Donald G Comb Lecture - Jo Handelsman

Sal Russello:
I am Sal Russello and I have the great pleasure of serving as just the third CEO here at New England Biolabs, my predecessor, Jim Ellard, who is here today, served in the role for 17 years and his predecessor was none other than our founder, Donald G. Comb, who served in the role for more than 30 years. Today we honor Don's enduring legacy with the Donald G. Comb Memorial Lecture. I'm pleased to welcome all of you here in Ipswich and those of you attending the lecture through our live stream. I'd also like to welcome several members of Don's family, many of whom are in the audience today, and several participating on the live stream. Don sadly passed away in 2020, and as we bring new people to NEB, fewer and fewer of us have had an opportunity to work directly with him. As such, these kinds of events become more and more important as they are an opportunity to reflect on his contribution to the life sciences, science advocacy, philanthropy, conservation, the arts and the environment.

Those of you that knew Don well knew him to be courageous, independent, unbelievably generous and fearless. I'd like to show a short memorial video to give all of you some perspective on Don. After that, Tom Evans, our executive director of research, we'll talk to you about this year's lectureship recipient, Jo Handelsman. As you'll be able to tell from the background of today's speaker, an accomplished microbiologist, science advocate and supporter of diversity in the life sciences. Don and Jo would've certainly had much in common and much to talk about. With that, please show the video. So first of all, Don, I want to thank you on behalf of everybody here for a few things. I first want to thank you for your love of science because you continue to inspire us.

Don Comb:
We purified the enzymes. These are called restriction enzymes, DNA in very specific ways. This was a time when your proteins were just getting started, and so these were very important enzymes, so I knew that. So we started isolating bacteria from all over the world with nice little enzymes. Now we have, I think 230 or so different research enzymes recognize different sequences of the four bases, the DNA sequences and cut the DNA so people can cut out a piece and clone it and it's very exciting...

Sal Russello:
I then want to thank you for your courage because if you hadn't had the courage to start going to the Biolabs, who knows where all these people would be right now? We wouldn't be here and I'm sure we wouldn't be as happy. Was the connection it seems between science and art, can you talk a little about that?

Don Comb:
I guess they're both trying to say something and science is trying to tell you the truth, are trying to present what they feel is the truth.

Sal Russello:
You're feeling that companies can kind of co-exist with this feeling of being environmentally conscious? You've done such a good job with the environment.

Don Comb:
Well, you certainly hope that that's the direction all companies will take. That's the direction they have to preserve biodiversity and they have to preserve the environment biodiversity. That's our main task.

Sal Russello:
Thirdly, I want to thank you for your generosity for sharing in the success of New England Biolabs with all of us. Okay? Not a lot of people would've done that. Last but not least, I want to thank you for the faith, for the faith you've put in me and everybody else here over the years to help grow and build New England Biolabs to fit your vision.

Don Comb:
Thank you all, and with the butterflies flying around house, I couldn't be better.

Tom Evans:
Morning, everyone, that's always such a powerful video for me, so many good images and even from the old building, which it's very special. So for those of you who don't know me, I'm Tom Evans. I'm executive director of research here at New England Biolabs, and it is my great honor to be able to introduce Dr. Jo Handelsman as the 2023 Donald G. Comb memorial Lecturer. So Jo is currently the director of the Wisconsin Institute for Discovery in Madison, I'm sorry, the University of Wisconsin Madison. She has a number of prestigious awards and appointments, so I'm just going to go through a couple here. She is the violist or a violist research professor at the University of Wisconsin. She's a Howard Hughes Medical Investigator Institute professor. She served as associate director for science at the White House office for Science and Technology, during the Obama administration. She was awarded the Presidential award of Excellence for science, Mathematics and Engineering Mentorship, and this was presented to her by President Obama in 2011, and just this year she was elected to the National Academy of Sciences. So congratulations on that latest award.
Jo is an advocate for fairness and opportunity. She's published a number of articles detailing gender biases in science hiring, and she founded an organization called The Tiny Earth Organization whose goal it is to increase diversity in science by allowing more people to do impactful science directly so that they can experience that thrill of discovery firsthand. As a scientist, Jo has made great strides in the understanding of microbial communities, the way they communicate, the way they interact, the way they create healthy soils, and the way those soils impact plants and other organisms that grow in the soil. In fact, one of our key papers coined the terms metagenomics. Hopefully I use that term a lot and hopefully I use it properly each time.

At dinner last night, we had a number of very interesting and fun conversations, the ones that I enjoyed the most actually related to the soil micro, the microbial community in the soil, how that soil is impacted by our farming practices and the erosion and how we as a community and farming can do things better to preserve such an important resource that feeds our population. But as I reflected on it, to distill it down, I decided Jo is a person who's passionate about science, passionate about people, and passionate about doing the right thing. And as Sal mentioned in his introduction, those are traits that Don definitely would've enjoyed, and she's excellent presenter for today's lecture. And so with that, please join me in welcoming Dr. Jo Handelsman.

Dr. Handelsman:
Well, thank you Tom for the lovely introduction and everyone being here to hear the Donald G. Comb lecture, especially his family. I really appreciate the chance to meet you since I never had the chance to meet Don himself. It's really an honor and a privilege and kind of humbling to be this lecturer, especially given how much I've learned about Don, how much he cared about his employees, the science they did, the basic science they do, the business they ran and the environment, and people across the world generally, he really left quite a legacy, so I am quite honored to represent his memory. So today I'm going to talk about some lifelong questions that have bothered me since actually before I even ordered my first EcoRI from New England Biolabs. This was something that goes back to graduate school. I was a graduate student in molecular biology, but rapidly became interested in the molecular basis for how communities function and how they interact.

And so some of the questions that immediately became evident when I started looking at very simple, and then later on much more complex communities of bacteria were, first of all, what's the role of bacterial signaling in community behavior? And it was clear that it looked like being a community was more than just an aggregate of organisms, which would suggest that they needed, the members needed a way to talk to each other and communicate. And if they were more than just an aggregate of organisms and represented a true community of interacting species, then we would think that communities have characteristics or phenotypes that wouldn't be predicted from each individual member in culture. And so I wanted to set out way back, I won't say quite how long, but a long time ago to answer these questions, but it was not a time to answer them. I remember my advisor in graduate school kind of laughing at me in a kind sort of way, but smirking at the thought that we could do molecular biology on an entire community simultaneously.

And I didn't know what I was thinking of. It took much smarter people to come up with the methods that we use today, but I knew that there had to be a way to begin to study the community as the unit of study rather than just the organism. And if communities truly are something bigger and different than just aggregates of organisms, we would never get to the properties of communities by studying the individuals. I realized after a while that maybe a model system would help us answer some of those questions because the communities I was interested in were in the soil on plant roots and then eventually in the human gut. Those are extremely complex communities, typically hundreds or in the case of soil, thousands of species coexist and form the community. And doing any kind of comprehensive analysis on that rich and complicated a community just seemed pretty impossible.

And so over the years, I became enamored of the idea of developing a model community. Now, it was time instead of the molecular biologists for the community ecologists to laugh at me, because they said you couldn't possibly have a model community because it would never represent real life. But I pointed back to all of the great communities that have given us the backbone of biology, way back to phage, and e coli, to the mouse, nematode tissue culture, fish and Arabidopsis more recently. Those have all been incredibly powerful in developing the concepts that we now think of as universal in biology. And certainly not everything transfers from a community or from an organism such as a mouse or E. coli. Not everything will be the same as in the more complicated system such as a human, but some of the basic principles, of course, are identical. And so we learned a lot, but we also learned what's different from these models.

So the question was, could we do the same for a community? There's nothing quite like this when I started thinking about it. So I really kind of had to start from scratch and I was not really sure where to start. So I want to introduce THOR, which stands for the Hitchhikers of the Rhizosphere. This is my model microbial community or microbiome, and it has today three members, and I'd like to introduce them one by one and show you how we built up the community to what it is today. And then I'll tell you some of the work we've done to do parallel understandings of all of the genes, all of the metabolites, all the transcripts in this community. So we started with Bacillus cereus, and that started from an undergraduate project in my lab. We had a collection of microorganisms, all bacteria from alfalfa roots across the state of Wisconsin, and we were using them for something else, but the undergraduate said, I want to look for, members of that collection that might suppress disease.
I was in a plant pathology department, didn't know much plant pathology, so I had to quickly learn something about plant pathogens. And the really important group of pathogens that fortunately due to a colleague's tutelage we ended up on were the oomycetes. This is an example of phytophthora. Oomycetes are lower protists. They were long thought to be fungi. In fact, people still say that a fungus caused the Irish potato famine because doesn't sound as good to say a protist caused the Irish potato famine, but we now know that it's most closely related to the golden brown algae and not to fungi at all. And so we studied these pathogens that have devastating effects in the field and in transport of agricultural products. This is an example of the kind of infection that phytophthora can cause on alfalfa. These are healthy alfalfa plants. This is an example of another Oomycete on cucumbers. By inoculating with just 500 spores of the pathogen, you can see that the entire cucumber is encompassed in this white fuzz, which is pythium growth.

And we wanted to set out to find an organism that would suppress these diseases. And so the student wanted to use the collection we had from alfalfa plants to find an organism that would keep a cucumber as healthy as this. This is an example of one, inoculated with the ultimate winner of that competition and would protect alfalfa and soybeans from their Oomycete diseases. And among the thousand, we only found one that met our very high standard for suppression of disease, which meant that under our conditions, all of the plants in the assay would survive. We wanted a hundred percent. In retrospect, that was kind of bizarre. People don't do assays where they require a hundred percent because you really rarely get a hundred percent of anything in biology. But I'm glad that we did because it was that stringent screen that gave us just one organism that survived through the sieve of a hundred percent plant survival.

And that was Bacillus cereus named it UW 85. And that has been an organism that has occupied me for the last almost four decades. Bacillus is a wonderful spore forming bacterium, and you can see here the endospores inside the mother cells. It also has some extraordinary capacities that even though it's one of the most common species in soil, there are about a hundred thousand Bacillus cereus cells in every gram of soil we've looked at. So this organism is ubiquitous. What had never been seen is the antibiotic that it produces. And it's not just UW 85 that produces it. It is many Bacillus cereus. About 20% of Bacillus cereus bacteria will produce this antibiotic called, Zwittermicin. My belief is that it was never discovered because it's hard to work with. It's an extremely polar molecule as you can see, and chemists don't particularly like polar molecules because everything purifies with them, all the sugars and amino acids.

So I finally found a very brave chemist, John Clardy then at Cornell, now at Harvard, who was willing to take this on more as a favor. And I think it started out as an indulgence because he wanted to help my lab and eventually became one of the most interesting projects in my lab. Zwittermicin turns out to be still even after many years, the only member of its class of amino-polyol antibiotics, and it has an absolutely extraordinary set of activities. Not only does it inhibit growth of the Oomycetes, but it also inhibits certain but not all bacteria and certain fungi. And it has a tremendous capacity to augment the activity of Bacillus thuringiensis toxin. Bacillus thuringiensis is the insecticidal bacterium that's used more widely than any bioinsecticide and tiny amounts of Zwittermicin in the nanomolar levels. Much lower concentrations that are needed to see inhibition of growth of bacteria will dramatically boost the activity of Bacillus thuringiensis, which is a story for another day.

So I think Zwittermicin was probably born to augment the activity of BT toxin and not be an antibiotic, but it serves quite well as an antibiotic as well. So we were all set to demonstrate that in fact, the way that Bacillus Cereus suppresses disease in the lab and in the field was through the production of Zwittermicin, which is highly inhibitory toward the Oomycetes. But a wonderful graduate student came along very early in my career, Greg Gilbert, who said, "I really don't think that's it at all. I think that it's the camouflage of the root that's important."

"I never heard of that. What's that?" His idea was that when we inoculated seeds with Bacillus cereus and then a root would eventually develop from that seed, that the Bacillus had the capacity to completely change the microbial community that was fostered on the root. And by changing that community in a particular direction, perhaps the root would be camouflaged from pathogens. And this was based on a long analysis of about 80 years of looking at microbial communities on roots when disease either was present or not. And Greg found this remarkable pattern that no one had seen before that when you looked at roots that were in some way protected, either through genetic resistance or through soil amendments or through biological control agents, or even in some cases chemicals, there was a shift in the community on the root to look more like a soil community and less like a root.

And so his idea was, well, if it doesn't look like a root, then the pathogen just swims on by and it can't find the root. And so he set out to test that hypothesis by isolating bacteria from field roots that had been inoculated or not with Bacillus cereus. He isolated, this is before 16S analysis and certainly long before metagenomics. So he did cultural analysis. He cultured organisms from those roots and then did classical physiological tests on those organisms and used a discriminant analysis statistical method to test his hypothesis to see if you could separate the communities from soil roots and roots that had been treated with UW85. And sure enough, he found in the discriminant analysis, you could see that these are the squares are untreated roots, the bulk soil are the circles, and Bacillus cereus treated roots are the triangles. There was a statistically significant separation of those three communities.

So Greg was right that in fact, there was a dramatic change in the microbial community despite the fact that we start out with large populations of the Bacillus on the seeds, but by the time the root was only a few days old, we can no longer detect our Bacillus. There were other Bacilli there. We no longer detect the inoculum on the roots. So we called it kind of the ghost. It had this ghost effect that even after Bacillus was gone, it was having this cascade of effects on the composition of the root community. And what was more significant was that Greg found that in fact, the treated root community looked more like bulk soil than did the untreated roots. So that was consistent with the camouflage hypothesis that in fact, making a root look more like bulk soil might be the basis for protecting it from the pathogen.

So some of the community changes that stand out, and I'm just going to summarize a few points here. The first was this change that made the root look more like a soil community. The second was that when we analyzed the communities that Greg cultured from the roots, we found this massive increase in organisms that look like what we call today Cytophaga. In those days, it was a flavobacterium group, and these are gliding bacteria, most of whom form yellow, very bright yellow colonies, thus the Flavobacterium name. And he found that sometimes as much as 30% of the community, which is extraordinary was the Cytophaga-like or flavobacterium-like bacteria.

And that's an awfully big percentage of a very balanced community that is normally quite diverse and doesn't have dominant members like that. So something was going on that made the inoculated roots conducive to Cytophaga. We found when, well, Greg found, I didn't accept this result for years, Greg found that when he put his bacillus, he would isolate bacillus serious from soybean roots grown in the field, whether they were inoculated with Bacillus serious or not, he could isolate bacillus serious just because it's a natural member of the soil community and gets onto roots.
When he isolated bacillus serious, he kept finding that he'd get these yellow slimy things as he liked to call them, growing out of the colonies of the bacillus. And I said, "Well, you're a plant pathologist. You're not a microbiologist. So clearly there's something wrong with your sterile technique. And these are contaminated colonies," the arrogance of the 26-year-old professor. So I learned better to listen to my students because in fact, it was a very repeatable phenomenon. And by the time the third student showed me these plates, I started believing that maybe there was a phenomenon here that was real and based on one very brave student, perhaps one that we could study. And Brooke Peterson eventually took this relationship apart and showed that any soybean root that you find in Wisconsin soils will have the same phenotype.

You can isolate bacillus serious from it. You then colony, purify, have a nice clean culture of bacillus and then put your pure colonies in the refrigerator for about two to three weeks. And she found that consistently 5% of the colonies would develop that yellow halo that Greg had seen, and it was very consistent 5%, which convinced me that maybe there was a phenomenon here that was real and could be studied. Once those organisms that had been grown in the cold and stimulated to grow separately came out of the bacillus culture, then we could colony, purify both and get them apart. That was not a problem. So we had a very interesting organism to begin to serve as perhaps another member of our model community. We called the organisms that came along with Bacillus from the field hitchhikers, and we found some general principles about them.

Again, about 5% of bacillus isolates will carry them, but interestingly, only the isolates from roots and not from the soil. And you can get plenty of bacillus serious from both. Only the root isolates would have this 5% frequency of carrying the hitchhikers. 83% of the co-isolates or the hitchhikers were from the Bacteroidetes group, and that is the same group that contains flavobacterium and Cytophaga. So began to explain some of the changes in the microbial community that Greg had seen. And it turns out that Bacteroidetes is not a really dominant member of the root community. Nobody's really a dominant member of the root community. And so it clearly was a selection from the root or soil community done by the hitchhiking process. And if you looked at the distribution of bacteria on root versus the co-isolates was totally different, much higher percentage of Bacteroidetes among the co-isolates. And then the other 17% were largely Pseudomonas and a few other proteobacteria, but they seemed a little bit more scattered.

So there was something unique about this relationship with the Bacteroidetes, and we began to, I think we might be understanding where this relationship evolved. When we tried to grow the flavobacterium, which has now changed its name Cytophaga, the flavobacterium in root exudates. So this is the material that roots will secrete while they're growing and you can collect it, make it sterile, and it will feed many organisms that normally grow on roots. So this is in the purple, you see the initial inoculum. In red, you see the inoculum where the inoculum is after three days, and then the yellow is that same culture of flavobacterium if we add Bacillus serious, and you can see that it doesn't grow on soybean root exudate, which is a pretty peculiar thing for an organism that was isolated from soybean roots or alfalfa roots in many cases. We found the same thing with alfalfa, no growth by itself.

And then when we add Bacillus, the flavobacterium grows fine. And so that suggested to us that maybe this hitchhiking phenotype arose from flavobacterium's need to be fed by Bacillus serious. Bacillus sloughs off pieces of cell wall. And it turns out that peptidoglycan, a component of the Bacillus cell wall is great food, makes a nice carbon source for flavobacterium, and that's what's missing in root exudate. There aren't normally carbon sources that can use from the root exudate, but when Bacillus is there, it gets all its other nutrients from the root, and then the peptidoglycan serves as its carbon source from Bacillus serious. So we had this nice proposal for symbiosis. Can't say that that is why it evolved, but it certainly is consistent with this evolution. Then we started finding out almost the opposite phenomenon. Very recently, we found that Bacillus might even be a hitchhiker under some conditions on flavobacterium.

First, we found that this gliding bacterium normally on a surface you see individual cells gliding on the surface. It's a very unusual and interesting form of motility, very different from the flagellar mechanism that most bacteria that are modal will use. But we observed a very odd kind of movement that if the individual cells were given a surface to attach to, they would attach and start what one of the physicists called a pinwheeling motion. They would start to spin. And that pinwheeling motion had sufficient power behind it to actually move a very large aggregate of cells across the surface. And so this was the discovery of the so-called zorbs, which are these aggregates of flavobacterium that have what my colleagues in engineering like to call the leg cells. These are the cells that pinwheel on the surface. So the zorb itself doesn't spin, it doesn't roll.

It literally walks across the surface on these individual pinwheeling cells. And this was the first evidence that a biofilm, which this appears to be, could actually have a motility function. We think of biofilms as sessile. They're actually used by bacteria to attach to surfaces and remain in a location. This seemed to be the opposite. A mechanism by which the bacteria can come together and move as a group. The zorbs could get quite big with thousands of cells. They had polysaccharide cores that we demonstrated biochemically, and that defined them as a biofilm, aggregates of bacteria held together by polysaccharides, basically a biofilm. So these were the first what we call moving or walking biofilms. It was the form of zorbs. And eventually then the zorbs would disaggregate. And all of this occurs in about 24 hours. So this is a time-lapse video that shows you several zorbs in the field.

So in the middle you see two zorbs that are meeting after moving toward each other and then fusing. They merge and form a larger zorb. And that seems to be a very common phenomenon that zorbs get bigger by fusing with other zorbs. But this is an example of zorb that just got bigger by itself. It didn't fuse that we could see anyway, but it did grow and it did move actually quite rapidly across surfaces. So that was our first discovery of zorbs. But then we found that the zorbs weren't always pure cultures of flavobacterium. When we added Bacillus cereus, the zorbs would have Bacillus inside of them. And so this is why I say it's almost a reverse hitchhiking. The red is flavobacterium and the greenish-yellow is Bacillus. And you can see that.

There we go. You can see that first. You can see the flavobacterium spheres forming, and then eventually you can start seeing the Bacillus cores inside. So it looks like a biofilm within a biofilm. The Bacillus cores will actually even maintain some of their structure even after the zorb disaggregates and explodes and the flavobacterium cells move off, the bacillus will actually remain as a core. So this suggested that maybe the bacillus was attaining something from living inside of flavobacterium. That relationship was beginning to look rich enough that we thought the flavobacterium would make a great second member for our model community. So bacillus is a great one because we know a lot about it. We know a lot about its metabolites. We have a very good genome sequence. We know a lot about its modes of gene expression and many other functions. And so that was a good start.

But its relationship with flavobacterium is really quite extraordinary and is unlike anything else described in the literature. It's this odd symbiosis of two organisms that can live separately, but in fact seemed to migrate toward this hitchhiker existence. And then one feeds the other, the peptidoglycan effect, and then they form zorbs, and those zorbs of flavobacterium will encapsulate bacillus. So we wondered what might be a benefit of the zorbs, and that led us to say we have to have a third member of the community and we want one that's very interactive with lots of small molecules. You may remember, very early interest was how to signaling play a role or secondary metabolites play a role in microbial communities. We did a hierarchy analysis of all of the members, the root isolates, and then the co-isolates that co-isolated with bacillus in our collection from the field. And we found one member, the Pseudomonas koreensis, CI, which means that it was co-isolated or it was a hitchhiker.

We found that in the hierarchy analysis, it was the most interactive with other species. We thought, okay, must have a lot of secondary metabolites or ways to signal or some other way to interact with other species. And my graduate student, Gabrielle Lizzano, found that it produces another novel antibiotic. This is why you might know, I've advocated for a long time, is the soil is not tapped out of secondary metabolites as many pharmaceutical companies would have us believe. Because every time we start looking at the secondary metabolites of a soil organism, we start finding novel antibiotics without even looking for them. This one popped up because it inhibits flavobacterium specifically. It's koreenceine A, and it falls into a class that actually is the closest to the coniine type molecules, coniceine, which are in plants. And in fact, that's the compound that was in hemlock that Socrates drank. Socrates or Aristotle? I always forget who it was. Somebody drank hemlock, that's highly toxic to people.

This one we don't know much about. Its toxicity to higher organisms, but it certainly is inhibitory to bacteria such as flavobacterium. We found several derivatives, but koreenceine B will turn out to be an important member of the story. We found flavobacterium, when it's grown in the presence of Pseudomonas, koreensis will not grow. So this is Flavobacterium alone. And then you can see here with Pseudomonas as soon as the Pseudomonas begins to produce, its koreenceine, the populations of flavobacterium drop. And we've shown that that is due to the koreenceine, not another molecule. But when we add Bacillus cereus, we found a very different picture. So again, the blue here is flavobacterium alone, the red is flavobacterium with Pseudomonas, and yellow is flavobacterium with Pseudomonas and Bacillus. And so what that shows is that when we added bacillus, it protected flavobacterium. It was completely immune to this apparent antibiotic from Pseudomonas koreensis.

So once again, a symbiotic relationship between bacillus and flavobacterium where bacillus is protecting flavobacterium. And it turned out that one of the elements of that protection appears to be that bacillus can catalyze the reaction of koreenceine A to koreenceine B. So simply removing this bond and koreenceine B is not toxic to flavobacterium. And so that might represent, again, one more way that bacillus and flavobacterium benefit each other. So we started thinking we had enough interactions that maybe this really was our very small, but definitely meets the definition of microbial community, three members. And that was when we started calling it Thor of the Hitchhikers of the Rhizosphere, so it was Bacillus, Flavobacterium and Pseudomonas. And the nagging question has been for a long time, what are the interactions that are unique to the community? We have a number of pairwise interactions, but what makes a community a community?

And if these three organisms do make a legitimate community, we would certainly predict that they have some interactions. So the two other members are both hitchhikers on Bacillus. They all produce interesting metabolites that affect one or another of the members of the community. And most of the phenotypes that we've looked at that one induces some other member can counteract. And so another one that I won't have time to show you is that the other two members can cause Bacillus to make these sort of dendritic growths from its colonies. And that can be either augmented by having two members there, or it can be diminished by having one.

The killing effects, the inhibitory effects are reversed with koreenceine. Then I'll go on to tell you a little bit about some of the genes and molecules that are typical of the community and not just of individuals. So once we had all three members, we wanted to profile THOR and look at it throughout time and under challenge by other organisms under physical challenged chemical challenge as it's facing the many challenges that a community would face to try to begin to understand how it responds to its environment of organisms, chemicals, and physical changes as a community, as a group of organisms rather than just massive individuals. So we started with InSeq, which is a massively parallel method that allows us to map mutations in all the organisms in a mixture using one collection of mutants. And we can look at the loss or gain of function by those insertions by looking for their frequency just by a sequencing approach within the mixture.

So our prediction was when we started this, that if there are genes that are truly community specific that determine life in a community and aren't required for other functions, many of them shouldn't have gene function assignments because of course most of the genetics that we have and functional analysis of genes and bacteria has been done in pure culture. And so we would expect that community-based genes, where until now nobody really has screened for mutants or gene function would have perhaps functions that are so far undescribed and perhaps some of those pesky, 20% of most genomes of bacteria that don't have known functions are in fact unfunctional because we don't have them under the right conditions, namely in communities. So the InSeq method involved sequencing a large community of mutants before and after a stress, and then looking for ones that are either a higher frequency in the mutants in the collection of sequences and ones that are a much lower frequency than the wild type.

And we found genes of both types, and we looked specifically in sand communities. So sand is one of the dominant substances on earth silicon dioxide, and it's one of the major constituents of soil. So we thought it was a good place to start when we were looking for surfaces that flavobacterium or others would attach to. We also looked at planktonic free floating communities and asked what genes are community specific in each? Well, we found 29 genes that contribute to fitness only in the community, that the organism is fine by itself. The mutants are fine when they're growing by themselves, but they are diminished in fitness when the community is present. We found 42 that reduce fitness in the community. So we found both directions that there are genes that will reduce lower fitness and then others that will increase fitness. Why an organism would carry so many genes that reduce fitness in the community?

We're not really sure, but it seems like an interesting phenomenon. Perhaps those are genes that are important under other circumstances when they don't have a community to contend with. And then there were 11 mutations that affected fitness, reduced fitness in the community, specifically on sand. And these were only in the community, so they had no effect of fitness when flavobacterium was grown alone. So they are community specific genes, and as we predicted, many of them were of unknown function, and in fact, many of them mapped in one cluster, an enormous cluster of genes of unknown function. So this cluster is actually since grown. I believe it's up to 70 kb. We don't know the operon structure yet, so we'll refer to it as cluster. And many of the genes that were hit by the transposons indicated by the asterisks that affect community life appear to be in glycosyltransferases or near those.

And so we suspect that the pathway that's encoded by this cluster is for synthesis of an extracellular polysaccharide. It's consistent with that, but it also seems to have some function in production of lipopolysaccharide, which is a very different kind of molecule in bacterial cell surfaces. So we asked, let's call these putative EPS mutants or extracellular polysaccharide mutants. We asked what was the phenotype in zorbing? And you can see here, this is just a still picture. You can see nice zorbs formed by the wild type, and none of the putative polysaccharide mutants formed zorbs, or if they did, they were very, very tiny that were sort of at the edge of the microscopic analysis. So that looked interesting because these seem to be quite effective genes in a community and in forming zorbs. And so I won't go into any more of the genes, but I can say that we found ones that were sand specific, like the putative polysaccharide pathway.

We found ones that were specific to planktonic growth. And one that was particularly interesting is an alarmone or stress respondent, which is called (p)ppGpp, formerly known as MagicSpot for many years in bacteriology. And this is a signal to the cell that there's some sort of stress and there needs to be a response. Turns out that MagicSpot also regulates flexirubin, which is the compound that makes flavobacterium yellow or orange. And without MagicSpot, you can see it's just a very, very pale yellow. With MagicSpot present, you get a lot more of the pigment produced. We don't know the biological significance in the community, but we think that's going to be a really interesting area to follow up once again where the community is regulating the production of a secondary metabolite. So if we go back to our model, we now have two hitchhikers, a hitchhike on Bacillus cereus.

One produces an antibiotic that is inactivated by the other one, but the antibiotic is effective against a third. They form many interesting structures that are specific to the community, including apparently these polysaccharides that are only expressed in the community. And we're working on the biochemistry of that to try to figure out what the polysaccharides are doing. We recently found that other polymers can actually compensate for the polysaccharides in zorb production. So that makes us, if some sort of polymer, maybe any polymer is needed for zorb production to initiate. And the engineers we work with have some physical explanations why that makes sense. I want to move on to metabolomics because we found one pretty extraordinary molecule looking for community-specific activity. We screened all of the metabolites in the community using mass spectrometry, and we found one particular phenomenon of interest, and that was this is a mass spec plot that only when all three members, the circle here is around the treatment with all three members in the community when only when all three members are present did we find this particular mass spec signal.
And so the rest are all singles or doubles. The pairs didn't synthesize this compound. The individuals certainly did not either. And so we were curious what this is and why it only forms when the community is there. This is a novel compound. We don't know of activity, but it comes, sorry, this is the structure of the compound, very complex compound that's related to Lokisin. We did not name Lokisin, but for those of you who are fans of Norse mythology, I did not know this, but my students made sure I did. Loki is Thor's brother. So it was particularly fortuitous that somebody had already named a compound from THOR as Lokisin. And so this is a compound that's well known. It's an antifungal agent produced by Pseudomonas, and the genes for its synthesis are carried in the Pseudomonas koreensis genome. But when Pseudomonas is around Bacillus, Bacillus modifies Lokisin, and if flavo is there, it further modifies Lokisin or modifies the modified Lokisin to create the community molecule.

And so we've done a molecular network analysis to look at the relatedness of these molecules and shown that in fact, they do have a clear mode of synthesis based on the chemistry as well as the biology of the organisms that converts Lokisin to the community molecule. So once again, we seem to have a community specific molecule and a function associated with forming that molecule that makes for a community that is different from just simply the sum of the parts. And finally, when we did RNASeq to look at all the transcripts in the community, we found that the profiles of gene expression were quite different in the community than when pairs were present or when the organisms were present in solitary culture. Eventually we want to take all of the data that I've talked about and link it together in mathematical models. We're working with a modeler and a network analysis expert who is looking to pull all of those data together and figure out how all of all these compounds and genes interact.

So one of the extraordinary elements that came from the RNASeq analysis, was that koreenceine has an outside impact. Nothing else we've seen so far has this kind of global impact on gene expression. So we compared gene expression in singles pairs and the triple organism community with and without koreenceine. We used a mutant that doesn't produce koreenceine and the wild type. And when we compared these, we found that in Bacillus cereus mixture with Bacillus cereus we found, and these are ordered according to the black, these are the genes that are affected by Pseudomonas in the order. They're just lined up in the order in which Pseudomonas affects gene expression. And so they're ordered by increasing gene expression with Pseudomonas. But when we knock out the koreensis pathway, the koreenceine from Pseudomonas, we see a reversal of that trend. And genes that were turned off in the presence of Pseudomonas are now on in the presence of the koreenceine mutant.

And you can see there's a pretty parallel but opposite relationship. With flavobacterium genes, it was even more so. You can see again, the genes are lined up in the levels of expression that is induced by Pseudomonas being mixed with the flavobacterium. And when we do the same lineup of those genes with the mutant Pseudomonas, so no koreenceine there, once again, we see a radically different picture. Genes that were turned on with the Pseudomonas are now strongly off with the Pseudomonas that lacks koreenceine. Those are just two examples. When we look at any combination of strains or the community itself, the one theme that is consistent across all of the interactions is that koreenceine is a major gene regulator. So when we start looking at the interactions among THOR organisms, we seem to have one element that keeps popping up as critical for regulating the functions of the community.

And we call that koreenceine, THOR's hammer because that was Thor's most powerful tool. And koreenceine seems like a pretty powerful tool in this community. Of all the things that we predicted, I never would've predicted that one molecule could have this global effect on gene expression. I thought it would be a much more subtle mixture of small and incremental gene effects, and it would take a lot longer time to sort out the network, but at least for the dominant member of that network, the force that really seems to be driving the changes in gene expression when the community members come together, koreenceine is the big player. So if we return to my original questions, I think we have pretty good evidence that microbial signaling with small molecules is important in community behavior. I haven't been able to talk about all the compounds, but we know of quite a few that are specific to the functions in the community or among the pairs of the members.

And the outcome can be changed such as with koreenceine, by adding a third member to a pairwise interaction. And so whether these are classical signals, which certainly koreenceine serves that purpose, or they serve as killer molecules, I'm not a big fan of the military model of antibiotics that they were evolved to kill other organisms because there's actually very little evidence for that in nature. And most of them aren't produced in nature at levels sufficiently high to kill other organisms. So some of these, at least I would expect, are just signals which may be a much more powerful kind of compound to have in a community than one that just knocks down populations much more subtle and perhaps much more important for responding to elements of the environment. And the second characteristic that we asked about was whether communities have characteristics that one wouldn't be predicted from solitary culture and by extension are specific to the community and would not do not have a role or appear in single culture.

And so when we look at all these interactions, many of them occur with pairs, but in fact, we found enhanced biofilm that I haven't had a chance to talk about the glyphosate, the global changes in gene expression and the community molecule that are all differentiated when the community is all together. So once again, these are characteristics we would not have predicted from just looking at the individual genomes or studying the individual organisms and culture. And we only see these functions when the three organisms are together. Suggesting, and I think this is some of the first evidence we have, that a community is in fact differentiated from other kinds of bacterial growth. That behaving in a community is a different phenomenon and the benefits from the community are perhaps quite significant compared with solitary growth.

And that's where of course much of our future work will go is toward characterizing the importance of these community functions, and particularly the ones that don't have whose genes don't have known homologous, because I think those are going to be really powerful in defining the uniqueness of microbial communities. So that is the Thor model as it stands today. And I think the future is going to be very revealing on both the specific biology of many of these phenomena, such as the formation of zorbs and how that happens, the production of many of these molecules, the activities of these molecules, but also about the nature of the community and what differentiates a community from an individual cell growth.

With that, I'd like to thank all of the many members of my lab that have contributed to the THOR story, and particularly my current group, Austin Hall, Natalia Rosario Melendez, Shruthi Magesh, Julia Nepper, Margaret Thairu, and Martel DenHartog, who contributed to a lot of the current data and particularly the RNASeq work that I talked about at the end. I've been lucky to have absolutely fantastic students all along the way from Greg Gilbert all the way to my most recent students who have taught me an enormous amount. And of course, the early lesson I learned, for those of you who are still young scientists, is don't think better than your students. Listen to them. And with that, I'd like to thank all of you for having me as your Donald G. Comb speaker, and I'd be happy to take questions.