In a report published in PLoS ONE, a journal published by the Public Library of Science on April 18, the researchers showed that Zoloft accumulates in the cell membranes of baker’s yeast cells and triggers autophagy, a protective mechanism whereby cells degrade their own parts and essentially eat themselves. Though this process can be triggered under natural conditions, autophagy as a response to Zoloft is a striking result that suggests the drug may have a number of unexpected side effects, especially given that yeast cells lack serotonin transporters, the usual target of Zoloft.
Perlstein first got involved with this work as a graduate student at Harvard, where he used yeast as a model system to study the genetic complexity of drug response to many compounds. While at Harvard, he introduced a variety of drugs to yeast cells and found that Zoloft was particularly potent at causing cytotoxicity for the yeast cells, a term Perlstein compared to an “overdose.”
“I like to say ‘overdose’ because I think it captures something about these psych drugs and what they are doing,” Perlstein said.
Upon arriving at the University in 2007, Perlstein’s lab began focusing on Zoloft and looking for mutant yeast cells that were resistant to the drug. He classified Zoloft as a selective serotonin reuptake inhibitor, a class of psychoactive drugs intended to be less toxic and safer than those of earlier generations of drugs.
“So these SSRIs were supposed to be safe, and yet we’re finding that these yeast are overdosing quite strongly in response to Zoloft,” he said.
Sara Haddock ’12 works in Perlstein’s lab, and her thesis work centered around finding yeast cells resistant to the drug and sequencing their genomes to determine mutations causing resistance. Such sequencing would determine cellular processes affected by the drug. Her work, like Perlstein’s recent paper, centers on the idea that Zoloft has non-serotonin-induced effects.
“The idea here is yeast do not contain the serotonin transporter, which is what the drug was designed to hit, so whatever Zoloft does in the yeast cell is an unintended effect,” Haddock said in an email. “This effect may also occur in the human body — but obviously, more research would need to be done to confirm that.”
Haddock said she will continue her research over the summer. Perlstein’s next steps are to test his findings in mice and explore other classes of drugs.
“The future direction of the entire approach my lab uses, which I call ‘evolutionary pharmacology,’ is to start off in yeast and then start to move up the complexity chain and look at more conventional, traditional disease models,” Perlstein said. “We’re using this paper as an opportunity to eventually get collaboration with mouse neuroscientists who have studied depression using traditional rodent models. We want to take this observation from yeast and integrate it with what they’ve been doing.”
Mark Rose, professor and undergraduate department representative for the molecular biology department, who also uses yeast as a model organism, said Perlstein’s work was interesting but needed further study to solidify its relevance.
“Perlstein makes a good case that the mechanism by which sertraline/Zoloft and other related compounds act is still not completely understood,” Rose said in an email. “To that extent it is certainly interesting that the yeast cells are sensitive to sertraline.”
Rose expressed concern about the concentration of Zoloft used for the various experiments in the study, saying that some results may not be clinically relevant. Perlstein added that his experiments following the publication of this paper were carried out at a lower concentration of Zoloft.

“It is not yet clear whether the effects that Perlstein is seeing in yeast will also be observed in human cells,” he said. “It will be very interesting to see how this research plays out over the next few years.”