Role of electrostatic interactions in SH2 domain recognition: salt-dependence of tyrosyl-phosphorylated peptide binding to the tandem SH2 domain of the Syk kinase and the single SH2 domain of the Src kinase

SH2 domains are small protein domains that bind specifically to tyrosyl-phosphorylated sequences. Because phosphorylation contributes a large part of the binding free energy, it has been postulated that electrostatic interactions may play an important role in SH2 domain recognition. To test this hypothesis, we have examined the salt dependence of the interaction between tyrosyl-phosphorylated peptides and SH2 domains. The dependence of the binding constant, K(obs), on [NaCl] was shown to be strong for binding of the tandem SH2 domain of the Syk kinase (Syk-tSH2) to doubly phosphorylated peptides derived from immune-receptor tyrosine activation motifs (dpITAMs): the slopes of plots of log(K(obs)) versus log [NaCl], designated SK(obs), ranged from -2.6 +/- 0.1 to -3.1 +/- 0.2. Binding of the single SH2 domain of the Src kinase to its consensus singly phosphorylated peptide (sequence pYEEI where pY indicates a phosphotyrosine) was also highly dependent on [NaCl] with a SK(obs) value of -2.4 +/- 0.1. The ability of salt to disrupt the interactions between Syk-tSH2 and dpITAM peptides was shown to be anion-dependent with the inhibitory effect following the order: phosphate > Cl(-) > F(-). For the Syk-tSH2 system, interactions in the pY-binding pockets were shown to be responsible for a large portion of the total salt dependence: removal of either phosphate from the dpITAM peptide reduced the magnitude of SK(obs) by 40-60% and weakened binding by 2-3 orders of magnitude. Consistent with this finding, binding of the single amino acid Ac-pY-NH(2) was characterized by a large salt dependence of binding and was also dependent on the identity of the perturbing anion. The role of peptide residues C-terminal to the pY, which are implicated in determining the specificity of the phosphopeptide-SH2 domain interaction, was next probed by comparing the binding of the Src SH2 domain to a peptide containing the pYEEI sequence with that of a lower affinity variant pYAAI peptide: the magnitude of SK(obs) for the variant peptide was reduced to -1.3 +/- 0.1 as compared to -2.4 +/- 0.1 for the pYEEI peptide, indicating that in addition to pY, residues conferring peptide binding specificity contribute significantly to the salt dependence of SH2 domain binding. This study shows that electrostatic interactions play important roles not only in mediating pY recognition and binding but also in contributing to the specificity of the interactions between tyrosyl phosphopeptides and SH2 domains.     

Alternative splicing affecting the SH3A domain controls the binding properties of intersectin 1 in neurons

Intersectin 1 (ITSN1) is a conserved adaptor protein implicated in endocytosis, regulation of actin cytoskeleton rearrangements and mitogenic signaling. Its expression is characterized by multiple alternative splicing. Here we show neuron-specific expression of ITSN1 isoforms containing exon 20, which encodes five amino acid residues in the first SH3 domain (SH3A). In vitro binding experiments demonstrated that inclusion of exon 20 changes the binding properties of the SH3A domain. Endocytic proteins dynamin 1 and synaptojanin 1 as well as GTPase-activating protein CdGAP bound the neuron-specific variant of the SH3A domain with higher affinity than ubiquitously expressed SH3A. In contrast, SOS1, a guanine nucleotide exchange factor for Ras, and the ubiquitin ligase Cbl mainly interact with the ubiquitously expressed isoform. These results demonstrate that alternative splicing leads to the formation of two pools of ITSN1 with potentially different properties in neurons, affecting ITSN1 function as adaptor protein.     

Disruption of the SH2-B gene causes age-dependent insulin resistance and glucose intolerance

Insulin regulates glucose homeostasis by binding and activating the insulin receptor, and defects in insulin responses (insulin resistance) induce type 2 diabetes. SH2-B, an Src homology 2 (SH2) and pleckstrin homology domain-containing adaptor protein, binds via its SH2 domain to insulin receptor in response to insulin; however, its physiological role remains unclear. Here we show that SH2-B was expressed in the liver, skeletal muscle, and fat. Systemic deletion of SH2-B impaired insulin receptor activation and signaling in the liver, skeletal muscle, and fat, including tyrosine phosphorylation of insulin receptor substrate 1 (IRS1) and IRS2 and activation of the phosphatidylinositol 3-kinase/Akt and the Erk1/2 pathways. Consequently, SH2-B-/- knockout mice developed age-dependent hyperinsulinemia, hyperglycemia, and glucose intolerance. Moreover, SH2-B directly enhanced autophosphorylation of insulin receptor and tyrosine phosphorylation of IRS1 and IRS2 in an SH2 domain-dependent manner in cultured cells. Our data suggest that SH2-B is a physiological enhancer of insulin receptor activation and is required for maintaining normal insulin sensitivity and glucose homeostasis during aging.     

Identification of a Src kinase SH3 binding site in the C-terminal domain of the human ErbB2 receptor tyrosine kinase

Overexpression of the ErbB2 receptor tyrosine kinase is associated with most aggressive tumors in breast cancer patients and is thus one of the main investigated therapeutic targets. Human ErbB2 C-terminal domain is an unstructured anchor that recruits specific adaptors for signaling cascades resulting in cell growth, differentiation and migration. Herein, we report the presence of a SH3 binding motif in the proline rich unfolded ErbB2 C-terminal region. NMR analysis of this motif supports a PPII helix conformation and the binding to Fyn-SH3 domain. The interaction of a kinase of the Src family with ErbB2 C-terminal domain could contribute to synergistic intracellular signaling and enhanced oncogenesis.     

Nitrotyrosine mimics phosphotyrosine binding to the SH2 domain of the src family tyrosine kinase lyn

The nitration of tyrosine residues in protein occurs through the action of reactive oxygen and nitrogen species and is considered a marker of oxidative stress under pathological conditions. The most active nitrating species so far identified is peroxynitrite, the product of the reaction between nitric oxide and superoxide anion. Previously, we have reported that in erythrocytes peroxynitrite irreversibly upregulates lyn, a tyrosine kinase of the src family. In this study we investigated the possible role of tyrosine nitration in the mechanism of lyn activation. We found that tyrosine containing peptides modelled either on the C-terminal tail of src kinases or corresponding to the first 15 amino acids of human erythrocyte band 3 were able to activate lyn when the tyrosine was substituted with 3-nitrotyrosine. The activity of nitrated peptides was shared with phosphorylated but not with unphosphorylated, chlorinated or scrambled peptides. Recombinant lyn src homology 2 (SH2) domain blocked the capacity of the band 3-derived nitrotyrosine peptide to activate lyn and we demonstrated that this peptide specifically binds the SH2 domain of lyn. We propose that nitropeptides may activate src kinases through the displacement of the phosphotyrosine in the tail from its binding site in the SH2 domain. These observations suggest a new mechanism of peroxynitrite-mediated signalling that may be correlated with the upregulation of tyrosine phosphorylation observed in several pathological conditions.     

Alternative splicing of RGS8 gene changes the binding property to the M1 muscarinic receptor to confer receptor type-specific Gq regulation

RGS proteins constitute a large family that modulates heterotrimeric G-protein signaling. We previously showed that RGS8 suppressed Gq signaling in a receptor type-specific manner. To elucidate molecular mechanisms underlying receptor-specific attenuation by RGS8, we examined whether RGS8 can interact with certain G-protein-coupled receptors. By pull-down assay, we showed that RGS8 directly binds to the third intracellular (i3) loop of M1 and M3 muscarinic acetylcholine receptors (mAChRs). The binding of RGS8S, a splice variant with a different N-terminus, was weaker. RGS8 could bind specifically to the C-terminal part of M1i3 (containing amino acids of 304-353 of i3 of human M1-mAChR), but RGS8S could not. Moreover, deletion of the N-terminal 9 amino acids and substitution of both Arg-8 and Arg-9 of RGS8 with Ala resulted in reduced binding to M1i3. BRET experiments revealed that RGS8 actually interacts with M1-mAChR, but RGS8S does not interact in the living cells. The RGS8 mutant, which had less binding ability to M1i3, showed a reduced inhibitory function of Gq signaling through M1-mAChR. These results demonstrated that RGS8 can directly interact with M1-mAChR via its N-terminus and the i3 loop of the receptor, and this binding must play an essential role in receptor-specific suppression by RGS8.     

Alternative splicing of the human MUC2 gene

Human colon cancers differ in amounts of MUC2 mucin synthesized. However, it is unclear whether MUC2 encodes a single protein. When clones of human colon cancer cells were assayed with antibodies against the TR2 mucin repeat or non-TR2 epitopes, differences in relative expression of MUC2 proteins suggested multiple immunoreactive forms. RT-PCR analysis detected the established 15kbp MUC2 cDNA and a novel form (designated MUC2.1) lacking the MUC2 TR2 repeat. Sequencing of cDNA and genomic DNA indicated that MUC2.1 results from an alternate splice donor. RT-PCR with splice-junction spanning primers confirmed the expression of MUC2.1 mRNA. Anti-MUC2.1 antibody stained colon cancer cells and normal colon in a pattern different from TR2-specific antibody. The presence of MUC2.1 mucin may help us to explain previous conflicting reports that have attempted to correlate the relative abundance of MUC2 protein and/or mRNA with the biological behavior of colon cancer cells.     

Essential role of PSM/SH2-B variants in insulin receptor catalytic activation and the resulting cellular responses

The positive regulatory role of PSM/SH2-B downstream of various mitogenic receptor tyrosine kinases or gene disruption experiments in mice support a role of PSM in the regulation of insulin action. Here, four alternative PSM splice variants and individual functional domains were compared for their role in the regulation of specific metabolic insulin responses. We found that individual PSM variants in 3T3-L1 adipocytes potentiated insulin-mediated glucose and amino acid transport, glycogenesis, lipogenesis, and key components in the metabolic insulin response including p70 S6 kinase, glycogen synthase, glycogen synthase kinase 3 (GSK3), Akt, Cbl, and IRS-1. Highest activity was consistently observed for PSM alpha, followed by beta, delta, and gamma with decreasing activity. In contrast, dominant-negative peptide mimetics of the PSM Pro-rich, pleckstrin homology (PH), or src homology 2 (SH2) domains inhibited any tested insulin response. Potentiation of the insulin response originated at the insulin receptor (IR) kinase level by PSM variant-specific regulation of the Km (ATP) whereas the Vmax remained unaffected. IR catalytic activation was inhibited by peptide mimetics of the PSM SH2 or dimerization domain (DD). Either peptide should disrupt the complex of a PSM dimer linked to IR via SH2 domains as proposed for PSM activation of tyrosine kinase JAK2. Either peptide abolished downstream insulin responses indistinguishable from PSM siRNA knockdown. Our results implicate an essential role of the PSM variants in the activation of the IR kinase and the resulting metabolic insulin response. PSM variants act as internal IR ligands that in addition to potentiating the insulin response stimulate IR catalytic activation even in the absence of insulin.     

Alternative splicing introduces a nuclear localization signal that targets multifunctional CaM kinase to the nucleus

Intracellular targeting may enable protein kinases with broad substrate- specificities, such as multifunctional Ca2+/calmodulin-dependent protein kinase (CaM kinase) to achieve a selectivity of action in vivo. We have examined the intracellular targeting of three delta-CaM kinase isoforms. The delta B-CaM kinase isoform is targeted to the nucleus in transfected cells while the delta A- and delta C-CaM kinase isoforms are cytosolic/cytoskeletal. A chimeric construct of alpha-CaM kinase containing the delta B-CaM kinase variable domain is rerouted to the nucleus while the native alpha-CaM kinase and chimeras of alpha-CaM kinase which contain the delta A- or delta C-CaM kinase variable domains are retained in the cytoplasm. Using site-directed mutagenesis, we have defined a nuclear localization signal (NLS) within an 11-amino acid sequence, likely inserted by alternative splicing, in the variable domain of delta B-CaM kinase. Isoform-specific nuclear targeting of CaM kinase is probably a key mechanism in the selective regulation of nuclear functions by CaM kinase. CaM kinase is a multimer that can be composed of several isoforms. We find that when cells express two different isoforms of CaM kinase, cellular targeting is determined by the ratio of the isoforms. When an excess of the cytoplasmic isoform of CaM kinase is coexpressed along with the nuclear isoform, both isoforms are localized in the cytoplasm. Conversely an excess of the nuclear isoform can reroute the cytoplasmic isoform to the nucleus. The nuclear isoform likely coassembles with the cytosolic isoform, to form a heteromultimeric holoenzyme which is transported into the nucleus. These experiments demonstrate isoform-specific targeting of CaM kinase and indicate that such targeting can be modified by the expression of multiple isoforms of the enzyme.     

Differential binding to and regulation of JAK2 by the SH2 domain and N-terminal region of SH2-bbeta

SH2-Bbeta has been shown to bind via its SH2 (Src homology 2) domain to tyrosyl-phosphorylated JAK2 and strongly activate JAK2. In this study, we demonstrate the existence of an additional binding site(s) for JAK2 within the N-terminal region of SH2-Bbeta (amino acids 1 to 555) and the ability of this region of SH2-B to inhibit JAK2. Four lines of evidence support the existence of this additional binding site(s). In a glutathione S-transferase pull-down assay, wild-type SH2-Bbeta and SH2-Bbeta(R555E) with a defective SH2 domain bind to both tyrosyl-phosphorylated JAK2 from growth hormone (GH)-treated cells and non-tyrosyl-phosphorylated JAK2 from control cells, whereas the SH2 domain of SH2-Bbeta binds only to tyrosyl-phosphorylated JAK2 from GH-treated cells. Similarly, JAK2 is present in alphaSH2-B immunoprecipitates in the absence and presence of GH, with GH substantially increasing the coprecipitation of JAK2 with SH2-B. When coexpressed in COS cells, SH2-Bbeta coimmunoprecipitates not only wild-type, tyrosyl-phosphorylated JAK2 but also kinase-inactive, non-tyrosyl-phosphorylated JAK2(K882E), although to a lesser extent. DeltaC555 (amino acids 1 to 555 of SH2-Bbeta) that lacks most of the SH2 domain binds similarly to wild-type JAK2 and kinase-inactive JAK2(K882E). Experiments using a series of N- and C-terminally truncated SH2-Bbeta constructs indicate that the pleckstrin homology (PH) domain (amino acids 269 to 410) and amino acids 410 to 555 are necessary for maximal binding of SH2-Bbeta to inactive JAK2, but neither region alone is sufficient for maximal binding. The SH2 domain of SH2-Bbeta is necessary and sufficient for the stimulatory effect of SH2-Bbeta on JAK2 and JAK2-mediated tyrosyl phosphorylation of Stat5B. In contrast, DeltaC555 lacking the SH2 domain, and to a lesser extent the PH domain alone, inhibits JAK2. DeltaC555 also blocks JAK2-mediated tyrosyl phosphorylation of Stat5B in COS cells and GH-stimulated nuclear accumulation of Stat5B in 3T3-F442A cells. These data indicate that in addition to the SH2 domain, SH2-Bbeta has one or more lower-affinity binding sites for JAK2 within amino acids 269 to 555. The interaction via this site(s) in SH2-B with inactive JAK2 seems likely to increase the local concentration of SH2-Bbeta around JAK2, thereby facilitating binding of the SH2 domain to ligand-activated JAK2. This would result in a more rapid and robust cellular response to hormones and cytokines that activate JAK2. This interaction between inactive JAK2 and SH2-B may also help prevent abnormal activation of JAK2.     

Alternative splicing of the actin binding domain of human cortactin affects cell migration

Cortactin is a filamentous actin (F-actin)-binding protein that regulates cytoskeletal dynamics by activating the Arp2/3 complex; it binds to F-actin by means of six N-terminal "cortactin repeats". Gene amplification of 11q13 and consequent overexpression of cortactin in several human cancers is associated with lymph node metastasis. Overexpression as well as tyrosine phosphorylation of cortactin has been reported to enhance cell migration, invasion, and metastasis. Here we report the identification of two alternative splice variants (SV1 and SV2) that affect the cortactin repeats: SV1-cortactin lacks the 6th repeat (exon 11), whereas SV2-cortactin lacks the 5th and 6th repeats (exons 10 and 11). SV-1 cortactin is found co-expressed with wild type (wt)-cortactin in all tissues and cell lines examined, whereas the SV2 isoform is much less abundant. SV1-cortactin binds F-actin and promotes Arp2/3-mediated actin polymerization equally well as wt-cortactin, whereas SV2-cortactin shows reduced F-actin binding and polymerization. Alternative splicing of cortactin does not affect its subcellular localization or growth factor-induced tyrosine phosphorylation. However, cells that overexpress SV1- or SV2-cortactin show significantly reduced cell migration when compared with wt-cortactin-overexpressing cells. Thus, in addition to overexpression and tyrosine phosphorylation, alternative splicing of the F-actin binding domain of cortactin is a new mechanism by which cortactin influences cell migration.     

A novel form of fibroblast growth factor receptor 2. Alternative splicing of the third immunoglobulin-like domain confers ligand binding specificity.

The fibroblast growth factor receptor 2 (FGFR2) gene is expressed as alternatively spliced mRNAs that encode bacterially expressed kinase, the keratinocyte growth factor receptor, or K-sam. We have now isolated a novel FGFR2 cDNA that is identical with the previously cloned human bacterially expressed kinase, except in the third immunoglobulin-like domain. The ligand binding properties of FGFR2 were studied by expressing the protein in rat L6 muscle myoblasts. Unlike human bacterially expressed kinase which binds acidic and basic FGF with similar affinities, FGFR2 bound acidic FGF with approximately 1000-fold higher affinity than basic FGF. These results indicate that alternative splicing of the FGFR2 gene in the region encoding the carboxyl-terminal half of the third immunoglobulin domain determines the ligand specificity of this group of receptors.     

Alternative splicing, the gene concept, and evolution

Alternative splicing allows for the production of many gene products from a single coding sequence. I introduce the concept of alternative splicing via some examples. I then discuss some current hypotheses about the explanatory role of alternative splicing, including the claim that splicing is a significant contributor to the difference in complexity between the human genome and proteosome. Hypotheses such as these bring into question our working concepts of the gene. I examine several gene concepts introduced to cope with processes such as alternative splicing. Next I introduce some hypotheses about the evolution of mechanisms alternative splicing in higher organisms. I conclude that attention to alternative splicing reveals that we adopt an attitude that developmental theorizing must inform evolutionary theorizing and vice versa.     

Mapping surface structures of the human insulin receptor with monoclonal antibodies: localization of main immunogenic regions to the receptor kinase domain

A panel of 37 monoclonal antibodies to the human insulin receptor has been used to characterize the receptor's major antigenic regions and their relationship to receptor functions. Three antibodies recognized extracellular surface structures, including the insulin binding site and a region not associated with insulin binding. The remaining 34 monoclonal antibodies were directed against the cytoplasmic domain of the receptor beta subunit. Competitive binding studies demonstrated that four antigenic regions (beta 1, beta 2, beta 3, and beta 4) are found on this domain. Sixteen of the antibodies were found to be directed against beta 1, nine against beta 2, seven against beta 3, and two against beta 4. Antibodies to all four regions inhibited the receptor-associated protein kinase activity to some extent, although antibodies directed against the beta 2 region completely inhibited the kinase activity of the receptor both in the autophosphorylation reaction and in the phosphorylation of an exogenous substrate, histone. Antibodies to the beta 2 region also did not recognize autophosphorylated receptor. In addition, antibodies to this same region recognized the receptor for insulin-like growth factor I (IGF-I) as well as the insulin receptor. In contrast, antibodies to other cytoplasmic regions did not recognize the IGF-I receptor as well as the insulin receptor. These results indicate that the major immunogenic regions of the insulin receptor are located on the cytoplasmic domain of the receptor beta subunit and are associated with the tyrosine-specific kinase activity of the receptor. In addition, these results suggest that a portion of the insulin receptor is highly homologous to that of the IGF-I receptor.     

Solid-phase binding assays of peptides using EGFP-Src SH2 domain fusion protein and biotinylated Src SH2 domain

Two solid-phase binding assays were designed and evaluated for their potential use in comparing the affinity of peptides to the Src SH2 domain. Resin beads attached to peptides were incubated with the enhanced green fluorescence protein(EGFP)-Src SH2 domain fusion protein or the biotinylated Src SH2 domain and extensively washed. The beads-attached tetrapeptides with high affinities to the EGFP-Src SH2 domain showed more fluorescence intensity than those beads containing tetrapeptides with weak binding affinities, as shown by fluorescence microscopy and fluorescence imaging system. Only the beads attached to pYEEI produced a dark purple color on incubation of the beads, respectively, with the biotinylated Src kinases SH2 domain, alkaline phosphatase-coupled streptavidin, and BCIP/NBT. These solid-phase binding assays may have potential applications for the screening of peptides for the Src kinases SH2 domains.     

Alternative splicing and gene structure of the transforming growth factor beta-activated kinase 1

We have identified a fourth splice variant of the TGF beta-activated kinase (TAK1), called TAK1-d, and identified an error in the previously published TAK1-c sequence. Our data shows that the c and d variants encode proteins whose carboxyl ends differ markedly from those of variants a and b. Analysis of the human TAK1 gene sequence, located at 6q16.1-q16.3, shows that the coding sequence is organised in 17 exons. The four splice variants result from alternative splicing of exons 12 and 16, the reading frame of exon 17 being determined by the presence or absence of exon 16. Study of the relative levels of expression of the four splice variants showed significant variations between tissues. Our evidence suggests that the alternative splicing of the TAK1 mRNA may have important functional implications.