They are invisible at first. In their Southeast Asian forest homes, they grow as thin strands of cells, foreign fibers sometimes more than 10 meters long that weave through the vital tissues of their vine hosts, siphoning nourishment from them. Even under a microscope, the single-file lines of cells are nearly indistinguishable from the vine’s own. They seem more like a fungus than a plant.

But when the drive to breed awakens them, the members of the Rafflesiaceae family erupt as immense, stemless, rubbery red “corpse flowers” covered in polka dots, with a putrid smell like rotting meat designed to draw pollinating carrion flies. The blooms of one species, Rafflesia arnoldii, are the largest flowers in the world — each one can be more than a meter across and weigh a whopping 10 kilograms, roughly the heft of a toddler.

More than a decade ago, Rafflesiaceae parasites caught the eye of Jeanmaire Molina, an evolutionary plant biologist at Long Island University in Brooklyn, who wondered if their genomes were as bizarre as their outward forms. Her initial investigations suggested they were. As she and her colleagues described it in a 2014 paper in Molecular Biology and Evolution, they successfully assembled the mitochondrial DNA from one Philippines species of Rafflesia. But they were unable to detect any functional genes from its chloroplasts. The plants seemed to have simply ditched their entire chloroplast genome.

That was almost unthinkable. Chloroplasts are best known for using light to make food, but like all the food-making organelles called plastids, they contain genes that are involved in many key cellular processes. Even malaria parasites still carry a plastid genome, Molina noted, and their last photosynthetic ancestor lived hundreds of millions of years ago.

This shocking finding has now been confirmed by an independent research team from Harvard University. The draft genome for another member of the Rafflesiaceae family that they recently published in Current Biology is full of surprises, showing how far parasites can go in shedding superfluous genes and acquiring useful new ones from their hosts. It also deepens mysteries about the role of highly mobile genetic elements that don’t encode proteins in enabling evolutionary changes. Perhaps the greatest lesson of the study is how much we still have to learn about genomics, particularly in plants, and in parasites — a category of organisms that includes more than 40% of all known species.

Losing to Win

Like Molina, Charles Davis, a professor of organismic and evolutionary biology at Harvard University and the curator of vascular plants in the Harvard University Herbaria, was drawn into studying the Rafflesiaceae because they are the most “charismatic and enigmatic of all the quarter-million species of flowering plants,” he said.

He has been trying to reveal their many secrets for nearly 15 years, but a nuclear genome sequence always proved elusive. Finally, his doctoral student Liming Cai (now a postdoctoral researcher in systematic biology at the University of California, Riverside) stepped up to spearhead the project, and with the help of the university’s informatics group and its director of bioinformatics, Timothy Sackton, the team was finally able to put together a draft genome for Sapria himalayana, a species with blooms the size of a human head.

Sapria’s genome follows several trends seen in many other parasitic plants (and in parasites more generally). Like them, Sapria has done away with many genes considered essential to its free-living relatives. Because parasites steal from their hosts, they essentially outsource the labor of metabolism, so they don’t need all the moving biochemical parts of an independent plant cell.

Still, Davis was shocked to see that nearly half of the genes widely conserved across plant lineages had disappeared from Sapria. That’s more than twice as many genes as are lost from the parasitic plants called dodders (genus Cuscuta), and four times the losses in cereal-killing witchweeds (genus Striga). “We knew that there would be loss,” he said, “but we didn’t think it would be on the order of 44% of its genes.”

And that’s in addition to the stunning deletion of the whole plastid genome that Molina’s work on Rafflesia had suggested. The only other organisms known to have jettisoned that genome are single-celled algae in the genus Polytomella, which gave up photosynthesis in favor of absorbing sustenance from the waters around them.

Molina said she found the confirmation of her team’s finding “comforting,” but also confusing, because Rafflesia still seem to make their plastid compartments. “When we did electron microscopy studies, we found plastids,” she said, “so it’s just quite bizarre that the plastids are empty.”

Sapria also seems to have cut other genetic corners. The plants have deleted the noncoding stretches of DNA within many genes. These regions, called introns, are interspersed among the parts of genes that code for the actual protein that is produced.

It might sound as though Sapria and its kin have simply made their genomes smaller and more efficient. But paradoxically, Sapria’s genome is big: an estimated 3.2 to 3.5 gigabases of DNA in total, roughly the same size as ours. What is filling up its genome?

A Life of Theft

For starters, it’s loaded with stolen genes. Davis’ team estimated that at least 1.2% of the plant’s genes came from other species, particularly its hosts, past and present. That might not sound impressive, but this kind of horizontal gene transfer is considered exceptionally rare outside of bacteria. So even a single percent of genes arising this way raises eyebrows.

Because these parasites have been stealing genes for millennia, Cai noted, their genome is like “a huge graveyard of DNA.” By carefully digging through that graveyard and comparing its contents to the genomes of 10 types of vines that seemed like potential hosts, Cai and her colleagues were able to peer back in time. “These horizontally transferred genes are serving as DNA fossils,” she said.

From these fossils, they unearthed “an extinct host-parasite association that dates back to maybe the mid-Cretaceous,” she said. Today, the roughly four dozen known species of Rafflesiaceae all infest vines from a single genus, Tetrastigma. But long before the parasites infested Tetrastigma, they seem to have infested and stolen from peppervines (genus Ampelopsis). This kind of ecological history is all but impossible to deduce from stony fossils: The parasite’s flowers don’t last long, and the thin, threadlike remains of its vegetative body are unlikely to fossilize.

Evolution on Repeat

Yet stolen genes represent only a paltry fraction of Sapria’s huge genome. The vast majority of it consists of copies of DNA sequences called transposable elements (also known as transposons or “jumping genes”). “The genome of this plant is something like 90% repeat elements,” said Sackton.

That high level of repetition is in fact why Davis struggled so long to assemble a draft genome for Sapria. Until about the past decade, genome sequencing technologies were easily stymied by DNA with too many indistinguishable, repetitive sequences. “It’s like trying to do a puzzle of a completely clear blue sky where every piece is exactly the same shape,” Sackton said. “There’s just no way to do it.”

But Cai and colleagues were able to take advantage of current sequencing technologies, which can handle much longer (and therefore more distinctive) stretches of DNA. Even so, they were only able to reconstruct what they estimate is 40% of the Sapria genome — the rest was still too repetitive.

This abundance of transposable elements is striking, says Saima Shahid, a plant biologist with the Donald Danforth Plant Science Center in St. Louis, who studies the functions of transposable elements in plants. It’s about twice what is seen in dodders. And in the other plant parasites sequenced to date, the dominant elements are “retrotransposons,” which move within the genome by first being transcribed into RNA. Sapria, however, is mostly filled with DNA transposons that repeatedly copy and paste themselves into the genome directly. “That’s something very interesting and unusual,” Shahid said.

But why does Sapria have so many of these jumping genes in the first place? No one is yet sure, but the answer may transform our understanding of parasite genomics.

Transposable elements are considered “selfish” genes; they replicate even at the expense of the genome they occupy. For that reason, host genomes usually rein in their expression. “Most of the time, they’re targeted for silencing,” said Shahid. It seems that either regulation has somehow gone awry in the Rafflesiaceae, or the parasites find some benefit in letting these elements jump around.

Cai, Davis and Sackton have a hunch that the superabundance of transposons is a consequence of the isolated lives these parasites lead. Because the Rafflesiaceae only invade Tetrastigma vines, each patch of vines is practically a desert island isolating its parasite inhabitants. In small populations with restricted growth and little gene flow from the outside, even less helpful genetic features can become more commonplace through sheer chance. That may mean that even deleterious copies of transposable elements “accumulate through time,” said Cai, until “you end up with these highly atypical gene structures.”

Another possibility is that the parasites can’t stop their jumping genes from jumping. Some of these elements came from their hosts, and they might be different enough that the parasite’s genetic machinery fails to recognize and silence them immediately. “It’s like an invasive species, basically,” Sackton said.

It could also be that, given how much genetic material the parasites acquire, they evolved adaptations that raised their tolerance for the extra burden of useless DNA, so there simply isn’t enough selection pressure to rid them of these jumping genes.

But to Shahid, it doesn’t make sense that a genome showing so many signs of streamlining — with the loss of genes, noncoding sequences within genes and an entire plastid genome — would be blasé about genomic dead weight. Worse, transposable elements are dangerous: “You must spend a lot of energy in order to silence them,” she said, lest they lay waste to everything. She finds it more likely that these transposons are doing something for the parasite; the question is what.

Their presence might be tied to all those stolen genes. When transposons jump, Shahid explained, they often bring bits of nearby DNA with them. “These transposable elements might be helping them to carry gene fragments, and then insert them into their own genome,” she said.

The transposons might then be the engines driving horizontal gene transfers that the parasite needs to survive. For example, they might help the parasites steal some of the host’s important gene regulators.

In a 2018 paper in Nature, Shahid and her colleagues showed that the parasitic plants called field dodders export tiny microRNA molecules into surrounding host cells as “off” signals for some of the host’s genes — presumably, to shut down defenses that would interfere with the theft of the host’s resources. Those microRNA controls may have made their way into the parasite with the help of a jumping gene. (However, no one has yet investigated whether Sapria and its relatives export microRNAs during their dormant stage.)

Transposons can influence gene regulation in other ways. When inserted into introns, for instance, they can enhance a gene’s expression, or they can guide inhibitors to shut the gene down. In Sapria, not all the introns have been reduced in length: Some are expanded to nearly 100 kilobases, making them the longest known introns from any plant. And transposable elements are responsible for 74% of all the expansions.

Transposable elements can cause chunks of the genome to move around, too, which can be dangerously destabilizing — but can also lead to gene duplication and innovation, Shahid said. That might help parasites stay a step ahead of their hosts’ defenses. Transposons could also be responsible for some of the unique features of the Rafflesiaceae: Molina has started to wonder if they have “something to do with the large flowers.”

Without more information, it’s impossible to tell how much of the huge cache of transposable elements in the Rafflesiaceae is functional and how much is just genomic clutter. Shahid said that one way to get to the bottom of all this might be to take a closer look at where different kinds of transposable elements sit in relation to other genomic features. That could help reveal whether the elements are playing pivotal roles in gene expression. She would also like to see if (or more likely when) these transposable elements are being expressed, as that could also give clues to their potential functions in the genome.

Cultivating Curiosity

Further research on these floral oddities could teach us a lot about everything from plastids to jumping genes, but the scarcity of the plants makes probing these questions much harder. Finding them involves hours of trekking deep into often dangerous jungles, Molina explained. In the Philippines, they’re found in forests where armed rebels often hide out — “We have to coordinate with a mayor to make sure it’s safe,” she said. And grueling fieldwork aside, the countries where these flowers grow often restrict their export, in part because the plants are critically endangered.

Because of their rarity, Molina is working with the United States Botanic Garden in Washington, D.C., to cultivate these parasites and their host vines domestically. Coaxing them to grow and bloom in Washington would bring them into the public eye, and she thinks allowing people to see them in person would aid in conservation efforts and enable much more detailed research. For now, though, these parasites that stretch the definition of what plants can be are keeping their secrets.