Vincent Chaptal

Understanding polyspecificity within the substrate‐binding cavity of the human multidrug resistance P‐glycoprotein

Human P‐glycoprotein (P‐gp) controls drugs bioavailability by pumping structurally unrelated drugs out of cells. The X‐ray structure of the mouse P‐gp ortholog has been solved, with two SSS enantiomers or one RRR enantiomer of the selenohexapeptide inhibitor QZ59, found within the putative drug‐binding pocket (Aller SG, Yu J, Ward A, Weng Y, Chittaboina S, Zhuo R, Harrell PM, Trinh YT, Zhang Q, Urbatsch IL et al. (2009). Science 323, 1718–1722). This offered the first opportunity to localize the well‐known H and R drug‐binding sites with respect to the QZ59 inhibition mechanisms of Hoechst 33342 and daunorubicin transports, characterized here in cellulo. We found that QZ59‐SSS competes efficiently with both substrates, with KI,app values of 0.15 and 0.3 μm, which are 13 and 2 times lower, respectively, than the corresponding Km,app values. In contrast, QZ59‐RRR non‐competitively inhibited daunorubicin transport with moderate efficacy (KI,app = 1.9 μm); it also displayed a mixed‐type inhibition of the Hoechst 33342 transport, resulting from a main non‐competitive tendency (Ki2,app = 1.6 μm) and a limited competitive tendency (Ki1,app = 5 μm). These results suggest a positional overlap of QZ59 and drugs binding sites: full for the SSS enantiomer and partial for the RRR enantiomer. Crystal structure analysis suggests that the H site overlaps both QZ59‐SSS locations while the R site overlaps the most embedded location.

Quantification of Detergents Complexed with Membrane Proteins.

Most membrane proteins studies require the use of detergents, but because of the lack of a general, accurate and rapid method to quantify them, many uncertainties remain that hamper proper functional and structural data analyses. To solve this problem, we propose a method based on matrix-assisted laser desorption/ionization mass spectrometry (MALDI-TOF MS) that allows quantification of pure or mixed detergents in complex with membrane proteins. We validated the method with a wide variety of detergents and membrane proteins. We automated the process, thereby allowing routine quantification for a broad spectrum of usage. As a first illustration, we show how to obtain information of the amount of detergent in complex with a membrane protein, essential for liposome or nanodiscs reconstitutions. Thanks to the method, we also show how to reliably and easily estimate the detergent corona diameter and select the smallest size, critical for favoring protein-protein contacts and triggering/promoting membrane protein crystallization, and to visualize the detergent belt for Cryo-EM studies.

Structural analysis of B. subtilis CcpA effector binding site

Vincent Chaptal, Virginie Gueguen-Chaignon, Sandrine Poncet, Cécile Lecampion, Philippe Meyer, Josef Deutscher, Anne Galinier, Sylvie Nessler,* and Solange Moréra* Laboratoire d’Enzymologie et Biochimie Structurales, CNRS FRE 2930, Gif-sur-Yvette, France Laboratoire de Génétique des Microorganismes, INRA-CNRS URA 1925, Thiverval-Grignon, France Laboratoire de Chimie Bactérienne, CNRS UPR 9043, Institut de Biologie Structurale et Microbiologie, Marseille, France

Glycosyl-Substituted Dicarboxylates as Detergents for the Extraction, Overstabilization, and Crystallization of Membrane Proteins

To tackle the problems associated with membrane protein (MP) instability in detergent solutions, we designed a series of glycosyl-substituted dicarboxylate detergents (DCODs) in which we optimized the polar head to clamp the membrane domain by including, on one side, two carboxyl groups that form salt bridges with basic residues abundant at the membrane-cytoplasm interface of MPs and, on the other side, a sugar to form hydrogen bonds. Upon extraction, the DCODs 8u2009b, 8u2009c, and 9u2009b preserved the ATPase function of BmrA, an ATP-binding cassette pump, much more efficiently than reference or recently designed detergents. The DCODs 8u2009a, 8u2009b, 8u2009f, 9u2009a, and 9u2009b induced thermal shifts of 20 to 29u2009°C for BmrA and of 13 to 21u2009°C for the native version of the G-protein-coupled adenosine receptor A2A R. Compounds 8u2009f and 8u2009g improved the diffraction resolution of BmrA crystals from 6 to 4u2005Å. DCODs are therefore considered to be promising and powerful tools for the structural biology of MPs.

Atomic modelling and systematic mutagenesis identify residues in multiple drug binding sites that are essential for drug resistance in the major Candida transporter Cdr1.

The ABC (ATP-Binding Cassette) transporter Cdr1 (Candida drug resistance 1) protein (Cdr1p) of Candida albicans, shows promiscuity towards the substrate it exports and plays a major role in antifungal resistance. It has two transmembrane domains (TMDs) comprising of six transmembrane helices (TMH) that envisage and confer the substrate specificity and two nucleotide binding domains (NBDs), interconnected by extracellular loops (ECLs) and intracellular loops (ICLs) Cdr1p. This study explores the diverse substrate specificity spectrum to get a deeper insight into the structural and functional features of Cdr1p. By screening with the variety of compounds towards an in-house TMH 252 mutant library of Cdr1p, we establish new substrates of Cdr1p. The localization of substrate-susceptible mutants in an ABCG5/G8 homology model highlights the common and specific binding pockets inside the membrane domain, where rhodamines and tetrazoliums mainly engage the N-moiety of Cdr1p, binding between TMH 2, 11 and surrounded by TMH 1, 5. Whereas, tin chlorides involve both N and C moieties located at the interface of TMH 2, 11, 1 and 5. Further, screening of the in house TMH mutant library of Cdr1p displays the TMH12 interaction with tetrazolium chloride, trimethyltin chloride and a Ca2+ ionophore, A23187. In silico localization reveals a binding site at the TMH 12, 9 and 10 interface, which is widely exposed to the lipid interface. Together, for the first time, our study shows the molecular localization of Cdr1p substrates-binding sites and demonstrates the participation of TMH12 in a peripheral drug binding site.

Au courant computation of the PDB to audit diffraction anisotropy of soluble and membrane proteins

This data article makes available the informed computation of the whole Protein Data Bank (PDB) to investigate diffraction anisotropy on a large scale and to perform statistics. This data has been investigated in detail in “X-ray diffraction reveals the intrinsic difference in the physical properties of membrane and soluble proteins” [1]. Diffraction anisotropy is traditionally associated with absence of contacts in-between macromolecules within the crystals in a given direction of space. There are however many case that do not follow this empirical rule. To investigate and sort out this discrepancy, we computed diffraction anisotropy for every entry of the PDB, and put them in context of relevant metrics to compare X-ray diffraction in reciprocal space to the crystal packing in real space. These metrics were either extracted from PDB files when available (resolution, space groups, cell parameters, solvent content), or calculated using standard procedures (anisotropy, crystal contacts, presence of ligands). More specifically, we separated entries to compare soluble vs membrane proteins, and further separated the later in subcategories according to their insertion in the membrane, function, or type of crystallization (Type I vs Type II crystal packing). This informed database is being made available to investigators in the raw and curated formats that can be re-used for further downstream studies. This dataset is useful to test ideas and to ascertain hypothesis based on statistical analysis.

X-ray diffraction reveals the intrinsic difference in the physical properties of membrane and soluble proteins

Membrane proteins are distinguished from soluble proteins by their insertion into biological membranes. This insertion is achieved via a noticeable arrangement of hydrophobic amino acids that are exposed at the surface of the protein, and renders the interaction with the aliphatic tails of lipids more energetically favorable. This important difference between these two categories of proteins is the source of the need for a specific handling of membrane proteins, which transpired in the creation of new tools for their recombinant expression, purification and even crystallization. Following this line, we show here that crystals of membrane proteins display systematically higher diffraction anisotropy than those of soluble proteins. This phenomenon dramatically hampers structure solution and refinement, and has a strong impact on the quality of electron-density maps. A farther search for origins of this phenomenon showed that the type of crystallization, and thus the crystal packing, has no impact on anisotropy, nor does the nature or function of the membrane protein. Membrane proteins fully embedded within the membrane display equal anisotropy compared to the ones with extra membranous domains or fusions with soluble proteins. Overall, these results overturn common beliefs and call for a specific handling of their diffraction data.

Multidrug ABC transporter Cdr1 of Candida albicans harbors specific and overlapping binding sites for human steroid hormones transport

The present study examines the kinetics of steroids efflux mediated by the Candida drug resistance protein 1 (Cdr1p) and evaluates their interaction with the protein. We exploited our in-house mutant library for targeting the 252 residues forming the twelve transmembrane helices (TMHs) of Cdr1p. The screening revealed 65 and 58 residues critical for β-estradiol and corticosterone transport, respectively. Notably, up to 83% critical residues for corticosterone face the lipid interface compared to 54% for β-estradiol. Molecular docking identified a possible peripheral corticosterone-binding site made of 8/14 critical/non-critical residues between TMHs 3, 4 and 6. β-estradiol transport was severely hampered by alanine replacements of Cdr1p core residues involving TMHs 2, 5 and 8, in a binding site made of 10/14 critical residues mainly shared with rhodamine 6G with which it competes. By contrast, TMH11 was poorly impacted, although being part of the core domain. Finally, we observed the presence of several contiguous stretches of 3-5 critical residues in TMHs 2, 5 and 10 that points to a rotation motion of these helices during the substrate transport cycle. The selective structural arrangement of the steroid-binding pockets in the core region and at the lipid-TMD interface, which was never reported before, together with the possible rotation of some TMHs may be the structural basis of the drug-transport mechanism achieved by these type II ABC transporters.

Crystal structure and functional analysis identify the P-loop containing protein YFH7 of Saccharomyces cerevisiae as an ATP-dependent kinase.

Genome sequencing projects have revealed that P‐loop proteins are highly represented in all organisms and that many of them have no attributed function. They are characterized by a conserved nucleotide‐binding domain and carry different activities implicated in many cellular processes. Saccharomyces cerevisiae YFH7 is one of these P‐loop proteins of unknown function. In this work we tried to integrate bioinformatics, structure, and enzymology to discover the function of YFH7. Sequence analysis revealed that yeast YFH7 is a yeast‐specific protein showing weak similarity with the phosphoribulokinase/uridine kinase/bacterial pantothenate kinase (PRK/URK/PANK) subfamily of P‐loop containing kinases. A large insertion of about 100 residues distinguishes YFH7 from other members of the family. The 1.95 Å resolution crystal structure of YFH7 solved using the SAD method confirmed that YFH7 has a fold similar to the PRK/URK/PANK family, with the characteristic core, lid, and NMPbind domains. An additional α/β domain of novel topology corresponds to the large sequence insertion. Structural and ligand binding analysis combined with enzymatic assays suggest that YFH7 is an ATP‐dependent small molecule kinase with new substrate specificity. Proteins 2008.

X-ray structure of a domain-swapped dimer of Ser46-phosphorylated Crh from Bacillus subtilis.

Introduction. In Bacillus subtilis, approximately 10% of the genome is regulated by adenosine 5 -triphosphate (ATP)-dependent phosphorylation of Ser46 of HPr and its homolog Crh (for catabolite repression HPr). The two proteins exhibit 45% sequence identity but residue 15 is a phosphorylable histidine in HPr and a glutamine in Crh. PEP-dependent phosphorylation of HPr His15 by enzyme I is the first step of the sugar phosphotransferase system called PTS in bacteria. Crh does not participate in the PTS. In contrary, the ATP-dependent HPr kinase/phosphorylase (HprK/P) efficiently phosphorylates the conserved Ser46 of Crh. The seryl-phosphorylated PserHPr and PserCrh act as alternate corepressors of the catabolite control protein A (CcpA). CcpA is the central regulator of a fundamental bacterial signal transduction pathway called carbon catabolite repression. The X-ray structure of unphosphorylated Crh is a domain-swapped dimer. Nuclear magnetic resonance (NMR) studies showed that Crh and PserCrh are both in slow monomer/dimer equilibrium. The X-ray structure of complexed HprK/P with HPr or PserHPr shows that the hexameric enzyme binds six molecules of monomeric protein substrate. The structure of CcpA in complex with PserHPr reveals a PserHPr monomer bound to each subunit of the CcpA dimer.