The molecular microbiology and physics of bacterial flotation

Some aquatic bacteria can make intracellular chambers (made entirely of protein) that are permeable only to environmental gasses. The structures are called gas vesicles (GVs) and they form conglomerates (gas vacuoles) identifiable by phase contrast microscopy. The aquatic bacteria that make GVs may use them for the phenomenon of buoyancy, allowing upward flotation in a static water column. This ability can be useful for some photosynthetic bacteria (e.g. cyanobacteria) that need to rise in a stratified aquatic niche to access light of a specific wavelength, or to acquire nutrients or oxygen at the air-liquid interface, or perhaps to escape predators or competitors. The GVs usually comprise a major protein (GvpA) and a minor protein (GvpC) and form cylindrical structures with apical poles.

We recently discovered gas vesicles in strain ATCC39006 of the enterobacterium, Serratia ("related" to E. coli). The existence of GVs in this strain was a unique, and totally unexpected, observation. In addition to production of GV organelles and the capacity to float, this strain has other interesting traits. It makes two antibiotics. One antibiotic is antibacterial (a carbapenem) and another (prodigiosin) can kill protozoans and other microbes. ATCC39006 can swim via flagella (motility) and can swarm on solid surfaces and make surface detergent molecules (biosurfactants) enabling spreading and colonisation of new niches or host surfaces. The strain also rots plants (potato) by secreting plant cell wall degrading enzymes and it kills microscopic worms (Caenorhabditis elegans) and so it is also nematicidal.

We identified the cluster of 19 GV genes in strain ATCC39006 and we engineered E. coli strains that expressed the Serratia GV genes and allowed E. coli to float up to the air-liquid interface in static culture. Production of the GVs in Serratia was bacterial cell density-dependent in a process called "Quorum Sensing" controlled by a diffusible chemical signal that moves between cells. The signalling molecule is essential for production of the GVs in Serratia; quorum-sensing mutants don't float. Therefore, in this bacterium, an intercellular chemical signal controls the development of intracellular organelles and thus the chemical communication signal is also a morphogen. Quorum sensing also controls the production of the antibiotics in this strain and so these toxic molecules are made at the same time as the GVs are assembled. We showed that GV production was also up-regulated by oxygen limitation, implying that GVs may allow flotation to the liquid surface to acquire oxygen.

In collaboration with Professor Raymond Goldstein in the Department of Applied Mathematics and Theoretical Physics (DAMTP) in Cambridge, we have been investigating the phenomenon of buoyancy in this bacterium. Based on our knowledge of the genetics, physiology and physics of mobility in this bacterium we have developed a testable working hypothesis as to why GV production, and flotation, is under quorum sensing control; responsive to oxygen levels, and developmentally preferred to flagellar motility. Our model also predicts why the production of the two antibiotics has evolved to be co-incident with the development of the GVs, and flotation. We will now investigate the fascinating connections between bacterial cell population density, motility, bioconvection, GV development, buoyancy and antibiotic production. This study has wide ramifications. It impinges on areas such as ecological adaptation to environmental stress cues in microbes; intercellular chemical communication in bacteria; bacterial organelle morphogenesis; and the fitness value of microbial conflict and niche defence. Our understanding of the evolution of some of these biological processes will be significantly enhanced by an appreciation of the underlying mathematical physics that describes their behaviour; an exciting interdisciplinary study where microbiology meets mathematics!

Serratia sp ATCC39006 (S39006) makes gas vesicles (GVs)- the only example of intracellular organelles existing naturally in a member of the Enterobacteriaceae. GVs form conglomerate light-refractive vacuoles under phase contrast microscopy. GVs are gas-permeable organelles providing buoyancy in bacteria. The ability to float upwards in a static water column can allow bacteria to gain access to nutrients, or access light, or air, or evade competitors. GVs were originally associated with marine and aquatic organisms such as cyanobacteria and halophilic Archea.

The GV locus has 19 genes, divisible into a left hand subcluster (L) encoding structural proteins and an R (right hand) subcluster that is regulatory. GV development is under quorum sensing (QS) control, via the signaling molecule (BHL) - a morphogen. Air-restricted cultures make more GVs and float better. The post-transcriptional regulator system, RsmAB also regulates GV morphogenesis. We will dissect the requirement for each of the 19 GV genes using non-polar mutants, assessing colonial impacts, GV architecture and protein composition, and flotation capacity. We will examine flotation functionality of apparently redundant structural genes encoding isoforms of GvpA. We will study the signal transduction pathways that sense QS (cell density) and oxygen levels and that use post-transcriptional control to link environmental stress to the developmental biology of GV biogenesis. We will test the hypothesis that the strain makes GVs as an adaptive response to avoid the consequences of flagellum-driven bioconvection processes seen during motility towards air-liquid interfaces. We will exploit various mutants of S39006 to describe mathematically the competing processes of motility and flotation and model the biophysical implications of GV production. Our aim is to compare the mathematical predictions with the experimental reality to test our view of the likely adaptive, evolutionary purpose of GV morphogenesis.

Who will benefit from this research?
Potential beneficiaries include biotechnology industries using culture of mammalian cells or potentially generating hybrid vaccines using gas vesicles as scaffolds with antigen presentation features. GVs could have utility for such applications. Purified GVs have potential uses in mammalian cell culture systems because they represent a "gentle" mode of gas delivery to labile cells which can be susceptible to aeration damage.

We considered the fact that GVs engineered with hybrid structural proteins, including viral peptide antigens, are potentially immunologically protective and this could be a technology for the development of vaccines. As our system involves oxygen and QS-mediated control, this could be useful for possible translation in bioprocessing for control of gene expression in industrial microbiology. QS systems and GVs have clear synthetic biology applications as molecular bio-bricks for regulation control systems (and for flotation). Indeed our work has already been highlighted on an iGEM web blog site with this in mind.

How will they benefit from this research?
The biotechnological potential of GV research and QS research needs industrial collaboration and translation with financial investment that would extend beyond this project. The PDRA to be appointed will benefit enormously from acquiring the molecular microbiology skills involved plus his/her exposure and involvement in the biological physics and mathematics of modeling bacterial behavior. He/she will also gain skills and experience in teaching students and in science communication.

What will be done to ensure that they have the opportunity to benefit from this research?
The research outcomes will be disseminated to scientists (universities, institutes and commercial organisations) via international publications, lectures and posters at international symposia. Open access journals ensure global accessibility of research knowledge generated from this study. We have a track record of BBSRC-CASE studentships (two recently completed with UK companies) and collaboration with UK research institutes. We will try to extend such associations to generate enhanced funding for our research programmes and bring in added value by synergy of research interests with collaborators. This university encourages commercial liaisons and spin-out, after the establishment of solid IP positions. Our track record shows that we actively consider filing patent applications e.g. on carbapenem antibiotics and cryptic gene activation systems e.g. Salmond et al (US Patent 5821077 - issued 1998); Salmond et al (WO/1995/032294). We filed a patent application on antiviral abortive infection systems in 2008.

When we discovered GV production in Serratia and engineered the GV locus genes in E. coli, we considered filing a patent on the strains and technology because GVs have utility e.g. see above. However, after taking advice about IP from Cambridge Enterprise, we dropped the idea because pre-existing patents on GVs were sufficiently broadly written to make protection for our GV system very difficult. Furthermore, a very recent (December 2011) solid state NMR analysis of GVs from a group at MIT now suggests that GVs may have amyloid properties - perhaps not an ideal protein system for immunization! So our initial desire for IP protection was well intentioned, but pragmatically restricted. Nevertheless, we remain open to the possibility of securing IP on any aspects of our technology.

Finally, the PI and the Co-I are also involved in other impacts in national governance roles (e.g. Governing Boards of Research Institutes, Advisory Councils and Learned Societies). They are also involved in outreach activities. The Co-I has presented Science Week activities in Cambridge and school liaison presentations and the PI gives talks to student societies and to the University of the 3rd Age.