Structural Adaptations that Modulate Monosaccharide, Disaccharide, and Trisaccharide Specificities in Periplasmic Maltose-Binding Proteins

Periplasmic binding proteins comprise a superfamily that is present in archaea, prokaryotes, and eukaryotes. Periplasmic binding protein ligand-binding sites have diversified to bind a wide variety of ligands. Characterization of the structural mechanisms by which functional adaptation occurs is key to understanding the evolution of this important protein superfamily. Here we present the structure and ligand-binding properties of a maltotriose-binding protein identified from the Thermus thermophilus genome sequence. We found that this receptor has a high affinity for the trisaccharide maltotriose (K_d < 1 μM) but little affinity for disaccharides that are transported by a paralogous maltose transport operon present in T. thermophilus. Comparison of this structure to other proteins that adopt the maltose-binding protein fold but bind monosaccharides, disaccharides, or trisaccharides reveals the presence of four subsites that bind individual glucose ring units. Two loops and three helical segments encode adaptations that control the presence of each subsite by steric blocking or hydrogen bonding. We provide a model in which the energetics of long-range conformational equilibria controls subsite occupancy and ligand binding.     

Crystal structures of the bacterial solute receptor AcbH displaying an exclusive substrate preference for β-D-galactopyranose

Solute receptors (binding proteins) are indispensable components of canonical ATP-binding cassette importers in prokaryotes. Here, we report on the characterization and crystal structures in the closed and open conformations of AcbH, the solute receptor of the putative carbohydrate transporter AcbFG which is encoded in the acarbose (acarviosyl-1,4-maltose) biosynthetic gene cluster from Actinoplanes sp. SE50/110. Binding assays identified AcbH as a high-affinity monosaccharide-binding protein with a dissociation constant (K_d) for β-d-galactopyranose of 9.8 ± 1.0 nM. Neither galactose-containing di- and trisaccharides, such as lactose and raffinose, nor monosaccharides including d-galacturonic acid, l-arabinose, d-xylose and l-rhamnose competed with [¹⁴C]galactose for binding to AcbH. Moreover, AcbH does not bind d-glucose, which is a common property of all but one d-galactose-binding proteins characterized to date. Strikingly, determination of the X-ray structure revealed that AcbH is structurally homologous to maltose-binding proteins rather than to glucose-binding proteins. Two helices are inserted in the substrate-binding pocket, which reduces the cavity size and allows the exclusive binding of monosaccharides, specifically β-d-galactopyranose, in the ⁴C₁ conformation. Site-directed mutagenesis of three residues from the binding pocket (Arg82, Asp361 and Arg362) that interact with the axially oriented O4-H hydroxyl of the bound galactopyranose and subsequent functional analysis indicated that these residues are crucial for galactose binding. To our knowledge, this is the first report of the tertiary structure of a solute receptor with exclusive affinity for β-d-galactopyranose. The putative role of a galactose import system in the context of acarbose metabolism in Actinoplanes sp. is discussed.     

The crystal structure of a thermophilic glucose binding protein reveals adaptations that interconvert mono and di-saccharide binding sites

Periplasmic binding proteins (PBPs) comprise a protein superfamily that is involved in prokaryotic solute transport and chemotaxis. These proteins have been used to engineer reagentless biosensors to detect natural or non-natural ligands. There is considerable interest in obtaining very stable members of this superfamily from thermophilic bacteria to use as robust engineerable parts in biosensor development. Analysis of the recently determined genome sequence of Thermus thermophilus revealed the presence of more than 30 putative PBPs in this thermophile. One of these is annotated as a glucose binding protein (GBP) based on its genetic linkage to genes that are homologous to an ATP-binding cassette glucose transport system, although the PBP sequence is homologous to periplasmic maltose binding proteins (MBPs). Here we present the cloning, over-expression, characterization of cognate ligands, and determination of the X-ray crystal structure of this gene product. We find that it is a very stable (apo-protein Tm value is 100(+/- 2) degrees C; complexes 106(+/- 3) degrees C and 111(+/- 1) degrees C for glucose and galactose, respectively) glucose (Kd value is 0.08(+/- 0.03) microM) and galactose (Kd value is 0.94(+/- 0.04) microM) binding protein. Determination of the X-ray crystal structure revealed that this T. thermophilus glucose binding protein (ttGBP) is structurally homologous to MBPs rather than other GBPs. The di or tri-saccharide ligands in MBPs are accommodated in long relatively shallow grooves. In the ttGBP binding site, this groove is partially filled by two loops and an alpha-helix, which create a buried binding site that allows binding of only monosaccharides. Comparison of ttGBP and MBP provides a clear example of structural adaptations by which the size of ligand binding sites can be controlled in the PBP super family.     

The Crystal Structure of the Hinge Domain of the Escherichia coli Structural Maintenance of Chromosomes Protein MukB

MukB, a divergent structural maintenance of chromosomes (SMC) protein, is important for chromosomal segregation and condensation in γ-proteobacteria. MukB and canonical SMC proteins share a characteristic five-domain structure. Globular N- and C-terminal domains interact to form an ATP-binding cassette-like ATPase or “head” domain, which is connected to a smaller dimerization or “hinge” domain by a long, antiparallel coiled coil. In addition to mediating dimerization, this hinge region has been implicated in both conformational flexibility and dynamic protein-DNA interactions. We report here the first crystallographic model of the MukB hinge domain. This model also contains approximately 20% of the coiled-coil domain, including an unusual coiled-coil deviation. These results will facilitate studies to clarify the roles of both the hinge and the coiled-coil domains in MukB function.     

Crystal Structure of the Periplasmic Component of a Tripartite Macrolide-Specific Efflux Pump

In Gram-negative bacteria, type I protein secretion systems and tripartite drug efflux pumps have a periplasmic membrane fusion protein (MFP) as an essential component. MFPs bridge the outer membrane factor and an inner membrane transporter, although the oligomeric state of MFPs remains unclear. The most characterized MFP AcrA connects the outer membrane factor TolC and the resistance-nodulation-division-type efflux transporter AcrB, which is a major multidrug efflux pump in Escherichia coli. MacA is the periplasmic MFP in the MacAB-TolC pump, where MacB was characterized as a macrolide-specific ATP-binding-cassette-type efflux transporter. Here, we report the crystal structure of E. coli MacA and the experimentally phased map of Actinobacillus actinomycetemcomitans MacA, which reveal a domain orientation of MacA different from that of AcrA. Notably, a hexameric assembly of MacA was found in both crystals, exhibiting a funnel-like structure with a central channel and a conical mouth. The hexameric MacA assembly was further confirmed by electron microscopy and functional studies in vitro and in vivo. The hexameric structure of MacA provides insight into the oligomeric state in the functional complex of the drug efflux pump and type I secretion system.     

The CcmC:Heme:CcmE Complex in Heme Trafficking and Cytochrome c Biosynthesis

A superfamily of integral membrane proteins is characterized by a conserved tryptophan-rich region (called the WWD domain) in an external loop at the inner membrane surface. The three major members of this family (CcmC, CcmF, and CcsBA) are each involved in cytochrome c biosynthesis, yet the function of the WWD domain is unknown. It has been hypothesized that the WWD domain binds heme to present it to an acceptor protein (apoCcmE for CcmC or apocytochrome c for CcmF and CcsBA) such that the heme vinyl group(s) covalently attaches to the acceptors. Alternative proposals suggest that the WWD domain interacts directly with the acceptor protein (e.g., apoCcmE for CcmC). Here, it is shown that CcmC is only trapped with heme when its cognate acceptor protein CcmE is present. It is demonstrated that CcmE only interacts stably with CcmC when heme is present; thus, specific residues in each protein provide sites of interaction with heme to form this very stable complex. For the first time, evidence that the external WWD domain of CcmC interacts directly with heme is presented. Single and multiple substitutions of completely conserved residues in the WWD domain of CcmC alter the spectral properties of heme in the stable CcmC:heme:CcmE complexes. Moreover, some mutations reduce the binding of heme up to 100%. It is likely that endogenously synthesized heme enters the external WWD domain of CcmC either via a channel within this six-transmembrane-spanning protein or from the membrane. The data suggest that a specific heme channel (i.e., heme binding site within membrane spanning helices) is not present in CcmC, in contrast to the CcsBA protein. We discuss the likelihood that it is not important to protect the heme via trafficking in CcmC whereas it is critical in CcsBA.     

Discoidin I from Dictyostelium discoideum and Interactions with Oligosaccharides: Specificity, Affinity, Crystal Structures, and Comparison with Discoidin II

Discoidin I (DiscI) and discoidin II (DiscII) are N-acetylgalactosamine (GalNAc)-binding proteins from Dictyostelium discoideum. They consist of two domains: an N-terminal discoidin domain and a C-terminal H-type lectin domain. They were cloned and expressed in high yield in recombinant form in Escherichia coli. Although both lectins bind galactose (Gal) and GalNAc, glycan array experiments performed on the recombinant proteins displayed strong differences in their specificity for oligosaccharides. DiscI and DiscII bind preferentially to Gal/GalNAcβ1-3Gal/GalNAc-containing and Gal/GalNAcβ1-4GlcNAcβ1-6Gal/GalNAc-containing glycans, respectively. The affinity of the interaction of DiscI with monosaccharides and disaccharides was evaluated using isothermal titration calorimetry experiments. The three-dimensional structures of native DiscI and its complexes with GalNAc, GalNAcβ1-3Gal, and Galβ1-3GalNAc were solved by X-ray crystallography. DiscI forms trimers with involvement of calcium at the monomer interface. The N-terminal discoidin domain presents a structural similarity to F-type lectins such as the eel agglutinin, where an amphiphilic binding pocket suggests possible carbohydrate-binding activity. In the C-terminal H-type lectin domain, the GalNAc residue establishes specific hydrogen bonds that explain the observed affinity (K_d = 3 × 10^− 4 M). The different specificities of DiscI and DiscII for oligosaccharides were rationalized from the different structures obtained by either X-ray crystallography or molecular modeling.     

Orchestration of Haemophilus influenzae RecJ Exonuclease by Interaction with Single-Stranded DNA-Binding Protein

RecJ exonuclease plays crucial roles in several DNA repair and recombination pathways, and its ubiquity in bacterial species points to its ancient origin and vital cellular function. RecJ exonuclease from Haemophilus influenzae is a 575-amino-acid protein that harbors the characteristic motifs conserved among RecJ homologs. The purified protein exhibits a processive 5′-3′ single-stranded-DNA-specific exonuclease activity. The exonuclease activity of H. influenzae RecJ (HiRecJ) was supported by Mg^2 + or Mn^2 + and inhibited by Cd^2 +, suggesting a different mode of metal binding in HiRecJ as compared to Escherichia coli RecJ (EcoRecJ). Site-directed mutagenesis of highly conserved residues in HiRecJ abolished enzymatic activity. Interestingly, substitution of alanine for aspartate 77 resulted in a catalytically inactive enzyme that bound to DNA with a significantly higher affinity as compared to the wild-type enzyme. Noticeably, steady-state kinetic studies showed that H. influenzae single-stranded DNA-binding protein (HiSSB) increased the affinity of HiRecJ for single-stranded DNA and stimulated its exonuclease activity. HiSSB, whose C-terminal tail had been deleted, failed to enhance RecJ exonuclease activity. More importantly, HiRecJ was found to directly associate with its cognate single-stranded DNA-binding protein (SSB), as demonstrated by various in vitro assays. Interaction studies carried out with the truncated variants of HiRecJ and HiSSB revealed that the two proteins interact via the C-terminus of SSB protein and the core-catalytic domain of RecJ. Taken together, these results emphasize direct interaction between RecJ and SSB, which confers functional cooperativity to these two proteins. In addition, these results implicate SSB as being involved in the recruitment of RecJ to DNA and provide insights into the interplay between these proteins in repair and recombination pathways.     

M. tuberculosis Pantothenate Kinase: Dual Substrate Specificity and Unusual Changes in Ligand Locations

Kinetic measurements of enzyme activity indicate that type I pantothenate kinase from Mycobacterium tuberculosis has dual substrate specificity for ATP and GTP, unlike the enzyme from Escherichia coli, which shows a higher specificity for ATP. A molecular explanation for the difference in the specificities of the two homologous enzymes is provided by the crystal structures of the complexes of the M. tuberculosis enzyme with (1) GMPPCP and pantothenate, (2) GDP and phosphopantothenate, (3) GDP, (4) GDP and pantothenate, (5) AMPPCP, and (6) GMPPCP, reported here, and the structures of the complexes of the two enzymes involving coenzyme A and different adenyl nucleotides reported earlier. The explanation is substantially based on two critical substitutions in the amino acid sequence and the local conformational change resulting from them. The structures also provide a rationale for the movement of ligands during the action of the mycobacterial enzyme. Dual specificity of the type exhibited by this enzyme is rare. The change in locations of ligands during action, observed in the case of the M. tuberculosis enzyme, is unusual, so is the striking difference between two homologous enzymes in the geometry of the binding site, locations of ligands, and specificity. Furthermore, the dual specificity of the mycobacterial enzyme appears to have been caused by a biological necessity.     

Crystal structure of a glutamate/aspartate binding protein complexed with a glutamate molecule: structural basis of ligand specificity at atomic resolution

The crystal structure of a periplasmic l-aspartate/l-glutamate binding protein (DEBP) from Shigella flexneri complexed with an l-glutamate molecule has been determined and refined to an atomic resolution of 1.0 Å. There are two DEBP molecules in the asymmetric unit. The refined model contains 4462 non-hydrogen protein atoms, 730 water molecules, 2 bound glutamate molecules, and 2 Tris molecules from the buffer used in crystallization. The final R_cryst and R_free factors are 13.61% and 16.89%, respectively. The structure has root-mean-square deviations of 0.016 Å from standard bond lengths and 2.35° from standard bond angles.The DEBP molecule is composed of two similarly folded domains separated by the ligand binding region. Both domains contain a central five-stranded β-sheet that is surrounded by several α-helices. The two domains are linked by two antiparallel β-strands. The overall shape of DEBP is that of an ellipsoid approximately 55 Å × 45 Å × 40 Å in size.The binding of ligand to DEBP is achieved mostly through hydrogen bonds between the glutamate and side-chain and main-chain groups of DEBP. Side chains of residues Arg24, Ser72, Arg75, Ser90, and His164 anchor the deprotonated γ-carboxylate group of the glutamate with six hydrogen bonds. Side chains of Arg75 and Arg90 form salt bridges with the deprotonated α-carboxylate group, while the main-chain amide groups of Thr92 and Thr140 form hydrogen bonds with the same group. The positively charged α-amino group of the l-glutamate forms salt bridge interaction with the side-chain carboxylate group of Asp182 and hydrogen bond interaction with main-chain carbonyl oxygen of Ser90. In addition to these hydrogen bond and electrostatic interactions, other interactions may also play important roles. For example, the two methylene groups from the glutamate form van der Waals interactions with hydrophobic side chains of DEBP.Comparisons with several other periplasmic amino acid binding proteins indicate that DEBP residues involved in the binding of α-amino and α-carboxylate groups of the ligand and the pattern of hydrogen bond formation between these groups are very well conserved, but the binding pocket around the ligand side chain is not, leading to the specificity of DEBP. We have identified structural features of DEBP that determine its ability of binding glutamate and aspartate, two molecules with different sizes, but discriminating against very similar glutamine and asparagine molecules.     

Crystal Structure Determination and Functional Characterization of the Metallochaperone SlyD from Thermus thermophilus

SlyD (sensitive to lysis D; product of the slyD gene) is a prolyl isomerase [peptidyl-prolyl cis/trans isomerase (PPIase)] of the FK506 binding protein (FKBP) type with chaperone properties. X-ray structures derived from three different crystal forms reveal that SlyD from Thermus thermophilus consists of two domains representing two functional units. PPIase activity is located in a typical FKBP domain, whereas chaperone function is associated with the autonomously folded insert-in-flap (IF) domain. The two isolated domains are stable and functional in solution, but the presence of the IF domain increases the PPIase catalytic efficiency of the FKBP domain by 2 orders of magnitude, suggesting that the two domains act synergistically to assist the folding of polypeptide chains. The substrate binding surface of SlyD from T. thermophilus was mapped by NMR chemical shift perturbations to hydrophobic residues of the IF domain, which exhibits significantly reduced thermodynamic stability according to NMR hydrogen/deuterium exchange and fluorescence equilibrium transition experiments. Based on structural homologies, we hypothesize that this is due to the absence of a stabilizing β-strand, suggesting in turn a mechanism for chaperone activity by ‘donor-strand complementation.’ Furthermore, we identified a conserved metal (Ni²⁺) binding site at the C-terminal SlyD-specific helical appendix of the FKBP domain, which may play a role in metalloprotein assembly.     

Recycling of the Posttermination Complexes of Mycobacterium smegmatis and Escherichia coli Ribosomes Using Heterologous Factors

In eubacteria, ribosome recycling factor (RRF) and elongation factor G (EFG) function together to dissociate posttermination ribosomal complexes. Earlier studies, using heterologous factors from Mycobacterium tuberculosis in Escherichia coli revealed that specific interactions between RRF and EFG are crucial for their function in ribosome recycling. Here, we used translation factors from E. coli, Mycobacterium smegmatis and M. tuberculosis, and polysomes from E. coli and M. smegmatis, and employed in vivo and in vitro experiments to further understand the role of EFG in ribosome recycling. We show that E. coli EFG (EcoEFG) recycles E. coli ribosomes with E. coli RRF (EcoRRF), but not with mycobacterial RRFs. Also, EcoEFG fails to recycle M. smegmatis ribosomes with either EcoRRF or mycobacterial RRFs. On the other hand, mycobacterial EFGs recycle both E. coli and M. smegmatis ribosomes with either of the RRFs. These observations suggest that EFG establishes distinct interactions with RRF and the ribosome to carry out ribosome recycling. Furthermore, the EFG chimeras generated by swapping domains between mycobacterial EFGs and EcoEFG suggest that while the residues needed to specify the EFG interaction with RRF are located in domains IV and V, those required to specify its interaction with the ribosome are located throughout the molecule.     

Crystal Structure of Bacillus cereusd-Alanyl Carrier Protein Ligase (DltA) in Complex with ATP

d-Alanylation of lipoteichoic acids modulates the surface charge and ligand binding of the Gram-positive cell wall. Disruption of the bacterial dlt operon involved in teichoic acid alanylation, as well as inhibition of the DltA (d-alanyl carrier protein ligase) protein, has been shown to render the bacterium more susceptible to conventional antibiotics and host defense responses. The DltA catalyzes the adenylation and thiolation reactions of d-alanine. This enzyme belongs to a superfamily of AMP-forming domains such as the ubiquitous acetyl-coenzyme A synthetase. We have determined the 1.9-Å-resolution crystal structure of a DltA protein from Bacillus cereus in complex with ATP. This structure sheds light on the geometry of the bound ATP. The invariant catalytic residue Lys492 appears to be mobile, suggesting a molecular mechanism of catalysis for this superfamily of enzymes. Specific roles are also revealed for two other invariant residues: the divalent cation-stabilizing Glu298 and the β-phosphate-interacting Arg397. Mutant proteins with a glutamine substitution at position 298 or 397 are inactive.     

Binding Hot Spot in the Weak Protein Complex of Physiological Redox Partners Yeast Cytochrome c and Cytochrome c Peroxidase

Transient protein interactions mediate many vital cellular processes such as signal transduction or intermolecular electron transfer. However, due to difficulties associated with their structural characterization, little is known about the principles governing recognition and binding in weak transient protein complexes. In particular, it has not been well established whether binding hot spots, which are frequently found in strong static complexes, also govern transient protein interactions. To address this issue, we have investigated an electron transfer complex of physiological partners from yeast: yeast iso-1-cytochrome c (Cc) and yeast cytochrome c peroxidase (CcP). Using isothermal titration calorimetry and NMR spectroscopy, we show that Cc R13 is a hot-spot residue, as R13A mutation has a strong destabilizing effect on binding. Furthermore, we employ a double-mutant cycle to illustrate that Cc R13 interacts with CcP Y39. The present results, in combination with those of earlier mutational studies, have enabled us to outline the extent of the energetically important Cc-CcP binding region. Based on our analysis, we propose that binding energy hot spots, which are prevalent in static protein complexes, could also govern transient protein interactions.     

Crystal Structures of A. acidocaldarius Endoglucanase Cel9A in Complex with Cello-Oligosaccharides: Strong − 1 and − 2 Subsites Mimic Cellobiohydrolase Activity

Alicyclobacillus acidocaldarius endoglucanase Cel9A (AaCel9A) is an inverting glycoside hydrolase with β-1,4-glucanase activity on soluble polymeric substrates. Here, we report three X-ray structures of AaCel9A: a ligand-free structure at 1.8 Å resolution and two complexes at 2.66 and 2.1 Å resolution, respectively, with cellobiose obtained by co-crystallization and with cellotetraose obtained by the soaking method. AaCel9A forms an (α/α)₆-barrel like other glycoside hydrolase family 9 enzymes. When cellobiose is used as a ligand, three glucosyl binding subsites are occupied, including two on the glycone side, while with cellotetraose as a ligand, five subsites, including four on the glycone side, are occupied. A structural comparison showed no conformational rearrangement of AaCel9A upon ligand binding. The structural analysis demonstrates that of the four minus subsites identified, subsites − 1 and − 2 show the strongest interaction with bound glucose. In conjunction with the open active-site cleft of AaCel9A, this is able to reconcile the previously observed cleavage of short-chain oligosaccharides with cellobiose as main product with the endo mode of action on larger polysaccharides.     

A LIM-9 (FHL) / SCPL-1 (SCP) Complex Interacts with the C-terminal Protein Kinase Regions of UNC-89 (Obscurin) in Caenorhabditis elegans Muscle

The C. elegans gene unc-89 encodes a set of mostly giant polypeptides (up to 900 kDa) that contain multiple immunoglobulin (Ig) and fibronectin type 3 (Fn3), a triplet of SH3-DH-PH, and two protein kinase domains. The loss of function mutant phenotype and localization of antibodies to UNC-89 proteins indicate that the function of UNC-89 is to help organize sarcomeric A-bands, especially M-lines. Recently, we reported that each of the protein kinase domains interacts with SCPL-1, which contains a CTD-type protein phosphatase domain. Here, we report that SCPL-1 interacts with LIM-9 (FHL), a protein that we first discovered as an interactor of UNC-97 (PINCH) and UNC-96, components of an M-line costamere in nematode muscle. We show that LIM-9 can interact with UNC-89 through its first kinase domain and a portion of unique sequence lying between the two kinase domains. All the interactions were confirmed by biochemical methods. A yeast three-hybrid assay demonstrates a ternary complex between the two protein kinase regions and SCPL-1. Evidence that the UNC-89/SCPL-1 interaction occurs in vivo was provided by showing that over-expression of SCPL-1 results in disorganization of UNC-89 at M-lines. We suggest two structural models for the interactions of SCPL-1 and LIM-9 with UNC-89 at the M-line.