In a paper published in Nature in 1999, Anurag Agrawal, Christian Laforsch and Ralph Tollrian showed that non-lethal exposure of Daphnia and wild radish to their predators, in addition to inducing defences, led to production of offspring that were better defended than those from unattacked individuals. This study was one of the first to find strong evidence, in both plants and animals, for suc a “transgenerational effect” or “maternally-induced defence”. Seventeen years after the paper was published, I spoke to Anurag Agrawal about how this study came about, his memories of doing these experiments, and what we have learnt since about transgenerational induction.
Citation: Agrawal, A. A., Laforsch, C., & Tollrian, R. (1999). Transgenerational induction of defences in animals and plants. Nature, 401(6748), 60.
Date of interview: 8th September 2016 (via Skype)
Hari Sridhar: Correct me if I’m wrong, but it seems like you did the work presented in this paper towards the end of your PhD. What was your motivation to do this piece of work, in relation to the rest of your PhD?
Anurag Agrawal: You’re right on the mark that it was at the end. When I started getting interested in induced response to herbivory in wild radish, I came up with the idea about transgenerational effects, and it seemed like a bit of a fringe or crazy idea, so it was on the back-burner. What got us to do the experiments is pretty interesting. There’s two pieces of the story I suppose. One is that I had read an obscure paper in The New Zealand Journal of Agriculture that reported that oilseed rape plants, when damaged by aphids, produced more defensive glucosinolates in the seeds. And that was a problem (for human consumption) because glucosinolates in the oil being pressed from the seeds gave it a bad flavour. This gave me the idea that there might be some transfer of information from what the parent plant experiences to what goes into the seed, which henceforth might shape the seedlings’ defences. The other thing is that there was an undergraduate student who came along named Joel Kniskern who was interested in this just as an initial research project. When you have interested undergraduates around, it’s an opportunity to try out some of the more risky ideas, and I proposed to him several different projects. This was the one that he chose. So I think this little bit of literature information as well as this undergraduate’s interest motivated us to do the initial experiment, and I replicated it and then it went from there.
HS: How did this group of three authors come together?
AA: That’s an interesting story as well. And you know I didn’t meet either of those authors for many years until after the paper was published. So what had initially happened is that I had the idea, Joel was on board, we were doing the preliminary experiments –at the same time I was writing a chapter for a book on inducible defences. Ralph Tollrian, who is a co-author of this paper, was one of the editors of that book, and I knew at the time that Daphnia were some of the most well-studied organisms for inducible defences – they had been studied for decades, much more so than plants. And so at the beginning of our work on the plants – on the wild radishes – I emailed Ralph Tollrian. I’d already been in contact with him about our book chapter – and said ‘Dear Ralph, we are starting these experiments on maternal effects or transgenerational induction in wild radish plants; have you heard of anything similar in animals, especially Daphnia’. I just thought there might be a part of the literature that I was unaware of. He immediately wrote back and said, ‘you know, I haven’t heard of any such things but let me look into it, I’ll get back to you’. Six months later he wrote to me, ‘Dear Anurag, I have a bit of an awkward situation that I want to mention to you. After I got your email about transgenerational effects, I mentioned it to a student in my lab [Christian Laforsch] and he’s been doing experiments for the last couple of months and he’s finding the kinds of effects that you had suggested to me’. His question at the end of the email was, ‘what do you think we should do about this?’I really appreciated the email and honesty, but it was not clear how to proceed. We had exchanged ideas by email, and he had started a research project based on that, but felt a little uneasy about just going off on his own to publish that, given that the idea came from our conversation. So at that point I suggested to Ralph that that we might put our work together, and perhaps it will make for a stronger story because the effects are shown for the first time in an animal and plant. That was just really serendipitous, but to me I think it points to two really important things, and they have to do with knowing the literature. It was this obscure paper that got me a little bit excited, and then it was in discussions with colleagues, looking for other literature and ideas from other kinds of systems, which eventually led us to put it together.
HS: Do you remember how you came across the paper in the New Zealand journal?
AA: I don’t remember. You know, at the time, maternal effects were quite, I’d say, popular, but kind of a fringe area of ecology and evolution, and there were several excellent reviews written in the mid-eighties about maternal effects in plants. It’s possible it was referenced in one of those.
HS: When Ralph Tollrian wrote back to you, were all the experiments already done – experiments with the wild radish and with the Daphnia? Was it just a matter of writing the paper from there on?
AA: I don’t exactly remember, but I don’t think so. We replicated the wild radish experiments after the initial work – and certainly Ralph had increased the replication in his lab, and we ended up doing a bunch of analyses, but I’d say it wasn’t long after that. We also did some chemical analyses of the seeds that were done after our initial experiments.
HS: Was the term “transgenerational induction” used for the first time in this paper?
AA: I would guess that transgenerational induction is used for the first time in the paper, but that ‘transgenerational’ as a term referring to cross-generational effects was certainly in use before. It’s interesting, in the abstract of the paper we coined the term “maternally-induced defence” but that’s not a term that has that has caught on in any way, it has not persisted in the literature.
HS: Did the work that you did for this paper go into your PhD?
AA: It did, it was a chapter of the thesis, yes.
HS: I just wanted to sort of step back a bit and also ask you about your PhD. Can you tell us a little about what your motivation was to work on this topic for your PhD?
AA: Sure. I had arrived to Rick Karban’s lab in the fall of 1994 and he was just completing a book co-authored with Ian Baldwin on induced responses to herbivory. Before I even started graduate school – the summer before – he had sent me a print-out of the manuscript. That was really an incredible source of inspiration. The book wasn’t published until 1997,but I had the manuscript version of both Rick Karban’s summary of the field and also his view of what the most pressing questions were. As an early graduate student, I had read the book a couple of times, and Rick had mentioned a few pearls of wisdom. One of his pearls was that if you’re interested in fitness, if you study an annual plant where you can measure lifetime reproduction, that will be much more impactful than if you work on a perennial plant. That was very simple advice, but also very powerful. Until that point, there were many studies of induced response to herbivory, but relatively few that had worked on annual plants with lifetime measures of fitness. Also, reading his book, it was very clear that although Rick himself was interested in the community ecology of induced responses, that hadn’t been very well-studied. And so as part of my thesis, I did experiments using four different herbivores and their impacts on the plant’s induction – two specialists and two generalists – and I think just having that replication and that community ecology approach was really beneficial. And then lastly, Rick was always, as an adviser, somebody who was very interested in and advocated novel or different ways of looking at systems in biology. So I think this transgenerational research seemed exciting because it was a kind of biology that hadn’t been well-explored.
HS: If you don’t mind my asking, how come Rick wasn’t an author on this paper? I’ve done quite a few interviews now, and I find this was quite common around that time but it’s not so common today, so could you tell us a little more about that?
AA: Yeah, that’s a good question and it’s a question that we got asked at the time that the study was published as well. There were some reporters and other people that asked why wasn’t your PhD supervisor a part of this study. Something you said is really true – there’s been a historical and temporal change in authorship guidelines, especially in our field of ecology and evolutionary biology. There was a time when most PhD papers were single authored by the people doing the work. Currently, most PhD-related papers are not single-authored for two reasons. In general our work is much more collaborative, so there’s often the adviser, but also other people as co-authors. The science has changed; it’s much more interdisciplinary. Also, PhD theses currently maybe a little bit less independent from advisors that in the past. One of the things that Rick Karban said to me early on is that it would be strange if we didn’t co-author some papers together during my PhD time. I think what he meant by that was he wanted to be involved in some of the projects – intellectually involved, not just providing money or edits on the manuscript, but to come out into the field or to come up with some ideas together. Indeed, he and I co-authored several papers together. But he didn’t impose his name on all the papers. In fact, at least some of the papers that I gave to him, as a manuscript to read, had his name on the author list. He would sometimes cross his name off. I think that was an amazing gift, but also a way of training to do some projects together and some projects independently. The last thing I’ll say about this is in relation to graduate students of my own. Here at Cornell University, most of my graduate students – not all of them – have had a paper or two that don’t have me as a co-author, and those are projects that were largely done independently. So I try to use the same thing that Rick taught me – to do some projects together, but not to necessarily put my name on all papers coming out of the lab.
HS: Why wild radish? Was Rick already working on this plant?
AA: No. In fact, I struggled a lot in my first two years of graduate school on the particular questions to ask and the particular systems to study. I had many failed attempts, but in my second year, I started interacting and collaborating with Sharon Strauss. She’s currently a professor at UC Davis, was an assistant professor at the time, had been working on wild radish, and she had developed that as a system for studying plant-herbivore and plant-pollinator interactions over the previous decade. I had asked her about this as a possible study system and she was very encouraging. She and I collaborated on a project together that was published in 1999 in Evolution on the costs of induced responses to herbivory. This was the first project I had done on wild radish, but then after that study, Sharon allowed me to continue to work on it. She continued to work on radish as well, and she and I continue to collaborate in different ways, but it was a tremendous amount of generosity. After our first project together, she gave me a lot of freedom to do what I wanted with it.
HS: Were the greenhouse and the population in the greenhouse established by her?
AA: Well, it’s a little bit complicated; there’s a really interesting history there. The original population of seeds was actually collected in the mid-1980sby Jeff Conner, who is currently at Michigan State University. Jeff had collected those for part of his post-doc work, here at Cornell University with Sara Via! The seeds were collected in Ithaca, New York, which is kind of amazing! Jeff and Sharon had then been collaborating in the late 80s and early 90s.The first set of experiments that I did with Sharon was basically with seeds that were descendants from that Ithaca population. We did our original study – that was that 1999 paper in Evolution – and those were then the basis for the seed lines we used to study transgenerational effects. So the seeds have an interesting history that traces back to different people at different institutions.
HS: Would you know what happened to this set up after you moved on? Is it still being used for experiments?
AA: Well, Jeff had collected the seeds from the field in Ithaca, and then had primarily been doing greenhouse experiments on heritability and selection. Some field experiments were done in Illinois by Sharon and Jeff, and some field experiments were actually done in California by Sharon and me. Since I’ve moved to Ithaca, I’ve certainly seen wild radish populations in the field, although it’s not very common. Although I’ve thought a lot about Jeff collecting those seeds in the mid 80s, we actually haven’t been able to go back to that specific population.
HS: What about the greenhouse in Davis? Is that still being used for experiments?
AA: I don’t know. The last time I was in Davis those greenhouses still existed. Some of them had been upgraded. The seed lines still exist in all kinds of places -Jeff has seed lines, I still have some seed lines that I haven’t used for a while, and Sharon must have some as well. Some of the greenhouse structures are, I guess, still there.
HS: Are they on the university campus?
AA: Yes they’re on campus, but on the edge of campus. We used to bicycle back and forth between our academic buildings and the greenhouses.
HS: That’s something I wanted to ask you about. During the time when you were doing these experiments can you give us a sense of what your daily routine was like? Which year were the experiments done?
AA: I can only assume they were done in 1998, probably, early in the year. There was probably about a year of experiments and then the paper was published in ‘99. I think that graduate school is some of the most free and exciting periods of our academic lives. There are two things that stand out to me about my daily routine in Davis as a graduate student. One, I got up every day and decided what I wanted to do that day. Some days it was spending hours in the library looking through the literature; other days, it was experiments all day; other days it was going to seminars and discussion groups. I was very much immersed in all aspects of the research – reading, writing, hands-on work, and discussions with other people. It was just something I did all the time. I lived with people that were doing science and that was really exciting for me. The other thing that stands out about that particular time – and I’d say, to some extent, I’ve encouraged my students to try to do things that are similar – is to always have my hands in some experiments most of the time across the year. In other words, we would do greenhouse experiments and laboratory analyses in the winter and we would do field and greenhouse experiments in the spring and summer. So I really had that as a strong part of my graduate school time, doing experiments year-round.
I’ll just also add that Davis was a really exciting place because, at that time, there were seminars that we would go to three out of five days of the week. There was entomology, ecology, population biology, behaviour, plant sciences – they each had their own seminar series and there were a lot of discussions that were going on among the graduate students and faculty; a lot of interaction.
HS: Where did you get the caterpillars for this experiment from? Were they wild-caught caterpillars?
AA: I honestly can’t remember for that experiment, but my guess is yes. The caterpillars were cabbage white butterflies or Pieris rapae and we were wild collecting them each year. Any of the Brassica that are cultivated or wild would frequently have Pieris rapae on them. I myself went out and caught some, and there was also a butterfly ecologist at Davis – Arthur Shapiro -he was always generous, and would often come back from the field and give us butterflies for the colonies.
HS: One thing I want to do is to go over the acknowledgments to get a better sense of how these people helped. Could we do that?
We could try, I don’t know if I’ll remember everybody, but let’s do it.
HS: You thank the plant-herbivore group at Davis, can you tell us a little more about that?
AA: Yeah, we had a weekly discussion group that was organized, I think, by Sharon Strauss and Rick Karban. At the time, we just got together every week and discussed papers or our own research on plant-herbivore interactions. I’d say it was mostly focused around ecology and evolution, although there were aspects of chemistry or agriculture that were part of those discussions. The one thing I would add is that we’ve developed that and continued it at Cornell University. There are about seven or eight labs that are active in plant-herbivore interactions at Cornell, and we have a weekly discussion group that we call PIG -Plant Interactions Group. We meet every week, there’s 30 or so people that attend; it’s a spectacularly interactive group which came out of what was going on at Davis in the late 1990s.
HS: The next name in the acknowledgements is W. Gabriel.
AA: That’s Wilfried Gabriel, he is a Daphnia Biologist that was a colleague of Ralph Tollrian is in Munich.
HS: M. Morra and V. Borek.
Yes, they were two people I was collaborating with at the University of Idaho. There’s very little chemistry reported in this study, but they had been helping with some glucosinolate analyses of the radish seeds.
HS: RJ Mercader
AA: He was another undergraduate at the time. He participated in some of the experiments, along with help from Joel Kniskern.
HS: You also thank people for comments on the manuscript. Can we go over that list as well?
HS: LS Adler.
AA: So that’s Lynn Adler. She’s currently a professor at the University of Massachusetts at Amherst. She was a graduate student colleague at the time working with both Sharon and Rick.
HS: S Diehl.
AA: Yeah, I’m pretty sure that is Sebastian Diehl. I think he is a Daphnia biologist that was suggested by Tollrian.
HS: And then, H. Dingle.
AA: Hugh Dingle is somebody that studied insect life-history evolution. He was a professor at Davis. I’d asked him to read the manuscript.
HS: RE Lenski.
AA: That’s Richard Lenski. He studies microbial evolution at Michigan State University. He was spending a week at Davis at the time, giving a workshop. It was probably a little bit presumptuous of me, but because he was spending several days at Davis, I had asked him to read the manuscript to provide comments.
HS: TW Schoener.
AA: Thomas Schoener. He is a well-known community ecologist at Davis. He’s now, I think, close to retirement. I just asked him to read the manuscript and I remember he made several important suggestions. He was the one that suggested calling this‘maternally-induced defence’ in the abstract
HS: JS Thaler.
AA: Jennifer Thaler. She was a graduate student colleague at Davis. She’s also a professor at Cornell now, she’s an entomologist and happens to be my wife. She read and commented on the paper
HS: DA Thiede.
AA: Denise Thiede. She was a postdoc at Davis, at the time, working with Maureen Stanton, and was an expert in areas of ecological genetics, and really understood maternal effects. So, I asked her to read and comment on the manuscript
HS: TG Whitham.
AA: Yes, that’s Thomas Whitham. He is at Northern Arizona University. He was just a senior ecologist that I’d gotten to know from meetings and showed a lot of interest. I had asked him to read the manuscript.
HS: You mentioned earlier that you didn’t meet your co-authors until later. When did you eventually meet them?
AA: You know, frankly, I don’t know if I’ve ever met Christian Laforsch. We’ve definitely corresponded by email. I don’t know where he is these days. Ralph Tollrian, I met at least once but probably a couple of times. In either 1999 or 2000 we were in Wageningen together, in the Netherlands, at a meeting or some kind of a conference. It was just a real pleasure to meet him and talk.
HS: Can you can you tell us a little about the writing of this paper itself. Do you remember how long the writing took and where and when you did most of the writing?
AA: Boy, you ask the hard questions. I don’t remember, I believe it was in Davis. If I had to characterize two aspects of my writing, I would say that, on the one hand, I write relatively fast, in terms of trying to crank out a manuscript quickly. But then, I tend to be patient and edit things very iteratively. That’s why that list of acknowledgements is long in terms of who read the paper. My guess is that I’d asked each one of those people to read it one at a time. I didn’t send it to all of those people at once; I was kind of working away and improving the manuscript by getting comments from somebody, and then editing and working on it, sending it to the next person. I tend to be kind of patient in getting a lot of comments and improving and honing the study. That’s something I still do.
HS: Would you remember how you discussed and shared drafts of this paper with your co-authors? Was it all over email, or were there Skype and telephone conversations as well?
AA: We were not using Skype, and we were not on the telephone; it was all by email. We didn’t have a Google doc or a working folder either. We were just passing it back and forth as attachments.
HS: Did this paper have a relatively smooth ride through peer review? I’m guessing Nature was the first place you submitted this?
AA: If I remember correctly, we submitted it to Science first, and it didn’t go out for review. Then we submitted it to Nature and it went pretty quickly. One of the reviewers was Erkki Haukioja,and I learned of that only because he then wrote a commentary in Nature and revealed that he was one of the reviewers. He titled his commentary ‘Bite the mother, fight the daughter’ which I thought was very clever.
HS: At the time the paper came out, did it attract a lot of attention, both within academia and in the popular press?
AA: I think there was a little bit, yes. A week or two, probably 10 phone calls, often articles in the newspaper or in other places talking about how it was similar to, or reminded people of Lamarckian inheritance. That kind of thing. Although, academically, maternal effects had been studied some, I think what made this paper more interesting is that the maternal effects were associated with mediating interesting ecological interactions.
HS: What kind of an impact would you say this paper had on your career and the subsequent trajectory of your research?
AA: Having a paper in Nature never hurts one’s career. I would say that in terms of early career work, this was a door-opening paper. It gave me some credibility in terms of thinking about new avenues of research frontiers. As an assistant professor at the University of Toronto I did two or three more projects on transgenerational induction in wild radish. I think that was important for my career because when I started a new job, I was, on the one hand, working on new and different projects, but it was helpful to work on an ongoing project. The other thing is whenever we start research projects and work on them for a while, sometimes they go dormant. A full 10 years had gone by when I hadn’t done any work on it. Then, I had a Swiss postdoctoral fellow in my lab named Sergio Rasmann, who came to me and said he’d like to do some experiments on transgenerational induction with Arabidopsis. Initially, I was quite discouraging for various reasons. But, I was very glad that he persisted because, using the power of Arabidopsis, with all of the mutants that we have now, he led a collaborative effort with Georg Jander’s lab as well as folks from Penn State including Gary Felton. It is a very interesting story, because a similar thing to what happened with Daphnia and radish happened here too, but this time with tomato and Arabidopsis. I think that opened the door to a whole wave of studies that came out right around that time, using Arabidopsis, and using more mechanistic epigenetic approaches that has revived interest in transgenerational induction. So,yeah, I think it has affected my own research trajectory. We’re doing experiments right now in broccoli plants to look at transgenerational induction as potentially an agricultural approach to more resistant crops
HS: This papers been cited over 600 times. Do have a sense of what it mainly get cited for, i.e. and whether the citations are specifically for the results, or like many early papers in a new research area,is it sort of used as a general overview citation?
AA: I think some of both. I haven’t really looked at all the citations and how many there are and what they are being cited for, but I think the fact that it was early and had showed the effects in an animal and plant, makes it a useful paper to cite to say these kinds of effects of predators or herbivores occur and can be common. But, certainly, I think some of the papers that have come out since then have built up that work, both in aquatic invertebrates like Daphnia and in plants, and I assume that in that case there’s some of the specific results that are getting cited.
HS: Today, it’s 17 years since this paper was published.Would you say that the main conclusions from this paper still hold true,more or less? If you wereto redo these experiments today, would you do anything differently, given the advances in technology, in our knowledge, in the techniques available for analysis etc.?
AA: Yeah,I think the results hold, and that’s nice to be able to say. My sense is that they were preliminary experiments, but due to my own work on radish for a couple years after that, as well as work other people have done in other systems, this seems to be a general phenomenon.
I’m not sure what I would do differently. I mean, I think there was a lot of serendipity at the time, e.g. getting in touch with Ralph Tollrian as we talked about earlier, that’s not something you can really plan; it just happened. But certainly, chemistry was a small part of the early experiments. In my laboratory now, we’re doing much more =chemical analyses. I suppose we could do some more chemistry. There are measures of methylation that you can get, in terms of the genome, asking whether you can see signatures of epigenetic effects. But that’s not where I would start personally. I’ll probably still start with something pretty similar, but the standards for publishing something like that would be higher; it would require more diverse kinds of data and perhaps something more definitive.
HS: In the paper you have an unpublished observation: “Herbivory on the maternal plant also induced changes in the number of trichomes per leaf in seedlings”. Wasthis published subsequently?
AA: I think in a paper in 2000 in Ecology, we certainly looked at maternally-induced trichomes. My guess is that we didn’t use the same data, but an expanded data set.
HS: Towards the end of your paper you say that several other systems are strong candidates for having maternally-induced defences. Do you know if people have gone out and looked for and found such defences in these systems?
AA: Definitely yes. Those were all systems where there was some hint that there might be something like that going on. At the time I had ideas about seed size in plants – should large seeded species or small seeded species be more likely to show maternal effects? I don’t know whether those have been tested, but certainly things like Arabidopsis, which have tiny seeds have been tested. So yeah, I think people have gone on and looked at other systems, including some vertebrates. Maternal effects are everywhere, and the epigenetic basis has now been revealed in several systems.
HS: Since the time this paper was published, have you ever read the paper again?
AA: Since it was published certainly; in the last 10 years, I don’t think so! There have been times when people have asked for data or clarifications on things, for which I’ve gone back and read it.
HS: If you compare this paper to papers you write today, do you notice any striking differences?
AA: I don’t think so. Because of being published in Nature, it has a very narrative kind of style – there’s methods at the end and interspersed throughout; all of that. So, from that perspective, I rarely write a paper like this. But I don’t think that my actual style, in terms of how to approach writing, has changed that much.
HS: This might be a difficult question to answer given the number of papers you have published but would you count this as one of your favourites?
AA: I suppose one of my favourites, sure. It stands out I think in two ways to me: one, in the two different systems that we ended up using to test the same question; that’s just kind of neat. Two, it stands out as being an early demonstration of something that people are much more interested in now, epigenetics. So from that perspective, it’s kind of cool, because we more or less were bumbling around looking for new and different things to do, we stumbled on something and got together this with Ralph Tollrian’s group .It was less than complete, but more than ordinary.
HS: What would you say to a student who is about to read this paper today. What should he or she take away from this paper? Would you add any caveats?
AA: Yeah, I think I would say, as a means to contextualize where the field was, flip through Karban and Baldwin’s 1997 book. That was the state of the field then, and you can then see how and why this paper was an advance at that time. Another thing I would say is to read a couple o fthese newer papers in the last three or four years that have been on transgenerational induction, just to see how the field has changed. In a way, the original paper is very singular in its contribution.It is a proof of concept showing that this stuff can happen in these two groups of organisms. On the one hand, we’ve come a long way in terms of understanding the mechanisms; on the other hand, we haven’t come that far in terms of understanding how important these effects are for population dynamics and evolution. So, yes, I think the important thing is to try to contextualize where we were at that time and where we are now.