Building Beauty

Deconstructing flowers yields the secrets of petals, scents and hue

Flowers are essentially variations on a single theme: Come hither. Instead of lipstick and lace, flowers advertise with vivid petals and ultraviolet stripes. Some plants offer a legitimate exchange of goods — visitors are rewarded with protein-rich pollen or sweet nectar. Other flowers deceive, mimicking the scent, color and feel of a rotting carcass to entice carrion flies looking for a suitable place to start a family. Even plants pollinated by the wind dress for success, taunting the air with copious, lightweight pollen.

THE ABC MODEL | The ABC model of floral development emerged in the late 1980s from research on flowering plants with genetic mutations. The model originally described three classes of mutant plants, linking genes categorized as A, B or C to predictable mistakes in flower structure. A flower normally is made up of four concentric layers, or whorls, of parts: sepals form the outermost whorl, then petals, stamens and carpels. In plants without functioning A, B or C class genes, floral organs developed in the wrong whorl or didn’t develop at all. The revised ABC model, shown above, includes two additional classes of genes (D and E). It shows how the genes work in various combinations to trigger the specialization of tissues into the whorls of a flower. A and E class genes are needed to make sepals; A, B and E for petals. To make stamens, B, C and E genes are active. C and E are required for carpels, and C, D and E for ovules inside the carpels. The model has held up well, but may change as scientists probe floral development in more species. Illustration by Gina Mikel

STOP AND SMELL THE ROSES | Scientists are zeroing in on the genes and enzymes that help build the hundreds of compounds that go into a rose’s fragrance. Adamkaz/istockPhoto; Source: Gang/Annual Review of Plant Biology 2005

THE COLOR BLUE Metals, pH and cellular architecture all can contribute to floral blues. Image Plan/Corbis

Beneath all the superficial charm of flashy petals or intoxicating scents is the serious business of sperm and eggs, reproduction, the continuance of the family line. Flowering plants excel at being fruitful and multiplying (heck, they invented it).

Although they are evolutionarily younger than their nonflowering relatives, the flowering plants — aka the angiosperms — are far more successful, conquering every continent and diversifying into some 250,000 species. Their evolution is credited with altering

Earth’s landscape in a way that allowed the rise of other life, such as insects, mammals and eventually primates.

Today people use flowers to woo, apologize or express sympathy or blessings. And, of course, people eat them. Flowers that have matured into fruits provide nearly 70 percent of human food. Rice, wheat, beans and corn are all technically seeds from fruits — the end point, and point of, flowers. Given what’s at stake, it’s not surprising that plants construct such elaborate advertisements — or that scientists are bent on figuring floral structures out. Such research could lead to improved crops, designer scents or novel compounds for drugs.

Recent genetic and molecular studies suggest there’s a curious simplicity underlying a flower’s structure — although ongoing work keeps adding new twists. Flowers’ wide array of petal designs, scents and colors appears to stem in part from a rich genetic toolkit that gives plants developmental flexibility. Many flowering plants use the same basic genes to direct the development of petals, though the control of other floral parts appears more complex. Investigations of scent compounds also implicate multiple copies of identical or closely related genes —exemplified by work detailing how some roses lost their namesake fragrance. The complexities of color are also being unraveled. For example, scientists are looking beyond pigments to learn how metals, cell shape and pH may give flowers the blues.

Floral ABCs

Plants are extremely opportunistic — stick a leaf from a jade plant in some sand and it may sprout roots. The roots may grow a shoot and, with the right light, the shoot may grow flowers (imagine planting a piece of tail and growing a squirrel). That capacity reflects an enormous developmental flexibility, one that sets plants apart from animals, says Beverley Glover of the University of Cambridge in England.

“A zoologist would say that plants have to have this flexibility to cope with the design flaw of being sessile, staying in one place,” says Glover. “I would argue that animals need mobility to cope with the design flaw of a limited genetic and developmental program.”

Plants’ flexibility appears to arise from a genetic pantry that’s extremely well-stocked. The genomes of some plants are outrageously large: The book of genetic instructions for the ornamental lily Fritillaria is nearly 40 times the size of the human genome. And many plants have multiple copies of identical or closely related genes.

When petals are made, “It is the same sorts of genes doing the same things in a slightly different way,” Glover says. “Maybe I am naive, but I think it is surprising.”

Flowers have four concentric whorls of parts: sepals on the outside, then petals, then stamens and carpels at the center. In the 1980s scientists working with mutant plants — Enrico Coen with snapdragons and, independently, Elliot Meyerowitz with Arabidopsis—described how three classes of genes (A, B and C) orchestrate the correct development of these floral parts. Called the ABC model of development, the idea hearkened back to a concept put forth in 1790 by German poet Johann Wolfgang von Goethe, who in botanical circles may be as famous for leaves as he is for Faust. His treatise on plant metamorphosis proclaimed, “All is leaf.”

Two hundred years later the notion that sepals, petals, stamens and carpels are all modified leaves got its molecular validation. Both Coen and Meyerowitz noticed that in plants in which certain genes were disabled, some floral organs were made in the wrong place or weren’t made at all, and in some cases, leaves were made instead.

Plants missing A class genes, for example, did not grow petals. B class mutants didn’t grow petals or stamens, instead developing two whorls of sepals and two of carpels. The model, since expanded to include D and E class genes, predicted that mutant plants would have particular variations on the normal floral arrangement, predictions that have largely been borne out by research. Yet snapdragon and Arabidopsis are but two of thousands of flowering plants. So researchers are still probing where the model holds up and where there are variations on the ABC theme.

New work investigating flower building in the buttercup family backs up the role B class genes play in petals, overturning some long-held beliefs about the family’s dizzying array of floral forms, says Harvard University’s Elena Kramer.

The buttercup family includes delphinium, monkshood, anemone and columbine. Looking at the family tree, it appears that petals have been lost and gained many times in this group’s evolution. And many of the flowers in this family have a second, inner whorl of petals that look suspiciously like modified stamens, a flower’s pollen-bearing male structures. Then there are the sepals, which in many plants are green and diminutive. Not so in the buttercups, many of which have glammed up their sepals with pigments and nectar spurs.

Still, in the January American Journal of Botany Kramer reports that no matter the size, shape or color, buttercups and their relatives all recruit the same genes to direct the development of petals.

“Perhaps there aren’t so many ways to skin a cat,” says David Baum of the University of Wisconsin–Madison.

It isn’t clear which genes lead to the stamenlike petals, Kramer notes. And the outermost whorl, the sepals, remains a puzzle. When this outer whorl becomes showy and petal-like, as in delphinium, it sometimes recruits petal genes. But that’s not so in other plants, such as asparagus, suggesting the ABC model will continue to be revised.

“What’s going on with sepals is anybody’s guess,” Kramer says.

Chemical billboards

Ostentatious petals or sepals are just one asset in the portfolio of floral attractants. Smell lilacs, jasmine or a lily and it’s clear that plants also advertise through the air. Plants manufacture a huge repertoire of chemicals: The structures of nearly 50,000 plant compounds have been elucidated, says David Gang of the University of Arizona in Tucson.

Besides helping to attract pollinators, metabolic compounds may ward off grazers, defend the plant from fungi or suppress the growth of neighboring plants.

“Plants can’t get up and run away,” says Gang. “So they can make things like spines to defend themselves, or they can make poisons.” (Gang’s enthusiasm prompts visions of plants with a wary eye on their surroundings — stinging hairs cocked, cyanide at the ready.)

Some plant biochemical pathways — the series of chemical reactions that generate these compounds — are extremely ancient. The floral scent pathway that produces terpenoids (a large class of aromatic chemicals) is also present in animals and fungi, Gang notes. Most enzymes — the proteins that work on the assembly line as molecules go by (add a sugar, take off a methyl group) — in these paths are very similar across the plant kingdom. But look at the enzymes farther along the scent pathway and the diversity explodes, says Gang.

Scientists believe much of this diversity has arisen through duplications of the genes that encode enzymes in the scent pathways. A duplicated gene may accumulate mutations that give it a slightly different function than the original gene. The result can change the scent produced by the biochemical pathway. This appears to have happened in modern roses, scientists reported last year in the Proceedings of the National Academy of Sciences.

Roses have long been recognized for their fragrance — some 300 to 400 compounds go into a rose’s scent. Many modern roses result from extensive breeding of two main lines: European varieties, which contributed resistance to cold and pests, and Chinese varieties, which contributed repeated flowering. Each variety also contributed its characteristic scent. (Never mind Juliet waxing about them smelling as sweet, no matter the name.)

“One of the big things that happened when we domesticated roses is we eliminated scent,” says Gang. “The ones you find in a garden with the most fragrant smells have very few petals but smell heavenly.”

Damask roses, a common European variety, are typically thought of as rose-scented, says Gang. Their fragrance is dominated by rose oxide, a compound prized by the perfume and flavor industries. But the modern tea roses that dominate the coolers of florists across the United States are a hybrid of the European and Chinese varieties. They often smell more like tea than eau de rose.

A major contributor to the tea aroma is a compound known as 3,5-dimethoxytoluene, or DMT. Two related enzymes, known as OOMT1 and OOMT2, are responsible for the reactions that produce DMT. Due to a gene duplication event, Chinese roses have both OOMT1 and OOMT2. European varieties have only OOMT2, which isn’t active in their petals. This means that European roses smell like, well, roses. Hybrids, on the other hand, inherited both OOMT enzymes from their Chinese relatives and so produce the DMT-tea odor.

The study is a nice example of how a single duplication event can lead to big changes in scent, says Gang. And figuring out the details of the duplication may lead to new rose varieties that smell as sweet as their name.

“We’re at a point where perhaps we could rebreed the rose scent back into the flowers with all the petals,” he says.

Blooming blues

While tantalizing perfumes and artfully arranged petals do their part to attract, something’s got to grab the attention of a pollinator from afar. Studies suggest that visual cue is often color. Just three major pigments are responsible for most of the color seen in flowers: betalins, carotenoids and anthocyanins.

When orange, red, blue or purple flowers catch your eye, it is most likely the work of the anthocyanins. These pigments are well characterized and often have other functions, such as acting as a sunscreen in leaves. So flower color can be strongly affected by what’s happening elsewhere in the plant — genes in charge of cranking up an herbivore deterrent in the leaves may yield flowers of a different shade, for example, says Mark Rausher of Duke University in Durham, N.C.

In a May 2008 paper in Evolution, Rausher and coauthor L. Caitlin Coberly reported that, in the common morning glory, versions of a gene that encode for white flowers are favored in some populations. Yet white flowers never dominate in these populations, perhaps because shutting down the flow of purple pigment in flowers would also shut down the pathway in the rest of the plant. This could hinder other important functions provided by the pigment, including protection from UV light and interactions with helpful soil fungi, says Rausher, who also explored trends in floral color shifts in a 2008 paper in the International Journal of Plant Sciences.

The local pigment environment — the architecture of the pigmented cells or their pH — also affects how pigments are perceived from outside the flower, evidence suggests.

For example, the rose cultivar Rhapsody in Blue changes from red-purple to bluish-purple with age. This shift in hue results from the accumulation of anthocyanins in petal cells. Similarly, in 2007 Glover and her colleagues reported in Arthropod-Plant Interactions that a mutation altering the surface cells of a snapdragon’s petals shifted light’s path within the pigment cells, making them appear pink rather than purple. And in more alkaline conditions, pigments in some red or pink flowers become blue.

Recent work also implicates metals in the path to the blues. In 2005 Japanese scientists reported that iron, magnesium and calcium work together with the plant compounds anthocyanin and flavone to produce the blue of cornflowers. Another group in Japan has shown that the sky blue of Himalayan poppies, Meconopsis, results from the bonding of iron and magnesium to an anthocyanin and a flavenol. The blue of the dayflower, Commelina, comes from a complex formed by magnesium, anthocyanin and flavone, a team reported last year. And the blue hue of hydrangeas is due in part to the buildup of aluminum in the sepals. The metal is less available in more alkaline conditions, so for pinker flowers gardeners can add lime, which is highly alkaline, to the soil.

Advances in understanding flower color will undoubtedly continue to emerge — the market for ornamental plants and cut flowers totals over $70 billion annually in sales, and the generation of new colors is a hot area of research for the industry. New fragrances — not only from flowers, but also leaves (such as new varieties of sweet basil) — are also being explored, and some of the investigated compounds hold promise as potential new drugs. As scientists come closer to deconstructing flowers, they will undoubtedly get better at constructing them as well. But nature’s got a good lead — one that will be hard to beat.