The vasopressor angiotensin II regulates vascular contractility and blood pressure through binding to type 1 angiotensin II receptors (AT1; refs 1, 2). Bradykinin, a vasodepressor, is a functional antagonist of angiotensin II (ref. 3). The two hormone systems are interconnected by the angiotensin-converting enzyme, which releases angiotensin II from its precursor and inactivates the vasodepressor bradykinin. Here we show that the AT1 receptor and the bradykinin (B2) receptor also communicate directly with each other. They form stable heterodimers, causing increased activation of G alpha(q) and G alpha(i) proteins, the two major signalling proteins triggered by AT1. Furthermore, the endocytotic pathway of both receptors changed with heterodimerization. This is the first example of signal enhancement triggered by heterodimerization of two different vasoactive hormone receptors.

Several examples of functional G-protein-coupled receptor heterodimers have been identified. However, it is not known whether receptor heterodimerization is involved in the pathogenesis of human disorders. Here we show that in preeclamptic hypertensive women, a significant increase in heterodimerization occurs between the AT(1)-receptor for the vasopressor angiotensin II and the B(2)-receptor for the vasodepressor bradykinin. AT(1)-B(2)-receptor heterodimerization in preeclampsia correlated with a 4-5-fold increase in B(2)-receptor protein levels. Expression of the AT(1)-B(2) heterodimer increased the responsiveness to angiotensin II and conferred resistance in AT(1)-receptors to inactivation by reactive oxygen species raised in normotensive and preeclamptic pregnancies. We suggest that AT(1)-B(2) heterodimers contribute to angiotensin II hypersensitivity in preeclampsia. Moreover, we identify preeclampsia as the first disorder associated with altered G-protein-coupled receptor heterodimerization.

stimulate mainly G␣ q/11 and G␣ i/o proteins (1-5). By contrast, G-protein coupling, signaling cascades, and biological effects of AT 2 receptors are less defined. Many in vivo studies with genetic deletion or transgenic overexpression of AT 2 receptors show that the AT 2 receptor inhibits AT 1 receptor-mediated functions (6 -8). The inhibitory effect of the AT 2 receptor has often been attributed to the AT 2 -mediated activation of protein phosphatases (9). However, in several systems AT 2 receptor-mediated signal inhibition persists in the presence of AT 2 -specific antagonists (10, 11) suggesting that the AT 2 receptor mediates additional functions independent of an agonist-AT 2 interaction. We previously demonstrated that AT 1 receptors can form stable heterodimers with another receptor, the bradykinin B 2 receptor (12). This receptor heterodimerization is agonist-independent and modifies AT 1 receptor-mediated responses (12, 13). Therefore we asked whether AT 1 receptors formed heterodimers with AT 2 receptors, and whether AT 1 /AT 2 receptor heterodimerization was responsible for agonist-independent effects of the AT 2 receptor causing inhibition of AT 1 receptor stimulation.We report here that the AT 2 receptor antagonizes AT 1 receptor-mediated responses by direct binding to the AT 1 receptor. Heterodimerization of the AT 1 with the AT 2 receptor was detected on transfected cells expressing AT 1 and AT 2 receptors, on fetal fibroblasts, and on myometrial biopsies from nonpregnant and pregnant women. The AT 2 -mediated antagonism of the AT 1 receptor was released upon decrease of AT 2 receptor expression, or when AT 1 /AT 2 receptor heterodimerization was inhibited. Thus, the AT 2 receptor antagonizes the activation of the AT 1 receptor by direct binding, and is therefore the first identified example of a G-protein-coupled receptor which acts as a receptor-specific antagonist. EXPERIMENTAL PROCEDURESCell Culture and Transfections-PC-12 cells and rat fetal fibroblasts were cultured and transfected as described (13,14).Construction of AT 2 Receptor Mutants-Site-directed mutagenesis of the rat AT 2 receptor-cDNA was performed by polymerase chain reaction similarly as described previously (14).Functional Assays-Inositol phosphate levels of PC-12 cells and of fetal fibroblasts was determined as described previously (14). The cellular content of cAMP was determined by radioimmunoassay (Immunotech) in the presence of 100 M isobutylmethylxanthine (14).Receptor-stimulated Redistribution of G-proteins-The activation of different G-proteins in cells and in tissues was determined by the agonist-stimulated release of G␣ proteins into the cytosol and subsequent detection of the cytosolic G␣ i/o , G␣ q/11 , or G␣ s proteins in immunoblot similarly as described (12, 13). For the detection of cytosolic G␣, proteins were separated on 7.5% SDS-containing polyacrylamide gels under nonreducing conditions. As a control, the amount of G␣ subunits in the total cell lysates was detected in immunoblot after separation of the prote...

Many G protein-coupled receptors form dimers in cells. However, underlying mechanisms are barely understood. We report here that intracellular factor XIIIA transglutaminase crosslinks agonist-induced AT1 receptor homodimers via glutamine315 in the carboxyl-terminal tail of the AT1 receptor. The crosslinked dimers displayed enhanced signaling and desensitization in vitro and in vivo. Inhibition of angiotensin II release or of factor XIIIA activity prevented formation of crosslinked AT1 receptor dimers. In agreement with this finding, factor XIIIA-deficient individuals lacked crosslinked AT1 dimers. Elevated levels of crosslinked AT1 dimers were present on monocytes of patients with the common atherogenic risk factor hypertension and correlated with an enhanced angiotensin II-dependent monocyte adhesion to endothelial cells. Elevated levels of crosslinked AT1 receptor dimers on monocytes could sustain the process of atherogenesis, because inhibition of angiotensin II generation or of intracellular factor XIIIA activity suppressed the appearance of crosslinked AT1 receptors and symptoms of atherosclerosis in ApoE-deficient mice.

Only so-called "professional" antigen-presenting cells (APCs) of hematopoietic origin are believed capable of inducing T lymphocyte responses. However, fibroblasts transfected with viral proteins directly induced antiviral cytotoxic T lymphocyte responses in vivo, without involvement of host APCs. Fibroblasts induced T cells only in the milieu of lymphoid organs. Thus, antigen localization affects self-nonself discrimination and cell-based vaccine strategies.