武藤 裕

Background: The AU binding homolog of enoyl-CoA hydratase (AUH) is a bifunctional protein that has two distinct activities: AUH binds to RNA and weakly catalyzes the hydration of 2-trans-enoyl-coenzyme A (enoyl-CoA). AUH has no sequence similarity with other known RNA binding proteins, but it has considerable sequence similarity with enoyl-CoA hydratase. A segment of AUH, named the R peptide, binds to RNA. However, the mechanism of the RNA binding activity of AUH remains to be elucidated. Results: We determined the crystal structure of human AUH at 2.2 Angstrom resolution. AUH adopts the typical fold of the enoyl-CoA hydratase/isomerase superfamily and forms a hexamer as a dimer of trimers. Interestingly, the surface of the AUH hexamer is positively charged, in striking contrast to the negatively charged surfaces of the other members of the superfamily. Furthermore, wide clefts are uniquely formed between the two trimers of AUH and are highly positively charged with the Lys residues in alpha helix H1, which is located on the edge of the cleft and contains the majority of the R peptide. A mutational analysis showed that the lysine residues in alpha helix H1 are essential to the RNA binding activity of AUH. Conclusions: a helix H1 exposes a row of Lys residues on the solvent-accessible surface. These characteristic Lys residues are named the "lysine comb." The distances between these Lys residues are similar to those between the RNA phosphate groups, suggesting that the lysine comb may continuously bind to a single-stranded RNA. The clefts between the trimers may provide spaces sufficient to accommodate the RNA bases.

Proteins that contain two or more copies of the RNA-binding domain [ribonucleoprotein (RNP) domain or RNA recognition motif (RRM)] are considered to be involved in the recognition of single-stranded RNA, but the mechanisms of this recognition are poorly understood at the molecular level. For an NMR analysis of a single-stranded RNA complexed with a multi-RBD protein, residue-selective stable-isotope labeling techniques are necessary, rather than common assignment methods based on the secondary structure of RNA. In the present study, we analyzed the interaction of a Drosophila Sex-lethal (Sxl) protein fragment, consisting of two RBDs (RBD1-RBD2), with two distinct target RNAs derived from the tra and Sxl mRNA precursors with guanosine and adenosine, respectively, in a position near the 5'-terminus of a uridine stretch. First, we prepared a [5-H-2]uridine phosphoramidite, and synthesized a series of H-2-labeled RNAs, in which all of the uridine residues except one were replaced by [5-H-2]uridine in the target sequence, GU(8)C. By observing the H5-H6 TOCSY cross peaks of the series of H-2-labeled RNAs complexed with the Sxl RBD1-RBD2, all of the base H5-H6 proton resonances of the target RNA were unambiguously assigned. Then, the H5-H6 cross peaks of other target RNAs, GU(2)GU(8), AU(8), and UAU(8), were assigned by comparison with those of GU(8)C. We found that the uridine residue prior to the G or A residue is essential for proper interaction with the protein, and that the interaction is tighter for A than for G. Moreover, the H1' resonance assignments were achieved from the H5-H6 assignments. The results revealed that all of the protein-bound nucleotide residues, except for only two, are in the unusual C2'-endo ribose conformation in the complex.

Hu antigen C (HuC) has three RNA-binding domains (RBDs). The N-terminal two, RBD1 and RBD2, are linked in tandem and bind to the AU-rich elements (AREs) in the 3'-untranslated region of particular mRNAs, The solution structures of HuC RBD1 and RBD2 were determined by NMR methods. The HuC RBD1 and REDS structures are quite similar to those of SxI RBD1 and RBD2, respectively. The individual RBDs of HuC, RBD1 and RBD2 in isolation can interact rather weakly with the minimal ARE motif, AUUUA, while the didomain fragment, RBD1-RBD2, of HuC binds more tightly to a longer ARE RNA, UAUUUAUUUU. Chemical shift perturbations by the longer RNA on HuC RBD1-RBD2 were mapped on and around the two beta-sheets and on the C-terminal region of RBD1. The HuC RBD1-RBD2 residues that exhibited significant chemical shift perturbations coincide with those conserved in Sxl RBD1-RBD2, These data indicate that the RNA-binding characteristics of the HuC and Sxl didomain fragments are similar, even though the target RNAs and the biological functions of the proteins are different.

Hu antigen C (HuC) has three RNA-binding domains (RBDs). The N-terminal two, RBD1 and RBD2, are linked in tandem and bind to the AU-rich elements (AREs) in the 3'-untranslated region of particular mRNAs. The solution structures of HuC RBD1 and RBD2 were determined by NMR methods. The HuC RBD1 and RBD2 structures are quite similar to those of Sxl RBD1 and RBD2, respectively. The individual RBDs of HuC, RBD1 and RBD2 in isolation can interact rather weakly with the minimal ARE motif, AUUUA, while the didomain fragment, RBD1-RBD2, of HuC binds more tightly to a longer ARE RNA, UAUUUAUUUU. Chemical shift perturbations by the longer RNA on HuC RBD1-RBD2 were mapped on and around the two beta-sheets and on the C-terminal region of RBD1. The HuC RBD1-RBD2 residues that exhibited significant chemical shift perturbations coincide with those conserved in Sxl RBD1-RBD2. These data indicate that the RNA-binding characteristics of the HuC and Sxl didomain fragments are similar, even though the target RNAs and the biological functions of the proteins are different.

In Bacillus subtilis, tryptophan biosynthesis is regulated by a mechanism called attenuation. The new crystal structure of the 'trp RNA binding attenuation protein', TRAP, in complex with RNA has provided new structural insights into how proteins can bind RNA to regulate transcription and translation.

The large subunit of the human U2 small nuclear ribonucleoprotein particle auxiliary factor (hU2AF(65)) is an essential RNA-splicing factor required for the recognition of the polypyrimidine tract immediately upstream of the 3' splice site. In the present study, we determined the solution structures of two hU2AF65 fragments, corresponding to the first and second RNA-binding domains (RBD1 and RBD2, respectively), by nuclear magnetic resonance spectroscopy. The tertiary structure of RBD2 is similar to that of typical RNA-binding domains with the beta 1-alpha 1-beta 2-beta 3-alpha 2-beta 4 topology. In contrast, the hU2AF65 RBD1 structure has unique features: (i) the al helix is elongated by one turn toward the C-terminus; (ii) the loop between al and beta 2 (the alpha 1/beta 2 loop) is much longer and has a defined conformation; (iii) the beta 2 strand is (188)AVQIN(192), which was not predicted by sequence alignments; and (iv) the beta 2/beta 3 loop is much shorter. Chemical shift perturbation experiments showed that the U2AF-binding RNA fragments interact with the four beta-strands of RBD2 whereas, in contrast, they interact with pr, beta 3 and beta 4, but not with beta 2 or the alpha 1/beta 2 loop, of RBD1, The characteristic alpha 1-beta 2 structure of the hU2AF65 RBD1 may interact with other proteins, such as UAP56.

The Sex-lethal (Sxl) protein of Drosophila melanogaster regulates alternative splicing of the transformer (tra) messenger RNA precursor by binding to the tra polypyrimidine tract during the sex-determination process. The crystal structure has now been determined at 2.6 Angstrom resolution of the complex formed between two tandemly arranged RNA-binding domains of the Sxl protein and a 12-nucleotide, single-stranded RNA derived from the tra polypyrimidine tract. The two RNA-binding domains have their beta-sheet platforms facing each other to form a V-shaped cleft. The RNA is characteristically extended and bound in this cleft, where the UGUUUUUUU sequence is specifically recognized by the protein. This structure offers the first insight, to our knowledge, into how a protein binds specifically to a cognate RNA without any intramolecular base-pairing.

The interactions of the second RNA-binding domain of the Drosophila melanogaster Sex-lethal protein (Sxl RBD2) with the oligoribonucleotides, GUUUUUUUU (GU(8)) and CUAGUG, representing the sequences surrounding an alternative 3'-splicing site of the transformer pre-mRNA (GU(8)CUAGUG), were studied using heteronuclear two-dimensional NMR techniques. The H-1 and N-15 chemical shifts of the backbone amide resonances upon titration of Sri RBD2 with each of these RNAs were recorded. It was found that Sxl RBD2 can bind not only to the polyuridine tract, GU(8), but also to the downstream 3' splice-site sequence, CUAGUG, with similar affinities. In contrast, a nonspecific sequence, Cg, did not bind to Sxl RBD2. This result is consistent with previous in vitro RNA-selection and UV-cross-linking results which indicated that the Sex-lethal protein binds to the uridine stretch and the AG dinucleotide in the consensus sequence, AU(n)N(n)AGU. In both cases, the chemical-shift perturbations were significant for almost the same amino acid residues, including the two central beta-strands formed by the RNP2-motif and RNP1-motif with the two highly conserved aromatic residues (Y214 and F256) in the middle. As the first RNA-binding domain of Sex-lethal (Sxl RBD1) has a characteristic aliphatic residue at one of the two corresponding positions (I128 and F170), Y214 of Sri RBD2 was replaced by ne using site-directed mutagenesis. On the one hand, the H-1 and N-15 chemical-shift perturbations indicated that GU(8) binds to the same interface of mutant Sxl RBD2 as of wildtype Sxl RBD2, although its binding affinity was decreased significantly. On the other hand, the specific binding of Sri RBD2 to CUAGUG was abolished almost completely by the Y-->I mutation. Taken together, the present results indicate that the interface residues that bind with GU(8) and CUAGUG are much the same, but the role of the Y214 residue is clearly different between these two target sequences.

The Sex-lethal (Sxl) protein from Drosophila melanogaster has two RNA-binding domains (RBDs). As the amino-terminal RBD (RBD1) of the Sxl protein exhibits low sequence homology to the typical RBDs, particularly at the putative functional residues, it was difficult to unambiguously locate the RNP1 and RNP2 motifs. Therefore, in the present study, we defined the amino and carboxy-terminal borders of the first RNA-binding domain (RBD1) of the Sxl protein by limited tryptic digestion. By replacement of Phe166 by Tyr, we constructed a highly soluble mutant, which exhibits the same RNA-binding properties as those of the wild-type. Using this mutant protein, we performed NMR measurements, and elucidated the secondary and tertiary structures of the Sxl RBD1 in solution. The beta alpha beta beta alpha beta folding pattern is conserved in the solution structure of the Sxl RBD1, as in other reported RBD structures. This allowed us to identify both the RNP1 and RNP2 motifs of the Sxl RBD1 unambiguously. Intriguingly, the RNP2 motif of the Sxl RBD1 has an he residue at the second position, which is generally occupied by an aromatic amino acid residue in RBDs and has been suggested to be involved in their RNA binding. Furthermore, the loop region between beta 2 and beta 3 of the Sxl RBD1 has an exceptional cluster of aromatic amino acid residues, in place of the normal basic amino acid cluster. In contrast, the second RBD of Sxl does not exhibit these characteristic features. (C) 1997 Academic Press Limited.

The backbone H-1, C-13, and N-15 resonances of the c-Ha-Ras protein [a truncated version consisting of residues 1-171, Ras(1-171)] bound with GMPPNP (a slowly hydrolyzable analogue of GTP) were assigned and compared with those of the GDP-bound Ras(1-171). The backbone amide resonances of amino acid residues 10-13, 21, 31-39, 57-64, and 71 of Ras(1-171).GMPPNP, but not those of Ras(1-171).GDP, were extremely broadened, whereas other residues of Ras(1-171).GMPPNP exhibited amide resonances nearly as sharp as those of Ras(1-171).GDP. The residues exhibiting the extreme broadening, except for residues 21 and 71, are localized in three functional loop regions [loops L1, L2 (switch I), and L4 (switch II)], which are involved in hydrolysis of GTP and interactions with other proteins. From the temperature and magnetic field strength dependencies of the backbone amide resonance intensities, the extreme broadening was ascribed to the exchange at an intermediate rate on the NMR time scale. It was shown that the Ras(1-171) protein bound with GTP or GTP gamma S (another slowly hydrolyzable analogue of GTP) exhibits the same type of broadening. Therefore, it is a characteristic feature of the GTP-bound form of Ras that the LI, L2, and L4 loop regions, but not other regions, are in a rather slow interconversion between two or more stable conformers. This phenomenon, termed a ''regional polysterism'', of these loop regions may be related with their multifunctionality: the GTP-dependent interactions with several downstream target groups such as the Raf and RalGDS families and also with the GTPase activating protein (GAP) family. In fact, the binding of Ras(1-171) GMPPNP with the Ras-binding domain (residues 51-131) of c-Raf-l was shown to eliminate the regional polysterism nearly completely. It was indicated, therefore, that each target/regulator selects its appropriate conformer among those presented by the ''polysteric'' binding interface of Ras. As the downstream target groups exhibit no apparent sequence homology to each other, it is possible that one target group prefers a conformer different from that preferred by another group. The involvement of loop L1 in the regional polysterism might suggest that the negative regulators, GAPs, bind to the polysteric binding interface (loops L2 and L4) of Ras and cooperatively select a conformer suitable for transition of the GTPase catalytic center, involving loops L1 and L4, into the highly active state.

It has been reported that a 183 residue fragment, consisting of the two RNA-binding domains (RBD1-RBD2) of the Drosophila melanogster Sex-lethal (Sri) protein, strongly binds an oligonucleotide of the target RNA sequence (5'-GUUUUUUUUC-3') that regulates alternative splicing, and forms four or five hydrogen bonds with the imino groups of the RNA, In the present study, we used site-directed mutagenesis to improve the solubility of the didomain fragment of Sri, and confirmed that this mutant fragment forms hydrogen bonds with the target RNA in the same manner as that of the wild-type fragment, The mutant fragment was shown to bind the cognate RNA sequences GUUUUUUUUC and AUUUUUUUUC more tightly than UUUUUUUUC, By using a [3-N-15]uridine phosphoramidite, we synthesized a series of N-15-labeled target RNAs, in which one of the uridine residues was specifically replaced by [3-N-15]uridine, By observing the imino H-1-N-15 coupling of the labeled uridine residue, we assigned all four of the hydrogen-bonded imino protons to U1, U2, U5 and U6, respectively, of the target RNA, The imino protons of U2 and U6 exhibited nuclear Overhauser effects with aliphatic protons of the protein, All these results indicate that the A/G, U1, U2, U5 and U6 residues in the target sequence of (G/A)UUUUUUUU are specifically recognized by the two RNA-binding domains of the Sri protein.

The chemical shifts of ring protons of the beta-Man residue in a triantennary complex type N-linked glycan chain having a GlcNAc beta 1-->6(GlcNAc beta 1-->2)Man alpha 1-->6Man beta sequence were unambiguously determined by two-dimensional proton nuclear magnetic resonance (H-1-NMR) spectroscopic methods, The chemical shift of H4 (3.84 ppm) of the beta-Man residue was for the first time revealed to be different from those (similar to 3.77 ppm) of biantennary and alternative type of triantennary glycans having a GlcNAc beta 1-->2Man alpha 1-->6Man beta sequence, but quite close to that (3.86 ppm) of a pentaantennary glycan containing a GlcNAc beta 1-->6 residue on the Man alpha 1-->6Man beta sequence, Thus, the addition of GlcNAc beta 1-->6 residue on the Man-4' residue, whose formation is catalyzed by GlcNAc transferase V, is considered to cause a down-field shift of beta-Man H4 in the complex-type N-glycan chains, One possible explanation of this phenomenon is that the conformation of Man alpha 1-->6 arm is folded back toward the proximal core region, as is the case with the complex-type N-glycan chains with the bisecting GlcNAc residue.

This study represents the first detailed investigation of the nature of highly sulfated (keratan-sulfate-like) complex-type asparagine-linked glycans having a tetraantennary core structure and shows the effectiveness of fast-atom-bombardment mass spectrometric (FAB-MS) methods incorporating derivatization and mild methanolysis for analyzing such complex types of sulfated glycans. The structure of the N-glycan chains was unambiguously established by a combination of compositional analysis, methylation analysis, mild methanolysis for desulfation, hydrazinolysis/nitrous acid deamination, enzymatic (endo-beta-galactosidase and peptide:N-glycosidase F) digestions, and instrumental analyses (H-1-NMR spectroscopy and FAB-MS) which revealed the novel repeating sulfated carbohydrate sequences, +/-Gal beta 1-->4Gal beta 1[-->4(HSO3-->6)GlcNAc beta 1-->3(+/-Gal beta 1-->4)Gal beta 1]-->(see Structure I; p+q+r+s approximate to 14). This sequence is unique in: (a) the skeletal structure is similar to that of keratan sulfate but is completely devoid of 6-O-sulfated Gal residues and (b) the presence of branched Gal residues in the sequence -->4GlcNAc beta 1-->3(Gal beta 1-->4)Gal beta 1-->.

5-Methylaminomethyluridine (mnm(5)U) exists in the first position of the anticodon (position 34) of Escherichia coli tRNA(4)(Arg) for codons AGA/AGG. In the present study, the temperature dependence of the ribose-puckering equilibrium of pmnm(5)U was analyzed by proton NMR spectroscopy. Thus, the enthalpy difference (Delta H) between the C2'-endo and C3'-endo forms was obtained as 0.65 kcal . mol(-1). By comparison of the Delta H values of pU and pmnm(5)U, the 5-substitution was found to increase the relative stability of the C3'-endo form over the C2'-endo form significantly (by 0.56 kcal mol-l). Furthermore, this conformational ''rigidity'' was concluded to depend on the 5'-phosphate group, because nucleoside U exhibits only a negligible change in the ribose-puckering equilibrium upon the 5-methylaminomethyl substitution. Further NMR analyses and molecular dynamics calculations revealed that interactions between the 5-methylaminomethyl and 5'-phosphate groups of pmnm(5)U restrict the conformation about the glycosidic bond to a low anti form, enhancing steric repulsion between the 2-carbonyl and 2'-hydroxyl groups in the C2'-endo form. This intrinsic conformational rigidity of the mnm(5)U residue in position 34 may contribute to the correct codon recognition.

On the basis of the nucleotide sequence of Tetrahymena group I intron, we constructed a 31 residue RNA that has the P7 stem and the 3'-terminal guanosine residue (3'-G) with a putative stem-loop structure (P9.0) intervening between them, For this model RNA (P7/P9.0/G), four residues around the guanosine binding site (GBS) in the P7 stem were found to exhibit much lower sensitivities to ribonuclease V-1 than those of a variant RNA having adenosine in place of the 3'-G, suggesting that the 3'-G contacts around the GBS, NMR analyses of the imino proton resonances of the P7/P9.0/G RNA indicated that the base pairing in the GBS is retained on the interaction with the 3'-G, and that the two base pairs of the putative P9.0 stem-loop are definitely formed. Comparison of the RNA with its variants with either A (3'-A) or a deletion in place of the 3'-G suggested that the stability of the P9.0 stem-loop is affected by the GBS-3'-G interaction. The melting temperatures of the P9.0 stem-loop were determined from the UV absorbances of these RNAs, which quantitatively indicated that the P9.0 stem-loop is significantly stabilized by the interaction of the GBS with the 3'-G, rather than the 3'-A, and also by direct interaction with divalent cations (Mg2+, Ca2+ or Mn2+), Upon replacement of the G . C base pair by C . G in the GBS of the P7/P9.0/G RNA, the specificity was switched from 3'-G to 3'-A, as in the case of the intact intron.

For the application of multidimensional NMR spectroscopy to larger proteins, it would be useful to perform selective labeling of one of the 20 amino acids. For some amino acids, however, amino acid metabolism drastically reduces the efficiency and selectivity of labeling in in vivo expression systems. In the present study, a cell-free protein synthesis system was optimized, so that highly efficient and selective stable isotope labeling of proteins can be achieved in the absence of amino acid metabolism. The productivity of the E. coli cell-free coupled transcription-translation system was first improved, by about fivefold, by using the T7 RNA polymerase for transcription and also by improving the translation conditions. Thus, about 0.1 mg protein per 1 ml reaction mixture was synthesized. Then, this improved cell-free system was used for Asp- or Ser-selective N-15-labeling of the human c-Ha-Ras protein. With a 15 ml cell-free reaction, using less then 1 mg of N-15-labeled amino acid, 1 mg of the Ras protein was obtained. H-1-N-15 HSQC experiments confirmed that the Ras protein was efficiently labeled with high selectivity. These results indicate that this cell-free protein synthesis system is useful for NMR studies.

A novel carbohydrate-rich sialoglycopeptide of apparent molecular mass similar to 6 kDa was isolated from the fertilized eggs of Fundulus heteroclitus (euryhaline killi fish). This glycopeptide is a member of the L-hyosophorin family, characterized by its high content of carbohydrate (80-90% by weight) and formed by depolymerization of the precursor glycopoly-protein (H-hyosophorin) upon fertilization, The structures of the N-glycan chains were unambiguously established by a combination of compositional analysis, methylation analysis, selective chemical degradation (periodate oxidation-Smith degradation and hydrazinolysis-nitrous acid deamination), enzymatic (peptide:N-glycosidase F, several beta-galactosidases, beta-hexosaminidase and alpha-galactosidase) digestions and instrumental analyses (H-1-NMR and fast atom bombardment mass spectrometry) to have the novel and unique carbohydrate sequences, Gal alpha 1-->3(Gal beta 1-->4) Gal beta 1-->4GlcNAc beta 1--> and Gal alpha 1-->3(+/-GalNAc beta 1-->4GlcNAc beta 1-->3Gal beta 1-->4)Gal beta 1-->4GlcNAc beta 1-->. This study represents the first detailed investigation of the nature of bulky complex asparagine-linked penta-antennary glycans with a bisecting GlcNAc residue in glycoproteins. Expression of such bulky multiantennary glycan units on proteins may be essential during early embryogenesis.

The chemical shifts of all the ring protons of the three Man residues in a pentaantennary, glycan chain have been unambiguously assigned by two-dimensional proton nuclear magnetic resonance (H-1-NMR) spectroscopic methods. The study, using chemical shift and J values on the conformation of the trimannosyl unit, revealed that the rotamer about the C5-C6 bond of the alpha 1-->6 linkage in the sequence of Man alpha 1-->6Man beta 1--> is predominantly confined to a gauche-gauche rotamer (omega = 180 degrees, omega = O6-C6-C5-H5) and not to a gauche-trans rotamer (omega = -60 degrees). We do not know of any previous demonstration that the dihedral angle omega (O6-C6-C5-H5) in Man alpha 1-->6Man beta 1--> is preferentially 180 degrees in complex-type N-linked glycans having no bisecting GlcNAc residue.

Ile-23 of human epidermal growth factor (hEGF) has been indicated, by mutagenesis and NMR studies, to be directly recognized by the receptor. In the present study, an unnatural phenylalanine analog, either 2-azaphenylalanine (2aF), 3-azaphenylalanine (3aF), 4-azaphenylalanine (4aF), or 4-fluorophenylalanine (4fF), was incorporated by an in vivo protein synthesis system into position 23 of [Phe(23)] hEGF, which retains appreciable receptor-binding affinity (about 20% of the wild type). We compared the receptor-binding affinities of the variants with that of [Phe(23)]hEGF and found that substitution of Phe-23 with 2aF or 3aF raised the affinity, while substitution with 4aF or 4fF remarkably reduced the affinity. The tertiary structure of [Phe(23)] hEGF was not significantly affected by the substitution of Phe-23 with 2aF, as shown by the two-dimensional nuclear magnetic resonance analysis. In addition, the substitution of residue 23 with His or Tyr produced an hEGF variant with a slightly higher receptor-binding affinity than that of [Phe(23)]hEGF. Our results suggest that the receptor has an asymmetric hydrophobic pocket for recognition of the side chain in position 23 of hEGF. Furthermore, on the receptor surface, this pocket seems to be adjacent to a less hydrophobic region with a hydrogen-bond acceptor and donor. Thus, the use of unnatural amino acids in addition to the 20 natural ones allows analyses of the structure-function relationship of a protein at a higher resolution than conventional site-directed substitution by only natural amino acid residues.

KDN-gp, which is the unique glycoprotein of the rainbow trout egg envelope, was shown to have a small amount of N-linked oligosaccharide units in addition to a large number of O-linked glycan units. Structural analysis based on chemical analysis in combination with 400 MHz H-1 NMR spectroscopy revealed the presence of fully KDNosylated bi- and triantennary complex-type oligosaccharide chains, mostly fucosylated at the innermost GlcNAc residue and bisected by the GlcNAc residue linked beta 1-->4 to the beta-Man residue. The structures thus determined represent the first demonstration of N-linked glycan unit containing the KDN residues in the KDN-containing glycoproteins (see Chart 1). The KDN-gp of the rainbow trout egg envelope is a molecule that is present in the second layer of the vitelline envelope but is exposed to the outer surface around the micropyle through which sperm can get in at fertilization. Like human hematopoietic cell surface glycoproteins such as glycophorin A and leukosialin, KDN-gp, which is now characterized to contain N-linked complex-type glycan chains as minor components, is heavily O-glycosylated with alpha 2-->8-linked oligo/polyKDN-containing glycan units attached O-glycosidically to Ser/Thr residues. Although little is known about the functional roles of these glycan chains, KDN-gp appears to form a model for further study on the function of cell surface receptor for sperm in fertilization.

In a previous report (Kitajima, K., Inoue, S., and Inoue, Y. (1989) Dev. Biol. 132, 544-553), we found the presence of a heavily glycosylated polyprotein, ''H-hyosophorin,'' isolated from the unfertilized eggs of Oryzias latipes. We now report our detailed analysis of the structure of the N-glycan chain in L-hyosophorin, the smallest repeating unit of H-hyosophorin, which was isolated from the fertilized eggs of O. latipes and formed from H-hyosophorin upon fertilization. The N-glycan structures were defined by a combination of compositional analysis, methylation analysis, selective chemical degradation (i.e. mild methanolysis, periodate-Smith degradation, and hydrazinolysis nitrous acid deamination), enzymatic (endo-beta-galactosidase, peptide:N-glycanase and Newcastle disease virus sialidase) digestion, and instrumental analyses (one- and two-dimensional proton nuclear magnetic resonance spectroscopy and fast atom bombardment mass spectrometry) which revealed novel and unique features: (a) the presence of highly branched poly-N-acetyllactosamino pentaantennary structures; (b) the presence of a beta-galactosylated Lewis X antigenic epitope, Gal beta 1 --> 4 Gal beta 1 --> 4 (Fuc alpha 1 --> 3)GlcNAc beta 1 -->; (c) the presence of a beta-galactosylated sialyl Lewis X structure, Gal beta 1 --> 4(Neu5Ac alpha 2 --> 3)Gal beta 1 --> 4(Fuc alpha 1 --> 3)GlcNAc beta 1 -->; (d) the presence of Gal beta 1 --> 4Gal beta 1 --> and Gal beta 1 -->; 4Gal beta 1 --> 4Gal beta 1 --> as the major and minor groupings, respectively; and (e) the presence of the branched Gal residues, --> 4GlcNAc beta 1 --> 3(Gal beta 1 --> 4) Gal beta 1 -->. This study represents the first detailed investigation regarding the nature of highly branched complex asparagine-linked pentaantennary glycans in glycoproteins. The unique expression of such bulky multiantennary glycan units on proteins could be essential during early embryogenesis.

The Tyr residues in positions 32 and 40 of human c-Ha-Ras protein were replaced by site-directed mutagenesis (Y32F, Y32W, Y40K, and Y40W) to examine their roles in the signal-transducing activity and the sensitivity to the GTPase activating protein (GAP). The signal-transducing activity of the oncogenic Ras protein in PC12 cells was lost upon mutations Y32F and Y40K, but retained upon mutations Y32W and Y40W. These results suggest that residues 32 and 40 are both required to have aromatic groups and residue 32 is further required to have a hydrogen donor. On the other hand, three mutations (Y32F, Y32W, and Y40W) caused no appreciable reduction in either GAP-binding affinity or GAP sensitivity. By the Y40K mutation, GAP-binding affinity was slightly lowered, while GAP sensitivity was drastically impaired. Therefore, for residues 32 and 40 of Ras, interactions with GAP appear to be different from those with the target of signal transduction in the PC12 cell. As for the Y32W-Ras protein bound with an unhydrolyzable GTP analogue (GMPPNP), the Trp32 fluorescence is appreciably red-shifted, weaker, and more susceptible to KI quenching as compared to that of the GDP-bound form. Two-dimensional NMR spectroscopy with selectively deuterated Ras proteins revealed fewer and weaker nuclear Overhauser effects on the aromatic protons of Trp32 in the GMPPNP-bound form than in the GDP-bound form. This indicates that the side chain of Trp32 is more exposed to the solvent in the GMPPNP-bound form than in the GDP-bound form. In contrast, fluorescence and NMR analyses for the Y40W-Ras protein indicate that the microenvironment of Trp40 is only slightly affected upon GDP-GMPPNP exchange. Accordingly, the aromatic residue in position 32, rather than that in position 40 of the Ras protein, becomes exposed to the surface of the protein upon GDP-GTP exchange, allowing hydrogen bonding with a target molecule in the signal-transduction pathway.

The three-dimensional solution structure of the colicin E3 immunity protein (84 residues) was determined by distance geometry calculations. The hydrophilic side of a four-stranded antiparallel beta-sheet constitutes a part of the surface of the protein, and two loops lie on the hydrophobic side of the sheet. All the three specificity-determining residues, which are included in the center of the beta-sheet, display their side groups on the protein surface.

Two acidic glycosphingolipids were isolated and purified from rainbow trout ovarian fluid. They were designated as ovarian fluid gangliosides ofg-2a and ofg-2b. Both of these glycolipids were found to contain glucose, galactose, and N-acetylgalactosamine in a molar ratio of 1:2:1, but they differ by the presence of 2 mol of deaminated neuraminic acid (KDN; 2-keto-3-deoxy-D-glycero-D-galacto-nononic acid) in ofg-2a and 1 mol each of KDN and 9-O-acetyl-KDN in ofg-2b. On the basis of composition analysis, methylation analysis, mild acid hydrolysis, fast atom bombardment mass spectrometry (FABMS), 400-MHz H-1 nuclear magnetic resonance spectroscopy, and immunochemical analysis using a monoclonal antibody (mAb.kdn3G), the complete structures of these gangliosides were determined to be KDNalpha2-->3Galbeta1-->3GalNAcbeta1-->4(KDNalpha2-->3)Galbeta1-->4Glcbeta1--->Cer for ofg-2a [(KDN)G(D1a)] and 9-O-AcKDNalpha2-->3Galbeta1-->3GalNAcbeta1-->4(KDNalpha2-->3)Galbeta1-->4Glcbeta1-->Cer for ofg-2b [(KDN)G(D1a)-(OAc+)]. The ceramide moieties (Cer) in both ofg-2a [(KDN)G(D1a)] and ofg-2b [(KDN)G(D1a)(OAC+)] were found by combining of the results from fatty acid analysis and FABMS measurements to be made up of 4-sphingenine and mainly a C24:1 fatty acyl chain (nervonate). The structures of ofg-2a and ofg-2b are novel, and they represent the second example of naturally occurring KDN-gangliosides. Mild acid hydrolysis of both ofg-2a and ofg-2b resulted in formation of (KDN)G(M1a).

All the backbone H-1 and N-15 magnetic resonances (except for Pro residues) of the GDP-bound form of a truncated human c-Ha-ras proto-oncogene product (171 amino acid residues, the Ras protein) were assigned by N-15-edited two-dimensional NMR experiments on selectively N-15-labeled Ras proteins in combination with three-dimensional NMR experiments on the uniformly N-15-labeled protein. The sequence-specific assignments were made on the basis of the nuclear Overhauser effect (NOE) connectivities of amide protons with preceding amide and/or C(alpha)protons. In addition to sequential NOEs, vicinal spin coupling constants for amide protons and C(alpha)protons and deuterium exchange rates of amide protons were used to characterize the secondary structure of the GDP-bound Ras protein; six beta strands and five helices were identified and the topology of these elements was determined. The secondary structure of the Ras protein in solution was mainly consistent with that in crystal as determined by X-ray analyses. The deuterium exchange rates of amide protons were examined to elucidate the dynamic properties of the secondary structure elements of the Ras protein in solution. In solution, the beta-sheet structure in the Ras protein is rigid, while the second helix (A66-R73) is much more flexible, and the first and fifth helices (S17-124 and V152-L171) are more rigid than other helices. Secondary structure elements at or near the ends of the effector-region loop were found to be much more flexible in solution than in the crystalline state.

Imino N-15 and H-1 resonances of Escherichia coli tRNA(1)(Ile) were observed in the absence and presence of E coli isoleucyl-tRNA synthetase. Upon complex formation of tRNAPe with isoleucyl-tRNA synthetase, some imino N-15-H-1 resonances dis appeared, and some others were significantly broadened and/or shifted in the H-1 chemical shift, while the others were observed at the same N-15-H-1 chemical shifts. It was indicated that the binding of tRNA(1)(Ile) with IleRS affect the following four regions: the anticodon stem, the junction of the acceptor and T stems, the middle of the D stem, and the region where the tertiary base pair connects the T, D, and extra loops. This result is consistent with those of chemical footprinting and site-directed mutagenesis studies. Taken together, these three independent results reveal the recognition mechanism of tRNA(Ile)(1) by IleRS: IleRS recognizes all the identity determinants distributed throughout the tRNA(1)(Ile) molecule, which induces changes in the secondary and tertiary structures of tRNA(1)(ILe).

A truncated human c-Ha-Ras protein that lacks the C-terminal 18 amino acid residues and the truncated Ras protein with the amino acid substitution Gly --> Val in position 12 were prepared by an E. coli overexpression system. The truncated Ras protein showed the same guanine-nucleotide binding activity and GTPase activity as those of the full-length Ras protein. Further, the same extent of GTPase activity enhancement due to GTPase-activating protein was observed for the truncated and full-length Ras proteins. In fact, two-dimensional proton NMR analyses indicated that the tertiary structure of the truncated Ras protein (GDP-bound or GMPPNP-bound) was nearly the same as that of the corresponding catalytic domain of the full-length Ras protein. Moreover, a conformational change around the effector region upon GDP --> GMPPNP exchange occurred in the same manner for both proteins. These observations indicate that the C-terminal flanking region (18 amino acid residues) of the Ras protein does not appreciably interact with the catalytic domain. Therefore, the truncated Ras protein is suitable for studying the molecular mechanism involved in the GTPase activity and the interaction with the GTPase-activating protein. On the other band, an active form of the truncated Ras protein, unlike that of the full-length Ras protein, did not induce neurite outgrowth of rat pheochromocytoma PC12 cells. Thus, membrane anchoring of the Ras protein through its C-terminal four residues is not required for the interaction of Ras and GAP, but may be essential for the following binding of the Ras-GAP complex with the putative downstream target.

By performing H-1-H-1 and H-1-N-15 two-dimensional (2D) nuclear magnetic resonance (NMR) experiments, the complete sequence-specific resonance assignment was determined for the colicin E3 immunity protein (84 residues; ImmE3), which binds to colicin E3 and inhibits its RNase activity. First, the fingerprint region of the spectrum was analyzed by homonuclear H-1-H-1 HOHAHA and NOESY methods. For the identification of overlapping resonances, heteronuclear H-1-N-15 (HMQC-HOHAHA, HMQC-NOESY) experiments were performed, so that the complete H-1 and N-15 resonance assignments were provided. Then the secondary structure of ImmE3 was determined by examination of characteristic patterns of sequential backbone proton NOEs in combination with measurement of exchange rates of amide protons and 3J(HN-alpha) coupling constants. From these results, it was concluded that ImmE3 contains a four-stranded antiparallel beta-sheet (residues 2-10, 19-22, 47-49, and 71-79) and a short alpha-helix (residues 31-36).