Drug resistance is a very important problem in the treatment of bacterial or viral (such as HIV) infections, and in cancer. For example, many bacteria have become resistant to antibiotics such as penicillin, and some bacteria (“superbugs”) are simultaneously resistant to many antibiotics, which makes it very difficult to kill them.
The reason is that pathogenic organisms have a natural ability to evolve and adapt to harsh conditions caused by therapy. Such micro-evolution is driven by random mutations or by exchange of genetic material between microbes.
Although majority of mutations is deleterious, some of them make cells more resistant to drugs. Frequent application of any drug causes the population of pathogenic cells to become enriched in resistant strains, because susceptible pathogens are killed whereas resistant ones thrive. Therefore, any drug becomes inefficient after some time. An important question is how long does it take?
I investigate how bacteria become resistant to antibiotics. In particular, I study how spatial inhomogeneities such as drug
gradients affect the rate of adaptation of bacteria. I have recently published (with P. Greulich and R. Allen) a theoretical paper about this problem and I am currently making some experiments to confirm our theory.
Animation. Evolution of drug resistance takes a completely different course for a uniformly distributed drug (left) than for a drug gradient (right). Uniform drug distribution causes the black strain (susceptible) to evolve into the blue one (resistant) at random places in the environment. This strain may then evolve to a more resistant one (green) etc. In contrast, drug gradient cause the population to evolve in waves of increasingly better adapted mutants and to extend its range in a step-wise manner. The most resistant mutant (red) evolves much faster in the presence of a drug gradient (left, drug concentration increases from bottom to top) than for a uniform drug distribution (right).
I am also interested in simple models (experimental and theoretical) of bacterial infections. For example, we have used the PEE Machine (shown in the picture below) as an experimental model of a urinary tract infection.