Nature Biotechnology pp 295 - 299

Puddings, custards, and other foods containing starch often lose their consistency when frozen and thawed because, during the freezing and thawing process, starch paste undergoes separation into solid and liquid phases. The resulting deterioration in product quality is a problem for the food industry, which routinely freezes merchandise during storage and transport. However, in the March issue of Nature Biotechnology, researchers have used a genetic technology called "antisense" to alter the starch composition and structure in a potato, creating a starch that can withstand up to five freeze-thaw cycles.

Potato starch is made up of two glucan polymers: amylose, a long and essentially unbranched chain of glucose molecules; and amylopectin, a relatively short and more highly branched glucose chain. Building on knowledge gained from the manipulation of various genes involved in starch synthesis, Stephen Jobling and colleagues hypothesized that a potato with mostly short-chained amylopectin would reduce separation in potato starch during the freeze-thaw process.

The researchers modified the starch structure by removing the long chains of amylose and reducing the branch length of short-chain amylopectin, by simultaneously inhibiting three starch synthase enzymes: one responsible for synthesizing amylose, and two that affect branch chain length of amylopectin. This was achieved by introducing into potatoes DNA sequences encoding messenger RNA in opposite orientation (antisense) to natural RNA produced by the three corresponding starch synthase genes, thereby inhibiting formation of the three enzymes. The result was a potato strain containing amylopectin with shorter branches and very little amylose; after five freeze-thaw cycles, the starch from this strain remained clear—a sign that separation had not occurred.

Stabilizing starch this way is environmentally friendly and potentially more economical than chemical modification, the current alternative.

Nature Biotechnology pp 256 - 263 and pp 235 - 236

Scientists have coopted immune cells as agents for delivering anticancer gene therapy. Although the system is far from testing in humans, the approach could complement more traditional cancer treatments such as chemotherapy.

Although cancer gene therapy is a potentially powerful adjunct to conventional cancer therapies, its use in clinical settings has been severely limited due to the undesirable foreign gene expression in tissues other than the tumor. To create a gene therapy delivery system more specific to tumors, researchers led by Richard Vile have employed T cells, a class of immune cells that preferentially targets tumors as a means of delivering a "suicide" gene to cancer cells.

The researchers first set about tailoring T cells to be specific to tumors by incorporating into their DNA a gene expressing a receptor that binds to a tumor-specific cell-surface protein called carcinoembryonic antigen (CEA). To turn the T cells into potent cancer-killing agents, they were engineered to produce a virus carrying a toxic gene (thymidine kinase). This gene, which originates from herpes simplex virus and is not normally present in cells, encodes a protein that converts a relatively harmless prodrug, ganciclovir, into a cancer killing poison. To add an even greater level of control, they also placed the thymidine kinase gene under the control of a regulatory element (the CEA promoter) that is active only in tumor cells.

The authors demonstrate the potency of their approach both in tissue culture as well as in mice with lung and liver metastases. Further studies—for example, to make sure that enough virus is produced by the T cells during their lifetime in the body—will be required before this system can be tested in humans.

Nature Biotechnology pp 275 - 281 and pp 234 - 235

The presence of "alien" sugars on the outer surface of various pathogens warns the human body of infection, raising alarm bells for the immune system and prompting the subsequent generation of antibodies. This presents a problem for pathogen diagnostics because carbohydrates are highly complex structures and no fast and robust methods are currently available for assessing sugars. All this might change, however, if a prototype carbohydrate microarray reported in this issue could be optimized and refined.

Denong Wang and colleagues spotted 20,000 microbial glycoconjugates—:molecules containing sugars with proteins, lipids, or other sugars—on glass slides, checking that they remained functional in this state. They then applied a small sample (one microliter) of human serum to the microarrays and tested for the presence of sugar-specific antibodies against, for example, Escherichia coli and Pneumococcus pathogens.

As well as being used as fast, sensitive detectors of infectious agents—both the commonplace and the more rare and sinister bioweapons—carbohydrate "chips" could provide valuable tools for investigating the function of sugars in healthy and diseased cells.