“Karma” means a lot of things to a lot of different people. For Rob Martienssen, a pioneer in epigenetics and professor at Cold Spring Harbor Laboratory, it refers to a specific transposon – a DNA sequence that can change its position within a genome – that can mean the difference between a plentiful crop and ten years of wasted labor.
Some of the biggest names in plant science have guided Martienssen throughout his career, from his Cambridge University doctoral mentor David Baulcombe – who, with Andrew Hamilton, discovered small interfering RNA – to his postdoctoral mentors, William Taylor and Michael Freeling, at the University of California, Berkeley. As a junior faculty member at Cold Spring Harbor Laboratory, Martienssen even had the chance to work with Barbara McClintock, whose discovery of mobile genetic elements won her the Nobel Prize in 1983.
Today, Martienssen, who has received several NSF awards for his work, is building on that legacy as a professor at Cold Spring Harbor and Howard Hughes Medical Institute. Martienssen recently discussed his lab’s research on epigenomic modification of Karma transposons in the African oil palm in a BIO distinguished lecture titled, “Germline reprogramming and epigenetic inheritance: How to avoid BadKarma.” Afterward, he spoke to us how epigenetics shapes the world around us and just what “bad Karma” means to a plant scientist.
OAD: What is “epigenetics”? How is it different from the genetics we learn about in high school biology?
A: The concept of epigenetics actually has a really long history. Throughout most of the Middle Ages there was this controversy about developmental biology: you could either imagine a germ cell as being very naïve and having to be programmed to make the next generation, or in those days, there was also this idea that maybe there was a homunculus in the sperm that was fully formed and simply had to grow into a baby. Aristotle was very much in the former camp and William Harvey used the term “epigenesis” as a way to talk about this programming that happened to very naïve germ cells to allow them to become a new body – and of course, that’s how we think about development now.
Then in the 1940s, Conrad Waddington wrote a famous article and subsequently a book on something he called “epigenetic landscapes.” He went a bit further with this idea, saying that there was this underlying genetic program, but that depending how it was interpreted in every cell, you would get different fates for those cells. So, this was the idea that you don’t access all the information in the genome in every single cell, you only access some of it – and that was determined by epigenetics. He was using the term because it was sort of superimposed on genetics – it was “above” genetics.
OAD: Did he have a concept of how we interpret epigenetics today?
A: He’s often regarded as the father of epigenetics – he really did have a good idea of what was going on. He had done experiments in Drosophila where he’d selected different wing shapes over multiple generations and was able to select new forms without making mutations – or without making mutations that he could readily identify as a geneticist. So, he really did get it, and was the first person to propose this sort of transgenerational inheritance being based on genes.
At the same time, Barbara McClintock and Alexander Brink were maize geneticists working in the 1940s and 50s, and they actually came up with real examples of traits controlled by genes affecting plant color – like the color of kernels on a corn cob – that were under epigenetic control. They could go in one direction or another from one generation to the next, and Barbara was convinced that all of this was controlled by her “controlling elements” – transposable elements. So those were really the first definitive examples of “transgenerational” epigenetics.
OAD: So, epigenetics led to changes in these traits – wing shape and corn kernel color – without any mutations in the genes that encode them. And even though the change wasn’t genetic, it was able to be shared from one generation to the next. Do we know how this happens?
A: With the discovery of DNA as the genetic material and the composition of chromosomes over the next several decades, mechanisms that might explain this started to come forward and that’s how we really think about epigenetics today. These mechanisms are ways of modifying the chromosomes without changing the DNA sequence. Primary among these – at least in plants and mammals – is DNA methylation, which is very widespread. You can methylate and demethylate DNA chemically using enzymes in the cell, and you can also just replicate DNA without methylation and that will remove methylation passively.
This turned out to be one of the major mechanisms of epigenetic inheritance, but it wasn’t the only one. Chromosomes are not only composed of DNA but also histones, which are the proteins that DNA is wrapped around – you can think of it like a spool of string or wool – and those histones can also be chemically modified in a reversible way. Those modifications are just as important as the ones on DNA, but they have to be associated with specific genes, and so how these modifications only end up on specific genes has been a major part of research in the last few decades.
OAD: What are the prevailing theories for how that happens – how does the cell know where to put those modifications?
One of the things I was involved in about 15 years ago was the realization that some of that instruction about where these modifications were made – both DNA methylation and histone modifications – was through RNA, and in particular small RNA, which David Baulcombe discovered in plants. It’s not the only way – there are lots of DNA-binding proteins that also instruct the cell where to make these epigenetic modifications – but RNA turns out to be very important. More and more we believe that in plants and in some animals like C. elegans, this RNA can actually pass from generation to generation, and could actually have this transgenerational epigenetic effect.
OAD: Much of your work focuses on the African oil palm. Why that plant specifically? What are the “real world” ramifications of the questions you study?
A: The African oil palm is propagated by cloning, and this was done starting in the 1970s. There are some very good reasons for that – it allows you to clone elite germplasm [living seeds or tissues kept as genetic resources for breeding plants or animals] without having to breed it any further, because the clones are all supposedly the same and they’re all going to have the same elite properties. But it turns out that epigenetics raises its ugly head, and in fact these clones are not identical. The reason they’re not is because of transposons that lose DNA methylation in the cloning process – and we’re trying to understand why that happens. When they lose methylation in the cloning process, the next generation has really nasty phenotypes that are really important economically – they develop abnormal flowers and dry, shriveled fruit that yield much less oil.
The reason the oil palm is so important from this point of view is that it only grows in very sensitive parts of the world where it competes with rainforest, and if you transplant an oil plant from the nursery into the plantation, you have to do that before it fruits. It isn’t until 10 years later that you realize there’s a problem – that it has fruit that won’t yield any oil because of this epigenetic change – and you’re sort of stuck, so the temptation is to burn down a bit more of the rainforest and plant a bunch more. Now for big palm oil companies, they can afford to not do that, but for a small holder, it’s a different matter altogether. This is a major problem in Malaysia and Indonesia, and so I’ve been very fortunate to work with the Malaysian government, with the MPOB – the Malaysian Palm Oil Board – and some biotech companies in the U.S. – and I should say in full disclosure that I’m a founder of one – for the last ten years to figure out what was going on.
OAD: What does “bad Karma” mean in this context?
A: We discovered that there was indeed a single transposon that’s responsible for this phenotype. It’s in a gene that is very well known in other plants – and I should say if we hadn’t had all of that basic research in Arabidopsis [a flowering plant used as a model organism in plant biology], we would have no idea what this gene actually does. We were able to develop a very simple test that can predict the phenotype of palm trees that are cloned. The test is based on a transposon called “Karma”, so we call it “good karma” when it’s methylated and you have normal fruit, and “bad Karma” when it’s unmethylated and you have this horrible phenotype. It’s a nice story that goes from really basic principles in epigenetics in model systems to a real-world consequence.
OAD: You work in a field that has – like many others – evolved rapidly over the course of your career, due in no small part to significant technological advances over the last few decades. What is it like to work in your field now versus when you were a graduate student at Cambridge?
A: I’m sure everyone has stories like this, but I got my Ph.D. based on about 1.5 kilobases of DNA sequence, and it’s just amazing when you think about it. The idea of doing a whole genome wasn’t really seriously discussed until the 1990s, and now an individual graduate student or postdoc can easily knock off a genome. It’s amazing. Having the genome sequencing projects¬ has really changed how we do epigenetics and that’s been very exciting. We used to have to grind up a whole plant to get enough DNA to do anything, and now you can literally look at a single pollen grain and really get a good idea of what the epigenetic and genetic makeup of that pollen grain is.
But at the same time, it’s interesting that the same questions are still there and despite all of this fabulous technology, we don’t have a unified concept of what epigenetics really means. And I think that’s very exciting – that’s why we do it.