Genes encoding enzymes in the threonine/methionine biosynthetic pathway were cloned and used to investigate their transcriptional response to signals known to affect gene expression on the basis of enzyme specific-activities. Four major responses were evident: strong repression by methionine of MET3, MET5 and MET14, as previously described for MET3, MET2 and MET25; weak repression by methionine of MET6; weak stimulation by methionine but no response to threonine was seen for THR1, HOM2 and HOM3; no response to any of the signals tested, for HOM6 and MES1. In a BOR3 mutant, THR1, HOM2 and HOM3 mRNA levels were increased slightly. The stimulation of transcription by methionine for HOM2, HOM3 and THR1 is mediated by the GCN4 gene product and hence these genes are under the general amino acid control. In addition to the strong repression by methionine, MET5 is also regulated by the general control.

The MET14 gene of Saccharomyces cerevisiae, encoding APS kinase (ATP:adenylylsulfate-3'-phosphotransferase, EC 2.7.1.25), has been cloned. The nucleotide sequence predicts a protein of 202 amino acids with a molecular mass of 23,060 dalton. Translational fusions of MET14 with the beta-galactosidase gene (lacZ) of Escherichia coli confirmed the results of primer extension and Northern blot analyses indicating that the ca. 0.7 kb mRNA is transcriptionally repressed by the presence of methionine in the growth medium. By primer extension the MET14 transcripts were found to start between positions -25 and -45 upstream of the initiator codon. Located upstream of the MET14 gene is a perfect match (positions -222 to -229) with the previously proposed methionine-specific upstream activating sequence (UASMet). This is the same as the consensus sequence of the Centromere DNA Element I (CDEI) that binds the Centromere Promoter Factor I (CPFI) and of two regulatory elements of the PHO5 gene to which the yeast protein PHO4 binds. The human oncogenic protein c-Myc also has the same recognition sequence. Furthermore, in the 270 bp upstream of the MET14 coding region there are several matches with a methionine-specific upstream negative (URSMet) control element. The significance of these sequences was investigated using different upstream deletion mutations of the MET14 gene which were fused to the lacZ gene of E. coli and chromosomally integrated. We find that the methionine-specific UASMet and one of the URSMet lie in regions necessary for strong activation and weak repression of MET14 transcription, respectively. We propose that both types of control are exerted on MET14.

A met4 mutant of Saccharomyces cerevisiae was unable to transcribe a number of genes encoding enzymes of the methionine biosynthetic pathway. The sequence of the cloned MET4 gene allowed the previously sequenced flanking LEU4 and POL1 genes to be linked to MET4 into a 10,327 bp contiguous region of chromosome XIV. From the sequence and mapping of the transcriptional start points, MET4 is predicted to encode a protein of 634 amino acids (as opposed to 666 amino acids published by others) with a leucine zipper domain at the C-terminus, preceded by both acidic and basic regions. Thus, MET4 belongs to the family of basic leucine zipper trans-activator proteins. Disruption of MET4 resulted in methionine auxotrophy with no other phenotype. Transcriptional studies showed that MET4 was regulated by the general amino acid control and hence by another bZIP protein encoded by GCN4. GCN4 binding sequences are present between the divergently transcribed MET4 and LEU4 genes. Over-expression of MET4 resulted in leaky expression from the otherwise tightly regulated MET3 promoter under its control. The presence of consensus sequences for other potential regulatory elements in the MET4 promoter suggests a complex regulation of this gene.

WHZ2 mRNA levels were followed through the growth cycle in WHZ2 mutant and wild-type cells of Saccharomyces cereuisiae. Levels were high during the first (glucose) phase of growth, and were reduced sharply during the second (ethanol) phase of growth. Transcript levels of the glycolytic genes PDCl and PYKl were also measured; they each showed a pattern similar to that of WHZ2, whereas transcript levels of the CDC7 gene remained constant throughout the cycle, showing that a decrease in transcription is not a general feature of genes. These results make it unlikely that the WHZ2 product acts as an inybitor of wll proliferation which is activated upon carbon starvation. No difference was observed between the pattern of expression of mutant and wild-type strains, showing that the mutant phenotype was not the result of a change in regulation at the transcriptional level.

The MET3 gene of Saccharomyces cerevisiae was cloned and its restriction map was found to differ in the upstream region from an earlier published map (Cherest et al. Gene 34, 269-281, 1985) and nucleotide sequence (Cherest et al. Mol. Gen. Genet. 210, 307-313, 1987). Southern blot analysis of genomic DNA from strains S288C and FL100 (the genetic backgrounds from which these different copies of the gene had been cloned) showed that our clone from a S288C-based library had the same restriction map as the chromosomal DNA from both of the strains. Comparison of the nucleotide sequence of the two clones indicated that the earlier published clone probably represented a cloning artifact. In our clone, we found upstream of MET3, the nucleotide sequence of the TDH2 gene (Holland and Holland, J. Biol. Chem. 255, 2596-2605, 1980). The chromosomal orientation of the two genes was determined to be MET3-TDH2-CEN10.