Role of the third intracellular loop of the angiotensin II receptor subtype AT2 in ligand-receptor interaction

Angiotensin II (Ang II) receptor subtypes AT1 and AT2 share 34% overall homology, but the least homology is in their third intracellular loop (3rd ICL). In an attempt to elucidate the role of the 3rd ICL in determining the similarities and differences in the functions of the AT1 and the AT2 receptors, we generated a chimeric receptor in which the 3rd ICL of the AT2 receptor was replaced with that of the AT1 receptor. Ligand-binding properties and signaling properties of this receptor were assayed by expressing this receptor in Xenopus oocytes. Ligand-binding studies using [125I-Sar1-Ile8] Ang II, a peptidic ligand that binds both the AT1 and the AT2 receptor subtypes, and 125I-CGP42112A, a peptidic ligand that is specific for the AT2 receptor, showed that the chimeric receptor has lost affinity to both ligands. However, IP3 levels of the oocytes expressing the chimeric receptor were comparable to the IP3 levels of the oocytes expressing the AT1 receptor, suggesting that the chimeric receptors could couple to phospholipase C pathway in response to Ang II. We have shown previously that the nature of the amino acid present in the position 215 located in the fifth transmembrane domain (TMD) of the AT2 receptor plays an important role in determining its affinity to different ligands. Our results from the ligand-binding studies of the chimeric receptor further support the idea that the structural organization of the region spanning the 5th TMD and the 3rd ICL of the AT2 receptor has an important role in determining the ligand-binding properties of this receptor.     

Identification of the region of AT2 receptor needed for inhibition of the AT1 receptor-mediated inositol 1,4,5-triphosphate generation

Increase in the intracellular inositol triphosphate (IP3) levels in Xenopus oocytes in response to expression and activation of rat angiotensin II (Ang II) receptor AT1 was inhibited by co-expression of rat AT2 receptor. To identify which region of the AT2 was involved in this inhibition, ability of three AT2 mutants to abolish this inhibition was analyzed. Deletion of the C-terminus of the AT2 did not abolish this inhibition. Replacing Ile249 in the third intracellular loop (3rd ICL) of the AT2 with proline, corresponding amino acid in the AT1, in the mutant M6, resulted in slightly reduced affinity to [125I]Ang II (K(d)=0.259 nM), however, did not abolish the inhibition. In contrast, replacing eight more amino acids in the 3rd ICL of the AT2 (at positions 241-244, 250-251 and 255-256) with that of the AT1 in the mutant M8, not only increased the affinity of the AT2 receptor to [125I]Ang II (K(d)=0.038 nM) but also abolished AT2-mediated inhibition. Interestingly, activation of the M8 by Ang II binding also resulted in increase in the intracellular IP(3) levels in oocytes. These results imply that the region of the 3rd ICL of AT2 spanning amino acids 241-256 is sufficient for the AT2-mediated inhibition of AT1-stimulated IP3 generation. Moreover, these nine mutations are also sufficient to render the AT2 with the ability to activate phospholipase C.     

Ligand-dependent complex formation between the Angiotensin II receptor subtype AT2 and Na+/H+ exchanger NHE6 in mammalian cells

Involvement of Angiotensin II (Ang II) in the regulation of sodium levels by modulating the Na+/H+ exchangers is demonstrated in many tissues. Screening of a mouse 17-day fetus cDNA library with the Angiotensin II receptor AT2 as the bait in yeast two-hybrid assay led us to identify an AT2-interacting mouse fetus peptide that shared 98% amino acid identity with the corresponding region of the human NHE6. NCBI Blast search showed that the clone 6430520C02 (GenBank Accession # AK032326) of the mouse genome project carried the complete sequence of this new mouse NHE6 isoform. The human and mouse NHE6 peptides share 97% overall homology. Further analysis showed that the region spanning the third intracellular loop and C-terminal cytoplasmic tail of the AT2 directly interacted with a 182 amino acid region that spans the predicted 5th intracellular loop and the initial part of the C-terminus of the mouse NHE6 in yeast two-hybrid assay. This 182-amino acid region that interacted with the AT2 also shares 98% homology with the corresponding region of rat NHE6 and therefore is highly conserved across species. We detected widespread expression of this NHE6 isoform in several rat tissues including 10-day fetus, 17-day fetus, and 30-day post-natal tissues of heart, brain, kidney and muscle. Moreover, the AT2 co-immunoiprecipitated with a hemagglutinin tagged NHE6 when expressed in human cell line MCF-7, and activated by AngII. This ligand-dependent complex formation between the AT2 and NHE6 suggests that the hormone Ang II may act as a regulator of NHE6, and Ang II-mediated direct protein-protein interaction between AT2 and NHE6 could be a mechanism for modulating the functions of the ubiquitously expressed NHE6 in different tissues.     

Multiple ramp domains are required for generation of amylin receptor phenotype from the calcitonin receptor gene product

Calcitonin (CT), calcitonin gene-related peptide (CGRP), amylin, and adrenomedullin constitute a family of structurally related peptides that signal via either the calcitonin receptor-like receptor or the CT receptor, with receptor phenotype determined by coexpression of one of the three receptor activity-modifying proteins (RAMPs). The nature of the interaction between the receptor and RAMP was investigated using chimeras between RAMP1 and RAMP2 where the amino-terminal domain of RAMP1 was attached to the transmembrane domain and carboxy terminus of RAMP2 and called RAMP1/2, and vice versa for RAMP2/1. Cotransfection of wild-type or chimeric RAMPs with the insert-negative isoform of the human CT receptor (hCTR(I1-)) into COS-7 cells resulted in the expression of (125)I-rat amylin binding sites. Highest specific binding was observed when either RAMP1 or RAMP2/1 were cotransfected, indicating the importance of the RAMP transmembrane domain and/or carboxy terminus for the degree to which amylin receptors are expressed. In contrast, the phenotype generated was primarily determined by the amino terminus, with similar RAMP1- and RAMP1/2-induced receptor phenotypes that had higher affinity for human CGRPalpha and lower affinity for human calcitonin than the RAMP2- and RAMP2/1-induced receptors.     

Characterization of an interaction between insulin receptor substrate 1 and the insulin receptor by using the two-hybrid system.

Insulin receptor substrate 1 (IRS-1) is a major substrate of the insulin receptor and has been implicated in insulin signaling. Although IRS-1 is thought to interact with the insulin receptor, the nature of the interaction has not been defined. In this study, we used the two-hybrid assay of protein-protein interaction in the yeast Saccharomyces cerevisiae to study the interaction between human IRS-1 and the insulin receptor. We demonstrate that IRS-1 forms a specific complex with the cytoplasmic domain of the insulin receptor when both are expressed as hybrid proteins in yeast cells. We show that the interaction is strictly dependent upon receptor tyrosine kinase activity, since IRS-1 shows no interaction with a kinase-inactive receptor hybrid containing a mutated ATP-binding site. Furthermore, mutation of receptor tyrosine 960 to phenylalanine eliminates IRS-1 interaction in the two-hybrid assay. These data suggest that the interaction between IRS-1 and the receptor is direct and provide evidence that the juxtamembrane domain of the receptor is involved. Furthermore, we show that a 356-amino-acid region encompassed by amino acids 160 through 516 of IRS-1 is sufficient for interaction with the receptor in the two-hybrid assay. Lastly, in agreement with our findings for yeast cells, we show that the insulin receptor is unable to phosphorylate an IRS-1 protein containing a deletion of amino acids 45 to 516 when expressed in COS cells. The two-hybrid assay should provide a facile means by which to pursue a detailed understanding of this interaction.     

The critical role of carboxy-terminal amino acids in ligand-dependent and -independent transactivation of the constitutive androstane receptor

The mouse constitutive androstane receptor (CAR) is a unique member of the nuclear receptor superfamily, for which an inverse agonist, the testosterone metabolite 5alpha-androstan-3alpha-ol (androstanol), and an agonist, the xenobiotic 1,4-bis[2-(3, 5-dichloropyridyloxy)] benzene, are known. In this study the role of the transactivation domain 2 (AF-2) of CAR was investigated, which is formed by the seven most carboxy-terminal amino acids of the receptor. The AF-2 domain was shown to be critical for the constitutive activity by mediating a ligand-independent interaction of CAR with coactivator (CoA) proteins. In addition this domain increased and decreased contact with CoAs in the presence of agonist and inverse agonist, respectively. In analogy to classical endocrine nuclear receptors, in CAR the charge clamp between K187 (in helix 3) and E355 (within the AF-2 domain) was expected to be critical for its interaction with CoAs. However, the hydrophobic amino acids L352, L353, and I356 on the surface of the AF-2 domain were found to be more important for this protein-protein interaction. Moreover, these amino acids and C357 were shown to be involved in the response of CAR to androstanol. Interestingly, the cysteine at position 357 appears to block classical endocrine responsiveness of CAR to agonists, since mutagenesis of this amino acid both reduced CoA interaction in the absence of ligand and drastically increased inducibility by 1,4-bis[2-(3, 5-dichloropyridyloxy)] benzene. We showed that this blockade is not due to an intramolecular disulfide bridge, but is probably caused by an interaction between C357 and Y336.     

The third intracellular loop stabilizes the inactive state of the neuropeptide Y1 receptor

Constitutively active G-protein-coupled receptors (GPCRs) can signal even in the absence of ligand binding. Most Class I GPCRs are stabilized in the resting conformation by intramolecular interactions involving transmembrane domain (TM) 3 and TM6, particularly at loci 6.30 and 6.34 of TM6. Signaling by Gi/Go-coupled receptors such as the Neuropeptide Y1 receptor decreases already low basal metabolite levels. Thus, we examined constitutive activity using a biochemical assay mediated by a Gi/Gq chimeric protein and a more direct electrophysiological assay. Wild-type (WT-Y1) receptors express no measurable, agonist-independent activation, while mu-opioid receptors (MOR) and P2Y12 purinoceptors showed clear evidence of constitutive activation, especially in the electrophysiological assay. Neither point mutations at TM6 (T6.30A or N6.34A) nor substitution of the entire TM3 and TM6 regions from the MOR into the Y1 receptor increased basal WT-Y1 activation. By contrast, chimeric substitution of the third intracellular loop (ICL3) generated a constitutively active, Y1-ICL3-MOR chimera. Furthermore, the loss of stabilizing interactions from the native ICL3 enhanced the role of surrounding residues to permit basal receptor activation; because constitutive activity of the Y1-ICL3-MOR chimera was further increased by point mutation at locus 6.34, which did not alter WT-Y1 receptor activity. Our results indicate that the ICL3 stabilizes the Y1 receptor in the inactive state and confers structural properties critical for regulating Y receptor activation and signal transduction. These studies reveal the active participation of the ICL3 in the stabilization and activation of Class I GPCRs.     

Mutational analysis of the cellular receptor for poliovirus.

To identify sequences of the cellular poliovirus receptor (PVR) involved in viral infection, mutant PVR cDNAs were constructed and assayed for biological activity in mouse L cells. To confirm that mutant PVRs reached the cell surface, an immunological tag, consisting of part of CH3 from human immunoglobulin G1, was engineered into the PVR. Deletion of PVR amino acids 256 to 320 or 385 to the carboxy terminus yielded receptors that were able to support poliovirus infection. PVRs lacking amino acids 40 to 136 or 137 to 256 were expressed at the cell surface but were not active as receptors for poliovirus. The results show that immunoglobulin-type domain 3 and the extreme carboxy terminus of the PVR are not required for viral receptor function, but sequences within the two amino-terminal domains contribute to the initiation of poliovirus infection.     

Interaction of transcriptional intermediary factor 2 nuclear receptor box peptides with the coactivator binding site of estrogen receptor alpha

The activation function 2/ligand-dependent interaction between nuclear receptors and their coregulators is mediated by a short consensus motif, the so-called nuclear receptor (NR) box. Nuclear receptors exhibit distinct preferences for such motifs depending both on the bound ligand and on the NR box sequence. To better understand the structural basis of motif recognition, we characterized the interaction between estrogen receptor alpha and the NR box regions of the p160 coactivator TIF2. We have determined the crystal structures of complexes between the ligand-binding domain of estrogen receptor alpha and 12-mer peptides from the Box B2 and Box B3 regions of TIF2. Surprisingly, the Box B3 module displays an unexpected binding mode that is distinct from the canonical LXXLL interaction observed in other ligand-binding domain/NR box crystal structures. The peptide is shifted along the coactivator binding site in such a way that the interaction motif becomes LXXYL rather than the classical LXXLL. However, analysis of the binding properties of wild type NR box peptides, as well as mutant peptides designed to probe the Box B3 orientation, suggests that the Box B3 peptide primarily adopts the "classical" LXXLL orientation in solution. These results highlight the potential difficulties in interpretation of protein-protein interactions based on co-crystal structures using short peptide motifs.     

Site-directed polyclonal antibodies inhibit binding of activated estrogen receptor to DNA

We have generated three polyclonal antisera to the DNA-binding domain of the human estrogen receptor (hER). Antiserum AT2A was generated against a peptide spanning amino acids 231-245 of hER, while antisera AT3A and AT3B were generated against a peptide spanning amino acids 247-261 of hER. The interaction of these three antisera with ER has been characterized by sucrose density gradient analysis. The antisera bound to the unactivated (8S), salt-activated (4S), and heat transformed (5S) ER complex. All the antisera appeared to be site-specific since binding of salt-activated ER to the antisera was blocked by the presence of excess free synthetic peptides. Antisera AT3A and AT3B inhibited the binding to DNA of the KCl-activated 4S ER and the heat-transformed 5S ER. Although antiserum AT2A bound to ER, it did not inhibit DNA binding of activated ER complexes. The ability of antisera AT3A and AT3B to inhibit ER binding to DNA was concentration dependent. Once bound to the DNA, ER complexes were not significantly affected by incubation with the antisera, suggesting that binding of DNA to ER inhibits antibody ER interaction and renders that domain inaccessible to the antibodies. These results demonstrate that site-directed antibodies to ER inhibit binding of activated ER complexes to DNA in vitro.     

The interleukin 1 (IL-1) receptor accessory protein Toll/IL-1 receptor domain: analysis of putative interaction sites in vitro mutagenesis and molecular modeling

The Toll/interleukin 1 (IL-1) receptor family plays an important role in both innate and adaptive immunity. These receptors are characterized by a C-terminal homology motif called the Toll/IL-1 receptor (TIR) domain. A principal function of the TIR domain is mediating homotypic protein-protein interactions in the signal transduction pathway. To suggest interaction sites of TIR domains in the IL-1 receptor complex, we modeled the putative three-dimensional structure of the TIR domain within the co-receptor chain, IL-1 receptor accessory protein. The model was based on homology with the crystal structures of human TLR1 and TLR2. The final structure of the IL-1 receptor accessory protein TIR domain suggests the conserved regions box 1 and 2, including Pro-446, as well as box 3 within the C-terminal alpha-helix as possible protein-protein interaction sites due to their exposure and their electrostatic potential. Pro-446, corresponding to the Pro/His mutation in dominant negative TLR4, is located in the third loop at the outmost edge of the TIR domain and does not play any structural role. Inhibition of IL-1 responsiveness seen after substitution of Pro-446 by charged amino acids is due to the loss of an interaction site for other TIR domains. Amino acids 527-534 as part of the loop close to the conserved box 3 are critical for recruitment of myeloid differentiation factor 88 and to a lesser extent for IL-1 responsiveness. Modeling suggests that native folding of the TIR domain may be approached by the responsive deletion mutants delta528-534 and delta527-533, whereas the C-terminal beta-strand and/or alpha-helix is displaced in the nonresponsive mutant delta527-534.     

Internalization of the Hm1 muscarinic cholinergic receptor involves the third cytoplasmic loop

The m1 muscarinic receptor was previously shown to stimulate phosphatidyl inositol (PI) turnover and to internalize rapidly upon agonist activation. Three receptor mutants with large deletions of the third cytoplasmic loop (i3) of human Hm1, leaving only 11 and 8 amino acids at the amino and carboxy terminal junctions of i3, respectively, retained full ability to stimulate PI turnover, when expressed in U293 cells, but receptor internalization was greatly reduced in two mutants with deletions reaching close to the NH2 terminal of i3. We propose that a receptor domain located toward the amino terminal junction of i3 plays a role in Hm1 internalization.     

Roles of the intracellular regions of angiotensin II receptor AT2 in mediating reduction of intracellular cGMP levels

We have shown previously that the angiotensin II (Ang II) receptor AT2 reduces the intracellular levels of cGMP in Xenopus oocytes when activated by ligand binding, and the C-terminal cytoplasmic tail of the AT2 acts as a negative regulator of this function. Here we report the effects of mutations in the 2nd and 3rd intracellular loops of AT2 on AT2-mediated cGMP reduction. Mutating the highly conserved DRY motif (D141G-R142G-Y143A) of the 2nd ICL implicated in activating G(alpha) subunit of trimeric G-proteins did not affect AT2-mediated cGMP reduction. Moreover, anti-Gialpha antibody or phosphodiesterase inhibitor IBMX did not inhibit AT2-mediated cGMP reduction, suggesting that Gialpha activation and subsequent phosphodiesterase activation are not involved in this function. In contrast, mutations T250R-R251N and L255F-K256R located in the C-terminus of the 3rd ICL of AT2 retained ligand-binding properties of the wild-type AT2, and its ability to interact with the ErbB3 in yeast two-hybrid assay, but abolished AT2-mediated cGMP reduction. Similarities in the roles of ICLs of AT2 in AT2-mediated cGMP reduction in oocytes, and AT2-mediated SHP1 activation in COS-7 cells, (need of 3rd ICL for both functions and lack of involvement of DRY motif), suggest that the cascade of events in these two signaling mechanisms could be similar, and that an oocyte-specific SHP1-like protein may be involved in AT2-mediated cGMP reduction in these cells.     

Structure-Function Analysis of Nucleolin and ErbB Receptors Interactions

Background

The ErbB receptor tyrosine kinases and nucleolin are major contributors to malignant transformation. Recently we have found that cell-surface ErbB receptors interact with nucleolin via their cytoplasmic tail. Overexpression of ErbB1 and nucleolin leads to receptor phosphorylation, dimerization and anchorage independent growth.

Methodology/Principal Findings

In the present study we explored the regions of nucleolin and ErbB responsible for their interaction. Using mutational analyses, we addressed the structure–function relationship of the interaction between ErbB1 and nucleolin. We identified the ErbB1 nuclear localization domain as nucleolin interacting region. This region is important for nucleolin-associated receptor activation. Notably, though the tyrosine kinase domain is important for nucleolin-associated receptor activation, it is not involved in nucleolin/ErbB interactions. In addition, we demonstrated that the 212 c-terminal portion of nucleolin is imperative for the interaction with ErbB1 and ErbB4. This region of nucleolin is sufficient to induce ErbB1 dimerization, phosphorylation and growth in soft agar.

Conclusions/Significance

The oncogenic potential of ErbB depends on receptor levels and activation. Nucleolin affects ErbB dimerization and activation leading to enhanced cell growth. The C-terminal region of nucleolin and the ErbB1 NLS-domain mediate this interaction. Moreover, when the C-terminal 212 amino acids region of nucleolin is expressed with ErbB1, it can enhance anchorage independent cell growth. Taken together these results offer new insight into the role of ErbB1 and nucleolin interaction in malignant cells.     

Neuregulin signaling via ErbB receptor assemblies in the nervous system

Neuregulins (NRG) play important roles in the development, maintenance, and repair of the nervous system, with influences on neuronal migration, synaptogenesis, receptor subunit composition, and the proliferation/survival of oligodendrocytes and Schwann cells. However, the precise detail of how the NRGs signal through ErbB receptors, particularly at central synapses, is incomplete. The receptor kinase domain provides sites for association with adaptor proteins. In addition, evidence from recent reports suggests that ErbB2/4 receptors, through their C-terminal amino acids, can form specific associations with scaffolding proteins. The existence of such assemblies expands the range of signaling cascades available to the NRGs.     

Intensification of growth factor receptor signalling by phorbol treatment of ligand-primed cells implies a dimer-stabilizing effect of protein kinase C-dependent juxtamembrane domain phosphorylation.

Protein kinase C (PKC) phosphorylates the juxtamembrane domain of many growth factor receptors, but the physiologic effect of this modification on ligand signalling and desensitisation is unclear. Here we show that PKC-dependent transmodulation of EGFR and ErbB2 signalling is schedule-specific: prolonged pre-treatment of A431 cells with the PKC agonist phorbol dibutyrate potently inhibits subsequent ligand-induced EGFR signalling as expected, but EGF pre-treatment reverses the inhibitory effect of phorbol. The agonist activity of PKC on receptor signalling is even more apparent when cells are treated with phorbol in the presence of a tyrosine phosphatase inhibitor. Because these findings suggested a synergistic interaction between tyrosine- and PKC-dependent phosphorylation events, we sought to define the interactions of tyrosine-phosphorylated and PKC-modified ErbB2 subsets within EGF-inducible hetero-oligomers. Growth factor-dependent PKC transphosphorylation takes place exclusively within endocytosed tyrosine-phosphorylated receptor oligomers. Moreover, phorbol differentially affects two ErbB2 C-terminal autophosphorylation sites: whereas phosphorylation of Tyr1222 is reduced, phosphorylation of Tyr1139 is increased. These results suggest that PKC-dependent phosphorylation of the juxtamembrane domain may contribute positively to both internalisation and signalling of ligand-activated receptors, simultaneously accelerating termination of growth factor action. We propose that transient PKC-dependent signal amplification results from enhanced stability of liganded receptor oligomers due to phosphorylation-dependent juxtamembrane domain interactions, analogous to the protein-protein binding now known to be induced by serine-threonine phosphorylation of CREB and SMAD.     

Studies with antibodies against the conserved cysteine region of progesterone receptor

Polyclonal antibodies were generated against two synthetic peptides corresponding to sequences from the DNA-binding domain of steroid receptors. The sequence for peptide 1 (13 amino acids) lies between the two putative metal-binding loops of the conserved cysteine region while the sequence for peptide 2 (12 amino acids) lies within one loop. Peptide antibodies were generated by injecting rabbits with peptide conjugated to bovine serum albumin. By Western blot analysis, antibodies to peptide 2 recognized chick and human progesterone receptor and human glucocorticoid receptor, but peptide 1 antibodies did not. No cross-reactivity with native chick progesterone receptor was detected with either anti-peptide. These findings suggest that the epitopes for peptide 2 antibodies, and possibly for peptide 1 antibodies, are inaccessible to antibody in the native receptor.