|Home | About | Journals | Submit | Contact Us | Français|
The histone genes are an important group of cell cycle regulated genes whose transcription is activated during the G1/S transition and repressed in early G1, late S, and G2/M. The HIR complex, comprised of Hir1, Hir2, Hir3 and Hpc2, regulates three of the four histone gene loci. While relief of repression at the G1/S boundary involves the HIR complex, as well as other cofactors, the mechanism by which this derepression occurs remains unknown. To better understand how transcriptional repression contributes to periodic expression in the cell cycle, we sought to identify the cell cycle signals required to alleviate HIR-mediated repression of the histone genes. By measuring histone gene transcription in strains with various combinations of clb mutations, we found that the mitotic Clb1/Clb2 cyclins are required to alleviate Hir-mediated repression during the G1/S transition and that Clb2 physically interacts with the HIR complex. While the HIR complex regulates histone genes transcription in combination with two other histone H3/H4 chaperones, Asf1 and Rtt106, our data demonstrate that the mitotic Clb cyclins are necessary to specifically alleviate the repressive action of the HIR complex itself in order to allow proper expression of the histone genes in late G1/early S phase.
Progression through the eukaryotic cell cycle is accompanied by periodic fluctuations in the expression of numerous genes . The products of most of these genes are either required for cell cycle progression or the control of the various cell cycle transitions [2-6]. A number of transcription factors that control periodic transcription have been identified in yeast and mammals [2, 3, 7]. In most cases, the function of these factors is controlled by different cyclin-dependent kinases (CDKs, e.g. Cdk1/Cdc28) , generally by phosphorylation. In mammals, for example, Cyclin A/Cdk2-mediated phosphorylation stimulates the activity of p53 and B-myb and inhibits the activity of E2F . In budding yeast, a genome-wide analysis revealed that half of the 800 cell cycle regulated genes are responsive to either elevated Cln3-Cdk1 or Clb2-Cdk1 activity, or to both activities . For example, the SBF and MBF transcription factors that regulate genes required for G1 and S phase progression are activated by Cln3/Cdk1, but inhibited by Clb2/Cdk1 kinase activity [6, 8, 9]. Other transcription factors in yeast with roles in periodic transcription, such as Mcm1/SFF and Swi5, are subject to positive and negative regulation, respectively, by Clb2-Cdk1 [6-8, 10-13].
In budding yeast, most transcription factors regulated by CDKs are activators, and transcriptional activation is postulated to be the predominant mechanism controlling periodic transcription. In contrast, transcriptional repression is a major mechanism of gene regulation in the vertebrate cell cycle. For example, a number of genes required for S-phase entry are transcribed at the G1/S transition, and periodic control is achieved via E2F binding sites in their promoters [14-17]. The binding of the retinoblastoma tumor suppressor pRb to E2F alters its activity, and as a result, E2F-regulated genes are not transcribed [14, 15, 17]. Association of pRb with E2F negatively affects transcription in two ways: (i) it prevents E2F from activating transcription at some promoters and (ii) when recruited by E2F, pRb actively represses transcription at others [15, 18]. E2F-RB complexes are prevalent in G0 and early G1 phase of the cell cycle and are disrupted in late G1. This is attributed primarily to the phosphorylation of pRb by the G1 CDK’s, cyclin D/cdk4,6 and cyclin E/cdk2 [19, 20]. The disruption of E2F-pRB complexes at the G1/S transition leads to both E2F-dependent trans-activation and relief of pRB-dependent repression. While the regulation of periodic transcription in yeast appears to be regulated mainly via transcriptional activation, it is interesting to note that the recently characterized protein Whi5 acts as an SBF-associated transcriptional repressor [21, 22]. Inactivation of WHI5 leads to premature activation of G1-specific transcription in G1 phase. Whi5 associates with G1-specific promoters in an SBF-dependent manner and is released from the DNA concurrent with transcriptional activation. Similar to CDK phosphorylation of pRb leading to activation of E2F-responsive genes, Cln3-associated Cdk1 phosphorylation of Whi5 promotes its release from SBF, resulting in derepression of SBF-dependent transcription [1, 21, 22].
The histone genes are an important group of cell-cycle-regulated genes. Transcription of these genes is tightly regulated and occurs during late G1 and S phases of the cell cycle to produce large quantities of new histones during DNA replication in S phase (reviewed in [23, 24]). Aberrant expression of the histone genes outside of S phase can be toxic and lead to alterations in chromatin structure . In budding yeast, both positive and negative regulation contributes to the pattern of histone gene transcription during the cell cycle. In addition to conserved upstream activation elements (UAS) in their promoters, six of the eight histone genes (HTA1-HTB1, HHT1-HHF1 and HHT2-HHF2) contain a negative regulatory site, named the NEG or CCR region, which is in close proximity to the UAS elements [25, 26]. The cell cycle regulated UAS elements are required to activate transcription at the G1/S transition , while the negative element is required to repress transcription in early G1, late S and G2 and M phases of the cell cycle [25, 26, 28]. In the absence of the negative element, while histone mRNA levels still peak in early S-phase, transcription occurs inappropriately in G1, G2 and M phases [26, 28]. The HTA1-HTB1 negative site functions in an UAS- and activator-independent manner and confers periodic transcription to a constitutive promoter [26, 28].
Several trans-acting factors that act at the negative site to repress transcription were identified through genetic screens [29, 30]. Among these were the four evolutionarily conserved HIR1, HIR2, HIR3 and HPC2 genes [28, 30-32]. We have previously shown that these four proteins stably assemble to form the HIR corepressor complex, which constitutes a novel histone chaperone complex ([33, 34] and reviewed in ). Although it was initially reported that the Hir/Hpc proteins did not possess intrinsic DNA binding activity [32, 36], the HIR complex was later observed to bind to DNA in a non-specific manner . Thus, the HIR complex is postulated to be recruited to the negative site of the histone gene promoters by an as yet uncharacterized sequence-specific DNA binding factor . However, the constitutively expressed Hir1 and Hir2 proteins can bypass the requirement for the NEG site when artificially recruited to DNA and repress transcription at the appropriate time in the cell cycle . Recently, we have shown that the CDI domain located within the N-terminal region of Hpc2 is essential for the recruitment of the HIR complex to the HIR-dependent histone gene loci , strongly suggesting that the Hpc2 CDI domain could directly bind to the unknown NEG sequence-specific DNA binding factor. Neither the levels of the HIR proteins nor their presence at the histone gene promoters appear to be cell cycle-regulated [28, 32, 39], which suggests that the repressive activity of the HIR complex is antagonized during the cell cycle to allow the histone genes to be transcribed .
The role of the HIR complex in repressing transcription of the histone genes outside of S phase has recently been linked to its association and interplay with two other H3/H4 histone chaperones – Asf1 and Rtt106 . The recruitment of Asf1 to the histone genes is dependent on the HIR complex, while Rtt106 recruitment is dependent on both Asf1 and the HIR complex . This suggests that it is the concerted actions of the HIR complex, Asf1 and Rtt106 that is required to maintain a repressive chromatin structure in order to prevent inappropriate transcription of the histone genes outside of S phase. We have recently shown that Rtt106 also plays an essential role in the recruitment of both SWI/SNF and RSC chromatin remodeling complexes to the HIR-dependent histone genes . While RSC is present at the histone genes outside of S phase concomitant with their repression , SWI/SNF is recruited to the histone genes in late G1/early S phase  in order to activate their expression . Therefore, Rtt106 is both a co-repressor and a co-activator of histone gene transcription .
To gain a better understanding of the mechanisms by which transcriptional repression contributes to periodic expression in the cell cycle, we identified the cell cycle signals required to alleviate HIR-mediated repression of the histone genes at the G1/S transition. We found that the mitotic Clb1/Clb2 cyclins are required for maximal transcription of the histone genes by alleviating Hir-mediated repression at the HTA1-HTB1 promoter. Furthermore, we showed that Clb2 physically interacts with the HIR complex in vivo. We propose that the role of the mitotic Clb cyclins is to specifically alleviate repression exerted by the HIR complex at the histone gene loci, as the cyclins do not appear to relieve repression mediated by Asf1 or Rtt106 at the HTA1-HTB1 promoter or to regulate the recruitment of the SWI/SNF complex to this promoter.
Yeast strains were grown in rich YEPD medium (1% yeast extract, 2% peptone and 2% dextrose) or in minimal synthetic dextrose (SD) medium (0.67% yeast nitrogen base, 2% dextrose and appropriate drop-out mixtures of amino acids and bases) and genetically manipulated by standard techniques. Cells were transformed with plasmids using the DMSO-enhanced whole cell transformation protocol as described by Hill and coworkers .
The yeast strains used in this study are listed in Table 1. All strains are isogenic or congenic to W303-1A or BY4741.
Plasmids for HIR1-HA, HIR2-HA and Clb2-HA fusion genes were constructed by sequential cloning of a PCR amplified product from the native promoter (500 bp upstream of +1 ATG) and coding sequences into YCp:URA3 (pRS416) or YCp:LEU2 (pRS415) vectors that contain three HA-tag sequences followed by an ADH1 terminator sequence. All plasmids were verified by DNA sequencing. Details of constructs used are available upon request.
For cell synchronization, cells were grown in YEPD medium and ~5×106 cells/ml were incubated with α-factor (Sigma) at 5 μg/ml for BAR1+ strains or 0.05 μg/ml for bar1Δ strains for 2.5 − 3 hours at 24°C, or until >97% of cells exhibited a characteristic G1 arrest morphology. Mating pheromone was removed by centrifugation, the cells were washed once with dH20, resuspended in appropriate prewarmed medium in the presence of 5mM CaCl2 and 40 μg/ml pronase, and then incubated at 24°C or 37°C. To assay the degree of synchronization, the percentage of budded cells was scored over time after G1 release by phase-contrast microscopy and the expression of the cell cycle regulated CLN2 gene was monitored by RT-qPCR. Samples for RNA or protein analysis were removed at regular intervals after release and placed in tubes containing 100 μl of a 10% sodium azide solution. The cells were centrifuged and the resulting cell pellets were frozen in dry-ice/ethanol bath and stored at −80°C.
Total RNA was extracted from synchronized or exponentially growing cells as previously described [44, 45] and used in S1 nuclease protection assays. S1-nuclease hybridization probes to detect HTB1, HTB2 and RP51A mRNAs were prepared by end-labeling 0.7 Kb SacI/HindIII (HTB1), 0.7 Kb HindIII/AccI (HTB2), or 0.6 Kb AvaII/SalI (RP51A) double-stranded DNA fragments with 32P ATP in the presence of T4 polynucleotide kinase as previously described . Ten nanograms of each probe were hybridized to 20 μg of total RNA. Forty μg of total RNA were used when expression levels were predicted to be low. Hybridization was performed at 47°C for at least 12 hours, and S1 nuclease digestion was performed for 1 hour at 15°C at an enzyme concentration of 1000 U/ml as previously described . Products were resolved on 8M urea-4%polyacrilamide gels and exposed to X-ray film for autoradiography. Quantitation was performed on a FUJI Bio-Imaging analyzer using MACBAS software.
Total RNA was extracted from cells using the E.Z.N.A. Total RNA Kit I (Omega, R6834-02) as described in the manufacturer’s instructions. cDNA was then synthesized from 400ng of total RNA using the High Capacity cDNA Reverse Transcription Kit (Applied Biosystems product # 4368814). Quantitative PCR was used to analyze the cDNA using Power SYBR PCR Master Mix (Applied Biosystems, product # 4367659). qPCR was carried out using the MyiQ (BioRad) machine. The amplified DNA was normalized to DNA amplified from ACT1 or SCR1 RNA samples, as indicated.
Chromatin immunoprecipitation (ChIP) assays were performed as described previously , except that samples taken at the indicated time-points after release at 24°C or 37°C from α-factor arrest were diluted 1:1 with Milli-Q water at room temperature or 10°C, respectively, to obtain the same temperature in the samples prior to addition of formaldehyde. Samples were cross-linked for 1 h at room temperature. SWI/SNF complex in cross-linked chromatin extract was immunoprecipitated using an antibody specific for the N-terminal end of Swi2/Snf2 by over-night incubation at 4°C, followed by 1 h incubation with Protein A beads. Relative levels of immunoprecipitated HTA1-HTB1 promoter sequence were quantified by qPCR. An intergenic region of chromosome V (INTV) was used as a non-target control. Primer sequences are available upon request.
Yeast whole cell extracts were prepared in lysis buffer (40 mM HEPES-KOH pH 7.5, 350 mM NaCl, 10% Glycerol, 0.1% Tween-20, 1 μg/ml pepstatin, 2 μg/ml leupeptin, 0.5 mM DTT and 1 mM PMSF). One milligram of protein extract was immunoprecipitated overnight at 4°C with 15 μl of IgG Sepharose beads (GE Healthcare Life Science, product # 17096901) that binds to the Protein A tag of the TAP tagged subunit. The beads were washed three times with lysis buffer and samples were prepared for Western blot analysis. The immunoprecipitates were resolved on 8 or 10% SDS-PAGE and blotted to a PVDF membrane (Thermo Scientific, product # 88518). Hybridization was carried out using antibodies against HA (Roche product # 12013819001) or rabbit PAP (Peroxidase-Anti-Peroxidase) antibody (Sigma-211 Aldrich, product # P1291) to detect the TAP tag.
All four histone loci are transcriptionally activated in late G1 [46, 47], and the transcription of the HTA1 gene has been shown to require a functional CDC4 gene product ([32, 48], and Osley, unpublished observations). The Cdc4, Cdc34, Cdc53 and Skp1 proteins form a complex, SCFCDC4, which is required for the ubiquitin-dependent proteolysis of a number of proteins that regulate cell cycle progression [49-53]. CDC4 is critical for the progression of yeast cells into S phase, as, in the absence of CDC4 function, cells perform ‘START’-related events but fail to perform subsequent events such as DNA replication, spindle formation, and cytokinesis (for a review, see reference ). These observations raised the possibility that both the initiation of DNA replication and histone synthesis may depend on the same regulatory step. We asked whether CDC4 played a role in transcription of the two HTA-HTB loci coding for histones H2A and H2B. We examined the transcription of HTA2 and HTB2 genes in a strain carrying a temperature-sensitive (Ts-) allele of CDC4 and compared it to HTA1 and HTB1 transcription in the same cdc4ts mutant. Although the HTA2 and HTB2 genes do not contain a negative element in their promoter and are not regulated by the HIR complex [24, 55], both are activated at the same time in late G1 as the HTA1 and HTB1 genes (Fig. 1A). As shown in Fig. 1B, after release of a cdc4ts mutant from α-factor arrest into nonpermissive (37°C) conditions, both HTA2 and HTB2 mRNA levels were greatly reduced and showed delayed accumulation when compared to permissive conditions (24°C), in which a normal periodic accumulation of transcripts occurred in S phase. As previously reported , HTA1 and HTB1 mRNA levels were reduced by at least 3 fold in S phase when cdc4ts cells were released at the restrictive temperature (Fig. 1B). However, a low level of HTA-HTB transcription persisted in cells blocked at the CDC4-dependent step (Fig.1B, 37°C). This is consistent with the fact that at the non-permissive temperature, cdc4ts cells arrest in late G1/early S phase as monitored by the expression of CLN2, which is activated but then remains highly transcribed (data not shown). In contrast, in a wild-type strain, there was no difference in the levels of HTA1, HTB1 and HTB2 mRNAs between 24°C and 37°C, expect for a slight reduction in the levels of HTA2 mRNAs. However, the timing of the first cell cycle in the wild-type strain appears to be slightly earlier at 37°C than at 24°C (Fig. 1A).
These results showed that activation of both the HTA1-HTB1 and HTA2-HTB2 loci is dependent on events regulated by CDC4. However, the low level of transcripts and delayed expression of the histone genes seen in the cdc4ts mutant at the nonpermissive temperature suggested that there is also a CDC4-independent step in the regulation of these loci.
A major role for the SCFCDC4 complex is the ubiquitination of p40SIC1, which targets it for degradation . p40SIC1 is a potent Cdk1 inhibitor, and its destruction at the G1/S transition leads to the activation of all six Clb/Cdk1 complexes (reviewed in [56-58]. To determine whether the CDC4-dependence of HTA-HTB transcription reflected a requirement for Clb kinases, we monitored the expression of the two loci in synchronized cells of different clb mutants. The coupling of histone synthesis to DNA replication suggested that the S-phase cyclins CLB5 and CLB6, which are involved in the initiation of DNA replication [59-61], might also regulate histone gene transcription. We examined HTB1 and HTB2 mRNA levels during the cell cycle in a strain carrying deletions of both cyclin genes. Such a strain is viable due to the redundancy among Clb cyclins, but shows a delayed entry into S-phase [52, 61, 62]. As can be seen in Figure 2A and 2B, there was no difference between the wild type and clb5Δclb6Δ strains in either the timing or levels of HTB1 and HTB2 mRNAs. The only difference was a delay of 15-30 min in HTB expression during the second cell cycle in the mutant strain, which corresponded to its delayed entry into S-phase (Figure 2C and 2D) . Thus, these results show that the Clb5/Clb6 cyclins and their associated kinase activities are not directly involved in the transcriptional control of either HTA-HTB locus.
New histone synthesis occurs at the G1/S boundary and is required for the G2/M transition [23, 24]. Cells unable to synthesize histones H2B and H4 arrest in G2 presumably due to the abnormal chromatin structure of their chromosomes, which are duplicated in the absence of histone synthesis [63, 64]. Interestingly, a strain carrying mutations in all four mitotic cyclin genes, CLB1, CLB2, CLB3, and CLB4, arrests in G2 with duplicated but unsegregated chromosomes [8, 65]. We therefore asked if the CDC4-dependence of histone gene transcription reflected a specific requirement for the mitotic Clb cyclins.
We monitored histone gene expression during the cell cycle in a strain deleted of CLB1, CLB3, and CLB4 genes and carrying a temperature sensitive allele of CLB2  after release of cells from α-factor arrest to permissive (24°C) and nonpermissive temperatures (37°C) (Fig. 3). The mRNA levels produced from all four HTA and HTB genes were reduced between 2.7- to 5-fold in S phase when cells were shifted to 37°C compared to 24°C, but were not completely eliminated (Fig. 3 and data not shown). We also observed a similar reduction in the expression of the HHT1, HHF1 and HHT2 genes when cells were shifted to 37°C (data not shown). As discussed above, the synthesis of new histones is imperative for passage through the G2/M transition [23, 24]. Therefore, as expected, cells placed at the nonpermissive temperature (37°C) suffer from G2/M arrest, which is apparent by the absence of peak expression during the second cell cycle. The expression of CLN2, which was used as a control for cell cycle progression, showed that cells were able to enter S phase at 37°C, but arrested shortly thereafter as indicated by the constitutive levels of CLN2 RNA compared to permissive conditions (24°C) (Fig. 3D). These results indicate that the mitotic Clb cyclins are required for the full transcriptional activation of both the HTA1-HTB1 and HTA2-HTB2 loci.
Clb1/Clb2 and Clb3/Clb4 constitute two distinct pairs of mitotic cyclins, each of which is presumed to confer specialized functions upon association with the cyclin-dependent kinase Cdk1 [65-67]. However, due to the redundancy among the mitotic cyclins [52, 61, 62], the results shown in Fig. 3 could not distinguish between a role for one or the other of the two mitotic cyclin pairs. We therefore measured expression of both HTA-HTB loci during the cell cycle in a strain carrying deletions of the CLB3 and CLB4 genes. No significant differences in either HTA1, HTB1 or HTA2 expression were observed between the clb3Δclb4Δ mutant and a wild type (WT) strain (Fig. 4A), suggesting that Clb3/Clb4 are not responsible for histone gene expression.
We then monitored expression of the histone genes during the cell cycle in a strain deleted for the CLB1 gene and carrying a temperature sensitive allele of CLB2, but containing functional CLB3/CLB4 cyclin genes (Fig. 4B). There was a 2- to 3-fold reduction in the levels of HTA1, HTB1 and HTA2 RNAs in the clb1Δclb2ts mutant at the nonpermissive temperature compared to the permissive temperature (Fig. 4B). However, the reduction in the HTA and HTB RNA levels in the clb1Δclb2ts mutant strain was not totally equivalent to the reduction observed in the clb1Δ clb2ts clb3Δ clb4Δ mutant strain (Fig. 3). We propose that this could be due to the redundancy among the mitotic cyclins [52, 61, 62], and thus Clb3/Clb4 may be able to partially compensate for the loss of Clb1 and Clb2 at the nonpermissive temperature in the clb1Δclb2ts mutant strain. Nevertheless, because the expression of the histone genes was not affected in the clb3Δ clb4Δ mutant, these results strongly suggest that the Clb1/Clb2 cyclins play a major role in the control of transcription of both the HTA1-HTB1 and HTA2-HTB2 loci. Because the reduction in mRNA production from the HTA1-HTB1 and HTA2-HTB2 loci in the absence of functional mitotic Clb cyclins resembled the effect observed in the cdc4ts mutant, we conclude that both the Cdc4- and the Clb1/Clb2-dependence of their transcription reflects a requirement for an active mitotic Clb1/Clb2-Cdk1 kinase.
Unpublished observations by Dimova and Osley showed that the HTA1-HTB1 negative element confers mitotic Clb cyclin-dependence on expression of a heterologous promoter . We therefore asked whether the HIR complex, which negatively regulates HTA1-HTB1 transcription through the negative site, is responsible for the Clb1/Clb2-dependence of this locus. We disrupted the HIR1 gene in a clb1Δclb2tsclb3Δclb4Δ strain and examined HTA1 and HTB1 mRNA levels in synchronized cultures at both 24°C and 37°C (Fig. 5A and B). As previously shown [26, 29](and data not shown), deletion of HIR1, similar to removal of the negative site, resulted in a higher level of mRNA accumulation in G1, G2, and M phases, although peak levels were still achieved in early S-phase under permissive conditions (Fig. 5A and B, 24°C). Strikingly, under nonpermissive conditions (37°C) (Fig. 5A and B, 37°C), HTA1 and HTB1 mRNA accumulated to the same levels as under permissive conditions (24°C) in the first cell cycle in a clb1Δclb2tsclb3Δclb4Δhir1Δ strain. However, due to subsequent G2/M arrest at the nonpermissive temperature, as shown by continuous CLN2 expression (Fig. 5D), no HTA1 or HTB1 RNAs accumulated during the second cell cycle at 37°C. Contrary to the effect on HTA1-HTB1 expression, HIR1 deletion did not restore HTA2 mRNA levels at the nonpermissive temperature (Fig. 5C). This was expected, as transcription of the HTA2-HTB2 locus is HIR-independent [29, 30]. These results indicate that deletion of HIR1, and hence disruption of the HIR complex, can completely bypass the requirement for the mitotic Clb cyclins to restore the expression of HTA1-HTB1 under nonpermissive conditions. Thus, these results indicate that Clb1/Clb2 is required for the relief of HIR-mediated repression of the HTA1-HTB1 locus to allow maximal transcription of these genes to occur at the proper time in the cell cycle.
Because of the functional interactions between the HIR complex and the mitotic Clb cyclins, we asked whether Clb2 and the HIR complex might be physically associated in yeast using a co-immunoprecipitation assay with epitope tagged Clb2, Hir1 and Hir2 proteins. As shown in Fig. 6A, immunoprecipitated Hir1-TAP and Hir2-TAP specifically associated with Clb2-HA (Fig. 6A, lane 3 and 5 of the IP panel), while no Clb2-HA was detected in an untagged strain (Fig. 6A, lane 1 IP panel). We performed the reciprocal immunoprecipitation, and showed that Clb2-TAP co-precipitated Hir1-HA and Hir2-HA (Fig. 6B and C, IP panel, lanes 3). These data demonstrate that the HIR complex specifically associates with the mitotic Clb2 cyclin in vivo, and suggest that the association might reflect a subsequent modification of the complex by the Clb1/Clb2-Cdk1 kinase. Clb1/Clb2 dependent relief of HIR-mediated repression at the HTA1-HTB1 locus might occur, for example, through the targeting of the HIR co-repressor complex for inactivation by the Clb1/Clb2-dependent Cdk1 kinase or through the recruitment of the SWI/SNF chromatin remodeling complex which is required for activation of the histone genes .
We have recently shown that the SWI/SNF complex is recruited in late G1 to the HTA1-HTB1 promoter by Rtt106  and is required for the activation of histone gene transcription at the G1/S phase boundary . We therefore asked whether the Clb1/Clb2 cyclins could alleviate HIR-mediated repression by controlling the recruitment of SWI/SNF to the HTA1-HTB1 locus. We monitored the presence of the SWI/SNF complex at the HTA1-HTB1 promoter by chromatin immunoprecipitation (ChIP) using an antibody against the N-terminal domain of the Swi2/Snf2 catalytic subunit in both a wild-type and clb1Δclb2tsclb3Δclb4Δ strain released from G1 arrest at both the permissive (24°C) and nonpermissive temperature (37°C). As seen in Figure 7, there were no differences in the recruitment of SWI/SNF to the HTA1-HTB1 promoter in the clb1Δclb2tsclb3Δclb4Δ strain compared to a wild-type strain at either temperature. However, we observed a slight delay and decrease in SWI/SNF recruitment to this promoter at 37°C in both strains compared to 24°C (Fig. 7). Thus, these results indicate that the decrease in expression of the HTA1-HTB1 genes at the nonpermissive (37°C) temperature in the clb1Δclb2tsclb3Δclb4Δ strain (Fig. 3) is not due to an alteration in recruitment of the SWI/SNF complex. These data suggest that the function of the mitotic Clb cyclins could be to directly turn off HIR-mediated repression in late G1 to allow expression of HTA1-HTB1 locus.
As discussed in the Introduction, the HIR complex associates with the H3/H4 histone chaperones Asf1 and Rtt106 , both of which play a role in the transcriptional repression of the HIR-dependent histone genes [40, 69]. Fillingham and colleagues  reported that the presence of Rtt106 at the HTA1-HTB1 promoter was dependent on the HIR complex and Asf1, while the presence of Asf1 was only dependent on the HIR complex, and that all of these factors were present at the HIR-dependent histone genes throughout the cell cycle. Interestingly, deletion of ASF1, RTT106 or genes encoding HIR complex subunits results in nucleosome depletion at the HTA1-HTB1 promoter , suggesting that the repressive chromatin structure established by Asf1/HIR/Rtt106 represents the main mechanism to repress the histone genes. To better understand the mechanism by which the mitotic Clb cyclins alleviate HIR-dependent repression of the HTA1-HTB1 locus, we sought to determine which of these three factors is the main target of the mitotic Clb cyclins. We first asked whether deletion of RTT106 or ASF1 in the clb1Δclb2tsclb3Δclb4Δ strain could bypass the requirement for the mitotic Clb cyclins at the nonpermissive temperature as observed when HIR1 was deleted (Fig 5). As shown in Figure 8, deletion of either RTT106 or ASF1 in the quadruple clb strain did not significantly restore the peak of expression of HTA1 as observed upon deletion of HIR1 in the clb1Δclb2tsclb3Δclb4Δ strain at the nonpermissive temperature (Fig. 8 compare panels B and C to D). However, a slight but persistent accumulation of HTA1 RNA was observed after S phase when the asf1Δ or rtt106Δ cells entered G2/M phase at the nonpermissive temperature (Fig. 8B and C, 60, 75 and 90min) when compared to the clb1Δclb2tsclb3Δclb4Δ strain (Fig. 8 panel A). Similar data were obtained for HTB1 RNA levels (data not shown). Because the presence of the HIR complex at the HTA1-HTB1 promoter has been shown to be independent of ASF1 and RTT106 , we conclude that the reduced HTA1 expression in lateG1/early S phase at the nonpermissive temperature in the clb1Δclb2tsclb3Δclb4Δ strain primarily reflects the repressive action exerted by the HIR complex alone. Taken together, these results strongly suggest that the role of the mitotic Clb cyclins, and in particular Clb1/Clb2, is to alleviate the repressive effects of the HIR complex itself at the HTA1-HTB1 locus in late G1/early S phase to allow proper histone gene expression. Thus, our results strongly suggest that in addition to the repressive activity of Asf1 and Rtt106, likely through their roles in nucleosome assembly, the HIR complex represses histone gene transcription outside of S phase by an independent mechanism that is alleviated by the mitotic Clb cyclins in late G1 to allow the full expression of HIR-dependent histone genes (Fig. 9).
The yeast histone genes are subject to both positive and negative regulation to achieve maximal transcription in S phase of the cell cycle. Conserved cell cycle regulated UAS elements found in all of the histone gene promoters are required to activate transcription at the G1/S transition through the binding of activators that include SBF, MBF, Spt10 and Spt21 [24, 26, 27, 70-77]. Negative regulation is achieved by the constitutive binding of the evolutionarily conserved HIR complex to a negative regulatory element in the promoters of three of the four histone gene loci (reviewed in [24, 35]). In addition to the action of activators, HIR-dependent repression of these promoters must be relieved at the G1/S boundary to achieve maximal levels of histone gene-specific RNAs in S phase. In this study, we have shown that the transcriptional activation of the four genes encoding histones H2A and H2B (the HTA1-HTB1 and HTA2-HTB2 loci) is dependent upon completion of late G1 events controlled by the CDC4 gene product. Furthermore, our results indicate that in late G1/early S phase, full transcriptional activation of the HIR-dependent histone genes HTA1, and HTB1, as well as the HIR-independent histone genes HTA2 and HTB2, requires the mitotic Clb1/Clb2 cyclins, which is likely to account for the Cdc4-dependence of histone gene transcription. Additionally, we found that the two other Hir-dependent histone gene loci, HHT1-HHF1 and HHT2-HHF2, also require the mitotic Clb cyclins as well as Cdc4 to become fully expressed in G1/S phase (data not shown). A strain carrying mutations in all four mitotic cyclin genes, CLB1, CLB2, CLB3, and CLB4, arrests in G2 with duplicated but unsegregated chromosomes [8, 65]. G2 arrest is also seen in cells that are unable to synthesize histones H2B and H4, and this has been attributed to an abnormal chromatin structure of the duplicated chromosomes [63, 64]. Thus, our data showing that the mitotic Clb cyclins are necessary for the correct expression of all four histone gene loci in late G1/early S phase may partially explain the G2 arrest phenotype observed in the absence of functional mitotic Clb cyclins in a clb1Δclb2tsclb3Δclb4Δ strain [8, 65].
The observation that the transcription of the histone genes is dependent upon the mitotic Clb cyclins was at first surprising given that histone gene transcription is coupled to S phase and the mitotic Clb cyclin-associated Cdk1 kinase does not become fully active until G2 phase [52, 66, 78]. However, at the G1/S transition when histone gene transcription is activated , the Clb1/Clb2-associated Cdk1 kinase has been shown to be active, although kinase levels are significantly lower than in G2 phase [79, 80]. Interestingly, we found that in vivo Clb2 associates with the HIR complex (Fig. 6). Moreover, our preliminary data show that the Clb2-Cdk1 complex can phosphorylate all four subunits of the HIR complex in vitro (data not shown). Additionally, a global study to identify substrates of Clb2-Cdk1 reported that Hpc2 was among the 200 proteins found in this screen , thus further corroborating our findings. Analysis of the protein sequences of the four subunits of the HIR complex reveals that they all contain at least one RXL motif, a feature found in Cdk1 kinase substrates that mediates an interaction with the hydrophobic patch found on cyclins  (data not shown), as well as multiple putative Cdk1 phosphorylation sites . In addition, the human homolog of Hir1 and Hir2, HIRA, was reported to be phosphorylated in S phase by cyclin A-cdk2 . Consistent with the notion that human HIRA is itself a cell cycle regulator, ectopic expression of HIRA causes cell cycle arrest in S phase . Confirmation of the yeast HIR complex as a direct target of the Clb-Cdk1 kinase and identification of the HIR complex residues phosphorylated by this kinase will ultimately shed light on the role played by the Clb cyclins in the function of the HIR complex to control transcription of the histone genes. Interestingly, we also found that the expression of the HIR-independent histone genes, HTA2 and HTB2, is also dependent on the Clb1/Clb2 cyclins. However, no repressors for HTA2-HTB2 locus are currently known, so this suggests that the mitotic Clb cyclins could be required either to alleviate the repression of an unknown repressor or to directly activate the transcription of this gene pair. Thus, additional work is needed to understand the specific role of the mitotic Clb cyclins in the transcriptional regulation of the HTA2-HTB2 locus.
We have shown here that inactivation of the negative regulatory system by deletion of the HIR1 gene relieved the dependence of HTA1-HTB1 transcription on the mitotic Clb1/Cb2 cyclins (Fig. 5). The HIR regulatory system includes not only the HIR complex proteins but several additional factors that associate with the HIR complex at histone gene promoters to regulate periodic transcription. These include the histone chaperones Asf1 and Rtt106, which establish a repressive chromatin structure at the histone gene loci, and the chromatin remodelers RSC and SWI/SNF (reviewed in ). While the former remodeler represses histone gene transcription in G2/M and G1, the latter acts as a transcriptional co-activator at the G1/S phase boundary. We showed that the recruitment of the SWI/SNF complex to a HIR-dependent promoter is not affected by the inactivation of the Clb cyclins (Fig. 7). More importantly, unlike the deletion of HIR1, deletion of ASF1 or RTT106 in the absence of functional Clb cyclins only partially relieved the defect in histone gene transcription, leading to an intermediate phenotype (Fig. 8). These results strongly suggest that in addition to the repressive activity of Asf1 and Rtt106, likely through their roles in nucleosome assembly, the HIR complex represses histone gene transcription outside of S phase by an independent mechanism that is alleviated by the mitotic Clb cyclins in late G1 to allow the full expression of HIR-dependent histone genes (Fig. 9). These results additionally suggest that the HIR complex itself, and not its associated factors, is the primary target of the mitotic Clb cyclins. Thus, the putative regulation of the HIR complex by Clb cyclin-dependent phosphorylation is unlikely to play a role in the association of these factors with histone gene promoters. Instead, the activity of some of these factors could be affected by HIR complex modification. We have previously shown that the HIR complex, when bound to a nucleosome template in vitro inhibits SWI/SNF chromatin remodeling activity . This inhibition could result from: (i) a direct inhibition of the SWI/SNF catalytic activity by the HIR complex; or (ii) the non-specific DNA binding of the HIR complex to nucleosomes, which generates a distinct protein/DNA complex that is resistant to remodeling by SWI/SNF. Thus, a potential mechanism for the mitotic Clb1/Clb2 cyclins and the putative Clb2-Cdk1-dependent phosphorylation of the HIR complex could be to alleviate this HIR-dependent inhibition of SWI/SNF chromatin remodeling activity. Future studies exploring these different hypotheses will shed light on understanding the specific mechanisms by which the mitotic Clb cyclins control histone gene transcription during the cell cycle, which is essential for cell viability, as well as further our understanding of the mechanism by which the HIR complex represses gene transcription.
The authors thank Amanda Raff for her excellent technical support for some of the experiments. This work was supported by grants NIH GM40118 to M.A.O. and NIH P20 RR016475 from the Kansas INBRE Program of the National Center for Research Resources to P.P. This work was also supported by a Leukemia and Lymphoma Society Special Fellowship to P.P., the Kansas Bioscience Authority (KBA), and the Kansas Medical Center Research Institute (KMCRI).
Publisher's Disclaimer: This is a PDF file of an unedited manuscript that has been accepted for publication. As a service to our customers we are providing this early version of the manuscript. The manuscript will undergo copyediting, typesetting, and review of the resulting proof before it is published in its final citable form. Please note that during the production process errors may be discovered which could affect the content, and all legal disclaimers that apply to the journal pertain.