OK, you just don’t see results like this. A tripling of grain yield in the greenhouse? Translating to a 50% increase out in the field, for real? Are you kidding me?
Something profound has been hit upon here. Not only is it a jaw-dropping result, but it arises from an approach that’s totally different. I think we might be talking about this for a long time.
We’re running out of arable land, but the population keeps growing. We’ll need to get far better crop yields on the land we have in the coming decades if we’re to avert mass starvation.
Lots of people have been busting their butts for decades trying to further improve crop yields by breeding, genome editing, big-data computational approaches, studying the regulatory networks within plants, trying to improve photosynthetic efficiency — you name it — and with only modest success. This has led many to believe that genetic approaches are running out of steam.
So along come Chuan He of the University of Chicago, Guifang Jia of Peking University, and several colleagues. They put a single gene from another species into plants and waltz into a bonanza like this. Their research was published July 22 in Nature Biotechnology.
The picture above is of rice from the greenhouse, but they also tried the same thing in potato, which is not closely related to rice ... and the same thing happened. Here is a comparison from the field:
Chuan He alludes to even more success they haven’t reported yet:
“The change really is dramatic. What’s more, it worked with almost every type of plant we tried it with so far, and it’s a very simple modification to make.”
I don’t mean to say they were merely lucky. They’ve been meticulously studying plants and poring over new approaches for years, and they admit they were surprised by their own result. But as with anything, if you keep at it, sometimes you just get a hold of one….
So let’s take a look at the new approach they used, what’s so different about it, and what it means for future work.
First of all, the “other species” they got the gene from is none other than the humble Homo sapiens. Of course, you can’t develop commercial crops with human genes in them, but that’s not the intention here. The point is that this gene has a function that plants hadn’t quite seen before, and because it’s from a non-plant, the plant doesn’t know how to regulate it, so it’s free to do its thing. This exposed the phenomenon for us, and now we’ve got to figure out other ways to achieve it.
This approach is so different because it ultimately relies on “junk” DNA to achieve very broad effects on the plant. Most approaches have been focused — and quite logically so — on changing specific components of say, photosynthesis, or how sugar moves around, or how flowers form. But here we change many things by changing the very nature of the plant’s chromosomes. Instead of altering one thing, we alter thousands of things, all at once.
So it all starts with the structure of DNA. A gene can’t do anything if it’s not accessible, and in a higher organism like a plant or a human, a lot of DNA does get tucked away, wound around spools called histones. Packed-up DNA is called heterochromatin, while more-open DNA is called euchromatin:
The plant has several ways to pack and unpack DNA like this, and it does so according to its own agenda. DNA is negatively charged, so it repels itself, but histones are positively charged, and they can ease this repulsion and allow DNA to compact together. One way to open up the DNA is to stick acetyl groups (basically vinegar molecules) onto the histones. That makes histones less positively charged and allows the DNA to self-repel some more and opens things up:
A lot of DNA information gets copied to RNA. Then a lot of that RNA goes off to other parts of the cell to be translated into proteins or do other things. But some of that RNA just hangs around in the nucleus near the DNA and doesn’t have any obvious function. It’s very stable and hangs around for a long time unless it’s actively degraded. It has repeated sequences that don’t seem to code for anything. It’s made from what we think of as “junk” DNA, which is interspersed throughout the genome for no apparent reason. This “junk” DNA, though, makes up about half of our genomes.
It was realized in 2014 that if you take this repetitive RNA away from euchromatin, it collapses into heterochromatin. That was elaborated upon in 2019 with the realization that it’s simply the negative charge of RNA that allows it to counteract the positive charge of histones and open up DNA. I’ll try to pick one figure from these pretty extensive studies that shows this clearly.
On the left side (-) we have human cells with a stain that lights up DNA; heterochromatin shows up especially well. On the right side (+) are the same kind of cells that got perforated and loaded up with DNase I, an enzyme that eats DNA. In the top row, not surprisingly, DNase I eats up the DNA, and we don’t get much staining on the (+) side.
But in the second row, we add RNase A (an enzyme that eats RNA) first, then we add DNase I. When we get rid of all the RNA in the cell this way, suddenly DNase I can’t eat the DNA anymore, because it’s densely packed into heterochromatin, as you can see by the bright staining on the (+) side. The third and fourth rows are pretty interesting, because there we eat up all the RNA first, but then we put some RNA back in. The funny thing is that it doesn’t even matter whether we add human RNA or E. coli RNA. Once we do that, the DNA opens up again and is able to be eaten by DNase I. So, to convert heterochromatin to euchromatin, literally any old RNA will do. We just have to have lots of it hanging out in the nucleus near the DNA.
I said earlier that this “junk” RNA will hang around in the nucleus for a long time unless it’s actively degraded. One way it does get degraded is if methyl (—CH3) groups get attached to it. The plant has enzymes that come along and add these methyl groups to RNA in whatever pattern the plant decides is appropriate. Then other enzymes look for methylated RNA and chew it up, again however the plant decides is appropriate.
The human gene that got added to these plants, FTO, takes these methyl groups off of the RNA, specifically when they’re attached to adenosine. (RNA, like DNA, is a chain made up of four different kinds of bases, and adenosine is one of them.)
FTO’s removal of methyl groups makes the RNA more stable so that it can persist longer. FTO seems to prefer to do this to the repetitive RNA that hangs around the nucleus. More RNA in the nucleus for a longer time generally means more-open DNA, and more genes being activated. Plants can remove methyl groups from RNA, but they don’t have any gene that looks very much like FTO, and FTO’s pattern of preference seems to be different from that of plants. There’s just something serendipitously very good about it.
When Drs. He, Jia, and the others took a look at what genes were active in their superplants, they found that over 11,000 of them were turned up! Eleven thousand! It’s like, just make more of everything! It’s an unconventional — even kooky — thing to suggest as a crop yield approach, but there it is.
They found that the plants weren’t taller, and that the rice grains didn’t look any different, but that there were just a lot more of them, and more tillers (grain-bearing branches) per plant. They also found that the plants converted CO2 to sugar faster, presumably to keep up with all these new mouths to feed. A lot of times when you succeed in making plants with more tillers or seeds or whatever, you don’t end up increasing grain yield because many of the seeds that do form get aborted. Didn’t happen that way here.
As if all of this weren’t interesting enough, there is more to say about FTO. In humans, it is strongly linked to obesity. Certain mutations in it lead to higher body mass index (BMI). When you turn up the level of FTO in mice, they attain greater weight, regardless of the diet they’re fed. FTO in humans and mice appears mostly in the brain. In the hypothalamus, which controls food intake, its level is turned down when essential amino acids are absent. So, FTO appears to be a signal that amino acids are plentiful, and hence that the body should use its available resources to add new body mass.
Not only that, but the cell growth thing goes a little further. FTO abundance is linked to gastric cancer, acute myeloid leukemia, and glioblastoma. Cancer cells are ones whose growth is out of control, and lots of FTO is an enabler of that. It must crank up a lot of genes that are associated with central metabolism; that is, associated with adding mass to cells.
So it sort of makes sense that FTO could act as a signal for plants to grow as well. You wouldn’t necessarily expect it to have a similar effect in organisms as radically different as humans and plants, but its universality makes it awfully compelling, wouldn’t you say?
I ought to close with some kind of hopeful statement that sums all of this up, but I’ll leave that to UChicago professor Michael Kremer, who took home a 2019 Nobel Prize in economics for his work in fighting global poverty.
“This is a very exciting technology and could potentially help address problems of poverty and food insecurity at a global scale—and could also potentially be useful in responding to climate change.”
I think this can indeed be a significant part of the solution. We just need to understand it a little better. OK, plant scientists, that means the challenge is in front of you. Let’s get it!
■□■□■□■□■
And as a hat tip to our friends in Chicago, here’s Sufjan Stevens with “Chicago”.
All things grow, all things grow…
