The alpha-mating pheromone receptor encoded by the STE2 gene of the yeast Saccharomyces cerevisiae is a G protein-coupled receptor (GPCR) that is homologous to the large family of GPCRs that mediate multiple types of signal transduction in mammals. We have screened libraries of mutant receptors to identify dominant negative alleles that are capable of interfering with the function of a co-expressed normal receptor. Two dominant negative alleles have been recovered in this manner. In addition, we find that previously isolated loss-of-function mutations in the alpha-factor receptor exhibit dominant negative effects. Detection of the dominant effects requires high-level expression of the mutant receptors but does not require a high ratio of mutant to normal receptors. Cellular levels of the normal receptors are not affected by co-expression of the dominant negative alleles. Expression of the mutant receptors does not interfere with constitutive signaling in a strain that lacks the G protein alpha subunit encoded by GPA1, indicating that interference with signaling occurs at the level of the receptor or the interacting G protein. Expression of increased levels of G protein subunits partially reverses the dominant negative effects. The dominant negative behavior of the mutant receptors is diminished by deletion of the SST2 gene, which encodes an RGS (Regulator of G protein Signaling) protein involved in desensitization of pheromone signaling. The most likely explanation for the dominant negative effects of the mutations appears to be the existence of an interaction between unactivated receptors and the trimeric G protein that titrates the G protein away from the normal receptors or renders the G protein insensitive to receptor activation. This interaction appears to be mediated by the SST2 gene product.

The alpha-factor receptor of the yeast Saccharomyces cerevisiae is a member of the superfamily of G protein-coupled receptors that mediate signal transduction in response to sensory and chemical stimuli. All members of this superfamily contain seven predicted transmembrane segments. We have created a series of genes encoding alpha-factor receptors with amino- or carboxyl-terminal truncations at each of the loop regions connecting transmembrane segments. Split receptors containing a discontinuity in the peptide backbone were synthesized by coexpressing pairs of truncated receptor fragments in yeast. Complementary pairs of fragments split at sites within each of the cytoplasmic and extracellular loops were capable of assembling and transducing a signal in response to alpha-factor binding. One pair of noncomplementary fragments containing a deletion in the second intracellular loop of the receptor also yielded a functional receptor. Coexpression of certain combinations of overlapping fragments containing supernumerary transmembrane segments also led to formation of functional receptors, apparently because of proteolytic trimming of overlapping regions. Coexpression of truncated receptor fragments with full-length receptors had no effect on signaling by the full-length receptors. These results demonstrate the following: (1) Correct folding of the alpha-factor receptor does not require a covalent connection between any pair of transmembrane segments that are adjacent in the sequence. (2) Most of the second intracellular loop of the receptor is not required for function. (3) The structure of the receptor cannot, in most cases, tolerate the presence of extra transmembrane segments. (4) None of the truncated fragments of the alpha-factor receptor can efficiently oligomerize with normal receptors in such a way as to inhibit receptor function.

The ␣6␤4 integrin is a laminin 332 (LN332) receptor central to the formation of hemidesmosomes in epithelial layers. However, the integrin becomes phosphorylated by keratinocytes responding to epidermal growth factor in skin wounds or by squamous cell carcinomas that overexpress/hyperactivate the tyrosine kinase ErbB2, epidermal growth factor receptor, or c-Met. We show here that the ␤4-dependent signaling in A431 human squamous carcinoma cells is dependent on the syndecan family of matrix receptors. Yeast two-hybrid analysis identifies an interaction within the distal third (amino acids 1473-1752) of the ␤4 cytoplasmic domain and the conserved C2 region of the syndecan cytoplasmic domain. Via its C2 region, Sdc1 forms a complex with the ␣6␤4 integrin along with the receptor tyrosine kinase ErbB2 and the cytoplasmic kinase Fyn in A431 cells. Engagement of LN332 or clustering of the ␣6␤4 integrin with integrin-specific antibodies causes phosphorylation of ErbB2, Fyn, and the ␤4 subunit as well as activation of phosphatidylinositol 3-kinase and Akt and their assimilation into this complex. This leads to phosphatidylinositol 3-kinase-dependent cell spreading and Akt-dependent protection from apoptosis. This is disrupted by RNA interference silencing of Sdc1 but can be rescued by mouse Sdc1 or Sdc4 but not by syndecan mutants lacking their C-terminal C2 region. This disruption does not prevent the phosphorylation of ErbB2 or Fyn but blocks the Fynmediated phosphorylation of the ␤4 tail. We propose that syndecans engage the distal region of the ␤4 cytoplasmic domain and bring it to the plasma membrane, where it can be acted upon by Src family kinases.The ␣6␤4 integrin is a laminin 332 (LN332, 2 also known as LN5 or kalinin) receptor that forms hemidesmosomes in epithelial cells (reviewed in Refs. 1-4). It engages LN332 linked to collagen VII anchoring fibrils in the extracellular matrix (5) and simultaneously engages cytoplasmic proteins (e.g. plectin and BP230) and the transmembrane protein BP180 via the long (ϳ1,000-amino acid) cytoplasmic domain of the ␤4 integrin subunit. These cytoplasmic interactions involve two pairs of fibronectin type III (FNIII) repeats in the ␤4 tail and the connecting segment joining these pairs. This couples the integrin to the intermediate filament cytoskeleton and provides a stable anchorage that resists frictional forces on the epithelium.In contrast to this stabilizing role of the ␣6␤4 integrin, phosphorylation of the ␤4 cytoplasmic domain causes hemidesmosome disassembly and activation of ␣6␤4 signaling. Skin wounding causes relocalization of the integrin to lamellipodia of invading keratinocytes in response to EGF or macrophagestimulating factor (6). In tumor cells, overexpression of the integrin or overexpression and/or hyperactivation of the receptor tyrosine kinase c-Met (hepatocyte growth factor receptor), ErbB1 (EGFR), or ErbB2 causes phosphorylation of the integrin and promotes the proliferation, survival, and invasion of the tumor cells (7-9). The sites targeted by these kin...