Summary
· In the late 1960s, the idea that certain marine luminescent bacteria used a self-produced autoinducer to induce light production was developing.
· Autoinduction of luminescence in marine bacteria allows individuals in a community to cooperate. It controls a social behavior.
· Research in the last 30 years has led to discovery of several chemically distinct families of signals used for bacterial communication.
· The next decade will see interest in the evolution of quorum sensing systems and in bacteria as models to study the biology of cooperative behavior.


I first learned about cell-density dependent expression of luminescence in Vibrio harveyi and Vibrio fischeri in 1973, when I took the summer microbiology course at the Marine Biology Laboratories (MBL) in Woods Hole, Mass. Ken Nealson and Anatol Eberhard were among the instructors and, together with Woody Hastings, they had uncovered a phenomenon that they termed autoinduction.

During early log phase, the two luminescent species emit very low levels of light, and light increases nearly 1 million-fold during a short time in late-log phase. If experiments were set up properly, adding a little late-log culture fluid to a fresh culture could shorten the lag before luminescence increases. Thus, Nealson, Eberhard, and Hastings believed that the bacteria made an inducer, an autoinducer, that accumulated in the culture fluid until it was at a level required to stimulate luminescence.

Nealson and Eberhard explained to me and the rest of my classmates that summer that the autoinducer enabled the light-emitting bacteria to discriminate their low-density, free-living seawater environment, where light would be an ill-afforded waste of energy, from their high-population-density symbiotic environments (for example, within light organs of fish), where light was beneficial to the symbiosis. I thought this phenomenon was amazing. The idea that bacteria could make such "decisions" was not then accepted part of the thinking in microbiology. What was also extraordinary was that the autoinducers had species specificity. For example, V. harveyi culture fluid had no influence on the timing of V. fischeri luminescence and vice versa. These bacteria could distinguish themselves from the other species.

My Woods Hole MBL Summer Experience Set a Research Path

That summer in Woods Hole influenced my research path in two other major ways. First, I began to learn about work on bacterial motility and chemotaxis being done primarily in the labs of Julius Adler at the University of Wisconsin, Madison; Dan Koshland at the University of California, Berkeley; Howard Berg, then at the California Institute of Technologly in Pasadena; and Mel Simon, then at the University of California, San Diego. Bacteria somehow measure concentrations of attractants and repellents, altering their behavior to move to better conditions. The mechanism involves a rudimentary memory system that compares chemoattractant concentrations at present with those from the immediate past. Bacteria had memories and used them as the basis for behaviors.

Second, that summer opened my eyes to the fact that I was not the only person in the world who would rather spend Friday night at a lecture on ants than anywhere else. The MBL Friday Night Lecture Series each summer brings scientists at the top of their fields to Woods Hole to talk about their work. The town is primarily a town of scientists, mostly biologists, most of whom also preferred the Friday night lecture to other more typical Cape Cod weekend activities. I was in my element.

It was such a lecture where I heard E. O. Wilson of Harvard University describe his work on the social activities of ants. Wilson explained to the Friday night crowd that one could learn about genetic and environmental influences on the social activities of animals other than ants, humans included, by applying the sorts of methods that he used to study ants. He called this area of investigation sociobiology. His lecture provided the last piece in a confluence of information that led me to believe that autoinduction of luminescence in V. harveyi and V. fischeri facilitated social behavior and that I wanted to work on it. But I was just starting my second year as a graduate student at the University of Massachusetts, and I would be working on spirochetes with Ercole Canale-Parola for a few more years before I could return to that line of research.

After Graduate Work, Back to Bioluminescence and Autoinducers

My graduate research was enormously satisfying and productive. The unique trait of spirochetes is their motility system, in which flagella are tucked inside the outer membrane. Although I convinced Canale-Parola to let me study spirochete motility and chemotaxis, those efforts are part of another story. Skipping ahead, near the end ofmygraduate student years I wrote to Hastings at Harvard and Nealson, who was then at the Scripps Institute of Oceanography, to inquire whether they might be able to take me as a postdoc. It was my very good fortune that both were interested in having me work on autoinduction, but which lab should I choose? I knew I couldn't go wrong.

In the end I went to work with Hastings at Harvard in 1977, in part because the move from Amherst to Cambridge was easier than to La Jolla and in part because there were several sociobiologists at Harvard. When I started my postdoc, Nealson was studying V. fischeri, while Hastings and I focused on V. harveyi. The aims of my postdoctoral project were to work out a good quantitative bioassay for the V. harveyi autoinducer, and then purify and identify it.

My first discovery was that V. harveyi not only responded to a signal it made itself, but also to signals produced by some marine bacteria but not others. V. fischeri did not make the inducer. We called the induction by other species alloinduction instead of autoinduction. I struggled to identify the signal and ultimately decided that V. harveyi might respond to lots of different molecules. The problem seemed too hard.

Meanwhile, Nealson and Eberhard showed the V. fischeri signal was an acylated amino acid-specifically, 3-oxo-hexanoyl (HSL). Shortly thereafter JoAnne Engebrecht, working with Nealson and Mike Silverman, cloned the genes responsible for signal production and signal reception along with the V. fischeri luminescence genes. By the time the V. fischeri luminesecence genes were cloned, I was an assistant professor at Cornell University in Ithaca, and I decided to switch to studying V. fischeri. Not much later, Bonnie Bassler arrived to work with Silverman, and they switched to V. harveyi.

Bassler and Silverman showed genetically that there are two parallel pathways for autoinduction of V. harveyi luminescence, while Ted Meighan showed that one of the pathways uses an acyl-HSL signal, 3-hydroxybutyryl-HSL. In the summer of 1996 Bassler came to my laboratory for a few weeks to try to purify the other signal and to determine which of the two signals was made by other bacteria. She and Ann Stevens in my lab showed that the alloinducer was the unidentified signal and very few species made hydroxybutyryl HSL. But we made no progress determining the chemical nature of the alloinducer other than ruling out various possibilities.

Unraveling the Chemical Language of Bacteria

When Bassler moved to Princeton, she and her students quickly discovered that the mysterious V. harveyi signal was made not only by a variety of marine bacteria but also by nonmarine bacteria such as Escherichia coli. Eventually in 2002, she and her colleagues identified the V. harveyi alloinducer, which she calls autoinducer-2. The identification was a heroic effort. First they identified the gene required for signal synthesis and then, from bioinformatics, predicted what the substrates might be for the gene product. They showed the purified gene product produced signal from the substrates but still they could not determine the chemical nature of the signal. They solved the structure of the signal receptor, and the resolution was high enough to deduce the structure of the bound signal.

By the mid 1990s, research on acyl-HSL quorum sensing was opening up. Paul Williams, the late Gordon Stewart, and their collaborators at the University of Nottingham in the United Kingdom reported that a variety of bacteria, including Pseudomonas aeruginosa, make 3-oxo-C6-HSL. Although not quite right, their findings, together with the following key discoveries, stimulated interest in the field. At the same time, homologs for the V. fischeri acyl-HSL synthase and receptor in other Proteobacteria were being discovered, and it was demonstrated that 3-oxo-C8-HSL regulated gene expression in Agrobacterium.

An advance came for us when Barbara Iglewski and her group at the University of Rochester discovered that P. aeruginosa uses a homolog of the V. fischeri signal receptor to control virulence gene expression. We collaborated with her and enlisted the help of Eberhard to show that P. aeruginosa had two acyl-HSL autoinduction systems, one that uses 3-oxo-C12 HSL as a signal and another that uses C4 HSL. These findings constituted the graduate work of Jim Pearson, who wrote the best M.S. thesis I have ever seen. This rapid progress led to an invite from the editor of the Journal of Bacteriology to write a minireview, "Quorum Sensing: the LuxR-LuxI Family of Cell-Density Responsive Transcriptional Regulators," which I coauthored with Steve Winans and Clay Fuqua.

Comments (0)

Subscribe to this article's comment feed

Show/Hide comment form

Name

Email

Website

Title

Comment

smaller | bigger

Subscribe via email (Registered users only)

I have read and agree to the Terms of Usage.

Add Comment