The Path of Yeast Resistance

June 15, 2012

When I started my lab at Princeton in the Fall of 2007, I set out to understand why yeast cells react to the widely prescribed SSRI antidepressant sertraline, which is better known by its Christian name, Zoloft. In those heady days of preliminary data and overreaching models, I encountered the same incredulous reaction to my cocktail party research summary both from informed laypeople  – think monthly subscribers to Scientific American – and expert academic colleagues, which went something like this:

Yeast don’t have a brain or a serotonin transporter or tissue of any type! Why are you giving antidepressants to yeast cells?

Their skepticism was not without reason. At first blush, a drug that was pharmaceutically honed to be more selective than the overdose-prone, preceding generation of antidepressants known as tricyclics had no business being biologically active in the simple yeast cell. So imagine the look of consternation when I proceeded to explain that the inaugural experiment of my lab was a selection for mutant yeast that are resistant to sertraline-induced “overdose.”

I wasn’t the first to observe the unexpected antifungal properties of SSRIs. That honor belongs to a group of Austrian scientists who published back-to-back papers on the subject in 2001. The first paper recounted a serendipitous clinical observation: three women suffering from premenstrual syndrome (PMS) and a Candida yeast infection were given Zoloft. During the course of Zoloft treatment the yeast infection cleared, but the Candida returned after discontinuation of the drug. The second paper went on to show that another yeast species (Aspergillus) was sensitive not only to sertraline but to other SSRIs, too. However, the authors cautioned that the concentration of SSRIs required to kill Candida cells in the lab were much higher than their serum concentrations in patients taking the drug. They went on to speculate that yeast cells might be hypersensitive to Zoloft inside the human body, and they implied that the concentration of Zoloft varies across tissues. The straightforward therapeutic take-home message was: let’s repurpose antidepressants into antifungal agents.

Two years later in 2003, scientists at Pfizer, the pharmaceutical company that originally developed Zoloft, responded to the Austrian studies. The Pfizer group first dutifully observed that yeast cells lack a serotonin system entirely, including therefore the conventional drug target of all SSRIs – the human serotonin transporter, or hSERT. Next, they made several key observations about the underlying structure-activity relationships. First, they found no correlation between hSERT binding and antifungal activity by examining a panel of Zoloft analogs, including diastereomers of Zoloft. Second, they found a strong, Overton-Meyeresque correlation between hydrophobicity and antifungal potency. They concluded the paper by stating:

The present study indicates that the antifungal effect of SSRIs is mediated through a non-specific mechanism related to the lipophilicity of the agents, and is associated with significant toxic effects on human cells at concentrations required for activity against fungal pathogens.

In the final analysis I believe that the Pfizer group’s conclusions were shortsighted, because they didn’t actually provide evidence to support the claim that the antifungal effect is “non-specific.” They reached that conclusion because the antifungal activity of SSRIs is not stereo-specific, merely suggesting that the putative antifungal drug target (a) is not a protein and (b) is very hydrophobic. Sounds a lot like cellular membrane to me…

Skip ahead to 2010 when my lab published in the journal Genetics the first installment of the Zoloft saga, describing the isolation and genetic characterization of Zoloft-resistant yeast mutants. For more on that study, please check out this more technical companion Featured post, which begins to chronicle the history of drug resistance screening in yeast.