Why do bacteria need to take up DNA in the first place? And how do they do it? There are three ideas about why a bacterium might want to acquire DNA in the first place: for genetic exchange, for use as a template to repair its own DNA, and even for food. But a bacterium can't take up DNA at just any time. It must physiologically ready itself for a state called competence.
Competence has been studied in gram-positive bacteria--notably using the pathogen Streptococcus pneumoniae and the soil bacterium Bacillus subtilis as models. The ability to be transformed, or to uptake DNA, was originally observed in S. pneumoniae by Oswald Avery and his colleagues in 1944. It was then documented that the bacteria's competence phase only occurred during a certain window during the growth phase, but it wasn't until 1965 that Alexander Tomasz discovered that competence was, in fact, the product of cell-to-cell communication. In both S. pneumoniae and B. subtilis, the regulatory cascade that leads to competence begins with a small pheromone peptide called an alarmone because it conveys a stress signal. As the bacteria grow to exponential phase, more alarmone is produced. This is sensed by a receptor on the bacterium's surface which activates the cascade that turns on transcription of genes that will make a DNA-uptake apparatus. Interestingly enough, competence also triggers bacterial fratricide in non-competent cells.
In a July 2006 Science report by Prudhomme et al., an additional idea was proposed as a reason for competence. Not only is extracellular DNA taken up to help repair DNA in times of stress, but it may even be a fitness-enhancing strategy helping to drive bacterial evolution. To test this idea, Prudhomme et al. stressed S. pneumoniae by adding mitomycin C, a chemical that damages DNA. Competence was monitored by hooking a luciferase gene to one of the competence genes and by assaying photon emission. So increased light production from the bacteria meant that they were becoming competent. The researchers' experiments showed that not only mitomycin C but also several other antibiotics such as kanamycin and streptomycin induced competence in S. pneumoniae.
The ability to gain competence in a stressful environment--such as the sudden introduction of antibiotics--has some obvious benefits for the bacteria. When a bacterium senses the antibiotic in its surroundings, the competence genes are switched on. It begins producing and exporting alarmone which is sensed by neighboring bacteria. They too begin producing alarmone, increasing the number of competent cells in the population. At the same time, the competence genes produce lytic enzymes and the proteins that will make up the DNA-uptake apparatus. The enzymes are the death knell to any bacteria that do not or cannot become competent--once the non-competent bacteria are lysed, in effect spilling their guts, the other bacteria take up in a Donner-party-like fashion the released nutrients and DNA.
Taking up other bacteria's DNA is particularly useful if the population of bacteria is diverse. In turn, the competent bacteria will be more genetically diverse. After all, some of that DNA might contain a useful gene for antibiotic resistance or conferring an advantage for the bacteria in another challenging environment to allow it to persist in places where it originally couldn't before. Competence, particularly in gram-positive bacteria like S. pneumoniae, probably plays a major role in how strains acquire antibiotic resistance over time. And perhaps, this observation is an "alarmone" for us not to overuse antibiotics as these bacteria fully have the ability to out maneuver us with this particular evolutionary trick.