Nature Biotechnology pp 537 - 542 and pp 528 - 529

Scientists have long been using "display" techniques to help them find enzymes, antibodies, and proteins of potential interest for industrial and medical applications. Now a new technique, which exploits the ability of bacteria to express proteins in the space between their cell wall and membrane, should both expand the range and accelerate the discovery of novel proteins as well as allowing screening for drug molecules that bind to them.

In standard protein display technologies, proteins are displayed on the surface of viruses or cells. The gene encoding a protein of interest is fused to a gene in a virus or cell in such a way that the fusion protein is transported to the surface of the particle. Millions of random variants of the protein of interest, each displayed on its own particle, are then screened for their ability to bind a desired target. However, many proteins can't be displayed in this way (e.g., because fusion to the virus coat may disrupt protein activity), limiting the diversity of the library and reducing the chances of finding a tight-binding protein.

In this issue, George Georgiou and his colleagues have developed an approach that avoids the need to "display" proteins on the surface of viruses or cells. They have expressed a collection of proteins in the bacterial "periplasm," the space between the bacterium's outer and inner membranes—which provides good conditions for natural protein folding and activity and avoids the need for fusion to the displaying system (e.g., a virus or cell).

To find molecules that bound to the expressed proteins, the researchers manipulated the experimental conditions to make the bacterial wall permeable to binding molecules labeled with a fluorescent chemical. After bacteria had been washed, any bugs containing bound fluorescently labeled molecules could then be detected by an automated fluorescent cell sorter. Using this method, the researchers were able to evolve an antibody that bound its target twice as tightly as the starting antibody. The method should work equally well for finding enzymes with improved functionality in industrial applications.

Nature Biotechnology pp 548 - 552

A spud containing a double-action vaccine against debilitating bacterial disease could be on the menu if a new idea takes root. Researchers have produced transgenic potatoes that, when fed to mice, simultaneously immunizes them against two deadly stomach bugs, rotavirus and enterotoxigenic E. coli.

Edible vaccines are nothing new, but their use thus far has met with limited success. A big part of the problem has been figuring out how to deliver the vaccine into the gut tissues before digestive juices start doing damage. Jie Yu and Bill Langridge of LLU have cleared that hurdle by designing a neat package of three vaccines that harnesses the otherwise deadly ability of cholera toxin (CT) to target the gut lining without being degraded in the gut. They hooked snippets of CT to bits from two other virulent pathogens: rotavirus and enterotoxigenic Escherichia coli. Then they grew potato plants that expressed the combo vaccine in their tubers. Female mice that ate these potatoes made antibodies against all the nasty stomach bugs and continued to show a measurable immune response for as long as two months after inoculation. In addition, at least some immunity was passed on to the pups of the inoculated mice. These results show promise for the creation of effective food-based vaccines to protect against either single or multiple pathogens.

Nature Biotechnology pp 543 - 547 and pp 527 - 528

Although elegant in concept, vaccinating animals using naked DNA to encode antigen has proven rather disappointing in practice. In general, it is difficult to get sufficient DNA into the right cells—:so-called antigen-presenting cells (APCs)—to trigger an immune reaction. Now, Harriet Robinson and her colleagues report that they can trick APCs into taking up naked DNA—by turning it into "lunch".

APCs are the housekeepers of the immune system, clearing up debris from sickly and dying cells. The APCs then process antigens from the debris, presenting them to white blood cells that generate a protective immune response. In the present study, the researchers manipulated the APCs penchant for "dead meat" by triggering an early death in local cells, linking the production of the "naked" DNA with that for a DNA encoding an enzyme (caspase) that promotes cell death when activated. Because caspases swiftly lead to the cell's death, the authors engineered a mutant caspase that slows down the process to provide time for the vaccine antigen to be synthesized. Sure enough, when the caspase was activated, naked DNA was produced before the cell dies, APCs flocked to the death scene and ingested antigen, triggering an immune response.

This approach, albeit indirect, evoked a stronger immune response against an influenza protein than had been achieved by putting naked DNA directly into the APC. Furthermore, the technique might be safer as the DNA-carrying cells self destruct.