In a paper published in PNAS in 1996, Nancy Moran showed that endosymbiotic bacterial lineages showed faster rates of evolution of 16S rDNA and increased rate of fixation of deleterious mutations, compared to free-living lineages. These results, Moran concluded, were evidence for the Muller’s ratchet phenomenon in combination with mutational bias. Twenty-three years after the paper was published, I asked Nancy Moran about how she got interested in this topic, her memories of carrying out this work, and what we have learnt since then about endosymbionts.
Citation: Moran, N. A. (1996). Accelerated evolution and Muller’s rachet in endosymbiotic bacteria. Proceedings of the National Academy of Sciences, 93(7), 2873-2878.
Date of interview: Questions sent by email on 15th January 2019; responses received by email on 20th January 2019.
Hari Sridhar: I would like to start by asking you about the origins and the motivation for the work presented in this paper. Your early work, including your PhD, seems to have been mainly about aphid-host plant relationships. Around 1990, you started working on aphid-endosymbiont relationships. How did you get interested in this topic, and, more specifically, what was the motivation to do the particular work presented in the 1996 PNAS paper?
Nancy Moran: I had long known about the endosymbionts because I was familiar with a remarkable book by Paul Buchner, a German embryologist who had described and illustrated hundreds of animal-symbiont systems. Buchner and his students did hundreds of studies, mostly before WWII, and his book summarizing these was translated into English in 1965. My advisor, Bill Hamilton, and I would discuss these sometimes, at University of Michigan. I was working on aphids—I chose aphids originally in an effort to study the evolution of sex. Aphids are one of the few animal groups that can alternate between sexual and asexual reproduction. In the course of pursuing this, I learned everything I could about aphids, and ended up with a thesis mostly on the evolution of host plant specificity. But aphids are famous for their symbiont associations; these take up many pages in the Buchner book.
Of course, I did my thesis in the ancient days — before DNA-based information was available for any organisms except a few lab models. As a result, it was very difficult to study endosymbionts, which typically do not grow in lab culture. The only options were microscopy and antibiotics, which are very limited on their own. So I didn’t work on symbionts. After my thesis, I was doing some work on evolution of phenotypic plasticity, another prominent feature of aphids. But phenotypic plasticity was a messy field (in my opinion then and still) and was also before its time, as this was the pre-genomics era, so development was mostly a black box outside of a few model species.
All of this changed, when the phone rang one day in my office at University of Arizona, and it was Paul Baumann, of UC Davis, very excited about using PCR, cloning and sequencing to explore aphid endosymbioses. This was 1988 or 1989, so PCR was somewhat new, and the overall exploration of the uncultured microbial world was taking off, with work from Carl Woese, Norm Pace, etc. This was a huge development in our understanding of the evolution of life, and symbioses are just one part of it.
HS: Stepping back a bit, how did you get interested in biology and research? It is interesting to see that your very early work included species as different as the Indian Honey bee and feral pigeons!
NM: The pigeon work was my undergrad thesis, and my advisor was Nancy Burley, a grad student at University of Texas at Austin, who went on to study sexual selection throughout her career. I was lucky that I had to do a senior thesis, as this got me into research. But birds were not for me, and I had always been attracted to plants and insects, since childhood. However, I also loved evolutionary biology, which I discovered for the first time as an undergraduate, partly in a class by Eric Pianka (who is now a colleague, here at UT). And the insect bug was encouraged in a field class by Larry Gilbert (who is also now a colleague here). So I decided to study evolution, and went to University of Michigan, to study with Richard Alexander in the Museum of Zoology. By luck, the Museum had a semester-long visiting professorship, and visitors in my first three years were John Maynard Smith, Bill Hamilton, and George Williams, all of whom were interested in the evolution of sexual recombination. So I wanted to work on that, and chose aphids. And Bill ended up moving to Michigan and became my advisor, along with Dick Alexander.
Actually I only started work on honey bees around 2007. But honey bees are the pinnacle of social evolution, so of course I knew about their biology, having worked with Bill Hamilton and others interested in social behavior.
HS: Was the choice of the five endosymbionts you used in this study dictated primarily by availability of information at that time? Has this work been extended to other endosymbionts subsequently?
NM: Yes, I used all the cases I could find in GenBank. There were so few data back then. Still those few sequences were sufficient to see this distinctive pattern in the evolution of endosymbiont genes.
And yes, the pattern has since been abundantly extended to dozens of vertically transmitted symbionts, with many whole genome sequences. In fact, some are far more extreme than Buchnera and Wolbachia. The accelerated evolution of coding genes, genome-side, is a diagnostic signal for vertically transmitted, and thus clonal, endosymbionts.
HS: If you think back to the time when you did this work, what are your most striking memories (of lab work, discussions, analysis etc.)? Where was this work done, and would you know what happened to that workspace after you stopped using it? When you moved to Yale in 2010, did you take your study organisms with you?
NM: I worked closely with Paul Baumann for a lot of the work on endosymbionts, starting with that phone call in 1988 or so. We send floppy disks by FedEx to exchange data! We found that the symbionts and hosts showed matching phylogenies, implying an ancient association and faithful vertical transmission over millions of years. Paul was primarily interested in how they worked and what they did for hosts, less in molecular evolution. But a 1993 paper, where we tried to date the origin of the aphid-Buchnera association using aphid fossils and other information, did give results suggesting that the symbionts were evolving faster than related free-living bacteria. However, this was only based on ribosomal RNA sequences. Later, Paul generated sequences for protein-coding genes; these give more power to do tests based on shifts in rate of protein evolution versus rates of DNA evolution. I did some simple analyses, which convinced me that there was something distinctive and interesting going on with these symbionts, as there was a clear shift in their pattern of gene evolution. I wrote the paper by myself, but talked about the idea with a few people, especially a postdoc in another lab, Liz Waters. She was working on heat shock genes in plants and knew molecular evolution literature well. She pointed out that if proteins were unstable (due to deleterious amino acid replacements), then a compensatory mechanism might be to up-regulate heat shock proteins including chaperonin. And up-regulation of chaperonin (GroEL) was a prominent feature of aphid endosymbiont.
HS: Could you tell us a little more about the people you acknowledge – how did you know them and how did they help with this study?
- Baumann: As I mentioned above, Paul was the fundamental reason I got into endosymbionts, and we worked together getting sequences for various genes, and reconstructing phylogenies, way before whole genome sequencing made things easy. He was the first person to use PCR and DNA sequencing to characterize endosymbiotic bacteria that could not be cultured. He did this in 1988 before I was involved. He showed that aphid endosymbionts were in the Gammaproteobacteria, along with E. coli, something completely unknown previously. Paul read everything on symbiosis and was a real microbiologist (which I was not). So I learned a tremendous amount from him. He was actually skeptical about this PNAS paper, as his view was that the aphid symbiosis seemed to work just fine – so why was I arguing that there was a build-up of deleterious mutations! I think later he shifted on this.
- Waters: Liz was a friend and someone who taught me a lot about molecular evolution. Specifically she had the idea that heat shock proteins might be elevated as a compensation for thermal instability of proteins, something that has been supported by a number of later studies that showed that high expression of chaperonin (GroEL) can mask amino acid replacements that are otherwise deleterious. This has been shown in plants, flies, and bacteria.
- Kurland: Chuck Kurland was interested in bacterial genomes and had come through Tucson, where we talked about some of these ideas.
- Seger: Jon Seger was a long-time friend and one of the most knowledgeable people around about evolutionary genetics. We had overlapped in Bill Hamilton’s group back at Michigan, and he was a biology professor at University of Utah. I don’t remember in detail, but almost certainly we talked about these ideas while I was writing the paper, probably at a meeting.
- Normark and B. Sullender and C. von Dohlen: Ben, Barry and Carol were postdocs in the lab. Ben and Carol became professors, at U Mass and Utah State respectively. Both have continued to do research on insect evolution.
HS: Would you remember where and when you did most of the writing for this paper, and approximately how long did it take you?
NM: One prominent condition of my life at the time is that I was a single mother, and had two kids living in the house, 4 years old (my daughter) and 14 years old (my stepdaughter), plus a demanding day-to-day job at University of Arizona where I was teaching. So basically I had very little time, and would write at night after my daughter’s bedtime, in my home office, a sort of screened-in back porch in an old house in central Tucson.
HS: How did you decide to submit this paper to PNAS? What do you remember about the peer-review of this paper?
NM: I was excited about the paper and convinced by the evidence that this was really something new and unrecognized. So I thought it should be in a journal that published papers of broad interest, and I chose PNAS. But at that time, one had to directly ask an NAS member to act as Editor, to handle the review process (this old system was changed long ago; most papers are now submitted directly to the journal). I didn’t know any suitable NAS member, so I randomly wrote, first to Robert Selander, who never replied, and then to Michael Clegg, who generously agreed to do it. And the reviews were very positive, especially one from Mike Lynch (who signed his).
HS: At the time it was published, what kind of attention did this paper get, in academia and in the popular press?
NM: I don’t remember in detail, except that some population geneticists didn’t like certain elements at first, due to my muddiness regarding the distinction between Muller’s ratchet and genetic drift. However, one supporter was Tomoko Ohta, which was nice for me, since I admired her. She has cited the evidence from endosymbionts as support for the view that slightly deleterious mutations are important in evolution (I agree). And numerous others were also very supportive.
HS: Today, 22 years after the paper was published, I would like you to reflect on the validity and relevance of your paper’s main takeaway:
“These observations are best explained as the result of Muller’s ratchet within small asexual populations, combined with mutational bias. In light of this explanation, two observations reported earlier for Buchnera, the apparent loss of a repair gene and the overproduction of a chaperonin, may reflect compensatory evolution. An alternative hypothesis, involving selection on genomic base composition, is contradicted by the observation that the speedup is concentrated at non synonymous sites.”
NM: The general conclusion has turned out to be supported strongly, by massively more data and more cases. Plus, we have an enormous number of additional insights into the process of genome degradation in endosymbionts. Probably the biggest extension is the finding that these genomes are highly reduced in size and number of genes, due to mutations inactivating non-essential genes, which then are deleted over time. This was revealed for Buchnera in 2000, when the first endosymbiont genome sequence was published by Shuji Shigenobu and others in the Ishikawa lab in Japan. We now have many many more cases of this kind of genome reduction in endosymbionts from different bacterial phyla and from other microbial groups.
HS: Given the advances in methods, theory, availability of information since this study was published, if you were to do this study today, would you change anything about it?
NM: Yes, today I would use sequences of several hundred genomes, representing more clades of endosymbionts, instead of a handful of single genes from just a couple of groups. One can get >1000x data for less effort today.
HS: Could you give us a sense of how this work was taken forward, by you or other research groups, subsequently?
NM: Lots of researchers have been studying symbiont genomes for the last 15 years. Some are former members of my lab group, such as John McCutcheon and Colin Dale and Jennifer Wernegreen and Gordon Bennett. But there are many others, worldwide. I think the reason is that there has been so much to discover; really an unknown world was opened up by sequencing and genomics. And a large proportion of animals, especially insects, have these kinds of associations, which turn out to be important in their biology.
HS: Did this paper have any kind of direct impact on your career?
NM: Well, probably because of this paper, I was asked to give a talk at the 1997 Gordon conference on Microbial Population Biology, and there I met Howard Ochman, an evolutionary biologist who worked on bacteria (and still does). We ended up getting married rather soon after. Besides obviously affecting my personal life, this has had a huge impact on my science for the last 20 years.
HS: In the last para of the paper you say: “Endosymbiosis is an evolutionary innovation without which many animal groups would not exist (20). The patterns described for Buchnera raise the possibility that, once “captured,” endosymbionts undergo long-term deterioration due to accumulation of mutations, possibly limiting long-term fitness of hosts.”
Could you give us a sense of what we have learnt in this regard in the years following this study?
NM: Again, we have learned a tremendous amount. We now have whole-genome views, which show how extreme this degenerative evolution has been in many endosymbiont groups. Even at the time, I was attracted to the idea that some host lineages might be dragged down by their captive symbionts as the latter become less and less viable. While hard to prove, this seems even more likely in view of current knowledge. For example, instability of symbiont gene products can limit thermal range, and this can limit the geographic range of hosts, making their populations small and subject to extinction.
Selection on hosts to support their degenerative symbionts can perhaps counter the demise of host lineages. A relatively new development comes from evidence that host genomes evolve to support their long-time obligate symbionts; a particularly interesting finding, from several insect groups, is that sometimes host genomes acquire new genes from bacteria in order to support or control their symbionts. Also, there are now numerous documented cases of ancient degenerative symbionts being replaced in host lineages by newly acquired bacterial lineages that possess robust genomes and capabilities. These cases, which are detected using phylogenetics and genomics, include weevils, aphids, scale insects, spittlebugs, cicadas, leafhoppers, etc.
HS: In the 22 years following this paper, have you ever read this paper again? If yes, in what context? Are there any striking differences in the way you wrote then and now?
NM: I gave it a quick read right now. It’s a little weird to read one’s own old writing! Probably the main change in my style is to write shorter, simpler sentences. This is likely a response to the ever-shrinking attention span of readers.
HS: Would you count this as a favourite among all the papers you have published?
NM: Yes, it’s a favorite, because I felt at the time that it was something new, and something of general importance… and mostly this has held up. The general phenomenon applies to organelles, and other endosymbionts, and more generally to clonally replicated genomes broadly, including Y chromosomes, etc.
And maybe a more fundamental reason I still like the paper is that I had always been interested in the limits of adaptation. Most biologists, and most people generally, are adaptationists; it’s easy to come up with a story for why almost any feature (such as a small genome, or introns, or repetitive element proliferation) is adaptive. But obviously natural selection has its limits: alleles can be fixed due to chance, and some features of organisms and genomes reflect those limits. I was trained by some of the most extreme adaptationists around (Alexander and Hamilton), and I tend to question authority. So, being able to identify limits to selection was of interest to me, and finding where selection falls short, by looking at the DNA level using a comparative framework, felt like a major revelation. Of course some people had been doing these kinds of things in the molecular evolution field, but this field was new to me at the time as it was fairly separate from most of evolutionary biology at that time.
HS: What would you say to a student who is about to read this paper today? Would you guide his or her writing in anyway? Would you point them to other papers they should read alongside? Would you add any caveats?
NM: There have been many extensions of knowledge of genome evolution in clonal organisms and symbionts particularly. Perhaps a somewhat recent review paper, e.g., one I wrote with Gordon Bennett called “The Tiniest Tiny Genomes”. (Ann Rev Microbiol) would be a fairly easy read and would bring a student mostly up to date. Also one remarkable development that represents an extreme case of this kind of evolution is the fragmentation of genomes in cicada symbionts; the cicada cases really illustrate the crazy “rabbit holes” to which obligate symbioses can lead. A 2018 paper giving a good picture of this is M Campbell et al. “Changes in Endosymbiont Complexity Drive Host-Level Compensatory Adaptations in Cicadas” (mBio).