Blue roses may be the horticultural Holy Grail, but despite careers-worth of effort no plant breeder's been able to come closer than lavender. This is due to a fundamental quirk of rose biology that's now been "fixed" with genetic engineering.
Plant color is largely determined by three classes of pigment chemicals: chlorophylls (photosynthesizing greens), carotenoids (yellows, oranges and reds), and anthocyanins (red, purples and blues). Each one of these three classes contains many chemicals which differ slightly from each other in molecular structure, and therefore chemical properties (e.g. hue). Each of these unique chemicals is produced and modified by an array of enzymes. Ultimately, the color of a given plant tissue is determined by what combination of these enzymes are present.
The headwaters of the anthocyanin biosynthetic pathway begin when an enzyme (chalcone synthase) smooshes together two chemicals (coumaroyl CoA and malonyl CoA) to create a third chemical (a chalcone). Another enzyme (chalcone isomerase) converts this chemical into naringenin (which contributes to the bitter taste of some citrus fruits). As more and more enzymes begin to work, this pathway quickly branches into many different streams, leading to many chemicals that play important roles in plant food flavor and nutrition.
Anthocyanidins pool at the end of a few of these streams with names that are often indicative of their color and origin of discovery: peonidin, cyanidin and delphinidin. The stream that leads to delphinidin (a blue chemical) requires several specific enzymes, including F3'5'H (flavonoid 3'5'-hydroxylase), which smooshes a molecule of water onto a specific position on the pigment molecule.
Since no rose varieties contain a functional copy of this enzyme (F3'5'H), no amount of clever plant breeding has been successful in creating a blue rose.
An Australian biotech company, Florigene, has succeeded in this long-sought goal by using genetic engineering techniques to move a working copy of this enzyme from a pansy into a rose.
Let me know if you see this recently-licensed variety at your local garden center!
Tanaka and Ohmiya. 2008. Seeing is believing: engineering anthocyanin and carotenoidbiosynthetic pathways. Current Opinion in Biotechnology. 19:190.