近藤 次郎

Parasitic infections recognized as neglected tropical diseases are a source of concern for several regions of the world. Aminoglycosides are potent antimicrobial agents that have been extensively studied by biochemical and structural studies in prokaryotes. However, the molecular mechanism of their potential antiprotozoal activity is less well understood. In the present study, we have examined the invitro inhibitory activities of some aminoglycosides with a 6-hydroxy group on ringI and highlight that one of them, 6-hydroxysisomicin, exhibits promising activity against a broad range of protozoan parasites. Furthermore, we have conducted X-ray analyses of 6-hydroxysisomicin bound to the target ribosomal RNA A-sites in order to understand the mechanisms of both its antibacterial and antiprotozoal activities at the molecular level. The unsaturated ringI of 6-hydroxysisomicin can directly stack on G1491, which is highly conserved in bacterial and protozoal species, through interaction and fits closer to the guanidine base than the typically saturated and hydroxylated ringI of other structurally related aminoglycosides. Consequently, the compound adopts a lower energy conformation within the bacterial and protozoal A-sites and makes pseudo pairs to either A or G at position1408. The A-site-selective binding mode strongly suggests that 6-hydroxysisomicin is a potential lead for the design of next-generation aminoglycosides targeting a wide variety of infectious diseases.

Parasitic infections recognized as neglected tropical diseases are a source of concern for several regions of the world. Aminoglycosides are potent antimicrobial agents that have been extensively studied by biochemical and structural studies in prokaryotes. However, the molecular mechanism of their potential antiprotozoal activity is less well understood. In the present study, we have examined the invitro inhibitory activities of some aminoglycosides with a 6′-hydroxy group on ringI and highlight that one of them, 6′-hydroxysisomicin, exhibits promising activity against a broad range of protozoan parasites. Furthermore, we have conducted X-ray analyses of 6′-hydroxysisomicin bound to the target ribosomal RNA A-sites in order to understand the mechanisms of both its antibacterial and antiprotozoal activities at the molecular level. The unsaturated ringI of 6′-hydroxysisomicin can directly stack on G1491, which is highly conserved in bacterial and protozoal species, through π-π interaction and fits closer to the guanidine base than the typically saturated and hydroxylated ringI of other structurally related aminoglycosides. Consequently, the compound adopts a lower energy conformation within the bacterial and protozoal A-sites and makes pseudo pairs to either A or G at position1408. The A-site-selective binding mode strongly suggests that 6′-hydroxysisomicin is a potential lead for the design of next-generation aminoglycosides targeting a wide variety of infectious diseases. © 2013 WILEY-VCH Verlag GmbH &amp Co. KGaA, Weinheim.

Sisomicin with an unsaturated sugar ring I displays better antibacterial activity than other structurally related aminoglycosides, such as gentamicin, tobramycin, and amikacin. In the present study, we have confirmed by X-ray analyses that the binding mode of sisomicin is basically similar but not identical to that of the related compounds having saturated ring I. A remarkable difference is found in the stacking interaction between ring I and G1491. While the typical saturated ring I with a chair conformation stacks on G1491 through CH/pi interactions, the unsaturated ring I of sisomicin with a partially planar conformation can share its pi-electron density with G1491 and fits well within the A-site helix.

Nucleotide bases are recognized by amino acid residues in a variety of DNA/RNA binding and nucleotide binding proteins. In this study, a total of 446 crystal structures of nucleotide-protein complexes are analyzed manually and pseudo pairs together with single and bifurcated hydrogen bonds observed between bases and amino acids are classified and annotated. Only 5 of the 20 usual amino acid residues, Asn, Gln, Asp, Glu and Arg, are able to orient in a coplanar fashion in order to form pseudo pairs with nucleotide bases through two hydrogen bonds. The peptide backbone can also form pseudo pairs with nucleotide bases and presents a strong bias for binding to the adenine base. The Watson-Crick side of the nucleotide bases is the major interaction edge participating in such pseudo pairs. Pseudo pairs between the Watson-Crick edge of guanine and Asp are frequently observed. The Hoogsteen edge of the purine bases is a good discriminatory element in recognition of nucleotide bases by protein side chains through the pseudo pairing: the Hoogsteen edge of adenine is recognized by various amino acids while the Hoogsteen edge of guanine is only recognized by Arg. The sugar edge is rarely recognized by either the side-chain or peptide backbone of amino acid residues.

Incorporation of an hydrophobic (phenethylamino) ethyl ether at C2 '' of N1-(HABA)-3',4'-dideoxyparomomycin led to a novel analog with an excellent antibacterial profile against a host of resistant bacteria. (C) 2010 Elsevier Ltd. All rights reserved.

Previously, a geometric nomenclature was proposed in which RNA base pairs were classified by their interaction edges (Watson-Crick, Hoogsteen or sugar-edge) and the glycosidic bond orientations relative to the hydrogen bonds formed (cis or trans). Here, base pairs and pseudo pairs observed in RNA-ligand complexes are classified in a similar manner. Twenty-one basic geometric families are geometrically possible (18 for base pairs formed between a nucleic acid base and a ligand containing heterocycle and 3 families for pseudo pairs). Of those, 16 of them have been observed in X-ray and/or NMR structures. Copyright (C) 2009 John Wiley & Sons, Ltd.

Patterson maps, which have a peak for intermolecular vectors between two molecules linked by a pseudo-translation, are widely used for structure solution. However, these maps may contain other peaks that indicate additional important information. In particular, if a molecule has internal symmetry, the Patterson maps may have a peak even when the relation between two molecules is other than a pure translation. A special and frequent case is a crystal that consists of molecules with pseudo-helical symmetry, like RNA or DNA, packed more or less parallel to each other. For such pairs of molecules, the Patterson peak does not simply link the molecular centres but is shifted along the helical axis. The shift is proportional to the rotation between the two molecules. The known step of the helix may be sufficient to obtain a system of linear equations whose solution gives an approximate position for the molecule. This technique may provide crucial information when molecular replacement fails to find a solution or suggests a number of candidates. Instead of repeating numerous molecular-replacement trials varying the model, a model may be positioned directly in place and be modified and refined directly. This Patterson-based search for molecular position has been tested with several solved crystals and has assisted in structure solution of RNA duplexes containing Homo sapiens cytoplasmic or mitochondrial ribosomal decoding sites (A sites).

The A site of the small ribosomal subunit participates in the fidelity of decoding by switching between two states, a resting off state and an active decoding on state. Eight crystal structures of RNA duplexes containing two minimal decoding A sites of the Homo sapiens mitochondrial wild-type, the A1555G mutant or bacteria have been solved. The resting off state of the mitochondrial wild-type A site is surprisingly different from that of the bacterial A site. The mitochondrial A1555G mutant has two types of the off states; one is similar to the mitochondrial wild-type off state and the other is similar to the bacterial off state. Our present results indicate that the dynamics of the A site in bacteria and mitochondria are different, a property probably related to the small number of tRNAs used for decoding in mitochondria. Based on these structures, we propose a hypothesis for the molecular mechanism of non-syndromic hearing loss due to the mitochondrial A1555G mutation.

The crystal structure of the complex between oligonucleotide containing the bacterial ribosomal decoding site (A site) and the synthetic poromomycin analogue 1, which contains the gamma-amino-alpha-hydroxybutyryl (L-haba) group at position N1 of ring II (2-DOS ring), and an ether chain with an O-phenethylaminoethyl group at position C2" of ring III, is reported. Interestingly, next to the paromomycin analogue I specifically bound to the A site, a second molecule of I with a different conformation is observed at the crystal packing interface which mimics the A-minor interaction between two bulged-out adenines from the A site and the codon-anticodon stem of the mRNA-tRNA complex. Improved antibacterial activity supports the conclusion that analogue 1 might affect protein synthesis on the ribosome in two different ways: 1) specific binding to the A site forces maintenance of the "on" state with two bulged out adenines, and 2) a new binding mode of I to an A-minor motif which stabilizes complex formation between the ribosome and the mRNA-tRNA complex regardless of whether the codon-anticodon stem is of the cognate or near-cognate type.

The lack of absolute prokaryotic selectivity of natural antibiotics is widespread and is a significant clinical problem. The use of this disadvantage of aminoglycoside antibiotics for the possible treatment of human genetic diseases is extremely challenging. Here, we hove used a combination of biochemical and structural analysis to compare and contrast the molecular mechanisms of action and the structure-activity relationships of a new synthetic aminoglycoside, NB33, and a structurally similar natural aminoglycoside apramycin. The data presented herein demonstrate the general molecular principles that determine the decreased selectivity of apramycin for the prokaryotic decoding site, and the increased selectivity of NB33 for the eukaryotic decoding site. These results are therefore extremely beneficial for further research on both the design of new aminoglycoside-based antibiotics with diminished deleterious effects on humans, as well as the design of new aminoglycoside-based structures that selectively target the eukaryotic ribosome.

In previous studies, it was reported that DNA fragments with the sequence d(gcGXYAgc) (where X = A or G and Y = A, T or G) form a stable base-intercalated duplex (Bi-duplex) in which the central X and Y residues are not involved in any base-pair interactions but are alternately stacked on each other between the two strands. To investigate the structural stability of the Bi-duplex, the crystal structure of d(gcGAACgc) with a point mutation at the sixth residue of the sequence, d(gcGAAAgc), has been determined. The two strands are associated in an antiparallel fashion to form two types of bulge-containing duplexes (Bc-duplexes), I and II, both of which are quite different from the Bi-duplex of the parent sequence. In both Bc-duplexes, three Watson-Crick G-C base pairs constitute the stem regions at the two ends. The A(4) residues are bulged in to form a pair with the corresponding A(4) residue of the opposite strand in either duplex. The A(4)center dot A(4)* pair formation is correlated to the orientations of the adjacent A(5) residues. A remarkable difference between the two Bc-duplexes is seen at the A(5) residue. In Bc-duplex I, it is flipped out and comes back to interact with the G(3) residue. In Bc-duplex II, the A(5) residue extends outwards to interact with the G(7) residue of the neighbouring Be-duplex I. These results indicate that trans sugar-edge/Hoogsteen (sheared-type) G(3)center dot A(6)* base pairs are essential in the formation of a Bi-duplex of d(gcGXYAgc). On the other hand, the alternative conformations of the internal loops containing two consecutive bulged A residues suggest molecular switching.

Oligonucleotides containing 5-(N-aminohexyl)carbamoyl-modified uracils have promising features for applications as antigene and antisense therapies. Relative to unmodified DNA, oligonucleotides containing 5-(N-aminohexyl)carbamoyl-2'-deoxyuridine (U-N) or 5-(N-aminohexyl)carbamoyl- 2'-O-methyluridine (U-N(m)), respectively exhibit increased binding affinity for DNA and RNA, and enhanced nuclease resistance. To understand the structural implications of U-N and U-N(m) substitutions, we have determined the X-ray crystal structures of DNA:DNA duplexes containing either U-N or U-N(m) and of DNA:RNA hybrid duplexes containing U-N(m). The aminohexyl chains are fixed in the major groove through hydrogen bonds between the carbamoyl amino groups and the uracil O4 atoms. The terminal ammonium cations on these chains could interact with the phosphate oxygen anions of the residues in the target strands. These interactions partly account for the increased target binding affinity and nuclease resistance. In contrast to U-N, U-N(m) decreases DNA binding affinity. This could be explained by the drastic changes in sugar puckering and in the minor groove widths and hydration structures seen in the U-N(m) containing DNA:DNA duplex structure. The conformation of U-N(m), however, is compatible with the preferred conformation in DNA:RNA hybrid duplexes. Furthermore, the ability of U-N(m) to render the duplexes with altered minor grooves may increase nuclease resistance and elicit RNase H activity.

The crystal structure of the tetragonal form of d(gcGAAAgc) has been revised and reasonably refined including the disordered residues. The two DNA strands form a base-intercalated duplex, and the four duplexes are assembled according to the crystallographic 222 symmetry to form an octaplex. In the central region, the eight strands are associated by I-motif of double A-quartets. Furthermore, eight hydrated-magnesium cations link the four duplexes to support the octaplex formation. Based on these structural features, a proposal that folding of d(GAAA)(n), found in the non-coding region of genomes, into an octaplex can induce slippage during replication to facilitate length polymorphism is presented.

The decoding A site of the small ribosomal subunit is an RNA molecular switch, which monitors codon-anticodon interactions to guarantee translation fidelity. We have solved the crystal structure of an RNA fragment containing two Homo sapiens cytoplasmic A sites. Each of the two A sites presents a different conformational state. In one state, adenines A1492 and A1493 are fully bulged-out with C1409 forming a wobble-like pair to A1491. In the second state, adenines A1492 and A1493 form non-Watson-Crick pairs with C1409 and G1408, respectively while A1491 bulges out. The first state of the eukaryotic A site is, thus, basically the same as in the bacterial A site with bulging A1492 and A1493. It is the state used for recognition of the codon/anticodon complex. On the contrary, the second state of the H.sapiens cytoplasmic A site is drastically different from any of those observed for the bacterial A site without bulging A1492 and A1493.

DNA fragments with the sequences d(gcGX[Y] n Agc) (n = 1, X = A, and Y = A, T, or G) form base-intercalated duplexes, which is a basic unit for formation of multiplexes such as octaplex and hexaplex. To examine the stability of multiplexes, a DNA with X = Y = G and n = 1 was crystallized under conditions different from those of the previously determined sequences, and its crystal structure has been determined. The two strands are coupled in an anti-parallel fashion to form a base-intercalated duplex, in which the first and second residues form Watson-Crick type G:C pairs and the third and sixth residues form a sheared G:A pairs at both ends of the duplex. The G(4) and G(5) bases are stacked alternatively on those of the counter strand to form a long G column of G(3)-G(4)-G(5)*-G(5)-G(4)*-G(3)*, the central four Gs being protruded. In addition, the three duplexes are associated to form a hexaplex around a mixture of calcium and sodium cations on the crystallographic threefold axis. These structural features are similar to those of the previous crystals, though slightly different in detail. The present study indicates that mutation at the 4th position is possible to occur in a base-intercalated duplex for multiplex formations, suggesting that DNA fragments with any sequence sandwiched between the two triplets gcG and Agc can form a multiplex.

DNA fragments containing the sequence d(GCGAAAGC) prefer to adopt a base-intercalated (zipper-like) duplex in crystalline state. To investigate effects of point mutation at the 5th residue on the structure, two crystal structures of d(GCGAGAGC) and d(GCGATAGC) have been determined by X-ray crystallography. In the respective crystals, the two octamers related by a crystallographic two-fold symmetry are aligned in an anti-parallel fashion and associated to each other to form a duplex, suggesting that the base-intercalated duplex is stable even when the 5th residue is mutated with other bases. The sheared G(3):A(6) pair formation makes the two phosphate backbones closer and facilitates formation of the A-X*-X-A* base-intercalated motif. The three duplexes are assembled around the three-fold axis, and their 3rd and 4th residues are bound to the hexamine cobalt chloride. The central 5th residues are bound to another cation.

A DNA fragment d( GCGAAAGC), postulated to adopt a stable mini- hairpin structure on the basis of its extraordinary properties, has been X- ray analyzed. Two octamers related by a crystallographic twofold symmetry are aligned in an antiparallel fashion and associate to form a duplex, which is maintained by two Watson - Crick G . C base pairs and a subsequent sheared G . A pair at both ends. The central two A residues are free from base- pair formation. The corresponding base moieties of the two strands are intercalated and stacked on each other, forming a long column of G(1)- C-2- G(3)- A(4)- A(5)*-A(5)-A*(4)- G*(3)-C*(2)-G*(1) ( asterisks indicate the counter- strand). The Watson - Crick and major- groove sites of the four stacked adenine bases are exposed to the solvent region, suggesting a functional role. Since this structural motif is similar to those found in the nonamers d(G(Br)CGAAAGCT) and d(G(I)CGAAAGCT), the base- intercalated duplex may be a stable form of the specific sequence. Electrophoresis results suggest that the octamer has two states, monomeric and dimeric, in solution depending on the Mg2+ concentration. The present duplex is preferred under the crystallization conditions, which correspond to physiologically allowed conditions.

A DNA fragment d(GCGAAAGCT), known to adopt a stable mini-hairpin structure in solution, has been crystallized in the space group I4(1)22 with the unit-cell dimensions a = b = 53.4 Angstrom and c = 54.0 Angstrom, and the crystal structure has been determined at 2.5 Angstrom resolution. The four nucleotide residues CGAA of the first half of the oligomer form a parallel duplex with another half through the homo base pairs, C-2:C-2(+) (singly-protonated between the Watson- Crick sites), G(3):G(3) (between the minor groove sites), A(4):A(4) (between the major groove sites) and A(5):A(5) (between the Watson-Crick sites). The two strands remaining in the half of the parallel duplex are split away in different directions, and they pair in an anti-parallel B-form duplex with the second half extending from a neighboring parallel duplex, so that an infinite column is formed in a head-to-tail fashion along the c-axis. It seems that a hexa-ammine cobalt cation supports such a branched and bent conformation of the oligomer. One end of the parallel duplex is stacked on the corresponding end of the adjacent parallel duplex; between them, the guanine base of the first residue is stacked on the fourth ribose of another duplex.