A major problem in the chemotherapy of cancer is the appearance of drug resistant tumor cells. Often these drug resistant tumors exhibit broad range tolerance to several structurally unrelated cytotoxic drugs. This ability to tolerate a variety of different drugs is referred to as multidrug resistance.

     One of the best understood mechanisms behind the acquistion of multidrug resistance involves the overproduction of an ABC transporter protein designated MDR1. This membrane protein acts as an ATP-dependent drug pump and prevents the accumulation of toxic levels of target compounds. Saccharomyces cerevisiae also exhibits loci that can be altered to give rise to multidrug resistance (referred to in this organism as pleiotropic drug resistance). Studies by many groups have identified a range of genes involved in Pdr encoding several different proteins including zinc finger transcription factors (PDR1, PDR3) and ABC transporter proteins (YOR1, PDR5). Two different drugs have been used to define Pdr genes. Cycloheximide is a translation inhibitor while oligomycin is an Fo ATPase inhibitor and is only toxic to cells growing on a non-fermentable carbon source. YOR1 (yeast oligomycin resistance 1) only contributes to oligomycin tolerance while PDR5 only contributes to cycloheximide resistance. PDR1 and PDR3 affect resistance to both of these agents.

     Our studies initially focused on establishing the link between the transcriptional regulatory proteins Pdr1p and Pdr3p and the ABC transporter-encoding genes YOR1 and PDR5. We found that both Pdr1p and Pdr3p bind to an element which was designated the Pdr1p/Pdr3p response element (or PDRE). The PDRE consensus sequence is TTCCGCGGAA and can be found in the promoter region of all genes under control of Pdr1p and/or Pdr3p.

     Originally, the Pdr phenotype in S. cerevisiae was defined by mutant forms of PDR1 giving rise to resistance to a wide range of toxic drugs, including both cycloheximide and oligomycin. Experiments in other labs first showed that these dominant mutant forms of PDR1 caused PDR5 mRNA to dramatically increase compared to cells carrying a wild-type PDR1 allele. Genetic studies in our lab and others demonstrated that loss of PDR5 eliminated the ability of Pdr1p to confer cycloheximide resistance on cells but had no effect on Pdr1p-mediated oligomycin resistance. These data suggested that a second target gene existed that was responsible for Pdr1p-mediated oligomycin resistance. To identify this oligomycin resistance gene, we screened a high-copy plasmid library for genes that would elevated oligomycin tolerance when present in multiple copies. Two loci were recovered from this screen: YOR1 and PDR13. YOR1 was found to encode an ABC transporter protein whose expression was required for Pdr1p to confer oligomycin resistance on cells. We found that PDR13 encodes a novel Hsp70 protein of previously unknown function. The available data suggest that Pdr13p acts by stimulating the function of the Pdr1p transcriptional regulatory protein.

Current Research Efforts

     Although the work described above provided a reasonable understanding of the details by which Pdr1p and Pdr3p regulated target gene expression, little was known of the possible means of control of the activity of these transcription factors. Single amino acid substitution mutations have been described in both the PDR1 and PDR3 genes that cause the resulting mutant factors to behave as hyperactive regulators of both gene expression and drug resistance. These findings suggest the possibility that the mutations relieve some negative regulation that would normally depress the activity of these proteins. To explore this possibility, we carried out a genetic screen searching for transposon insertion mutations that would elicit overproduction of PDR5 mRNA with associated cycloheximide hyper-resistance. We recovered two different insertions that fulfill both of these criteria. Loss of the FZO1 or OXA1 genes dramatically elevated both cycloheximide tolerance and PDR5 expression. FZO1 encodes a large GTPase involved in mitochondrial fusion and is required for normal maintenance/transmission of the mitochondrial genome. Cells that lack FZO1 rapidly lose their mitochondrial genome and are subsequently referred to as rho zero (rho0) cells. Oxa1p is required for assembly of both the cytochrome c oxidase complex and the F1Fo ATPase but importantly oxa1 mutants are not rho0. This finding was the first indication that the signalling in oxa1 and rho0 cells was likely to have unique properties. Further experiments have demonstrated that depletion of the the Fo subcomplex of the mitochondrial ATPase seems to be a key feature in generating the mitochondrial signal that ultimately leads to activation of PDR5expression. Intriguingly, this mitochondrial signal is exclusively received by Pdr3p while Pdr1p does not participate in this regulatory system.

     We are currently attempting to determine the nature of the signal that is detected by Pdr3p. Additionally, genetic searches for additional participants in the pathway leading from the mitochondria to the nucleus are underway to learn how the loss of normal mitochondrial function is transmitted to the nuclearly-localized Pdr3p. It is important to note that a similar phenomenon of loss of mitochondrial function leading to strong drug resistance has been described for the pathogenic yeast, Candida glabrata. Study of this pathway in the genetically tractable yeast Saccharomyces cerevisiae will provide rapid progress towards understanding the physiological relevance of the connection between mitochondria and PDR gene expression.