ABOVE: © nicolle fuller, sayo studio

Back in the spring of 2012, Duke University graduate student Fay-Wei Li was struggling with his dissertation research. He had set out to uncover the evolutionary history of neochrome, a strange photoreceptor found in ferns that allows the plants to sense and grow toward light sources. It was a big innovation for the plant clade as it enabled life in the shade, and helped them thrive as their flowering cousins rapidly took over the planet during the Cretaceous period some 144 million to 65 million years ago. Indeed, scientists hypothesize that neochrome was a big part of why ferns, which reproduce via spores, not only survived in the face of their new floral competition, but actually flourished, now boasting more than 10,000 species—about four times the biodiversity of any other group of non-flowering land plants.

It was a dream project for Li, as he says he’s always adored ferns and even credits them with inspiring his career in science. The only problem was that he was nearly two years into the project and had little to show for it, as he was struggling to find homologs of the neochrome gene or pieces of it in the available genome data—clues that might shed light on how the protein evolved. “I remember I was sitting in my advisor Kathleen Pryer’s office, very depressed . . . really worried about my prospects as a scientist,” he says.

Then a fellow grad student helped him get early access to the wealth of transcriptome data being amassed by the One Thousand Plant Transcriptomes Initiative. If other plant species had similar photoreceptor genes, their sequences were sure to be in there. So Li designed a bioinformatic pipeline for detecting genetic code for photoreceptors and started scanning the data for neochrome homologs. Then one day, while he sat on the floor of his Durham, North Carolina, duplex watching TV with his laptop running the program on his coffee table, he got a match—from a hornwort. “The first thing I did was Google ‘what is a hornwort’,” Li, now an evolutionary biologist at the Boyce Thompson Institute in New York, recalls, laughing.

Hornworts are one of the most ancient plant lineages still around today, he quickly learned. They and their bryophyte kin split from other plants, including ferns, more than 600 million years ago, by some estimates. But when Li looked in the genomes or transcriptomes of club mosses and other lycophytes, which are much more closely related to ferns, he found no trace of neochrome, making him suspicious of the result. Perhaps the sample that yielded the putative hornwort sequence was contaminated with fern material somehow, he recalls thinking. But within a few days, his program had uncovered more neochrome-like proteins in hornworts, and a quick phylogenetic analysis showed that the sequences were uncannily similar, confirming that they were indeed homologous to those of fern neochromes. He began to realize he was looking at genes that had jumped from one plant group to the other. “I’m going to have a heart-attack!!” he wrote in an email to his advisors in March 2012. In the end, he concluded that roughly 178 million years ago, some fern took in a bit of hornwort DNA and boom, an entire lineage became poised to radiate rapidly even as flowering plants exploded in diversity and came to dominate the planet’s terrestrial flora.

I just think the lesson is that DNA really gets around.

—Nancy Moran, University of Texas at Austin

If Li had been studying bacteria, finding this kind of lateral or horizontal gene transfer (HGT) wouldn’t have been so unbelievable. Almost 100 years earlier, researchers had first demonstrated that bacteria can pass one another genetically encoded traits. With the advent of DNA sequencing in the 1980s and ’90s, scientists were not only able to see DNA move between microbial species, they also uncovered numerous mechanisms that allow it to do so, providing what’s seen as irrefutable evidence that horizontal transfer plays a dominant role in prokaryotic evolution. But the trading of DNA among complex eukaryotic genomes is a whole other matter, and something “not [thought] possible until fifteen, twenty years ago,” says Antonis Rokas, an evolutionary biologist at Vanderbilt University in Tennessee. 

The story of Li’s research arc is not unique, however. There are now numerous instances of likely HGT in protists and unicellular algae, as well as in fungi, plants, and animals—including vertebrates—and researchers are coming around to the viewpoint that lateral gene flow plays an important role in eukaryotic evolution, too. How these organisms managed to swap DNA with distantly related species is unknown; it could be that a virus or a bubble of membrane ferried the DNA, although there are other possibilities. (See illustration below.) But whatever the delivery vehicle, one thing is clear: in contrast to the simplicity of evolution from a common ancestor as commonly depicted, the branches of the tree of life appear to be inextricably tangled, and scientists are only just beginning to understand the extent of this complexity. 

“As our sampling increases, so does our power to detect horizontal gene transfers,” says Rokas. “So as we sequence more and more and more diverse lineages, I think we’re going to find more and more cases.”

Possible Mechanisms of Eukaryotic HGT

Genetic studies have made it clear that eukaryotic horizontal gene transfer can and does happen. Exactly how, though, remains speculative. For foreign DNA to make it into a eukaryotic cell’s genome, DNA must first enter the cell, then cross the nuclear envelope, and finally insert itself into the genome. Below are a number of proposed mechanisms by which this may unfold.

     Infographic showing possible mechanisms of eukaryotic HGT
© nicoLle fuller, sayo studio


     Number 1


In addition to inserting their own genetic material into their hosts, viruses can pick up and carry genes from the various species they infect and may therefore serve as ferries for HGT. One 2018 study showed that geminiviruses can move host genes from one plant to another.

     Number 2

Extracellular vesicles 

These small bubbles of membrane are known to transport molecules between cells and can carry chunks of DNA. In lab studies, extracellular vesicles have been implicated in the introduction of foreign DNA into cultured cells, including the integration of cow DNA from the fetal bovine serum in culture media into mouse cells.

     Number 3


Microbial eukaryotes that engage in phagotrophy to consume microbes might receive gene transfers from the DNA in their meals. Evidence for this idea is seen in the numerous reported laterally transferred genes in phagotrophic protists.

     Number 4

Endosymbionts or endoparasites

Studies are accumulating that suggest the rate of gene transfers is higher to and from cytoplasm-dwelling species. Malaria parasites, for instance, take in and express human DNA when living inside blood cells. And bacteria that live inside insect cells have transferred numerous genes to their hosts, allowing the parasites to survive with extremely small genomes.

     Number 5

Environmental DNA

Although the mechanisms are unclear, cells are known to pull in external DNA molecules under the right conditions. That means DNA in the environment—from shed skin, mucus, spawned gametes, or other sources—may find its own way into cells and thus act as source material for transfers, as is suspected in the transfer of an ice binding protein between arctic fishes.


     Number 6


Viruses, especially DNA viruses, are often able to sneak genetic material into their host’s nucleus. The 2018 study on geminiviruses found that genes acquired from one host were transcribed in another after infection, indicating the viral minicircles containing host DNA made it into the new host’s nucleus.

     Number 7

Endogenous processes

Studies have found that DNA injected into the cytoplasm can make its way into the nucleus. One way this can occur is through the association with proteins that have nuclear localization signals, such as transcription factors or histones; the proteins essentially drag the DNA with them as they’re ferried through the nuclear pores. Other, as of yet undescribed mechanisms likely exist.


     Number 8


Many kinds of viruses, including retroviruses, have can add DNA to chromosomes—hence their use in certain transfection methods. However, it has not been conclusively shown that wild viruses have inserted host-derived genes into other hosts’ genomes. 

     Number 9

Endogenous repair

DNA in the nucleus may get incorporated during repair processes. For instance, when DNA breaks are induced using CRISPR/Cas9, researchers have found that unintentional additions of DNA can occur.

     Number 10

Jumping genes

Once foreign DNA is in the nucleus, it may get incorporated into the genome with the help of transposons, which excise and insert themselves through well-documented mechanisms. Indeed, recently transferred sequences are often flanked by such jumping genes.

See full infographic: WEB | PDF

Microbes being microbes

The first study demonstrating the transfer of traits between bacteria was published in 1928, before DNA had even been discovered. London physician Frederick Griffith injected mice with living nonpathogenic bacteria and dead pathogenic ones at the same time and saw that the mere presence of deceased bacteria that were once virulent was enough to grant gentler microbes the ability to cause grave harm. In the 1940s, further research confirmed Griffith’s findings and pointed to DNA as the means of trait transfer, and a steady trickle of bacterial HGT papers followed. Then, during the genomics revolution of the 1990s, an overwhelming amount of evidence—including detailed molecular mechanisms explaining how DNA travels between cells—established HGT as a major mechanism of evolution in prokaryotes. 

Still, when the first studies hinting at the possibility of HGT in eukaryotic organisms came out in the 2000s, researchers were hesitant to trust the results, citing numerous barriers that seemed insurmountable. For example, while foreign DNA can easily access bacterial genomes free-floating in the cytoplasm, in eukaryotes it would have to cross through highly regulated nuclear pores to enter the genome’s nuclear home. Also, because eukaryotic genomes are organized into pairs of homologous chromosomes that need to line up properly during meiosis, some researchers argued that large insertions would simply be too physically disruptive and would impede gamete production. 

To this day, the evidence for eukaryotic HGT is far sparser than that for bacterial HGT, and some scientists remain skeptical of its prevalence or importance in the evolution of diverse taxa. But Andrew Roger, a comparative genomicist at Dalhousie University in Nova Scotia, says that those who work with single-
celled eukaryotes have readily embraced the idea. Genomic data—particularly long-read sequencing data, which establishes the genomic context for putatively foreign genes—has been unequivocal, he says. “The more genomes you get, the more you realize: here’s this chunk of chromosome from this organism in this one over here. And it’s just clearly a transfer.” In his mind, he adds, there’s little doubt that horizontal gene transfer has played a large role in the evolution of protists.

As we sequence more and more and more diverse lineages, I think we’re going to find more and more cases.

—Antonis Rokas, Vanderbilt University

Take Pygsuia biforma, for example, which lives in the oxygen-poor, soft mud of Cape Cod’s Prince Cove.  This amoeba-like organism was apparently given a genetic gift from another microbe. Specifically, the single-celled eukaryote acquired a gene that encodes an enzyme key to making rhodoquinone, a molecule that can take oxygen’s place as the final electron acceptor in the production of ATP. Whether Pygsuia biforma got the gene from a bacterium directly isn’t clear, because genetic comparisons revealed that all kinds of hypoxia-tolerant protists possess highly similar genes. These are “completely unrelated lineages that independently adapted to low oxygen,” Roger explains. “So, I think . . . the enzyme came into one of them, and then subsequently, it’s been passed from one to the other.”

Other examples of functional HGT events abound in single-celled eukaryotes. Studies suggest genes involved in digesting complex carbohydrates, surviving icy temperatures, and tolerating extremely salty environments have found their way into protist genomes from foreign sources. Yeast genomes, too, appear to be rife with transfers. “Fungi seem to be a real mess genomically, and clearly, there’s been a lot of gene transfer going on,” says Roger.  

Rokas points to brewer’s yeast (Saccharomyces cerevisiae) as a well-documented case of eukaryote-to-eukaryote gene transmission. Among many commercial applications, the yeast is used in winemaking, where it comes into close contact with microbial contaminants that find their way into barrels. Research has uncovered that three large, fermentation-relevant chunks of the S. cerevisiae genome—more than 120 kilobases in all—were copied from distantly related species of contaminating yeasts. Rokas also has numerous examples from his own research, including a nectar-dwelling yeast that obtained an iron-binding compound from bacteria that live inside insect guts, and a sometimes-pathogenic fungus with dozens of foreign metabolic genes that likely facilitate its flexible lifestyle. Rokas’s team and others have even spotted entire pathways that have jumped organisms, facilitated by fungi’s penchant for grouping genes involved in metabolic pathways together in the genome.

Rokas says that when he and his team first started looking for such whole-cluster transfers a little more than a decade ago, he was skeptical that they’d find any. But his then-postdoc Jason Slot was more open-minded about the idea, Rokas says, and the pair reasoned that if such transfers did happen, they’d be more likely to stick around than single genes or snippets of them, because clusters can transfer novel abilities in one fell swoop. So, the researchers developed a pipeline to look for clusters taken in one fell swoop, and it soon yielded a candidate: a 57-kilobase collection of 24 genes for making the potent toxin sterigmatocystin, which appeared to have been swapped between two filamentous fungi, Aspergillus nidulans and Podospora anserina. “I was surprised,” Rokas says. “I was delighted to be wrong.”

Now, Rokas says, he thinks that DNA is not only moving around the tree of life, but it’s moving a lot more frequently than anyone had imagined, and the cases of HGT that have been identified are just a small sample of the total transfers that have taken place over evolutionary time. “My sense is there is orders of magnitude more integration than retention,” Rokas says. “So there is a lot more DNA coming into fungal genomes, and only a tiny fraction of that DNA is adaptive or selected enough to be maintained.”

Others have pointed out that evolutionary utility isn’t actually required for HGTs to persist. Back in 1998, evolutionary biologist W. Ford Doolittle noted that even if an inserted gene simply acts as a duplicate of an organism’s existing gene, it could take over the current gene’s role. Just like when a gene is accidentally duplicated during replication, he reasoned, it doesn’t really matter to the genome which copy gets retained and which mutates into something new or into oblivion. So chance alone can cause a newly inserted foreign gene to take over the functionality of a gene and be preserved. The more “tries” a gene has to enter a genome, the more likely it is to be retained—an idea backed by the high gene transfer rate from endosymbiont-derived organelles such as mitochondria into the nuclear genome. 

Doolittle detailed this “ratchet” favoring HGT accumulation specifically in regard to microbial eukaryotes and their food. He suggested that for unicellular eukaryotes that eat other microbes, the foreign DNA from the consumed has at least some chance of being incorporated into the eater’s genome, and therefore provides an endless supply of new genes for evolution to tinker with. But as researchers have learned over the past 25 years since Doolittle’s work, there are many other ways that foreign DNA could end up inside eukaryotic cells, and these include the cells of multicellular organisms. (See “Genetics Gifts” below.)

      Cow and snake photos
GENETIC GIFTS: Horizontally transferred genes play significant roles in eukaryotic genomes
© istock.com, DanVostok; © istock.com, Milan Simonovic

Crossing kingdoms

A little more than a decade ago, University of Texas at Austin evolutionary biologist Nancy Moran was studying small, phloem-feeding pea aphids (Acyrthosiphon pisum), when she stumbled upon a very unexpected finding regarding how the insects got their green and red colors. Researchers thought the animals lacked the ability to produce the carotenoid pigments that provided the color, though a variety of animal species were known to sequester carotenoids from what they eat to alter their coloration. Because carotenoids are lipid-soluble, however, they aren’t generally found in the aqueous phloem sap the aphids drink, so Moran didn’t think that was likely, she recalls. Alternatively, bacteria inside the insects could be producing the pigments, but genetic studies by Moran and others had previously failed to find carotenoid biosynthesis genes in the aphids’ endosymbionts, and the pattern of color inheritance didn’t match the matrilineal passage of symbionts. That left the tantalizing option that aphids had somehow evolved the ability to make their own carotenoids. 

As luck would have it, the pea aphid genome had just been sequenced and published, so Moran decided to comb it for genes similar to ones involved in bacterial carotenoid synthesis. Lo and behold, she found matches. And when she then reversed the search to look for published sequences most similar to the aphid genes, the closest matches were in a yeast and multicellular molds. “It turned out even the arrangements of the genes in the genome was the same as in the fungi,” says Moran. “It all made sense, and basically was just a proof” of cross-kingdom HGT, she says.

Moran says she didn’t receive any pushback on her claim of a fungus-to-animal gene transfer—“it’s just a really cut-and-dried case,” she notes. But she says it was initially seen as a “weird thing that happened” until other cases of cross-kingdom transfers in insects and other eukaryotes came to light. The nematode Pristionchus pacificus, for example, appeared to have gotten cellulase genes from slime molds and algae that help them digest biofilm-forming bacteria, according to research from Ziduan Han, a biologist with the Max Planck Institute for Biology Tübingen in Germany, and colleagues. And just last year, Chinese Academy of Agricultural Sciences researcher Youjun Zhang’s team made headlines with its finding that whiteflies (Bemisia tabaci) use a plant gene to detoxify compounds in the plants they consume—the second HGT event to be detected in the species. Now, aphids are just one pixel in an emerging picture of HGT as a regular contributor to evolution in all forms of life, says Moran. “I just think the lesson is that DNA really gets around.” 

Nevertheless, the idea that HGT is a significant contributor to the evolution of eukaryotes remains contentious, says Roger. Particularly “in the animal genome community, there’s a lot of skepticism still,” he notes. Part of that may be the lingering memory of early, bold claims for HGT in animals that turned out to be the result of poor assemblies and contamination, says Li—such as the 2001 Nature paper reporting the first human genome sequence, which asserted that it was rife with bacterial genes. “That kind of put people off,” he says. Still, Li says he thinks “with more and more papers coming out, people are [becoming] more open to this idea.”

Now, the question is shifting from whether HGT occurs in eukaryotes to exactly how common or rare it is in different groups and why. Université Paris-Saclay evolutionary biologist Clément Gilbert is particularly interested in determining the frequency of HGT in multicellular organisms. For a 2020 Nature Communications paper, Gilbert and colleagues sought a broad sense of transfer frequency in vertebrates—a group where lasting transfers have long been thought to be almost nonexistent. By examining the 307 vertebrate genomes that were available on GenBank at the time, the researchers found evidence for more than 975 independent transfers of transposable elements among vertebrates. The vast majority of these were between fishes and other aquatic vertebrates, with relatively few seen in terrestrial animals. That could suggest that the environment or certain lifestyle traits play a role in facilitating HGT: many fishes have external fertilization, so their germlines may simply be more exposed. Gilbert notes that the analyses didn’t look for transfers of other kinds of sequences, or from other sources, so these are likely only some of the HGT events in vertebrates. 

Taking a different tack to examine transfer frequency, Gilbert dove more deeply into the whitefly genome with his colleague Florian Maumus and found evidence for a minimum of 24 separate transfer events from plants that make up the whitefly’s diet, according to the bioRxiv preprint the duo posted in January—a “remarkable” finding, says Gilbert. Meanwhile, they didn’t find any genes from plants in Drosophila melanogaster, insects that also feed on plants but which simply munch on external bits or secretions such as nectar, whereas whiteflies pierce into a plant’s tissues to drink its internal fluids. Gilbert says the results may suggest that certain kinds of interactions facilitate HGT events. Indeed, it tracks that the more closely organisms interact, and the greater the degree of ecological intimacy between species, the more opportunities there will be for the DNA of one to find its way into another. In addition to the high rates of transfer between endosymbionts and their hosts, for instance, there’s some evidence that HGT rates are higher in parasites, which may even play a role in shuttling DNA between species. “There are a lot of questions that can be addressed, and there’s lots of room for many, many researchers to contribute here in this field,” Gilbert says. 

“We have the tools and the data now that allow us to quantify these transfers and quantify the impact they’ve had on evolution of eukaryotic genomes,” he adds. “I think the [whitefly] study that we did prompts us and prompts many others to start conducting such systematic analyses now. And we may come to realize that these transfers are not that not as rare as we thought they were.”

Even if transfers in multicellular organisms are indeed less common than in microbes, Gilbert notes that evolutionary importance is not just a numbers game. “We want to think not [just] in terms of number, but also in terms of impacts. Perhaps just one transfer may have had a huge impact on the viability of some species.”

Moran agrees. “In many cases, [horizontally transferred DNA] seems really central to the ecology or lifestyle of that particular group. Even if it’s only a tiny part of the genome, it can still be a major influence.”