PerspectiveSignal Transduction

A New Thread in an Intricate Web

See allHide authors and affiliations

Science  30 Nov 2001:
Vol. 294, Issue 5548, pp. 1845-1847
DOI: 10.1126/science.1067418

Most neurotransmitters, hormones, and growth factors activate cellular signaling pathways by binding to specific membrane receptors. These signaling pathways subsequently modulate biochemical networks comprising various cytoplasmic kinases, phosphatases, and guanosine triphosphate (GTP)-binding proteins. Information flow through these complex signaling networks requires a precise spatiotemporal organization of the appropriate signaling partners, many of which interact promiscuously when isolated from their native cellular environment. Thus, a critical aspect of cellular signaling is a matter of molecular choreography: getting the right proteins to the right place at the right time. It is not surprising, then, that membrane-trafficking pathways—which determine the structure and biochemical composition of the specialized membrane compartments in eukaryotic cells—have important effects on cellular signal transduction and, conversely, that signaling events can modulate membrane trafficking. Indeed, there appear to be many functional interactions between the otherwise distinct processes of signaling and membrane trafficking. This realization, which emerged over the last decade (1), has motivated a convergence between traditionally separate fields of cell biology. From this convergence emanates the question: How are signaling and membrane trafficking related at the molecular level? On page 1939 of this issue, Zheng et al. (2) describe a protein, RGS-PX1, that may be a new molecular thread in the intricate web that links signaling and membrane trafficking events.

Zheng and colleagues mined sequence databases for candidate proteins containing RGS (regulators of G protein signaling) domains. RGS domains are conserved in diverse organisms and have profound effects on cellular signal transduction triggered by seven-transmembrane G protein-coupled receptors (GPCRs). GPCRs trigger signaling by prompting guanine nucleotide exchange on the α subunit of heterotrimeric G proteins. This results in conversion of the “inactive” guanosine diphosphate (GDP)-bound α subunit to the “activated” GTP-bound form. RGS proteins are crucial for accelerating the conversion of the activated G protein back to its inactive GDP-bound form by potentiating an intrinsic GTPase activity present in the α subunit (3) (see the figure). All known RGS proteins regulate Gi- or Gq-type heterotrimeric G proteins, whereas no RGS proteins characterized so far act on Gs. This is important because Gs mediates a spectrum of effects on cellular signaling (such as stimulating various isoforms of adenylyl cyclase) that are different from the effects of Gi- and Gq-type heterotrimeric G proteins. Zheng et al. demonstrate that the RGS domain in RGS-PX1 selectively promotes GTP hydrolysis on purified Gs and modulates Gs activity in intact cells, thus establishing RGS-PX1 as a member of the RGS protein family that regulates signaling through this heterotrimeric G protein.

Cycles within cycles.

(A) GPCRs promote binding of GTP to the α subunit of heterotrimeric G proteins. This leads to dissociation of the heterotrimer into the corresponding “activated” α subunit and βγ subcomplex. RGS proteins potentiate an intrinsic GTPase activity present in the α subunit that terminates the signaling activity of this subunit and promotes reassembly of the (inactive) heterotrimer. (B) The endosomal localization of RGS-PX1—which is dependent on binding of its PX domain to phosphatidylinositol 3-phosphate, PtdIns(3)P, in the endosome membrane—suggests that the biochemical cycle of Gs activation and inactivation may be linked to a physical cycle of α-subunit translocation between the plasma membrane and endosomes. The physical translocation of Gs is linked to a cycle of regulated depalmitoylation and repalmitoylation of the α subunit. (C) The proposed cycling of Gs α subunits between the plasma membrane and RGS-PX1-associated endosomes parallels a cycle of ligand-regulated endocytosis and recycling of many GPCRs. This suggests that endosomes may link the activities of the receptor and the G protein. One possible consequence of this linkage would be to physically organize distinct components of signal transduction triggered by activated GPCRs between a “peripheral” component (mediated by activated Gs) and a “deeper” component (dependent on protein kinases associated with phosphorylated receptors in endosome-associated signaling complexes). Alternatively, linkage could promote reassembly of receptor-Gs complexes in the plasma membrane after termination of signaling, thereby helping to restore responsiveness to a subsequent round of receptor activation.

In addition to identifying the unique specificity of the RGS domain in RGS-PX1, Zheng et al. note sequences outside of the RGS domain that are homologous to another family of cytoplasmic proteins called sorting nexins. Sorting nexins contain PX domains, a conserved protein module that binds to phosphatidylinositol 3-phosphate. This phospholipid is highly concentrated in endosomal membranes, and the interaction of PX domains with phosphatidylinositol 3-phosphate modulates the association of sorting nexins with endosomes (4). Some sorting nexins interact directly with endocytosed receptors, such as receptor tyrosine kinases activated by epidermal growth factor (EGF) (5), but probably they also have more general (although poorly defined) effects on endosomal traffic. Perhaps the best understood sorting nexin is the yeast protein Vps5p, which, together with Vps17p, Vps26p, Vps29p, and Vps35p, forms the “retromer,” a complex that retrieves proteins from the prevacuolar compartment (a type of endosome in yeast) and shuttles them back to the trans-Golgi network. Vps5p does not itself interact with specific membrane cargo, although other proteins in the retromer complex (such as Vps35p) do so (6).

Zheng et al. demonstrate that overexpression of RGS-PX1 in mammalian cells inhibits ligand-induced proteolysis of coexpressed EGF receptors. This finding suggests that RGS-PX1 inhibits the endocytic sorting of this receptor tyrosine kinase to lysosomes. Thus, RGS-PX1 may have an opposite effect to that of SNX1, a sorting nexin that enhances endocytic trafficking to lysosomes (5). Nevertheless, RGS-PX1 is associated with endosomal membranes in transfected cells, supporting the idea that this protein is a sorting nexin. Although RGS-PX1 has sequence homology with both sorting nexins (hence a previous designation as SNX13) and RGS domains (4), Zheng et al. make the critical observation that both domains in this protein are functional. They provide compelling evidence that these domains mediate, in intact cells, distinct functional effects on the signaling activity of a heterotrimeric G protein and the endocytic membrane trafficking of a receptor tyrosine kinase.

The idea of linking signaling and trafficking activities in the same protein may be a general theme in cell biology. For example, β-arrestins (nonvisual arrestins), a family of cytoplasmic proteins that interact with a variety of GPCRs after ligand-induced activation, are involved in both signaling and membrane trafficking (7). First, β-arrestins attenuate signal transduction by preventing coupling of receptors to heterotrimeric G proteins within seconds to minutes after initial receptor activation, a process often called rapid desensitization. Second, β-arrestins regulate membrane trafficking by promoting endocytosis (also called sequestration) of receptors via clathrin-coated pits. Third, sometimes β-arrestins appear to participate in a distinct signaling pathway activated by receptors that have been endocytosed; components of this signaling pathway include endosome-associated protein tyrosine kinases or mitogen-activated protein kinase cascades. The diverse biochemical effects of the same β-arrestin can be understood in terms of its capacity for coordinating the complex functional itinerary of the same GPCR. The effects of RGS-PX1 are more difficult to rationalize because this protein modulates pathways that do not appear to be closely linked in function. The existence of such a linkage between apparently disparate signaling and trafficking events is exciting because it challenges us to explore previously unanticipated possibilities.

A possible rationale for linking a Gs-specific RGS domain to a sorting nexin derives from the unusual biology of Gs. Inactive GDP-bound Gs is associated with the inner leaflet of the plasma membrane by a covalently attached palmitoyl moiety and by binding to membrane-anchored β and γ subunits. Upon activation by an appropriate GPCR, the GTP-bound α subunit of Gs dissociates from the βγ subcomplex; a fraction of the activated α subunit then becomes depalmitoylated and moves away from the plasma membrane and into the peripheral cytoplasm. In contrast, α subunits from other heterotrimeric G proteins contain a stably attached myristoyl modification and remain attached to membranes after activation (8). The association of a Gs-selective RGS protein with endosomal membranes could provide a way to limit the range (or duration) of signaling induced by liberated Gs α subunits to a cytoplasmic region delineated by RGS-PX1-associated endosomes (see the figure). By superimposing on this model of G protein cycling the itinerary of certain Gs-coupled GPCRs (such as the β2 adrenergic receptor), which cycle through endosomes after ligand-induced activation (9), one can envisage that localizing RGS activity could physically separate distinct signals. Thus, signals initiated by receptors at the plasma membrane and mediated by activation of Gs could be separated from signals emanating from internalized receptors and mediated by endosome-associated kinases (7). It is also conceivable that localized inactivation of α subunits near endosomes could facilitate regeneration of Gs heterotrimers in the vicinity of recycling GPCRs. This would help in the reassembly of receptor-Gs complexes in the plasma membrane (or perhaps on endosomes that later recycle to the plasma membrane), thereby rendering the receptor-Gs signaling system competent to be retriggered by a subsequent round of ligand-induced activation at the cell surface (see the figure).

Of course, these and other hypotheses about RGS-PX1 remain to be tested, and many questions relating to the structure and function of this Gs-selective RGS protein remain to be addressed. The Zheng et al. study should stimulate efforts to define additional molecular threads in the complex web linking signaling and membrane-trafficking events, and to investigate the functional significance of the connections thus revealed.


Navigate This Article