BEACON Researchers at Work: Plasmid evolution is the key to fighting antibiotic resistance

This week’s BEACON Researchers at Work blog post is by University of Idaho graduate student Julie Hughes.

We are in the midst of a medical crisis. Even though we have more antibiotics on the market than ever before, our ability to effectively combat antibiotic-resistant pathogens is constantly decreasing. While many people know that bacteria are becoming increasingly resistant to antibiotics, it is not common knowledge just how they are adapting so rapidly. One of the key weapons in the bacterial arsenal is the plasmid. Plasmids are DNA molecules that can be transferred between different bacteria through a process known as horizontal gene transfer. Most plasmids can code for a variety of traits, including the ability to degrade organic compounds, virulence, and, of course, resistance to antibiotics. Moreover, one plasmid may contain multiple such genes, providing its bacterial host with a full range of armor against various types of antibiotics at once. Therefore, if we want to combat bacterial pathogens, we need to learn more about plasmids and how they interact with their bacterial hosts over evolutionary time. We need to get to know the enemy and start looking for chinks in its armor for antibiotic use to be successful in the future.

Fortunately, a bacterial population won’t necessarily maintain plasmids forever. Just as wearing armor will protect a knight during battle, plasmids are useful to their hosts under certain environmental conditions (like in the presence of antibiotics, or an organic compound that could be used in metabolism). After the battle, a knight will shed his armor because it is no longer needed and it only serves to slow him down. Likewise, the cost of producing plasmid-encoded protein products and replicating the plasmid can retard host growth. Plasmid-free cells would therefore have a competitive edge over plasmid-bearing cells, and so eventually only plasmid-free cells will persist in the population in the absence of selection for plasmid maintenance. Additionally, to be maintained in a population, plasmid and host proteins must interact properly to make sure that copies of the plasmid are made before cell division and to ensure that each daughter cell gets at least one of those copies. Otherwise a plasmid may be lost from a population over time – that is, it may have poor stability.

When a plasmid is first introduced into a new host, it may exhibit poor stability. If selective pressures mandate plasmid maintenance for survival or competitive success over evolutionary time, either the plasmid, the host, or both can adapt to each other to improve plasmid stability or reduce the cost of plasmid-carriage. However, it is not known how quickly plasmid-host adaptation occurs or what evolutionary dynamics are involved in this process. The goals of my research are to address these two questions.

I am currently working with a mini-replicon, pMS0506. This plasmid was originally constructed from a natural plasmid that was isolated from Bordetella pertussis, the causative agent of whooping cough and exactly the type of host we don’t want plasmids giving antibiotic resistance to. The host I work with is much more benign, as Shewanella oneidensis MR-1 was originally isolated from lake sediment, and is well-loved by many for its ability to reduce uranium to more stable forms. It was previously shown that after long-term evolution of pMS0506 in this host (over 1000 generations, or 100 days of growth), our plasmid improved its stability in strain MR-1 through mutations in the gene that codes for the replication initiation protein, TrfA1. The tradeoff of adapting to this new host was that it could no longer replicate in the human pathogen Pseudomonas aeruginosa. This is called a host range shift.

My research involves determining the tempo and evolutionary mechanisms of pMS0506 evolution. While we knew that plasmid adaptation occurred within 1,000 generations we now know that these populations exhibit improved stability within 200 generations (only 20 days!). Keep in mind that this is when the population reached high stability – individual plasmids mutated well before then. It’s quite possible that even during the course of patient medication, plasmids are adapting to their bacterial hosts, which may mean that bacteria will remain resistant to certain antibiotics even after medication is completed. The longer a plasmid persists, the more likely it is that it will find its way to new hosts, including the next bug that’s going to make you sick…and so on.

Another aspect of my work involves the dynamics of evolution. I am interested in questions about how many different mutations may arise in the population, how do they interact with one another (if at all), etc. It was once thought that beneficial mutations would occur so rarely in a population that you would only ever have one mutant type affecting a given phenotype in a population at a given time in an asexual population. Now it is clear that in large populations with high mutation rates, it is possible to have more than one variant present at once. Since they are asexual, bacteria don’t recombine their genes all that often, and so two different beneficial mutations on two different cells are likely to stay in different cells. These cells will then compete with each other (and the ancestral types as well) until one of them eventually “wins,” or outcompetes the other types. This process, called clonal interference, has some important implications for evolutionary dynamics. If a given mutant (let’s call it Mutant A) was competing only with the ancestor, it would be relatively easy to dominate the populations. If you add another strong competitor into the mix (named Mutant B), A doesn’t have to just outdo the ancestor, but B as well! Since B has a beneficial mutation he’s no pushover, so he’s hard to beat, so they have to duke it out for a little while. In the meantime Mutant C comes along and joins the fun, making it even hard for A or B to take over the population quickly. This will continue until one of the best mutations is generated and gains dominance, eventually outcompeting everyone else. One of the consequences of clonal interference, therefore, is that it takes longer for a mutation to fix in the population (that is, outcompete everyone else). Another consequence is that it’s usually the best mutants that eventually dominate the population. It turns out that this appears to be exactly what happened in our case of plasmid evolution.

We see many different variants of this gene (some have a point mutation, others a deletion, or a duplication, some are in-frame mutations, some are frame-shift mutations….you name it, we’ve got it!). When I randomly chose ten clones from each of five evolved lineages from each of 11 time points (ancestral population through generation 1,000, in increments of 100 generations), we see up to nine different trfA genotypes present simultaneously, all within the first 400 basepairs of the gene. Talk about diversity! When we used Roche 454 pyrosequencing of whole plasmids with up to 1200X coverage we see over 100 trfA genotypes throughout the 1,000 generations of evolution in one of our lineages. What’s more, it looks like four of five populations were dominated by the same 129-basepair deletion by generation 1,000. It’s too early to be sure, but I wouldn’t be surprised if this particular mutation was just a slight bit bet

ter in some way than many of the other competitors, either that or it got lucky four out of five times.

So what does all that mean? First of all, like I mentioned before, these things adapt fast! So we need to find fast-acting solutions to our antibiotic crisis. On the other hand, we might want certain bacteria to hold on to certain plasmids for a while (think bioremediation!). If we learn more about the tempo and dynamics of plasmid evolution maybe we could use that knowledge to coerce bacteria to do what we wanted them to do. To keep us from getting too cocky, though, the knowledge of how many variants of one gene region can be present in a given population should warn us that bacteria (and plasmids!) have options, and they might not always chose the option that we think is best. In general, though, the more we know about plasmids and how (and when!) they evolve, the better chances we’ll have of combating the spread of antibiotic resistance among pathogenic bacteria.

For more information about Julie’s work, contact her at nich5271 at

This entry was posted in BEACON Researchers at Work and tagged , , , , . Bookmark the permalink.

Comments are closed.