Eukaryotic genomes contain euchromatic regions, which are transcriptionally active, and heterochromatic regions, which are repressed. These domains are separated by "barrier elements": DNA sequences that protect euchromatic regions from encroachment by neighboring heterochromatin. To identify proteins that play a role in the function of barrier elements we have carried out a screen in S. cerevisiae. We recovered the gene HHO1, which encodes the yeast ortholog of histone H1, as a high-copy modifier of barrier activity. Histone H1 is a linker histone that binds the outside of nucleosomes and modifies chromatin dynamics. Here we show that Hho1p reinforces the action of several types of barrier elements, and also inhibits silencing on its own.

---

THE genome of eukaryotes is organized in euchromatic and heterochromatic regions. This spatial pattern is stable enough to ensure the permanence of proper gene expression through cell divisions, yet it is also flexible and can be remodeled as a cell differentiates. One important question is to determine what sets the limits between active and inactive domains of the genome.

The yeast Saccharomyces cerevisiae is a useful model to study gene repression by heterochromatin (RUSCHE et al. 2003). In this organism, only three classes of loci can be heterochromatinized or silenced: the silent mating type loci HMR and HML, some of the ribosomal DNA (rDNA) repeats, and the subtelomeric regions. The actors in this process are well characterized. Central to the silencing process is the protein Sir2p, which can deacetylate histones and promote the formation of repressive chromatin. Sir2p binds Sir3p and Sir4p to form the SIR complex, which acts at HML, HMR, and the subtelomeric regions. In contrast, Sir3p and Sir4p are not required for rDNA silencing.

In yeast, heterochromatin spreads linearly from nucleation points called silencers. The SIR proteins are recruited to the silencers, deacetylate nearby histones, then bind the newly deacetylated histones and start a new cycle of deacetylation. The spreading of heterochromatin thus corresponds to a polymerization of the SIR complex over DNA. It is stopped by barrier elements (DHILLON and KAMAKAKA 2002).

Several approaches have been taken to understand the function of barrier elements in yeast. Some investigators have studied the endogenous barrier elements that occur around HMR (DONZE et al. 1999; DONZE and KAMAKAKA 2001) or in subtelomeric positions (FOUREL et al. 1999, 2001; KIMURA et al. 2002; SUKA et al. 2002). In other experiments, genetic screens were conducted to identify proteins that can form a barrier element (ISHII et al. 2002; OKI et al. 2004). In light of these results, barrier elements seem to fall into two classes (DONZE and KAMAKAKA 2002; ISHII and LAEMMLI 2003; KIMURA and HORIKOSHI 2004). The first type of barrier is of a structural nature: nucleosome gaps or interaction with the nuclear pores are able to block the spread of silencing. The second type of barrier is more dynamic. The local recruitment of transcription factors, nucleosome remodeling enzymes, or histone modifying enzymes can compete with the repressive SIR activity and prevent further spreading.

Histone H1 is a linker histone that binds the outside of nucleosomes and modifies chromatin dynamics (BUSTIN et al. 2005). S. cerevisiae has one histone H1 ortholog, Hho1p, whose role is unknown (PATTERTON et al. 1998). We have carried out a genetic screen to identify yeast proteins that can modulate the activity of a barrier element and have recovered Hho1p. We show that yeast histone H1 increases the activity of different barrier elements and that it can inhibit silencing on its own.

Plasmids:

The high-copy library we used contains yeast genomic DNA inserted into the LEU2-marked vector Yep13. It was built in the Nasmyth lab and kindly made available to us by Etienne Schwob. Library plasmid p133 was digested with BamHI, filled in, and religated to generate plasmid p134, in which there is a frameshift in HHO1. HHO1 with its endogenous promoter and terminator was amplified by PCR and cloned into Yep13 to form plasmid p172. The expression vectors for Gal4_1–147 and Gal4-ID are multicopy HIS3 plasmids and have been published (DEFOSSEZ and GILSON 2002). The expression vector for LexA (multicopy, HIS3) has also been described (BI et al. 2004). Plasmids used in this study are listed in Table 1.

View this table: In this window In a new window   TABLE 1 Plasmids used in this study

 

Yeast strains:

GF97 is a derivative of W303-1a that has already been described (FOUREL et al. 1999). This strain contains a silencing reporter construct integrated at telomere VIIL by recombination with the ADH4 gene. The reporter construct contains, closest to ADH4, an 850-bp fragment containing the TRP1 gene, linked to a 150-bp fragment containing four Gal4p binding sequences (UAS_Gal4), and finally a 1.3-kb fragment containing the URA3 gene. Both reporter genes are transcribed toward the telomere (see Figure 1A). The transcription start site of URA3 is 1.1 kb away from the telomeric repeats, while that of TRP1 is 2.1 kb away.

View larger version (70K): In this window In a new window Download PPT slide   FIGURE 1.— Yeast histone H1 increases the activity of an artificial subtelomeric barrier element. (A) The test strain GF97 contains two reporter genes inserted next to ADH4 at telomere VIIL. The activity of the silencing barrier present between the two genes is measured by the proportion of cells that grow on medium lacking tryptophan and containing 5-fluoro-orotic acid (–W+FOA). (B) Overexpression of HHO1 increases barrier activity in GF97. A chimeric protein with weak insulating activity, Gal4-ID, was transformed into strain GF97. Under these conditions, 1% of cells are TRP+ ura– (10-fold dilutions spotted on control or –HLW+FOA medium, top line). After transformation with a multicopy library of yeast genomic DNA, individual clones were tested for growth on –HLW+FOA. One of the plasmids recovered in the screen includes a genomic fragment that contains a fraction of NAN1, HHO1, and a fraction of TBF1. It increases barrier activity 10-fold relative to the control (second line). Third line: inactivation of HHO1 by a frameshift mutation cancels that effect. Fourth line: overexpression of HHO1 fully recapitulates the stimulation of barrier activity. The control plates select only for the presence of the plasmids, and lack histidine and leucine (–HL). (C) Deletion of HHO1 in GF97 decreases barrier activity. Gal4-ID was introduced in GF97 (top), or its derivative lacking HHO1 (bottom). Serial 10-fold dilutions of the cultures were spotted on the indicated media.

 
Strains H0, H2, and H10 have been described (BI et al. 2004). The strain KIY54 was a gift from Ulrich Laemmli (ISHII et al. 2002) and JS124 was from Jeff Smith (SMITH et al. 1998). HHO1 was deleted in various strains using an hho1::KanMX disruption cassette, and disruption was verified by PCR. For chromatin immunoprecipitation, Sir4p was tagged with nine copies of the Myc epitope by PCR and homologous recombination. Yeast strains used in this study are listed in Table 2.

View this table: In this window In a new window   TABLE 2 Yeast strains used in this study

 

Screening for multicopy modifiers of barrier activity:

We first transformed GF97 with an HIS3 plasmid expressing a domain of chicken CTC-binding factor (CTCF) (amino acids 641–728) fused to the DNA binding of Gal4p (amino acids 1–147). This strain was then transformed with the overexpression library of yeast genomic DNA. Cells containing both plasmids were selected on plates lacking leucine and histidine. Individual transformants were picked and grown to saturation in 250 µl of selective medium in 96-well culture plates. Ten microliters of a 20-fold dilution of each culture was spotted in duplicate on SC-HL as well as on SC-HLW+FOA. Candidates with an increased or decreased proportion of TRP+/FOAr cells were picked on the SC-HL plate and retested. The plasmids were rescued from the candidate strains by smash and grab (HOFFMAN and WINSTON 1987). The LEU2 plasmids were selectively recovered by complementation of the leuB6 mutation in bacterial strain HB101 (a gift of Agnès Bègue).

Generation of anti-Hho1p serum:

We synthesized two 15-amino acid peptides matching the sequence of Hho1p: APKKSTTKTTSKGKK and CVENGELVQPKGPSG. They were coupled to keyhole limpet hemocyanin (KLH) and used to immunize two rabbits (Eurogentec, Seraing, Belgium). Antibodies contained in the final bleed were affinity purified and used for Western blotting (dilution 1:5000) and chromatin immunoprecipitation.

Chromatin immunoprecipitation:

We followed a protocol given to us by Laurent Kuras (KURAS and STRUHL 1999). The cells were crosslinked for 15 min at room temperature in 1% formaldehyde, lysed with glass beads, and the DNA sonicated to an average size of 500 bp. The antibodies used to immunoprecipitate chromatin in the different figures were: anti-Gal4p (rabbit polyclonal SC-577, Santa Cruz Biotechnology), immunopurified rabbit serum directed against Hho1p, and anti-Myc tag (mouse monoclonal clone 9E10, Roche Diagnostics). The sequence of the primers used is available on request. For Figure 3, C and E, the DNA recovered after immunoprecipitation was quantified by real-time PCR on a LightCycler (Roche). Amplification of the recovered DNA was done in parallel with amplification of serial dilutions of input DNA, and we verified that the efficacy of the PCR reaction was high enough to permit quantitation. The amount of immunoprecipitated DNA is expressed as a percentage fraction of the input.

View larger version (47K): In this window In a new window Download PPT slide   FIGURE 3.— Histone H1 increases the barrier activity of Gal41–147. (A) In the absence of Gal41–147, the activity of the barrier is undetectable (top: no growth on –LW+FOA). Overexpression of HHO1 does not increase barrier activity in this context. Expression of Gal41–147 is not sufficient to generate detectable barrier activity (bottom). However, overexpression of HHO1 in the presence of Gal41–147 generates an active barrier, as evidenced by growth on –HLW+FOA medium. (B) Histone H1 overexpression does not affect the expression of Gal41–147. Protein extracts were prepared from strains containing the indicated plasmid combinations, and used for Western blotting with an antibody directed against Gal4p. (C) Histone H1 does not recruit Gal41–147 to the barrier. Gal41–147 abundance at the indicated loci was measured by ChIP followed by quantitative PCR. Strains harbored an HHO1 overexpression plasmid (2µ HHO1) or a control plasmid (WT). (D) Characterization of an anti-Hho1p antibody. Total protein extracts from the indicated yeast cells were used for Western blotting. (E) Gal41–147 does not recruit Hho1p to the barrier. Strain GF97 with endogenous or elevated levels of HHO1 was transformed with a control vector or an expression vector for Gal41–147 (+Gal4p). Abundance of Hho1p at the indicated loci was measured by ChIP followed by quantitative PCR.

 

Protein extraction and Western blotting:

Protein extracts were made according to a published protocol (KUSHNIROV 2000). Western blotting followed standard procedures. For Western blotting on Gal4p, we used antibody SC-577 (Santa Cruz Biotechnology) at 1:300 dilution.

A yeast screen for high-copy modifiers of barrier activity identifies histone H1:

To identify yeast proteins involved in transcriptional insulation we designed a genetic screen. It involved a previously described yeast strain in which barrier activity can be monitored (FOUREL et al. 2001). This strain, GF97, has two reporter genes, TRP1 and URA3, integrated close to the left telomere on chromosome VII and separated by four upstream activation sequences (UAS) that can recruit Gal4p (Figure 1A). The two reporter genes are subject to telomeric position effect, and their expression is repressed in most cells in a population. However, repression is not absolute and the genes can be expressed in some cells. Cells in which TRP1 is expressed while URA3 is repressed can be detected by their ability to form colonies on medium lacking tryptophan and containing 5-fluoro-orotic acid (FOA), a drug toxic to cells expressing URA3. These tryptophan–prototroph and FOA-resistant (TRP+/FOAr) cells are rare in a population, occurring at a frequency of 10^–6. This reflects the fact that the expression of TRP1 and URA3 is normally coupled: either both genes are silenced or both are expressed.

Chromatin barriers can block the spreading of repressive chromatin (DHILLON and KAMAKAKA 2002). If a peptide with barrier activity is fused to the DNA-binding domain of Gal4p and expressed in strain GF97, TRP1 can be shielded from the telomeric repression that acts on URA3. This results in an increase in the number of TRP+/FOAr cells, which can then represent as much as 10–20% of the population (FOUREL et al. 2001). Using this assay we have found that the vertebrate insulating protein CTCF can also act as an insulator in S. cerevisiae (DEFOSSEZ and GILSON 2002). To identify yeast proteins involved in this phenomenon we undertook a screen for high-copy modifiers of insulation by CTCF. A C-terminal domain of chicken CTCF (amino acids 641–728) was fused to Gal4_1–147 to form the chimera Gal4-ID (insulation domain), which was expressed in strain GF97. Gal4-ID has relatively weak barrier activity and made only 1% of the cells TRP+/FOAr (DEFOSSEZ and GILSON 2002). We then transformed this strain with a library of yeast genomic DNA fragments cloned in a multicopy vector. Individual transformants were picked, grown in 96-well plates, and spotted on selective plates to determine total cell number as well as the number of TRP+/FOAr cells. In the 2500 clones examined we found 11 plasmids that increased the fraction of TRP+/FOAr cells and 7 that decreased it. Plasmid p133 gave a strong increase in barrier activity and was selected for further work, which is reported here.

Plasmid p133 contains a 3.2-kb fragment of yeast chromosome XVI (coordinates 307660–310860). It encompasses HHO1, the gene encoding histone H1, as well as portions of the genes NAN1 and TBF1 (Figure 1B). We first tested whether overexpression of HHO1 was necessary to increase barrier activity. We modified p133 to create a frameshift after amino acid 62 in Hho1p, yielding a truncation that removes three-quarters of the protein. The resulting construct had no effect on barrier activity (Figure 1B), establishing that HHO1 was indeed necessary for the effect of plasmid p133. We then examined whether overexpression of HHO1 was in itself sufficient to increase insulation by the chimera. We cloned HHO1 with its promoter and terminator into the library plasmid. This construct fully recapitulated the effect of p133. From this we conclude that overexpression of yeast histone H1 is sufficient to enhance insulation by Gal4-ID. Next, we deleted the endogenous HHO1 gene in GF97. The barrier activity of Gal4-ID was weaker in hho1 than in a wild-type strain (Figure 1C). This shows that endogenous histone H1 contributes to insulation by Gal4-ID. Therefore, yeast histone H1 is a dose-dependent enhancer of insulation by Gal4-ID.

An alternative explanation for these results would be that HHO1 acts directly on the reporter genes to increase TRP1 expression and/or decrease URA3 expression. We undertook four control experiments to examine this possibility. First, we determined the effect of HHO1 overexpression in the absence of the Gal4-ID protein. No TRP+/FOA-resistant cells appeared under these conditions (see Figure 3A, left). Second, we used a strain in which the positions of TRP1 and URA3 were reversed (Figure 2A). TRP1 was partially silenced in this strain, and overexpression of HHO1 did not alleviate silencing. The proportion of FOA-resistant cells was also unaffected. As expected, no cells were simultaneously TRP+ and FOA-resistant in this strain, and no such cells appeared after HHO1 overexpression. Third, we eliminated telomeric silencing in strain GF97 by deleting SIR4. In this context, overexpression of HHO1 had no effect on the expression of TRP1 or URA3 (Figure 2B). Fourth, we asked whether the effect of HHO1 deletion observed in Figure 1C might be due to a general change in telomere position effect. To address this question, we deleted HHO1 in GF97 and monitored reporter gene expression (Figure 2C). The deletion did not affect TRP1 or URA3 expression. This result is consistent with earlier findings that failed to detect an effect of HHO1 deletion on the silencing of subtelomeric reporter genes (ESCHER and SCHAFFNER 1997; PATTERTON et al. 1998). From this set of experiments we conclude that HHO1 does not act directly on the reporter genes, nor does it affect telomeric silencing. Rather, HHO1 stimulates the activity of the artificial barrier.

View larger version (62K): In this window In a new window Download PPT slide   FIGURE 2.— Histone H1 does not act directly on the reporter genes or affect telomeric silencing. (A) HHO1 overexpression has no effect on a reporter strain in which TRP1 and URA3 are swapped. Serial 10-fold dilutions of yeast cultures transformed with the indicated plasmid were spotted on selective plates as in Figure 1. (B) HHO1 overexpression has no effect in a sir4 strain that lacks telomeric silencing. (C) Deletion of HHO1 does not cause loss of telomeric silencing in strain GF97.

 

Overexpression of histone H1 increases insulation by Gal4_1–147:

After finding that histone H1 increased the barrier activity of Gal4-ID, we carried out additional control experiments. First, we tested whether the effect of the histone required the presence of the Gal4-ID protein. There were no TRP+/FOAr cells in GF97 in the absence of Gal4-ID, and overexpression of histone H1 in that context did not increase barrier activity (Figure 3A, top). Next, we tested whether Hho1p would have an effect on Gal4_1–147 in the absence of the fused insulation domain. We expressed Gal4_1–147 in GF97. This did not generate any detectable insulation (Figure 3A, bottom left). However, overexpression of histone H1 in the presence of Gal4_1–147 gave rise to a large number of TRP+/FOAr cells. This suggests that Hho1p stimulates the barrier activity of Gal4-ID simply by acting on the Gal4_1–147 moiety. The DNA-binding domain of LexA has been shown to have barrier activity (BI et al. 2004), but to our knowledge this is the first report that the DNA-binding domain of Gal4p also has barrier-forming potential.

We then tested three possible mechanisms by which histone H1 could increase the barrier activity of Gal4_1–147. First, we tested the possibility that HHO1 overexpression might increase the expression level of Gal4_1–147. This was ruled out; there was no difference in the level of Gal4_1–147 in cells overexpressing HHO1 relative to control cells, as judged by Western blotting (Figure 3B). A second possibility was that HHO1 might facilitate the binding of Gal4_1–147 to its target sites. We addressed this by chromatin immunoprecipitation (ChIP). Gal4_1–147 was crosslinked to DNA and immunoprecipitated. We then carried out quantitative PCR on the recovered DNA to examine the abundance of two sequences. The first one, located in the GAL1 promoter, contains a natural cluster of high affinity Gal4p binding sites (REN et al. 2000). The second sequence contains the Gal4p binding sites located between the two reporter genes TRP1 and URA3 in GF97. This strain contains the endogenous Gal4p protein but it is present in very low amounts relative to Gal4_1–147, which is expressed from the ADH1 promoter on a high-copy vector. Overexpression of HHO1 did not change the amount of Gal4_1–147 bound to either site (Figure 3C). Therefore we conclude that HHO1 does not increase barrier activity by enhancing recruitment of Gal4_1–147. A third possible mechanism would be that histone H1 is a limiting factor for barrier activity, and that it is recruited by Gal4_1–147. In that case, more histone H1 should be present at the subtelomeric region in the presence of Gal4_1–147 than in its absence. Again, we used ChIP to address this possibility. Our first approach was to use tagged versions of Hho1p. However, versions of Hho1p tagged at the C-terminus with 3xHA or 3xMyc tags failed to reproduce the effect of the wild-type protein. We therefore raised Hho1p antisera in rabbits. The antibodies recognize only one band by Western blotting on a total yeast extract. The band matches the predicted mass of histone H1 and disappears in a hho1 strain (Figure 3D), validating the use of the serum for ChIP. Two loci were examined: the telomeric sequence containing Gal4p binding sites and the ORF of ACT1 to serve as a control (Figure 3E). First we examined the endogenous histone H1. The protein was more abundant at the subtelomeric site than at ACT1. The abundance of each site was not modified by the overexpression of Gal4_1–147. Then we overexpressed histone H1 in the same two strains. Binding to ACT1 and the subtelomeric locus was increased, in line with the idea that histone H1 is present in limiting amounts in the cell (FREIDKIN and KATCOFF 2001). Overexpression of Gal4_1–147 did not enhance binding of histone H1 to either site.

This set of experiments shows that histone H1 does not recruit Gal4_1–147, and that Gal4_1–147 does not recruit histone H1. Therefore, the two proteins seem to act independently to form barriers, and we believe their synergistic action is due to a threshold effect.

Histone H1 synergizes with different types of barrier elements and can also act alone:

Because HHO1 seemed to act independently of Gal4_1–147 we asked whether it would also synergize with barrier elements that did not involve Gal4_1–147. First we tested whether it might also potentiate the effect of LexA, a DNA-binding protein that has barrier activity (BI et al. 2004) and is unrelated to Gal4p. In the strain we used, URA3 is separated from an inverted HML-I by two target sites for LexA. Binding of LexA to these sites has been shown to generate a barrier (BI et al. 2004). We observed that overexpression of HHO1 increased the activity of the barrier in this context as well (Figure 4A). We then examined a strain in which URA3 is separated from the inverted HML-I by four CGGNN repeats (Figure 4B). These sequences have a low propensity to form nucleosomes and interfere with the spread of silencing (BI et al. 2004). Overexpression of HHO1 increased barrier activity in that strain (Figure 4B), while deletion of HHO1 decreased barrier activity (Figure 4C). The observations made on these two strains reinforce the fact that the effect we first noted on Gal4_1–147 was not specific, and that HHO1 might more generally contribute to barrier activity.

View larger version (67K): In this window In a new window Download PPT slide   FIGURE 4.— Histone H1 stimulates the activity of different transcriptional barriers, and also inhibits silencing on its own. (A) Overexpression of HHO1 increases the activity of a LexA barrier next to HML-I but also acts in the absence of LexA. (B) Overexpression of HHO1 increases the barrier activity of CCGNN repeats but also inhibits silencing in the absence of the barrier. (C) Deletion of HHO1 increases URA3 silencing in strains H10 and H0.

 
These experiments yielded another important result. Indeed, we observed in all three strains (H0, H2, and H10) that histone H1 overexpression decreased the silencing of URA3 even in the absence of the barrier element (Figure 4, A and B). Conversely, deletion of HHO1 increased the silencing of URA3 even in the absence of a barrier (Figure 4C, bottom).

Overexpression of Histone H1 decreases the abundance of SIR proteins at a silenced locus:

Next we asked whether the effect of HHO1 was indeed a direct effect on silencing. For this we measured the abundance of Sir4p at different positions in a silenced locus, in the presence or absence of overexpressed histone H1 (Figure 5). The experiment was performed in a derivative of strain H10, in which a 9xMyc epitope tag was added to Sir4p. We started by verifying that URA3 silencing was not affected by the modification of Sir4p and also that HHO1 overexpression decreased URA3 silencing as it did in the parent strain (Figure 5A). The amount of Sir4p was measured by ChIP at three loci: HML-I, and sites a and b, which are increasingly distant from HML-I. In addition we used a nonsilenced locus, ACT1, as a negative control for Sir4p binding. In the absence of overexpressed HHO1 we detected binding of Sir4p to HML-I and to sites a and b (Figure 5B). This is consistent with the silencing of URA3 that is observed under these conditions. When HHO1 was overexpressed, the binding of Sir4p to all three test sites was decreased. The decrease followed a gradient and was least pronounced at HML-I and most pronounced at the greatest distance from HML (Figure 5B). This supports the notion that histone H1 overexpression inhibits silencing directly, by impeding the propagation of the SIR complex away from silencers.

View larger version (49K): In this window In a new window Download PPT slide   FIGURE 5.— Decreased binding of the SIR complex to a silenced locus upon HHO1 overexpression. (A) Characterization of a reporter strain containing a tagged version of Sir4p. URA3 silencing is maintained in the strain, and is decreased by HHO1 overexpression (decreased growth on –L+FOA medium). (B) Measurement of SIR complex binding to the silenced locus. Strain H10 Sir4-Myc was transformed with an HHO1 overexpression vector, or with a control vector. The SIR complex was immunoprecipitated with an anti-Myc antibody, and the abundance of different sequences in the recovered DNA was tested by PCR. ACT1 is a negative control that is not bound by the SIR complex. Ab, antibody.

 

Overexpression of histone H1 decreases silencing at some loci but not all:

After observing an effect for histone H1 overexpression outside of HML-I, we tested whether this situation also occurred at other silenced loci. First we examined a strain containing two reporter genes embedded within HML (ISHII et al. 2002). Histone H1 overexpression did not protect URA3 or ADE2 against silencing (Figure 6A), suggesting that it can only act outside, and not within, the silenced domain. Because silencing at HMR often behaves differently from silencing at HML, we also tested GCY317, a reporter strain in which ADE2 is inserted within HMR. In this strain, increasing degrees of HMR silencing translate into increasing frequency and color intensity of pink sectors in the yeast colonies (CUPERUS and SHORE 2002). We could not detect any difference between colonies of GCY317 containing an empty vector and colonies containing an HHO1 overexpression vector (data not shown). Therefore, as for HML, HHO1 does not affect silencing within the silenced HMR domain.

View larger version (69K): In this window In a new window Download PPT slide   FIGURE 6.— The effect of histone H1 against silencing is context dependent. (A) Histone H1 overexpression does not modify silencing of ADE2 or URA3 within HML. Gal41–147 was expressed in the reporter strain KIY54. (B) Histone H1 overexpression does not modify silencing of URA3 close to a telomere. (C) Histone H1 overexpression slightly decreases the silencing of TRP1 and URA3 at a distance from the telomere. (D) Histone H1 overexpression has no effect on rDNA silencing.

 
Next, we asked whether HHO1 overexpression could affect telomeric silencing. We tested a strain in which URA3 is located in close proximity to a telomere (Figure 6B). Overexpression of histone H1 did not relieve the silencing of URA3 in this strain. To test whether HHO1 might decrease silencing of a reporter gene placed farther away from the telomere, we used the strain GF97 again. Overexpression of HHO1 in GF97 had a modest inhibitory effect on the silencing of both on URA3 and TRP1 (Figure 6C). These results suggest that histone H1 can oppose telomeric silencing, but only weakly and in certain contexts.

Last, we asked whether HHO1 could affect silencing at the rDNA by using a strain in which URA3 integrated in the rDNA array (Figure 6D). We failed to detect an effect of HHO1 overexpression in that strain. We controlled the behavior of the strain by overexpressing SIR2 and observed, as expected, an increase in silencing (not shown). We can therefore conclude that HHO1 overexpression does not affect rDNA silencing.

Interaction with Histone H2A.Z:

The variant histone H2A.Z can protect genes against the spread of silencing (MENEGHINI et al. 2003). We asked whether the effect of HHO1 required the presence of H2A.Z. For this we deleted HTZ1, the gene encoding H2A.Z, in the reporter strain H0. We found that HHO1 overexpression in this mutant strain still decreased the silencing of URA3, but to a lesser extent than in the wild-type strain, suggesting that HHO1 is partially dependent on HTZ1 for function (Figure 7).

View larger version (39K): In this window In a new window Download PPT slide   FIGURE 7.— Histone H2A. Z is not required for histone H1 to inhibit silencing outside of HML. The effect of HHO1 overexpression was tested in strain H0, or in its derivative lacking HTZ1, the gene that encodes histone H2A.Z.

 

Here we report that histone H1 of S. cerevisiae influences the activity of barrier elements; elevated expression of Hho1p increases the function of Gal4p-generated barriers, while loss of Hho1p decreases their activity. In addition, Hho1p directly inhibits silencing, at least outside of HML. However, this effect is not general, as overexpression of Hho1p had no effect on silencing within HML and HMR, at telomere-proximal positions, or at the rDNA. It had a subtle effect on genes placed at a distance from a telomere, and a stronger effect on genes located outside of HMR-I.

We do not know the reason for that sensitivity of Hho1p effect to context. It could be quantitative: the silencing may be too strong to be opposed in certain regions, such as the telomere ends. Alternatively, it could be qualitative, and the determinants of silencing at the different loci may respond differently to histone H1. Along similar lines, histone acetyltransferases have different effects when targeted to different silenced loci (JACOBSON and PILLUS 2004).

These results contrast with earlier reports that failed to detect any effect of Hho1p on silencing (ESCHER and SCHAFFNER 1997; PATTERTON et al. 1998). We believe there are two reasons for these discrepancies. First, histone H1 is not abundant: the most recent estimate is 6500 molecules/cell (GHAEMMAGHAMI et al. 2003), well under the estimated number of nucleosomes, which is in the order of 50,000. In addition, histone H1 is spread over different sites in the genome; in agreement with previous reports (FREIDKIN and KATCOFF 2001; DOWNS et al. 2003), we were able to detect some histone H1 bound within the rDNA (not shown) as well as in the coding sequence of an expressed gene, ACT1. Therefore, only a fraction of the total histone H1 resides around silenced regions, probably explaining why the deletion has only a subtle effect, while the overexpression has a larger effect. The second reason may be technical: some of the tagged versions of Hho1p that have been used (C-terminally tagged with HA or Myc) do not reproduce the effect of the wild-type protein, at least in our assays.

Finally, how might Hho1p impede silencing? There are two appealing possibilities. One would be that the linker histone influences nucleosomal histone modification, for instance by activating histone acetylation by Sas2p or impeding histone deacetylation by the SIR complex. Another possibility would be that Hho1p competes with the SIR complex for binding to nucleosomes. We have observed decreased binding of the SIR proteins to a silenced locus upon overexpression of Hho1p, suggesting that the second model may be true, but the first possibility cannot be ruled out at this point.

We thank the following investigators for the gift of plasmids and strains: Agnès Bègue, Geneviève Fourel, Susan Gasser, Ann Ehrenhofer-Murray, Ulrich Laemmli, Su-Ju Lin, Sophie Loeillet, Alain Nicolas, Etienne Schwob, David Shore, and Jeff Smith. Work in the Defossez lab was supported by Fondation pour la Recherche Médicale and by Centre National de la Recherche Scientifique (programme ATIP). X.B. acknowledges the support of the National Institutes of Health (RO1 grant GM-62484). The laboratory of E.G. is supported by the Ligue Nationale contre le Cancer (équipe labelisée).

BI, X., Q. YU, J. J. SANDMEIER and Y. ZOU, 2004 Formation of boundaries of transcriptionally silent chromatin by nucleosome-excluding structures. Mol. Cell. Biol. 24: 2118–2131.[Abstract/Free Full Text]

BUSTIN, M., F. CATEZ and J. H. LIM, 2005 The dynamics of histone H1 function in chromatin. Mol. Cell 17: 617–620.[CrossRef][Medline]

CUPERUS, G., and D. SHORE, 2002 Restoration of silencing in Saccharomyces cerevisiae by tethering of a novel Sir2-interacting protein, Esc8. Genetics 162: 633–645.[Abstract/Free Full Text]

DEFOSSEZ, P. A., and E. GILSON, 2002 The vertebrate protein CTCF functions as an insulator in S. cerevisiae. Nucleic Acids Res. 23: 5136–5141.

DHILLON, N., and R. T. KAMAKAKA, 2002 Breaking through to the other side: silencers and barriers. Curr. Opin. Genet. Dev. 12: 188–192.[CrossRef][Medline]

DONZE, D., and R. T. KAMAKAKA, 2001 RNA polymerase III and RNA polymerase II promoter complexes are heterochromatin barriers in Saccharomyces cerevisiae. EMBO J. 20: 520–531.[CrossRef][Medline]

DONZE, D., and R. T. KAMAKAKA, 2002 Braking the silence: how heterochromatic gene repression is stopped in its tracks. BioEssays 24: 344–349.[CrossRef][Medline]

DONZE, D., C. R. ADAMS, J. RINE and R. T. KAMAKAKA, 1999 The boundaries of the silenced HMR domain in Saccharomyces cerevisiae. Genes Dev. 13: 698–708.[Abstract/Free Full Text]

DOWNS, J. A., E. KOSMIDOU, A. MORGAN and S. P. JACKSON, 2003 Suppression of homologous recombination by the Saccharomyces cerevisiae linker histone. Mol. Cell 11: 1685–1692.[CrossRef][Medline]

ESCHER, D., and W. SCHAFFNER, 1997 Gene activation at a distance and telomeric silencing are not affected by yeast histone H1. Mol. Gen. Genet. 256: 456–461.[CrossRef][Medline]

FOUREL, G., E. REVARDEL, C. E. KOERING and E. GILSON, 1999 Cohabitation of insulators and silencing elements in yeast subtelomeric regions. EMBO J. 18: 2522–2537.[CrossRef][Medline]

FOUREL, G., C. BOSCHERON, E. REVARDEL, E. LEBRUN, Y. F. HU et al., 2001 An activation-independent role of transcription factors in insulator function. EMBO Rep. 2: 124–132.[CrossRef][Medline]

FREIDKIN, I., and D. J. KATCOFF, 2001 Specific distribution of the Saccharomyces cerevisiae linker histone homolog HHO1p in the chromatin. Nucleic Acids Res. 29: 4043–4051.[Abstract/Free Full Text]

GHAEMMAGHAMI, S., W. K. HUH, K. BOWER, R. W. HOWSON, A. BELLE et al., 2003 Global analysis of protein expression in yeast. Nature 425: 737–741.[CrossRef][Medline]

HOFFMAN, C. S., and F. WINSTON, 1987 A ten-minute DNA preparation from yeast efficiently releases autonomous plasmids for transformation of Escherichia coli. Gene 57: 267–272.[CrossRef][Medline]

ISHII, K., and U. K. LAEMMLI, 2003 Structural and dynamic functions establish chromatin domains. Mol. Cell 11: 237–248.[CrossRef][Medline]

ISHII, K., G. ARIB, C. LIN, G. VAN HOUWE and U. K. LAEMMLI, 2002 Chromatin boundaries in budding yeast: the nuclear pore connection. Cell 109: 551–562.[CrossRef][Medline]

JACOBSON, S., and L. PILLUS, 2004 Molecular requirements for gene expression mediated by targeted histone acetyltransferases. Mol. Cell. Biol. 24: 6029–6039.[Abstract/Free Full Text]

KIMURA, A., and M. HORIKOSHI, 2004 Partition of distinct chromosomal regions: negotiable border and fixed border. Genes Cells 9: 499–508.[Abstract/Free Full Text]

KIMURA, A., T. UMEHARA and M. HORIKOSHI, 2002 Chromosomal gradient of histone acetylation established by Sas2p and Sir2p functions as a shield against gene silencing. Nat. Genet. 32: 370–377.[CrossRef][Medline]

KURAS, L., and K. STRUHL, 1999 Binding of TBP to promoters in vivo is stimulated by activators and requires Pol II holoenzyme. Nature 399: 609–613.[CrossRef][Medline]

KUSHNIROV, V. V., 2000 Rapid and reliable protein extraction from yeast. Yeast 16: 857–860.[CrossRef][Medline]

MENEGHINI, M. D., M. WU and H. D. MADHANI, 2003 Conserved histone variant H2A.Z protects euchromatin from the ectopic spread of silent heterochromatin. Cell 112: 725–736.[CrossRef][Medline]

OKI, M., L. VALENZUELA, T. CHIBA, T. ITO and R. T. KAMAKAKA, 2004 Barrier proteins remodel and modify chromatin to restrict silenced domains. Mol. Cell. Biol. 24: 1956–1967.[Abstract/Free Full Text]

PATTERTON, H. G., C. C. LANDEL, D. LANDSMAN, C. L. PETERSON and R. T. SIMPSON, 1998 The biochemical and phenotypic characterization of Hho1p, the putative linker histone H1 of Saccharomyces cerevisiae. J. Biol. Chem. 273: 7268–7276.[Abstract/Free Full Text]

REN, B., F. ROBERT, J. J. WYRICK, O. APARICIO, E. G. JENNINGS et al., 2000 Genome-wide location and function of DNA binding proteins. Science 290: 2306–2309.[Abstract/Free Full Text]

RUSCHE, L. N., A. L. KIRCHMAIER and J. RINE, 2003 The establishment, inheritance, and function of silenced chromatin in Saccharomyces cerevisiae. Annu. Rev. Biochem. 72: 481–516.[CrossRef][Medline]

SMITH, J. S., C. B. BRACHMANN, L. PILLUS and J. D. BOEKE, 1998 Distribution of a limited Sir2 protein pool regulates the strength of yeast rDNA silencing and is modulated by Sir4p. Genetics 149: 1205–1219.[Abstract/Free Full Text]

SUKA, N., K. LUO and M. GRUNSTEIN, 2002 Sir2p and Sas2p opposingly regulate acetylation of yeast histone H4 lysine16 and spreading of heterochromatin. Nat. Genet. 32: 378–383.[CrossRef][Medline]

Communicating editor: L. PILLUS

International Access Link

Online ISS 1091-6490
Copyright 2008 by the Genetics Society of America
phone: 412-268-1812   fax: 412-268-1813   email: genetics-gsa{at}andrew.cmu.edu