The Voyage of the Petunia

We humans have a biased view of sex. To make offspring, we think, it takes two: male and female get together and swap DNA. But what if, like a plant, you can't move? Having your lower extremities buried securely in dirt complicates things slightly. You need to fool insects and hummingbirds into carrying your DNA around for you, trading food for vicarious sex. Problem solved, it seems, except that some plants, like the petunia, never managed to separate into two: the male anthers and female pistils both reside on a single plant. And since inbreeding tends to make a species less diverse and thereby more vulnerable to environmental threats, there are distinct advantages to making sure you don't accidentally have a relationship with yourself.

Though some plants (of lesser moral fiber, perhaps) don't care whose genes they get, the stiff-necked petunia does its best to refuse any advances it might make upon itself. It was Charles Darwin who first observed that some plants self-pollinate and that others (like the petunia) do not. Since then, over 130 years ago, advances in the field haven't exactly come at a breakneck pace; a solution wasn't proposed until the early 1980s, and wasn't directly proved until Penn State's Teh-hui Kao tracked down and bagged the gene responsible in 1994. Though Darwin knew nothing about DNA, he did notice what happens superficially during pollination. When the sperm-bearing pollen reaches the pistil, it forms a pollen tube that burrows deep into the flower, seeking out the ovaries. If the pollen is foreign, the tube extends the full length of the pistil, allowing the sperm to reach the ova, or eggs. But if the pollen is self-made, the tube only extends about two-thirds of the distance before it mysteriously stops and is destroyed.

This discrimination and destruction is known as self-incompatibility, and researchers have theorized that a single gene is responsible. Such an S (for self-incompatible) gene would likely produce a corresponding S protein in the pistil, which somehow stops the pollen. If these ideas sound vague, it's because they are. A probable S protein was identified in the 1980s, but all proofs of its nature were circumstantial or indirect. Not until Kao tackled the problem in petunias did direct proof of the function of the S protein become available.

Darwin, like Kao, entered his field of study in a roundabout, accidental way. Darwin was fourth in line for the post aboard the H.M.S. Beagle, accepting only after all the candidates had refused the job (including himself, though he later reversed his decision). And the job was not that of naturalist—Robert MacCormick, ship's surgeon, filled that role. No, Darwin wound the clocks. The Beagle was a mapping vessel, taking precise longitudinal measurements, and Darwin was responsible for the ship's 22 chronometers—not the expected job for the father of modern biology.

Kao, too, began his voyage in a different field. As a molecular biology post-doc at Cornell, he didn't picture petunias in his future. But in 1984, he joined a team that was examining self-incompatibility in brassica cabbages. He could clone genes, something no one else in the group could do, and soon became a valuable member of the team. "It was a very productive collaboration," says Kao; so productive, in fact, that it initiated his decade-plus journey into plant research. In 1985 he published his first paper on self-incompatibility with the Cornell team, and a year later accepted an appointment as assistant professor at Penn State. Not wanting to compete with his former colleagues, he turned from cabbages to plants like tobacco, tomatoes, potatoes, and the aforementioned petunias—the flowers that yielded up a host of genetic secrets as well as a cover story in the 10 February, 1994 Nature.

No voyage of discovery, no creation, it seems, can begin without an act of destruction, no matter how small. Rather than by shattering the traditional champagne bottle, Kao's journey was christened by grinding up petunia pistils. The pistils were known to contain the S protein, but the S gene had yet to be found. It, too, presumably was in the pistils, waiting to be isolated. After grinding, the proteins in the pulverized pistils were separated by their molecular weight, and the S proteins identified from among these based on their segregation with respective S alleles. That is, Kao's team—including postdoctoral researcher Hyun-Sook Lee and graduate researchers Shihshieh Huang and Balasulojini Karunanandaa—found which gene was always present when a certain protein was present, and which gene was always absent when that protein was absent. "But that evidence was only indirect," Kao explains, "—circumstantial." Since an organism's DNA is made up of many genes linked together in a chain, it was possible that the genes which seemed to correspond to the S proteins were merely nearby links in the chain—neighbors of the S gene, and not the gene itself. "We needed to find direct evidence," he says.

Direct evidence meant two proofs: a gain of function, and a loss of function. If adding the gene to a plant that lacked it caused self-incompatibility, they proved gain of function. And if removing the gene from a plant allowed it to pollinate itself, they proved loss of function. But with this simple-sounding solution travels fundamental problems—like how to get the S gene into and out of a plant's DNA sequences.

All petunias have the S gene, in one form or another. Yet like the genes that determine eye color, the S gene is polymorphic, with many different slight variations of the same gene. Each variation, called an allele, and its corresponding protein, is labeled as S1, S2, S3, and so on. Each petunia has two alleles, and therefore contains two types of S protein—for example, S1 and S2 (this plant would be said to have an S1S2 genotype). Since the sperm and egg each contribute half of the genetic material for the offspring, the sperm-bearing pollen from the S1S2 petunia will have either the S1 or the S2 protein. These two types of pollen—whether from the plant's own anthers or from a different plant that also makes S1 or S2 pollen—are the only two that the flower will not allow to fertilize its ova.

The S1S2 petunia, then, was resistant to S1 and S2 pollen, but receptive to S3 pollen. Kao needed now to make it turn away the S3 pollen as well. If the S3 gene were introduced into the S1S2 plant and the plant then resisted S3 pollination attempts, it would prove that the suspected gene was actually the long-sought S gene and not merely a neighbor.

The matter of actually installing an S3 gene into an S1S2 plant's DNA sequence was a tricky problem. Part of the solution was old hat for Kao. Since he needed to get the S3 gene before he could place it into another plant, and since you can't exactly pluck it out with tweezers, the way to do it, Kao knew, was through gene cloning. He isolated DNA from the leaf tissue, then used enzymes to break the DNA into smaller fragments. To clone the DNA, he inserted each gene fragment into the bacteria of a separate colony of E. coli. The colonies grew, making larger amounts of the original DNA fragments, though Kao didn't yet know which colony held which gene. But he had a way to find out. From RNA, which, in cells, takes DNA codes and then makes proteins, you can reconstruct partial copies (called cDNA) of the original DNA. His team had previously created S3 cDNA, so Kao "labeled" it with phosphorus-32, a radioactive isotope. Since the cDNA and original genes are so similar, they anneal, or group together when mixed. In this case, it marked the S3-bearing colony with radioisotope, leaving it "hot" and clearly visible on X-ray film.

Kao now had his cloned S3 gene. But for its introduction into the S1S2 plant, Kao, the molecular biologist, had to turn to a genetic engineer with a great deal more experience: the agrobacterium. Admittedly, his partner had an unsavory reputation, with a history of abuse to its discredit. Agrobacterium would prey on wounded plant roots, infecting its hapless, immobile victims and striking at their very chromosomes. For deep within the dark heart of this single-celled burglar lies a structure known as a TI-plasmid. The initials stand for "Tumor-inducing," and the T-DNA within the plasmid insinuates itself into the infected plant's own DNA, causing tumors to grow. Not the most pleasant of company, perhaps, but apart from its morality, one thing can be said for the agrobacterium—it's born to be a genetic engineer. The agrobacterium is also a simpler organism, and easier to manipulate than the petunia. The S3 gene can be inserted without too much difficulty, and its harmful components turned off. The agrobacterium then becomes not the scourge of weakened plants, but a rather useful mechanic, a genetic gelatin capsule.

Kao plucked a leaf from the S1S2 genotype petunia, and with a scalpel, slashed it to ribbons. He soaked it in a solution teeming with S3-gene-carrying agrobacteria, then waited. "Plant tissues aren't like animal tissues," he says, noting a truism that makes the whole experiment possible. For after the leaf was left to soak, its infected cells began to dedifferentiate. That means they became less specialized, eventually forming a callus, a mass of undifferentiated cells. But the plant-animal differences don't end there, for the cells soon began to redifferentiate, growing as if from a seed. Small green shoots sprouted, then roots, and within four months, an entire transgenic plant had grown from the callus—a whole petunia, grown not from a seed, bulb, or even from germ cells, but from an infected leaf. "A transgenic plant is indistinguishable from a normal plant," Kao says, "except that our transgenic petunia also contained the S3 gene." After grinding up pistils from the engineered petunia, Kao's team saw that, in addition to the S1 and S2 proteins present in the original plant, the new petunia also carried S3 protein, giving it an S1S2S3 genotype. And, when exposed to S3 pollen, it resisted fertilization, proving that S protein is sufficient for self-incompatibility. With the gain-of-function proved, he and his colleagues were halfway to their goal.

Removing a gene is an entirely different matter from inserting one. Gene cloning makes a copy of the gene, but it doesn't take it out of the original sequence. And there are no agrobacteria that remove genetic material, only ones that introduce their own. Kao needed a new tactic.

For a protein to be made in a cell it needs the help of not only DNA but also RNA. DNA is a molecule made of two complementary strands, whereas RNA boasts only a single strand. One strand of DNA is normally used as a template to transcribe a complementary RNA strand (called sense RNA, because it makes genetic "sense"), which in turn makes proteins. But if a section of the gene is flipped by 180 degrees, the normally unused strand of DNA becomes the template. The antisense RNA this flipped template produces is an exact opposite to ordinary sense RNA, and when both are present, they join up and cancel each other out, making it impossible to manufacture any proteins.

To eliminate S3 proteins from the petunia, Kao and his team cloned the S3 gene and reversed its sequence, so it would produce antisense S3 RNA. They then introduced the antisense S3 gene into the plant—this time, an S2S3 genotype plant—via the agrobacterium and again waited for a new petunia to grow.

After a pistil grinding and protein analysis of the new plant, the researchers found that the S3 protein was, as expected, missing. Since the S3 gene had been turned off, it failed to produce any S3 protein. Further, when S3 pollen reached the pistil, the flower failed to fight it off, resulting in fertilization. Kao and his colleagues had completed the second proof, loss of function, demonstrating that S protein is likewise necessary for self-incompatibility. "At this point," says Kao, "we finally had direct evidence. We're confident that we're working on the right protein."

Even though he's found the true S gene, it doesn't mean he's out of a job. Kao and his colleagues have moved on, getting involved in two new problems of self-incompatibility. "We wonder what the molecular basis of self/nonself discrimination is," he says, "and what the biochemical mechanism of self-rejection is." For though they've proven that the pistils do recognize S proteins, they don't yet understand how it's done. And as for the latter, the mechanism of rejection, Kao and Co. are already well on their way.

S proteins, it turns out, are ribonucleases—that is, they're enzymes that degrade and digest RNA. In this way, Kao hypothesizes, the S proteins could destroy the pollen tubes during their growth in the pistil. To prove the necessity of ribonuclease activity, Kao simply had to take that activity away. He created a mutant S3 gene by replacing a key amino acid that ribonuclease activity requires. The S3 mutant had all the properties of a regular S3, save that it was no longer capable of digesting RNA. The team installed it via the usual agrobacterium method into an S1S2 plant, and grew a transgenic S1S2S3 petunia. Ordinarily, such a plant would behave exactly as the first transgenic plant he had created, destroying the S3 pollen tubes. But this time, with the mutant S3 in place, the plant was unable to stop the advancing tubes and was successfully fertilized with S3 pollen.

Kao's work is not just a satisfying solution to a puzzling mystery, it is a promise of simpler, cheaper agriculture. Since most commercial crops today are hybrids, controlling self-pollination is an important issue. Hybrid seeds are made by crossing two plant lines, and if either of the parent plants self-pollinate, the resulting seed will not be a hybrid and the cross will have failed. Currently, growers must take active measures to prevent such accidental pollination, increasing the effort and cost of hybrid seed production. But if the plants could be made self-incompatible by Kao's methods, then all seeds produced would be guaranteed hybrids, requiring no further assurances.

On the flipside, Kao's discoveries could also make growing crops like apples much cheaper and easier. Apple blooms are self-incompatible, and require the pollen of a genetically different neighbor to bear fruit. The logistics of planting compatible trees near each other while still growing different varieties of apples can be tricky. But with Kao's techniques, each tree could be made self-compatible, and thereby self-sufficient. Planting what they want, where they want, sweetens the harvest for both the growers and the consumers.

Kao, as he continues his journey of discovery, glimpsing strange new lands within the flower of the petunia, is an example of how strong a role fate plays in science. For, like Darwin, if you wind up in a job you never imagined you'd do, you just might change the world. And while Kao's discovery doesn't carry quite the same significance as Darwin's, he has finally completed a journey that the famed naturalist began more than 130 years ago.

Teh-hui Kao, Ph.D., is an associate professor in the department of biochemistry and molecular biology, 403 Althouse Lab, University Park, PA 16802; 814-863-1042. Hyun-Sook Lee, a former postdoctoral researcher in plant molecular biology, participated in this project. Graduate students who have worked on the project include Shihshieh Huang and Balasulojini Karunanandaa.

The paper "S proteins control rejection of incompatible pollen in Petunia inflata," by Hyun-Sook Lee, Shihshieh Huang, and Teh-hui Kao, appeared in the 10 February, 1994 issue of Nature. This work was funded by the National Science Foundation and the U.S. Department of Agriculture.

Matthew Holm is a former Research/Penn State intern.

Last Updated June 01, 1996