Sequence-specific DNA glycosylase found in a restriction-modification system

Restriction-modification systems consist of genes that encode a restriction enzyme and a cognate modification methyltransferase. It was believed that restriction enzymes are sequence-specific endonucleases that introduce double-strand breaks at specific sites by catalyzing the cleavages of phosphodiester bonds. R.PabI is a type II restriction enzyme from a hyperthermophilic archaea Pyrococcus abyssi that recognizes 5'-GTAC-3' sequence and cleaves DNA duplexes without the addition of a divalent cation. The structural and mutational analyses of R.PabI in our previous work showed that R.PabI forms a homodimer and has a novel DNA-binding fold called a "half-pipe," which consists of a highly curved anti-parallel β-sheet. Because the structure of R.PabI shares no structural similarity to any other protein, the structural basis of the sequence-recognition and DNA-cleavage mechanisms remained unclear. In this study, we report the crystal structure of R.PabI in complex with a DNA duplex containing the R.PabI recognition sequence. The structure of the R.PabI-DNA complex shows that R.PabI unwinds a DNA duplex at a 5'-GTAC-3' site and flips the guanine and adenine bases out of the DNA helix to recognize the sequence. The electron-density map of the R.PabI-DNA complex shows that R.PabI releases adenine bases from the R.PabI recognition sequence. Biochemical assays using HPLC and MALDI-TOF MS spectrometry also support the observation that R.PabI releases adenine bases by hydrolysis. These results show that R.PabI is not an endonuclease but a sequence-specific adenine DNA glycosylase. R.PabI is the first example of a restriction enzyme that shows DNA glycosylase activity. Mutational analysis reveals the active site of the adenine DNA glycosylase activity of R.PabI. The two opposing apurinic/apyrimidinic (AP) sites generated by R.PabI are cleaved by heat promoted β elimination and/or by endogenous AP endonucleases of host cells to introduce a double-strand break.

Degenerate Sequence Recognition by AdpA

AdpA is the central transcriptional factor in the A-factor regulatory cascade of Streptomyces griseus and activates hundreds of genes required for both secondary metabolism and morphological differentiation, leading to onset of streptomycin biosynthesis as well as aerial mycelium formation and sporulation. It has been shown that AdpA binds to over 500 operator regions with the loosely conserved consensus sequence, 5'-TGGCSNGWWY-3' (S: G or C; W: A or T; Y: T or C; and N: any nucleotide). However, it is still obscure how AdpA can control hundreds of genes. To reveal the molecular basis of the low nucleotide sequence specificity, we have determined the crystal structure of the complex of DNA-binding domain of AdpA and a 14-mer duplex DNA with two-nucleotide overhangs at 5'-ends at 2.95-Å resolution. The crystal structure and electrophoretic mobility-shift assays showed that only two arginine residues, Arg262 and Arg266, are involved in the sequence recognition and determine the nucleotide specificity/preference of continuous five base-pairs of positions 1-5 in the consensus sequence. These results partially explain how AdpA directly controls hundreds of genes.

A special halide-stabilizing pair and enantioselectivity mechanism of DatA

A novel haloalkane dehalogenase DatA from Agrobacterium tumefaciens C58 belongs to the HLD-II subfamily and hydrolyzes brominated and iodinated compounds, leading to the generation of the corresponding alcohol, a halide ion, and a proton. DatA possesses a unique Asn-Tyr residue pair instead of the Asn-Trp residue pair conserved among the subfamily members, thus the structural basis for its reaction mechanism merits elucidation. In addition, DatA is potentially useful for pharmaceutical and environmental applications, though several crystal structures of HLD-II dehalogenases have been reported so far, the determination of the DatA structure will provide an important contribution to those fields. This work provided insight into the reaction mechanism of DatA via a combination of X-ray crystallographic and computational analysis. The crystal structures of DatA and the Y109W mutant were determined at 1.70 Å [1] and 1.95 Å, respectively. The location of the active site was confirmed by using its novel competitive inhibitor, CHES. The structural information from these two crystal structures and the docking simulation with 1,3-dibromopropane revealed that the replacement of the Asn-Tyr pair with the Asn-Trp pair increases the binding affinity for 1,3-dibromopropane, due to the extra hydrogen bond between Trp109 and halogenated compounds; and that the key residue to bind halogenated substrate is only Asn43 in the wild-type DatA, while those in the Y109W mutant are the Asn-Trp pair. Furthermore, docking simulation using the crystal structures of DatA and some chiral compounds indicated that enantioselectivity of DatA toward brominated alkanes is determined by the large and small spaces around the halogen binding site.

Crystal structure of the M14R mutant of oryctin in complex with trypsin

Oryctin is a 66-amino-acid protein purified from the larval haemolymph of the coconut rhinoceros beetle Oryctes rhinoceros, which shows no sequence similarity to any other protein known. We determined the solution NMR structure of oryctin, and found that oryctin had a similar backbone fold to the turkey ovomucoid domain 3, OMTKY3, a Kazal-type serine protease inhibitor [1]. Based on the structural similarity, we tested the serine protease inhibitory activity of oryctin, and found that oryctin does inhibit some serine proteases, such as α-chymotrypsin, endopeptidase K, subtilisin Carlsberg, and leukocyte elastase [1]. However, oryctin cannot inhibit trypsin at all. In this study, we have introduced point mutations to the putative inhibition loop of oryctin to obtain oryctin mutants that can inhibit trypsin. Then, we have solved the crystal structure of such an oryctin mutant, M14R-oryctin with a Ki value of 3.4 ¹ 0.8 nM, in complex with trypsin to reveal how it binds to and inhibits trypsin. As predicted, the putative inhibition loop lay on the substrate binding cleft of trypsin. Particularly, the side chain of R14 fit into the S1 pocket of trypsin by forming hydrogen/ionic bonds with D191, S192 and G216 at the bottom of the S1 pocket and G195, D196, S197 and S212 at its entrance. In addition, R65 located in the C-terminal α-helix of M14R-oryctin formed hydrogen bonds with S40 and F44 of trypsin. The latter interaction, which is unique to oryctin, enhances its binding affinity to trypsin.

A novel six-coordinated ferric ion binding mode of TtFbpA

Iron is an essential element for the growth and survival of nearly all living organisms. However, it is difficult for most organisms to get enough iron from the environment, because of the extremely low solubility of ferric ion. One of the strategies for iron acquisition is to use the ATP-binding cassette (ABC) transport system. In Gram-negative bacteria, a typical iron uptake ABC transporter consists of a ferric ion-binding protein (Fbp) located in periplasm (FbpA), two transmembrane proteins that form a pathway for ferric ions (FbpB), and two peripheral ATP-binding proteins located at the cytoplasm side of the inner membrane (FbpC). TtFbpA is a ferric ion-binding protein of a putative iron uptake ABC transporter from Thermus thermophilus HB8. Here we report the crystal structures of the apo-form and ferric ion-bound form of TtFbpA at 1.8-Å and 1.7-Å resolutions, respectively [1]. The crystal structure of the ferric ion-bound TtFbpA shows that a ferric ion binds to a specific site of TtFbpA to form a six-coordinated complex by three tyrosine residues, two bicarbonates and a water molecule, revealing a novel mode of coordination to a ferric ion. Another crystal structure of ferric ion-bound TtFbpA reported earlier showed the bound ferric ion is five-coordinated by three tyrosine residues and a carbonate bound in the bidentate manner [2]. The different modes of the coordination would probably result from the different pHs used for crystallization: pH 5.5 (six-coordinated) vs. pH 7.5 (five-coordinated). The Gram negative bacterium T. thermophilus HB8 can live in a wide pH range of 3.4-9.6. We propose that TtFbpA, a periplasmic protein of T. thermophilus HB8, can act as a ferric ion-binding protein over the wide pH range by taking at least two different coordination manners to a ferric ion depending on pH. This is the first example of a periplasmic ferric iron-binding protein that can coordinate a ferric ion via multiple types of coordination complex formation.

Molecular mechanism of strigolactone perception by DWARF14

The terpenoid, small-compound strigolactones (SLs) are plant hormones that regulate plant shoot branching, which is an important agronomic trait that determines crop yields. An α/β-hydrolase protein, DWARF14 (D14), has been recognized to be an essential component of plant SL signaling. Recently, it has been demonstrated that D14 interacts with a gibberellin (GA)-signaling repressor SLR1 in an SL-dependent manner [1], which suggests that SLR1 mediates crosstalk between the SL and GA signalings in the regulation of plant shoot branching. Although D14 functions in SL perception to promote the interaction with SLR1, its molecular mechanism remains unclear. Here, we report the crystal structure of D14 in the complex with 5-hydroxy-3-methylbutenolide (D-OH), which is a reaction product of SLs. The structure was solved at a 2.10-Å resolution when an SL synthetic analogue, (–)-ent-2'-epi-GR7, was soaked into D14 crystals [1]. In the complex structure, D-OH was located at a site far from the catalytic residues including H297 and appeared to function as a plug for the catalytic cavity to induce an overall hydrophobic surface with a hydrophilic patch between the two α-helices in the cap structure of D14. In the binding site, the indole amine of Trp205 formed a hydrogen bond with the oxygen atom of the C2' hydroxy group, which arose from the catalytic reaction of D14, instead of a water molecule in the structure of apo D14. In addition, the side chain of Phe245 moved 1.3 Å toward D-OH. Mutational analyses of D14 showed that the interaction between D14 and SLR1 required an enzymatic activity of D14 and the residues Trp205 and Phe245 were essential for the SL-dependent SLR1-binding of D14. These results suggest that the D14–D-OH complex mediates the interaction with SLR1 in which the D-OH-induced surface and/or structural change is crucial.

Brazzein and structurally similar proteins: structural/functional comparisons

Brazzein, a 6.5-kDa protein consisting of 54 amino acids and four disulfide bonds, is the smallest sweet-tasting protein yet isolated from the wild African plant Pentadiplandra brazzeana. Brazzein has various desirable properties for use as a low-calorie sweetener in the diets of individuals suffering from diabetes, obesity, and metabolic syndrome. For example, brazzein has a high water solubility and a high thermostability. In addition, brazzein is 2000-times sweeter than sucrose on a weight basis. Both the solution and crystal structures of brazzein have been reported. In the crystal structure [1], brazzein has a defensin-like fold containing two α-helices and a three-stranded antiparallel β-sheet. Defensins are small cysteine-rich cationic proteins found in both animals and plants, which function by binding to the microbial cell membrane, and, once embedded, forming pore-like membrane defects that allow efflux of essential ions and nutrients. In fact, Yount and Yeaman reported that brazzein has antimicrobial activity against Gram positive (Bacillus subtilis and Staphylococcus aureus) and negative (Escherichia coli) bacteria and a fungus (Candida albicans) at pH 7.5 rather than pH 5.5 [2]. A search for proteins with a similar backbone fold to brazzein using the DALI server shows that structurally similar proteins to brazzein include plant defensins, scorpion neurotoxins (K+ channel blockers), arthropod defensins, mollusc defensins, mold defensins, and a plant trypsin inhibitor. These proteins commonly have a γ-core sequence. Here we compare their sequences, structures and functions, which has led to a conclusion that the C-terminal half of brazzein is important for its antimicrobial activity, brazzein will not have a neurotoxin activity, and it will not act as a trypsin inhibitor.