A Brief History of Drug Resistance

June 17, 2012

The inaugural experiment of my Princeton lab was my homage to classical microbial genetics, specifically the technique of selecting for drug resistant mutants. For the uninitiated: mutations that confer resistance to a lethal compound often occur in what turn out to be its molecular target (or targets). As I’ll elaborate in the next post in this series, if a “genetic target” is independently confirmed by techniques that measure physical binding of a drug , like autoradiography or affinity chromatography, then it has been fully credentialed, as they say in the business.

The drug resistance approach works spectacularly well with pharmacologically simple compounds, e.g., high-affinity binders to a single protein target.  However, the number of mutational paths to resistance appears to scale with the complexity of molecular interactions with drug targets at or above the lethal drug dose. Two specific cases studies from the annals of yeast genetics provide rich justification for a yeast-based “psych drug overdose” resistance approach, which my lab debuted in a 2010 Genetics paper, hereafter Rainey et al. Both examples involve the work of unsung heroine and second-wave yeast geneticist named Norma Neff.

One of the major results of Rainey et al was our isolation of sertraline-resistant mutations in three subunits of the vacuolar ATPase complex (V-ATPase): VMA1, VMA3 and VMA9. But chemical screens of yeast deletion collections in the early 2000s showed that V-ATPase mutants were non-specifically hypersensitive to drugs, the opposite of what we found. To make sense of this confusion, we need to go back to 1988. Norma Neff’s group at Sloan-Kettering was interested in the mode of action of trifluoperazine,  a psychoactive cationic amphipath, and so chemical relative of sertraline. Specifically, trifluoperazine is a phenothiazine antipsychotic that descends from chlorpromazine (Thorazine®), the first popularized antipsychotic drug. Neff’s group published a paper in which they cloned a novel trifluoperazine-resistance gene, TFP1. Based on sequence homology to a subunit of the energy-producing  mitochondrial F1/F0 ATPase, Neff’s lab correctly concluded that TFP1 was (a component of) a proton transporter, as shown in a reproduction of Figure 7 from the paper: (right). Therefore, sertraline resistance and trifluoperazine resistance share a common genetic modifier, the V-ATPase.

The only criticism I have with this otherwise well-executed paper is the interpretation of why mutations in TFP1/VMA1 resulted in trifluoperazine resistance. The paper concludes:

The regulation of the activity of such ion pumps may contribute to intracellular traffic and organelle function, because local metabolic activity and protein targeting may be influenced by changes in the different electrochemical potentials across the different membranes which enclose cellular compartments.”

This gets it half right, I think. Neff clearly understood that acidification was the salient physiological process underlying trifluoperazine resistance. But the most parsimonious explanation is lysosomotropism, as I argued in Rainey et al. For more recent literature on lysosomotropism in a mammalian cell model, see the excellent work of Jeff Krise’s lab at the University of Kansas.

The second, totally unexpected finding of Rainey et al was the observation that the reduced fitness caused by a point mutation in the heavy chain of clathrin (CHC1E292K) is suppressed by low doses of sertraline. In other words, the same drug used to select for drug-resistant mutants at high doses enhances the growth of one particular sick drug-resistant mutant at low doses, as originally shown in our paper: (right)

Turns out Neff and associates, working in the lab of genetics impresario David Botstein (now at Princeton, then at MIT), published in 1985 a lovely Genetics paper describing almost exactly the same phenomenon, except in their case they selected for mutants resistant to the microtubule depolymerizer benomyl. Benomyl insinuates itself into tubulin and prevents yeast cells from segregating their chromosomes, a lethal blow at high doses. One benomyl-resistant mutant in particular, tub2-150, exhibited “benomyl dependence,” that is this benomyl-resistant mutant actually grew better in the presence of benomyl, especially in the cold. They rationalized benomyl dependence as follows:

The benomyl-dependent mutations described here fit this model well. They could have a defect that renders their microtubules more stable than wild type, as has been demonstrated for a tubulin mutation in A. nidulans. Such a defect could be compensated for by a chemical destabilizing agent, such as benomyl, or by a physical one, such as low temperatures.”

If you substitute clathrin for microtubules and sertraline for benomyl, one can begin to understand why sub-lethal doses of sertraline rescue the fitness defect caused by altered clathrin function. But how exactly sertraline accumulation in vesiculogenic membranes affects clathrin function in yeast cells is an ongoing focus of research efforts in the lab.

  • http://twitter.com/MerWright13 Meredith Wright

    Really interesting! It never occurred to me that organisms could mutate such that a drug against them will actually allow them to grow better.

    • http://twitter.com/eperlste Ethan Perlstein

      Thanks for reading! It didn’t occur to me either when we started, so I was gratified to find an independent example involving a different drug and a different drug target.