Posted by: Dan | June 20, 2007

Regulation by GEFs

In the same vein as my recent post on kinase structures and autoinhibition of multimodular signaling proteins, last week’s Nature had a post on Rewiring cellular morphology pathways with synthetic guanine nucleotide exchange factors (Yeh et al., 2007). Here, instead of kinases as the signaling regulators inside the cell, we’re talking about a class of proteins called guanine nucleotide exchange factors (GEFs). But before discussing Yeh et al (2007), I might answer the question, “Why does guanine need to be exchanged?,” first.

GTPases (“G” proteins) represent a family of hydrolase enzymes that can bind and hydrolyze guanine triphosphate (GTP) into the diphosphate form (GDP). The GTP-bound form is generally the “active” form, whereas GDP-bound G proteins are “inactive.” Thus, GTP hydrolysis turns the cell off, until the GDP dissociates from the G protein. This makes a GTPase a metaphorical switch inside the cell (in reality, this is just a conformational shift in the protein structure, which influences the interaction kinetics with other proteins – it’s not really a binary switch).

So the kinetics of GTPase signals are influenced by the rate of GTP hyrolysis, which is, in turn influenced by other proteins. GTPase-activating proteins (GAPs) accelerate the GTP hydrolysis, GDP dissociation inhibitors (GDIs) inhibit GDP dissociation, and GEFs accelerate the exchange of GDP for GTP.


Because the kinetics of this GTP hydrolysis impacts a number of very important cellular processes, like cell shape, motility, gene expression, and cell division, altering those kinetics for a given GTPase (and it’s regulatory modules) could have potent effects on cell behavior. Now, on to Yet et al. (2007)…

Yeh and colleagues examine the potential for GEFs of a particular subfamily of G proteins, the Rho family of monomeric GTPases. Rho GTPases include Rho, Rac and Cdc42, which all effect the cytoskeleton (and therefore cell shape) in various ways – facilitating stress fiber, lamellipodia and filopodia formation, respectively. Adding the GEFs specific for these proteins will exacerbate their normal functions, generating “super” stress fibers, lamellipodia and filopodia. It’s a relatively simple thing to do, but potentially a powerful way to manipulate cell behavior, and informative to understanding both signaling networks and degrees of evolutionary plasticity.

[Fig. 3A: Microinjection of constitutively active intersectin DH-PH (and Cdc42) induced filopodia in REF52 cells. Constitutely active Trio DH (and Rac1) induced lamellipodia.]

Abstract below the fold:

Eukaryotic cells mobilize the actin cytoskeleton to generate a remarkable diversity of morphological behaviours, including motility, phagocytosis and cytokinesis. Much of this diversity is mediated by guanine nucleotide exchange factors (GEFs) that activate Rho family GTPases-the master regulators of the actin cytoskeleton. There are over 80 Rho GEFs in the human genome (compared to only 22 genes for the Rho GTPases themselves), and the evolution of new and diverse GEFs is thought to provide a mechanism for linking the core cytoskeletal machinery to a wide range of new control inputs. Here we test this hypothesis and ask if we can systematically reprogramme cellular morphology by engineering synthetic GEF proteins. We focused on Dbl family Rho GEFs, which have a highly modular structure common to many signalling proteins: they contain a catalytic Dbl homology (DH) domain linked to diverse regulatory domains, many of which autoinhibit GEF activity. Here we show that by recombining catalytic GEF domains with new regulatory modules, we can generate synthetic GEFs that are activated by non-native inputs. We have used these synthetic GEFs to reprogramme cellular behaviour in diverse ways. The GEFs can be used to link specific cytoskeletal responses to normally unrelated upstream signalling pathways. In addition, multiple synthetic GEFs can be linked as components in series to form an artificial cascade with improved signal processing behaviour. These results show the high degree of evolutionary plasticity of this important family of modular signalling proteins, and indicate that it may be possible to use synthetic biology approaches to manipulate the complex spatio-temporal control of cell morphology.

  • Yeh BJ, Rutigliano RJ, Deb A, Bar-Sagi D, Lim WA. Rewiring cellular morphology pathways with synthetic guanine nucleotide exchange factors. Nature. 2007 May 31;447(7144):596-600.


  1. Good article. I just have one question.

    In GST pulldown assays where the RBD of a GEF is chemically linked to GST, are we pulling down GDP loaded GTPases or GTP loaded GTPAse? If GEF catalyzes the replacement of GDP by GTP, then what we are pulling down must be the former?

    Thanks for your help.

  2. Hi Cecille,
    You’d probably want to check that experimentally, but I’d assume that any bound GTP would be hydrolyzed to GDP, and that any GEF activity would be negated by the fact that outside of the cell, fresh GTP is unavailable.

    This might be negated by supplementing the pulldown buffers with GTP, or even better, by supplementing with nonhydrolyzable GTP (GTPgS) before or after cell lysis.


%d bloggers like this: