A position effect in the control of transcription at yeast mating type loci

Gene activation and re-expression of a Trypanosoma brucei variant surface glycoprotein

The expression of the Trypanosoma brucei variant surface glycoprotein AnTat 1.1 proceeds by a mechanism that transfers a duplicated gene copy into a new genomic environment, the so-called expression site, where it will be expressed. We have isolated a genomic fragment containing the region spanning the expression site-transposon junction, and the 5' half of the coding sequence. Comparing this DNA segment with its template copy (basic copy) allowed us to identify the exact breaking point and indicated a base sequence which could be involved in initiating the transposition event. Sequencing data also indicated that the co-transposed segment 5' to the coding sequence is 430 bp in length. The extreme 5' end of the mRNA is derived from a region in the expression site not immediately adjacent to the transposed DNA segment. This particular sequence exists in multiple copies in the genome and is common to the mRNA of all variant surface glycoproteins so far analysed.

Plasmid-directed synthesis of hepatitis B surface antigen in monkey cells

We introduced the gene encoding the hepatitis B virus surface antigen (HBsAg) into simian virus 40 (SV40)-based plasmids capable of autonomously replicating in both Escherichia coli and permissive monkey cells. After introduction into monkey cells by transfection, these plasmids directed the synthesis of high levels of HBsAg, as determined by immunofluorescence, radioimmunoassays, and identification by sodium dodecyl sulfate-polyacrylamide gel electrophoresis of the polypeptides comprising the antigen. Expression was dependent upon the presence of an SV40 promoter, with both the early and late promoters able to effectively initiate transcription. Using expression of HBsAg to assay promoter function, we demonstrated that an intact copy of the SV40 72-base pair repeat, which constitutes an essential element of the SV40 early promoter during the lytic SV40 cycle and which can enhance the transcriptional activity of heterologous promoters, was not required for HBsAg expression, suggesting that the hepatitis genome contains an enhancer element capable of complementing that provided by the 72-base pair repeat element of SV40. The antigen appears to be glycosylated after synthesis in transfected cells and is apparently secreted, as evidenced by the localization of [35S]cysteine-labeled antigen to the medium of transfected cultures. Using constructions in which the first ATG sequence appearing in HBsAg mRNA was that corresponding to the gene encoding the mature form of the antigen, we demonstrated that these post-translational events could occur without the involvement of a putative precursor peptide suggested by the DNA sequence of the viral genome. In view of the inability of hepatitis B virus to propagate in vitro, this strategy offers a convenient approach for further characterizing the biosynthesis of this antigen and may provide a means to identify additional polypeptides encoded by this virus.

Characterization of a regulatory region upstream of the ADR2 locus of S. cerevisiae

We have used in vitro mutagenesis to identify a DNA sequence that is required for glucose regulation of the ADR2 locus of Saccharomyces cerevisiae. This region, located between 200 and 1,000 base pairs upstream of the structural gene for ADH II, has been identified by genetic and biochemical analysis as the site of ADR3, a cis-acting regulatory locus. If this region is deleted, ADR2 is no longer repressed by glucose. Moreover, if the sequence is excised and ligated in front of the ADC1 gene, it puts that gene under glucose control.

The yeast his3 promoter contains at least two distinct elements

Phenotypic analysis of 65 mutations indicates that the yeast his3 promoter is composed of at least two separate regions of DNA. Each is necessary, but neither is sufficient for wild-type levels of his3 expression. Deletion mutations that destroy either promoter element express his3 poorly or not at all. The upstream element is located between 112 and 155 base pairs before the site of transcriptional initiation (nucleotides -112 to -155). A comparison of derivatives strongly suggests that the downstream element maps somewhere between nucleotides -32 and -52 and includes a sequence between nucleotides -45 and -52. This location coincides with sequences conserved before most eukaryotic genes(the TATA box region). By using derivatives in which his3 sequences are replaced by a small fragment of coliphage M13 DNA, three properties of the his3 promoter were established. First, his3 TATA box deletions fail to express his3 because they lack specific sequences and not because they disrupt spacing relationships between other sequences. Second, the TATA box region can be replaced functionally by the one orientation of the M13 DNA fragment that contains a TATA-like sequence. Third, the distance between the two elements (normally 90 base pairs) can be varied between 40 and 160 base pairs without markedly affecting promoter function. These results strongly suggest that yeast RNA polymerase II, unlike its Escherichia coli counterpart, does not bind simultaneously to both promoter elements, and they add further support to the view that the upstream element is not part of a transcriptionally competent binding site. This ability of the his3 upstream promotor element to act at a long and variable distance is similar to properties of viral enhancer sequences and is reminiscent of position effects in yeast.

Early role of the esc+ gene product in the determination of segments in Drosophila

Recent studies of the esc+ gene suggest that its product acts principally in the initiation, but not the maintenance, of segmental determination. We have tested this hypothesis by examining when the esc+ gene product is required during embryogenesis, and whether the esc+ gene product is required for the stable determination of imaginal disc cells passaged in in vivo culture. We find that the esc+ gene product is required during a discrete period of embryogenesis. If the gene product is absent or inactive during this period, most segments develop like the eighth abdominal segment. In contrast, absence of active gene product before or after this period has relatively little effect on segmental development. These results support the hypothesis that the esc+ gene product has a discrete early function in the initiation of segmental determination.

rme1 Mutation of Saccharomyces cerevisiae: map position and bypass of mating type locus control of sporulation

Sporulation in Saccharomyces cerevisiae normally occurs only in MATa/MAT alpha diploids. We show that mutations in RME1 bypassed the requirements for both a and alpha mating type information in sporulation and therefore allowed MATa/MATa and MAT alpha/MAT alpha diploids to sporulate. RME1 was located on chromosome VII, between LEU1 and ADE6.

Structure of the Schizosaccharomyces pombe cytochrome c gene

The cytochrome c gene of the fission yeast Schizosaccharomyces pombe has been cloned by using the Saccharomyces cerevisiae iso-1-cytochrome c gene as a molecular hybridization probe. The DNA sequence and the 5' termini of the mRNA transcripts of the gene have been determined. The DNA sequence has confirmed, with two exceptions, the previously determined protein sequence. The nonrandom distribution of silent third base differences which was observed between the two cytochrome c genes of S. cerevisiae does not extend to the S. pombe cytochrome c gene, suggesting that there are no constraints other than protein function and codon usage which have acted to conserve the cytochrome DNA sequences of the two yeasts. Introduction of the S. pombe cytochrome c gene on a yeast plasmid into a S. cerevisiae mutant which lacked functional cytochrome c transformed that recipient strain for the ability to grow on a nonfermentable carbon source. This implies that the S. pombe cytochrome c gene has all the regulatory signals which are required for its expression in S. cerevisiae, and that none of the amino acid differences between the cytochrome c proteins of the two yeasts has a drastic effect on the function of the protein in vivo.

Preliminary characterization of the transcriptional and translational products of the Saccharomyces cerevisiae cell division cycle gene CDC28

The CDC28 gene was subcloned from a plasmid containing a 6.5-kilobase-pair segment of Saccharomyces cerevisiae DNA YRp7(CDC28-3) by partial digestion with Sau3A and insertion of the resulting fragments into the BamHI sites of YRp7 and pRC1. Recombinant plasmids were obtained containing inserts of 4.4 and 3.1 kilobase pairs which were capable of complementing a cdc28(ts) mutation. R-loop analysis indicated that each yeast insert contained two RNA coding regions of about 0.8 and 1.0 kilobase pairs, respectively. In vitro mutagenesis experiments suggested that the smaller coding region corresponded to the CDC28 gene. When cellular polyadenylic acid-containing RNA, separated by agarose gel electrophoresis after denaturation with glyoxal and transferred to nitrocellulose membrane, was reacted with labeled DNA from the smaller coding region, and RNA species of about 1 kilobase in length was detected. Presumably, the discrepancy in size between the R-loop and electrophoretic determinations is due to a segment of polyadenylic acid which is excluded from the R-loops. By using hybridization of the histone H2B mRNAs to an appropriate probe as a previously determined standards, it was possible to estimate the number of CDC28 mRNA copies per haploid cell as between 6 and 12 molecules. Hybrid release translation performed on the CDC29 mRNA directed the synthesis of a polypeptide of 27,000 daltons, as determined by polyacrylamide gel electrophoresis in sodium dodecyl sulfate. This polypeptide was not synthesized when mRNA prepared from a cdc28 nonsense mutant was translated in a parallel fashion. However, if the RNA from a cell containing the CDC28 gene on a plasmid maintained at a high copy number was translated, the amount of in vitro product was amplified fivefold.

Sporulation and rna2 lower ribosomal protein mRNA levels by different mechanisms in Saccharomyces cerevisiae

In Saccharomyces cerevisiae, the levels of ribosomal protein mRNAs are regulated coordinately. Vegetative strains carrying the temperature-sensitive rna2 mutation exhibit a dramatic decrease in the levels of most ribosomal protein mRNAs at the restrictive temperature. Similarly, in wild-type cells induced to sporulate by nitrogen starvation, there is a fivefold reduction in the relative synthesis rate of ribosomal proteins. Using Northern gel analysis and cloned ribosomal protein genes, we compared the way in which ribosomal protein mRNA is affected under these two conditions. In vegetative rna2 cells, incubation at 34 degrees C led to the disappearance of ribosomal protein mRNAs and the accumulation of higher-molecular-weight precursor RNAs. A different phenotype was observed during sporulation. Although sporulating conditions led to a significant reduction in the relative abundance of ribosomal protein mRNA, there was no detectable accumulation of precursor RNAs even in rna2/rna2 diploids at 34 degrees C. A suppressor of rna2 and of other rna mutations, SRN1, at least partially relieved the block in the splicing of the ribosomal protein 51 intron in vegetative rna2 cells but did not detectably affect the level of ribosomal protein mRNA in sporulating cells. We concluded that the rna2 mutation and sporulation conditions affected ribosomal protein mRNA metabolism in two quite different ways. In vegetative cells the mutant rna2 effected a block which occurred primarily in post-transcriptional processing, whereas in sporulating cells the ribosomal protein mRNA levels were decreased by some other mechanism, presumably a change in the relative rate of transcription or mRNA turnover. Furthermore, the data suggest that the mutation rna2 has no additional effect on ribosomal protein mRNA metabolism in sporulating cells.

Reversion of a promoter deletion in yeast

Promoter function in yeast has been examined by obtaining revertants of a his3 promoter deletion in vivo. Events which provide a new promoter for the his3 gene include insertion of the transposable element Ty1, rearrangements of the plasmid vector, and chromosomal mutations. A role for dicentric chromosomes as a source of the plasmid rearrangements is discussed.

Transcriptional control signals of a eukaryotic protein-coding gene

Transcriptional control signals of a model eukaryotic protein-coding gene have been identified by a new procedure of in vitro mutagenesis. This method allows small clusters of nucleotide residues to be substituted in a site-directed manner without causing the addition or deletion of other sequences. Transcription assays of a systematic series of these clustered point mutants have led to the identification of three distinct control signals located within the 105-nucleotide residues immediately upstream from the point where transcription begins.

RNA splicing is required to make the messenger RNA for a variant surface antigen in trypanosomes

The expression of the gene for variant surface glycoprotein (VSG) 118 in Trypanosoma brucei is activated by transposing a DNA segment containing the gene and 1-2 kb in front of it to an expression site elsewhere in the genome. By S1 nuclease protection and RNA blotting experiments we show here the presence of several minor transcripts in trypanosomes synthesizing VSG 118, one of which covers the entire transposed segment. Comparison of the sequence of the 5' terminal segment of VSG 118 messenger RNA (mRNA), determined by primed reverse transcription, and the corresponding region of the 118 VSG gene, shows that the 5' terminal 34 nucleotides of the mRNA are not encoded in the 118 VSG gene contiguous with the remainder of the mRNA. We conclude that synthesis of a VSG mRNA involves splicing of a much longer primary transcript, which may start outside the transposed segment.

The establishment of genomic DNA libraries for the human malaria parasite Plasmodium falciparum and identification of individual clones by hybridisation

The DNA of Plasmodium falciparum has been purified and fragmented with the restriction endonucleases EcoRI and HindIII. The fragments have been incorporated in vitro into derivatives of bacteriophage lambda to make libraries in which most of the parasite DNA is represented. By Southern hybridisation we have been able to recover from these libraries specific clones containing (a) repetitive DNA sequences, (b) rRNA gene(s) and (c) sequences homologous to an actin gene probe. Parasite DNA from two independent sources differs markedly in the pattern of its repetitive DNA visualised by hybridisation to our repetitive clone. By contrast, the rRNA genes of the two isolates prove to be carried on identically sized fragments.

Structure and expression of a Trypanosoma brucei gambiense variant specific antigen gene

The expression-linked copy of the T. b. gambiense variant specific antigen gene LiTat 1.6 is transposed in a 20 kb DNA region devoid of restriction sites, located near a chromosome end. This expression site is very similar to that of T. b. brucei variants 117, 118 (1) and AnTat 1.8. In the basic copy, the transposable element (TE) is flanked by repetitive sequences; it includes the gene copy as well as a sequence of 0.9 to 2.1 kb (probably around 1.1 kb) long, upstream from the gene. Probes derived from the 5' part of the TE specifically reveal three polyadenylated transcripts of 4.2, 1.45 and 0.85 kb, respectively, distinct from the 2.1 kb mRNA. The amount of the 4.2 kb sequence is probably less than 0.01% of total trypanosome RNA. Whereas the mRNAs coding for the three isotypic antigens AnTat 1.8 (T. b. brucei), 12.2 (T. b. rhodesiense) and 3.3 (T. evansi) are recognized by LiTat 1.6 probes extending into the 3' half of the transposed sequence, the 5' genomic probes do not hybridize with any of these RNAs. These observations suggest that the LiTat 1.6 gene could be first transcribed in a large precursor molecule. This precursor would be rapidly processed, loosing a large portion of less conserved sequence from its 5' half. Our data are compatible with a model in which the promoter would be provided by the expression site.

Preliminary characterization of the transcriptional and translational products of the Saccharomyces cerevisiae cell division cycle gene CDC28

The CDC28 gene was subcloned from a plasmid containing a 6.5-kilobase-pair segment of Saccharomyces cerevisiae DNA YRp7(CDC28-3) by partial digestion with Sau3A and insertion of the resulting fragments into the BamHI sites of YRp7 and pRC1. Recombinant plasmids were obtained containing inserts of 4.4 and 3.1 kilobase pairs which were capable of complementing a cdc28(ts) mutation. R-loop analysis indicated that each yeast insert contained two RNA coding regions of about 0.8 and 1.0 kilobase pairs, respectively. In vitro mutagenesis experiments suggested that the smaller coding region corresponded to the CDC28 gene. When cellular polyadenylic acid-containing RNA, separated by agarose gel electrophoresis after denaturation with glyoxal and transferred to nitrocellulose membrane, was reacted with labeled DNA from the smaller coding region, and RNA species of about 1 kilobase in length was detected. Presumably, the discrepancy in size between the R-loop and electrophoretic determinations is due to a segment of polyadenylic acid which is excluded from the R-loops. By using hybridization of the histone H2B mRNAs to an appropriate probe as a previously determined standards, it was possible to estimate the number of CDC28 mRNA copies per haploid cell as between 6 and 12 molecules. Hybrid release translation performed on the CDC29 mRNA directed the synthesis of a polypeptide of 27,000 daltons, as determined by polyacrylamide gel electrophoresis in sodium dodecyl sulfate. This polypeptide was not synthesized when mRNA prepared from a cdc28 nonsense mutant was translated in a parallel fashion. However, if the RNA from a cell containing the CDC28 gene on a plasmid maintained at a high copy number was translated, the amount of in vitro product was amplified fivefold.

Structure of the Schizosaccharomyces pombe cytochrome c gene

The cytochrome c gene of the fission yeast Schizosaccharomyces pombe has been cloned by using the Saccharomyces cerevisiae iso-1-cytochrome c gene as a molecular hybridization probe. The DNA sequence and the 5' termini of the mRNA transcripts of the gene have been determined. The DNA sequence has confirmed, with two exceptions, the previously determined protein sequence. The nonrandom distribution of silent third base differences which was observed between the two cytochrome c genes of S. cerevisiae does not extend to the S. pombe cytochrome c gene, suggesting that there are no constraints other than protein function and codon usage which have acted to conserve the cytochrome DNA sequences of the two yeasts. Introduction of the S. pombe cytochrome c gene on a yeast plasmid into a S. cerevisiae mutant which lacked functional cytochrome c transformed that recipient strain for the ability to grow on a nonfermentable carbon source. This implies that the S. pombe cytochrome c gene has all the regulatory signals which are required for its expression in S. cerevisiae, and that none of the amino acid differences between the cytochrome c proteins of the two yeasts has a drastic effect on the function of the protein in vivo.

Characterization of the DNA duplication-transposition that controls the expression of two genes for variant surface glycoproteins in Trypanosoma brucei

The genome of Trypanosoma brucei carries over a hundred genes coding for different variants of the major surface glycoprotein. Activation of some of these genes is accompanied by a duplication and transposition of the gene (the basic copy) to another region in the genome where it is transcribed. We present here physical maps of the basic and transposition-activated genes for two surface glycoproteins of Trypanosoma brucei, stock 427. In both cases the transposed segment starts 1-2 kb in front of the coding region and ends within the 3'-terminal region of the gene. The DNA segments flanking both transposed genes are indistinguishable and share a 6-kb stretch upstream and a 8-kb stretch downstream of the transposed segment not cut by several restriction endonucleases. The 5' borders of the two transposed segments are homologous and contain sequences present in many copies in the genome. A different repeated sequence has previously been found at the 3' edge of the transposed segment. The replicative transposition may, therefore, involve a unidirectional gene conversion initiated by base pairing between the edges of the transposed sequence and a single expression site elsewhere in the genome.

Multiple arrangements of viral DNA and an activated host oncogene in bursal lymphomas

Proviruses of avian leukosis virus (ALV) are located in the vicinity of a putative cellular oncogene (c-myc) in ALV-induced bursal lymphomas. Enhanced expression of c-myc occurs in association with proviruses found in any of three configurations: (I) on the 5' side ('upstream') of c-myc in the same transcriptional orientation; (II) on the 3' side ('downstream') of c-myc in the same orientation; (III) upstream, in the transcriptional orientation opposite to that of c-myc. Thus, activation of adjacent cellular genes by retroviral DNA can involve mechanisms other than provision of a transcriptional promoter.

Activation of trypanosome surface glycoprotein genes involves a duplication-transposition leading to an altered 3' end

Expression of the genes for variant surface glycoproteins 117 and 118 in Trypanosoma brucei is accompanied by the appearance of an extra copy of these genes, the expression-linked copy, which differs in the surrounding restriction enzyme sites from the corresponding basic copy of the genes. We present direct evidence that the expression-linked copy is the one used for messenger RNA synthesis. By S1-nuclease-protection experiments we show that cloned basic-copy genes contain the nucleotide sequence of the corresponding messenger RNA except for the last 100 to 150 nucleotides before the poly(A) tail. Comparison of the 3'-terminal sequence of the 117 basic-copy gene and the 117 complementary DNA shows that this region differs by multiple point mutations, insertions and deletions, the differences starting within the coding sequence. Genomic blots demonstrate that a Bsp I site in the 3'-terminal part of the 118 complementary DNA is present in the expression-linked copy but not in the basic-copy gene. We conclude that expression-linked copies are the active genes, and that the generation of expression-linked copies involves a duplication--transposition in which the 3' end of the gene is replaced.

rme1 Mutation of Saccharomyces cerevisiae: map position and bypass of mating type locus control of sporulation

Sporulation in Saccharomyces cerevisiae normally occurs only in MATa/MAT alpha diploids. We show that mutations in RME1 bypassed the requirements for both a and alpha mating type information in sporulation and therefore allowed MATa/MATa and MAT alpha/MAT alpha diploids to sporulate. RME1 was located on chromosome VII, between LEU1 and ADE6.

Analysis of the DNA and RNA changes associated with the expression of isotypic variant-specific antigens of trypanosomes

Using specific (32P) labelled cDNA probes, we compared the mRNAs and the genomic DNA sequences coding for the synthesis of two pairs of serologically related variant-specific antigens (VSAs) of trypanosomes: AnTat 1.1 and AnTat 1.1b, both from the strain 1125 of T.b.brucei and AnTat 1.8 and LiTat 1.6 from T.b.brucei and T.b. gambiense, respectively. Within each pair, large similarities were observed in the coding sequence, except in the 3' region which appears to be highly variable. However, a low level of cross-hybridization can be detected between all sequences, in the 3' region only. The expression of these VSAs is linked to a similar duplication-transposition mechanism. The insertion locus of the transposition unit is the same both in AnTat 1.1 and AnTat 1.1b DNAs. In both pairs, the transposition unit seems to comprise at least about 200 bp upstream of the 5' extremity of the coding sequence. The significance of these results, regarding the structure and synthesis of the VSAs, is discussed.

A position-effect control for gene transposition: state of expression of yeast mating-type genes affects their ability to switch

Mating-type switches of the yeast Saccharomyces cerevisiae occur by unidirectional transposition of copies of unexpressed mating-type genetic information, residing at HML and HMR loci, into the expressed MAT locus. The HML and HMR loci remain unchanged. In contrast, in appropriate strains where the silent loci are also allowed to express, for example in mar mutants, efficient switches of HML and HMR are shown to occur at rates equivalent to those observed for MAT. Thus the position-effect control on the direction of transposition is affected by the state of expression of the locus under study the expressed loci switch regardless of their location.

Structural comparison of two yeast tRNA Glu 3 genes

DNA sequences in a 1.7 kb Pst fragment from yeast have been determined. This fragment is part of a yeast 7.4 kb Hind III segment cloned ino pBR322 (pY 5). The fragment carries a single gene for a glutamate tRNA. The coding portion of this gene is identical in sequence to that of the tRNA Glu 3 gene from pY 20 [1]. The flanking regions differ in their sequences, but possible secondary structures within the 5'-flanking regions bear similar features. Sequence homologies between pY 5 and pY 20 were detected far outside the tRNA genes. More surprisingly, extended sequence homologies were seen between the flanking regions of the pY 20 tRNA Glu 3 gene and a tRNA Ser gene [2,3]. We have also checked the known tRNA genes for structural similarities. Hybridization studies indicate that portions of the Pst fragment are repeated within the yeast genome.

Deletion mapping of sequences essential for in vivo transcription of the iso-1-cytochrome c gene

The 5' termini of yeast CYC1 RNA molecules have been mapped, by nuclease S1 digestion of mRNA . DNA duplexes, to seven locations from 29 to 93 base pairs upstream from the initiating ATG codon. When the CYC1 gene is introduced into yeast in plasmid YEp13, substantially the same transcripts are made. Using this system to study in vivo gene expression, we measured the capacity of enzymatically produced DNA deletions to form the normal set of RNAs. Four regions of 5'-flanking DNA were identified as functional. Sequences within the region -242 to -139 are required for maximal CYC1 transcript formation; their deletion reduces transcription by a factor of 15 but does not change the pattern of 5' ends observed. Deletion of the sequence between -242 and -99 does not further change the overall transcript level but does affect the specificity of CYC1 mRNA starting. A deletion that extends from -242 to -75 causes both an additional shift in the pattern of 5' ends observed and a further large decrease (factor of 10--20) in CYC1 RNA level. Deletions that extend from -242 to -43, and particularly two deletions that extend still closer to the initiating ATG, cause the appearance of an abundant transcript which starts upstream of position -1078 and of minor transcripts starting in the region -325 to -245.

Regulation of transcription in expressed and unexpressed mating type cassettes of yeast

The genes that control the a, alpha and a/alpha cell types in Saccharomyces are carried on transposable elements known as a and alpha cassettes which reside at three different chromosomal loci. Examination of the transcripts by R-looping and filter hybridization indicates that each cassette is capable of producing two divergent transcripts. Cassettes at the MAT locus are transcribed constitutively. Transcription of cassettes at HML and HMR is prevented by trans-acting negative regulators.

Homology and non-homology at the yeast mating type locus

Four mutations of the alpha mating type locus of Saccharomyces cerevisiae have been analysed to determine their relationship to the a mating type locus. Mat alpha+ recombinations are produced by mat alpha 2-/MAT but not by mat alpha 1-/MATa diploids. MAT alpha and MATa thus contain regions of homology (coding for at least part of MAT alpha 2) and regions of non-homology (coding for at least part of MAT alpha 1)-the genetic determinant for cell type is larger than the non-homologous sequence seen by DNA-DNA heteroduplexes and genetic analysis. The segment transposed in mating type interconversion includes both types of sequence.