Ultrasound imaging maps structures through the scattering of very high-frequency acoustic waves, which enables the imaging of objects as small as 10-100 micrometres. Researchers have investigated whether injected microbubbles - lipid structures that can trap gas – could be used as ultrasound contrast agents that enhance the scattering in the specific region of interest. However, these microscale objects are too large to move from the blood stream into the body. They are also unstable.

"If you asked me to make gas-containing structures that were nanometres in size and stable, it would be very difficult," says Shapiro. The big problem, he continues, is the surface tension: as the particle size is decreased, the larger surface tension reduces the particles' stability.

Nature’s solution

"We needed a structure that has a certain density and elasticity, and is also a protein, so we started looking in nature for something that would fit the bill," explains Shapiro. "As soon as we found that microorganisms make these gas vesicles we thought we had to try them."

The gas vesicles are genetically encoded gas-containing structures with dimensions of tens to hundreds of nanometres. They are produced by micro-organisms to control buoyancy, which in turn helps to improve their access to light and nutrients. Gas vesicles have been known about for over a century, but have not yet been exploited in nanotechnology.

Shapiro describes how these natural structures have avoided the problem of high surface tension in small particles. By excluding water but allowing gases in and out, there is no pressure difference between the inside and outside of the vesicle. "They are a fascinating bit of machinery," he adds.

Genetically encoding properties

What is more, studies of gas vesicles produced by two different micro-organisms reveal how differences in genetic coding result in different properties. Shapiro's team, in collaboration with colleagues at the University of California at Berkeley and the University of Toronto, found that vesicles produced by Halobacterium NRC-1 (Halo) demonstrate excellent back scatter from higher harmonics of the incident acoustic wave, which is a useful property for contrast specificity. On the other hand, gas vesicles produced by Anabaena flos-aquae (Ana) do not produce higher harmonic signals, but are much more stable under increased pressures.

The greater resistance to collapse made the Ana gas vesicles easier to handle, allowing the researchers to functionalize them with the protein biotin. Another protein streptavidin has a very strong affinity to biotin, so the presence of streptavidin causes the vesicles to form clusters, which in turn enhances the ultrasound signal. Shapiro believes that this clustering response could be used in methods analogous to the use of aggregation-dependent nanoparticle reporters that are widely used in magnetic resonance imaging. The species-dependent differences in the gas vesicles also offer the ability to control their properties.

Producing gas vesicles in the body

Further work is needed to optimize the signal strength and functionalization to target specific cells such as those that form tumours. In addition, while no acute toxicity was noted during in vivo demonstrations on mice over 48 hours, tests are needed to see whether this remains the case for other organisms. It also remains to be seen whether immune responses develop over longer periods.

Shapiro points out that using the gas vesicles to improve ultrasound imaging is just the starting point. The longer term aim will be to make cells in the body produce gas vesicles with tailored properties. "From a nanotechnology point of view, we have something genetically encoded, so we can change the size and shape and chemical make-up, and they are easy to make," he explains. "These are really beautiful structures nature has created."

The research is published in Nature Nanotechnology