Haloarchaea and the formation of gas vesicles

Abstract

Halophilic Archaea (Haloarchaea) thrive in salterns containing sodium chloride concentrations up to saturation. Many Haloarchaea possess genes encoding gas vesicles, but only a few species, such as Halobacterium salinarum and Haloferax mediterranei, produce these gas-filled, proteinaceous nanocompartments. Gas vesicles increase the buoyancy of cells and enable them to migrate vertically in the water body to regions with optimal conditions. Their synthesis depends on environmental factors, such as light, oxygen supply, temperature and salt concentration. Fourteen gas vesicle protein (gvp) genes are involved in their formation, and regulation of gvp gene expression occurs at the level of transcription, including the two regulatory proteins, GvpD and GvpE, but also at the level of translation. The gas vesicle wall is solely formed of proteins with the two major components, GvpA and GvpC, and seven additional accessory proteins are also involved. Except for GvpI and GvpH, all of these are required to form the gas permeable wall. The applications of gas vesicles include their use as an antigen presenter for viral or pathogen proteins, but also as a stable ultrasonic reporter for biomedical purposes.

Figure 1

Colonies (a) and cells of Halobacterium (Hbt.) salinarum producing gas vesicles (b,c). (a) Colonies on solid media grown for one week at 40 °C and three weeks at room temperature. Vesicle (Vac⁺) cells form pink white colonies, whereas colonies of Vac⁻ mutants are red and transparent. (b) Cells grown in liquid media observed by phase-contrast light microscopy. (c) Cells of a Vac⁺ colony investigated by transmission electron microscopy. The pleomorphic shape of the cells grown for three months on solid media differs from the rod-shaped cells seen in liquid culture.

Figure 2

Arrangement of gvp genes in p-vac of Hbt. salinarum PHH1 (a) and a comparison of the intergenic regions separating P_D and P_A (b). (a) The arrows depicting genes are colored as follows: dark green, encoding structural proteins of the A-J-M family and GvpC; light green, encoding accessory Gvp; red, encoding regulator proteins. Black arrows mark the start sites of transcription. (b) Comparison of the intergenic regions separating P_A and P_D in p-, mc- and c-vac. A 22-nt insertion occurs in c-vac adjacent to BRE_D. The 20-nt sequence required for GvpE-mediated activation is underlined (8-nt elements separated by 4 nt of unimportant sequences) in the case of UAS_A and marked by an arrow. Similar activation elements are found with UAS_D, pointing in the opposite direction. The TATA-box and BRE sequences (italics) are shaded in grey.

Figure 3

Activation of P_A-P_D by GvpE (a) and repression by GvpD (b). Schematic representation of the region between gvpA and gvpD and the two oppositely-oriented promoters P_A and P_D. TATA-box and BRE are shown in grey, and the two UAS elements are partly overlapping in the center in light grey. The reading frames gvpD and gvpA are represented by dark arrows. (a) Activation of transcription by GvpE should involve binding of GvpE, presumably as a dimer, at the respective UAS element and recruitment of TFB, TBP and of the RNA polymerase. (b) In the presence of GvpD, the interaction of GvpE-GvpD leads to a strong reduction in the amount of GvpE and the repression of gas vesicle formation.

Figure 4

Sequence alignment of GvpA proteins derived from Haloarchaea and methanogens. Identical amino acid residues are highlighted in the following colors: red, negatively charged; blue, positively charged; green, polar, uncharged; yellow, small and variable; grey, aromatic; white, non-polar residues. The bar on top marks the highly conserved central 51 amino acids. The difference in the C-terminus of GvpA (gvp1, NRC-1) to the sequence of pGvpA (p-vac, PHH1) is due to a missing G nucleotide close to the 3'-terminus of gvpA in gvp1. pGvpA and cGvpA are derived from Hbt. salinarum PHH1 and mcGvpA from Hfx. mediterranei.

Figure 5

Amino acid sequence of GvpA and the structure derived from the in silico modelling. (a) Amino acid sequence of pGvpA and the proposed 2D structure (model). H denotes α-helices and E β-strands. The α-helices, H1 and H2, and the anti-parallel β-strands are indicated on top. (b) Electron micrograph of a Vac⁻ ∆A + A_mut transformant, showing that gas-filled compartments are indeed lacking in such mutants. (c) Structure of GvpA, highlighting some positions of ala-substitutions leading to gas vesicle negative ∆A + A_mut transformants [63]. The respective single amino acids that are altered are shown in bold in (a).