Abstract Cyclophilin B belongs to the family of yeast cyclophilin-related peptidyl-prolyl cis-trans isomerases (PPIases), enzymes that catalyse the isomerization of prolyl-peptide bonds in vitro. The protein was originally identified as the deduced translation product of a gene cloned as a homolog of a human cyclophilin A-encoding cDNA. Yeast cyclophilin B is directed to the ER/secretory pathway by a 20 amino acid hydrophobic signal sequence, where it may function to catalyse the folding of secretory proteins. Cyclophilin B is also involved in the heat shock response. Although purified recombinant cyclophilin B protein possesses cyclosporin A (CsA)-sensitive PPIase activity in vitro, cyclophilin B does not play a role in CsA cytotoxicity in yeast, which is mediated instead by cytosolic cyclophilin A.

FK506 and cyclosporin A (CsA) are potent immunosuppressive agents that display antifungal activity. They act by blocking a Ca2+-dependent signal transduction pathway leading to interleukin-2 transcription. Each drug forms a complex with its cognate cytosolic immunophilin receptor (i.e., FKBP12-FK506 and cyclophilin-CsA) which acts to inhibit the Ca2+/calmodulin-dependent protein phosphatase 2B, or calcineurin (CN). We and others have defined the Saccharomyces cerevisiae FKS1 gene by recessive mutations resulting in 100–1000-fold hypersensitivity to FK506 and CsA (as compared to wild type), but which do not affect sensitivity to a variety of other antifungal drugs. The fks1 mutant also exhibits a slow-growth phenotype that can be partially alleviated by exogenously added Ca2+ [Parent et al., J. Gen. Microbiol. 139 (1993) 2973–2984]. We have cloned FKS1 by complementation of the drug-hypersensitive phenotype. It contains a long open reading frame encoding a novel 1876-amino-acid (215 kDa) protein which shows no similarity to CN or to other protein phosphatases. The FKS1 protein is predicted to contain 10 to 12 transmembrane domains with a structure resembling integral membrane transporter proteins. Genomic disruption experiments indicate that FKS1 encodes a nonessential function; fks1::LEU2 cells exhibit the same growth and recessive drug-hypersensitive phenotypes observed in the original fks1 mutants. Furthermore, the fks1::LEU2 allele is synthetically lethal in combina- tion with disruptions of both of the nonessential genes encoding the alternative forms of the catalytic A subunit of CN (CNA1 and CNA2). These data suggest that FKS1 provides a unique cellular function which, when absent, increases FK506 and CsA sensitivity by making the CNs (or a CN-dependent function) essential.

The yeast TOR1 and TOR2 proteins were previously discovered as putative targets of the immunosuppressive drug rapamycin, Although their cellular function is unknown, they are predicted to be at least 215 kDa in size and possess a C-terminal phosphatidylinositol (PI) kinase-related domain. We previously identified a conserved Ser residue, within the PI kinase-related domain of both yeast TOR proteins (Ser(1972) in TOR1; Ser(1975) in TOR2), as being the site of missense mutations conferring dominant rapamycin resistance. The Ser(1972/1975) res. residue of yeast TOR is conserved in mammalian TOR homologs. One possibility is that this residue is critical for a direct interaction between TOR and the FKBP12-rapamycin complex. There is very recent biochemical evidence for an interaction between mammalian TOR and FKBP12-rapamycin (Brown, E. J., Albers, M. W., Shin, T. B., Ichikawa, K., Keith, C. T., Lane, W. S., and Schreiber, S. L. (1994) Nature 369, 756-758; Sabatini, D. M., Erdjument-Bromage, H., Lui, M., Tempst, P., and Snyder, S. H. (1994) Cell 78, 35-43). Using the yeast two-hybrid system, we now have obtained genetic proof of a physical interaction between FKBP12-rapamycin and TOR and have demonstrated that this interaction requires the conserved Ser residue. We have found that a small fragment of wild-type yeast TOR2 spanning Ser(1975) is capable of interacting with human FKBP12 in the presence of rapamycin, whereas an Arg(1975) mutant fails to interact, This effect is dependent upon rapamycin and is antagonized by FK506.

The yeast TOR1 (DRR1) and TOR2 (DRR2) proteins are putative targets of the immunosuppressive drug rapamycin (Rm), defined by dominant drug-resistance mutations. They share a large C-terminal domain that exhibits sequence similarity to the 110-kDa subunit of phosphatidylinositol (PI) 3-kinases. In this report, we present an amino acid (aa) sequence alignment of TOR1 (DRR1) and TOR2 (DRR2) and identify conserved and nonconserved motifs within the N-terminal domain that are indicative of possible nuclear localization. We also show that the mutations responsible for Rm resistance in four independent drr 2(dom) alleles alter the identical aa (Ser(1975)-->Arg) previously identified in drr1(dom) mutants (Ser(1972)-->Arg or Asn). Models for TOR (DRR) protein function are discussed.

Rapamycin (Rm) is a macrolide antifungal agent related to FK506 that exhibits potent immunosuppressive properties which are mediated through interaction with specific cytoplasmic receptors (FKBPs or RBPs, for FK506- and Rm-binding proteins, respectively). These proteins possess peptidyl-prolyl cis-trans isomerase (PPIase) activity in vitro which is inhibited by the binding of Rm and FK506. In Saccharomyces cerevisiae, Rm sensitivity (Rms) is mediated by binding of the drug to RBP1, a homolog of the 12-kDa human FK506-binding protein (FKBP12); null mutations in the yeast RBP1 gene result in a recessive drug resistance phenotype. To identify missense mutations that define amino acid (aa) residues in RBP1 involved in drug sensitivity, we selected and genetically characterized over 250 independent Rmrrbp1 mutants and screened them for both RB1-specific mRNA and protein expression. Whereas all rbp1 mutants expressed abundant levels of RBP1 mRNA, stable RBP1 protein production was detected in only one mutant strain. The RBP1 gene was PCR-generated (in triplicate) from several rbp1 mutants and independent clones were sequenced. Most of the immunoblot-negative alleles were found to contain various types of null mutations; however, some alleles contained specific missense mutations that apparently affect protein stability in vivo. The single immunoblot-positive allele was found to contain a mutation altering a specific residue (Tyr89) which is conserved among the known FKBPs, and which, based on the solution and x-ray structures of human FKBP12, has been proposed to be part of a hydrophobic drug-binding pocket for FK506 and Rm. Analysis of [3H]Rm binding in rbp1 mutant vs. wild-type yeast cell extracts indicated that the Tyr89 → Asp (Y89D) change results in an approx. 95% loss of binding activity. These data confirm the assignment of this Tyr residue as being involved in Rm binding.

Rapamycin is a macrolide antifungal agent that exhibits potent immunosuppressive properties. In Saccharomyces cerevisiae, rapamycin sensitivity is mediated by a specific cytoplasmic receptor which is a homolog of human FKBP12 (hFKBP12). Deletion of the gene for yeast FKBP12 (RBP1) results in recessive drug resistance, and expression of hFKBP12 restores rapamycin sensitivity. These data support the idea that FKBP12 and rapamycin form a toxic complex that corrupts the function of other cellular proteins. To identify such proteins, we isolated dominant rapamycin-resistant mutants both in wild-type haploid and diploid cells and in haploid rbp1::URA3 cells engineered to express hFKBP12. Genetic analysis indicated that the dominant mutations are nonallelic to mutations in RBP1 and define two genes, designated DRR1 and DRR2 (for dominant rapamycin resistance). Mutant copies of DRR1 and DRR2 were cloned from genomic YCp50 libraries by their ability to confer drug resistance in wild-type cells. DNA sequence analysis of a mutant drr1 allele revealed a long open reading frame predicting a novel 2470-amino-acid protein with several motifs suggesting an involvement in intracellular signal transduction, including a leucine zipper near the N terminus, two putative DNA-binding sequences, and a domain that exhibits significant sequence similarity to the 110-kDa catalytic subunit of both yeast (VPS34) and bovine phosphatidylinositol 3-kinases. Genomic disruption of DRR1 in a mutant haploid strain restored drug sensitivity and demonstrated that the gene encodes a nonessential function. DNA sequence comparison of seven independent drr1dom alleles identified single base pair substitutions in the same codon within the phosphatidylinositol 3-kinase domain, resulting in a change of Ser-1972 to Arg or Asn. We conclude either that DRR1 (alone or in combination with DRR2) acts as a target of FKBP12-rapamycin complexes or that a missense mutation in DRR1 allows it to compensate for the function of the normal drug target.

Rapamycin (Rm) and FK506 are macrolide antifungal agents that exhibit potent immunosuppressive properties in higher eukaryotes which are mediated through interaction with specific receptor proteins (FKBPs or RBPs, for FK506- and Rm-binding proteins, respectively). These proteins possess peptidyl-prolyl cis-trans isomerase (PPIase) activity in vitro which is inhibited by the binding of Rm and FK506. We previously isolated a gene encoding an RBP from Saccharomyces cerevisiae, and demonstrated that null mutations in this gene (called RBP1) result in a recessive Rm-resistant (Rm(R)) phenotype. We now have cloned the Candida albicans RBP1 gene via complementation of the Rm(R) phenotype in S. cerevisiae. The predicted C. albicans RBP exhibits 61%, 52% and 49% amino acid (aa) sequence identity with RBPs (FKBPs) from S. cerevisiae Neurospora crassa and human cells (FKBP-12), respectively. Furthermore, several of the aa residues identified as being important for drug binding in human FKBP-12 are conserved within the C. albicans RBP.

Cyclophilins (Cyps) constitute a highly conserved family of proteins present in a wide variety of organisms. Historically, Cyps were first identified by their ability to bind the immunosuppressive agent cyclosporin A (CsA) with high affinity; they later were found to have peptidyl-prolyl cis-trans isomerase (PPIase) activity, which catalyzes the folding of oligopeptides at proline-peptide bonds in vitro and may be important for protein folding in vivo. Cells of Saccharomyces cerevisiae contain at least two distinct Cyp-related PPIases encoded by the genes CYP1 and CYP2. A yeast strain (GL81) containing genomic disruptions of three known yeast PPIase-encoding genes [CYP1, CYP2 and RBP1 (for rapamycin-binding protein); Koltin et al., Mol. Cell. Biol. 11 (1991) 1718–1723] was previously constructed and found to be viable. Soluble fractions of these cells possess residual CsA-sensitive PPIase activity (2–5% of that present in wild-type cells as assayed in vitro). We have purified an approx. 18-kDa protein exhibiting PPIase activity from a soluble fraction of GL81 cells and determined that its N-terminal amino acid (aa) sequence exhibits significant homology (but nonidentity) to the Cyp1 and Cyp2 proteins. We designate the gene for this new protein, CYP3. Using a degenerate oligodeoxyribonucleotide (oligo) based on the N-terminal aa sequence, plus an internal oligo homologous to a conserved region within the portion of CYP1 and CYP2 that had been deleted in the genome, a CYP3-specific DNA fragment was generated by the polymerase chain reaction (PCR) using GL81 genomic DNA as a substrate. This PCR fragment was used as a probe to isolate CYP3 genomic and cDNA clones. Chromosome blots indicate that CYP3 maps to chromosome VIII, and therefore is unlinked to either CYP1 or CYP2. The aa sequence of the deduced protein (Cyp3) exhibits 70% and 48% identity with Cyp1 and Cyp2, respectively. Like Cyp2, Cyp3 contains an N-terminal extension which may play a role in subcellular localization. Haploid cells carrying a genomic disruption of CYP3 are viable, indicating that it is nonessential gene. Furthermore, haploid cells carrying a genomic disruption of CYP1, CYP2, CYP3 and RBP1 are also viable, suggesting either that these enyzmes do not carry out an essential function, or that additional PPIases exist which can compensate for their specific absence.

Rapamycin is a macrolide antifungal agent with structural similarity to FK506. It exhibits potent immunosuppressive properties analogous to those of both FK506 and cyclosporin A (CsA). Unlike FK506 and CsA, however, rapamycin does not inhibit the transcription of early T-cell activation genes, including interleukin-2, but instead appears to block downstream events leading to T-cell activation. FK506 and CsA receptor proteins (FKBP and cyclophilin, respectively) have been identified and shown to be distinct members of a class of enzymes that possess peptidyl-prolyl cis-trans isomerase (PPIase) activity. Despite the apparent differences in their mode of action, rapamycin and FK506 act as reciprocal antagonists in vivo and compete for binding to FKBP. As a means of rapidly identifying a target protein for rapamycin in vivo, we selected and genetically characterized rapamycin-resistant mutants of Saccharomyces cerevisiae and isolated a yeast genomic fragment that confers drug sensitivity. We demonstrate that the resonse to rapamycin in yeast cells is mediated by a gene encoding a 114-amino-acid, approximately 13-kDa protein which has a high degree of sequence homology with human FKBP; we designated this gene RBP1 (for rapamycin-binding protein). The RBP1 protein (RBP) was expressed in Escherichia coli, purified to homogeneity, and shown to catalyze peptidyl-prolyl isomerization of a synthetic peptide substrate. PPIase activity was completely inhibited by rapamycin and FK506 but not by CsA, indicating that both macrolides bind to the recombinant protein. Expression of human FKBP in rapamycin-resistant mutants restored rapamycin sensitivity, indicating a functional equivalence between the yeast and human enzymes.

Cells of Saccharomyces cerevisiae contain a major cytosolic cyclophilin (Cyp)-related peptidyl-prolyl cis-trans isomerase (PPIase) which is the target for cyclosporin A (CsA) cytotoxicity and which is encoded by the CYP1 gene [Haendler et al., Gene 83 (1989) 39-46]. We recently identified a second Cyp-related gene in yeast, CYP2 [Koser et al., Nucleic Acids Res. 18 (1990) 1643] which predicts a protein with a hydrophobic leader sequence. A sequence lacking 33 codons from the 5'-end of the CYP2 open reading frame was generated by the polymerase chain reaction and engineered for expression in Escherichia coli. The corresponding recombinant truncated protein was purified and found to exhibit PPIase activity which was inhibited by CsA. The CYP2 gene is genetically unlinked to CYP1. As with CYP1, genomic disruption of CYP2 had no effect on haploid cell viability. Disruption of all three of the known yeast PPIase-encoding genes [CYP1, CYP2, and RBP1 for rapamycin-binding protein; Koltin et al., Mol. Cell. Biol. 11 (1991) 1718-1723] in the same haploid cell also resulted in no apparent cellular phenotype, suggesting either that none of these enzymes have an essential function or that additional PPIases can compensate for their specific absence. Whereas cells containing a genomic disruption of CYP1 exhibited a CsA-resistant phenotype, genomic disruption of CYP2 had no effect on CsA sensitivity. This suggests that the CYP1 gene product is the primary cellular target for CsA toxicity in yeast. Since both purified Cyps display CsA sensitivity in vitro, our data suggest that Cyp1 and Cyp2 differ in terms of their cellular function and/or localization.