Treasure Your Exceptions: An Interview with 2017 George Beadle Award Recipient Susan A. Gerbi

Susan A. Gerbi

The Genetics Society of America’s George W. Beadle Award honors individuals who have made outstanding contributions to the community of genetics researchers and who exemplify the qualities of its namesake. The 2017 recipient is Susan A. Gerbi, who has been a prominent leader and advocate for the scientific community.

In the course of her research on DNA replication, Gerbi helped develop the method of Replication Initiation Point (RIP) mapping to map replication origins to the nucleotide level, improving resolution by two orders of magnitude. RIP mapping also provides the basis for the now popular use of λ-exonuclease to enrich nascent DNA to map replication origins genome-wide. Gerbi’s second area of research on ribosomal RNA revealed a conserved core secondary structure, as well as conserved nucleotide elements (CNEs). Some CNEs are universally conserved, while other CNEs are conserved in all eukaryotes but not in archaea or bacteria, suggesting a eukaryotic function. Intriguingly the majority of the eukaryotic-specific CNEs line the tunnel of the large ribosomal subunit through which the nascent polypeptide exits.

Gerbi has promoted the fly Sciara coprophila as a model organism ever since she used its enormous polytene chromosomes to help develop the method of in situ hybridization during her PhD research in Joe Gall’s lab. The Gerbi lab maintains the Sciara International Stock Center and manages its future, actively spreading Sciara stocks to other labs. Gerbi has also served in many leadership roles, working on issues of science policy, women in science, scientific training, and career preparation.

An abridged version of this interview was published in the December 2017 issue of GENETICS.

How did you get involved with the March for Science?

As scientists, we can enjoy doing science because those of us fortunate enough to have research grants from the NIH and NSF receive them through tax dollars. So, we have an obligation to share with the public what our science is about. It’s important for scientists to learn how to speak to the public because the worst thing we can do is speak in such technical terms that their eyes will glaze over and they say, “this is why I didn’t want to study science in the first place!” Of course, this has always been true, but it seems especially true in the current era. There seems to be a disregard for science as a methodology. I was really spurred on by [GSA President] Lynn Cooley at the fly meeting, where she challenged me when she was presenting me with the Beadle Award. She mentioned that I had played a role in public policy through the American Society for Cell Biology [ASCB] and through FASEB [Federation of American Societies for Experimental Biology], as well as through the AAMC [American Association of Medical Colleges]. And then she said, “we need you now!”

I went home and I thought: yes, the field needs people to be actively involved in public policy at this particular time in history. So, with some difficulty, I found the local leaders for the March for Science in Rhode Island and then played an active role in mobilizing the Brown community. The underpinning of the March for Science was applauding the importance of science and the scientific approach. I enjoyed that experience and I really thank Lynn for her challenge to me, as well as her inspirational writings about how the community of geneticists really needs to be vocal.

Even though the March for Science itself was amazingly successful, it must go on beyond that. We need to speak to our congressional representatives, we need to speak to the general public, and to our neighbors about what we do, why it’s exciting, and why it’s important for advances in our society.

What inspired you to become a scientist?

Years ago, there was a study reported at the ASCB [American Society for Cell Biology] that found that many prominent women scientists have their fathers as their role model—that was certainly true for me. My father was a physician scientist. He grew up in Italy and went to medical school there, but he also was involved in renal hypertension research. During World War II he came to this country and was ultimately affiliated with Columbia University College of Physicians and Surgeons. When I was a youngster he would bring me to lectures at the New York Academy of Sciences, which was terribly exciting. I would be learning about things in high school biology and then would get to hear talks by the people making the discoveries. Holley spoke about the structure of tRNA, for example, and Palade about ribosomes, and Nirenberg about cracking the genetic code.

What drew you to studying chromosomes?

I became interested in chromosomes in high school after reading a Scientific American article by J. Herbert Taylor who had discovered that replication of chromosomes was semiconservative, which temporally paralleled the discovery by Matt Meselson at the DNA level. Then when I went to Barnard College I had the opportunity to take a molecular genetics course with Herb Taylor, and that confirmed my interest in chromosomes and replication. I knew I wanted to do a PhD on chromosomes, and one of the emerging leaders in the field at the time was Joe Gall. I was planning to apply to the University of Minnesota, where he was at the time, but I learned he was going to join the faculty at Yale, so I applied to Yale. The rest is history, as they say!

It was a fortuitous time to be in his lab because the method of molecular hybridization had just emerged from the work of Spiegelman, where radiative probes are hybridized to DNA captured on nitrocellulose filters. It was a no-brainer to try to expand that to the chromosome level. Joe Gall went to a meeting in South America where several scientists brainstormed about how they might best apply this method. They all went home to their labs and got hung up on the controls. But Gall, being a fabulous biologist, said he was going to use a system where he knew what the biological answer should be and then he would work things out from there.

He and my fellow grad student Mary-Lou Pardue worked out the initial method of in situ hybridization. They used the stage of meiosis in Xenopus oocytes where you find thousands of nucleoli, and everything in the field pointed to the fact they contained amplified genes for ribosomal RNA, and indeed that turned out to be the case. The next step was to apply the method to chromosomes themselves rather than amplified nucleoli, and I was part of that effort. We did the first in situ hybridization to chromosomes using the gigantic polytene chromosomes from the salivary glands of the lower dipteran Sciara, as well as Drosophila. Sciara polytene chromosomes are a bit larger than those of Drosophila because they have a few more rounds of endoduplication.

How did your interest in ribosomes begin?

The probe we used in the in situ hybridizations was ribosomal RNA labeled with tritiated uridine, and we used Xenopus rRNA because it was available from tissue culture cells. I wondered how Xenopus RNA could hybridize to fly chromosomes. I thought there must be some sequences that have been retained during evolution, and that started me on the long path of studying eukaryotic ribosomal RNA using evolution as a guide.

So, we reasoned that sequences with strong functional consequences should be evolutionarily conserved. We started with Xenopus rRNA because it was the first eukaryotic gene ever cloned. By hybridization we found there were regions of conservation even between bacteria and eukaryotes. Then we produced the first rRNA sequence from a metazoan. We modeled the secondary structure of rRNA using principles of compensatory base changes—where base-pairing in hairpin stem regions would be retained even if the sequence changes—and we found that there was a core structure that was conserved between Xenopus, yeast, and E. coli.

Our subsequent studies found that eukaryotic rRNA is larger because of insertions that were highly variable in sequence length and nature. We called them expansion segments, and initially people thought they were a remnant of evolution and didn’t have any function, but current studies by John Woolford making mutations in yeast and by other groups doing X-ray crystallography and cryo-electron microscopy are starting to zero in on whether they may indeed be playing functional roles.

There are now an enormous number of rRNA sequences that have become available across the three domains of life. We did a bioinformatic study and confirmed that there were some sequences that were universally conserved, and in addition we discovered a new category: sequences that are fully conserved within one domain of life, such as eukaryotes, but not present in that sequence composition in the other two domains. That points to the possibility that they carry out a domain-specific function. Intriguingly the majority of them line the tunnel of the large ribosomal subunit through which the nascent polypeptide exits. Whether it plays a regulatory role feeding back to the nearby peptidyl transferase center is something worthy of future study.

What can we learn from understanding Sciara re-replication?

DNA re-replication leading to gene amplification is a hallmark of many cancers, but the underlying mechanism isn’t fully understood. Whether re-replication is an alternate or a primary mechanism that subsequently leads to breakage and rejoining and recombination hasn’t been studied. One cannot induce amplification in cells in a way that allows you to study the initiating events; you only see the final outcomes of amplification. So, it became very desirable to look for model systems where this is a natural part of development.

There are two known cases of developmentally programmed locus-specific re-replication: Drosophila follicle cells, and salivary gland polytene chromosomes from the end of Sciara larval life. We want to understand how these origins of replication bypass normal cellular controls. Once we figure that out, this may serve as a paradigm to understand whether the same thing is happening in cancer cells.

What is the function of developmentally programmed re-replication?

The areas that undergo re-replication in the Sciara polytene chromosomes are called DNA puffs (to contrast them from Drosophila RNA puffs). The DNA puffs have undergone extra rounds of replication, and are templates for a massive amount of transcription that is translated into the proteins needed to make the pupal case in the next stage of development.

In both Sciara late larvae and in Drosophila follicles there’s a very short window in which a massive amount of protein is needed. In Drosophila it’s for the chorion that forms the egg shell, and in Sciara it’s for the pupal coat. And so the strategy in both systems is gene amplification. You might ask why other cell types don’t use the same strategy. The problem is that once you’ve undergone re-replication you now have nested replication forks and a structure called an onion-skin that is potentially very unstable when the cell tries to divide. But in both Sciara polytene chromosomes and the polyploid cells of Drosophila follicle cells there is no mitosis, so the onion-skin structure is not damaging. In addition, both tissues are destined to be destroyed soon after the re-replication event, so they wouldn’t have to live with the consequences anyway. If such onion-skin structures occur in dividing cells—such as in the cells that become cancerous—this might lead to breakage and recombination and eventually lead to amplification.

What have you learned about re-replication?

The first thing we had to do is to understand what an origin of replication looks like at the sequence level. This has been a very elusive target for the replication community because no specific origin of replication sequence has emerged for any organism except for budding yeast. Other organisms seem to have initiation zones rather than point origins, and no specific sequence. We developed a method that we called Replication Initiation Point (RIP) mapping. This was done with Anja Bielinsky, who was a postdoc in my lab. We needed an enriched population of newly replicated DNA to start with, and for this we popularized the use of the enzyme λ-exonuclease. This will digest DNA from its 5′ end in an exonucleolytic fashion, but not if there’s an RNA primer at the end, such as there is after re-replication. We first piloted the method using SV40 and then using yeast ARS1 [an origin of replication]. The ARS1 structure was very well established, but it wasn’t known whether there was a specific nucleotide where initiation starts, or whether it involves a larger area. We were able to show that indeed DNA replication begins at a unique start site. We were then able to identify where the Sciara DNA puff re-replication starts at the nucleotide level. There too we saw a unique start site for synthesis, even though there’s an apparently larger initiation zone seen by 2-D gels. Maybe at each end of the initiation zone there are preferred sites to start DNA synthesis, and that gives rise to the appearance of a zone.

Once we established where DNA synthesis starts in re-replication, we could look at the surrounding sequence and see if anything jumped out at us that might be a regulatory element. Directly adjacent to the start site, where the origin of replication complex binds, we found a potential binding site for an ecdysone receptor. This is the master regulator of insect development, and it was the first transcription factor ever discovered. We’re trying to test whether it is also acting as a replication factor. If so, the question is whether —in hormonally sensitive cancers such as breast cancer—the estrogen receptor might also serve as an amplification factor.

What role do you think the ecdysone receptor might play in re-replication?

We don’t have direct evidence it is a replication factor, only smoking gun evidence. But there is some precedence for transcription factors also acting as replication factors in certain animal viruses. In the case of Sciara we imagine a couple of scenarios. One possibility is that the ecdysone receptor is interacting with some of the replication machinery that’s sitting adjacent to it on the chromosome, keeping it in an “on” state. Another possibility of course is that it’s acting only as a transcription factor and triggering a cascade of events that lead to re-replication. The difficulty is you would expect this to impact all origins in the genome, not specific subsets. In Sciara there are 18 DNA puffs; what distinguishes them is still a mystery, but to me it suggests there’s something in the local environment—either at the sequence level or the chromatin level or in neighboring proteins such as the ecdysone receptors.

You are a great advocate for Sciara. What’s so compelling about this species?

Sciara is an amazing model organism with many unique biological features. Geneticists usually figure out how things work by making mutations. But, if you will, the unique features in Sciara are like God-given mutations; they are variations of canonical processes that can shed light on the underlying mechanism.

Around 1914 Charles Metz decided to study Sciara for his PhD thesis at Columbia. He captured it in the pigeon house at Cold Spring Harbor Laboratory on the suggestion of a friend. It took him quite a number of years to figure out the chromosome mechanics, but he ultimately succeeded and ended up dedicating his career to studying Sciara.

In the 1930s geneticists had a meeting at Cold Spring Harbor and realized that they would make more progress if they all worked on the same organism. They discussed which to choose, and the two finalists were Sciara and Drosophila. We all know who won! The reason Drosophila was chosen was because geneticists of the 1930s relied on making mutations by X-irradiation, and Sciara turns out to be extremely resistant to X-irradiation. This is another of its unique biological features, but it was not good at the time. Sciara surfaced again in 1970–71 when Sydney Brenner spent two years in the library trying to figure out a good model system for developmental biology. Sciara made his final shortlist of six organisms, but the winner of that competition was the nematode worm C. elegans.

Fast forward to the current time, and of course now we don’t have to rely on X-irradiation for mutation. Thanks to genome sequencing and other methods there’s been an explosion of emerging model systems where one can now reap the benefits of studying unique aspects of biology. In our lab, a senior staff member, Yutaka Yamamoto, has succeeded in developing a germline transformation method for Sciara. Moreover, former graduate student John Urban has sequenced and assembled the Sciara genome. Thus, we’ve established the toolbox of a genome sequence and a methodology to transform Sciara, so the time is now ripe for the scientific community to study all the unique features of Sciara. I’ve been trying to encourage other lab groups to start to work with Sciara. I’m thrilled that several labs have already started and others are on the horizon. We give a 1–2 day workshop in my lab for anyone who wants to learn how to work with Sciara.

What are some of the unique features of Sciara?

One is sex determination. There’s no Y chromosome, and sex is determined by the mother. There are two types of females: those with two copies of the X have only sons; those with one X and one X’ (an X with a long paracentric inversion) have only daughters. How that works is if the haploid egg came from an X/X’ mother, then the egg (when fertilized) will become a daughter. Whereas if it’s an egg from a mother that was X/X, the fertilized egg will become a son. Something, possibly in the cytoplasm, is conditioned by the mother at an early stage prior to meiosis when the X’ and X separate.

Spermatogenesis is also unique. In the first meiotic division in males there’s a monopolar spindle. Monopolar spindles have been studied in cases where they’re induced, but in Sciara it’s a normal occurrence. The chromosomes move from what looks like prophase, skipping metaphase, directly into an anaphase-like configuration, and ultimately telophase. What’s remarkable is in the anaphase-like configuration all the paternally-derived homologs move towards the nonpolar end of the spindle. That’s instead of it being random whether a maternally derived homologue will go to one pole end of the spindle or the other. This was the first example of imprinting, where the cell can recognize the paternal or maternal origin of chromosomes. It was noticed by Helen Crouse, who worked with Metz for a few years, and she coined the term imprinting in her 1960 GENETICS paper. It was later studied much more in-depth in mammalian systems, but it is not yet clear whether in Sciara it occurs by modification systems, such as methylation, as it does in mammals.

So, all of the paternally derived homologs move away from the single pole and are then discarded in a little bud of cytoplasm. In a way, this is a system en route to parthenogenesis because—at least in sperm—it’s not using the paternally-derived chromosomes of the previous generations. The chromosomes that move towards the single pole are maternally derived, and of course, how chromosomes move to this pole is a fascinating subject that is worthy of study in itself.

Then in meiosis II a bipolar spindle is established, though there’s only a single centrosome at one end, the one that came from the previous monopolar spindle. So now the chromosomes do align on a metaphase plate and then segregate, with the exception of the X. The X instead stays locked into the single centrosome, and the result is two products of meiosis II: one is nullo-X and the other has two copies of X (the X dyad). The nullo-X product is also encapsulated in a small bud of cytoplasm and degenerates. So, the only product of spermatogenesis is a single cell that has two copies of an X and is haploid for the autosomes. At fertilization, you have one X from the egg and two from the sperm, and the zygote ends up with three copies. But, of course, you can’t keep doing this every generation! You would accumulate more and more X chromosomes. So, in an early cleavage division some of the X chromosomes are eliminated.

And this is where sex determination comes into play: if the offspring is going to be male, it eliminates one of the three Xs; if the offspring is going to be female, it eliminates two of the three Xs. The final chromosomal complement in the soma of males is a single X, and females are either X/X or X/X’. Now imprinting comes into play. The eliminated Xs are always paternally derived. The X chromosomes that will be eliminated line up on the metaphase plate and start to separate—in that their centromeres disjoin and start to be pulled to one pole or the other—but the arms of the Xs fail to separate. So, it’s as if there’s a chromosome-specific failure of the cohesins to dissolve.

It turns out there’s a region that was genetically identified by Crouse that she called the controlling element (CE). It governs the X dyad nondisjunction in meiosis II, as well as the X chromosome elimination in embryogenesis. You can move the CE locus to any of the three autosomes by reciprocal translocations, and now you’ve fooled the cell into treating the autosomes as if they were the X: The translocation autosome will undergo non-disjunction in meiosis II and chromosome elimination in early embryogenesis, and the X that now lacks the controlling element no longer undergoes those unique behaviors.

What is the controlling element and how does it regulate these processes? We’d like to know more. The controlling element is located within the tandem array of 50 copies of ribosomal RNA genes—it’s right in the middle of the array and is flanked by translocation breakpoints. So, we would like to be able to zero in on it with long read sequencing and terrific genome assemblies. We know already know there is some non-rDNA sequence within the tandem array which in itself is interesting – then the question is what part of that is functional and how does it function. Is it, for example, like the XIST locus, which creates an RNA that coats the entire chromosome? That’s one hypothesis because the controlling element acts on the chromosome on which it’s sitting.

In addition to the sex determination mechanism and the unusual behaviors imparted by the controlling element, Sciara also has germline-limited chromosomes called the L chromosomes, whose roles are totally unknown. And, in addition, Sciara has locus-specific re-replication in DNA puffs of polytene chromosomes and other unique features.

Who have been your most important mentors?

Joe Gall without a doubt, and I’m still in very close touch with him. He’s most important cell biologists of our generation. He was always fascinated by the biology uniquely offered by particular eukaryotic species, including less well-studied organism, and that was one reason I went to study with him. He wasn’t wedded to one biological system, but being such a well-rounded biologist he would ask, what is the best biological system to study the question at hand? Rather than the other way around, which is, I have this biological system, now what questions can I ask with it? I think that’s what makes him quite remarkable and unique. A sequel lesson learned from that is it’s safest if we do experiments in a biological context—rather than try to dissect everything through test-tube biology or even cells in culture. I’ve always tried to do experiments in the biological system itself because then you’re less likely to have changes in unknown parameters that will give you the wrong answer.

What’s the best advice you ever received?

My colleague at Brown Art Landy once gave me some advice when we were worried about being scooped. He said if you know the answer to the experiment you are doing, you can jump ahead to where that was going to take you and plan the next experiment. If others arrive at a conclusion that you trust, then you can simply fast forward to the next logical question.

What advice would you give to younger scientists?

Treasure your exceptions. Sciara is an exception to the way things normally happen, but it can give you an enormous amount of insight into the basic canonical mechanisms that are shared by most other organisms. If you get a result in the lab that is unexpected, don’t throw up your hands in despair and say, things aren’t working, and I must have done something wrong, and this is not what the field would have predicted. You may in fact have opened up a whole new line of pursuit! After you repeat it and do the appropriate controls, it could change the mindset of the field and let people know that the current hypothesis or model might need some tweaking.