Properly breeding a plant for the traits you want is hard. It requires hundreds of plants and hundreds of cross-breeding attempts between them to obtain desired traits, isolate them, and cross them into more vigorous cultivars with other wanted traits.
That was how it was done for most of human history and it’s a long, slow process. As knowledge of genetics has grown, one of the things we’ve discovered is that plants have far more hidden traits than they openly express in the plant’s phenotype. The problem is that these traits have been, for one reason or another, turned off by the plant cells in a process known as DNA methylation.
How To Methylate DNA
The way this system works is that a methyl group is added to the end of a nucleotide, acting as a kind of signal tag to prevent transcription of that part of the genome. Thus, that gene isn’t expressed by the organism, even if it has the genetic code and overall capability to express it. This control at a higher level than the genes themselves is a part of the system of epigenetics.
The primary form of this methylation occurs on the cytosine nucleotides, though adenines are also possible. For plants, the common three sequence orders that end up methylated are when cytosine is paired with a guanine next to it, when there is any nucleotide in between cytosine and a guanosine, and the last is when the cytosine is followed by two non-guanosine nucleotides. And, yes, I know that covers basically every combination that a cytosine will ever show up in in the genome.
Methylation is commonly used to turn off elements of the genome that would cause disruption to other processes or even possibly cell death if carried out. This includes repeat elements and transposons, but it also includes many useful trait-based genes that ended up turned off and yet could greatly benefit the plant and plant breeders as a whole.
Making Demethylated Offspring
One of the ways we can see these traits expressed directly is by crossing a wild-type plant with a mutant that is deficient in being able to conduct DNA methylation called met1-deficient null mutants. For the null mutants without the option to methylate, all of the genes end up turned on by default.
The mosaically expressed wild-type/mutant offspring of the two are referred to as epigenetic recombinant inbred lines (epiRILs) and feature a mixture of methylation due to the wild-type parent having at least one working copy of the methylation system.
But making the null mutants in order to perform such a cross is difficult, as such a mutant would in most cases die from having lethal genes turned on. That makes this source of study of methylated genes complicated to accomplish and other sources to control the amount of methylation have been sought after. Many techniques have come and gone over the years and plenty have been at least partially successful in their control of gene activation, but as new genetic tools have become available, new methods have come online as well.
Researchers at the University of Georgia have come up with an alternative they have termed epimutagenesis in order to make the epiRILs they need. Working with the model organism Arabidopsis thaliana, the scientists introduced a transgene for the human TET1 protein, an enzyme involved in DNA methylation that converts a methylated cytosine (5-mC) into a hydroxymethylated version (5-hmC) through oxidation. This transgene was added in with the common cauliflower mosaic virus promoter to ensure strong transcription. Whole genome sequencing was then used on two independent plants that had both undergone the procedure.
Overall, their sequencing showed a drop in methylation of CG sequences from 18% of the genome to 9 and 7%, respectively. The other two sequence types of methylations did not see much of a change. An important part of the test was that different amounts of demethylation were seen even with plants that went through the process, meaning it was possible to demethylate different parts of the genome through repetition, allowing eventual access to all of the locked away traits.
To show precisely why demethylation had occurred, they did a more complex global sequencing to check for the presence of 5-hmC, as the original sequencing wasn’t capable of distinguishing between the two states of methylated and hydroxymethylated. They ultimately found no meaningful levels of 5-hmC, implying that the oxidation conversion by the TET1 protein either destabilized the methylation at those locations or somehow marked them in the cellular system to have their methyl group removed. Active removal of oxidized products like 5-hmC seems likely.
The majority (53.7%) of the demethylated areas were in genes, with an additional 7.6% being promoters and the rest being intergenic sequences. Therefore, the main target of the demethylation appears to be genes themselves, which is what the researchers desired from the experiment. The next step was to test and ensure that the demethylation process didn’t interfere with plant growth in some way or at least that the growth matched what happened with the null mutants as well.
Creating Consistent Offspring Generations
The Arabidopsis they grew to maturity did have a delay in moving from the vegetative to the flowering state for reproduction, but this is likely due to the demethylation of a flowering locus and is also seen in the met1-deficient null mutants. They didn’t observe any changes beyond that.
Next, they checked whether the demethylation by the TET1 transgene is kept between generations. To prevent any cross-breeding variables being introduced, the plants were self-pollinated in order to make the T2 generation. This process also allowed for the removal of the transgene in the offspring. Unexpectedly, the T2 plants completely reverted to wild-type, going back to the prior methylation levels and not retaining the changes.
A closer whole genome inspection showed that some of the genes at particular loci kept their demethylation, but most of the genome had remethylation occur, especially with transposons. This implies that there is an active counter process in the plant cells pushing for methylation and, without the transgene, there was nothing to keep the offspring from mostly returning to wild-type. Less than a third of the demethylated sites retained that status in the T2 generation.
They next tested whether overexpressing the transgene in the meristem region would increase the amount of demethylation transmission to the offspring without the transgene. They successfully proved that this was indeed the case, with a significant increase in demethylation retention. With that, they had found a mostly successful method of creating generational transgene-free plants with variable demethylated regions for rare trait expression.
Exposing Hidden Abilities
The primary benefit of this, as noted at the beginning of the article, is that scientists no longer have to generate methyltransferase-deficient mutants in order to produce cross-bred demethylated offspring for trait research. The production of them is a long and drawn out process with many failures and now that is no longer needed.
The creation of variable demethylated plants is perfect for not just traits, but also for improving genetic diversity, since not needing to use breeding selection techniques means that they no longer inherently narrow the gene pool by only going with the most desired plant line. Additionally, the researchers suggest that this epimutagenesis technique could be combined with CRISPR-based systems used to bind to DNA, such as catalytically dead Cas9 (dCas9), in order to more precisely direct which genes are demethylated.
The impact this will have on plant research, agriculture, and more in conjunction with the other genetic tools of marvel in recent years is hard to pin down. But if used to its full potential, we could end up seeing some quite interesting plants on the market in coming years with trait combinations we would have never expected.
Photo CCs: Müürlooga õie stigma from Wikimedia Commons