A p-loop motif and two basic regions in the regulatory protein GvpD are important for the repression of gas vesicle formation in the archaeon Haloferax mediterranei

Institut für Mikrobiologie und Genetik, TU Darmstadt, Schnittspahnstr. 10, D-64287 Darmstadt, Germany 1 Author for correspondence: Felicitas Pfeifer. Tel: +49 6151 162957. Fax: +49 6151 162956. e-mail: pfeifer{at}bio.tu-darmstadt.de D transformants containing all 14 gvp genes of Haloferax mediterranei required for gas vesicle formation except for gvpD are gas vesicle overproducers (Vac ++ ), whereas D/D transformants containing the gvpD reading frame under ferredoxin promoter control on a second construct in addition to D did not form gas vesicles (Vac - ). The amino acid sequence of GvpD indicates three interesting regions (a putative nucleotide-binding site called the p-loop motif, and two basic regions); these were altered by mutation, and the resulting GvpD mut proteins tested in D/D mut transformants for their ability to repress gas vesicle formation. The exchange of amino acids at conserved positions in the p-loop motif resulted in Vac ++ D/D mut transformants, indicating that these GvpD mut proteins were unable to repress gas vesicle formation. In contrast, a GvpD mut protein with an alteration of a non-conserved proline in the p-loop region (P41A) was still able to repress. The repressing function of the various GvpD proteins was also investigated at the promoter level of the gvpA gene. This promoter is only activated during the stationary phase, depending on the transcriptional activator protein GvpE. Whereas the Vac ++ D transformants contained very high amounts of gvpA mRNA predominantly in the stationary growth phase, the amount of this transcript was significantly reduced in the Vac - transformants D/D and D/D P41A . In contrast, the Vac ++ D/D mut transformants harbouring GvpD mut with mutations at conserved positions in the p-loop motif contained large amounts of gvpA mRNA already during exponential growth, suggesting that this motif is important for the GvpD repressor function during this growth phase. The GvpD mutants containing mutations in the two basic regions were mostly defective in the repressing function. The GvpD mut protein containing an exchange of the three arginine residues 494RRR496 to alanine residues was able to repress gas vesicle formation. No gvpA mRNA was detectable in this transformant, demonstrating that this GvpD protein was acting as a strong repressor. All these results imply that the GvpD protein is able to prevent the GvpE-mediated gvpA promoter activation, and that the p-loop motif as well as the two basic region