Solution NMR structure of the myosin phosphatase inhibitor protein CPI-17 shows phosphorylation-induced conformational changes responsible for activation.

Contractility of vascular smooth muscle depends on phosphorylation of myosin light chains, and is modulated by hormonal control of myosin phosphatase activity. Signaling pathways activate kinases such as PKC or Rho-dependent kinases that phosphorylate the myosin phosphatase inhibitor protein called CPI-17. Phosphorylation of CPI-17 at Thr38 enhances its inhibitory potency 1000-fold, creating a molecular on/off switch for regulating contraction. We report the solution NMR structure of the CPI-17 inhibitory domain (residues 35-120), which retains the signature biological properties of the full-length protein. The final ensemble of 20 sets of NMR coordinates overlaid onto their mean structure with r.m.s.d. values of 0.84(+/-0.22) A for the backbone atoms. The protein forms a novel four-helix, V-shaped bundle comprised of a central anti-parallel helix pair (B/C helices) flanked by two large spiral loops formed by the N and C termini that are held together by another anti-parallel helix pair (A/D helices) stabilized by intercalated aromatic and aliphatic side-chains. Chemical shift perturbations indicated that phosphorylation of Thr38 induces a conformational change involving displacement of helix A, without significant movement of the other three helices. This conformational change seems to flex one arm of the molecule, thereby exposing new surfaces of the helix A and the nearby phosphorylation loop to form specific interactions with the catalytic site of the phosphatase. This phosphorylation-dependent conformational change offers new structural insights toward understanding the specificity of CPI-17 for myosin phosphatase and its function as a molecular switch.

NMR structure of Streptomyces killer toxin-like protein, SKLP: further evidence for the wide distribution of single-domain betagamma-crystallin superfamily proteins.

A protein isolated from the culture supernatant of the soil bacterium, Streptomyces sp. F-287, exhibits cytocidal effects for both budding and fission yeasts, and causes morphological changes of yeasts and filamentous fungi. This protein, which was the first killer toxin-like protein for yeasts identified in the Streptomyces microorganism, was named SKLP (Streptomyces killer toxin-like protein). Since the amino acid sequence of the protein, as determined by sequential Edman degradations, seemed to be unique, we determined the structure by NMR spectroscopy. Although the actual target of SKLP in yeasts has not been determined yet, the structure might give us a clue to characterize the targets. The solution structure of SKLP determined by NMR, however, turned out to be a single-domain crystallin-like protein, with two Greek key motifs and a short extra beta-strand at the N terminus. The final ensemble of 20 NMR structures overlaid onto their mean coordinate with rmsd values of 0.32(+/-0.06) A for the backbone atoms involved in the secondary structure elements. As a yeast killer toxin, WmKT, isolated from the yeast strain Williopsis mrakii also has a Greek key beta-barrel fold, we have made a detailed comparison of the structural features of SKLP with the other crystallin superfamily proteins. It is very interesting that SKLP has a unique electrostatic potential distribution on the molecular surface. Namely, one surface of the beta-barrel fold in SKLP has a large negatively charged region, with an isolated positive charge of the Arg62 side-chain at the center. The edge of this surface is surrounded by positively charged residues, including Arg31, Arg65 and Arg74. The salient features of the charge distribution on this surface and the cluster of Arg residues might be related to the target binding of SKLP.

Structural comparison between wild-type and P25S human cystatin A by NMR spectroscopy. Does this mutation affect the alpha-helix conformation?

The effect of substituting Pro25, located in the alpha-helical region of the cystatin A structure, with Ser has been studied. The structures of wild type and P25S cystatin A were determined by multidimensional NMR spectroscopy under comparable conditions. These two structures were virtually identical, and the alpha-helix between Glu15-Lys30 exists with uninterrupted continuity, with a slight bend at residue 25. In order to characterize the possible substitution effects of Pro25 with Ser on the alpha-helix, the chemical shifts of the amide nitrogens and protons, the generalized order parameters obtained by the analyses of the 15N-1H relaxation data, the amide proton exchange rates, and the NOE networks among the alpha-helical and surrounding residues were carefully compared. None of these parameters indicated any significant static or dynamic structural differences between the alpha-helical regions of the wild-type and P25S cystatin A proteins. We therefore conclude that our previous structure of the wild-type cystatin A, in which the alpha-helix exhibited a sharp kink at Pro25, must be revised. The asymmetric distribution of hydrophobic interactions between the side-chain residues of the alpha-helix and the rolled beta-sheet surface, as revealed by NOEs, may be responsible for the slight bend of the alpha-helix in both variants and for the destabilized hydrogen bonding of the alpha-helical residues that follow Pro25/Ser25, as evidenced by increased amide exchange rates.