Insulin and Growth Factor Signaling
- 832 Downloads
Expression of drug-metabolizing enzymes may be altered in response to development, aging, gender, genetics, nutrition, pregnancy, disease states such as diabetes, long-term alcohol consumption, and inflammation, and by xenobiotics. Although the mechanisms by which xenobiotics regulate drug-metabolizing enzymes have been intensively studied, relatively little is known regarding the cellular mechanisms by which drug-metabolizing enzymes are regulated in response to endogenous factors such as hormones and growth factors. The first major section of the chapter defines the major insulin- and growth factor-mediated signaling pathways implicated in regulating drug-metabolizing enzyme expression, including those involving mitogen-activated protein and phosphatidyl inositol 3-kinases, small G proteins, and phosphatases. The second major section of the chapter presents a summary and evaluation of methods for determination of the role and function of signaling pathways that are involved in the regulation of drug-metabolizing enzyme gene and protein expression, including methods for determination of kinase activity and phosphorylation, the use of kinase inhibitors and dominant-negative protein kinase constructs, and the application of new RNA interference methods.
Key WordsCytochrome P450 extracellular signal-regulated kinase glutathione S-transferase G proteins green fluorescent protein insulin insulin receptor jun N-terminal kinase kinase assays kinase inhibitors mammalian target of rapamycin (mTOR) microsomal epoxide hydrolase mitogen-activated protein kinase pathway p38 mitogen-activated protein kinase p70 ribosomal protein S6 kinase p70 S6 kinase phosphatase and tensin homologue deleted on chromosome ten pleckstrin homology primary culture protein kinases protein phosphatases rat hepatocytes receptor tyrosine kinase small interfering RNA (siRNA) small temporal RNA stress-activated protein kinases sulfotransferase UDP-glucuronosyltransferase
The body is equipped with several mechanisms to ensure that xenobiotics do not accumulate. Polar molecules are often poorly absorbed into the body, but when absorbed they are readily eliminated via the kidneys. In contrast, nonpolar molecules are a special problem because of their affinity for membranes. Increasing the polarity of small, nonpolar molecules through metabolism generally promotes the excretion of the metabolites. Xenobiotic biotransformation is the principal mechanism for maintaining homeostasis during exposure of organisms to small foreign molecules and occurs predominantly in the liver, although it also occurs in nasal mucosa, intestine, kidneys, lungs, placenta, and skin.
The reactions catalyzed by xenobiotic-, or drug-metabolizing enzymes are generally divided into two groups, phase I and phase II. Phase I reactions introduce a functional group that increases hydrophilicity and are mediated by cytochrome P450 (CYP), flavin-containing monooxygenase, xanthine oxidase, prostaglandin H synthase, amine oxidase, alcohol dehydrogenase, aldehyde dehydrogenase, epoxide hydrolase, and esterase. Phase II reactions include glucuronidation, sulfation, methylation, glutathione conjugation, and amino acid conjugation. In general, these reactions, with the exception of methylation, result in a large increase in xenobiotic hydrophilicity.
It is generally recognized that the expression of drug-metabolizing enzymes may be altered in response to development, aging, gender, genetic factors, nutrition, pregnancy, and pathophysiological conditions such as diabetes, long-term alcohol consumption, inflammation, and protein-calorie malnutrition. The expression may also be altered by xenobiotics. Although the mechanisms by which xenobiotics regulate drug-metabolizing enzymes have been intensively studied, relatively less is known regarding the cellular mechanisms by which drug-metabolizing enzymes are regulated in response to endogenous factors such as hormones and growth factors. Recent findings, however, have revealed that hormones and growth factors play an important role in the regulation of drug-metabolizing enzyme expression. Furthermore, the cellular signaling pathways involved in hormone- and growth factor-mediated regulation of drug-metabolizing enzymes are currently being studied.
Effect of Diabetes, Insulin and Glucagon on Drug-Metabolizing Enzyme Expression and/or Activity
Restored by insulin
Increase(18,20,21)/ Decrease (5,19)
Because these pathophysiological states all result in altered hormone (insulin, glucagon, growth hormone) secretion, these hormones may be etiologic factors affecting the expression of hepatic drug-metabolizing enzymes. It has been reported that insulin or growth hormone administration to chemically induced or spontaneously diabetic rats restores drug-metabolizing enzyme activity and expression to control values (Table 1) (5,6,9,24, 25, 26, 27, 28, 29). Our laboratory and others have demonstrated that the activity and/or expression of hepatic drug-metabolizing enzymes such as CYP2B, CYP2E1, CYP2C11, CYP2A5, GST alpha class, GST pi class, UGT, and mEH are regulated by insulin and glucagon (Table 1) (30, 31, 32, 33, 34, 35, 36, 37, 38, 39, 40). These results indicate that changes in drug-metabolizing enzyme mRNA or protein levels observed in pathophysiological conditions may be attributed to alterations in these hormone levels. Thus, it is of interest to identify which cellular signaling pathways are involved in regulating the expression of these genes in response to hormones.
In addition to altered expression of drug-metabolizing enzymes by patho-physiologic conditions, the pattern of drug-metabolizing enzyme expression is changed during development and aging and occurs in an organ-, sex-, and species-specific manner. Such observations also suggest that a cellular or organ context regulates the expression of drug-metabolizing enzymes. Growth factors, including epidermal growth factor (EGF) and hepatocyte growth factor (HGF), have a role in regulating drug-metabolizing enzyme gene expression. HGF results in decreased CYP2C11 expression in primary cultured rat hepatocytes (41) and decreased CYP1A1/2, 2A6, 2B6, and 2E1 activities in primary cultured human hepatocytes (42). In primary cultured rat hepatocytes, addition of EGF suppresses constitutive and xenobiotic-inducible CYP expression including CYP2C11, CYP2C12, CYP1A1, CYP2B1/2 (43, 44, 45) and CYP2E1 (Woodcroft et al., unpublished data). EGF has also been reported to increase alpha and pi class GST expression (46,47).
In this chapter, an overview of the signaling pathways implicated in regulating drug-metabolizing enzyme expression in response to insulin and growth factors is described along with the methods for identifying which signaling pathways and components are involved in mediating this regulation.
2 Insulin-and Growth Factor-Mediated Signaling Pathways
The numerous and varied biological functions of insulin and growth factors are mediated by their corresponding cell surface receptors. The insulin receptor and growth factor receptors belong to the large family of receptor tyrosine kinase (RTK) cell surface receptors possessing intrinsic tyrosine kinase activity. After binding of their corresponding agonist, these receptors undergo autophosphorylation of tyrosine residues in the cytoplasmic domain and initiate a complex series of intracellular signaling cascades that ultimately result in diverse cellular responses.
Insulin and growth factors stimulate the recruitment of a family of lipid kinases known as class I PI3Ks to the plasma membrane. There, PI3Ks phosphorylate the glycerophospholipid phosphatidylinositol (PI) 4,5-bisphosphate at the D-3 position of the inositol ring, converting it to PI 3,4,5-triphosphate (PI[3,4,5]P3). Recent evidence indicates that serine/threonine protein kinase Akt/PKB (protein kinase B), atypical protein kinase C (PKC), and p70 S6 kinase mediate many of the downstream events controlled by PI3K.
2.1 Insulin Receptor
The insulin receptor is a transmembrane heterotetramer consisting of two α- (extracellular; 135 kDa) and two β- (transmembrane; 95 kDa) subunits linked by disulfide bonds (49). During transport to the cell surface, a single high-molecular-weight proreceptor is proteolytically cleaved at a tetrabasic amino acid sequence (arginine-lysine-arginine-arginine) located at the junction of the α- and β-subunits, and oligosaccharide chains are added at specific sites of glycosylation. Interactions between the two α-subunits, and between the a and β subunit, are stabilized by disulfide bridges (50,51).
Unoccupied α-subunits on the cell membrane surface inhibit the intrinsic tyrosine kinase activity of the β-subunit and may be viewed as a regulatory subunit of the catalytic intracellular subunit (52). The β-subunit is composed of a short extracellular domain, a transmembrane domain, and a cytoplasmic domain that possesses intrinsic tyrosine kinase activity. The cytoplasmic domain contains the ATP binding site and autophosphorylation sites. Binding of insulin to the α-subunits of the receptor induces conformational changes leading to activation of the RTK activity, resulting in transphosphorylation of the β-subunits and endocytic internalization of the receptor via clathrin-coated vesicles (51). Some of the tyrosine-phosphorylated residues of the β-subunits of the receptor present binding sites for the subsequent recruitment of signaling molecules. The insulin receptor uses a family of soluble adaptors or scaffolding molecules, such as insulin receptor substrates (IRSs 1–4) and Shc molecules, to initiate its signaling cascade through other effectors.
Whereas the IRSs lack intrinsic catalytic activity, they have pleckstrin homology (PH) and phosphotyrosine binding (PTB) domains, and multiple phosphorylation motifs. The PH domains are globular protein domains of about 100–120 amino acids found in more than 150 proteins to date. PH domains are primarily lipid-binding modules, although they are also involved in mediating protein–protein interactions. The PTB domain of IRS binds to phosphorylated NPXP motifs in the insulin receptor and are subsequently phosphorylated on multiple tyrosine residues by the activated insulin receptor kinase. Following phosphorylation, IRS attracts and binds additional effector molecules to the vicinity of the receptor, thereby serving to increase the diversity of the signaling pathways initiated by the insulin receptor (53,54). The primary effector that binds to IRSs in response to insulin receptor activation is the lipid kinase PI3K that produces PI(3,4,5)P3 and subsequently activates Akt, PKC, and p70 S6 kinase (Fig. 3).
The adaptor protein Shc exists in p46, p52, and p66 isoforms and possesses Src homology-2 (SH2) and PTB domains and three tyrosine-phosphorylation sites. In response to extracellular signals, Shc is phosphorylated on tyrosine residues and binds the growth factor receptor binding protein 2 (Grb2), which is constitutively associated with the guanine nucleotide exchange factor Son of Sevenless (SOS) (55). Recruitment of the Grb2–SOS complex to the vicinity of Shc induces exchange of GDP to GTP on the membrane-bound GTPase Ras, thereby activating Ras. Activated Ras binds Raf and activates the serine/threonine Raf/MAPK kinase (MKK)/MAPK signaling pathway (Fig. 3) (56).
2.2 Growth Factor Receptor
The growth factor receptors are also members of a large family of cell surface receptors, including the EGF receptor; ErbB2, 3, and 4; and c-MET, which exhibit intrinsic protein tyrosine kinase activity. The EGF receptor is synthesized from a 1210-residue polypeptide precursor; after cleavage of the N-terminal sequence and glycosylation, the 1186-residue protein is inserted into the cell membrane (57). The glycosylated extracellular domain of the EGF receptor contains conserved cysteine-rich clusters that comprise the ligand-binding domain. Within the intracellular domain, the juxtamembrane region is required for feedback attenuation by PKC, followed by the tyrosine kinase domain and C-terminal regulatory tail. The C-terminal tail contains the tyrosine autophosphorylation sites and motifs for internalization and degradation of the receptor. Binding of EGF to the receptor results in receptor dimerization and autophosphorylation. The resulting conformational change creates specific docking sites for recruitment and activation of additional signaling proteins that contain SH2 and PTB domains, including Shc and Grb2. In addition, the EGF receptor can be phosphorylated by other kinases, such as PKC and Src kinase, which regulate the distribution and kinase activity of the EGF receptor (58,59). Following ligand binding and activation, EGF receptor dimers are recruited into clathrin-coated pits, which initiates a rapid endocytosis and degradation of the EGF receptor.
2.3 Pl3K/Akt/p70 S6 Kinase/Atypical PKCs
There are four major classes of PI3Ks, designated class I–IV; class I is also subdivided into Ia and Ib subsets. Class IV PI3Ks are not known to possess lipid kinase activity, but are serine/threonine kinases. The different classes of PI3Ks catalyze phosphorylation of the 3′-OH position of phosphatidyl myoinositol lipids, generating different 3′-phosphorylated lipid products that act as second messengers. Class Ia PI3Ks are primarily responsible for production of 3′-OH phosphoinositides in response to insulin and growth factors (60). Class Ia enzymes are dimers composed of a 110-kDa catalytic subunit that is associated nonconvalently to an 85- or 55-kDa regulatory subunit (Fig. 3). The catalytic subunit in subclass Ia is subdivided into p110 α-, β-, and δ. The regulatory subunit maintains the catalytic subunit in a low-activity state in quiescent cells and mediates its activation through interactions between SH2 domains of the regulatory subunit and phosphotyrosine residues of activated growth factor receptors or adaptor proteins, such as the IRSs (61). The single class Ib PI3K is the p110 γ catalytic subunit complexed with a p101 regulatory protein and mainly activated by heterotrimeric G protein-based signaling pathways. Direct binding of p110 γ to activated Ras plays an important role in the stimulation of PI3K in response to growth factor (62), but the physiological significance of this interaction in insulin-mediated PI3K signaling is not entirely clear.
Following the recruitment of PI3K to the plasma membrane, the lipid kinase phosphorylates the 3′-OH position of the inositol ring to generate PI(3,4,5)P3, PI(3,4)P2, and PI(3)P. The preferred substrate of class I PI3Ks appears to be PI(4,5)P2. These events occur within the first minute of insulin binding to its receptor and resulting lipid products then interact with a number of signaling proteins with PH domains, resulting in their membrane targeting and/or modulation of their enzyme activity.
The rapid increase in PI(3,4,5)P3 concentration in response to insulin activates several protein kinases, such as phosphatidylinositide-dependent kinase 1 (PDK1), Akt, PKC isoforms, and p70 S6 kinase (63, 64, 65). Among the PI(3,4,5)P3-dependent kinases, Akt has received much attention. Akt/PKB was identified as a protein kinase with a high degree of homology to PKA and PKC, and is the cellular homologue of the viral oncoprotein v-Akt. Akt is a 57-kDa serine/threonine kinase with a PH domain and the three known isoforms of Akt (Akt1, 2, 3) are widely expressed (66).
Akt exists in the cytoplasm of unstimulated cells in a low-activity conformation. The activation of Akt1 by insulin and growth factors is accompanied by its phosphorylation on threonine-308 in the kinase domain (T-loop) and serine-473 in the C-terminal regulatory domain (hydrophobic motif). Activation of Akt and phosphorylation of both these residues are abolished by pretreatment of cells with PI3K inhibitors such as wortmannin and LY294002 (67). After activation of PI3K, association of PI(3,4,5)P3 at the membrane brings Akt and PDK1 into proximity through their PH domains and facilitates phosphorylation of Akt at threonine-308 by PDK1 (65). The mechanism mediating serine-473 phosphorylation remains to be clarified.
p70 S6 kinase catalyzes the phosphorylation of the S6 protein, a component of the 40S subunit of eukaryotic ribosomes, and thus plays a role in protein synthesis (68,69). p70 S6 kinase participates in the translational control of mRNA transcripts that contain a polypyrimidine tract at their transcriptional start site. Although these transcripts represent only 100–200 genes, most of these transcripts encode components of the translational apparatus. The initial step in p70 S6 kinase activation appears to involve a phosphorylation-induced conformational change in the C-terminal domain, revealing additional phosphorylation sites. Subsequently, phosphorylation of the newly exposed sites (threonine 229, 389, and serine 371) occurs, which is dependent on both PI3K and the mammalian target of rapamycin (mTOR), based on wortmannin and rapamycin sensitivity, respectively.
Although expression of a constitutively membrane-anchored and active Akt variant induces the activation of p70 S6 kinase (70), Akt does not appear to represent the immediate upstream effector of p70 S6 kinase. Conus et al. (71) suggested that p70 S6 kinase activation could be achieved independent of Akt. Dufner et al. (72) demonstrated that a constitutively active wortmannin-resistant form of Akt was sufficient to induce glycogen synthase kinase-3 and eIF4E-binding protein 1 phosphorylation, but not phosphorylation and activation of p70 S6 kinase. The data suggest that p70 S6 kinase activation by membrane-targeted forms of Akt may be an artifact of membrane localization and that Akt resides on a parallel PI3K-dependent signaling pathway to that described for p70 S6 kinase.
Recent findings indicate that atypical PKC isoforms (ζ, rat) and (λ, mouse) serve as downstream effectors for PI3K (73). Increased activity of PKCζ/λ results from PDK1-dependent phosphorylation of the catalytic domain, via threonine-410 in rat PKCζ and threonine-411 in mouse PKCλ, followed by autophosphorylation of threonine-560 in rat PKCζ and threonine-563 in mouse PKCλ (64,74). PI(3,4,5)P3 may interact with the N-terminal lipid-binding domain of PKCζ to facilitate the interaction of threonine-410 with the catalytic site of PDK1 (74). PI(3,4,5)P3 also stimulates autophosphorylation of PKCζ and relieves the autoinhibition exerted by the N-terminal pseudosubstrate sequence on the C-terminal catalytic domain of PKCζ (74,75). Insulin-stimulated glucose transport and protein synthesis are dependent on PI3K/PKCζ activity (73,76). The latter is consistent with the observation that dominant-negative PKCζ antagonizes activation of p70 S6 kinase (77). However, it is not known whether PKCζ can directly phosphorylate p70 S6 kinase or which residue(s) is/are involved.
Mammalian cells contain three different Ras genes that give rise to four Ras small GTPases—H-Ras, N-Ras, KA-Ras and KB-Ras—which are key regulators of signal transduction pathways controlling cell proliferation, differentiation, survival, and apoptosis (78,79). In response to a great variety of extracellular stimuli, including hormones, growth factors, cell–extracellular matrix interactions, and oxidative stress, Ras proteins are activated through the GDP/GTP nucleotide exchange factor SOS, which induces the exchange of GDP for GTP, and thereby converts Ras to its active form. Ras cycles between the inactive GDP-bound and active GTP-bound states through the controlled activity of GTP nucleotide exchange factors and GTPase-activating proteins. After activation of insulin and growth factor receptors through agonist binding, the link between RTKs and Ras is provided by the GTP exchange factor SOS that exists in a complex with the adaptor protein Grb2 in the cytosol. Phosphorylated tyrosine residues in insulin and growth factor receptors are docking sites for Grb2. In addition, the interaction between Grb2/SOS and the receptors can be mediated by the adaptor protein Shc, which becomes tyrosine phosphorylated when recruited to the cytoplasmic domains of the activated receptors. This process brings SOS to the plasma membrane in close proximity to Ras, where it can promote GDP/GTP exchange. GTP-bound activated Ras recruits and activates three main classes of effector proteins, Raf kinases, PI3K, and RalGDS (62).
Three genes encode for the Raf family of serine/threonine kinases found in mammalian cells: A-Raf, B-Raf, and Raf-1 (c-Raf). The large majority of studies regarding the role of Raf in ERK activation have been performed with Raf-1. In resting cells, Raf-1 is located in the cytoplasm and is stabilized by a 14-3-3 scaffold protein dimer binding to phosphorylated serines 259 and 621, which are phosphorylated in resting cells (80). The binding of Raf to Ras and translocation to the plasma membrane can displace 14-3-3 from phosphoserine 259, which makes it accessible to dephosphorylation and activation by protein phosphatase 2A (PP2A) (81), although the role of dephosphorylation of serine-259 in Raf-1 activation was recently challenged (82).
The activation of Raf-1 is required for the subsequent multistep events to occur at the plasma membrane following the relief from autoinhibition. Agonists such as insulin and growth factors stimulate the phosphorylation of several residues, including serine-338, tyrosine-341, tyrosine-491, and serine-494 (83). Phosphorylation at serine-338 and tyrosine-341 is a critical step for Raf activation (83) and serine-338 phosphorylation appears to be a good qualitative indicator of Raf-1 activation.
MEK1 (MKK1) and MEK2 (MKK2) contain a proline-rich sequence necessary for the interaction of MEK with Raf-1 (84). MEKs are phosphorylated by Raf-1 on two serine residues (serine-217 and -221), which are necessary for full activation. MEK1 and MEK2 activate ERK1 (p44 MAPK) and ERK2 (p42 MAPK) via phosphorylation of a threonine–glutamate–tyrosine motif in the activation loop. ERK is a proline-directed serine/threonine kinase at the end of this pathway with more than 50 identified substrates, including transcription factors, MAPK-activated protein kinase-2, and the p90 ribosomal S6 kinase (85). ERK activation has traditionally been associated primarily with cell proliferation.
The stress-activated protein kinases (SAPKs) such as JNK, p38 MAPK, and ERK5 are slightly activated by insulin or growth factors but vigorously activated by stress signals (UV irradiation, heat or cold shock, osmotic stress, mechanical shear stress, oxidative stress), cytokines, and G protein–coupled receptor agonists (86). The SAPKs are involved in the regulation of growth arrest, apoptosis, and proliferation. The SAPKs are activated through a similar kinase cascade as ERK, although some different mechanisms have been noted. MKK4/SEK1 and MKK7 phosphorylate and activate JNK, whereas p38 MAPK is activated by MKK3 and MKK6. At the level of the MKKK, many kinases activating either or both JNK and p38 MAPK have been identified by overex-pression or dominant-negative experiments (87).
2.5 Crosstalk Between PI3K and MAPKs
A number of studies have suggested that the MAPK and the PI3K pathways cross-talk on several levels (88, 89, 90, 91, 92, 93, 94, 95, 96). But the interrelationship between MAPK and PI3K signaling pathways has been controversial. This may reflect differences associated with cell context and activators of the signaling cascades.
Studies of the relevance of PI3K signaling for the activation of ERK are inconclusive, with both increased and decreased ERK activation being reported. Although ERK1/2 activation may occur independently of PI3K (97, 98, 99), inhibitors of PI3K have been reported to inhibit insulin- and growth factor-induced increases in ERK1/2 activity in a number of cell types such as the rat skeletal muscle cell line L6, rat adipocytes, and hepatic stellate cells (100, 101, 102). In contrast, Akt phosphorylates serine-259, located in the regulatory domain of Raf-1, resulting in the inactivation of Raf-1 (103,104). Moelling et al. (92) showed that the PI3K/Akt pathway inhibited the Ras/Raf-1/MEK/ERK pathway at the level of Raf-1 and Akt. Thus, the PI3K-dependent signaling pathway can either stimulate or inhibit ERK activation and this likely depends on the cell context and the type of stimuli, as well as the concentration and period of treatment (88,92).
JNK activity is elevated in obesity and an absence of JNK1 results in decreased adiposity, significantly improved insulin sensitivity and enhanced insulin receptor signaling capacity in two different models of mouse obesity (105). Lee et al. (96) showed that insulin-stimulated JNK associated with IRS1 and phosphorylated IRS1 at serine-307 in mouse embryo fibroblasts and 3T3-L1 adipocytes, and that this interaction inhibited insulin signaling. These results suggest that prolonged activation of JNK inhibits IRS-associated PI3K activity and can be a crucial mediator of insulin resistance.
The phosphorylation of tyrosine residues in proteins by kinases plays a key role in the regulation of cell signaling and gene expression. The level of tyrosine phosphorylation in receptors and their downstream substrates is dynamically and precisely regulated by two types of enzymes, protein tyrosine kinases, which catalyze the phosphorylation of tyrosine residues, and protein tyrosine phosphatases (PTPs), which dephosphorylate the phosphotyrosine residues (106,107). Disregulated PTP activity may lead to aberrant tyrosine phosphorylation, which may contribute to disease, including cancer and diabetes (108,109). PTPs can be divided into tyrosine-specific and dual-specific subfamilies, based on their substrate specificity. The dual-specificity subfamily recognizes phosphotyrosine, phosphothreonine and phosphoserine residues. Tyrosine-specific PTPs can be further classified as intracellular and receptor-like PTPs. Intracellular PTPs possess a single conserved catalytic domain, and the N- or C-terminus that appears to play a regulatory or targeting role. PTP-1B and SHP-2, an SH2-containing PTP-2, belong to this subfamily and are key regulators that control the intracellular phosphotyrosine level. Receptor-like PTPs, exemplified by CD45 and PTPα, contain one or two cytoplasmic catalytic domains, a single transmembrane region and an extracellular domain. The extracellular domains have structures found in cell-adhesion molecules, suggesting a role for this subfamily of PTPs in cell–cell and/or cell–extracellular matrix interactions. Dual-specificity PTPs include the MAPK phosphatases (MKPs) and cell cycle regulator Cdc25 phosphatases.
PTP-1B, the first mammalian PTP to be purified to homogeneity (110), is widely expressed and localized predominantly to the endoplasmic reticulum through a cleavable proline-rich C-terminal segment (111). Cleavage of the C-terminal 35 amino acids of PTP-1B appears to release this phosphatase from the endoplasmic reticulum and increase its specific activity (112). PTP-1B deficiency in mice results in enhanced insulin sensitivity, as demonstrated by an increased insulin-stimulated phosphorylation of the insulin receptor in muscle and liver, an improved glucose clearance in glucose and insulin tolerance tests, and a significant reduction in fed glucose levels (113). This study suggests that PTP-1B is a negative regulator of the insulin-stimulated signal transduction pathway by dephosphorylating the phosphotyrosine residues of the insulin receptor kinase (114). The association between insulin receptor and PTP-1B has been demonstrated using the substrate trapping method, immunoprecipitation, and immunoblot analysis (115,116). The mechanism(s) for regulation of PTP-1B activity is unclear. Recently, it has been reported that insulin-stimulated intracellular hydrogen peroxide production may reversibly oxidize PTP-1B, resulting in inhibition of PTP-1B and enhancement of the insulin signaling cascade (117,118). Several groups have reported that phosphorylation of PTP-1B by the insulin receptor and other protein kinases affects PTP-1B enzyme activity (119, 120, 121). In addition to the insulin receptor, PTP-1B has other targets such as EGF receptor and Src kinase (122,123).
The SHP-2 phosphatase contains two tandem SH2 domains that mediate the binding of SHP-2 to phosphorylated tyrosine residues on other molecules. In the resting state, SHP-2 activity is repressed, but its mechanism of activation is unclear. SHP-2 plays a positive and/or negative role in transducing signals relayed from RTKs. For example, introduction of the catalytically inert SHP-2 markedly inhibited activation of ERK in response to EGF and insulin stimulation (124,125). Chen et al. (126) also reported that overexpression of dominant-negative SHP-2 resulted in a modest impairment of insulin-stimulated glucose transporter 4 translocation, suggesting SHP-2 may play a minor role as a positive modulator of the metabolic effects of insulin. In contrast, Ouwens et al. (127) reported that expression of SHP-2 in cells resulted in a negative regulation of IRS-1 phosphorylation, PI3K activation, and stimulation of glycogen synthesis in response to insulin. Thus, it remains to be seen whether SHP-2 plays a major physiological role in insulin signaling.
The dual specificity MKPs are able to dephosphorylate both phosphotyrosine and phosphothreonine residues in the activation loop of MAPKs and inactivate them. In mammalian cells, at least 10 MKPs have been identified and individual MKPs display differential selectivity toward different MAPK family members and MKPs are localized to the nucleus or cytoplasm (128, 129, 130).
In general, serine/threonine protein phosphatases can be classified into the phosphoprotein phosphatase (PPP) and Mg2+-dependent protein phosphatase (PPM) gene families on the basis of similarity in the primary amino acid sequence between the different catalytic subunits (131). The PPP family includes the most abundant protein phosphatases—PP1, PP2A, and PP2B (calcineurin)—as well as more recently cloned enzymes such as PP4 (also known as PPX), PP5, PP6, and PP7. Five PPC2 isoforms, together with the pyruvate dehydro-genase phosphatase, constitute the gene family PPM. The catalytic subunit of PP1 has four mammalian isoforms and this phosphatase is inhibited by the cell-permeable toxins okadaic acid and calyculin A and the membrane-impermeable agent microcystin (132,133). PP2A is spontaneously active and inhibited by the inhibitors of PP1 whereas PP2B, a Ca2+-dependent protein, is not inhibited by these inhibitors (134). In response to insulin, PP1 is activated and catalyzes the insulin-mediated dephosphorylation of metabolic enzymes. The glycogen-associated form of PP1 dephosphorylates both glycogen synthase (resulting in enzyme activation) and glycogen phosphorylase (resulting in inactivation) to provide insulin-mediated coordination of glycogen metabolism (135). PP2A contributes to the dephosphorylation and regulation of MAPKs (81,136).
As discussed in Subheading 2.3., many different phosphorylated derivatives of PI play diverse roles in cellular signaling. The versatility of these molecules as cellular signals results from PH domain specificity in recognizing particular configurations of the inositol phosphate headgroup. Much attention has focused on PI(3,4,5)P3 as an intracellular second messenger produced rapidly via the action of PI3K in response to many divergent cellular stimuli. Phosphatase and tensin homologue deleted on chromosone 10 (PTEN) was discovered in 1997 as a new tumor suppressor and serves as an unusual phosphatase whose primary target is PI(3,4,5)P3. The 3-phosphorylated inositol lipids, PI(3,4,5)P3 and PI(3,4)P2, are the most efficient substrates for PTEN, which removes phosphate from the D-3 position of the inositol ring.
Evidence suggests that protein stability, localization, and transcription of the PTEN gene regulate the function of PTEN. The regulation of stability and localization appears to be achieved through the C-terminal tail of PTEN via phosphorylation of multiple serine and threonine residues. Recently, it has been shown that PI3K activation stimulates PTEN phosphorylation, suggesting that one of the kinases activated by PI3K is likely to be involved in PTEN phosphorylation (137).
3 Methods for Determination of Signalting Pathways Involved in the Regulation of Drug Metabolizing Enzyme Gene and Protein Expression
3.1 Methods for Examination of Protein Kinase Activity and Phosphorylation
Phosphorylation plays an essential role in the regulation of most protein kinases. Phosphorylation of specific residues is a major determinant of protein kinase activity. For example, the activation of Akt by insulin or growth factors is accompanied by phosphorylation on threonine-308 in the kinase domain and serine-473 in the C-terminal regulatory domain. Activation of Akt and phosphorylation of both these residues are abolished by treatment with the PI3K inhibitors, wortmannin or LY294002, prior to stimulation with an agonist such as insulin.
3.1.1 Immunoblot Analysis
Since first reported by Ross et al. (138), antibodies reactive with phosphoresidues (e.g., phosphotyrosine, phosphoserine, and phosphothreonine) have become invaluable tools for isolating phosphorylated proteins and examining phosphorylation states. Phospho-specific antibodies for many protein kinases have been developed and these antibodies can be used for immunoblot analysis, immunoprecipitation, immunocytochemistry, and flow cytometry. In general, immunoblot analysis, coimmunoprecipitation, and kinase activity assays are the most frequently used methods for examination of protein kinase activation. Phospho-specific antibodies against a number of kinases and receptors are commercially available, and are used in standard immunoblotting procedures. If phospho-specific antibodies are not available, the kinase or receptor can be immunoprecipitated followed by immunoblotting with antiphosphotyrosine/serine/threonine antibodies.
Protein kinase activation usually occurs within a few minutes of agonist binding to a receptor (e.g., insulin and growth factors), suggesting that the phosphorylation state is dynamically regulated. Thus, inhibition of phosphatase activity is very important during preparation of cell lysates. Treatment of cells with phosphatase inhibitors may result in activation of kinases. Generally, cell lysis buffer contains phosphatase inhibitors such as sodium ortho-vanadate, sodium fluoride, ethylenebis(oxyethylenenitrilo)tetraacetic acid, and okadaic acid, to prevent dephosphorylation of protein kinases and other phosphoproteins. In immunoblot analysis, Laemmli sample buffer that contains sodium dodecyl sulfate and dithiothreitol can be used directly for making cell lysates.
3.1.2 Kinase Activity Assays
For protein kinase assays, phospho-specific antibodies to protein kinases have been used for selectively immunoprecipitating activated protein kinases from cell lysates. This method depends on the availability of a specific immunopre-cipitating antibody that does not interfere with the kinase activity, but many of these are commercially available. Protein kinase activity can be assayed by incorporation of phosphate from ATP into a synthetic peptide substrate based on the sequences of the phosphorylation sites on the target substrate protein. Many protein kinases phosphorylate the short peptide substrate with kinetic parameters similar to those of the native target proteins. In some cases, however, protein kinases that recognize or require an aspect of 3-D structure for their target, in addition to the primary sequence, will phosphorylate synthetic peptides poorly or not at all. Kinases that fall into this class must be assayed using the native protein target as substrate, or at least an expressed domain of the substrate that contains the requisite recognition features. Most protein kinase assays using synthetic peptides use radioactive ATP, resulting in a radiolabeled phosphopeptide that can be quantified by scintillation counting. Recently, nonradioactive kinase assays employing phospho-specific antibodies to the substrate protein have been developed and allow detection and quantification of kinase activity following immunoprecipitation of an active kinase.
3.2 Chemical Inhibitors of Protein Kinases
The inhibitors PD98059 and U0126 bind to the inactive form of MEK, preventing its activation by Raf-1 and other upstream activators (141,142). These inhibitors do not compete with ATP and do not inhibit the phosphorylation of MEK, and thus are likely to have a distinct binding site on MEK. Quantitative evaluation of the steady-state kinetics of MEK inhibition by these compounds shows that U0126 has higher affinity than PD98059 (142). In a comparison of multiple kinase inhibitors, the MEK1 and MEK2 inhibitors appeared to be the most specific kinase inhibitors tested (143). But both of these inhibitors have recently been shown to inhibit activation of the ERK5 pathway through direct effects on MEK5 (144). It is recommended that PD98059 or U0126 be added to cells at a concentration of 50–100 μM or 5–25 μM, respectively.
SB203580 and SB202190, a class of pyridinyl imidazoles, are relatively specific inhibitors of p38 MAPK α and β, but not p38 MAPK γ and δ, at a concentration of 10 μM (145,146). However, these inhibitors were reported to inhibit the activation of PDK1 and its downstream effectors, including Akt and p70 S6 kinase (147,148), although PDK1 activity remained unaffected by in vitro incubation with SB203580 or SB202190 (143). We have found that in primary cultured rat hepatocytes, these p38 MAPK inhibitors failed to affect insulin-mediated Akt phosphorylation (40). These compounds bind the ATP-binding cleft of the low-activity p38 MAPK, which binds ATP poorly (149). As a consequence of binding the unphosphorylated form, these inhibitors appear to interfere with the activation of p38 MAPK. Generally, SB203580 and SB202190 completely inhibit p38 MAPK at a concentration of 10 μM.
SP600125, an anthrapyrazolone inhibitor of JNK1, JNK2 and JNK3, has been reported to inhibit JNKs through a reversible ATP-competition (150). A number of studies have reported that the compound prevents the expression of several anti-inflammatory genes in cell-based assays and the activation of AP1 in synoviocytes (150,151). The inhibitor is starting to be used more widely as a JNK inhibitor. However, Bain et al. (152) recently reported that SP600125 was a relatively weak inhibitor of JNK isoforms and also inhibited other protein kinases with similar or greater potency. Care must be used, therefore, when employing this inhibitor and in the interpretation of resulting data. For inhibition of JNKs, SP600126 has been used at a concentration of 10–25 μM.
SU6656 (153) and the related pyrazolopyrimidine, PP1 (154), were developed as inhibitors of the Src family of enzymes. PP1 was originally described as a selective, ATP-competitive inhibitor of Src family kinases and has been widely used to investigate the contribution of Src kinases to a number of biological functions (154). It is recommended that SU6656 be added to cells at a concentration of 1–5 μM.
Rapamycin, a potent immunosuppressant, rapidly inactivates p70 S6 kinase and prevents the activation of this kinase by all known agonists (155,156). Rapamycin binds to the immunophilin FK506 binding protein 12, and the resultant complex interacts with the protein kinase mTOR/FKBP 12-rapamycin-associated protein, thereby inhibiting it. This leads to the dephosphorylation and inactivation of p70 S6 kinase (157). Generally, rapamycin completely inhibits p70 S6 kinase at a concentration of 100 nM.
GF109203X (bisindolylmaleimide I; Gö6850) and Ro-31-8220 are bisin-dolylmaleimides that differ from each other in two functional groups and are analogues of staurosporine (158,159). These inhibitors, which compete for the ATP binding site on PKC, have approx 100-fold selectivity for PKC over PKA (160). They are both potent inhibitors of the α, β, and γ isoforms of PKC with IC50 values in the nanomolar range in vitro. However, micromolar concentrations of GF109203X are required to inhibit atypical PKCs (161). These classes of compounds may also have the ability to selectivity inhibit PKC isoforms. Gö6976, another staurosporine-related compound, inhibits α- and β1-PKCs when utilized at nanomolar concentrations, but fails to inhibit δ-, ε-, and ζ-PKC isoforms (161). It is recommended that GF109203X be added to cells at a concentration of 1–10 μM.
Wortmannin and LY294002 are cell-permeable inhibitors of PI3K (162,163). Wortmannin, an irreversible inhibitor, alkylates a lysine residue at the putative ATP binding site of p110α of PI3K and LY294002 is a pure competitive inhibitor of ATP. It is recommended that wortmannin or LY294002 be added to cells at a concentration of 100–500 nM or 10–20 μM, respectively. At higher concentrations, wortmannin inhibits a number of other kinases, including the class 2 PI3K (164,165). If a longer incubation time is required, LY294002 is the inhibitor of choice rather than wortmannin, because of its higher stability in aqueous solution.
For longer treatment durations in highly metabolically competent cells, such as primary hepatocytes, the concentrations of protein kinase inhibitor recommended earlier may not be sufficient to inhibit each target protein kinase activity. Thus, higher concentrations of most of these inhibitors are often required to offset metabolism of the inhibitor, and care must therefore be exercised in the interpretation of these data.
3.3 Dominant-Negative Protein Kinase Constructs
The activity of a protein kinase can be interfered with by expression of a dominant-negative mutant. The generation of dominant-negative mutants involves the design of an inactive form of the protein that can sequester interacting proteins. Some knowledge of the mechanism of regulation or function of the protein of interest is helpful when designing these molecules. In general, the activity of protein kinases requires the phosphorylation of specific residues for activation and the binding of ATP to a conserved protein motif for phosphorylation of effector proteins. Thus, the point mutation of the phosphorylation site or ATP-binding region can produce an inactivated kinase or kinase-dead mutant, respectively (166,167). Overexpression of an inactive form of the kinase may act as a dominant-negative by sequestering interacting proteins or cofactors and thus inhibiting the activity of the endogenous wild-type kinase. Many protein kinases are inactive in resting cells and this basal inhibition is achieved by interaction with a regulatory protein or an inhibitory domain within the same polypeptide. Thus, in some cases, overexpression of a regulatory protein or an inhibitory domain can reduce or inhibit the ability of the pathway to stimulate the endogenous protein. Similarly, overexpression of a pseudosubstrate domain that can bind the enzyme but cannot be converted to product can often result in inhibition of signaling, as it will compete with the endogenous substrate (168).
DNA constructs encoding inactive kinase mutants must be transported through the cell membrane and into the nucleus, to inhibit signaling pathways through their expression. There are several well-established techniques that allow transient transfection of recombinant DNA into cells in culture. These methods generally involve the permeabilization of cell membranes by chemical or electrical means, or the use of viral constructs that can recognize specific receptors on the cell surface, resulting in cellular uptake. A variety of viral systems, including adenoviruses and retroviruses, have become available for transporting recombinant DNA into cells (169). The DNA can either be incorporated into the viral genome or be chemically linked to the exterior of the virion. After transfection of adenovirus into a mammalian cell, viral production may be monitored with green fluorescent protein (GFP), which is encoded by a gene incorporated into the viral backbone (170). The most common methodologies have been reviewed in detail elsewhere (170,171).
In 1998, Fire and Mello described a new technology that was based on the silencing of specific genes by double-stranded RNA (dsRNA) and termed RNA interference (RNAi) (172). RNAi consists of the presentation of a “triggering” dsRNA that is subsequently processed into 21–25 base-pair small interfering RNAs (siRNAs) through the action of the Dicer enzyme (RNase III endonu-clease) (173, 174, 175). siRNAs with 2-nucleotide 3′-end overhangs are then incorporated into a multisubunit RNA-induced silencing complex, which targets their complementary RNA transcript for enzymatic degradation (176). The siRNA-induced degradation of mRNA is highly sequence-specific, to the extent that even a one- or two-nucleotide difference in the targeting recognition sequence hampers RNAi function.
In contrast to siRNAs, small temporal RNA (stRNA) molecules, which represent a large group of small transcripts called micro-RNAs, mediate gene suppression by inhibiting translation of target mRNA (177,178). In common with siRNAs, Dicer is also involved in the processing of the 21- to 23-nucleotide stRNAs from approx 70-nucleotide stable hairpin precursors (179). But stRNAs are stem-loops that are processed into an imperfect complementary dsRNA that inhibit protein translation of an imperfectly matched target sequence, which is almost invariably located at the 3′-untranslated region of the target mRNA (180).
In mammalian somatic cells, dsRNAs longer than 30 nucleotides activate an antiviral defense mechanism that includes the production of interferon and activation of dsRNA-dependent protein kinase, resulting in inhibition of protein synthesis initiation and stimulation of apoptosis (181,182). One mechanism for dealing with these nonspecific dsRNA responses is to create dsRNA triggers of fewer than 30 base pairs in length. Both siRNA and stRNA are long enough to induce sequence-specific suppression, but short enough to evade the host defense response. Although the use of siRNAs to silence genes in vertebrate cells was reported only a few years ago, the emerging literature indicates that most vertebrate genes can be studied with this technology.
Several laboratories demonstrated that synthesized dsRNAs induced sequence-specific gene silencing when transiently transfected into mammalian cells (183,184). Factors that could ultimately limit the usefulness of siRNAs include a relatively short and transient period of activity. The longevity of silencing is dependent on abundance of mRNA and protein, stability of protein, the half-life of the silencing complex, and cell division rate. Generally the siRNA directs rapid reduction in mRNA levels that is readily observed in 18 h or less and siRNA-mediated RNAi lasts for 3–5 d for most cell lines (185).
Recently a number of studies have reported the success of using RNA polymerase III promoters, such as U6 or H1, to direct in vivo synthesis of functional siRNAs (186, 187, 188, 189, 190, 191). These siRNAs have been expressed in two ways. In the first case, hairpin constructs are expressed from a single RNA polymerase III promoter. The resulting RNAs are predicted to form hairpins containing 19- to 29-nucleotide stems that match target sequences precisely, three- or nine-nucleotide loops and 3′ overhangs of four or fewer uridines. It is believed that these hairpin RNAs are processed by Dicer to active siRNAs in vivo (192). In the second case, coding and noncoding strands of a potential siRNA are driven from separate promoters and the expressed transcripts anneal in the cell nucleus. The hairpin siRNA strategy appears to inhibit gene expression more efficiently than the duplex siRNAs expressed from two separate plasmids (192). The use of a plasmid-based RNA polymerase III promoter system to intracellularly produce siRNAs could allow for a longer period of expression as compared with exogenously added siRNAs.
An alternative approach to prolong siRNA-mediated inhibition of gene expression is the introduction of modified nucleotides into chemically synthesized RNA. Amarzguioui et al. (193) reported that siRNA generally tolerated mutations in the 5′-end, while the 3′-end exhibited low tolerance. An siRNA with two 2′-O-methyl RNA nucleotides at the 5′-end and four methylated monomers at the 3′-end was as active as its unmodified counterpart and led to a prolonged silencing effect in cell culture (193).
The effectiveness of an siRNA is likely to be determined by the accessibility of its target sequence in the intended substrate. It has been suggested that the first 50–100 nucleotides of an mRNA sequence, downstream of the translation initiation sequence, should be used to target a gene and that 5′- or 3′-untranslated regions, as well as highly conserved domains (i.e., catalytic, ligand binding, etc.), should be avoided, as they are likely to contain regulatory protein binding sites (190). However, successful gene inhibition has been reported for siRNAs targeting various sequences, including the 3′-untranslated region (194). There are no reliable ways to predict or identify the “ideal” sequence for an siRNA. However, targeting different regions of a given mRNA might give different results (185). Generally, siRNAs become susceptible to RNase H; therefore, the degree of the RNase H sensitivity of a given probe reflects the RNase H accessibility of the chosen sites. In practical terms, it might be just as easy to construct and test several siRNAs.
It is becoming increasingly clear that endogenous factors, including hormones and growth factors, play an important role in the regulation of drug-metabolizing enzyme expression in both physiological and pathophysiological conditions. Our laboratory has used phospho-specific antibodies and chemical inhibitors of protein kinases to define the signaling pathways involved in insulin- and glucagon-mediated regulation of several drug-metabolizing enzymes (40,48). Small molecule chemical inhibitors of protein kinases used for this purpose in many studies, however, have been reported to have specificity problems, although many chemicals have been considered to be reasonably selective inhibitors for each target protein kinase. As with all pharmacological tools, interpretation of experiments with these protein kinase inhibitors requires caution. It is advisable to conduct experiments with at least two pharmacologically distinct inhibitors wherever possible. Dominant-negative kinase constructs allow for more kinase-specific inhibition. Recently, RNAi methods have opened new opportunities for investigators to study cell signaling pathways by leading to a highly specific mRNA degradation. Furthermore, retrovirus or adenovirus vectors have been developed for use in carrying dominant-negative kinase constructs or siRNA-expressing DNA templates into cells to mediate gene-specific silencing in cells or animals. Thus, RNAi using siRNAs to silence specific genes is a very promising method for determination of cell signaling pathways involved in protein expression in response to hormones and growth factors.
- 21.Raza H, Ahmed I, Lakhani MS, Sharma AK, Pallot D, Montague W. Effect of bitter melon (Momordica charantia) fruit juice on the hepatic cytochrome P450-dependent monooxygenases and glutathione S-transferases in streptozotocin-induced diabetic rats. Biochem Pharmacol 1996;52:1639–1642.PubMedCrossRefGoogle Scholar
- 44.De Smet K, Loyer P, Gilot D, Vercruysse A, Rogiers V, Guguen-Guillouzo C. Effects of epidermal growth factor on CYP inducibility by xenobiotics, DNA replication, and caspase activations in collagen I gel sandwich cultures of rat hepatocytes. Biochem Pharmacol 2001;61:1293–1303.PubMedCrossRefGoogle Scholar
- 54.White MF. IRS proteins and the common path to diabetes. Am J Physiol 2002; 283:E413–E422.Google Scholar
- 73.Farese RV. Insulin-sensitive phospholipid signaling systems and glucose transport. Update II. Exp Biol Med (Maywood) 2001;226:283–295.Google Scholar
- 74.Standaert ML, Bandyopadhyay G, Perez L, et al. Insulin activates protein kinases C-zeta and C-lambda by an autophosphorylation-dependent mechanism and stimulates their translocation to GLUT4 vesicles and other membrane fractions in rat adipocytes. J Biol Chem 1999;274:25308–25316.PubMedCrossRefGoogle Scholar
- 97.Yamamoto-Honda R, Tobe K, Kaburagi Y, et al. Upstream mechanisms of glycogen synthase activation by insulin and insulin-like growth factor-I. Glycogen synthase activation is antagonized by wortmannin or LY294002 but not by rapamycin or by inhibiting p21ras. J Biol Chem 1995;270:2729–2734.PubMedCrossRefGoogle Scholar
- 100.Cross DA, Alessi DR, Vandenheede JR, McDowell HE, Hundal HS, Cohen P. The inhibition of glycogen synthase kinase-3 by insulin or insulin-like growth factor 1 in the rat skeletal muscle cell line L6 is blocked by wortmannin, but not by rapamycin: evidence that wortmannin blocks activation of the mitogen-activated protein kinase pathway in L6 cells between Ras and Raf. Biochem J 1994;303:21–26.PubMedGoogle Scholar
- 101.Standaert ML, Bandyopadhyay G, Farese RV. Studies with wortmannin suggest a role for phosphatidylinositol 3-kinase in the activation of glycogen synthase and mitogen-activated protein kinase by insulin in rat adipocytes: comparison of insulin and protein kinase C modulators. Biochem Biophys Res Commun 1995; 209:1082–1088.PubMedCrossRefGoogle Scholar
- 107.Asante-Appiah E, Kennedy BP. Protein tyrosine phosphatases: the quest for negative regulators of insulin action. Am J Physiol 2003;284:E663–E670.Google Scholar
- 147.Lali FV, Hunt AE, Turner SJ, Foxwell BM. The pyridinyl imidazole inhibitor SB203580 blocks phosphoinositide-dependent protein kinase activity, protein kinase B phosphorylation, and retinoblastoma hyperphosphorylation in inter-leukin-2-stimulated T cells independently of p38 mitogen-activated protein kinase. J Biol Chem 2000;275:7395–7402.PubMedCrossRefGoogle Scholar