Knut Langsetmo

Structural basis of protein phosphatase 1 regulation

The coordinated and reciprocal action of serine/threonine (Ser/Thr) protein kinases and phosphatases produces transient phosphorylation, a fundamental regulatory mechanism for many biological processes. The human genome encodes a far greater number of Ser/Thr protein kinases than of phosphatases. Protein phosphatase 1 (PP1), in particular, is ubiquitously distributed and regulates a broad range of cellular functions, including glycogen metabolism, cell-cycle progression and muscle relaxation. PP1 has evolved effective catalytic machinery but lacks substrate specificity. Substrate specificity is conferred upon PP1 through interactions with a large number of regulatory subunits. The regulatory subunits are generally unrelated, but most possess the RVxF motif, a canonical PP1-binding sequence. Here we reveal the crystal structure at 2.7 Å resolution of the complex between PP1 and a 34-kDa N-terminal domain of the myosin phosphatase targeting subunit MYPT1. MYPT1 is the protein that regulates PP1 function in smooth muscle relaxation. Structural elements amino- and carboxy-terminal to the RVxF motif of MYPT1 are positioned in a way that leads to a pronounced reshaping of the catalytic cleft of PP1, contributing to the increased myosin specificity of this complex. The structure has general implications for the control of PP1 activity by other regulatory subunits.

A structure-based interpretation of E.coli GrpE thermodynamic properties.

GrpE is the nucleotide exchange factor for the Escherichia coli molecular chaperone DnaK, the prokaryotic homologue of Hsp70. Thermodynamic properties of GrpE structural domains were characterized by examining a number of structural and point mutants using circular dichroism, differential scanning calorimetry and analytical ultracentrifugation. These structural domains are the long paired N-terminal helices, the central four-helix bundle, and the C-terminal compact beta-domains. We show that the central four-helix bundle (t(m) approximately 75 degrees C) provides a stable platform for the association of the long paired N-terminal helices (t(m) approximately 50 degrees C), which can then function as a temperature sensor. The stability of the N-terminal helices is linked to the presence of the C-terminal compact beta-domains of GrpE, providing a potential mechanism for coupling of DnaK-binding activity of GrpE with temperature. On the basis of our thermodynamic analysis of E.coli GrpE, we present a structure-based model for the melting properties of the nucleotide exchange factor, wherein the long paired helices function as a molecular thermocouple.

Isoforms of the small non-catalytic subunit of smooth muscle myosin light chain phosphatase

Chicken gizzard smooth muscle myosin light chain phosphatase is composed of a approximately 37 kDa catalytic subunit, a approximately 110 kDa myosin binding or targeting subunit and a approximately 20 kDa subunit (MPs) whose function is as yet undefined. It was reported previously that a cloned chicken gizzard MPs cDNA encodes a protein of 186 amino acids (aa) [Y.H. Chen, M.X. Chen, D.R. Alessi, D.G. Gampbell, C. Shanahan, P. Cohen, P.T.W. Cohen, FEBS Lett. 356 (1994) 51-55]. More recently, we obtained by PCR amplification another MPs cDNA that encodes a protein of only 161 aa [Y. Zhang, K. Mabuchi, T. Tao, Biochim. Biophys. Acta 1343 (1997) 51-58]. In this work we obtained cDNAs corresponding to both sequences using a different set of PCR primers, indicating that the two sequences correspond to isoforms that most likely arose from alternative splicing of the same gene. Using two polyclonal antibodies, one raised against the recombinant 161 aa isoform of chicken gizzard MPs and the other against a C-terminal polypeptide that is present only in the 186 aa isoform, we found that while the 161 aa isoform is the predominant one in chicken gizzard, in chicken aorta it is the 186 aa one; in chicken stomach both isoforms are present, and in mammalian tissues such as ferret and rat only the 186 aa isoform is detected. Furthermore, we purified the MPs associated with the chicken gizzard myosin light chain phosphatase holoenzyme and determined its molecular weight, amino acid composition and six residues of its C-terminal sequence. The results from these analyses showed conclusively that the predominant isoform in chicken gizzard is the 161 aa one.