...the signal peptide? Interesting. I'll start at the beginning.
One of the few bright spots regarding the problem of antibiotic resistance is that resistance typically infers a fitness cost to the bacterium, at least initially. In other words, the resistant strain usually grows slower than a nearly identical sensitive strain*. While compensatory mutations can lower or eradicate this 'cost of resistance', it is thought that resistance can't increase initially without favorable selective conditions--antibiotic use--due to the cost of resistance.
We'll need a little background about antibiotic resistance genes. One important class of resistance genes is known as the beta-lactamases. Beta-lactamase genes encode beta-lactamases (funny how that works). Beta-lactamases protect bacteria against some or all beta-lactam antibiotics (depending on the particular beta-lactamase), which are the antibiotics that begin with "cef" or end with "-cillin." The most widespread beta-lactamase is known as TEM-1 which confers resistance to ampicillin and amoxicillin. Other beta-lactamases, such as the CTX-M beta-lactamases, confer resistance to far more drugs, so it's always been puzzling why TEM-1 is still so prevalent.
Another broad-range beta-lactamase is known as SME-1, and it confers resistance to virtually every beta-lactam antibiotic including the cephalosporins and the carbapenemases. So why hasn't it spread? In lab culture, Marciano et al. found that plasmids (transferable 'mini-chromosomes' that often carry antibiotic resistance genes) with the SME-1 beta-lactamase had a considerable fitness cost: the cells with an SME-1 plasmid lysed (exploded) at a high frequency. After an initial innoculum of 1:1 SME-1 versus TEM-1, 24 hours later 98% of the cells were TEM-1. Doing some very clever genetic manipulations, the authors found that the lysis isn't due to the beta-lactamase protein, but due to its signal peptide; a 'hybrid' beta-lactamase with an SME-1 protein and a TEM-1 signal peptide didn't experience this lysis.
So what's a signal peptide? At the front end of many proteins is a section that is chopped off during the transport of a protein through a membrane. Traditionally, it's been thought not to be important; for those of you who remember cassette tapes, signal peptides were thought to be like the white part at the beginning of the tape. In my own work, my colleagues and I also found that the signal peptide of a fimbrial adhesin had a huge effect. Essentially, in that case, the protein took so long to get through the membrane because of the different signal peptide that the structure it was a part of grew longer, which has all sorts of effects**.
I know everyone is all excited about those mystery RNAs, but there's still a lot more we need to know about proteins.
*This isn't always slower growth rates per se. It can also involve increased cell lysis. At the population level, however, this manifested as a slower growth rate.
**The protein is known as an adhesin and is used to stick to various surfaces including the urinary tract. It sits at the end of a long protein structure known as a pilus on the outside of the cell (they look like little hairs). The pilus growth stops once the adhesin is transported to the end and 'caps' the pilus. What happens when the adhesin moves more slowly through the membrane is that the bacterium will have fewer, but longer pili. These pili are less likely to break under conditions of 'high shear flow'--in other words, when you pee. This is a good thing (for the bacterium) if it winds up in the urinary tract.