Protein network keeps stems cells at their peak

Harvard researchers have mapped a network of interacting proteins in mouse embryonic stem cells that seems to keep the cells in their early, unspecialized state, where they maintain their capacity to form any tissue in the body.

Revealing the regulatory circuit in stem cells could help scientists reach their goal of reprogramming already specialized adult cells to an embryonic state. This could provide human stem cells for research and medical uses, without the ethical controversies surrounding the current methods of creating embryonic stem cell lines.

Other researchers had already identified genes involved in maintaining stem cells. But to understand the complete regulatory process, they need to find out how the products of those genes work together in the cell.

The researchers, in the lab of Stuart Orkin at Harvard Medical School, constructed the map starting with the critical stem cell regulator protein, Nanog. To find other proteins in mouse embryonic stem cells that were in direct contact with Nanog, they used a specially tagged version of the protein as “bait” to fish for interacting proteins. Then they repeated the experiment using several of the newly discovered partner proteins as bait to expand the map of the network.

In the end, they found an interconnected web of 37 proteins, many of which were already known as players in early development. Much of the network seems to function by squelching the expression of genes that trigger stem cells to turn into more specialized cells.

Researchers have known that stem cells that have begun to specialize produce many of the proteins in the network at lower levels. This suggests that finding ways to boost the production of such proteins would be a useful strategy for reprogramming adult cells.

While the work was done using mouse cells, the researchers suspect the same basic network will be found in human cells.

The study appeared online earlier this week in Nature.

Computer program predicts protein structure

Harvard researchers reported this week the first computer simulation of the complete folding of a small protein molecule. The modeling program they developed not only predicted the final structure of a test protein almost perfectly, but also identified key intermediate structures during the folding process.

The biological activity of proteins depends on successful folding, and the accumulation of badly folded proteins is now believed to cause many neurological disorders such as Alzheimer’s and Parkinson’s diseases. Computer simulations could be a faster and simpler way of determining protein structure than existing laborious laboratory methods.

So far, the complexity of protein folding—which has millions of intermediate steps—has prevented the development of computer models for all but a few short, simple protein-folding schemes.

The new program uses a very efficient computing method to simulate protein-folding events over microsecond intervals. That’s about a thousand times longer than previous methods and close to the time it takes some proteins to fold in cells.

The program was good at predicting the final shape of a test protein based on its amino acid sequence. In 4,000 separate trials, the computed end structure closely matched the atomic structure solved previously by other techniques. By determining which intermediate structures showed up most often in the multiple simulations, the researchers also arrived at a likely chronology of folding in atomic-level detail.

The paper, from the lab of Eugene Shakhnovich of Harvard, appeared this week in the Proceedings of the National Academy of Science online.

Low-carb diets not harmful, says long-term study

Low-carbohydrate diets, like the popular Atkins regime, help people lose weight in the short term, but because they tend to be heavy in fat, some physicians have suspected they might not be heart healthy. A 20-year study from the Harvard School of Public Health suggests that this is not the case, at least in women. On the contrary, some high-carbohydrate diets may raise the risk of heart disease, the researchers found.

The study, published this week in the New England Journal of Medicine, analyzed the diets and health outcomes of 82,801 women in the Nurses Health Study, a long-running survey of lifestyle factors and disease. Based on data from food questionnaires the women filled out, the researchers divided the women into groups according to the percentage of calories they ate from carbohydrate, fat, and protein.

In the 20-year follow-up, nearly 2,000 of the women developed heart disease. There was no difference in the risk of heart disease for women on low-carbohydrate diets compared with women eating high-carbohydrate diets, even though they consumed more fat.

While the total amount of fat consumed did not seem to affect the women’s cardiovascular health, the kind of fat did. Among women with low carbohydrate intake, those who ate mostly vegetable fat had a 30 percent lower risk of heart disease than those who ate mainly animal fat, according to senior author Frank Hu.

And no matter what their diet, over the course of 20 years, women in every group gained enough weight to bump the average body mass index up a few points.