We have discovered a protein family termed Regulators of G-protein Signaling (RGS) that impair signal transduction through pathways that use seven trans- membrane receptors and heterotrimeric G proteins. Such receptors, when activated following the binding of a ligand such as a hormone or chemokine, trigger the G alpha subunit to exchange GTP for GDP; this causes the dissociation of G alpha and G beta-gamma subunits and downstream signaling. RGS proteins bind G alpha subunits and function as GTPase activating proteins (GAPs), thereby deactivating the G alpha subunit and facilitating their re-association with G beta-gamma. We have shown that RGS proteins modulate signaling through a variety of G-protein coupled receptors including chemokine receptors. RGS1 over-expressing B lymphocytes fail to migrate in response to the chemokine CXCL12. Conversely, Rgs1 -/- B cells obtained from mice in which the Rgs1 gene has been disrupted by gene targeting have an enhanced chemotaxic response to CXCL12 and fail to desensitize properly following exposure to chemokines. Furthermore, B cells from these mice enter into lymph nodes more easily, target better into lymph node follicles, and move more rapidly than do B cells from wild type mice. Likely as a consequence the Rgs1 -/- mice have impaired immune responses, altered lymphoid tissue architecture, an excessive germinal center response, and improper trafficking of plasma cells. We have also demonstrated that germinal center B lymphocytes and thymic epithelial cells strongly express another RGS protein, RGS13. To study the role of RGS13 as well as other RGS proteins we have developed constructs that express shRNAs to knock-down RGS1, RGS2, RGS3, RGS10, RGS13, RGS14, or RGS16 mRNA expression. Introduction of a shRGS13 construct into human B cell lines reduces RGS13 mRNA expression and enhances responses to CXCL13 and to CXCL12. Reduction of RGS1 expression in the same B cell line that expresses low levels of RGS1, only mildly increases responses to CXCL13 and CXCL12. However, reduction of both RGS1 and RGS13 dramatically augments the responses to CXCL12 and CXCL13. In addition, the double knock-down cells show an impaired ability to properly polarize following chemokine exposure and an inability to properly orient the leading edge of the cell. To complement these studies we have recently begun to examine immune function in mice in which the Rgs10 or the Rgs13 gene has been disrupted. While RGS proteins attenuate heterotrimeric G-protein signaling the loss of G alpha i subunits would be expected to dramatically alter chemokine receptor signaling. Pertussis toxin, which inactivates all three G alpha i subunits blocks lymphocyte responses to chemokine stimulation. Lymphocytes predominately express nearly equivalent amounts of G alpha i2 and G alpha i3. Surprisingly mice, which lack G alpha i2 (Gnai2-/-) yet express G alpha i3 have a major defect in chemokine receptor signaling. These mice have a thymocyte egress defect, and defective lymph node and Peyer?s patch development. Lymphocytes isolated from these mice respond very poorly to chemokines and home poorly to the spleen and lymph nodes in adoptive transfer experiments. In vivo imaging of adoptively transferred T and B cells from these mice revealed very poor adhesion to high endothelial venules (HEVs) and a marked reduction in their velocities within lymph nodes. These studies in conjunction with the studies of the Rgs1-/- mice suggest that the ratio between Gnai2 and Rgs1 plays a crucial role in the responsiveness of lymphocytes to chemokine signaling. Another RGS protein highly expressed in vascular smooth muscle, RGS5, acts as a potent GTPase activating protein for G alpha i and G alpha q and attenuates signaling triggered by angiotensin II, endothelin-1, and sphingosine-1-phosphate. To confirm the physiologic importance of RGS5, mice in which the RGS5 gene has been disrupted have been developed. These mice are viable, but significantly underweight versus controls. Preliminary analysis suggests that these mice are also hypotensive. Another RGS protein, RGS3 undergoes extensive mRNA splicing. One of the splice variants termed PDZ-RGS3 is widely expressed. A combination of confocal and video time-lapse microscopy revealed that cells overexpressing a PDZ-RGS3 GFP fusion protein failed to establish a functional midbody. The PDZ-RGS3 GFP fusion protein localized at the midbody during the late stages of the cell cycle. Furthermore, we identified an shRNA construct that reduced PDZ-RGS3 expression and its expression results in a similar phenotype. In addition we have found that PDZ-RGS3 co-immunoprecipitates with the Aurora B kinase, a kinase known to be involved in cytokinesis. We have produced mice with a disrupted Rgs3 allele. To date we have not identified any viable Rgs3-/- mice. Studies of embryos obtained from interrupted pregnancies indicate that the mice are dying around day 10 of gestation. To date pathological studies have not discovered the reason for the embryonic lethality. RGS14, a larger member of the RGS family, contains an RGS, Rap-interacting, and GoLoco domain. RGS14 targeting is known to cause very early embryonic lethality. Using RGS14-specific antibodies we found that RGS14 co-localized with a centrosome marker, gamma-tubulin in centrosomes. Further studies revealed that RGS14 is a nuclear-cytoplasmic shuttling protein. Reduction in RGS14 expression results in a decrease in microtubules and decreases in cell viability. To complement our in vitro studies we have begun a conditional gene targeting project, which should allow us to study the function of RGS14 in adult lymphocytes. We also have begun several studies to examine the expression of other proteins potentially involved in heterotrimeric G-protein signaling, but not associated with signaling through seven transmembrane receptors. These include certain G alpha subunits; RIC-8, a guanine nucleotide exchange factor for G alpha subunits; AGS4, a protein that contains 3 GoLoco domains, and G beta5. Initial immunoblotting experiments and/or RNA expression studies has shown that each of these proteins in well expressed in lymphocytes. To further facilitate our studies of B cell migration we have developed new imaging tools that allow us to study B cell migration and the interaction of B cells and dendritic cells in more detail. As a model of B cell-DC interactions we examined B cells (TgB) from hen egg lysozyme (HEL) transgenic mice and spleen-derived DCs pulsed with HEL (DC-HEL) in 3-dimensional collagen matrices. Analysis of the live-cell dynamics revealed autonomous movements and random encounters between TgB cells and DC-HEL best described by a ?kiss-run and engage? model that led to formation of micro- and macro-complexes. Antigen localized at contact sites between TgB cells and DC-HEL. Thus, B cells productively interact with DCs displaying their cognate antigen to form a stable microenvironment similar to the immune synapse between T cells and DC. We have also tested a number of specific inhibitors of signaling molecules on B-lymphocyte chemotaxis. These studies have revealed potential roles for PI-3 kinase, P38 kinase, and BTK kinase in B cell migration. Treatment of either mouse B cells or human B cells with inhibitors of each of theses kinases potently inhibits B cell chemotaxis and in vivo homing to lymph nodes. The PI-3 kinase inhibitor markedly reduces B cell sticking to high endothelial venules (HEVs) while the other two inhibitors mildly affect B cell adhesion.
Showing the most recent 10 out of 39 publications
