What had Silverman and Hastings learned by that point, and where did you come in?
B: Hastings and then Silverman worked on this bacterium called Vibrio fischeri, a bioluminescent marine bacterium that lives inside a squid. Hastings isolated the autoinducer, and then Silverman found some of the genes that control this system. They found that the autoinducer allows bacteria to count their neighbors. When a bacterium is alone, it dribbles out a couple of molecules and they just diffuse away. But if you have lots of bacteria together, the autoinducer will begin to build up. And when there’s enough, it turns on a whole host of genes, including those involved in bioluminescence. The first investigators called this process “density sensing,” and we now call it quorum sensing.
I worked on another bacterium, Vibrio harveyi. It’s bioluminescent, too, but it is a free-living bacterium, so it encounters all kinds of different environments. We knew it made a similar autoinducing molecule, so we decided to try and unravel its quorum-sensing machinery. I just assumed that it would work the same as it does in Vibrio fischeri. I did the same experiment a billion times — looked for mutants that didn’t make light, hoping I would have knocked out the quorum-sensing machinery. It should have been super simple. But the strategy of looking for dark mutants didn’t help me find the right genes. I had this crisis of confidence. You think there’s something wrong with you. And then, eventually, you say to yourself, “Perhaps I’m not thinking about this correctly.”
It occurred to me that maybe the reason I couldn’t get the correct genes is because there are two quorum-sensing molecules. It could be that when I knock out the genes for one, the other still works. That was an epiphany. If there were two molecules and you knocked one out, they’d still make some light. So I tweaked the experiment to look for mutants that were dim, not dark. My dim mutants were clearly lacking one molecule, but they still made something else. They were impaired, not broken. There had to be another molecule.
What was the second one doing?
B: I didn’t have a clue. Why do you need two? I knew there was a second molecule out there. The genetics told us that molecule existed. But I didn’t have that gene, and I didn’t know the chemical structure of the second molecule. We took a lot of flack for that. The idea that bacteria could talk with one molecule was only starting to percolate through the community. So the idea that there were multiple molecules involved in bacterial coordination was so out there.
Your work showing that there had to be a second quorum-sensing molecule helped you get a job at Princeton, and in 1994 you started your own lab there. Did you figure out the purpose of the second molecule?
B: When I got here, we finally found the gene that made the enzyme that made the second molecule. But it wasn’t just about finding the gene. We wanted to understand why Vibrio harveyi has this second molecule. So we sequenced the DNA of that gene. Our real hope was that our gene would look like some other known gene and give us a clue about what it did. There were about 40 genomes of bacteria sequenced at that time, and what you could do was compare your gene of interest to other genomes to see if they contained something similar. The computer would scan through all the known bacterial genomes and say, “Do any of them have that gene?”
That process took a long time back then. You type the DNA sequence of your gene into a database, and then you sit and you wait. I remember the screen filling up one by one. The database told us every bacterium that had been sequenced — not just the bioluminescent bacteria, but every bacterium — has it; they all make an identical molecule. And I said, “They’re talking to each other. That’s how they talk across species.” That idea had not occurred to us. So that was an amazing moment. I still get goose pimples.
Did other scientists buy your explanation?
B: At first, the quorum scientists had trouble getting others to believe that bacteria could speak within their species. And then we came forward with this idea that they could talk across species — lots of people thought I was nutty. The dogma had been that bacteria can’t communicate, so it was hard to accept that they could talk using two molecules, and even more difficult to imagine they could talk across species.
It takes a long time to change dogma. And there were still problems with the story. We had the gene, and the gene was in all these different bacteria, but we still did not have the identity of the molecule. So there was a chink in our armor.
In 2002, you finally identified the second molecule. And then you won the MacArthur “Genius” Fellowship, a grant awarded to individuals who show exceptional creativity. Did you feel vindicated?
B: That put me on the map. This prize is very coveted, and it also brands you as creative, not crazy. I’d been here at Princeton eight years, and I had never been able to adequately fund this lab. I was struggling to get money. I had been trying, trying, trying — writing five, six, seven grants a year. Most of my grants were getting rejected. And to get this thing that you didn’t try for — that was so important for my confidence. And it branded quorum sensing as the hottest, coolest, most creative science.
You’ve now found that there are actually three molecules involved in quorum sensing. The first, the one Hastings found in the ’70s, allows bacteria to count their siblings. The second allows them to detect other species. And the third, which is made by all bacteria in the Vibrio genus, allows bacteria to identify their “cousins,” or extended family, giving them even more information. How do bacteria use these molecules to communicate?
B: What they first do is they scan the environment. And they’re asking the simplest question: “Am I alone or am I in a group?” They just look for any quorum-sensing molecule. Then, the more sophisticated question that I think they ask is, “Who is that?”
They can say, “You are my absolute identical twin.” They can say, “You’re my extended family.” And then they say, “You’re some other species.” They’re not just counting. There’s information encoded in these molecules that tells a bacterium who that neighbor is — how related they are. And depending on the ratio of those three molecules, they understand whether their family is winning or losing.
Why would they need to know that? How does that help them?
B: Having that information is extremely useful for decision-making. Bacteria aren’t just swimming around. They live adhered to surfaces. Your skin, your scalp, your intestines — they’re all covered in communities of bacteria, called biofilms. In order to make a biofilm, they have to secrete this substance that glues them all together, which acts like a shield. That’s controlled by quorum sensing.
For these communities to be maximally productive, they can’t be willy-nilly. They have to use multiple molecules to discern who their neighbor is — self or other — and to direct what job each participant in the community will take on.
Do you think the language of bacteria is more complex than we realize?
B: In 20 years, my field has gone from thinking of bacteria as asocial recluses to seeing them as at least being trilingual. And there’s mounting evidence that this is going to be an inter-kingdom dialog. Humans and all higher organisms live in fantastic association with many species of bacteria. We speculate that the host makes molecules that tell the bacteria what to do, and the bacteria make molecules that the host is tuned into. It has got to be like that.
This field is only really 20 years old. And we just haven’t found all the molecules yet. In the lab, we shake the bacteria around in a flask, and each bacterium perceives an identical environment. It could be that there’s a whole set of molecules that they never deploy there. To find those, you have to put them in a much more realistic environment.