Structure of the YSPTSPS repeat containing two SPXX motifs in the CTD of RNA polymerase II: NMR studies of cyclic model peptides reveal that the SPTS turn is more stable than SPSY in water.

The carboxyl-terminal domain of RNA polymerase II, which is rich in phosphorylation sites, contains 17--52 tandem repeats with the consensus sequence of the heptapeptide, YSPTSPS. The repeat unit of the heptapeptide has two SPXX motifs showing potential beta-turns, SPTS and SPSY. NMR studies were performed in water at pH 4.0 for two cyclic peptides containing one and two repeat units, cyclo-[C(1)R(2)D(3)Y(4)S(5)P(6)T(7)S(8)P(9)S(10)Y(11)S(12)R(13)D(14)C(15)] (peptide 1) and cyclo-[C(1)R(2)D(3)Y(4)S(5)P(6)T(7)S(8)P(9)S(10)Y(11)S(12)P(13)T(14)S(15)P(16)N(17)Y(18)S(19)R(20)D(21)C(22)] (peptide 2), which are cyclized with a disulfide bridge of two Cys residues at the N- and C-termini. SP in 1 and 2 are predominantly in trans form. The following NMR parameters were detected: (1) lower temperature coefficients of amide proton chemical shifts of T7 and S8 in 1, and Tx (T7 or T14), Sx (S8 or S15), Tz (T14 or T7) and Sz (S15 or S8) in 2, (2) significantly large deviation of H(alpha) chemical shifts from its random coil value (Delta H(alpha)) of Pro preceding the Thr (P6 in 1, and Px and Pz in 2), (3) relatively large (3)J(HNH alpha) coupling constants (>8.7 Hz) of T7 in 1 and Tx and Tz in 2, and (4) NOE (d(NN) (i, i+1)) connectivities between the amide protons of T7-S8 and S10-Y11 in 1, and Tx-Sx, S10-Y11, Tz-Sz, and N17-Y18 in 2, although two Pro-Thr-Ser segments in 2 (each of these are annotated by 'x' and 'z') in the first and second repeat units were not distinguishable. Comparison of the NMR parameters between the cyclic peptides and the corresponding linear peptides indicates that cyclization promotes structural stabilization in water. The present NMR data were consistent with the presence of a beta-turn at both SPTS and SPSY: S(5)P(6)T(7)S(8) and S(8)P(9)S(10)Y(11) in 1, and SPxTxSx, SPzTzSz, SP(9)S(10)Y(11), SP(16)N(17)Y(18) in 2. However, the structure of the SPTS segment is more stable than that of the SPSY segment. Conformations consistent with NMR parameters including NOE distances were obtained through molecular dynamics and energy minimization methods. These calculations yielded two stable conformers for the SPTS segment. One of the two corresponds to a type I beta-turn.

Solution structures of the N-terminal domain of yeast calmodulin: Ca2+-dependent conformational change and its functional implication.

We have determined solution structures of the N-terminal half domain (N-domain) of yeast calmodulin (YCM0-N, residues 1-77) in the apo and Ca(2+)-saturated forms by NMR spectroscopy. The Ca(2+)-binding sites of YCM0-N consist of a pair of helix-loop-helix motifs (EF-hands), in which the loops are linked by a short beta-sheet. The binding of two Ca(2+) causes large rearrangement of the four alpha-helices and exposes the hydrophobic surface as observed for vertebrate calmodulin (CaM). Within the observed overall conformational similarity in the peptide backbone, several significant conformational differences were observed between the two proteins, which originated from the 38% disagreement in amino acid sequences. The beta-sheet in apo YCM0-N is strongly twisted compared with that in the N-domain of CaM, while it turns to the normal more stable conformation on Ca(2+) binding. YCM0-N shows higher cooperativity in Ca(2+) binding than the N-domain of CaM, and the observed conformational change of the beta-sheet is a possible cause of the highly cooperative Ca(2+) binding. The hydrophobic surface on Ca(2+)-saturated YCM0-N appears less flexible due to the replacements of Met51, Met71, and Val55 in the hydrophobic surface of CaM with Leu51, Leu71, and Ile55, which is thought to be one of reasons for the poor activation of target enzymes by yeast CaM.

NMR studies of model peptides of PHGGGWGQ repeats within the N-terminus of prion proteins: a loop conformation with histidine and tryptophan in close proximity.

The N-terminal region of the prion protein from human and mouse contains five tandem repeats with the consensus sequence of PHGGGWGQ. NMR studies were performed in water for two cyclic peptides, cyclo-[C(1)R(2)Q(3)P(4)H(5)G(6)G(7)S(8)W(9)G(10)Q(11)R(12)D(13)C(14)] (C1) and cyclo-[C(1)R(2)D(3)P(4)H(5)G(6)G(7)G(8)W(9)G(10)Q(11)P(12)H(13)G(14)G (15)G(16)W(17)G(18)Q(19)R(20)D(21)C(22)] (C2), which are cyclized by a disulfide bridge between the Cys residues at the N- and C-termini, and for their corresponding linear peptides (L1 and L2) which are formed by reduction. The patterns of the C(alpha)H chemical shift difference of these four peptide mimetics were very similar to those observed for the tandem repeats of human prion protein reported by other researchers. The medium-range NOE connectivities were found between the C(beta)H of the H5 and the proton of the W9 side chain for L1. The corresponding NOEs were also observed in H5-W9 and H13-W17 of L2 with ambiguity. These observations indicate that histidine (i) is in close proximity to tryptophan (i+4). d(alphaN) (i,i+2) NOE connectivities were observed between W9 and Q11 of L1 and L2, and d(NN) (i,i+1) NOE connectivities were also observed for G10-Q11 of L1 and L2 and for G18-Q19 of L2. Significantly lower temperature coefficients of amide proton chemical shifts were obtained for Q11 and Q19 of L2 and C2. Structure calculations for L1 showed that HGG(G/S)W and (G/S)WGQ adopt a loop conformation and a beta-turn, respectively. These results strongly suggest that the tandem repeats within prion protein adopt a non-random structure.

Chitin-binding proteins in invertebrates and plants comprise a common chitin-binding structural motif.

Tachycitin, a 73-residue polypeptide having antimicrobial activity is present in the hemocyte of horseshoe crab (Tachypleus tridentatus). The first three-dimensional structure of invertebrate chitin-binding protein was determined for tachycitin using two-dimensional nuclear magnetic resonance spectroscopy. The measurements indicate that the structure of tachycitin is largely divided into N- and C-terminal domains; the former comprises a three-stranded beta-sheet and the latter a two-stranded beta-sheet following a short helical turn. The latter structural motif shares a significant tertiary structural similarity with the chitin-binding domain of plant chitin-binding protein. This result is thought to provide faithful experimental evidence to the recent hypothesis that chitin-binding proteins of invertebrates and plants are correlated by a convergent evolution process.

Early stage-stress relaxation in compact bone.

The early stage of stress relaxation, up to 10 s after strain application, in compact bone was investigated in order to find the limit of the applicability of the empirical equation (Sasaki et al., 1993. Journal of Biomechanics 26, 1369-1376), E(t) = E0{A1exp[-(t/tau1)beta]+A2exp(-t/tau)}, A1+A2 = 1, 0

Calcium binding induces interaction between the N- and C-terminal domains of yeast calmodulin and modulates its overall conformation.

Calmodulin from the yeast Saccharomyces cerevisiae binds 3 mol of Ca2+ cooperatively. We report here lines of evidence supporting the intramolecular interaction between the N- and C-terminal domains which modulates the Ca2+ binding properties of yeast calmodulin. First, the sum of the Ca2+ binding curves of the N-terminal and the C-terminal half-molecule did not yield the Ca2+ binding curve of yeast calmodulin. Second, the mean residue CD of yeast calmodulin at 222 nm (-Delta epsilon222) decreased with increases in the concentration of Ca2+, whereas those of each half-molecule increased. Finally, the C2 proton of His107 in the C-terminal domain of yeast calmodulin showed three resonance peaks with increases in the concentration of Ca2+, each corresponding to the apo, the intermediate, and the Ca2+-saturated state. The intermediate peak could not be observed in the C-terminal half-molecule of yeast calmodulin. Computer simulation considering the macroscopic Ca2+ binding constants assigned this intermediate to a species consisting of the apo C-terminal domain and the N-terminal domain with at least one of the two sites occupied by Ca2+. Peptide segments spanning the defective fourth Ca2+ binding site may be involved in the interdomain interaction and the yeast-specific function of calmodulin.

Solution structure of an insect growth factor, growth-blocking peptide.

Growth-blocking peptide (GBP) is an insect growth factor consisting of 25 amino acid residues that retards the development of lepidopteran larvae at high concentration while it stimulates larval growth at low concentration. In this study, we determined the solution structure of GBP by two-dimensional 1H NMR spectroscopy. The structure contains a short segment of double-stranded beta-sheet involving residues 11-13 and 19-21 and a type-II beta-turn in the loop region (residues 8-11), whereas the N and C termini are disordered. This is the first report of the three-dimensional structure of the peptiderigic insect growth factor, and the structure of the well defined region of GBP was found to share similarity with that of the C-terminal domain of the epidermal growth factor (EGF). Because GBP has been reported to stimulate DNA synthesis of not only insect cells but also human keratinocyte cells at the same level with EGF, the structural similarity between GBP and EGF may lead to the interaction of GBP to EGF receptor.

NMR characterization of physicochemical properties of rat D-dopachrome tautomerase.

D-dopachrome tautomerase (DOPD) catalyzes the conversion of D-dopachrome to 5,6-dihydroxyindole. DOPD was found to have amino acid sequence homology with macrophage migration inhibitor factor (MIF), suggesting that DOPD functions as a proinflammatory cytokine. We here demonstrate the physicochemical properties of DOPD by nuclear magnetic resonance (NMR). The native molecular weight of rat DOPD was about 37 kDa as calculated from Sephadex G-100 column chromatography. Since the deduced molecular weight of its cDNA (117 amino acid residues) is 12.5 kDa, it is assumed that DOPD exists as a homotrimer in native form. Since several methyl proton resonances were observed in the high magnetic field region of the one-dimensional 1H-NMR spectrum, DOPD appeared to have a hydrophobic core in which methyl groups and aromatic groups are located close to each other. From a simple integration of one-dimensional 1H-NMR spectra, the contents of the alpha-helix, beta-strand, and random coil were calculated to be 16%, 68%, and 16%, respectively. Since the denaturation temperature of DOPD is exceedingly high at the range between 90-100 degrees C, it is considered that the high beta-strand content may contribute to its heat stability.

Secondary structure and Ca(2+)-binding property of the N-terminal half domain of calmodulin from yeast Saccharomyces cerevisiae as studied by NMR.

Using two- and three-dimensional NMR techniques, 1H and main-chain 15N resonances of the N-terminal half domain of yeast calmodulin (YCM0-N) in the presence of Mg2+ and Ca2+ (Mg(2+)-and Ca(2+)-forms) were assigned. The secondary structures of YCM0-N in both forms were determined. The NOESY and 15N-edited NOESY spectra of YCM0-N in each form indicate that there is a hydrophobic core and that two Ca(2+)-binding loops are connected by a short antiparallel beta-sheet. There are four helices (A, B, C, and D named from the N-terminus) for YCM0-N in the Mg(2+)-form. The B-helix is, however, not formed in the Ca(2+)-form. The Ca(2+)-binding of YCM0-N was monitored by (1H,15N)-HSQC at various Ca2+ concentrations. The observed spectral changes as a function of Ca(2+)-concentration can not readily be grouped into a small number of classes; each residue shows individual spectral change. There is no apparent relationship between the spectral change and the type or location of the amino acid concerned.

Establishment and characterization of transgenic mice expressing human platelet glycoprotein Ib alpha.

The platelet glycoprotein (GP) Ib/IX/V is a hetero-oligomeric receptor complex for von Willebrand factor (vWF) and mediates platelet adhesion and aggregation under high shear stress conditions. It is composed of alpha and beta chain of GP Ib, GP IX, AND and GP V. To establish transgenic mice carrying human GP Ib alpha, we injected into mouse zygotes a 6 kb DNA fragment containing human GP Ib alpha gene that included entire coding sequence and putative promoter region. One hundred and thirteen offsprings were screened, and only one was found to express human GP Ib alpha protein and has passed the human GP Ib alpha gene as well as the expression of the gene to next generation. The expression of human GP Ib alpha in transgenic mice was limited to platelets and megakaryocytes. Glycocalicin, a proteolytic fragment of human GP Ib alpha found in normal human plasma, was not detected in transgenic mouse plasma. Human vWF in the presence of ristocetin supported agglutination of transgenic mouse platelets, but not of control mouse platelets.

Glycine-15 in the bend between two alpha-helices can explain the thermostability of DNA binding protein HU from Bacillus stearothermophilus.

On the basis of sequence comparison of thermophilic and mesophilic DNA binding protein HUs, Bacillus stearothermophilus DNA binding protein HU (BstHU) seems to gain thermostability with a change in amino acid residues present on the molecular surface. To evaluate the contribution of exchange of each amino acid to the thermostability of BstHU, we constructed three mutants, BstHU-T13A (Thr13 to Ala), BstHU-G15E (Gly15 to Glu), and BstHU-T33L (Thr33 to Leu), in which the amino acids in BstHU were changed to the corresponding ones in Bacillus subtilis DNA binding protein HU (BsuHU). Stability of the mutant proteins was determined from thermal-denaturation curves. Replacement of Gly15 located in the turn region between alpha 1 and alpha 2 helices (HTH motif), with Glu (BstHU-G15E), resulted in a decrease in thermostability, and the Tm value was 54.0 degrees C compared to the Tm value of 63.9 degrees C for BstHU. The mutants, BstHU-T13A and BstHU-T33L, were, by contrast, slightly more stable (Tm values of 67.0 and 65.6 degrees C for BstHU-T13A and BstHU-T33L, respectively) than the wild type. We then generated the BsuHU mutant protein BsuHU-E15G, where Glu15 in BsuHU was in turn replaced by Gly, and we analyzed the thermostability. This substitution clearly enhanced the melting temperature by 11.8 degrees C (Tm value: 60.4 degrees C for BsuHU-E15G) compared to the value for BsuHU (Tm: 48.6 degrees C). Thus, Gly15 in the HTH motif of BstHU has an important role in the thermostability of BstHU. Characterization of the structure of the BstHU-G15E by 1H-NMR analysis showed that solvent accessibility of amide proton of Ala21 in the mutant was significantly increased compared with that of wild type, which means that the structure of the HTH motif in the N-terminal region in the mutant was changed to a more open conformation, thereby avoiding the interaction of Ala21 with either Ser17 by hydrogen bond or Ala11 by hydrophobic interaction.

The structure and physicochemical properties of rat liver macrophage migration inhibitory factor.

We expressed rat macrophage migration inhibitory factor (rMIF) in E. coli using the cDNA isolated from a rat liver cDNA library. rMIF specifically bound glutathione (dissociation constant = 500 microM). We purified rMIF homogeneously on SDS-PAGE by S-hexylglutathione Sepharose affinity column chromatography and Sephadex G-100 column chromatography. The amino-acid sequence of rMIF was highly homologous to that of human MIF from a T-cell line; only a single amino-acid residue was substituted if conservative amino-acid substitutions were involved. The molecular weight of rMIF was calculated to be 12.4 kDa and 23.6 kDa by SDS-PAGE and analytical ultracentrifugation, respectively. Thus, it was concluded that the native rMIF formed a homodimeric structure. Proton nuclear magnetic resonance (1H-NMR) study revealed that rMIF was less thermostable (the denaturing temperature was from 50-60 degrees C) than human MIF (the denaturing temperature is about 80 degrees C (Nishihira et al. (1993) Biochem. Mol. Biol. Int. 31, 841-850). The secondary structure of rMIF evaluated by 1H-NMR experiments revealed that the contents of alpha-helix, beta-strand, and coil were 13.8%, 55.6%, and 30.6%, respectively.

13C-NMR relaxation study on mobility of the DNA-binding arm of HU.

The mobility of the DNA-binding arm of HU protein was studied by 13C-NMR spectroscopy. The correlation times tau c of Phe47C alpha in the body and Gly60C alpha in the arm of HU were determined for HU and HU-DNA complex. The value of tau c of Phe47C alpha is 2-4 times larger than that of Gly60C alpha irrespective of the presence or absence of DNA. These results show that Gly60C alpha undergoes more rapid motion than Phe47C alpha. The increase in correlation time on addition of DNA is greater for Gly60C alpha than for Phe47C alpha. This suggests that the addition of DNA influences more significantly the motion of Gly60C alpha than that of Phe47C alpha. These results are in accord with the X-ray result, in which the top part of the arm is not visible. Gly60C alpha in the arm is thus more mobile than Phe47C alpha in the body, and the mobility of Gly60C alpha is reduced by the DNA binding.

Cloning and characterization of human type 2 and type 3 inositol 1,4,5-trisphosphate receptors.

We have cloned cDNAs coding for human type 2 and type 3 and part of type 1 inositol 1,4,5-trisphosphate receptors (IP3Rs). The complete nucleotide sequences for type 2 and type 3 receptors were determined and the pharmacological properties of the latter were characterized. Human type 2 and type 3 IP3Rs are 2701 amino acids and 2671 amino acids long, respectively, and have significant sequence homologies as well as structural similarities including the six membrane-spanning regions near the C-termini when compared with the rat or mouse counterpart. COS-7 cells transfected with human type 3 IP3R showed characteristic inositol 1,4,5-trisphosphate (IP3)-binding properties with Kd values of 28.8 nM. The order of potency of competition with IP3 was Ins(1,4,5)P3 (IP3) > Ins(2,4,5)P3 > Ins(1,3,4,5)P4 > Ins(1,2,3,4,5,6)P6. Type 2 and type 3 IP3Rs were mapped to human chromosomes 12p11 and 6p21, respectively, by in situ hybridization. cDNA cloning of the human IP3Rs allowed us to identify the types of the receptor expressed in various human hematopoietic and lymphoma cell lines. The type 3 receptor was present in all of cell lines tested, while the type 1 or 2 receptor was expressed in only particular cell types. The differential expression of the IP3R types could confer the cell-specific regulation on the IP3/Ca2+ signalling.

Mg2+ inhibits formation of 4Ca(2+)-calmodulin-enzyme complex at lower Ca2+ concentration. 1H and 113Cd NMR studies.

Our previous 1H NMR studies indicated that when mastoparan (MP) is added to Ca(2+)-half-saturated calmodulin (2Ca(2+)-CaM) in the absence of Mg2+ ions, Ca2+ ions transfer from the C-terminal-half domain of CaM not interacting with MP to the N-terminal-half domain of CaM interacting with MP at lower MP concentrations (Ohki, S., Yazawa, M., Yagi, K., and Hikichi, K. (1991b) J. Biochem. (Tokyo) 110, 737-742). As a consequence, the active form of 4Ca(2+)-CaM.MP complex is formed. In the present study, we studied the effect of Mg2+ ions on Ca2+ transfer. In the presence of Mg2+ ions, such Ca2+ transfer does not occur. The effects of Mg2+ ions are also studied by observing 113Cd NMR in the presence of M13, the 26-residue peptide of the CaM-binding region of myosin light chain kinase. The 113Cd NMR results show that Mg2+ ions prevent to form the active complex. Mg2+ plays an important role as an inactivating factor to CaM.

Identification of the hydrophobic ligand-binding region in recombinant glutathione S-transferase P and its binding effect on the conformational state of the enzyme.

Recombinant glutathione S-transferase P (GST-P) was purified in a homogeneous state. Fatty acid analysis of the enzyme revealed that the final enzyme preparation endogenously bound fatty acids, mostly palmitic acid or stearic acid, which were difficult to dissociate from the complex. Temperature-dependent analysis by 1H NMR indicated that the molecular motion of fatty acids was strongly restrained under physiological conditions, which was significantly different from that of serum albumin. On the other hand, there existed another hydrophobic ligand-binding region in GST-P, to which 1-amino-8-naphthalenesulfonic acid and bilirubin would bind with relatively lower affinity than the endogenously bound fatty acid. The hydrophobic ligand-binding region was determined to be around 141-156 residues from the N-terminus by procedures including association of the enzyme to fatty acid-linked Sepharose and affinity labeling with fluorescent fatty acid. Furthermore, circular dichroism analysis showed that the binding of hydrophobic ligand to GST-P produced a remarkable conformational change of the enzyme, which led to states devoid of transferase activity. In addition, the hydrophobic ligand binding caused a significant fluorescence quenching of tryptophan 38, which was assumed to be located at the active center of GST-P. It could be the result of a conformational change of the active center of the enzyme.

1H-NMR study of Ca(2+)-and Mg(2+)-dependent interaction between troponin C and troponin I inhibitory peptide (96-116).

The Ca(2+)-and Mg(2+)-dependence of the interaction between rabbit skeletal muscle troponin C (TnC) and a 21 residue peptide corresponding to 96-116 of troponin I (denoted as CN4) was examined by means of 1H-NMR spectroscopy. The spectral changes of TnC with 4 mol of Ca2+ (Ca4TnC) and TnC with 4 mol of Mg2+ (Mg4TnC) were observed as a function of CN4 concentration. As CN4 was added to Ca4TnC, resonances of the following residues changed in chemical shift: Tyr10, Phe23, Phe72, Ala106, Gly108, Tyr109, Ile110, His125, Gly144, Ile146, Phe102 or Phe151, and Phe148 located in the N- and C-domains of Ca4TnC. Such CN4-induced change was also observed for resonances of Phe19, 26, and 75 in the N-domain of Ca4TnC by means of NOESY and HOHAHA experiments. The presence of CN4 increased the native-to-unfolded transition temperature of the N-domain of Ca4TnC. On the basis of these results, we conclude that CN4 binds to both the C- and N-domains of Ca4TnC ([CN4]:[TnC] = 1:1) and stabilizes the structure of the N-domain. The CN4-binding constant was estimated to be 1.1 x 10(5) M-1. As CN4 was added to Mg4TnC, chemical shift change was observed for resonances of Phe99, Tyr109, and Ile110 in the C-domain, while no change was observed for resonances arising from the N-domain. The presence of CN4 did not change the thermal stability of the N- and C-domains of Mg4TnC. The CN4-binding constant of Mg4TnC was obtained as 0.9 x 10(4) M-1, which is one-tenth of that of Ca4TnC.(ABSTRACT TRUNCATED AT 250 WORDS)

Characterization of cysteine residues of glutathione S-transferase P: evidence for steric hindrance of substrate binding by a bulky adduct to cysteine 47.

Glutathione S-transferase P (GST-P) lost the enzymatic activity by 7-fluoro-4-sulfamoyl-2, 1, 3-benzodiazole (ABD-F), a thiol-group chemical modifier, but did not by methylmethanethiol-sulfonate. Both ABD-F and methylmethanethiolsulfonate reacted with Cys47 and Cys101. These two cysteine residues were site-directedly mutated with serine residues. Only the Cys101Ser lost the enzymatic activity by the treatment of ABD-F. On carbon 13 NMR experiments, a NMR signal of S-[13C]CH3 adduct to Cys47 did not show any change by the addition of S-hexylglutathione. These facts revealed that Cys47 did not locate at the active site, and a bulky adduct to Cys47 hindered the binding of substrates to the active site.