Histones as regulators of genes.
Histones were once dismissed as little more than packing material for nuclear DNA. In fact, these proteins can both repress and facilitate activation of many genes.
Biomedical subjects
Publications and source records attributed to M Grunstein.
Histones were once dismissed as little more than packing material for nuclear DNA. In fact, these proteins can both repress and facilitate activation of many genes.
Nucleosomes are positioned in the presence of the yeast repressor alpha 2 in minichromosomes containing the alpha 2 operator and on the promoters of a-cell-specific genes regulated by alpha 2. To investigate the possibility that alpha 2 directs nucleosome position through an interaction with a component of the core particle, we analyzed chromatin structures adjacent to the operator in alpha cells containing mutations in the amino-terminal region of histone H4. Deletion or point mutation of specific amino acids in histone H4 altered the location and/or stability of nucleosomes adjacent to the alpha 2 operator. These changes in chromatin structure were accompanied by partial derepression of a beta-galactosidase reporter construct under alpha 2 control, even though alpha 2 remained bound to its operator sequence. Our data suggest that complete repression by alpha 2 requires stable positioning of nucleosomes in promoter regions and this positioning involves the conserved amino-terminal region of histone H4.
Histone mRNA synthesis is tightly regulated to S phase of the yeast Saccharomyces cerevisiae cell cycle as a result of transcriptional and posttranscriptional controls. Moreover, histone gene transcription decreases rapidly if DNA replication is inhibited by hydroxyurea or if cells are arrested in G1 by the mating pheromone alpha-factor. To identify the transcriptional controls responsible for cycle-specific histone mRNA synthesis, we have developed a selection for mutations which disrupt this process. Using this approach, we have isolated five mutants (hpc1, hpc2, hpc3, hpc4, and hpc5) in which cell cycle regulation of histone gene transcription is altered. All of these mutations are recessive and belong to separate complementation groups. Of these, only one (hpc1) falls in one of the three complementation groups identified previously by other means (M. A. Osley and D. Lycan, Mol. Cell. Biol. 7:4204-4210, 1987), indicating that at least seven different genes are involved in the cell cycle-specific regulation of histone gene transcription. hpc4 is unique in that derepression occurs only in the presence of hydroxyurea but not alpha-factor, suggesting that at least one of the regulatory factors is specific to histone gene transcription after DNA replication is blocked. One of the hpc mutations (hpc2) suppresses delta insertion mutations in the HIS4 and LYS2 loci. This effect allowed the cloning and sequence analysis of HPC2, which encodes a 67.5-kDa, highly charged basic protein.
We have previously shown that nucleosome loss, obtained by repressing histone H4 mRNA synthesis, activates otherwise inactive PHO5, GAL1, and CYC1 gene promoters (fused to the bacterial beta-galactosidase [lacZ] reporter gene) to moderate levels of activity (approximately 2 to 15% of fully induced levels). We now report that nucleosome loss activates the expression of two additional promoters that are normally induced by independent mechanisms: CUP1 (induced by heavy-metal toxicity) and HIS3 (induced by amino acid starvation). Surprisingly, the level of CUP1-lacZ and HIS3-lacZ activation by nucleosome loss approximates fully induced levels of transcription. These CUP1 and HIS3 promoter activities are increased similarly from either episomal or genomic constructs. Our results emphasize the universality of the mechanism by which nucleosome loss activates yeast promoters. Moreover, a comparison of absolute levels of activation for different promoters suggests that activation by nucleosome loss results in a relatively constant level of activation, while levels obtained by normal induction vary considerably. These data argue that nucleosome loss may play a uniquely dominant role in the regulation of certain promoters.
Explore the source record for details and available documents.
To search for histone domains that may regulate transcription in vivo, we made deletions and amino acid substitutions in the histone N-termini of S. cerevisiae. Histone H4 N-terminal residues 4-23, which include the extremely conserved, reversibly acetylated lysines (at positions 5, 8, 12, and 16), were found to encompass a region required for the activation of the GAL1 promoter. Deletions in the H4 N-terminus reduce GAL1 activation 20-fold. This effect is specific to histone H4 in that large deletions in the N-termini of H2A, H2B, and H3 do not similarly decrease induction. Activation of the PHO5 promoter is reduced approximately 4- to 5-fold by these H4 deletions. Mutations in histone H4 acetylation sites and surrounding residues can cause comparable and, in some cases, even greater effects on induction of these two promoters. We postulate that the H4 N-terminus may interact with a component of the transcription initiation complex, allowing nucleosome unfolding and subsequent initiation.
Histones and nucleosomes are involved in the folding of DNA in the eukaryotic cell. Recent evidence suggests that they are also involved in a multistep process of DNA unfolding and gene regulation.
Repression of transcription from the silent mating loci (HML alpha and HMRa) is essential for mating ability in Saccharomyces cerevisiae. This silencing is known to require at least five proteins (SIR1, SIR2, SIR3, SIR4, and histone H4) and is accompanied by a change in chromatin structure. We show here that four positions of histone H4 (N-terminal residues 16, 17, 18, and 19) are crucial to silencing. HML alpha and HMRa are efficiently repressed when these positions are occupied by basic amino acids but are derepressed when substituted with glycine. These results suggest that acetylation of Lys-16 would lead to derepression of the silent mating loci. Three strong extragenic suppressors of the latter H4 mutations were isolated and determined to be located in SIR3. These suppressors allow high mating efficiencies in cells expressing either wild-type H4 or H4 containing single amino acid substitutions. They did not allow efficient mating in a strain that contained an H4 N-terminal deletion. These results indicate that the SIR3 mutations do not bypass the requirement for the H4 N terminus but, rather, allow repression in the presence of a less than optimal H4 N terminus. This provides a link between one of the SIR proteins and a component of chromatin.
Saccharomyces cerevisiae centromeric DNA is packaged into a highly nuclease-resistant chromatin core of approximately 200 base pairs of DNA. The structure of the centromere in chromosome III is somewhat larger than a 160-base-pair nucleosomal core and encompasses the conserved centromere DNA elements (CDE I, II, and III). Extensive mutational analysis has revealed the sequence requirements for centromere function. Mutations affecting the segregation properties of centromeres also exhibit altered chromatin structures in vivo. Thus the structure, as delineated by nuclease digestion, correlated with functional centromeres. We have determined the contribution of histone proteins to this unique structural organization. Nucleosome depletion by repression of either histone H2B or H4 rendered the cell incapable of chromosome segregation. Histone repression resulted in increased nuclease sensitivity of centromere DNA, with up to 40% of CEN3 DNA molecules becoming accessible to nucleolytic attack. Nucleosome depletion also resulted in an alteration in the distribution of nuclease cutting sites in the DNA surrounding CEN3. These data provide the first indication that authentic nucleosomal subunits flank the centromere and suggest that nucleosomes may be the central core of the centromere itself.
Yeast (Saccharomyces cerevisiae) histone mRNA synthesis is tightly regulated to the S phase of the cell division cycle as a result of both transcriptional and posttranscriptional regulation. We focused on the role of posttranscriptional control in histone H2B1 gene (HTB1) regulation and studied a portion of the HTB1 message required for cell-cycle-specific accumulation. The 3' end of the HTB1 gene containing a 17-amino-acid coding sequence and entire noncoding sequence was fused to the bacterial neomycin phosphotransferase II gene (neo) under control of the GAL1 promoter. The expression of the endogenous and chimeric HTB1 genes was analyzed during the yeast cell cycle. As yeast cells entered a synchronous cell cycle following release from alpha-factor arrest, the level of GAL1-promoter-controlled neo-HTB1 message increased approximately 12-fold during S phase and dropped to basal level when the cells left S phase. This indicates that the 3' end of the HTB1 mRNA is capable of conferring cycle-specific regulation on a heterologous message. Deletion analysis of the 3' end showed that the signal for cell cycle control of HTB1 mRNA includes contiguous coding and noncoding sequences surrounding the stop codon. This differs from the situation in mammalian cells, whose posttranscriptional regulation of histone genes is mediated through a short sequence containing a stem-loop structure near the very terminus of the untranslated 3' end.
Explore the source record for details and available documents.
A gene from Saccharomyces cerevisiae has been mapped, cloned, sequenced and shown to encode a catalytic subunit of an N-terminal acetyltransferase. Regions of this gene, NAT1, and the chloramphenicol acetyltransferase genes of bacteria have limited but significant homology. A nat1 null mutant is viable but exhibits a variety of phenotypes, including reduced acetyltransferase activity, derepression of a silent mating type locus (HML) and failure to enter G0. All these phenotypes are identical to those of a previously characterized mutant, ard1. NAT1 and ARD1 are distinct genes that encode proteins with no obvious similarity. Concomitant overexpression of both NAT1 and ARD1 in yeast causes a 20-fold increase in acetyltransferase activity in vitro, whereas overexpression of either NAT1 or ARD1 alone does not raise activity over basal levels. A functional iso-1-cytochrome c protein, which is N-terminally acetylated in a NAT1 strain, is not acetylated in an isogenic nat1 mutant. At least 20 other yeast proteins, including histone H2B, are not N-terminally acetylated in either nat1 or ard1 mutants. These results suggest that NAT1 and ARD1 proteins function together to catalyze the N-terminal acetylation of a subset of yeast proteins.
A 13-year-old girl with cystic fibrosis (CF), severe lung disease, purpura on the lower extremities and with elevated serum mixed cryoglobulins is described. The possible relationship of the purpura in CF with cryoglobulinaemia is discussed.
Nucleosome depletion can be made to occur in yeast by addition of glucose to strains containing the histone H4 gene under GAL promoter control. This leads to the activation of downstream promoter elements (TATA box and initiation, I, region) of three different regulated yeast promoters fused to the E. coli lacZ gene. Nucleosome loss activates the PHO5 downstream element in the presence or absence of the upstream activator sequences (UAS) through which PHO5 induction is normally mediated. The cytochrome C (CYC1) and galactokinase (GAL1) promoters are normally repressed by glucose through their UAS elements. However, when these UAS are deleted, the remaining downstream promoters are also activated by glucose-mediated nucleosome loss. These data suggest that nucleosome loss increases transcription initiation and subsequent elongation in vivo. They also indicate that the proteins which recognize the downstream promoter are activated and functional, at least in part, even in the absence of the UAS complex.
Yeast histone H4 function was probed in vivo by deleting segments of this extremely conserved 102 amino acid protein. Deletions in the hydrophobic core of H4 are lethal and block chromosomal segregation. In contrast, deletions at the hydrophilic N terminus (residues 4-28) and C terminus (residues 100-102) are viable. However, N-terminal deletion alters normal chromatin structure and lengthens the cell cycle, especially G2. Surprisingly, removal of the H4 N terminus also derepresses the silent mating type loci, HML alpha and HMRa, disrupting mating. This activation is specific since other regulated genes (GAL10, PHO5, CUP1) are repressed and induced normally in these cells. Deletions of the hydrophilic N termini of H2A or H2B do not show this effect on mating. These experiments allow us to define a unique H4 function that is not shared by other histones (H2A and H2B).
We have constructed a yeast strain (UKY403) in which the sole histone H4 gene is under control of the GAL1 promoter. This allows the activation of H4 mRNA synthesis on galactose and its repression on glucose. UKY403 cells, pre-synchronized in G1 with alpha-mating factor, have been used to show that glucose treatment results in the loss of approximately half the chromosomal nucleosomes. This depletion is only partially reversible when the H4 gene is reactivated on galactose. It was found that the resultant lethality manifests itself first in S phase, the period of nucleosome assembly, but leads to highly synchronous arrest in G2 and a virtually complete block in chromosomal segregation. Histone H4-depleted chromatin was analyzed for its efficiency as a template for all three RNA polymerases. Using pulse-labeling, we find no evidence for altered transcription by RNA polymerase I (25S, 18S and 5.8S rRNAs) or RNA polymerase III (5S rRNA, tRNAs). Northern blot analysis was used to measure levels of RNA polymerase II transcripts. There was little effect on the activation or repression of the CUP1 chelatin gene. While there may be some decrease in the level of certain mRNAs (e.g. HIS4, ARG4) other message levels (HIS3, TRP1) show little change upon glucose repression. Therefore, nucleosome loss certainly does not have a general effect on transcription.
We have previously constructed a yeast strain (UKY403) whose sole histone H4 gene is under control of the GAL1 promoter. This yeast arrests in G2 upon glucose treatment as a result of histone H4 depletion. The yeast PHO5 gene contains phase nucleosomes covering promoter (UAS) sequences in the PHO5 repressed state and it has been suggested that nucleosomes prevent the binding of positively acting factors to these UAS sequences. Using UKY403 we examined the length of polynucleosomes and nucleosome phasing in the PHO5 upstream region by the use of micrococcal nuclease and indirect end-labeling. It was found that glucose arrest led to a severe disruption in PHO5 chromatin structure and that most nucleosomes had their position altered or were lost from the PHO5 promoter region. Cell undergoing nucleosome depletion synthesized large quantities of accurate PHO5 transcripts even under repressive, high inorganic phosphate conditions. Histone H4 depletion did not appear to affect the repression or activation of another inducible yeast gene, CUP1. Arrest with landmarks in early G1 (in the cell division cycle mutant cdc28) or in various stages of G2 (in cdc15, cdc17 and cdc20) does not activate PHO5; nor does arrest due to chromosome topology changes (in top2 or the top1top2 topoisomerase mutants). cdc14, which has its arrest landmark at a similar point in the cell cycle as cdc15, does derepress PHO5. However, since it also leads to derepression of CUP1 it is probably functioning through an independent mechanism. Therefore, our data suggest that nucleosomes regulate PHO5 transcription.
To determine which cellular processes are dependent on histones, we blocked histone H2B mRNA synthesis in asynchronously growing yeast after fusing the H2B gene to a repressible GAL10 promoter. Chromosomal segregation, replication, and transcription were then examined. We found that the cells arrested in mitosis, with a cell division cycle (cdc) phenotype. Chromatin structure and nuclear segregation were disrupted. A full round of DNA replication took place after the repression of histone H2B mRNA synthesis. Active transcription and the induction of new transcripts also continued in the arrested cells.