Spatial cycles in G-protein crowd control

Abstract

The nature of living systems and their apparent resilience to the second law of thermodynamics has been the subject of extensive investigation and imaginative speculation. The segregation and compartmentalization of proteins is one manifestation of this departure from equilibrium conditions; the effect of which is now beginning to be elucidated. This should not come as a surprise, as even a cursory inspection of cellular processes reveals the large amount of energetic cost borne to maintain cell-scale patterns, separations and gradients of molecules. The G-proteins, kinases, calcium-responsive proteins have all been shown to contain reaction cycles that are inherently coupled to their signalling activities. G-proteins represent an important and diverse toolset used by cells to generate cellular asymmetries. Many small G-proteins in particular, are dynamically acylated to modify their membrane affinities, or localized in an activity-dependent manner, thus manipulating the mobility modes of these proteins beyond pure diffusion and leading to finely tuned steady state partitioning into cellular membranes. The rates of exchange of small G-proteins over various compartments, as well as their steady state distributions enrich and diversify the landscape of possibilities that GTPase-dependent signalling networks can display over cellular dimensions. The chemical manipulation of spatial cycles represents a new approach for the modulation of cellular signalling with potential therapeutic benefits.

Figure 1

Schematic representation of Rab5–Rab7 conversion on endocytic vesicles (coloured spheres). Rab5GEFs on donor membranes activate Rab5 and the activated GTP-loaded Rab5 is recruited to the endocytic vesicular membrane. As PI(3)P levels on the vesicular membrane build up, Mon1b recruitment occurs (blue flash), thereby driving the activation and recruitment of Rab7 and inhibition of the Rab5 positive feedback. Rab5GAPs inactivate Rab5 by catalysing GTP hydrolysis and Rab5-GDIs sequester inactivated Rab5 into the cytosol, allowing its diffusion to the PM, in which it may once more be activated on newly formed vesicles. Similarly, as the vesicles reach the late endosomal structures, Rab7GAPs on these membranes inactivate Rab7, and Rab7GDIs sequester it into the cytosol, allowing it to diffuse to vesicles undergoing the Rab5–Rab7 conversion event. Inset: schematic representation showing time dependence of Rab5 and Rab7 activation (and thus vesicular recruitment) in relation to PI(3)P levels. The reaction progression of PI(3)P generation is correlated with the spatial coordinates of vesicles.

Figure 2

Schematic representations of Rab5–Rab7 conversion models. A Rab5–RabX5 positive feedback and the presence of a cargo-mediated signal through adaptor proteins, such as AP-1, are common features of all models. (A) Cut-out switch model: Rab5 is activated by a cargo-mediated signal. A negative feedback between Rab5 and Rab7 competes with, and eventually overrides the positive feedback between Rab5 and RabX5, thus facilitating Rab conversion (B) Toggle-switch model: cargo-activated Type III PI(3) kinases generate PI(3)P on the vesicular membrane and act as master controllers. Here, PI(3)P inhibits Rab5 and activates Rab7, leading to Rab conversion. (C) Transistor model: The cargo-mediated and Rab5-mediated activation of Type III PI(3) kinases generates PI(3)P on the vesicular membrane. Accumulation of PI(3)P recruits Mon1b, which inhibits RabX5, thereby interrupting the Rab5–RabX5 positive feedback. Mon1b simultaneously activates Rab7 through the HOPS complex. This dual effect of Mon1b causes Rab conversion. The Rab7GEF that maintains activation and enrichment of Rab7 is as yet unknown, but could be the HOPS complex itself.

Figure 3

Ran gradients in the two states of the cell cycle. (A) RanGEF activity (red) is localized on chromatin, generating activated RanGTP in its vicinity. As RanGAP activity is excluded from the nucleus in interphase cells, Ran shuttles between the nucleus and cytoplasm in its GTPase cycle. Interactions of RanGTP with karyopherins allow the Ran GTPase cycle to function as a motor driving vectorial nuclear transport of cargo using the free energy of GTP hydrolysis. (B) During mitosis, the nuclear envelope is lost, and RanGAP activity is homogenously spread throughout the mitotic cytosol. RanGEF remains chromatin bound, generating high local RanGTP concentrations. This RanGTP is converted to RanGDP as it diffuses away from chromatin and stochastically encounters homogeneously distributed RanGAPs. The Ran cycle thus forms a gradient-based positioning system around chromatin, providing cues for guidance and self-assembly of the spindle and its regulatory processes.

Figure 4

The H/N-Ras spatial cycle uses an acylation cycle and vectorial transport to maintain localization and counter entropy increase. (A) Golgi-confined palmitoylation of Ras and transport to the PM through the secretory pathway, account for enrichment of H/N-Ras at the Golgi and PM, respectively. This palmitoylated Ras will slowly redistribute to all membranes (leakage). Ubiquitous depalmitoylation by thioesterases, such as APT1, converts mislocalized palmitoylated Ras to the fast-diffusing solely farnesylated form. Diffusion increases the probability of a Golgi encounter, in which it is (re)-palmitoylated and transported to the PM through the secretory pathway. (B) Perturbation of the spatial cycle through thioesterase inhibition, leads to eventual leakage from the PM, and equilibrium binding of palmitoylated Ras on all cellular membranes, effectively diluting Ras on the plasma membrane and attenuating its signalling response.