Theoretically

Don't peel that grape. A team at the University of Illinois at Chicago reckons that a substance called resveratrol, found in grape skins, may prevent cells turning cancerous and inhibit the spread of cells which are already malignant.

The team conducted hundreds of tests looking for anti-cancer compounds in foods that were widely available. In a study published in the latest Science magazine, they found that the grape came out best. But now the provisos: resveratrol has been tested only in cell cultures and laboratory animals. Still, the results offer the promise eventually of developing pills that will defend against cancer

The disease-causing effects of the BRCA2 breast cancer gene depend on where it is mutated, according to researchers at Addenbrooke's Hospital, Cambridge. Mutations toward the middle of the gene, discovered in 1995, predispose a woman more to ovarian than breast cancer.

The finding could help refine genetic tests, making them better predictors of whether a woman with a family history of breast cancer will get the disease. "This observation may ultimately be important in the counselling of women at risk of cancer due to BRCA2," said Simon Gayther in the January issue of Nature Genetics.

Perhaps PrP - the protein which in its "rogue" form causes mad cow disease and CJD - normally plays a role in the immune system, suggest a team at the Institute for Animal Health in Edinburgh. They compared ordinary mice with others that lacked the gene that codes for PrP, and found that normal mice produced up to twice as many T-cells (white blood cells essential to the immune operation) as those lacking prions.

Is the universe structured like a giant honeycomb, with clusters of galaxies dotted through a regular pattern of voids? That's the idea put forward by an international team whose findings - that there are star systems roughly every 391 million light years - have surprised others. The study, in last week's Nature, revealed "a quasi-regular, three-dimensional" pattern in the distribution of galaxies.

Albert Einstein, a scientist you may have heard of, apparently calculated the possibility of "gravitational lensing" - the perceived bending of light from a distant star by the gravity of a nearer, more massive object - in 1912, almost 24 years before he wrote a paper on the topic. A reconstruction of early research notes has found that Einstein discarded the idea as something that couldn't be confirmed empirically. But lensing is an important astrophysical consequence of his general theory of relativity, published in 1915; and it was confirmed by observation in 1979.