GEFs: structural basis for their activation of small GTP-binding proteins.

Small GTP-binding proteins of the Ras superfamily function as molecular switches in fundamental events such as signal transduction, cytoskeleton dynamics and intracellular trafficking. Guanine-nucleotide-exchange factors (GEFs) positively regulate these GTP-binding proteins in response to a variety of signals. GEFs catalyze the dissociation of GDP from the inactive GTP-binding proteins. GTP can then bind and induce structural changes that allow interaction with effectors. Representative structures of four main classes of exchange factors have been described recently and, in two cases, structures of the GTP-binding protein-GEF complex have been solved. These structures, together with biochemical studies, have allowed a deeper understanding of the mechanisms of activation of Ras-like GTP-binding proteins and suggested how they might represent targets for therapeutic intervention.

Crystal structures of GMP kinase in complex with ganciclovir monophosphate and Ap5G.

Guanosine monophosphate kinases (GMPK), by catalyzing the phosphorylation of GMP or dGMP, are of dual potential in assisting the activation of anti-viral prodrugs or as candidates for antibiotic strategies. Human GMPK is an obligate step for the activation of acyclic guanosine analogs, such as ganciclovir, which necessitate efficient phosphorylation, while GMPK from bacterial pathogens, in which this enzyme is essential, are potential targets for therapeutic inhibition. Here we analyze these two aspects of GMPK activity with the crystal structures of Escherichia coli GMPK in complex with ganciclovir-monophosphate (GCV-MP) and with a bi-substrate inhibitor, Ap5G. GCV-MP binds as GMP to the GMP-binding domain, which is identical in E. coli and human GMPKs, but unlike the natural substrate fails to stabilize the closed, catalytically-competent conformation of this domain. Comparison with GMP- and GDP-bound GMPK structures identifies the 2'hydroxyl of the ribose moiety as responsible for hooking the GMP-binding domain onto the CORE domain. Absence of this hydroxyl in GCV-MP impairs the stabilization of the active conformation, and explains why GCV-MP is phosphorylated less efficiently than GMP, but as efficiently as dGMP. In contrast, Ap5G is an efficient inhibitor of GMPK. The crystal structure shows that Ap5G locks an incompletely closed conformation of the enzyme, in which the adenine moiety is located outside its expected binding site. Instead, it binds at a subunit interface that is unique to the bacterial enzyme, which is in equilibrium between a dimeric and an hexameric form in solution. This suggests that inhibitors could be designed to bind at this interface such as to prevent nucleotide-induced domain closure. Altogether, these complexes point to domain motions as critical components to be evaluated in therapeutic strategies targeting NMP kinases, with opposite effects depending on whether efficient phosphorylation or inhibition is being sought after.

Crystallization and preliminary X-ray diffraction studies of the human erythrocyte bisphosphoglycerate mutase.

Bisphosphoglycerate mutase (EC 2.7.5.4) catalyzes the synthesis and breakdown of 2,3-diphosphoglycerate in red cells. The human enzyme, cloned and expressed in Escherichia coli has been crystallized in the rhombohedral space group R32 with a = b = c = 100.4 A and alpha = beta = gamma = 81.2 degrees. The asymmetric unit contains either a dimeric enzyme molecule, or a monomer.

Structural mimicry of DH domains by Arfaptin suggests a model for the recognition of Rac-GDP by its guanine nucleotide exchange factors.

Small G proteins cycle between an inactive form bound to GDP, and an active form bound to GTP. The two forms have different conformations and interact specifically with different partners, hence, the ability of G proteins to function as molecular switches. This view has been challenged by recent structural and biochemical studies of the Arfaptin/Por protein, which interacts equally well with the GDP- and GTP-bound forms of the G protein Rac. Here it is shown that the dimeric helical domain of Arfaptin superimposes with a monomeric helical domain from the Dbl homology domain of Tiam, a guanine nucleotide exchange factor (GEF) for Rac, in their respective complexes with Rac. This unexpected structural mimicry suggests that the Rac-GDP-Arfaptin complex resembles the low-affinity Rac-GDP-GEF complex that initiates the exchange reaction. This provides a model for the exchange mechanism where DH domains first dock onto Rac-GDP at the switch 2 before they undergo domain closure to catalyze GDP dissociation.

Structural consequences of a one atom mutation on aspartate transcarbamylase from E. coli.

Tyr-240 of the catalytic chain of aspartate transcarbamylase from E. coli has been substituted by Phe using site-directed mutagenesis. The regulatory mechanisms of the mutant enzyme have been shown to be slightly less effective than the wild-type enzyme. A study of the structural consequences of the mutation using solution X-ray scattering and computer simulations is reported here. No significant change from the wild-type enzyme is detectable in the quaternary structure. Simulations suggest that the only effect of the mutation is an increased mobility of the mutated side chain.

Cysteine-153 is required for redox regulation of pea chloroplast fructose-1,6-bisphosphatase.

Chloroplastic fructose-1,6-bisphosphatases are redox regulatory enzymes which are activated by the ferredoxin thioredoxin system via the reduction/isomerization of a critical disulfide bridge. All chloroplastic sequences contain seven cysteine residues, four of which are located in, or close to, an amino acid insertion region of approximately 17 amino acids. In order to gain more information on the nature of the regulatory site, five cysteine residues (Cys49, Cys153, Cys173, Cys178 and Cys190) have been modified individually into serine residues by site-directed mutagenesis. While mutations C173S and C178S strongly affected the redox regulatory properties of the enzyme, the most striking effect was observed with the C153S mutant which became permanently active and redox independent. On the other hand, the C190S mutant retained most of the properties of the wild-type enzyme (except that it could now also be partially activated by the NADPH/NTR/thioredoxin h system). Finally, the C49S mutant is essentially identical to the wild-type enzyme. These results are discussed in the light of recent crystallographic data obtained on spinach FBPase [Villeret et al. (1995) Biochemistry 34, 4299-4306].

Modelling allosteric processes in E coli aspartate transcarbamylase.

The allosteric properties of aspartate transcarbamylase from E coli have been investigated by a combination of genetic, biochemical and structural studies. Based on the X-ray structures of the enzyme in T and R state established by Lipscomb et al, we have analyzed the interactions between the 12 polypeptide chains and have identified subunit interfaces that play a major part in the allosteric mechanism: the c1c4 interface between the 2 catalytic trimers, and one of 2 different interfaces between catalytic and regulatory chains, the c1r4 interface, which exists only in T state. We have modelled mutations affecting these interfaces: mutation pAR5 in the gene coding for r chains concerns the c1r4 interface, mutation Tyr----Phe 240 in the gene coding for c chains, the c1c4 interface. Both mutant proteins have reduced cooperativity and/or allosteric regulation by CTP and ATP. Molecular mechanic simulations lead to specific proposals for the structural origin of these effects, and some of the proposals can be checked by site-directed mutagenesis. Finally, we have modelled substrates bound at the active site of the T state, which binds aspartate less tightly than the R state and for which X-ray structures of bound substrate analogs were not available.

On the action of Brefeldin A on Sec7-stimulated membrane-recruitment and GDP/GTP exchange of Arf proteins.

Arf (ADP-ribosylation factor) proteins form a special class of small GTP-binding proteins in that their activation by GDP/GTP exchange is coupled to their recruitment to membranes using a built-in structural mechanism. These coupled processes are stimulated by GEFs (guanine nucleotide-exchange factors) that carry a catalytic Sec7 domain, whose basic mechanism has been uncovered by biochemical and structural studies. Crystal structures of intermediates of the GDP/GTP exchange reaction, from which GDP has not dissociated, notably allowed a movie of the exchange reaction to be reconstituted. They showed that Sec7 domains secure Arf-GDP to membranes before they proceed to nucleotide dissociation, and thus are active participants to the coupling of membrane-recruitment to nucleotide exchange. The drug BFA (Brefeldin A) was used to trap the complex that initiates the exchange reaction, providing a structural basis for its inhibition of Arf and its action on the membrane-recruitment of isolated Sec7 domains. Based on the dissection of this basic mechanism, the survey of reported BFA effects in cells on large multidomain ArfGEFs of the BIG1/2 and GBF1 families shows that the levels and compartmental distribution of BFA-induced recruitment of ArfGEFs to membranes cannot be explained from isolated Sec7 domains acting as independent domains. This leads to the hypothesis that Sec7 activity is inhibited in these ArfGEFs by an intramolecular interaction, which would be released by interaction with a compartment-specific receptor.

Arf, Sec7 and Brefeldin A: a model towards the therapeutic inhibition of guanine nucleotide-exchange factors.

GEFs (guanine nucleotide-exchange factors), which stimulate GDP dissociation from small G-proteins, are pivotal regulators of signalling pathways activated by small G-proteins. In the case of Arf proteins, which are major regulators of membrane traffic in the cell and have recently been found to be involved in an increasing number of human diseases, GDP/GTP exchange is stimulated by GEFs that carry a catalytic Sec7 domain. Recent structural results captured snapshots of the exchange reaction, revealing that Sec7 domains secure Arf-GDP to membranes before nucleotide exchange takes place, taking advantage of a built-in structural device in Arf proteins that couples their affinity for membranes to the nature of the bound nucleotide. One of the Arf-Sec7 intermediates was trapped by BFA (Brefeldin A), an uncompetitive inhibitor of Arf activation that has been instrumental in deciphering the molecular principles of membrane traffic at the Golgi. BFA targets a low-affinity Arf-Sec7 intermediate of the exchange reaction. It binds at the Arf-GDP/Sec7 interface, thus freezing the complex in an abortive conformation that cannot proceed to nucleotide dissociation. In the cell, this results in the specific inhibition of Arf1 by a subset of its GEFs, and the efficient and reversible block of membrane traffic at the Golgi. The mechanism of BFA leads to the concept of 'interfacial inhibition', in which a protein-protein interaction of therapeutic interest is stabilized, rather than impaired, by a drug. Up-regulated activity of small G-proteins is involved in various human diseases, making their GEFs attractive candidates to interrupt specifically the corresponding signalling pathway. Interfacial inhibitors are proposed as an alternative to competitive inhibitors that may be explored for their inhibition.

RhoGDI-3, a promising system to investigate the regulatory function of rhoGDIs: uncoupling of inhibitory and shuttling functions of rhoGDIs.

rhoGDIs (Rho GDP dissociation inhibitors) are postulated to regulate the activity and the localization of small G-proteins of the Rho family by a shuttling process involving extraction of Rho from donor membranes, formation of inhibitory cytosolic rhoGDI/Rho complexes, and delivery of Rho to target membranes. However, the role of rhoGDIs in site-specific membrane targeting or extraction of Rho is still poorly understood. We investigated here the in vivo functions of two mammalian rhoGDIs: the specific rhoGDI-3 and the well-studied rhoGDI-1 (rhoGDI) after structure-based mutagenesis. We identified two sites in rhoGDIs, forming conserved interactions with their Rho target, whose mutation results in the uncoupling of inhibitory and shuttling functions of rhoGDIs in vivo. Remarkably, these rhoGDI mutants were detected at Rho-induced membrane ruffles or protrusions, where they co-localized with RhoG or Cdc42, probably identifying for the first time the site of extraction of a Rho protein by a rhoGDI in vivo. We propose that these mutations act by modifying the steady-state kinetics of the shuttling process regulated by rhoGDIs, such that transient steps at the cell membranes now become detectable. They should provide valuable tools for future investigations of the dynamics of membrane extraction or delivery of Rho proteins and their regulation by cellular partners.

Structure of the small G protein Rap2 in a non-catalytic complex with GTP.

We report a novel crystal form of the small G protein Rap2A in complex with GTP which has no GTPase activity in the crystal. The asymmetric unit contains two complexes which show that a conserved switch I residue, Tyr 32, contributes an extra hydrogen bond to the gamma-phosphate of GTP as compared to related structures with GTP analogs. Since GTP is not hydrolyzed in the crystal, this interaction is unlikely to contribute to the intrinsic GTPase activity. The comparison of other G protein structures to the Rap2-GTP complex suggests that an equivalent interaction is likely to exist in their GTP form, whether unbound or bound to an effector. This interaction has to be released to allow the GAP-activated GTPase, and presumably the intrinsic GTPase activity as well. We also discuss the definition of the flexible regions and their hinges in the light of this structure and the expanding database of G protein structures. We propose that the switch I and switch II undergo either partial or complete disorder-to-order transitions according to their cellular status, thus defining a complex energy landscape comprising more than two conformational states. We observe in addition that the region connecting the switch I and switch II is flexible in Rap2 and other G proteins. This region may be important for protein-protein interactions and possibly behave as a conformational lever arm, as characterized for Arf. Taken together, these observations suggest that the structural mechanisms of small G proteins are significantly driven by entropy-based free energy changes.

Protein-protein recognition analyzed by docking simulation.

Antibody-lysozyme and protease-inhibitor complexes are reconstituted by docking lysozyme as a rigid body onto the combining site of the antibodies and the inhibitors onto the active site of the proteases. Simplified protein models with one sphere per residue are subjected to simulated annealing using a crude energy function where the attractive component is proportional to the interface area. The procedure finds clusters of orientations in which a steric fit between the two protein components is achieved over a large contact surface. With five out of six complexes, the native structure of the complexes determined by X-ray crystallography is among those retained. Docked complexes are then subjected to conformational energy refinement with full atomic detail. With Fab HyHEL 5 and lysozyme, a native-like complex has the lowest refined energy. It can also be retrieved when starting with the X-ray structure of free lysozyme. However, some non-native complexes cannot be rejected: they form large interfaces, have a large number of H-bonds, and few unpaired polar groups. While these are necessary features of protein-protein recognition, they are not sufficient in determining specificity.

Protein-protein interaction: an analysis by computer simulation.

A survey of protein-protein interactions in structures derived by X-ray crystallography of protease-inhibitor and antigen-antibody complexes shows that they form close-packed interfaces from which water is excluded. The interfaces are of almost constant size, and they contain about ten hydrogen bonds. These features account for the stability of the complexes. To test whether they also account for specificity, we designed a computer simulation that searches for complementary surfaces on two protein molecules. In all cases tested, the simulation finds a number of complexes having interfaces and hydrogen bonds equivalent to those of the native complexes. These artificial complexes might represent secondary specificities, which can be detected when normal association is prevented by mutation or other means.

AlF3 mimics the transition state of protein phosphorylation in the crystal structure of nucleoside diphosphate kinase and MgADP.

Nucleoside diphosphate kinase reversibly transfers the gamma-phosphate of ATP onto its active site histidine. We have investigated the transition state of histidine phosphorylation with the high-resolution crystal structures of the enzyme from Dictyostelium discoideum with MgADP and either aluminium or beryllium fluoride. The bound aluminium fluoride species is the neutral species AlF3 and not the more common AlF4-. AlF3 forms a trigonal bipyramid that makes it an accurate analog of the transition state of the gamma-phosphate of ATP undergoing transfer to the catalytic histidine. Its axial ligands are a histidine nitrogen and a beta-phosphate oxygen. Beryllium fluoride also binds at the same position and with the same ligands but in a tetrahedral geometry resembling the Michaelis complex rather than the transition state. The two x-ray structures show explicit enzyme-substrate interactions that discriminate between the ground and the transition states of the reaction. They also illustrate the partially dissociative geometry of the transition state of phosphoryl transfer and demonstrate the potential applications of metallofluorides for the study of kinase mechanisms.

The structural GDP/GTP cycle of human Arf6.

The small GTP-binding protein Arf6 coordinates membrane traffic at the plasma membrane with aspects of cytoskeleton organization. This function does not overlap with that of other members of the ADP-ribosylation factor (Arf) family, although their switch regions, which are their major sites of interaction with regulators and effectors, have virtually identical sequences. Here we report the crystal structure of full-length, non-myristoylated human Arf6 bound to GTPgammaS. Unlike their GDP-bound forms, the active forms of Arf6 and Arf1 are very similar. Thus, the switch regions are discriminatory elements between Arf isoforms in their inactive but not in their active forms, a property that may generalize to other families of small G proteins. This suggests that GTP-bound Arfs may establish specific interactions outside the switch regions and/or be recognized in their cellular context rather than as isolated proteins. The structure also allows further insight into the lack of spontaneous GTPase activity of Arf proteins.

Rigid-body docking with mutant constraints of influenza hemagglutinin with antibody HC19.

An automatic docking algorithm has been applied to the modeling of the complex between hemagglutinin from influenza virus and the Fab fragment of a monoclonal antibody raised against this antigen. We have introduced here the use of biochemical information provided by mutants of hemagglutinin. The docking procedure finds a small number of candidate solutions where three sites of escape mutations are buried and form hydrogen bonds in the interface. The localization of the epitope is improved by additional biochemical data about mutants that do not affect antibody binding. Five candidate solutions with low energy, reasonably well-packed interfaces, and six to ten hydrogen bonds are compatible with mutant information. One of the five stands out as generally better than the others from these points of views.

The pAR5 mutation and the allosteric mechanism of Escherichia coli aspartate carbamoyltransferase.

Mutation pAR5 replaces residues 145'-153' at the C terminus of the regulatory (r) chains of Escherichia coli ATCase by a new sequence of six residues. The mutated enzyme has been shown to lack substrate cooperativity and inhibition by CTP. Solution X-ray scattering curves demonstrate that, in the absence of ligands, its structure is intermediate between the T form and the R form. In the presence of N-phosphonacetyl-L-aspartate, the mutant is similar to the wild type. An examination of the crystal structure of unligated ATCase reveals that the mutated site is at an interface between r and catalytic (c) chains, which exists only in the T allosteric form. A computer simulation by energy minimization suggests that the pAR5 mutation destabilizes this interface and induces minor changes in the tertiary structure of r chains. The resulting lower stability of the T form explains the loss of substrate cooperativity. The lack of allosteric inhibition may be related to a new electrostatic interaction made in mutant r chains between the C-terminal carboxylate and a lysine residue of the allosteric domain.

Model-building of Fnr and FixK DNA-binding domains suggests a basis for specific DNA recognition.

The DNA-binding C-terminal domains of the regulatory proteins Fnr from Escherichia coli and FixK from Rhizobium meliloti have been modelled on the basis of their homologies to the CAP protein from E. coli. Residues Glu181, Thr182 and Arg185 of CAP, which are exposed residues of the DNA-recognition helix alpha F, are conserved in Fnr and FixK. However, Arg180 and Gly184 are substituted by Val and Ser respectively in Fnr. We propose that this valine makes a Van der Waals' contact with the first thymine in the Fnr consensus TTGA-N6-TCAA, and that the serine contributes to the binding by displacing a thymine-bound water molecule. The corresponding residues in FixK, Ile and Ser allow the same interactions with a thymine. Therefore we predict that FixK may recognize the same sites as Fnr. This is supported experimentally by showing that Fnr can substitute for FixK in activating the fixN gene in E. coli.