Advertisement
JBC

HOME HELP FEEDBACK SUBSCRIPTIONS ARCHIVE SEARCH TABLE OF CONTENTS
QUICK SEARCH:   [advanced]


     

Originally published In Press as doi:10.1074/jbc.M600149200 on February 16, 2006

J. Biol. Chem., Vol. 281, Issue 17, 11805-11814, April 28, 2006
This Article
Right arrow Abstract Freely available
Right arrow Full Text (PDF)
Right arrow All Versions of this Article:
281/17/11805    most recent
M600149200v1
Right arrow Submit a Letter to Editor
Right arrow Alert me when this article is cited
Right arrow Alert me when eLetters are posted
Right arrow Alert me if a correction is posted
Right arrow Citation Map
Services
Right arrow Email this article to a friend
Right arrow Similar articles in this journal
Right arrow Similar articles in PubMed
Right arrow Alert me to new issues of the journal
Right arrow Download to citation manager
Right arrowRequest Permissions
Citing Articles
Right arrow Citing Articles via HighWire
Right arrow Citing Articles via Google Scholar
Google Scholar
Right arrow Articles by Zeeh, J.-C.
Right arrow Articles by Cherfils, J.
Right arrow Search for Related Content
PubMed
Right arrow PubMed Citation
Right arrow Articles by Zeeh, J.-C.
Right arrow Articles by Cherfils, J.
Social Bookmarking
Add to CiteULike   Add to Complore   Add to Connotea   Add to Del.icio.us   Add to Digg   Add to Reddit   Add to Technorati  
What's this?

Dual Specificity of the Interfacial Inhibitor Brefeldin A for Arf Proteins and Sec7 Domains*

Jean-Christophe Zeeh{ddagger}12, Mahel Zeghouf{ddagger}2, Cedric Grauffel§3, Bernard Guibert{ddagger}, Elyette Martin4, Annick Dejaegere§5, and Jacqueline Cherfils{ddagger}6

From the {ddagger}Laboratoire d'Enzymologie et Biochimie Structurales, CNRS, Avenue de la Terrasse, 91198 Gif sur Yvette, France and the §Biocomputing Group, Structural Biology and Genomics Department, Institut de Génétique et de Biologie Moléculaire et Cellulaire, BP 10142, F-67404 Illkirch Cedex, France

Received for publication, January 6, 2006 , and in revised form, February 7, 2006.


   ABSTRACT
TOP
 ABSTRACT
INTRODUCTION
 EXPERIMENTAL PROCEDURES
 RESULTS
 DISCUSSION
 REFERENCES
 
Guanine nucleotide exchange factors (GEFs), which activate small GTP-binding proteins (SMG) by stimulating their GDP/GTP exchange, are emerging as candidate targets for the inhibition of cellular pathways involved in diseases. However, their specific inhibition by competitive inhibitors is challenging, because GEF and SMG families comprise highly similar members. Nature shows us an alternative strategy called interfacial inhibition, exemplified by Brefeldin A (BFA). BFA inhibits the activation of Arf1 by its GEFs in vivo by stabilizing an abortive complex between Arf-GDP and the catalytic Sec7 domain of some of its GEFs. Here we characterize the specificity of BFA toward wild-type (ARNO and BIG1) and mutant Sec7 domains and toward class I, II, and III Arfs. We find that BFA sensitivity of the exchange reaction depends on the nature of both the Sec7 domain and the Arf protein. A single Phe/Tyr substitution is sufficient to achieve BFA sensitivity of the Sec7 domain, which is supported by our characterization of brefeldin C (BFC), a BFA analog that cannot interact with the Tyr residue, and by free energy computations. We further show that Arf1 and Arf5, but not Arf6, are BFA-sensitive, despite their having every BFA-interacting residue in common. Analysis of Arf6 mutants points to the dynamics of the interswitch, which is involved in membrane-to-nucleotide signal propagation, as contributing to, although not sufficient for, BFA sensitivity. Altogether, our results reveal the Tyr/Phe substitution as a novel tool for monitoring BFA sensitivity of cellular ArfGEFs and document the exquisite and dual specificity that can be achieved by an interfacial inhibitor.


   INTRODUCTION
TOP
 ABSTRACT
 INTRODUCTION
EXPERIMENTAL PROCEDURES
 RESULTS
 DISCUSSION
 REFERENCES
 
Guanine nucleotide exchange factors (GEFs),7 which activate small GTP-binding proteins (SMGs) by stimulating their GDP/GTP exchange, are increasingly being characterized as critical determinants in SMG signaling specificity (reviewed in Refs. 1 and 2). They are therefore emerging as candidate targets for the specific inhibition of SMG-regulated pathways involved in diseases (reviewed in Refs. 3 and 4). However, both SMGs and their GEFs exist as families within which members are highly similar, making the discovery of specific inhibitors a complex issue. Besides, SMGs can be activated by several GEFs depending on the nature of the triggering signals, and, conversely, many GEFs have been shown to be promiscuous for more than one SMG target. Competitive inhibitors of SMG/GEF interactions may therefore block multiple pathways, yielding detrimental cellular effects. Interfacial inhibitors, exemplified by the drug Brefeldin A (BFA), were recently proposed as an alternative to competitive inhibition (2, 5). Interfacial inhibitors are defined as inhibitors that target protein complexes that are undergoing a conformational change, taking advantage of the topology at the protein interface in the complex and of unbalanced energetic conditions resulting from the conformational changes (5). BFA is a paradigm of interfacial inhibition and a rare example of an inhibitor that is efficient at inhibiting the activation of an SMG in vivo (6-9). It blocks the activation of Arf, an essential regulator of membrane traffic at the Golgi (reviewed in Ref. 10), by its GEFs. BFA acts by an uncompetitive mechanism, trapping an early Arf·ArfGEF complex of the exchange reaction from which the GDP nucleotide has not yet dissociated (8, 11). Biochemical and structural studies have demonstrated that the drug binds in a cavity at the interface between Arf-GDP and the catalytic domain of the GEF (a conserved domain called the Sec7 domain), thus stabilizing a transient reaction intermediate (12-14).

There are five Arf isoforms in humans, which are grouped in three classes (Arf1/3 in class I, Arf4/5 in class II, and Arf6 in class III (15)), and over 12 ArfGEFs, which all carry the conserved catalytic Sec7 domain (reviewed in Refs. 16 and 17). However, not all Sec7 domains are sensitive to BFA. Sensitivity to the drug has been mapped to two sets of residues in the Arf-binding site, which are YS and/or DM in BFA-sensitive Sec7 domains, or FA and/or SP in BFA-insensitive Sec7 domains (8, 12, 13, 18). Much less is known, in contrast, about the contribution of Arf isoforms to the inhibition by BFA. Most reported effects so far involve inhibition of the activation of Arf1, a member of class I Arfs, explaining its effects on Golgi organization (6, 19, 20). BFA was reported to inhibit activation of Arf3 in an in vitro assay in which Golgi membranes and soluble factors are added, whereas Arf5 was not activated under these conditions and therefore not affected by BFA (21). More recently, decreasing the expression of both Arf1 and Arf4 by small interfering RNA, but not of either Arf1 or Arf4 alone, was shown to result in effects similar to those of BFA treatment, suggesting that both Arf1 and Arf4 are BFA targets in the cell (22). Finally, BFA has no effect on the function of Arf6 (the only class III Arf) at the plasma membrane (23, 24), and activation of Arf6 by exchange factors of the ARNO (Arf nucleotide-binding site opener) (25) or EFA6 (exchange factor for Arf6) (26) families is BFA-insensitive.

Here we analyze the respective contributions of Arf and ArfGEF proteins to BFA sensitivity using the Sec7 domain of ARNO, whose BFA sensitivity can be monitored by mutations, and Arf proteins from classes I, II, and III. Using biochemical and computational analysis, we map the contribution of the Sec7 domain to a unique amino acid in direct interaction with BFA, which is confirmed by a BFA analog in which interaction with this residue is impaired. We then find that Arf proteins contribute to the sensitivity to BFA, with Arf1 and Arf5 being BFA-sensitive and Arf6 being BFA-insensitive. Mutation of Arf6 residues located in the vicinity of the drug to their Arf1 counterparts points to the interswitch, a region that transmits conformational changes between the membrane and the nucleotide-binding site, as contributing to the Arf1/Arf6 differences toward BFA. Altogether, our results document the exquisite specificity that can be achieved by an interfacial inhibitor. They constitute a step toward the design of inhibitors of Arf pathways, whose deregulated activities have recently been associated with human diseases as diverse as cancer (27-29), viral (30, 31) and bacterial (32-35) infections, and neuronal diseases (36) (reviewed in Ref. 2).


   EXPERIMENTAL PROCEDURES
TOP
 ABSTRACT
 INTRODUCTION
 EXPERIMENTAL PROCEDURES
RESULTS
 DISCUSSION
 REFERENCES
 
Cloning, Protein Expression, and Purification—Point mutations were introduced using the QuikChangeTM site-directed mutagenesis kit (Stratagene, Amsterdam, The Netherlands). The F190Y (ARNO1M) and A191S/S198D/P208M (ARNO3M) mutants of the Sec7 domain of human ARNO (residues 52-242) were generated from the wild type Sec7 domain (ARNO) or a mutant carrying four BFA-sensitizing mutations (F190Y/A191S/S198D/P208M, called ARNO4M hereafter), respectively. The Sec7 domain of human BIG1 (residues 700-888, called BIG1 hereafter) was amplified by PCR and cloned in a modified pET-28a plasmid (Novagen), in which the KpnI and AgeI restriction sites were added for cloning purpose, to express an N-terminal His6 tag fusion protein (tag sequence MGHHHHHHGT). The cDNA encoding for the N-terminally truncated human Arf5 (residues 14-180, called [{Delta}14]Arf5 hereafter) was cloned in the modified pET-28a plasmid. Mutations K58S and K58S/N60T were introduced in human Arf6 truncated of the N-terminal helix (residues 13-175, called [{Delta}13]Arf6 hereafter). All constructs were confirmed by sequencing. [{Delta}17]Arf1, ARNO, and ARNO4M were a gift from P. Chardin, B. Antonny and S. Robineau (IPMC, CNRS, France); [{Delta}13]Arf6 and [{Delta}13]Arf6S38I were a gift from M. Franco (IPMC, CNRS, France); full-length human BIG1 and Arf5 were from P. Melançon (University of Alberta, Canada).

All proteins were overexpressed in BL21 Escherichia coli strains, except [{Delta}14]Arf5 and BIG1, which were expressed in Rosetta (DE3) pLysS strain (Novagen). All purification steps were performed at 4 °C. ARNO constructs were expressed and purified essentially as described in Ref. 37 except that cells were lysed by sonication, and protein purity was polished, after the ion exchange step on a Q-Sepharose fast flow column (Amersham Biosciences), by gel filtration on a Superdex 75 XK 16/90 column (Amersham Biosciences) equilibrated with a 50 mM Tris-HCl, pH 8, buffer containing 1 mM EDTA and 0.5 mM beta-mercaptoethanol. [{Delta}17]Arf1 and [{Delta}13]Arf6 proteins were expressed and purified as previously described in Refs. 37 and 38. BIG1 expression was induced by 0.5 mM of isopropyl 1-thio-beta-D-galactopyranoside for 4 h at37 °C. Cells were disrupted by sonication in buffer A (100 mM sodium phosphate buffer, pH 8, 300 mM NaCl, 2 mM dithiothreitol) supplemented with a protease inhibitor mixture (1 mg/ml each aprotinin, leupeptin, pepstatin, antipain, and 1 mM each 4-(2-aminoethyl)-benzenesulfonyl fluoride and benzamidine). DNase treatment (5 mg/ml) was then performed for 30 min in the presence of 3 mM MgCl2. Cell extracts were centrifuged at 300,000 x g for 2 h prior to being incubated with Ni2+-nitrilotriacetic acid superflow resin (Qiagen) for 1 h with slow rotation. The resin was then poured in an Econo Pack column (Bio-Rad) and washed with 20 column volumes, and the bound protein was eluted with buffer A containing 200 mM imidazol. BIG1 was further purified by gel filtration on a Superdex 75 XK 16/90 column equilibrated with 50 mM Hepes, pH 7.5, 30 mM NaCl, and 1 mM dithiothreitol. Expression of [{Delta}14]Arf5 was induced by 0.25 mM isopropyl 1-thio-beta-D-galactopyranoside for 5 h at 28 °C. Cells were lysed with 1 mg/ml lysozyme for 1 h in 50 mM Tris-HCl, pH 8, 1 mM EDTA, 0.5 mM beta-mercaptoethanol, 200 mM GDP supplemented with the protease inhibitor mixture. After DNase treatment, cell extracts were cleared by centrifugation for 15 min at 30,000 x g followed by 1 h at 300,000 x g prior to being applied on a Q-Sepharose fast flow column equilibrated with 50 mM Tris-HCl, pH 8, 50 µM GDP, 1 mM 4-(2-aminoethyl)-benzenesulfonyl fluoride, and 0.5 mM beta-mercaptoethanol. [{Delta}14]Arf5 was eluted with a linear NaCl gradient, precipitated at 60% saturation ammonium sulfate, and then further purified by gel filtration on a Superdex 75 XK 16/90 column equilibrated with 50 mM Tris-HCl, pH 8, 1 mM MgCl2, 100 mM NaCl, and 0.5 mM beta-mercaptoethanol. As judged by SDS-PAGE analysis, proteins were at least 90% pure. All proteins were concentrated to at least 20 mg/ml on a Vivaspin concentrator (molecular weight cut-off 5000; Vivascience AG) and stored at -80 °C.

Nucleotide Exchange and Inhibition Kinetic Assays—BFA was purchased from Fluka. Brefeldin C (BFC) was synthesized as described (39) and was a gift from S. Archambaut and A. Guingant (Faculté des Sciences et Techniques, Nantes, France). Stock solutions of BFA and BFC were prepared at 20 mM in ethanol. Spontaneous and exchange factor-stimulated GDP dissociation was monitored by tryptophan fluorescence. All fluorescence measurements were performed with an Aminco Bowman Serie 2 fluorimeter mainly as described in Ref. 40 using excitation and emission wavelengths of 292 and 340 nm, respectively. Briefly, protein solutions were incubated 5 min at 37 °C in 700 µl of reaction buffer (50 mM Tris-HCl, pH 8, 50 mM NaCl, 2 mM MgCl2,2mM beta-mercaptoethanol) with or without BFA or BFC. The concentration of Arf proteins was 1 µM in all assays. The concentration of Sec7 domains was 0.05 µM except for the determination of the apparent inhibition constant (Ki(app)) for all Arf6 constructs. Exchange reactions were initiated by injection of 100 µM GTP. The apparent activation rate constant of Arf proteins (kapp) was determined by fitting the fluorescence change to a single exponential. Specific exchange activities of the various exchange factor constructs were obtained as described in Ref. 37. BFA or BFC apparent inhibition constants (Ki(app)) were determined from the hyperbolic fit of kapp values plotted as a function of inhibitor concentration. All exchange measurements (spontaneous, GEF-stimulated, and inhibition) were done at least in triplicate.

Cell Culture and Immunofluorescence Microscopy—HeLa cells were grown as a monolayer in Dulbecco's modified Eagle's medium supplemented with 10% fetal calf serum, GlutaMAXTM, and an antibiotic/antimycotic mixture (Invitrogen) as described in Ref. 41. For drug treatment, cells were grown in individual wells on LabTek multichamber glass slides (Nalge Nunc International, Naperville, IL), and inhibitors were added in the culture medium as indicated in the legend to Fig. 3.

For immunofluorescence studies, cells were fixed for 10 min with 3% paraformaldehyde in phosphate-buffered saline at room temperature. Cells were glycinated, permeabilized, and blocked as described in Ref. 42. Incubation with anti-GM130 mouse monoclonal antibody (1:200) (BD Transduction Laboratories, Lexington, KY) was performed in phosphate-buffered saline containing 0.25% bovine serum albumin, 0.01% Tween 20, and 0.01% saponin overnight at 4 °C. After phosphate-buffered saline washes, staining with the secondary antibody was carried out for 20 min at room temperature with Alexa Fluor 488-conjugated goat anti-mouse IgG (Molecular Probes, Leiden, The Netherlands) and TRITC-labeled phalloidin (1:100) (Sigma). Images were collected with a Leica TCS SP2 upright laser-scanning confocal microscope (Leica Microsystems, Mannheim, Germany) with an oil x 63 (numerical aperture 1.32) objective. The different fluorochromes were detected sequentially line by line with the acousto-optical tunable filter system using laser lines 488 nm (Alexa Fluor 488) and 514 nm (TRITC).

Molecular Dynamics Simulations and Free Energy Analysis—Molecular dynamics (MD) used the Charmm27 force field (43) for the proteins, Mg2+, and GDP. Nonbonded forces were truncated using a switch function for van der Waals energy with a cut-off of 12 Å and with a shifted function for electrostatics with a cut-off of 13 Å. A dielectric constant of 1 was used, and the integration step was 1 fs. BFA and BFC bonded and Lennard-Jones parameters were derived from the Quanta (Accelrys Software Inc., San Diego, CA) force field. Atomic charges for BFA and BFC were from a Mulliken population analysis of an HF/6-31G* quantum calculation. Molecular dynamics simulations were run for ARNO4M and ARNO3M in complex with full-length nonmyristoylated Arf1 and either BFA or BFC. The experimental x-ray coordinates of the Arf1-GDP·ARNO4M·BFA complex (Protein Data Bank entry code 1R8Q; chains B and F) were used as a model to build the other three complexes. Hydrogen atoms were added using the Hbuild module of the CHARMM program (44). All structures were first energy-minimized using 1000 steps of steepest descent algorithm followed by 500 steps of adapted basis Newton-Raphson algorithm and harmonic constraints (50 kcal/mol Å2 for the side chains and 25 kcal/mol Å2 for the backbone) in order to minimize drifts from the x-ray structure. The complexes were then solvated in a water layer of 10 Å and were equilibrated by three cycles of 3 ps of heating, 5 ps of equilibration during which harmonic constraints were progressively removed, and finally 20 ps at 300 K using velocity scaling. At that stage, production was run for 1.2 ns, and coordinate frames were saved every 0.5 ps for analysis.

The free energy, {Delta}G, was estimated according to a simplified molecular mechanics-Poisson Boltzmann surface-accessible approach using the equation,

Formula 1(Eq. 1)
where {Delta}Eelec and {Delta}Evdwc are the electrostatic and van der Waals contributions to the energy of the complex formation, and Formula 1 and Formula 1 are the electrostatic and nonpolar contributions due to solvatation. Conformational entropy contributions to the thermodynamics of binding, as well as contribution to binding resulting from conformational changes between the bound and unbound complexes, are assumed to be similar in all complexes and are therefore neglected. This approximation permits a free energy decomposition into individual amino acid contributions and has been shown to be reliable in identifying dominant contributions to binding.8 Each contribution to the binding free energy entering Equation 1 was decomposed as a sum of individual contributions per residue. Van der Waals free energies of binding were computed using a 6-12 Lennard-Jones potential. The total electrostatic energy was partitioned into charge-charge and charge-solvent contributions. We used the program UHBD to solve the linearized Poisson-Boltzmann equation using a finite difference method. The electrostatic contribution was decomposed by residue according the model of Hendsch and Tidor (45, 46). The continuum electrostatics calculations used the same charges and radii for the proteins, GDP and Mg2+, as the molecular dynamics simulations. Radii for BFA and BFC were obtained by comparison with desolvation data obtained using the Jaguar Program (Jaguar version 4.1; Schrödinger Inc., Portland, OR). Several tests were performed using different radii and charge distributions to ensure the reliability of the results. The nonpolar contribution to the free energy of binding, Formula 1, was modeled as proportional to the loss in solvent-accessible surface area, which was calculated with CHARMM using a probe radius of 1.4 Å and a {gamma} constant of 5 cal·mol-1·Å-1 according to the equation,

Formula 1(Eq. 2)
The free energy decomposition was applied to a subset of conformations from the molecular dynamics trajectory, which were selected to maximize fluctuations in intermolecular electrostatic interaction energies and avoid artifacts due to the use of a single structure.8


   RESULTS
TOP
 ABSTRACT
 INTRODUCTION
 EXPERIMENTAL PROCEDURES
 RESULTS
DISCUSSION
 REFERENCES
 
BFA-sensitizing Mutations in ARNO Have No Impact on the Specific Exchange Activity—In order to analyze the specificity of BFA in terms of both the sequence of the Sec7 domain and the class of its Arf substrate, we first characterized a model system in which these parameters can be varied. All nucleotide exchange assays were performed using Trp fluorescence kinetics, taking advantage of changes in the environment of two conserved Trp residues from Arf proteins (Trp-66 and -78 in Arf1 and Arf5, Trp-62 and -74 in Arf6) upon going from their GDP- to their GTP-bound form and using Arf proteins truncated of their N-terminal helix whose exchange kinetics in solution closely mimic those of their myristoylated, full-length form in the presence of membranes (reviewed in Ref. 47). Furthermore, crystal structures show that binding of BFA to the Arf-GDP·Sec7 complex is identical whether or not the N-terminal helix is truncated (13).

We first measured the spontaneous exchange rates of human [{Delta}17]Arf1, [{Delta}14]Arf5, and [{Delta}13]Arf6 as representatives of class I, II, and III Arfs (Table 1, Fig. 1B). The spontaneous exchange rate of [{Delta}13]Arf6 was 5 times higher than that of [{Delta}17]Arf1, as previously reported (48); that for [{Delta}14]Arf5 was 3-fold higher. We then measured the specific exchange activity of the Sec7 domain of ARNO (ARNOWT) on all three Arf classes (Table 1). ARNOWT was 6 times less active on [{Delta}13]Arf6 compared with Arf1, as previously observed by others (25, 49), but equally efficient on [{Delta}17]Arf1 and [{Delta}14]Arf5. These data establish ARNO as an efficient exchange factor for all three Arf classes in vitro.


View this table:
[in this window]
[in a new window]
  		TABLE 1
Arf1, Arf5, and Arf6: Spontaneous exchange, ARNO-stimulated exchange, and effect of BFA-sensitizing mutations

 
ARNO is naturally insensitive to BFA (50), but it can be made BFA-sensitive by introducing four mutations (F190Y/A191S/S198D/P208M) in its Arf1-interacting region (8, 12). We therefore checked whether these mutations might affect the specific exchange activity on the three Arf classes (Table 1). In all cases, ARNO4M-specific exchange activity was very similar to that of ARNOWT. Likewise, we found that ARNO constructs containing only the F190Y (ARNO1M) or the A191S/S198D/P208M (ARNO3M) mutations were also essentially as active as ARNOWT on [{Delta}17]Arf1 (Table 2). We conclude that BFA-sensitizing mutations are silent with respect to the exchange activity in the absence of the inhibitor, regardless of the Arf substrate used. As a consequence, ARNOWT and its different mutants can be used as a model to analyze the contribution of Sec7 domain and Arf classes to BFA sensitivity.


View this table:
[in this window]
[in a new window]
  		TABLE 2
BFA-sensitizing residues of Sec7 domains: Contribution to inhibition of the exchange reaction by BFA and BFC All activities were measured for [{Delta}17]Arf1. The sequence numbering is for ARNO.

 


Figure 1
View larger version (27K):
[in this window]
[in a new window]
  		FIGURE 1.
Specificity of BFA for Arf and Sec7 domains. A, determination of Ki(app) by tryptophan fluorescence kinetics. Left, the exchange reaction is started by the addition of 100 µM GTP to N-terminally truncated Arf-GDP preincubated with the Sec7 domain and increasing concentrations of the inhibitor, from which acceleration rates (kapp) are calculated (here Arf1, ARNO4M, and BFA). Right, Ki(app) is determined from hyperbolic fit of kapp values plotted as a function of inhibitor concentration. B, spontaneous exchange rates of Arf1, Arf5, Arf6, and Arf6 mutants. Exchange rates were determined by using a 1 µM concentration of Arf isoforms after injection of 100 µM GTP in the absence of the Sec7 domain. C, contribution of BFA-sensitizing residues to Ki(app) of BFA (black) and BFC (white) on ARNO, ARNO mutants, and BIG1. All experiments were done with Arf1 as described under "Experimental Procedures." D, acceleration rates of Arf1, Arf5, Arf6, and Arf6 mutants by ARNOWT (black) and ARNO4M (white). Acceleration rates are calculated as the ratio of the Sec7-stimulated to the spontaneous exchange rates. 100% represents the acceleration rate of [{Delta}17]Arf1 by ARNOWT. E, specificity of BFA inhibition for wild-type and mutant Arf proteins. All experiments were done with ARNO4M.

 
Tyr-190 Is the Only Essential Residue in the Sec7 Domain for BFA Sensitivity—The crystal structures of BFA-inhibited Arf1-GDP·ARNO4M complexes (13) show that of the four mutations that confer BFA sensitivity, only Tyr-190 is in direct contact with BFA, whereas Ser-191 is not on the protein surface and Asp-198 and Met-208 are found in the second coordination shell of the inhibitor (Fig. 2, A and B). Tyr-190 forms a hydrogen bond with the 7-hydroxyl of BFA (referred to as OH7 hereafter), which in turn forms a second hydrogen bond with Trp-78 from Arf1 (Fig. 2, A and B). In order to determine the respective contribution of the BFA-interacting versus second shell residues, BFA apparent inhibition constants Ki(app), calculated as described under "Experimental Procedures," were obtained using [{Delta}17]Arf1 tryptophan fluorescence kinetics in the presence of Sec7 constructs carrying different combinations of sequences at the BFA site (Fig. 1A). The exchange reactions were initiated by the addition of GTP after a 5-min incubation of the Arf and Sec7 proteins with various concentrations of BFA. This incubation time was sufficient to reach the equilibrium of the abortive complex formation, with longer times having no effect on the observed rates (data not shown).

As a reference, we first analyzed the Ki(app) of BFA for the Sec7 domain of BIG1, a naturally BFA-sensitive ArfGEF (11, 51), which carries the YSDM sequence (Table 2). BIG1 was found to be slightly less active than ARNO for [{Delta}17]Arf1 and to be inhibited by BFA with a Ki(app) of 13 ± 9 µM. We then compared the BFA sensitivity of ARNO constructs carrying no BFA-sensitizing sequence (ARNOWT), the BFA-interacting mutation (ARNO1M), the second shell mutations (ARNO3M), or all four mutations (ARNO4M). We observed that the YSDM mutations in ARNO4M indeed confer BFA sensitivity with a Ki(app) of 12 µM, a value similar to that of the naturally BFA-sensitive BIG1. In contrast, the Ki(app) determined with ARNOWT was 20 times higher, as expected for a BFA-resistant exchange factor.


Figure 2
View larger version (53K):
[in this window]
[in a new window]
  		FIGURE 2.
Brefeldin A structure, binding site, and interactions. A, interfacial binding of BFA in the Arf1-GDP·ARNO4M·BFA complex (13). Arf1 is shown in green, and Sec7 is shown in blue. Residues from Sec7 domains previously reported to confer BFA sensitivity are shown in yellow. Hydrogen bonds are shown as dotted lines. B, schematic view of BFA interactions with Arf1 (in green) and ARNO4M (in blue). Hydrogen bonds are shown as dotted lines, and hydrophobic contacts (cut-off of 4 Å) are shown as half circles. Sec7 residues analyzed in this work are framed. C, chemical structures of brefeldin A and brefeldin C. D, sequence differences between Arf1 and Arf6 in the vicinity of BFA. The molecular surface of the Arf1-GDP·ARNO4M·BFA complex (13) is shown in brown, with Arf1 (green) and ARNO4M (blue) shown in transparency. Arf6-GDP (yellow) (2) is superimposed on Arf1 based on the interswitch region, showing the more charged or bulky residues (Lys-58 and Ser-60 in Arf6, Ser-62 and Thr-64 in Arf1) that could hinder access to the interfacial pocket.

 


Figure 3
View larger version (85K):
[in this window]
[in a new window]
  		FIGURE 3.
BFA and BFC effects on the Golgi in HeLa cells. HeLa cells were treated with 0.5% EtOH (A) or 100 µM BFA (B) or BFC (C) for 20 min at 37 °C prior to being analyzed by fluorescence microscopy as described under "Experimental Procedures." The cis Golgi was visualized by monoclonal antibody to the matrix protein GM130 (green), and the F-actin was stained by TRITC-phalloidin (red). Bar, 40 µm.

 
We then compared the BFA sensitivity of ARNO4M with those of ARNO1M and ARNO3M (Table 2, Fig. 1C). Surprisingly, ARNO3M was found to be BFA-insensitive, with a Ki(app) only slightly lower than that of ARNOWT within experimental errors. Conversely, ARNO1M was only slightly less sensitive to BFA than ARNO4M. We conclude that Tyr-190 is essential and sufficient to sensitize the Sec7 domain of ARNO to BFA, whereas the other three mutations have little, if any, effect in BFA sensitivity.

OH7 of BFA Is Critical for Its Inhibition of the Exchange Reaction— Since the main contribution of Tyr190 in the Arf1-GDP·BFA·ARNO4M complex is the formation of a hydrogen bond with OH7 of BFA (Fig. 2, A and B), we further investigated the contribution of this interaction by removing this hydroxyl from BFA. The corresponding compound is brefeldin C (Fig. 2C), which cannot form the hydrogen bond regardless of whether the Sec7 sequence is Tyr or Phe. We first compared the inhibitory properties of BFA and BFC using BFA-sensitive Sec7 constructs and [{Delta}17]Arf1. The Ki(app) rose from 12 µM for BFA to 201 µM for BFC in the case of ARNO4M and from 13 to 122 µM for the naturally BFA-sensitive BIG1 (Table 2, Fig. 1C). Since the DM mutations are not involved in binding the OH7 of BFA, we further investigated whether partial BFC inhibition may be recovered in the ARNO3M mutant. However, the Ki(app) values were again in the 200 µM range, indicating that BFC failed to inhibit this construct (Table 2). BFC was also ineffective on ARNOWT. Likewise, BFC did not inhibit the activation of [{Delta}13]Arf6 by either ARNOWT or ARNO4M (data not shown). All three human ArfGEFs acting at the Golgi are predicted to be the target of BFA in vivo (11, 20). Therefore, we finally checked whether BFC was able to induce BFA-like effects in HeLa cells. In EtOH-treated control cells, the cis/medial Golgi marker GM130 exhibited a perinuclear localization, as visualized by immunofluorescence staining (Fig. 3A). In contrast, BFA treatment led to the almost complete dissolution of the Golgi apparatus, and GM130 seemed completely dispersed throughout the cytoplasm (Fig. 3B). As expected from the ~10 times higher Ki(app) of BFC compared with BFA in vitro, the effect of BFC on HeLa cells (100 µM) was significantly attenuated compared with BFA, with cells retaining well defined perinuclear structures (Fig. 3C). However, BFC induced a significant dispersion of the marker that was still detectable even at 20 µM (data not shown), suggesting that the compound has an inhibitory activity in the cell.

The inability of BFC to inhibit any of the ARNO constructs, together with our comparison of the ARNO1M/ARNO3M mutants, thus establishes that the hydrogen bond observed in the crystal structures between the Sec7 domain (Tyr-190) and the inhibitor (OH7) is essential in conferring inhibition. Surprisingly, the Ki(app) values measured with BFA or BFC using ARNOWT or ARNO3M were similar, although BFA is able to form an additional direct interaction with Trp-78 in the switch 2 of Arf, which BFC cannot (see Fig. 2, A and C). This aspect will be discussed in more detail below in light of free energy computational analysis.

Arf Proteins Contribute to BFA Sensitivity—The crystal structures of the Arf1-GDP·BFA·ARNO4M complexes show that BFA is entirely buried at the Arf/Sec7 interface, where it makes contacts with both Arf and the Sec7 domain (13, 14). The interface area is formed with the Sec7 domain for one-third and with Arf1 for two-thirds, making Arf1 a major contributor to binding the inhibitor (Fig. 2, A and B). We therefore investigated the contribution of Arf proteins from classes I, II, and III to BFA sensitivity of the exchange reaction, using the BFA-sensitive ARNO4M mutant, which was shown to be active on all three truncated Arf constructs (Table 3). BFA was able to inhibit the activation of [{Delta}14]Arf5 by ARNO4M with a Ki(app) similar to the one determined for [{Delta}17]Arf1. In contrast, BFA did not inhibit [{Delta}13]Arf6 activation using this BFA-sensitive exchange factor. Thus, the presence of BFA-sensitizing mutations on the Sec7 domain is not sufficient to yield inhibition in vitro. Inhibition thus requires determinants that are carried by the Arf protein, with class I and II Arfs being BFA-sensitive and class III Arf6 being BFA-insensitive.


View this table:
[in this window]
[in a new window]
  		TABLE 3
Specificity of BFA for class I, II, and III Arf proteins

 
Toward Making Arf6 Sensitive to BFA—The BFA insensitivity of Arf6 in the context of a BFA-sensitive Sec7 domain is puzzling, since all Arf1 residues that interact directly with BFA are strictly conserved within Arf6. Sequence differences between Arf1 and Arf6 closest to BFA are located in the interswitch, where Arf6 carries residues that are either more charged or bulky compared with those of Arf1 and could hinder access to the interfacial pocket (Ser-62 in Arf1 versus Lys-58 in Arf6, Thr-64 in Arf1 versus Asn-60 in Arf6; see Fig. 2D). Another difference between Arf1 and Arf6 is located upstream the switch 1 (Ile-42 in Arf1 versus Ser-38 in Arf6) and had been shown earlier to be involved in the low affinity of Arf6-GDP for Mg2+; hence its fast nucleotide exchange rate (48). Although located at a distance from the BFA-binding site, this difference may impact on the exchange mechanism in such a way as to exclude the binding of the drug. In order to evaluate whether these sequence differences are involved in the insensitivity of Arf6 to BFA, mutants of [{Delta}13]Arf6 carrying either the K58S, K58S/N60T, or S38I mutations were characterized and tested in the BFA-sensitive ARNO4M assay.

We first characterized the spontaneous and ARNO-stimulated exchange rates of the different mutants (Table 4). All Arf6 mutants exchanged GDP at a slower rate compared with Arf6WT, with Arf6K58S being intermediate between Arf1 and Arf6, and Arf6S38I and Arf6K58S/N60T similar to Arf1 (Fig. 1B). Thus, both the previously described S38I mutation, which allows for a stabilizing interaction with Mg2+, and mutations in the interswitch remote from the nucleotide binding site make Arf6 closer to Arf1 in terms of spontaneous nucleotide exchange. We then characterized whether the mutations had an effect on ARNO-stimulated nucleotide exchange. Since Arf6 and its various mutants have different spontaneous exchange rates, the acceleration rate rather than the specific exchange activities were used for the purpose of comparison (Fig. 1D). We found that the S38I and K58S mutations are neutral with regard to ARNOWT acceleration rate compared with Arf6WT (5, 8, and 4%, respectively, of its acceleration rate for Arf1). In contrast, the acceleration rate for Arf6K58S/N60T was 24% that of Arf1, which is in the range of that for Arf5 (25%). Thus, only the double mutation in the interswitch results in an Arf chimera that is closer to Arf1 with regard to its activation by ARNO, whereas other mutations, although they reduce the spontaneous exchange rate, have no effect.


View this table:
[in this window]
[in a new window]
  		TABLE 4
Introducing Arf1 residues in Arf6: Effect on spontaneous exchange, on ARNO-stimulated exchange, and on the apparent inhibition constant of BFA (ARNO4M-stimulated exchange)

 
Acceleration rates measured with the BFA-sensitive ARNO4M were indistinguishable from those obtained with ARNOWT, allowing for the evaluation of the BFA responsiveness of the Arf6 mutants (Fig. 1E). All three Arf6 mutants had Ki(app) smaller than that of Arf6WT: 509 ± 48 µM for Arf6S38I, 558 ± 45 for Arf6K58S, and 450 ± 29 for Arf6K58S/N60T compared with 866 ± 77 µM for Arf6WT. Thus, all Arf6 mutants have a significantly improved response to BFA. However, they remain essentially insensitive to BFA compared with Arf1 (Ki(app) = 12 µM) and Arf5 (Ki(app) = 23 µM).

Comparison of Tyr/Phe and BFA/BFC by Free Energy Decomposition— The above biochemical analysis uncovers a critical role for both Arf and the Sec7 domain in conferring BFA sensitivity to the exchange reaction and points to the hydrogen bond between the Tyr residue in the Sec7 domain and the OH7 in BFA as an essential determinant of this sensitivity. In order to gain insight in the energetic components of these interactions, we took advantage of the Arf1-GDP·ARNO4M·BFA crystal structure to compute molecular dynamics trajectories and free energy decomposition of this complex and two other complexes characterized in this study: Arf1-GDP·ARNO3M·BFA and Arf1-GDP·ARNO3M·BFC. Although the free energy decomposition does not give a quantitative evaluation of binding thermodynamics, it gives important insight on the role of individual amino acids in the molecular recognition process. MD trajectories showed only small deviations from the crystal structures, with the smallest root mean square deviation observed in the BFA binding pocket (data not shown), thus validating the MD protocol for the analysis of inhibitor-protein interactions. In what follows, we focus on those amino acids that confer BFA sensitivity. A complete description of the data (including BFA-protein and protein-protein interactions) will be published elsewhere.

We first analyzed the respective energetic contributions of Arf1 and ARNO4M to the binding of BFA based on the Arf1-GDP·BFA·ARNO4M crystal structure (13). Overall, the interactions of Arf1 with BFA are more important than those of the Sec7 domain; they account for 65 and 35% of the free energy, respectively. In Arf1, interactions are dominated by the van der Waals terms. The most significant interactions are made by residues that are in direct contact with the drug, notably with the lactone moiety, and are conserved between Arf1 and Arf6. The free energy decomposition for Ile-42, Ser-62, and Thr-64, which correspond to the mutations introduced in Arf6 (see above), shows that Thr-64 has a van der Waals interaction of about 1 kcal/mol with the ligand but that the other two residues do not interact with the inhibitor (Fig. 4A). Favorable electrostatic interactions with BFA are made by Trp-66 and Trp-78 (Fig. 4B). Trp-66 is stacked above BFA and makes polar interactions, whereas Trp-78 is oriented perpendicular to BFA and makes a hydrogen bond between its ring NH and OH7 of BFA. In contrast, both van der Waals and electrostatic interactions contribute to a similar extent in ARNO4M (Fig. 4A). Of the four amino acids that have been reported to confer BFA sensitivity, only Tyr-190 has a significant energetic contribution. This contribution is entirely electrostatic (Fig. 4B), and it is due to the hydrogen bond with OH7 of BFA, which is maintained throughout the MD simulation.

We then investigated the impact of removing this hydrogen bond by the Tyr/Phe mutation (Arf1-GDP·ARNO3M·BFA complex). This resulted in only minor rearrangements in the BFA pocket during the MD run and essentially no change in the energetic decomposition except for the removed hydrogen bond (Fig. 4, A and B). In particular, the hydrogen bond between OH7 of BFA and Trp-78Arf was maintained throughout the simulation. The free energy decomposition thus points at the hydrogen bond between Tyr-190ARNO and OH7 of BFA as a major component of the affinity of BFA for the complex and its loss as the major explanation for the drop in BFA inhibition efficiency in ARNO3M.

A surprising result of the biochemical analysis is that BFA is not better than BFC at inhibiting the exchange reaction in assays using Sec7 domains that carry the Phe residue (ARNOWT and ARNO3M), although BFA can form a hydrogen bond with Trp-78 in Arf1, which BFC cannot form because of its missing OH7 group. We therefore compared the free energy decompositions of Arf1-GDP· ARNO3M·BFA and Arf1-GDP·ARNO3M·BFC, which differ by their ability or inability, respectively, to form this hydrogen bond. The free energy analysis confirms that the OH7/Trp-78 hydrogen bond has a favorable contribution to the overall electrostatic energy in the Arf1-GDP· ARNO3M·BFA complex, which is lost in the Arf1-GDP·ARNO3M·BFC complex (Figs. 5, A and C). The free energy decomposition for both ligands indicates that binding is dominated by van der Waals interactions and that electrostatics are overall unfavorable (Fig. 5, B and D). However, the desolvation cost of BFA upon binding to the Arf-Sec7 complex is larger than that of BFC, which is reflected in their relative electrostatic contribution to binding (Fig. 5D). The free energy analysis thus predicts that unfavorable desolvation energy in BFA compensates for the favorable electrostatic energy of its OH7-Trp-78 hydrogen bond. As a result, BFA is not favored over BFC in inhibiting the Arf1-ARNO3M complex, and only the ARNO4M construct has sufficient electrostatic interactions to overcome the desolvation cost of BFA.


Figure 4
View larger version (24K):
[in this window]
[in a new window]
  		FIGURE 4.
Free energy comparison of BFA binding to ARNO3M (Phe) and ARNO4M (Tyr). Shown is free energy decomposition for selected amino acids in the Arf1-GDP·ARNO4M·BFA and Arf1-GDP·ARNO3M·BFA complexes (for details, see "Experimental Procedures"). Amino acids from ARNO are indicated with an asterisk; residue 190* is Tyr in ARNO4M and Phe in ARNO3M. Only the contributions from amino acids discussed in this study are indicated. A, total contribution to the free energy (see Equation 1); B, electrostatic (including desolvation) contribution to the free energy. Units are kcal/mol.

 


Figure 5
View larger version (27K):
[in this window]
[in a new window]
  		FIGURE 5.
Free energy comparison of BFA and BFC binding to Arf1-GDP·ARNO3M. Shown is free energy decomposition for the ligand and selected amino acids in the Arf1-GDP·ARNO3M·BFA and Arf1-GDP·ARNO3M·BFC complexes (for details, see "Experimental Procedures"). A, total contribution to the free energy (see Equation 1) for Arf1-GDP. B,asin A, for the ligand (BFA/BFC). C, electrostatic (including desolvation) contribution to the free energy for Arf1-GDP. D,asin C, for the ligand (BFA/BFC). For the sake of clarity, amino acids from ARNO3M are omitted. Their contributions are similar in both complexes. Units are kcal/mol.

 

   DISCUSSION
TOP
 ABSTRACT
 INTRODUCTION
 EXPERIMENTAL PROCEDURES
 RESULTS
 DISCUSSION
REFERENCES
 
In this study, we characterized the respective contributions of Arf proteins from classes I, II, and III and from the sequence of the Sec7 domain to conferring sensitivity of the exchange reaction to interfacial inhibition by BFA or a BFA analog. Our analysis was based on kinetics analysis using tryptophane fluorescence and was further supported by free energy computational analysis. Kinetics data were analyzed following a Michealis-Menten formalism, yielding Ki(app) constants that were used to compare the different contributions. It should be noted here that this value should not be confused with a bona fide inhibition constant; however, it provides a convenient and robust way to compare the different experiments.

A Dual Contribution of Arf Proteins and Sec7 Domains to BFA Sensitivity—We first characterized the individual contributions of Sec7 residues that had been previously reported to confer BFA sensitivity to the exchange reaction. We first showed that whether these residues have a BFA-sensitive or BFA-insensitive sequence has no impact on Arf activation. These positions are thus likely to be silent as long as the system does not encounter an inhibitor and their sequences probably do not reflect a selection pressure. This is in agreement with the observation that ArfGEF orthologs from different species may have different sequences at BFA-sensitizing locations (16, 52). Combining site-directed mutagenesis and the use of BFC, a BFA analog, we then consistently showed that of the different positions that have been reported in the literature, only one has a strong impact on BFA sensitivity. This is consistent with crystallographic data, which show that this residue, Tyr-190 in BFA-sensitive ARNO mutants, is the only one that forms a direct contact with the inhibitor. Furthermore, our free energy decomposition suggests that this position is one of the two residues that provide significant contribution to the BFA/Sec7 interaction energy. The other one is an essentially invariant Met (Met-194 in ARNO; see Fig. 4A), which has a strong hydrophobic contribution to the binding of BFA and whose replacement by a branched residue (Ile) was used to create BFA-insensitive ArfGEF versions (53), probably through steric conflict with the drug. In the case of the Tyr residue, we suggest that the electrostatic interaction with the drug is not only essential to increase the interaction energy of the inhibitor with the Arf-Sec7 complex but may also help to drive the inhibitor toward the bottom of the interfacial cavity with a proper orientation.

On the other hand, the minor contribution of the Asp/Met sequence was unexpected, since these residues were previously reported to induce BFA sensitivity in ARNO (12) and in cytohesin 1 (18) and to potentiate the effect of the YS mutations in ARNO (12). It is possible that this discrepancy reflects that fact that none of the previously reported experiments assessed Ki(app) inhibition constants as reported here. Notably, the study reported in Ref. 12 focused on koff values (i.e. binding rather than inhibition). Thus, the contribution of the Tyr residue to the kon component may be significant, whereas that of DM would be negligible. Likewise, experiments in Ref. 18 reported BFA effects at equilibrium using high BFA concentrations, at which we also observed modest inhibition. Besides these differences in the type of biochemical parameter that was actually explored, we also noticed that inhibition efficiency was strongly influenced by the nature of the exchange buffer, with Hepes yielding about 2 times faster exchange rates compared with Tris for BIG1 and ARNO (data not shown). This may explain the somewhat faster specific activity of ARNO reported in Ref. 37 relative to the present study.

Arf1 is a major contributor to the interfacial cavity, burying two-thirds of the surface area of BFA in the crystal structures (13, 14). Yet, whether Arf proteins contribute to BFA sensitivity had not been previously investigated. Arf6 activation has been reported to be insensitive to BFA effects in cells (23, 24); however, whether Arf6 contributes to this insensitivity is not known, since the ARNO/cytohesin family of candidate GEFs for Arf6 is insensitive to BFA (25, 50), whereas individual insensitivity of EFA6 versus Arf6 cannot be distinguished, since EFA6 is active solely on Arf6 (26). Here, using an ARNO mutant that is active on all three Arf classes and is BFA-sensitive, we establish that class I and class II Arfs are BFA-sensitive in vitro, whereas Arf6 activation was not inhibited by BFA regardless of whether the Sec7 domain was BFA-sensitive or not. However, a puzzling observation is that every residue of Arf1 in interaction with BFA is identical in Arf6, together with all residues in contact with Sec7 domains, whatever the Arf/Sec7 intermediate considered (except for an Ile-49Arf1/Val-45Arf6 substitution that forms a loose hydrophobic interaction in Arf1).

In an effort to identify positions in Arf6 responsible for its biochemical differences with Arf1, we analyzed the spontaneous and Sec7-stimulated nucleotide exchange and the BFA response of Arf6 mutants in which selected residues were changed to their Arf1 counterparts. We focused on residue differences located in the interswitch in the BFA second coordination shell (K58S and N60T) (13) and on a residue change that was previously shown to make Arf6 closer to Arf1 in terms of spontaneous GDP dissociation and Mg2+ affinity (S38I) (48). For both types of mutants, spontaneous exchange was closer to that of Arf1, and the response to BFA was improved, although not to the point of actual inhibition. Surprisingly, only the double mutation in the inter-switch resulted in an Arf6 mutant with Sec7-stimulated nucleotide exchange closer to that of Arf1. The symmetrical single site mutations in Arf1 (S62K and I42S) were reported earlier to have no effect on improving Arf1 activation by an Arf6-specific exchange factor, EFA6 (49). Based on the available crystal structures, the absence of effect of the S38I mutation may be explained by the fact that the Mg2+-destabilizing interaction allowed by the Ser residue in Arf6 is disrupted as soon as the switch I is displaced, which is the initiating event of the exchange reaction. This also suggests that dissociation of Mg2+ is not rate-limiting in Sec7-stimulated nucleotide exchange, since no improvement is achieved regardless of the affinity for Mg2+. In contrast, the interswitch contributes to both spontaneous and Sec7-stimulated Arf1/Arf6 differences. The improved BFA response of the Arf6 mutants rules out that the insensitivity of Arf6 to BFA is due to its inability to form an interfacial pocket with Sec7 domains. This is consistent with the structural snapshots of the exchange reaction, which suggest that the pocket has a mechanistic role in driving the interswitch toggle (13). The correlation between slower spontaneous GDP dissociation kinetics and improved BFA response suggests that the mutations allow longer, rather than sterically or electrostatically more favorable, access to the interfacial pocket. Altogether, our results show that BFA has a dual specificity for each partner of the complex. They demonstrate the exquisite potential of interfacial inhibitors to select between highly similar targets and to sense structural and chemical characteristics of both protein components of the complex. Such selectivity may thus potentially be achieved by interfacial inhibitors for other families of SMGs and their GEFs, such as the Ras or Rho family, where many therapeutic targets can potentially be found (reviewed in Refs. 2-4).

Revisiting the Effects of BFA Analogs: Toward Inhibitors of Arf Activation—Activation of Arf proteins has recently been increasingly involved in human diseases, including their activation in bacterial (32-35) and viral (30, 31) infections as well as in cancer invasion and angiogenesis (27-29). They are therefore candidate targets for inhibition by therapeutic compounds, for which BFA may serve as a starting point for the design of interfacial inhibitors. Here we analyzed one of the simplest BFA analogs, BFC (39), in which a single hydroxyl is removed, thus impairing the formation of a pair of hydrogen bonds with Sec7 and Arf. This analog was characterized with both BFA-sensitive and BFA-insensitive Arf and Sec7 domains but failed to inhibit any of the exchange reactions tested. Interestingly, the fact that the BFC/PheSec7 pair was as ineffective as the BFA/PheSec7 pair was only partially explained by the removal of the targeted hydrogen bond, since BFA can still make an additional BFA/Arf hydrogen bond compared with BFC. These observations were, however, fully accounted for by the free energy decomposition, which uncovered a desolvation effect compensating for the loss of the hydrogen bond. Several other BFA analogs were previously analyzed for their effect on disrupting the Golgi apparatus (6) and more recently for their cytotoxic and apoptotic effects (54). Although none of these experiments assessed the inhibitory mechanism, they consistently showed that modifications of the 7-hydroxyl impair the efficiency of the drug. Epimerization and oxidation of this hydroxyl yielded analogs that were significantly less efficient than BFA at releasing beta-COP from the Golgi in cells, whereas larger modifications were completely inactive.9 Likewise, the 7-stereoisomer was ineffective at inducing apoptosis (54). Replacement of the 7-hydroxyl by 7-O-acetyl was the only modification that was neutral, but it is likely that this compound is hydrolyzed in the cell (54). Altogether, this identifies attractive interactions at the bottom of the interfacial cavity as an essential component of the binding of BFA, which will need to be fulfilled in future design of BFA derivatives or analogs aimed at BFA-insensitive Arf/ArfGEF pairs of therapeutic interest.

In conclusion, our analysis predicts that Sec7 domains that carry a Tyr residue are BFA-sensitive, even in the case of mixed sequences such as human GBF1 (Golgi-specific brefeldin A resistance factor 1; YADM). This is in agreement with recent data, which strongly suggest that GBF1 has the hallmarks of a BFA-sensitive GEF in vivo (20). Thus, introducing this mutation in BFA-insensitive GEFs should be sufficient to convert them into a BFA-sensitive version toward BFA-sensitive Arf of classes I and II. Since this mutation does not interfere with the exchange efficiency, it may now be used to investigate the cellular functions of BFA-insensitive GEFs that act on BFA-sensitive small G proteins, such as ARNO and its putative non-Arf6 substrates.


   FOOTNOTES
 
* This work was supported by grants from the Association pour la Recherche contre le Cancer and the French Research Ministry (to J. C.). All authors are members of the CNRS research network GDR2823. The costs of publication of this article were defrayed in part by the payment of page charges. This article must therefore be hereby marked "advertisement" in accordance with 18 U.S.C. Section 1734 solely to indicate this fact. Back

1 Supported by a grant from the French Ministry. Back

2 These authors contributed equally to this work. Back

3 Supported by a grant from CNRS/Région Alsace. Back

4 Supported by a grant from the Ligue contre le Cancer. Back

5 Supported by CNRS, INSERM, and the University Louis Pasteur de Strasbourg. Back

6 To whom correspondence should be addressed. Tel.: 33-1-69-82-34-92; Fax: 33-1-69-82-31-29; E-mail: cherfils{at}lebs.cnrs-gif.fr.

7 The abbreviations used are: GEF, guanine nucleotide exchange factor; BFA, brefeldin A; BFC, brefeldin C; SMG, small GTP-binding protein; MD, molecular dynamics; TRITC, tetramethylrhodamine isothiocyanate. Back

8 A. Dejaegere, unpublished results. Back

9 J. Donaldson, personal communication. Back


   ACKNOWLEDGMENTS
 
We thank B. Mouratou (LEBS, CNRS, France) for help with the modified pET28 vectors and the [{Delta}14]Arf5 construct, G. Daoust and V. Biou (LEBS, CNRS, France) for help with the BIG1-Sec7 construct; and B. Olofsson and E. Dransart (LEBS, CNRS, France) and the staff at the Imaging and cell biology facility (ISV, CNRS, France) for help. A. D. thanks CINES/IDRIS for the generous allowance of computer time.



   REFERENCES
TOP
 ABSTRACT
 INTRODUCTION
 EXPERIMENTAL PROCEDURES
 RESULTS
 DISCUSSION
 REFERENCES
 

  1. Marinissen, M. J., and Gutkind, J. S. (2005) Trends Biochem. Sci. 30, 423-426[CrossRef][Medline] [Order article via Infotrieve]
  2. Zeghouf, M., Guibert, B., Zeeh, J.-C., and Cherfils, J. (2005) Biochem. Soc. Trans. 33, 1265-1268[CrossRef][Medline] [Order article via Infotrieve]
  3. Walker, K., and Olson, M. F. (2005) Curr. Opin. Genet. Dev. 15, 62-68[CrossRef][Medline] [Order article via Infotrieve]
  4. Rossman, K. L., Der, C. J., and Sondek, J. (2005) Nat. Rev. Mol. Cell Biol. 6, 167-180[CrossRef][Medline] [Order article via Infotrieve]
  5. Pommier, Y., and Cherfils, J. (2005) Trends Pharmacol. Sci. 26, 138-145[CrossRef][Medline] [Order article via Infotrieve]
  6. Donaldson, J. G., Finazzi, D., and Klausner, R. D. (1992) Nature 360, 350-352[CrossRef][Medline] [Order article via Infotrieve]
  7. Helms, J. B., and Rothman, J. E. (1992) Nature 360, 352-354[CrossRef][Medline] [Order article via Infotrieve]
  8. Peyroche, A., Antonny, B., Robineau, S., Acker, J., Cherfils, J., and Jackson, C. L. (1999) Mol. Cell 3, 275-285[CrossRef][Medline] [Order article via Infotrieve]
  9. Jackson, C. L. (2000) Subcell. Biochem. 34, 233-272[Medline] [Order article via Infotrieve]
  10. Donaldson, J. G., Honda, A., and Weigert, R. (2005) Biochim. Biophys. Acta 1744, 364-373[Medline] [Order article via Infotrieve]
  11. Mansour, S. J., Skaug, J., Zhao, X. H., Giordano, J., Scherer, S. W., and Melancon, P. (1999) Proc. Natl. Acad. Sci. U. S. A. 96, 7968-7973[Abstract/Free Full Text]
  12. Robineau, S., Chabre, M., and Antonny, B. (2000) Proc. Natl. Acad. Sci. U. S. A. 97, 9913-9918[Abstract/Free Full Text]
  13. Renault, L., Guibert, B., and Cherfils, J. (2003) Nature 426, 525-530[CrossRef][Medline] [Order article via Infotrieve]
  14. Mossessova, E., Corpina, R. A., and Goldberg, J. (2003) Mol. Cell 12, 1403-1411[CrossRef][Medline] [Order article via Infotrieve]
  15. Lee, F. J., Stevens, L. A., Hall, L. M., Murtagh, J. J., Jr., Kao, Y. L., Moss, J., and Vaughan, M. (1994) J. Biol. Chem. 269, 21555-21560[Abstract/Free Full Text]
  16. Mouratou, B., Biou, V., Joubert, A., Cohen, J., Shields, D. J., Geldner, N., Jurgens, G., Melancon, P., and Cherfils, J. (2005) BMC Genomics 6, 20[CrossRef][Medline] [Order article via Infotrieve]
  17. Shin, H. W., and Nakayama, K. (2004) J. Biochem. (Tokyo) 136, 761-767[Abstract/Free Full Text]
  18. Sata, M., Moss, J., and Vaughan, M. (1999) Proc. Natl. Acad. Sci. U. S. A. 96, 2752-2757[Abstract/Free Full Text]
  19. Zhao, X., Lasell, T. K., and Melancon, P. (2002) Mol. Biol. Cell 13, 119-133[Abstract/Free Full Text]
  20. Niu, T. K., Pfeifer, A. C., Lippincott-Schwartz, J., and Jackson, C. L. (2005) Mol. Biol. Cell. 16, 1213-1222[Abstract/Free Full Text]
  21. Tsai, S. C., Adamik, R., Haun, R. S., Moss, J., and Vaughan, M. (1993) J. Biol. Chem. 268, 10820-10825[Abstract/Free Full Text]
  22. Volpicelli-Daley, L. A., Li, Y., Zhang, C. J., and Kahn, R. A. (2005) Mol. Biol. Cell. 16, 4495-4508[Abstract/Free Full Text]
  23. Peters, P. J., Hsu, V. W., Ooi, C. E., Finazzi, D., Teal, S. B., Oorschot, V., Donaldson, J. G., and Klausner, R. D. (1995) J. Cell Biol. 128, 1003-1017[Abstract/Free Full Text]
  24. Donaldson, J. G. (2003) J. Biol. Chem. 278, 41573-41576[Free Full Text]
  25. Frank, S., Upender, S., Hansen, S. H., and Casanova, J. E. (1998) J. Biol. Chem. 273, 23-27[Abstract/Free Full Text]
  26. Franco, M., Peters, P. J., Boretto, J., van Donselaar, E., Neri, A., D'Souza-Schorey, C., and Chavrier, P. (1999) EMBO J. 18, 1480-1491[CrossRef][Medline] [Order article via Infotrieve]
  27. Tague, S. E., Muralidharan, V., and D'Souza-Schorey, C. (2004) Proc. Natl. Acad. Sci. U. S. A. 101, 9671-9676[Abstract/Free Full Text]
  28. Ikeda, S., Ushio-Fukai, M., Zuo, L., Tojo, T., Dikalov, S., Patrushev, N. A., and Alexander, R. W. (2005) Circ. Res. 96, 467-475[Abstract/Free Full Text]
  29. Hashimoto, S., Onodera, Y., Hashimoto, A., Tanaka, M., Hamaguchi, M., Yamada, A., and Sabe, H. (2004) Proc. Natl. Acad. Sci. U. S. A. 101, 6647-6652[Abstract/Free Full Text]
  30. Faure, J., Stalder, R., Borel, C., Sobo, K., Piguet, V., Demaurex, N., Gruenberg, J., and Trono, D. (2004) Curr. Biol. 14, 1056-1064[CrossRef][Medline] [Order article via Infotrieve]
  31. Belov, G. A., Fogg, M. H., and Ehrenfeld, E. (2005) J. Virol. 79, 7207-7216[Abstract/Free Full Text]
  32. Kahn, R. A., and Gilman, A. G. (1986) J. Biol. Chem. 261, 7906-7911[Abstract/Free Full Text]
  33. Nagai, H., Kagan, J. C., Zhu, X., Kahn, R. A., and Roy, C. R. (2002) Science 295, 679-682[Abstract/Free Full Text]
  34. Balana, M. E., Niedergang, F., Subtil, A., Alcover, A., Chavrier, P., and Dautry-Varsat, A. (2005) J. Cell Sci. 118, 2201-2210[Abstract/Free Full Text]
  35. O'Neal, C. J., Jobling, M. G., Holmes, R. K., and Hol, W. G. (2005) Science 309, 1093-1096[Abstract/Free Full Text]
  36. Sheen, V. L., Ganesh, V. S., Topcu, M., Sebire, G., Bodell, A., Hill, R. S., Grant, P. E., Shugart, Y. Y., Imitola, J., Khoury, S. J., Guerrini, R., and Walsh, C. A. (2004) Nat. Genet. 36, 69-76[CrossRef][Medline] [Order article via Infotrieve]
  37. Beraud-Dufour, S., Robineau, S., Chardin, P., Paris, S., Chabre, M., Cherfils, J., and Antonny, B. (1998) EMBO J. 17, 3651-3659[CrossRef][Medline] [Order article via Infotrieve]
  38. Chavrier, P., and Franco, M. (2001) Methods Enzymol. 329, 272-279[CrossRef][Medline] [Order article via Infotrieve]
  39. Archambaud, S., Aphecetche-Julienne, K., and Guingant, A. (2005) Synlett 139-143
  40. Antonny, B., Beraud-Dufour, S., Chardin, P., and Chabre, M. (1997) Biochemistry 36, 4675-4684[CrossRef][Medline] [Order article via Infotrieve]
  41. Brunet, N., Morin, A., and Olofsson, B. (2002) Traffic 3, 342-357[CrossRef][Medline] [Order article via Infotrieve]
  42. Dransart, E., Morin, A., Cherfils, J., and Olofsson, B. (2005) J. Biol. Chem. 280, 4674-4683[Abstract/Free Full Text]
  43. MacKerell, J., A. D., Bashford, D., Bellott, M., Dunbrack, Jr., R. L., Evanseck, J. D., Field, M. J., Fischer, S., Gao, J., Guo, H., Ha, S., Joseph-McCarthy, D., Kuchnir, L., Kuczera, K., Lau, F. T. K., Mattos, C., Michnick, S., Ngo, T., Nguyen, D. T., Prodhom, B., Reiher, I., W. E., Roux, B., Schlenkrich, M., Smith, J. C., Stote, R., Straub, J., Watanabe, M., Wiorkiewicz-Kuczera, J., Yin, D., and Karplus, M. (1998) J. Phys. Chem. B 102, 3586-3616
  44. Brooks, B. R., Bruccoleri, R. E., Olafson, B. D., States, D. J., Swaminathan, S., and Karplus, M. (1983) J. Comput. Chem. 4, 187-217[CrossRef]
  45. Hendsch, Z. S., and Tidor, B. (1994) Protein Sci. 3, 211-226[Medline] [Order article via Infotrieve]
  46. Hendsch, Z. S., and Tidor, B. (1999) Protein Sci. 8, 1381-1392[Medline] [Order article via Infotrieve]
  47. Pasqualato, S., Renault, L., and Cherfils, J. (2004) in The ARF Book (Kahn, R. A., ed) pp. 23-48, Kluwer Academic Publishers, Norwell, MA
  48. Menetrey, J., Macia, E., Pasqualato, S., Franco, M., and Cherfils, J. (2000) Nat. Struct. Biol. 7, 466-469[CrossRef][Medline] [Order article via Infotrieve]
  49. Macia, E., Chabre, M., and Franco, M. (2001) J. Biol. Chem. 276, 24925-24930[Abstract/Free Full Text]
  50. Chardin, P., Paris, S., Antonny, B., Robineau, S., Beraud-Dufour, S., Jackson, C. L., and Chabre, M. (1996) Nature 384, 481-484[CrossRef][Medline] [Order article via Infotrieve]
  51. Morinaga, N., Adamik, R., Moss, J., and Vaughan, M. (1999) J. Biol. Chem. 274, 17417-17423[Abstract/Free Full Text]
  52. Cox, R., Mason-Gamer, R. J., Jackson, C. L., and Segev, N. (2004) Mol. Biol. Cell 15, 1487-1505[Abstract/Free Full Text]
  53. Steinmann, T., Geldner, N., Grebe, M., Mangold, S., Jackson, C. L., Paris, S., Galweiler, L., Palme, K., and Jurgens, G. (1999) Science 286, 316-318[Abstract/Free Full Text]
  54. Zhu, J. W., Nagasawa, H., Nagura, F., Mohamad, S. B., Uto, Y., Ohkura, K., and Hori, H. (2000) Bioorg. Med. Chem. 8, 455-463[CrossRef][Medline] [Order article via Infotrieve]

Add to CiteULike CiteULike   Add to Complore Complore   Add to Connotea Connotea   Add to Del.icio.us Del.icio.us   Add to Digg Digg   Add to Reddit Reddit   Add to Technorati Technorati    What's this?


This article has been cited by other articles:


Home page
 J. Biol. Chem.Home page
G. Moroy, E. Martin, A. Dejaegere, and R. H. Stote
Molecular Basis for Bcl-2 Homology 3 Domain Recognition in the Bcl-2 Protein Family: IDENTIFICATION OF CONSERVED HOT SPOT INTERACTIONS
J. Biol. Chem., June 26, 2009; 284(26): 17499 - 17511.
[Abstract] [Full Text] [PDF]

Home page
 Appl. Environ. Microbiol.Home page
P. Shivshankar, L. Lei, J. Wang, and G. Zhong
Rottlerin Inhibits Chlamydial Intracellular Growth and Blocks Chlamydial Acquisition of Sphingolipids from Host Cells
Appl. Envir. Microbiol., February 15, 2008; 74(4): 1243 - 1249.
[Abstract] [Full Text] [PDF]

Home page
 J. Biol. Chem.Home page
C. Browning, E. Martin, C. Loch, J.-M. Wurtz, D. Moras, R. H. Stote, A. P. Dejaegere, and I. M. L. Billas
Critical Role of Desolvation in the Binding of 20-Hydroxyecdysone to the Ecdysone Receptor
J. Biol. Chem., November 9, 2007; 282(45): 32924 - 32934.
[Abstract] [Full Text] [PDF]

Home page
 J. Physiol.Home page
C. Wang, Y. Gu, G.-W. Li, and L.-Y. M. Huang
A critical role of the cAMP sensor Epac in switching protein kinase signalling in prostaglandin E2-induced potentiation of P2X3 receptor currents in inflamed rats
J. Physiol., October 1, 2007; 584(1): 191 - 203.
[Abstract] [Full Text] [PDF]

Home page
 J. Biol. Chem.Home page
O. Ramaen, A. Joubert, P. Simister, N. Belgareh-Touze, M. C. Olivares-Sanchez, J.-C. Zeeh, S. Chantalat, M.-P. Golinelli-Cohen, C. L. Jackson, V. Biou, et al.
Interactions between Conserved Domains within Homodimers in the BIG1, BIG2, and GBF1 Arf Guanine Nucleotide Exchange Factors
J. Biol. Chem., September 28, 2007; 282(39): 28834 - 28842.
[Abstract] [Full Text] [PDF]

Home page
 J. Cell Sci.Home page
M. Tanaka, K. Sasaki, R. Kamata, and R. Sakai
The C-terminus of ephrin-B1 regulates metalloproteinase secretion and invasion of cancer cells
J. Cell Sci., July 1, 2007; 120(13): 2179 - 2189.
[Abstract] [Full Text] [PDF]

Home page
 Proc. Natl. Acad. Sci. USAHome page
J. Viaud, M. Zeghouf, H. Barelli, J.-C. Zeeh, A. Padilla, B. Guibert, P. Chardin, C. A. Royer, J. Cherfils, and A. Chavanieu
Structure-based discovery of an inhibitor of Arf activation by Sec7 domains through targeting of protein-protein complexes
PNAS, June 19, 2007; 104(25): 10370 - 10375.
[Abstract] [Full Text] [PDF]

Home page
 J. Biol. Chem.Home page
M. E. Yohe, K. L. Rossman, O. S. Gardner, A. E. Karnoub, J. T. Snyder, S. Gershburg, L. M. Graves, C. J. Der, and J. Sondek
Auto-inhibition of the Dbl Family Protein Tim by an N-terminal Helical Motif
J. Biol. Chem., May 4, 2007; 282(18): 13813 - 13823.
[Abstract] [Full Text] [PDF]

This Article
Right arrow Abstract Freely available
Right arrow Full Text (PDF)
Right arrow All Versions of this Article:
281/17/11805    most recent
M600149200v1
Right arrow Submit a Letter to Editor
Right arrow Alert me when this article is cited
Right arrow Alert me when eLetters are posted
Right arrow Alert me if a correction is posted
Right arrow Citation Map
Services
Right arrow Email this article to a friend
Right arrow Similar articles in this journal
Right arrow Similar articles in PubMed
Right arrow Alert me to new issues of the journal
Right arrow Download to citation manager
Right arrowRequest Permissions
Citing Articles
Right arrow Citing Articles via HighWire
Right arrow Citing Articles via Google Scholar
Google Scholar
Right arrow Articles by Zeeh, J.-C.
Right arrow Articles by Cherfils, J.
Right arrow Search for Related Content
PubMed
Right arrow PubMed Citation
Right arrow Articles by Zeeh, J.-C.
Right arrow Articles by Cherfils, J.
Social Bookmarking
Add to CiteULike   Add to Complore   Add to Connotea   Add to Del.icio.us   Add to Digg   Add to Reddit   Add to Technorati  
What's this?

HOME HELP FEEDBACK SUBSCRIPTIONS ARCHIVE SEARCH TABLE OF CONTENTS
 All ASBMB Journals   Molecular and Cellular Proteomics 
 Journal of Lipid Research   ASBMB Today 
Copyright © 2006 by the American Society for Biochemistry and Molecular Biology.
Advertisement
spacer
Advertisement
Advertisement