They mighty hornwort. Click to enlarge this picture. Trust me, you want to enlarge this picture. You can see the spores inside the horns up close. CC-by-2.0. Gorgeous photo by Jason Hollinger. Click image for license and source.

Scientists knew neochrome was odd before they started rooting around in its family tree. A union of independent proteins -- red-sensing phytochrome and blue-sensing phototropin -- the super-protein combines two already-great pieces into one fantastic whole that helps plants grow toward dim, filtered light. It's a slick little adaptation for living on the murky forest floor, a way of acknowledging: yeah, I know I'm not the tallest tree in the forest. But I'm gonna work with what I've got, and I'm gonna look d--- good doing it.

Until recently, neochrome was known from two places on Earth: a single species of green alga that is -- well, there's no nice way to put this -- pond scum, and a group of shade-loving ferns. But the pond scum and the ferns are separated by at least 400 million years of evolution, and maybe much more. That is a problem.

It's a problem because only three things could explain this, and all of them seem unlikely. First, neochrome was present in the common ancestor of the pond scum and the ferns, but has been lost from all other plants except for these two. Second, neochrome evolved independently in these two groups. And finally, the pond scum neochrome gene somehow got sucked into the genome of the ferns.

This last option is called horizontal gene transfer (HGT), and although bacteria have elevated it to an art form, the rest of life on Earth has always seemed averse to swapping genes with strangers. Extra-species DNA is like a box of chocolates . . . and, well, you know the rest.

As it turns out, hypothesis #2 was true, but the story was stranger still. Fern neochrome was not invented by ferns. They stole it. Just not from the pond scum.

Several clues hinted this. Within the ferns, neochrome is widespread but not universal. Two early-evolved orders of fern -- the Osmundales and the Schizaeales -- lack it entirely. But it's abundant in the Cyatheales and the Polypodiales. These later two groups evolved after the emergence of forests dominated by flowering trees with light-hogging broad leaves at the end of the Age of Dinosaurs and in the aftermath of The Asteroid. Scientists have speculated neochrome may have helped these ferns prosper in Earth's new, presumably darker forests. But where did the ferns get their neochrome? No one knew.

A team of scientists from Europe, North America, and China led by Fay-Wei Li at Duke University set out to find out. They scoured the genetic material of 474 plants and algae representing the spectrum of plant life. The scientists searched for and compared the DNA sequences of neochrome, phototropin, and phytochrome. When compared, the pattern of mutations in these genes tell a story about who is related to whom, and approximately when their ancestors parted ways.

Green algae neochrome had indeed evolved independently from fern neochrome. But unexpectedly, neochrome turned up not just in ferns and pond scum, but also in a quirky little group of plants related to mosses called hornworts. That meant neochrome had both evolved independently twice, but that it had *also* been subject to horizontal gene transfer. Fern neochrome had evolved from hornwort neochrome, and the computer programs that compared the sequences calculated that the evidence for this is very strong. At some point before being incorporated into neochrome, hornwort phototropin lost all its non-coding introns -- portions of the genetic code usually excised when making proteins from DNA -- and this loss of introns is plainly visible in fern neochrome as well. Though this isn't the only evidence for the transfer, it is very strong evidence indeed.

In this figure from the paper, you can see, at top, the structure of fern neochrome, and at the bottom, a family tree of fern neochrome made by comparing its sequence to the genes of neochrome, photoropin, and phytochrome in several other groups of plants.

Fig. 1 from Li et al. 2014. Click image for source.

Hornworts are plants so obscure that even I have never seen one, though that may be due more to my own ignorance than the fact I've never crossed one's tiny, creeping path. Like mosses and liverworts, they belong to a group of land plants that evolved early in Earth's history. These plants lack the sort of robust internal plumbing that the rest of land plants possess (botanists call this "vasculature", and plants that have it are "vascular plants" or "tracheophytes").

There are only about 100 known hornwort species. Superficially, their bodies resemble liverworts, named for their vaguely liver-shaped, lobed bodies. But unlike liverworts (who have their own menu of bizarre reproductive structures), hornwort spores are dispersed from long, tall horns that split from stem to stern. Inside are fibers called elaters that flex in response to changing humidity (I wrote about structurally different but functionally similar horsetail elaters earlier this year). Their contortions help boot the spores from the nest when the time is right.

A hornwort photographed from overhead reveals its liverwort and lichen-like appearance. Horns are upper-center. CC-by-2.0; Photo by Jason Hollinger. Click image for license and source.

The genetic evidence suggested that neochrome jumped from hornworts to ferns around 133-229 million years ago, some 200 million years after previously published estimates of the split between the last common ancestor of ferns and hornworts. Since horizontal gene transfer would by definition have taken place after the groups diverged, these dates are consistent with the HGT hypothesis.

Botanists are realizing that horizontal gene transfer between plant species is not so rare as has been previously thought. The cox1 "homing intron" has wormed its way from a fungus into the genomes of plants more than 1,000 times. It doesn't produce any functional difference in the plant, though. Neochrome does, and seems to be the first important example of this phenomenon, but perhaps not the last.

Carl Zimmer, in a column at the New York Times, somewhat provocatively -- and rightly, I think -- points out that this casual attitude toward uploading foreign DNA may be little different to what modern genetic engineers do when they transfer genes from one organism to another to make a genetically modified organism (GMO). Though many have decried this practice on the grounds it is "unnatural", it was plainly natural and useful to these ferns. So natural and useful, in fact, that they've made it a habit.

When the scientists constructed a family tree of neochrome genes found in ferns and compared it to a family tree of ferns based on the rest of their DNA, they got another shock: the trees looked completely different. If a single ancestral fern had acquired neochrome and then passed it on to its descendants in the usual (vertical) way, you would expect the two trees to be the same. But they are not. Here they are -- and I apologize for the size of the text. The colors correspond between the two sides and should help you comprehend the image:

Fig. S6 From Li et al. 2014. Click image for source.

Apparently, ferns -- even distantly related ones -- have been swapping neochrome like it's a social disease. That would account for the scrambled inheritance pattern of neochrome that you see at left above.

How might this be possible? The cox1 "homing intron" -- which has itself wormed its way from a fungus into the genomes of plants over 1,000 times -- seems to be aided by sequences called transposons that promote its movement between plants. Transposons are pieces of DNA that contain genes for enzymes that enable the sequence's own extraction and re-insertion at another location. Usually, that new location is the same organism. In this case, that new location may be a completely new species.

Li et al. speculated that neochrome may be associated with its own transposon-like mobile elements that promote movement between species. If that were the case, it would make this particular gene much more likely than other DNA to get around.

But there may be another reason ferns are especially suited for cross-cultural gene swapping. Ferns lead dual lives. When you see a frond-bearing fern in the forest, what you do not see is its small, diminutive twin. This twin -- the gametophyte -- is the form of the fern that contains one copy of all its DNA, while the fern you know has two copies (just like you and me). All plants maintain this dual life cycle, called "the alternation of generations". But only in ferns and one other plant group called lycopods are the two plants capable of living independently.

The lower half of the green thing in this picture is the tiny, gauzy fern gametophyte -- the fern you don't see:

Image by Vlmastra Cc-by-3.0. Click image for license and source.

Sprouting from it is a new frond-making fern. This baby fern grew from an egg fertilized by a fern sperm that had to swim over the plant or be launched by a falling rain drop to get there. Prior to fertilization, the structures that make sperm and egg lay naked and exposed on the little plant's body. Gametophyte ferns often grow in the same moist, low-profile habitat as plants like mosses and hornworts, and probably often find themselves living with each other on intimate terms. Mosses and hornworts, like ferns, also make swimming, expeditionary sperm that must negotiate the outside world to locate their targets. With the fern germline so exposed and in such close contact to comparable structures of mosses and hornworts, the authors hypothesize, who knows what crazy things could happen?


Li F.W., Villarreal J.C., Kelly S., Rothfels C.J., Melkonian M., Frangedakis E., Ruhsam M., Sigel E.M., Der J.P. & Pittermann J. & Horizontal transfer of an adaptive chimeric photoreceptor from bryophytes to ferns, Proceedings of the National Academy of Sciences, DOI:

CC-by-2.0. Gorgeous photo by Jason Hollinger. Click image for license and source