Yeast as a model to study the immunosuppressive and chemotherapeutic drug rapamycin
Rapamycin is a natural product of a soil bacterium and has potent immunosuppressive, and antiproliferative actions. First identified in 1975 as an antifungal drug, rapamycin languished in obscurity after it was found to cause bone marrow suppression [112]. Interest in rapamycin was later rekindled when it was discovered to be structurally related to the potent T-cell inhibitor FK506 [65]. Subsequent studies, conducted first in yeast and then in mammalian cells, revealed the molecular basis of therapeutic action. Rapamycin diffuses into the cell and binds to a small cellular protein, FKBP12, forming an FKBP12-rapamycin protein-drug complex that is the active intracellular agent. This complex then binds to and inhibits the Tor kinases, which function in nutrient sensing pathways that control cell growth and differentiation (Figure 1) (reviewed in [101]). The Tor kinases and FKBP12 are conserved from yeast to worms, flies, plants, and humans. Rapamycin has received FDA approval as an immunosuppressant, and more recently as an antiproliferative agent to inhibit restenosis following cardiac stenting [84]. Recent studies indicate that rapamycin and its analogs will find additional clinical applications as chemotherapeutic agents, topical immunosuppressive therapies in dermatology, and novel antifungal agents [33, 45, 83, 88, 116]. We review here the use of yeast as a model to elucidate the molecular basis of therapeutics for this exciting natural product. The history of the budding yeast Saccharomyces cerevisiae as a model to identify the targets and molecular basis of therapeutic action of rapamycin and other immunosuppressive drugs began with the discovery of rapamycin (sirolimus), cyclosporine A (CsA), and FK506 (tracrolimus) themselves. These immunosuppressive drugs were identified in three independent screens conducted at three different pharmaceutical companies. CsA was discovered at Sandoz Pharmaceuticals in a screen of soil samples for inhibitors of a mixed lymphocyte response (MLR) assay. CsA is produced by a fungus, Tolypocladium inflatum, which was isolated from a soil sample from northern Norway [13]. CsA received FDA approval in 1983 as an immunosuppressant to prevent and treat graft rejection in organ transplant recipients, revolutionized organ transplant therapy, and became the gold standard for immunosuppressive therapies. FK506 was subsequently discovered at Fujisawa Pharmaceuticals in 1987 in a soil sample taken from the Tsukuba region of northern Japan and found to be a macrolide produced by the soil bacterium S. tsukubaensis [65]. FK506 (tacrolimus) received FDA approval in 1994 and has gone on to have considerable impact in transplant medicine. Rapamycin was discovered from a screen for novel natural products in 1975 at Wyeth-Ayerst Pharmaceuticals [112, 125]. In this case, the screen was for antifungal activity, and the rapamycin producing bacterium, S. hygroscopicus, was discovered in an isolate from the beaches of Easter Island (Rapa nui). Rapamycin remains one of the most potent anti-Candida drugs ever discovered [5], however the early finding that it caused bone marrow suppression halted development as an antimicrobial agent. It was only following the discovery of FK506 in 1987, and the appreciation that FK506 and rapamycin are structurally related, that rapamycin was resurrected from the shelf and studies began again in earnest to understand its molecular targets and possible therapeutic applications. Studies in yeast that began in the late 1980s defined the molecular targets of rapamycin, contributed to elucidate the mechanism of action of FK506 and CsA, and fueled comparable studies in mammalian T-cells that led to considerable insights into the molecular basis of therapeutic action (reviewed in [20]). Early studies from Merck demonstrated that FK506 and rapamycin inhibit T-cell proliferation by blocking different signaling pathways [38, 39]. FK506, like CsA, blocked the T-cell antigen response pathway necessary for signaling cascades to drive expression of hundreds of genes required for T-cell activation. In contrast, rapamycin had no effect on the T-cell antigen response pathway, but instead blocked T-cell proliferation in response to interleukin-2 (IL-2). FK506 and rapamycin were found to act as reciprocal antagonists [38, 39], suggesting that the two exerted their actions via a common target. Concurrently, biochemical studies led to the identification of a small abundant cellular binding protein, FK506 binding protein of 12 kDa (FKBP12), which is bound with high affinity to either FK506 or to rapamycin [49, 117]. However, given that FKBP12 is an abundant, ubiquitous protein expressed in all cells in the human body, a general view was that its drug-binding activity might not be involved in a very specific action in lymphocytes. Studies in yeast resolved this dilemma, and established unequivocally the central role of the FKBP12 protein in rapamycin action. Around this time, an FKBP12 homolog that shared 54% amino acid sequence identity with the human FKBP12 protein was identified in the yeast S. cerevisiae [52, 132]. Subsequently the crystal structures were solved for both the yeast and the human protein and found to be essentially superimposable [104, 123]. The gene encoding the yeast FKBP12 protein was cloned and disrupted. Whereas wild-type yeast cells are exquisitely sensitive to growth inhibition by rapamycin, with a minimum inhibitory concentration (MIC) of 25 ng/ml, yeast cells lacking FKBP12 were completely viable and only slightly reduced in growth rate [52, 67, 132]. Thus inhibition of an essential function of FKBP12 could not explain the potent toxic action of FKBP12. Taken together, these findings support a model in which both the FKBP12 protein and its ligand rapamycin are both required to exert a toxic effect in yeast. These findings also established unequivocally that FKBP12 plays a central role in the action of rapamycin, and the next challenge then was to identify the molecular target of the FKBP12-rapamycin complex. The targets of the FKBP12-rapamycin complex, the products of the TOR1 and TOR2 genes (target of rapamycin), were discovered in a genetic screen in yeast searching for rapamycin resistant mutants [51]. Mutations in three different genes were identified. First, mutations in the FPR1 gene encoding FKBP12 were found to be recessive, and resulted from amino acid substitutions that based on structural studies of the FKBP12-rapamycin complex were predicted to be critical for rapamycin binding. Importantly, mutations in two other genes identified were genetically distinct from FPR1, mapped to two different genomic locations and conferred dominant, or semidominant, drug resistance in genetic crosses. Based on an unusual genetic behavior between alleles of these three genes, known as nonallelic noncomplementation, it was proposed that the three might form a physical complex. The TOR1 and TOR2 genes were subsequently cloned by the Hall and Livi laboratories, revealing that they encode extremely large proteins, ∼280 kDa, that Tor1 and Tor2 are homologs of each other, and that both share a C-terminal domain with homology to lipid and protein kinases [18, 54, 71]. Further studies demonstrated that FKBP12-rapamycin forms a physical complex with the yeast Tor1 and Tor2 proteins [22, 76, 118]. Later, work from five different groups converged to identify the mammalian Tor homolog (mTor) via its ability to bind the FKBP12-rapamycin complex [15, 27, 105, 106]. In addition, it was demonstrated that yeast Tor-mTor hybrid genes are capable of providing Tor function in yeast cells [2]. Tor homologs were later identified in other organisms; similar to S. cerevisiae two Tor proteins have been characterized in S. pombe, and a single Tor homolog has been identified in C. albicans, C. neoformans, D. melanogaster, A. thaliana, and H. sapiens [15, 32, 51, 71, 82, 92, 105, 131, 136]. © 2007 Springer.
