in such cells adenylyl cyclase activity is subject to dual regulation. Similarly, some K⁺ channel proteins in the heart are opened by binding of G_βγ (see Figure 21-41). In general, considerably higher concentrations of G_βγ than of G_α · GTP are required to modulate the activity of an effector protein. Regulation of certain effector enzymes by G_βγ and various G_α · GTP complexes contributes to the integration of cellular metabolism (Figure 20-43).

Glycogenolysis Is Promoted by Multiple Second Messengers

Earlier we saw that a rise in cAMP induced by epinephrine stimulation of β-adrenergic receptors promotes glycogen breakdown (see Figure 20-35). In both muscle and liver cells, other second messengers also produce the same cellular response.

Stimulation of muscle cells by nerve impulses causes release of Ca²⁺ ions from the sarcoplasmic reticulum and an increase in the cytosolic Ca²⁺ concentration. This rise not only triggers muscle contraction (Section 18.4) but also activates glycogen phosphorylase kinase (GPK), thereby stimulating degradation of glycogen to glucose 1-phosphate, which fuels prolonged contraction. Recall that phosphorylation by cAMP-dependent protein kinase also activates GPK. Thus this key regulatory enzyme in glycogenolysis is subject to dual regulation (Figure 20-44a).

In liver cells, hormone-induced activation of phospholipase C, which generates IP₃ and DAG, also regulates glycogen breakdown and synthesis by the two branches of the inositol-lipid signaling pathway (Figure 20-44b). DAG activates protein kinase C, which phosphorylates glycogen synthase, yielding the phosphorylated inactive form and thus inhibiting glycogen synthesis. IP₃ induces an increase in cytosolic Ca²⁺, which activates GPK as in muscle cells, leading to glycogen degradation.

The dual regulation of GPK results from its multimeric subunit structure (αβγδ)₄. The γ subunit is the catalytic protein; the regulatory α and β subunits, which are similar in structure, are phosphorylated by cAMP-dependent protein kinase; and the δ subunit is calmodulin. GPK is maximally active when Ca²⁺ ions are bound to the calmodulin subunit and at least the α subunit is phosphorylated; in fact, binding of Ca²⁺ to the calmodulin subunit may be essential to the enzymatic activity of GPK. Phosphorylation of the α and β subunits increases the affinity of the calmodulin subunit for Ca²⁺, enabling Ca²⁺ ions to bind to the enzyme at the submicromolar Ca²⁺ concentrations found in cells not stimulated by nerves. Thus increases in the cytosolic concentration of Ca²⁺ or cAMP, or both, induce incremental increases in the activity of GPK. As a result of the elevated level of cytosolic Ca²⁺ after neuron stimulation of muscle cells, GPK will be active even if it is unphosphorylated; thus glycogen can be hydrolyzed to fuel continued muscle contraction even in the absence of hormone stimulation.

Insulin Stimulation Activates MAP Kinase and Protein Kinase B

Insulin is a prime example of a hormone that can initiate multiple signaling pathways, inducing both immediate and long-term cellular responses. The immediate effects of this hormone include an increase in the rate of glucose uptake from the blood into muscle cells and adipocytes and modulation of the activity of various enzymes involved in glucose metabolism. These effects occur within minutes, do not require new protein synthesis, and occur at insulin concentrations of 10⁻⁹ to ¹⁰⁻¹⁰ M. Continued exposure to insulin produces longer-lasting effects including increased expression of liver enzymes that synthesize glycogen and of adipocyte enzymes that synthesize triacylglycerols; insulin also functions as a growth factor for many cells (e.g., fibroblasts). These effects are manifested in hours and require continuous exposure to ≈10⁻⁸ M insulin.

As noted earlier, the insulin receptor is an RTK, but unlike most RTKs it exists as a dimer in the absence of ligand. Binding of insulin can initiate two distinct signaling pathways: one that includes Ras and one that does not. Activation of Ras in the Ras-dependent insulin pathway, however, Ras differs somewhat from that for other RTKs depicted in Figure 20-23. Several lines of evidence suggest that insulin action via both the Ras-dependent and Ras-independent pathways depends on a 130-kDa polypeptide, called insulin receptor substrate 1 (IRS1). For instance, injection of antibodies to IRS1 into cultured cells blocks the normal proliferative response induced by insulin.

Insulin and Glucagon Work Together to Maintain a Stable Blood Glucose Level

During periods of stress, epinephrine plays a key role in inducing an increase in blood glucose. During normal daily living, however, the blood glucose level is under the dynamic control of insulin and glucagon. Both of these peptide hormones are produced by cells within the islets of Langerhans, cell clusters scattered throughout the pancreas. Insulin, which contains two polypeptide chains linked by disulfide bonds (see Figure 17-42b), is synthesized by the β cells in the islets; glucagon, a monomeric peptide containing 29 amino acids is produced by the α cells in the islets. Insulin acts to reduce the level of blood glucose, whereas glucagon acts to increase blood glucose. Each islet functions as an integrated unit, delivering the appropriate amounts of both hormones to the blood to meet the metabolic needs of the animal. Hormone secretion is regulated by a combination of neural and hormonal signals.

Insulin binding to receptors on muscle and adipocytes causes a rapid increase in the uptake of glucose and stimulation of glycogen synthesis via the Ras-independent pathway (see Figure 20-45). The glucagon receptor, found primarily on liver cells, is coupled to G_s protein, like the epinephrine receptor. Glucagon stimulation of liver cells activates adenylyl cyclase, leading to the cAMP-mediated cascade that leads to glycogenolysis and inhibition of glycogen synthesis (see Figure 20-35a).

The availability of glucose for cellular metabolism is regulated during periods of abundance (following a meal) or scarcity (following fasting) by the adjustment of insulin and glucagon concentrations in the blood (Figure 20-46). (Epinephrine is used only under stressful conditions.) After a meal, when blood glucose rises above its normal level of 80 – 90 mg/100 ml, the pancreatic β cells respond to the rise in glucose or amino acids by releasing insulin into the blood, which transports the hormone throughout the body. By binding to muscle and adipocyte cell-surface receptors, insulin causes glucose to be removed from the blood and stored in muscle cells as glycogen. Insulin also affects hepatocytes, primarily by inhibiting glucose synthesis from smaller molecules, such as lactate and acetate, and by enhancing glycogen synthesis from glucose. If the blood glucose level falls below ≈80 mg/100 ml, the pancreatic α cells start secreting glucagon. Glucagon binds to glucagon receptors on liver cells, triggering degradation of glycogen and the release of glucose into the blood.

Receptors for Many Peptide Hormones Are Down-Regulated by Endocytosis

The principal mechanism for down-regulating the receptors for many peptide hormones (e.g., insulin, glucagon, EGF, and PDGF) is ligand-dependent receptor-mediated endocytosis. In the absence of EGF ligand, for instance, the EGF receptor is internalized with bulk membrane flow. Binding to EGF induces a conformational change in the cytoplasmic tail of the receptor. This exposes a sorting motif that facilitates receptor recruitment into coated pits and subsequent internalization. After the receptor-hormone complex is internalized, the hormone is degraded in lysosomes — a fate similar to that of other endocytosed proteins, such as low-density lipoproteins (see Figure 17-64). Unlike the low-density lipoprotein (LDL) receptor, internalized receptors for many peptide hormones do not recycle efficiently to the cell surface.

In the presence of EGF, for instance, the average half-life of an EGF receptor on a fibroblast cell is about 30 minutes; during its lifetime, each receptor mediates the binding, internalization, and degradation of only two EGF molecules. Each time an EGF receptor is internalized with bound EGF, it has a high probability (about 50 percent) of being degraded in an endosome or lysosome. Exposure of a fibroblast cell to high levels of EGF for 1 hour induces several rounds of endocytosis, resulting in degradation of most receptor molecules. If the concentration of extracellular EGF is then reduced, the number of EGF receptors on the cell surface recovers, but only after 12 – 24 hours. Synthesis of new receptors is needed to replace those degraded by endocytosis, which is a slow process that may take more than a day.

The fewer hormone receptors present on the surface of a cell, the less sensitive the cell is to the hormone; as a consequence, a higher hormone concentration is necessary to induce the usual physiological response. A simple numerical example illustrates this important point. Suppose a cell has 10,000 insulin receptors on its surface with a K_D of 10⁻⁸ M. As noted earlier, in many cases only a fraction of the available receptors must bind ligand to induce the maximal physiological response (see Figure 20-7). If we assume only 1000 receptors must bind insulin to induce a physiological response (e.g., activation of glucose transport), we can calculate the insulin concentration [H] needed to induce this response from Equation (20-2) rewritten in the following form:

where R_T = 10,000 (the total number of insulin receptors), K_D = 10⁻⁸ M, and [RH] = 1000 (the number of insulinoccupied receptors). In this example, the necessary insulin concentrations is 1.1×10⁻⁹ M. If R_T is reduced to 2000/cell, then a ninefold higher insulin concentration (10⁻⁸ M) is required to occupy 1000 receptors and induce the physiological response. If R_T is further reduced to 1200/cell, an insulin concentration of 5 × 10⁻⁸ M, a 50-fold increase, is necessary to generate a response.

Experiments with mutant cell lines demonstrate that internalization of receptor tyrosine kinases plays an important role in regulating cellular responses to EGF and other growth factors. For instance, a mutation in the EGF receptor that makes it resistant to ligand-induced endocytosis or, in dynamin, that blocks formation of clathrin-coated endocytic vesicles substantially increases the sensitivity of cells to EGF as a mitogenic signal. Such mutant cells are prone to EGF- induced cell transformation. Interestingly, the mutant-dynamin inhibition of internalization also causes a qualitatively different pattern of phosphorylation of substrate proteins by the activated EGF receptor, as well as quantitative changes in the phosphorylation of known components in the EGF signaling pathways. Interestingly, internalized receptors can continue to signal from intracellular compartments prior to their degradation. This raises the intriguing possibility that receptor activity can be spatially controlled. Hence, internalization may modulate both the nature of RTK-transmitted signals, their magnitude, and location.

Studies with RTKs that bind PDGF suggest that PI-3 kinase plays an important role in the endocytosis and down-regulation of this class of receptors. Mutations that abolish the ability of the PDGF receptor to bind PI-3 kinase but not other enzymes (e.g., PLCγ) cause a reduction in the rate of receptor degradation. Although the mutant receptor is internalized, its sorting to the lysosome for degradation is blocked by an unknown mechanism. The observation that yeast cells expressing a mutant PI-3 kinase exhibit defective sorting of proteins to the vacuole (the yeast lysosome) raises the intriguing possibility that this enzyme plays an important role in membrane trafficking both in yeasts and mammalian cells.

Phosphorylation of Cell-Surface Receptors Modulates Their Activity

The ability of many cell-surface receptors to transmit signals is either increased or decreased by phosphorylation, leading to sensitization or desensitization of cells to various hormones. For instance, when cultured cells are exposed to epinephrine for several hours, several serine and threonine residues in the cytosolic domain of the β-adrenergic receptor become phosphorylated. The phosphorylated receptor can bind epinephrine, but ligand binding does not lead to activation of adenylyl cyclase or a rise in the cAMP level; thus the receptor is desensitized.

Four residues in the cytosolic domain of the β-adrenergic receptor are phosphorylated by a cAMP-dependent protein kinase (cAPK). The activity of this kinase is enhanced by the high cAMP level induced by epinephrine, explaining the receptor desensitization observed after prolonged exposure to epinephrine. Because the activity of all G_s protein – coupled receptors, not just the β-adrenergic receptor, is reduced by cAPK-catalyzed phosphorylation, this process is called heterologous desensitization. Other residues in the cytosolic domain of the β-adrenergic receptor are phosphorylated by a receptor-specific enzyme, called β-adrenergic receptor kinase (BARK), when the receptor is occupied by ligand. Because BARK phosphorylates only the β-adrenergic receptor, this process is called homologous desensitization. Prolonged treatment of cells with epinephrine or other agonists results in BARK-catalyzed phosphorylation of the β-adrenergic receptor and inhibition of its ability to activate G_s and adenylyl cyclase. In this case, a protein called β-arrestin has been shown to bind to the phosphorylated receptor and to sterically block interaction of the receptor with G_s, thus preventing activation of adenylyl cyclase following hormone binding. A similar protein binding to phosphorylated rho- dopsin has been identified in the visual system (Section 21.6).

Figure 20-47 illustrates the feedback loop for modulating the activity of the β-adrenergic and related G_s protein – coupled receptors. This loop permits a cell to adjust receptor sensitivity to the constant hormone level at which it is being stimulated, so as to maintain a normal physiological response. Because the phosphorylated receptors are constantly being dephosphorylated and resensitized by constitutive phosphatases, the number of phosphates per receptor molecule reflects how much ligand has been bound in the recent past (1 – 10 min). If the hormone level is increased, the resulting rise in the intracellular level of cAMP leads to phosphorylation and desensitization of more receptors, so that production of cAMP and hence the response remain relatively constant. If the hormone is removed, the receptor is completely dephosphorylated and “reset” to a high sensitivity, in which case it can respond to very low levels of hormone.

A similar feedback loop regulates the activity of the EGF receptor, which is an RTK. Phosphorylation of this receptor by protein kinase C decreases its affinity for EGF, thereby moderating the growth-stimulating effect of EGF. As noted earlier, protein kinase C is activated by DAG, which is generated by hormone stimulation of the inositol-lipid signaling pathway. We can summarize the feedback loop modulating the activity of the EGF receptor as follows: binding of EGF to its receptor → activation of PLC_γ → generation of DAG → activation of protein kinase C by DAG → phosphorylation of EGF receptor by protein kinase C → down-regulation of receptor activity.

Arrestins Have Two Roles in Regulating G Protein – Coupled Receptors

The observation that internalization of β-adrenergic receptors in response to agonists is stimulated by overexpression of BARK and β-arrestin suggested an additional function for β-arrestin in regulating cell-surface receptors. β-Arrestin binds not only to the β-adrenergic receptor but also to clathrin, which, in turn, promotes the formation of coated pits and vesicles. Immunofluorescence microscopy has shown that β-arrestin and clathrin are co-localized at the membrane in a punctate pattern with β-adrenergic receptors within 10 minutes of treating cultured cells with agonist. Endocytosis of the BARK-phosphorylated, inactive receptors is promoted by the co-localized β-arrestin, which acts as an adapter protein for clathrin polymerization and formation of clathrin-coated vesicles (see Figure 17-54). Internalized receptors become dephosphorylated in endosomes, β-arrestin dissociates, and the resensitized receptors recycle to the cell surface, similar to recycling of the LDL receptor discussed in Chapter 17 (see Figure 17-46). Regulation of other GPCRs also is thought to involve dual-acting arrestins.

SUMMARY

• Many RTKs and GPCRs activate multiple signaling pathways, and different second messengers sometimes mediate the same cellular response.

• Some activated RTKs are coupled to the Ras-MAP kinase pathway or inositol-lipid pathway in a tissuespecific manner.

• Eukaryotes possess multiple G_α, G_β, and G_γ subunits. Different G_α subunits activate various effector proteins, leading to production of specific second messengers (see Table 20-5).

• The activity of some effector proteins, including certain adenylyl cyclase isoforms, is regulated by G_βγ.

• Glycogen breakdown and synthesis is regulated by multiple second messengers induced by neural or hormonal stimulation (see Figure 20-44).

• The insulin receptor, a dimeric RTK, can act through a Ras-dependent pathway, leading to activation of MAP kinase, or through a Ras-independent pathway involving phosphoinositides, leading to activation of protein kinase B (see Figure 20-45).

• Insulin stimulation of muscle cells and adipocytes leads to activation of protein kinase B, which promotes glucose uptake and glycogen synthesis, resulting in a decrease in blood glucose.

• Binding of glucagon to its GPCR promotes glycogenolysis and an increase in blood glucose via the cAMP-triggered kinase cascade.

• Ligand binding frequently induces phosphorylation of the cytosolic domain of a cell-surface receptor, thereby modulating its activity.

• At high ligand concentration, some cell-surface receptors are internalized by endocytosis, reducing the number of receptors on the surface and making cells less sensitive to ligand.

• Many internalized RTKs are degraded in lysosomes. In this case, resensitization depends on synthesis of additional receptor molecules.

• Internalization of phosphorylated (inactive) GPCRs leads to receptor dephosphorylation, β-arrestin dissociation, and recycling of active receptors to the cell surface.