Glutathione reductase (NADPH+GSSG+H+-->NADP(+) + 2GSH) is a homodimeric flavoenzyme of known geometry. Each subunit contains four well-defined domains and contributes essential residues to the active sites; consequently, the monomer is expected to be inactive. As part of our program to develop dimerization inhibitors of human glutathione reductase (hGR) as antimalarial agents, we mutagenized the residues 446 and 447 which, together with their counterparts on the other subunit, represent the tightest contact between the subunits [Karplus, P. A., & Schulz, G. E. (1987) J. Mol. Biol. 195, 701-729]. Wild-type human glutathione reductase and mutants of this protein were produced in plasmid-transformed Escherichia coli SG5 cells. Active enzyme species, namely, wild-type hGR, N-terminally truncated delta(1-15)hGR, and the point mutant F447P-hGR, were purified by 2',5'-ADP-Sepharose chromatography and crystallization. Inactive mutants such as G446E-hGR or the double mutants G446E/F447P-hGR and G446P/F447P-hGR were isolated by immunoadsorption chromatography. G446E/F447P-hGR was studied in detail. This mutant behaved like a poorly folded monomeric protein, as indicated by the following properties: absence of the intersubunit disulfide bridge, Cys90-Cys90'; failure to bind FAD; failure to bind NADPH and analogues thereof; a short half-life (< 4 min) in E. coli cells; and high susceptibility to trypsin in vitro. The results suggest that the sequence around G446 can control dimerization as well as domain folding. This is unexpected since the FAD-binding domain and the NADPH-binding domain occur in many different enzymes and have been regarded as autonomous folding units.(ABSTRACT TRUNCATED AT 250 WORDS)

We investigated regulation of macrophage prostaglandin production during activation by interferon gamma (IFN-gamma) and lipopolysaccharide (LPS). An in vitro model was established using the mouse macrophage-like cell line RAW 264.7. Cells were cultivated in the presence of IFN-gamma and LPS for up to 48 h and changes in the secretion of nitric oxide (NO.) and tumor necrosis factor alpha (TNF-alpha) were observed as activation markers. Under these conditions a prompt and strong increase in PGE2 production was found in the first 8 h followed by nearly constant generation of PGE2 during the next 40 h. In contrast, the activity of prostaglandin endoperoxide synthase (PGHS), measured as PGE2 production of microsomal protein fractions, was also increased, but reached a clear maximum at 24 h. Recently a second form of PGHS was cloned (PGHS-2) and specific antibodies and mRNA probes for both isoforms are available. PGHS-2 enzyme was expressed maximally after 24 h of activation whereas PGHS-1 was not influenced. In the presence of IFN-gamma and LPS, PGHS-2 mRNA expression reached a maximum at 8 h but PGHS-1 mRNA was not induced during the whole time period. These data indicate that changes in PG synthesis following macrophage activation are due to regulation of PGHS-2 expression.

Human glutathione reductase (GR; which catalyzes the reaction NADPH + GSSG + H' -2 GSH + NADP') is an obligatory FAD-containing homodimer of known geometry. Native human GR, a potential target of antimalarial and cytostatic agents, cannot be dissociated by dilution or by means of subunitinterface mimetics, similarly to well-studied viral dimeric proteins. However, ab initio folding andlor dimerization of human GR can be inhibited by point mutations or by peptides corresponding to subunitinterface areas, for example synthetic peptide P I 1, which represents the intersubunit-contact helix HI 1. The structure of this peptide, which might assist inhibitor design, was solved by high-resolution NMR spectroscopy. Residues 440-453, were found to be a helical in the isolated peptide. To quantitate the efficacy of inhibitors such as P I 1, we developed the following unfolding/reactivation assay. The effects of various guanidine hydrochloride (Gdn/HCl) concentrations were studied by analytical ultracentrifugation. It was shown that human GR denatured by greater than 3 M Gdn/HCl is monomeric and free of FAD. Circular-dichroism experiments at 223 nm indicated a half-life of approximately 20 s at 20°C for the unfolding process. To optimize the reactivation yield, four parameters [protein concentration (x) in the range 0.3-10 pg/ml, cofactor supplementation, temperature (y; 0-32"C), and time (0-72 h)] were varied systematically, and a reactivation score z was given to each constellation of parameters. This type of analysis might be useful to optimize refolding and activation yields for other proteins. For human GR, the highest recovery was found not to occur at one of the corners of the x,y plain, but close to its center. Consequently, the optimal assay conditions for folding and dimerization inhibitors are as follows. The enzyme (at 300 pg/ml) is denatured by 5 M guanidine hydrochloride/5 mM dithiothreitol, then reactivated by dilution to 1 pg/ml at pH 6.9 and 20°C. In the absence of inhibitors, this procedure leads to 70% of the control activity within 8 h. Peptides representing the upper subunit interface (for instance residues 436-478) of human GR were found to inhibit refolding with EC,,,, values in the micromolar range, whereas fragments from other regions of the protein had no influence on this process. For peptide P11, the EC,,,, value was 20 pM. In conclusion, hGR, enzyme with a tight intersubunit contact area of 21 nm', appears to be suitable for studying protein folding, dimerization, and prosthetic-group complexation in the absence and presence of compounds that inhibit these processes. There is a shortage, at least for oligomeric enzymes of eukaryotes, of published systematic studies on protein (re)activation.Keywords: unfolding/reactivation ; human glutathione reductase; antiparasitic drug; protein-dimerization inhibitor; peptide-structure determination by NMR. Neuenheimer Feld 328, D-69120 Heidelberg, Germany. of glutathione, respectively ; Williams, 1992). Since each binding site for the substrate GSSG and each...

No abstract