Search

Serratia sp. ATCC 39006 (S39006) is a motile Gram-negative phytopathogenic bacterium known to produce two antibiotics – the β-lactam, 1-carbapen-2-em-3-carboxylic acid (a carbapenem) and the bioactive red pigment prodigiosin (a prodiginine). It is also the only known enterobacterium which naturally produces intracellular gas vesicles (GVs), enabling cells to float in static water columns. These hollow proteinaceous organelles are expressed from a 16.6 kb gene cluster comprised of 19 ORFs arranged in two contiguous operons separated by a 930-bp intergenic region. The left-hand gvpA1-gvpY operon mostly encodes GV structural components while the right-hand gvrA-gvrC operon is composed largely of regulatory genes. Other genetic regulators of GV biogenesis have been previously described which include the post-transcriptional regulator RsmA, the ribose operon repressor RbsR, the DeoR-family transcriptional regulator FloR, the sigma factor σ 54RpoN, and the transcription factor DksA. The regulation of GVs and secondary metabolites in S39006 can be coordinated but such pleiotropy is still poorly understood. PigP (an XRE family transcriptional regulator) and Rap (regulator of antibiotic production, a SlyA transcriptional regulator) are well-defined pleiotropic regulators of secondary metabolite production in S39006. However, their roles in the regulation of flotation have not been comprehensively investigated. In this study, quantitative proteomic analysis was used to compare the intracellular proteomes of pigP and rap mutants against the wild type. The abundances of proteins involved in antibiotic production, motility and GV biogenesis were consistent with the phenotypes of each mutant while known regulators also showed altered expression. Random transposon mutagenesis also identified two novel regulatory genes, draG (encoding an ADP-ribosylglycohydrolase) and xynR (encoding an IclR family transcriptional regulator), exhibiting diminished production of the carbapenem, prodigiosin, GVs and cellulase in S39006. Both mutants showed increased flagellar motility but attenuated virulence in planta and against C. elegans. Furthermore, by- pass mutagenesis of the draG mutant uncovered a putative inter-operonic sRNA located in the GV cluster (GV-sRNA) that repressed flotation of S39006. Finally, comparative quantitative proteomics was also undertaken on the draG, draG,GV-sRNA and xynR transposon mutants to investigate the wider impacts of each mutation and better understand their regulatory roles., Department of Science and Technology - Science Education Institute (DOST-SEI Philippines)

Serratia sp. ATCC 39006 (S39006) is a Gram-negative, rod-shaped enterobacterium known for the production two antibiotics; the ��-lactam, 1-carbapen-2-em-3-carboxylic acid (a carbapenem) and the red-pigmented tripyrrole, 2-methyl-3-pentyl-6-methoxyprodigiosin (prodigiosin; a prodiginine). It also produces plant cell wall degrading enzymes (PCWDEs) and is capable of flagellar-mediated swimming and swarming motility. S39006 is the only enterobacterium known to naturally produce gas vesicles (GVs). GVs are proteinaceous, intracellular organelles that increase the buoyancy of a cell and enable flotation upwards through the water column and colonisation of air-liquid interfaces. The production and regulation of GVs is a complex process with a range of regulatory inputs. GVs are expressed from a cluster of 19 genes arranged in two contiguous operons, gvpA1-gvpY and gvrA-gvrC. Prior to this study, three regulators encoded within the GV cluster, which are essential for GV production, had been described: GvrA, GvrB, and GvrC. Other regulators that have been identified include the post-transcriptional regulator RsmA, the repressor of the ribose operon, RbsR, and the DeoR-family transcriptional regulator, FloR. This study employed random transposon mutagenesis, visual screening of mutant phenotypes, cloning and sequencing, and bioinformatic analysis to identify novel regulators of GV production. This included mutants with transposon insertions in genes encoding an O-antigen ligase (waaL), the sigma factor ��54 (rpoN), and a transcription factor (dksA). The waaL mutant exhibited increased transcription and expression of GV genes but was unable to float. Pleiotropic effects of the transposon insertion included an increase in carbapenem production and a decrease in motility and virulence in a Caenorhabditis elegans model. The rpoN mutant showed a reduction in GV and pectate lyase production, swimming and swarming motility, and an increase in carbapenem production. The dksA mutant showed a decrease in GV and antibiotic production, motility, and virulence in C. elegans. A quantitative proteomic analysis was undertaken comparing the rpoN and dksA mutants against wild type S39006 to understand their regulatory roles. Expression of proteins involved in GV biogenesis, antibiotic production and motility as well as previously identified transcriptional regulators were significantly altered in each mutant, compared to wild type., Woolf Fisher Trust Cambridge Trust

Gas vesicles (GVs) are cylindrical or spindle-shaped protein nanostructures filled with air and used for flotation by various cyanobacteria, heterotrophic bacteria, and Archaea. Recently, GVs have gained interest in biotechnology applications due to their ability to serve as imaging agents and actuators for ultrasound, magnetic resonance and several optical techniques. The diameter of GVs is a crucial parameter contributing to their mechanical stability, buoyancy function and evolution in host cells, as well as their properties in imaging applications. Despite its importance, reported diameters for the same types of GV differ depending on the method used for its assessment. Here, we provide an explanation for these discrepancies and utilize electron microscopy (EM) techniques to accurately estimate the diameter of the most commonly studied types of GVs. We show that during air drying on the EM grid, GVs flatten, leading to a ~1.5-fold increase in their apparent diameter. We demonstrate that GVs’ diameter can be accurately determined by direct measurements from cryo-EM samples or alternatively indirectly derived from widths of flat collapsed and negatively stained GVs. Our findings help explain the inconsistency in previously reported data and provide accurate methods to measure GV dimensions.

The globally important marine diazotrophic cyanobacterium Trichodesmium is abundant in the New Caledonian lagoon (southwestern Pacific Ocean) during austral spring/summer. We investigated the cellular processes mediating Trichodesmium mortality from large surface accumulations (blooms) in the lagoon. Trichodesmium cells (and associated microbiota) were collected at the time of surface accumulation, enclosed under simulated ambient conditions, and sampled over time to elucidate the stressors and subcellular underpinning of rapid biomass demise (> 90 % biomass crashed within ∼ 24 h). Metatranscriptomic profiling of Trichodesmium biomass, 0, 8 and 22 h after incubations of surface accumulations, demonstrated upregulated expression of genes required to increase phosphorus (P) and iron (Fe) availability and transport, while genes responsible for nutrient storage were downregulated. Total viral abundance oscillated throughout the experiment and showed no significant relationship with the development or demise of the Trichodesmium biomass. Enhanced caspase-specific activity and upregulated expression of a suite of metacaspase genes, as the Trichodesmium biomass crashed, implied autocatalytic programmed cell death (PCD) as the mechanistic cause. Concurrently, genes associated with buoyancy and gas vesicle production were strongly downregulated concomitant with increased production and high concentrations of transparent exopolymeric particles (TEP). The rapid, PCD-mediated, decline of the Trichodesmium biomass, as we observed from our incubations, parallels mortality rates reported from Trichodesmium blooms in situ. Our results suggest that, whatever the ultimate factor, PCD-mediated death in Trichodesmium can rapidly terminate blooms, facilitate aggregation, and expedite vertical flux to depth.

In this study, we confirmed the effects of light and K+ concentration on the buoyancy of the cyanobacterium Microcystis aeruginosa (Kutzing) Lemmermann (NIES-843), and the relationship between the gas vesicle and buoyancy of M. aeruginosa was revealed through the culture experiment. The results showed that under laboratory conditions, light illumination and K+ concentration strongly affected the flotation and settlement of M. aeruginosa, and that the floating and settling cycle of M. aeruginosa could be reproduced in a test tube by controlling light illuminance. The ability of M. aeruginosa buoyancy weakened with the increase in the K+ concentration, and M. aeruginosa could not exhibit buoyancy at the K+ concentration of 0.03 mol/L. The phase-contrast microscope observation revealed that M. aeruginosa that floated on the water surface had gas vesicles, while none of the gas vesicles was detected for M. aeruginosa at the bottom of a test tube.

In order to effectively remove the Microcystis from drinking-water sources, pre-pressure treatment was first used to make the Microcystis lose buoyancy, and then it is easily removed by coagulation and sedimentation processes. The Microcystis-containing water from Taihu Lake was taken for the pre-pressure coagulation and sedimentation treatments in this study. Both intermittent laboratory experiment and continuous-flow field experiment were conducted. Experimental results showed that the optimum pre-pressure condition was pressuring at 0.6-0.8 MPa for at least 10 s, and 60 s was the best. Comparing with the pre-oxidation, pre-pressure could obviously increase the removal efficiency of Microcystis by following coagulation and sedimentation, and would not increase the dissolved microcystins. The mechanism of pre-pressure treatment was that the pre-pressure destroys the gas vesicles in Microcystis cells and the gas diffuses out of the cells, which leads the Microcystis to lose buoyancy and make them to sink. The recovery time of gas vesicles was longer than the sludge discharge period of sedimentation tank; therefore, the sinking Microcystis would not re-float in the sedimentation tank. In the practical application of drinking water treatment plant, the continuous-flow pressure device could be chosen, with the energy consumption of about 22.9 kw·h per 10,000 m

Previous investigations suggested that buoyancy state in response to light differed between individuals within natural populations of cyanobacteria. To understand the mechanisms of heterogeneity of buoyancy in different species/strains in relation to light, two types of colonial Microcystis in different buoyancy behavior were selected and used to compare their photosynthetic activity, gas vesicle volume, ballast mass, and migration at varying light regime. The photosynthesis-irradiance curve and F (v)/F (m) examination indicated that negatively buoyant strains were more adapted at high irradiance than buoyant ones. Transcription levels of gvp gene and gas vesicle volume decreased in buoyant strains, but increased in negatively buoyant ones at high irradiance. The results indicated that the combination effect of decrease in gas vesicle buoyancy and increase in carbohydrate contributed to the downward migration of buoyant strains, while the significant increase of gas vesicles provided sufficient buoyancy to negate ballast mass, resulting in the upward migration of negatively buoyant ones at high irradiance. In addition, either sinking or floating velocities were elevated in buoyant and negatively buoyant strains, coincidently with the colony enlargement of all strains at high irradiance, respectively. These findings suggest that this heterogeneity was associated with the interplay between gas vesicles, ballast, and colony size. The fact that different species/strains of Microcystis respond diversely to light depending on their physiological conditions presents a good example to understand heterogeneity of buoyancy in the field, and the presence of heterogeneity of buoyancy may be implicated in the dominancy and persistence of Microcystis bloom in ever-changing environment.

Species of Oscillatoria with abundant gas vesicles have been reclassified in the genus Planktothrix and the names of species distinguished partly by differences in pigmentation and filament width have been retained in the new combinations. We determined the variation in filament width (d f ) and gas vesicle critical pressure (P c ) in three collections of Planktothrix spp.: 56 green coloured isolates (P. agardhii) and 32 red isolates (P. rubescens) from Blelham Tarn, a shallow lake in the English Lake District; 19 isolates of P. rubescens from Lake Zurich, a deep lake in Switzerland; and 25 green and 15 red isolates from 21 Nordic lakes in the NIVA collection. The differences in filament width, which occur in Planktothrix populations from different lakes, are retained in culture. The difficulty in using d f to distinguish different taxa is that it forms a normally distributed continuum. The mean critical pressures of gas vesicles form a discontinuous distribution (with peaks at about 0.7, 0.9 and 1.1 MPa) correlated with three GV genotype groups (GU5/6, GV2/4 and GV3, respectively). The p c is not a useful taxonomic character: the full range of p c is encountered in strains of P. rubescens that do not differ in other respects. Though quantitative data is lacking, it has been proposed that p c is correlated with lake depth: the 1.1 MPa gas vesicles have been found only in Planktothrix from the deep Lake Zurich; only the 0.7 MPa gas vesicles occur in Planktothrix from the shallow Blelham Tarn. The presence (red isolates) or absence (green isolates) of phycoerythrin was stable in cultures maintained over many generations.

Gas vesicles are hollow, proteinaceous structures found in some strains of cyanobacteria. They have been used to increase the oxygen supply and improve the cultivation of shear-sensitive mammalian cells. However, the production and, especially, collection of cells and gas vesicles were laborious and ineffective. In this study we examined the use of the cationic polymer, polyethyleneimine (PEI), for improving the cell harvesting by flocculation and flotation. PEI was examined to determine the appropriate molecular weight, pH range, and dosage. The dose of 20–30 mg/l of PEI with molecular weight of 25,000 in the pH range of 6.0–8.5 was found to provide effective and efficient cell flocculation and flotation. As the PEI dose increased, the rate of flotation increased but the clearance (collection) efficacy declined slightly. The culture samples used in this study were taken from light-limiting continuous culture systems at different dilution rates (0.05–0.24 h−1). Without PEI addition, the cells at dilution rates lower than 0.1 h did not float while those at higher dilution rates would float slowly. With PEI addition, the flocculated cells at the dilution rate of 0.05 h−1 sank and those of higher dilution rates would float and the flotation rate increased with increasing specific growth rate. Nonetheless, PEI flocculation and flotation (or sedimentation) could be used to harvest cells at a wide range of growth states.

After freeze-fracturing/etching of specimens and unidirectional or rotational metal-shadowing of their fracture plane replicas, electron micrographs of gas vesicles of 11 strains (8 planktic species) of Cyanobacteria were taken. Micrographs were subsequently digitized and the periodicity of gas vesicle membrane striation was analyzed using FOURIER transform. In parallel with it, number of transversely fractured gas vesicles on 0.1 ľm2 of cross-fracture plane was counted. Different striation (rib) periodicity of gas vesicle membranes of different species (and of different strains of the same species) was established, ranging from 3.9 to 4.4 nm. Striation periodicity of gas vesicle membrane (though limited in range) appears independent of the number of gas vesicles per 0.1 ľm2 of cross-fracture plane and both parameters can be used for a closer characterization of gas vesicles. These results imply that the structure of gas vesicle membrane is neither species nor genus or family specific and that different cyanobacterial cells operate with different alleles of the genes for the gas-vesicle proteins.

It has been known for more than 40 years that vacuolate organisms can be induced to sediment with ultrasound. However, robust indicators are still needed to compare the efficacy of different treatments. A repeatable index is proposed that makes it possible to quantify the ultrasonic induced sedimentation. The procedure is used to monitor the long term sedimentation of Microcystis aeruginosa after sonication. Results reveal that the sedimentation process continues after gas vesicles have fully recovered, although at a slower rate.

We have previously shown that the gas-vesicle protein GvpC is present on the outer surface of the gas vesicle, can be reversibly removed and rebound to the surface, and increases the critical collapse pressure of the gas vesicle. The GvpC molecule, which contains five partially conserved repeats of 33 amino acids (33-RR) sandwiched between 18 N-terminal and 10 C-terminal amino acids, is present in a ratio of 1:25 with the GvpA molecule, which forms the ribs of the gas vesicle. By using recombinant techniques we have now made modified versions of GvpC that contain only the first two, three or four of the 33-amino-acid repeats. All of these proteins bind to and strengthen gas vesicles that have been stripped of their native GvpC. Recombinant proteins containing three or four repeats bind in amounts that give the same ratio of 33-RR:GvpA (i.e. 1:5) as the native protein, and they restore much of the strength of the gas vesicle; the protein containing only two repeats binds at a lower ratio (1:7.7), however, and restores less of the strength. Ancestral proteins with only two, three or four of the 33-amino-acid repeats would have been functional in strengthening the gas vesicle but the progressive increase in number of repeats would have provided strength with increased efficiency.

Summary Before 1988 only four sequences of halobacterial protein coding genes were known, but in recent years the number has increased dramatically. Today 130 protein coding genes have been sequenced, belonging to a limited set of protein families. Here a compilation of these genes is given. Most of the genes have the same codon preference. A halobacterial codon usage table is presented and its characteristic features are discussed. The genes of the gas vesicle gene clusters exhibit a different codon usage, which is also shown. The use of the halobacterial codon usage table for the identification of protein coding genes as well as its application as a guide in sequencing halobacterial DNA and construct oligonucleotides are exemplified.

The ArsA ATPase ofE. coliplays an essential role in arsenic detoxification. Published evidence implicates ArsA in the energization of As(III) efflux via the formation of an oxyanion-translocating complex with ArsB. In addition, eukaryotic ArsA homologues have several recognized functions unrelated to arsenic resistance. By aligning ArsA homologues, constructing phylogenetic trees, examining ArsA encoding operons, and estimating the probable coevolution of these homologues with putative transporters and auxiliary proteins unrelated to ArsB, we provide evidence for new functions for ArsA homologues. They may play roles in carbon starvation, gas vesicle biogenesis, and arsenic resistance. The results lead to the proposal that ArsA homologues energize four distinct and nonhomologous transporters, ArsB, ArsP, CstA, and Acr3.

Gas vesicles provide buoyancy in many aquatic cyanobacteria and planktonic bacteria. An apparatus comprising three capillary compression tubes is described for the determination of the gas vesicle volume in a cell suspension. It provides replicate measurements of the irreversible volumetric contraction when gas vesicles are collapsed by pressure. Potential errors in the measurement associated with thermal instability, calibration of the tubes and capillaries, the formation of gas bubbles and the effects of surface tension are discussed. Highly reproducible measurements were obtained with suspensions containing as little as 0.5 μl ml−1 of gas vesicle gas. The use of the apparatus is described in quantitative studies of the critical collapse pressures of gas vesicles, changes in gas vesicle content of cell suspensions, the proportion of cell volume occupied by gas space, the proportion of the gas vesicles required for neutral buoyancy, and the buoyant density of floating cells. These measurements are required in quantitative studies on buoyancy regulation and waterbloom formation by cyanobacteria (blue-green algae).

Antibodies were raised against intact gas vesicles of Anabaena flos-aquae, and against a synthetic peptide (GVPaNT) whose sequence is identical to the N-terminal region of the main gas vesicle protein, GVPa. A two-stage centrifugation procedure is described for separating gold-labelled antibodies bound to gas vesicles from unbound antibodies. The GVPaNT antibody bound to gas vesicles that had been previously rinsed with SDS to remove the outer gas vesicle protein, GVPc. Treatment with this antibody caused the gas vesicles to aggregate together end-to-end rather than side-by-side. The binding of the anti-GVPaNT-immunogold particles to the gas vesicle was restricted to the conical ends of the structure. These observations indicate that the sequence to which the GVPaNT antibodies were raised, residues 1 to 13 of the GVPa molecule, is exposed only at the outer surface of the cones and that it is normally obscured by GVPc. As GVPa forms both the conical ends and the cylindrical midsection of the gas vesicle, exposure of the N-terminal sequence only in the cones must be due to differences in the contact between adjacent GVPa molecules in the central cylinders and end-cones.

Gas vesicles are gas-filled prokaryotic organelles that provide buoyancy in many planktonic (cyano)bacteriaand halophilic archaea. Remarkably, more and more genomes of soil bacteria, especially those of actinomycetes,show gas vesicle gene (gvp) clusters often encoding homologues of atleast the eight gvp genes essential for gas vesicle formation in archaea.Here, I discuss characteristics specific to actinomycete Gvp proteins, their expression under stress conditions,and the apparent absence of a buoyancy phenotype in gvp mutant strains.Alternative functions for gas vesicles in actinomycetes are discussed in relation to their unusually complexdevelopmental life cycle.

Wild-type strains of the cyanobacterium Planktothrix rubescens have a cluster of gas vesicle ( gvp ) genes with repeats of alternating gvpA and gvpC . The gvpC occurs in three length variants, all with the same 3′-sequence, Ω C . Spontaneous non-buoyant mutants had lost some of the alternating gvpAC copies and their gvpC genes had a novel 3′-end sequence, Ψ C ; additional gvpC genes terminating in this sequence were also found in the wild-type and representatives of other GV genotypes. Alleles of gvpC terminating in Ψ C occurred only at the downstream ends of the gvpAC clusters investigated; all other gvpC s terminated in Ω C . Mutants of strains with the GV3 genotype produced only 30–50% of the gas vesicles present in the wild-type; their gas vesicles had lower mean critical pressures (0.70–0.78 MPa) than those in the wild-type (1.05–1.10 MPa).

The mean critical collapse pressure (p(c)) of gas vesicles in 81 strains of the cyanobacterium Planktothrix rubescens from Lake Zurich, Switzerland, was bimodally distributed between a minimum of 0.86 MPa and a maximum of 1.17 MPa. Measurements were made of the cylinder diameter (d) of gas vesicles isolated from seven of the strains. The mean diameter, which varied from 48 to 61 nm, was inversely related to p(c), in keeping with the theory of strength of thin-walled rigid cylinders. These measurements extended the range of p(c)-width relationship of gas vesicles, which can be described by the expression p(c) = 461(d/nm)(-1.53) MPa. p(c) was correlated with gas vesicle genotype (see the accompanying paper by S. J. Beard, B. A. Handley, P. K. Hayes & A. E. Walsby, Microbiology 145, 2757-2768): of the 81 strains investigated, all those with the gas vesicle genotype GV2 produced gas vesicles with a mean p(c) of less than 1.0 MPa, whereas those of GV3 had a mean p(c) of greater than 1.0 MPa. It is suggested that gas vesicles of the GV3 strains, which are narrower and stronger than any previously recorded in freshwater cyanobacteria, have evolved to withstand the high hydrostatic pressures during deep winter mixing in Lake Zurich.

Gas vesicles encoded by gvp genes provide buoyancy in many prokaryotes. In a recent Trends in Microbiology article entitled 'Gas vesicles in actinomycetes: old buoys in novel habitats?' van Keulen et al. documented the occurrence of gvp genes in soil-inhabiting actinomycetes but questioned whether any of them produce gas vesicles. We suggest that the protein encoded by gvpA in actinomycetes might be incompatible with the structure of the standard gas vesicle. Perhaps it has another role associated with the air-water interface.

SUMMARY: Previous studies have shown that gas vesicles isolated from the cyanobacterium Anabaena flos-aquae contain two types of protein, GvpA, a small hydrophobic protein that forms the main ribbed structure, and GvpC, a protein comprising five repeats of a 33-amino-acid-residue motif, which is located on the outer surface of the GvpA shell. GvpC was shown to increase the critical collapse pressure of the gas vesicles; it was thought to do this by forming a series of molecular ties that bind the ribs together. We now show that antibodies raised against GvpC label both the central cylinders and the conical end caps of native gas vesicles but fail to bind to gas vesicles that have been stripped of GvpC. The molar ratio of GvpA to GvpC has been calculated from amino acid analyses of gas vesicle hydrolysates by reference to the abundance of amino acids that occur predominantly or exclusively in one protein or the other; the molar ratio was found to be 25:1 in freshly isolated gas vesicles and 23:1 in gas vesicles saturated with GvpC. We have considered three ways in which the 33-residue repeats of GvpC might interact with the crystallographic unit cell of GvpA molecules in the ribs. The Anabaena GvpC will bind to and restore the strength of gas vesicles isolated from Aphanizomenon and Microcystis that lack their native GvpC.

The gas vesicles found in various planktonic prokaryotes are hollow, rigid structures permeable to gases. They collapse when the difference between the external hydrostatic pressure and internal gas pressure exceeds their critical pressure (usually about 0.6 MPa). It was found that dried gas vesicles would survive exposure to gas pressures considerably in excess of this value (4 MPa or more), because gas diffused into them as the pressure was raised and the pressure difference required to cause collapse was not established. They survived the most rapid rates of pressure rise, 0–4.6 MPa in less than 2.5 ms, to which they were exposed. From this it can be calculated that the gas permeability coefficient of the average gas vesicle ( α ) exceeds 22 x 10 3 s -1 and the permeability of the gas vesicle wall ( k ) is greater than 332 μm s -1 . Gas molecules may diffuse through fixed pores in the gas vesicle wall. Since a gas molecule of collision diameter 0.63 nm is known to penetrate the gas vesicle, this would be the minimum diameter of such a fixed pore. It is shown by kinetic theory that the permeability coefficient of an average gas vesicle with one pore of this size would be 2.1 x 10 3 s -1 : there would, therefore, have to be at least 11 such pores to account for the observed minimum permeability coefficient. Gas vesicles in aqueous suspension will also survive a rapid rise in the overlying gas pressure in excess of their critical pressure if they are near enough to the gas-water interface for sufficient gas to reach them by diffusion during the pressure rise. The distance from the interface at which the gas vesicles survive can be used to calculate the diffusivity of the gas through the suspension. A modification of this method can be used to measure the gas-permeability of certain types of cells containing gas vesicles.

Filaments without gas vacuoles arose spontaneously in the gas-vacuolate alga Anabaena flos-aquae. The non-vacuolate mutant was enriched by repeated sedimentation and subsequently cloned by microsyringe transfer. No revertants have been observed. In the gas-vacuolate wild-type alga the gas vesicle protein was clearly distinguished by gel electrophoresis as one of the ten most abundant protein species present in whole cell extracts. Electrophoresis indicated that the mutant had lost the ability to synthesize the gas vesicle protein. A second mutant partially defective in production of gas vacuoles and gas vesicle protein has been isolated.

Gas vesicle protein may account for more than 6% of the cell dry weight and over 10% of the total cell protein in the planktonic cyanobacterium Anabaena Jlos-aquae. The gas vesicles collapsed by rising cell turgor pressure during buoyancy regulation can therefore represent a significant protein pool within this organism. We have looked for evidence of gas vesicle protein recycling. Gas vesicles were isolated from cells of A.80s-aquae that had been pressurized and then cultured in a radioactively labelled medium. The new gas vesicles that formed were less highly labelled than the new vesicles formedin cells from a second culture that had not been pressurized. The results suggest that the collapsed gas vesicles are disaggregated and that the constituent gas vesicle protein is reassembled into new gas vesicles. Proof of this will require experiments using in vitro assembly systems.

The contribution of K+ accumulation to cell turgor pressure was investigated in the gas-vacuolate blue-green alga Anabaena flos-aquae. The cell turgor pressure, measured by the gas vesicle method, drops in cells suspended in culture medium depleted of K+ but rapidly rises again, by 100 kPa or more, when K+ is resupplied. A similar though rather slower rise in turgor pressure is supported by an equivalent concentration of Rb+. The internal K+ concentration rose from 66 to 91 mM when K+ was supplied at an external concentration of 0-4 mM. This rise was light-dependent. Greater increases in internal K+ concentration and turgor pressure occurred when K+ was supplied at a higher concentration, 3-6 mM. In both cases over 60% of the observed turgor pressure rise could be accounted for by accumulation of K+. The turgor pressure rise supported by light-stimulated K+ uptake can cause collapse of enough of the alga's gas vesicles to destroy its buoyancy. The effect of K+ availability on buoyancy regulation by planktonic blue-green algae is discussed.

One of the most extraordinary structures found in bacteria is the gas vesicle. It has the form of a shell of protein molecules enclosing a gas-filled space. The properties of this structure ― its rigidity, its contrasting inner and outer surfaces, its permeability to gases and impermeability to liquid water ― are all provided by its constituent proteins