Jos Cox

Jos A. Cox received a PhD from the University of Leuven, Belgium. Thereafter he spend four years doing field research in central Africa and teaching at the University of Lovanium, Congo-Kinshasa. In 1973 he moved to Geneva as a postdoc and later became group leader (Maître d'enseignement et de recherche) in the department of Biochemistry. He heads an independent research unit on calcium-binding proteins since 1984.

OBJECTIVES

Calcium-binding proteins (0.13% of human genes) have been exquisitely fine tuned to perform vital functions inside the cell. They either interfere with the Ca2+ fluxes (Ca2+ buffers), or constitute the primary sensors of the Ca2+ signal by activating target proteins (activators). All are characterized by greatly varying affinities for Ca2+, different selectivities for cations and complex binding patterns with positive and negative cooperativities. Moreover, when they interact with their target proteins, these affinities for Ca2+ increase together with a complication of the binding pattern. Elucidating this complexity in many subfamilies of the EF-hand containing calcium-binding proteins has been the main theme of our research since 1973. The aim of our research is to provide detailed descriptions of newly discovered interactions and of new conformational states. Our recent research focuses on different EF-hand proteins involved in keratinocyte differentiation and in the control of the levels of reactive oxygen species (ROS).

BACKGROUND

In the skin the transition from proliferation to differentiation is strictly controlled by extracellular Ca2+, which forms a gradient directed from the basal to the granular layers. In the finally differentiating cells a 15 nm-thick sheath of proteins is formed beneath the plasma membrane due to cross-linking of more than 10 different proteins by protein transglutaminases. A novel Ca2+-binding protein CLSP (for CaM-like skin protein), very specifically expressed in terminally differentiating keratinocytes, but not in proliferating ones, is a very good candidate regulator of the differentiation. Other Ca2+-binding proteins, S100A8 and 9, exclusively expressed in myeloid cells and differentiating keratinocytes, gained recently very much interest. The S100 subfamily contains about 20 members, each of which exhibits a unique pattern of tissue/cell type specific expression and often associated with metastatic phenotypes. Psoriasin (S100A7) is overexpressed in psoriatic skin. At present the idea prevails that heterodimeric S100A8+9 acts as an intracellular reservoir for arachidonic acid and as an extracellular shuttle for transport the fatty acid to the appropriate cells.

Reactive oxygen species (ROS) imply a vast cocktail of species that often leads to oxidation of proteins, lipids and nucleic acids with negative (aging, cancerogenesis), but also positive (gene expression, oxygen sensing, blood pression regulation) effects on the cells. . The generation of ROS most often starts with the superoxide radical .O2- or with nitric oxide radical .NO. Superoxide is generated by NADPH oxidases. In phagocytes NADPH oxidase generates an electron current that flows from intracellular NADPH to extracellular oxygen and in the ologomeric enzyme gp91phox is the catalytic subunit. The search for nonphagocytic NADPH oxidases led to the discovery of a family of gp91phox homologues including NOX1 (in colon), NOX3 (fetal kidney), NOX4 found (kidney cortex) and the distant relative NOX5 found in testis and lymphoid organs. Whereas NOX1 and NOX4 produce low amounts of superoxide in a constitutively active manner, NOX5 contains an N-terminal extension with three canonical EF-hands and is able to generate high levels of superoxide and to conduct H+ ions in response to cytosolic free Ca2+ elevations.

ONGOING PROJECTS

1. EF hands in skin: CLSP and S100 proteins
   Our interest is to obtain by means of point and deletion mutations precise information on the location of the high-affinity Ca2+/Mg2+-mixed sites and low affinity Ca2+-specific sites in CLSP and their role in the structure and function of CLSP. We also found that the two main folding units in CLSP are quite independent and behave as in the case of calmodulin. CLSP complexes the model peptide melittin, nevertheless it does not activate the target enzymes and structural proteins of calmodulin. Based on this homology we are screening promising peptides of several putative CLSP target enzymes for high affinity and selectivity. Peptide design will be carried out to unravel the rules of high-affinity, high-selectivity interactions of these peptides not only with CLSP, but also with calmodulin, centrin and NOX5-EF (see below). When the single thiols Cys42 (S100A8) and Cys3 (S100A9) are labeled with suitable fluorescent thiol-reactive probes, free resonance energy transfer (FRET) is likely to occur in the heterodimer and not in the homodimers. This may be a very convenient method to monitor how different factors (Ca2+, Zn2+, Cu2+ and arachidonic acid) influence the formation of heterodimer, which is the biologically active species of this protein pair. In the heterodimer S100A8+9 Trp88 is about 10 Å away from the putative arachidonic acid binding and its fluorescence should be very sensitive to AA binding. This approach should allow us to monitor the binding (affinity and kinetics) of arachidonic acid. When the single thiols Cys42 (S100A8) and Cys3 (S100A9) are labeled with suitable fluorescent thiol-reactive probes, free resonance energy transfer (FRET) is likely to occur in the heterodimer and not in the homodimers. This may be a very convenient method to monitor how different factors (Ca2+, Zn2+, Cu2+ and arachidonic acid) influence the formation of heterodimer, which is the biologically active species of this protein pair. In the heterodimer S100A8+9 Trp88 is about 10 Å away from the putative arachidonic acid binding and its fluorescence should be very sensitive to AA binding. This approach should allow us to monitor the binding (affinity and kinetics) of arachidonic acid.
2. NOX5 in ROS production and aging
   Recently we established that the N-terminal cytoplasmic domain of NOX5, called NOX5-EF, possesses not only three canonical EF-hands, but also a PEF EF-hand and that the recombinant segment binds with high affinity and selectivity four Ca2+ ions. Ca2+-binding leads to pronounced conformational changes and allows interaction with melittin and synthetic peptides. We will carry out similar studies with an extended construct that contains NOX5-EF plus a 12 residue-long Pro/Arg segment. The role of the PEF-type EF-hand and of the other Ca2+-binding sites will be assessed by inactivation mutations.Recently we established that the N-terminal cytoplasmic domain of NOX5, called NOX5-EF, possesses not only three canonical EF-hands, but also a PEF EF-hand and that the recombinant segment binds with high affinity and selectivity four Ca2+ ions. Ca2+-binding leads to pronounced conformational changes and allows interaction with melittin and synthetic peptides. We will carry out similar studies with an extended construct that contains NOX5-EF plus a 12 residue-long Pro/Arg segment. The role of the PEF-type EF-hand and of the other Ca2+-binding sites will be assessed by inactivation mutations.
   We assume that the mechanism of activation implies that NOX5-EF interacts with the C-terminal cytoplasmic flavin-binding segment of the enzyme, likely with the basic segment SKRLSRSVTMRKSQRSS, which is unique to isoform 5 of NOX. Therefore we intend to study NOX5-EF interaction with this synthetic peptide. Also the whole C-terminal cytoplasmic tail, containing the binding sites for FAD and NADPH must be expressed since it constitutes an independent folding unit and may contain the cofactors. It may be an excellent tool for the interaction with NOX5-EF. We will study 2 aspects of the enzymatic activity of NOX5 with a in vitro assay: First, compare the Ca2+ response of enzyme and search for modulators of the Ca2+-sensitivity. Second, screen NOX5-EF inhibitors (models used in CLSP research) for their potency to inhibit the enzyme. This may allow us to set up the rules for specific binding and inhibition of new inhibitors of NOX5 enzyme, that could be physiologically important in neurodegenerative diseases. We assume that the mechanism of activation implies that NOX5-EF interacts with the C-terminal cytoplasmic flavin-binding segment of the enzyme, likely with the basic segment SKRLSRSVTMRKSQRSS, which is unique to isoform 5 of NOX. Therefore we intend to study NOX5-EF interaction with this synthetic peptide. Also the whole C-terminal cytoplasmic tail, containing the binding sites for FAD and NADPH must be expressed since it constitutes an independent folding unit and may contain the cofactors. It may be an excellent tool for the interaction with NOX5-EF. We will study 2 aspects of the enzymatic activity of NOX5 with a in vitro assay: First, compare the Ca2+ response of enzyme and search for modulators of the Ca2+-sensitivity. Second, screen NOX5-EF inhibitors (models used in CLSP research) for their potency to inhibit the enzyme. This may allow us to set up the rules for specific binding and inhibition of new inhibitors of NOX5 enzyme, that could be physiologically important in neurodegenerative diseases.

RECENT PUBLICATIONS

1. New perspectives on S100 proteins: a multi-functional Ca2+-, Zn2+- and Cu2+- binding protein family. Heizmann, C.W., and Cox, J.A. (1998) in BioMetals 11, 383-397.
2. The binding of calcium to the B-repeat segment of SdrD, a cell surface protein of Staphylococcus aureus. Josefsson, E., O'Connell, D., Foster, T, Durussel, I., and Cox, J.A. (1998) J. Biol. Chem. 273, 31145-31152.
3. Reactivation of the AB site of rat parvalbumin yields a functional Ca2+-binding site. Cox, J.A., Durussel, I., Scott, D.J., and Berchtold, M.W (1999) Eur. J. Biochem. 264, 790-799.
4. Genomic structure of the amphioxus calcium vector protein. Yuasa, H.J., Cox, J.A., and Takagi, T. J. (1999) J. Biochem. 126, 572-577.
5. Two forms of the apoptosis linked protein ALG-2 with different Ca2+ affinities and target recognition. Tarabykina, S., Moller, A.L., Durussel, I., Cox. J.A., and Berchtold, M.W. (2000) J. Biol. Chem. 275, 10514- 10518.
6. S100A13: biochemical characterization and subcellular localization in different cell lines. Ridinger, K., Schäfer, B.W., Durussel, I., Cox, J.A., and Heizman, C.W. (2000) J. Biol. Chem. 275, 8686-8694.
7. Ligand-induced conformational and stability changes in Nereis sarcoplasmic calcium binding protein: Evidence that the apo state is a molten globule. Christova, P., Cox, J. A., and Craescu, C.T. (2000) Proteins 40, 177-184.
8. Cation- and peptide-binding properties of human centrin 2. Durussel, I., Blouquit, Y., Middendorp, S., Craescu, C.T., and Cox, J.A. (2000) FEBS Lett. 472, 208-212.
9. Calcium binding to the regulatory domain of calcium vector protein reveals structural asymmetry and is accompanied by large conformational changes. Théret, I., Baladi, S., Cox, J.A., Sakamoto, H., and Creascu, C.T. (2000) Biochemistry 39, 7920-7926.
10. S100A5 is a novel calcium-, zinc-, and copper-binding protein of the EF-hand superfamily. Schäfer, B.W., Fritschy, J.-M., Murmann, P., Durussel, I., Heizmann, C.W., and Cox, J.A. (2000) J. Biol. Chem. 275, 30623-30630.
11. The domain structure of calcium vector protein of amphioxus: structural and functional properties of the recombinant N- and C-teminal halves. Baladi, S., Tsvetkov, P.O., Petrova, T.V., Takagi, T., Sakamoto, H., Lobachov, V.M., Makarov, A.A, and Cox, J.A. (2001) Prot. Sci. 10, 771-778.
12. Biochemical characterization of the penta-EF-hand protein grancalcin and identification of L-plastin as a binding partner. Lollike, K., Johnsen, A.H., Durussel, I., Borregaard, N., and Cox, J.A. (2001) J. Biol. Chem. 276, 17762-17769.
13. Solution structure and backbone dynamics of the defunct domain of calcium vector protein. Théret, I., Baladi, S., Cox, J.A., Gallay, J., Sakamoto, H. and Craescu, C.T. (2001) Biochemistry 40, 13888-13897.
14. Cation- and peptide-binding properties of the human calmodulin-like skin protein CLSP. Durussel, I., Méhul, B., Bernard, D., Schmidt, R., and Cox, J.A. (2002) Biochemistry 41, 5439-5448.
15. Distinct translocation pathways of S100A13 and S100A6 in repsonse to increase of intracellular calcium levels in human endothelial cells. Hsieh, H.-L., Cox, J.A., Schäfer, B.W., and Heizmann, C.W. (2002) J. Cell Sci. 115, 3149-3158.

THE GROUP