Why cornflower blue is like a red, red rose

Roses are red, cornflowers are blue - and for almost a century scientists have struggled to work out why.

Although the flowers have strikingly different colours, they share the same pigment, the reason for which has baffled botanists since 1913.

The anthocyanin pigment was first detected in the blue cornflower 92 years ago,and later the same pigment was found in a red rose, leading to the mystery of how one pigment could produce two colours.

Various theories abounded. Perhaps the cells of one flower had a different pH - acidity was known to affect flower colour - or perhaps it was due to differences in metal ions taken up from the soil.

Now a team of scientists from Japan, led by Masaaki Shiono of Kyushu University, reveal in the journal Nature that the answer lies in the way the pigment forms an unusual "supermolecule" in the cells of the cornflower plant.

While being of scientific interest, the development will be followed closely by commercial flower growers who are spending huge sums on trying to nurture blue flowers from plants that do not naturally produce blooms of that colour.

The most famous example is the search to produce a true blue rose. So far the attempt has centred on breeding a rose that produces a blue pigment called delphinidin, which is found naturally in delphiniums, but never in roses.

The scientists in Japan discovered that the pigment in cornflowers forms a giant molecular complex consisting of six molecules of anthocyanin bound to another pigment called flavone, along with four metallic ions - one iron, one magnesium and two calciums. But in roses, the same supermolecule does not appear to form, which could account for why the same pigment can cause different colours in the two flowers, the scientists say. Cathie Martin, professor of biology at the John Innes Institute in Norwich, said that the Japanese study broke the rule that to get a blue flower you needed to add the delphinidin pigment. "People have been trying to turn certain flowers blue, such as carnations, roses and chrysanthemums. The idea was that blue could only be conferred by the delphinidin pigment. It has been the dogma for years," Professor Martin said.

"This study is not saying that the idea is wrong, it's just that there now seems to be another way to make blue flowers and the point about cornflower blue is that it is a true blue and exceptionally beautiful," she said.

Discovering how a pigment found in red roses can be a brilliant sky blue in the cornflower could alter the breeding strategies aimed at creating blue roses. "The basic chemistry of the pigment molecule is the same in both cornflowers and roses. It's the way it's organised in its chemical superstructure that is different," Professor Martin said.