The longevity and reversibility of quiescence in Schizosaccharomyces pombe are dependent upon the HIRA histone chaperone

Abstract

Quiescence (G0) is a reversible non-dividing state that facilitates cellular survival in adverse conditions. Here, we demonstrate that the HIRA histone chaperone complex is required for the reversibility and longevity of nitrogen starvation-induced quiescence in Schizosaccharomyces pombe. The HIRA protein, Hip1 is not required for entry into G0 or the induction of autophagy. Although hip1Δ cells retain metabolic activity in G0, they rapidly lose the ability to resume proliferation. After a short period in G0 (1 day), hip1Δ mutants can resume cell growth in response to the restoration of a nitrogen source but do not efficiently reenter the vegetative cell cycle. This correlates with a failure to induce the expression of MBF transcription factor-dependent genes that are critical for S phase. In addition, hip1Δ G0 cells rapidly progress to a senescent state in which they can no longer re-initiate growth following nitrogen source restoration. Analysis of a conditional hip1 allele is consistent with these findings and indicates that HIRA is required for efficient exit from quiescence and prevents an irreversible cell cycle arrest.

Keywords: G0; HIRA; MBF transcription factor; chromatin; histone chaperone; quiescence.

Figure 1.

HIRA is required for survival in G0. (a) the indicated strains were grown to mid logarithmic phase in EMM medium (Prolif) and then suspended in medium lacking a nitrogen source (EMM-N) at 30°C to induce quiescence (−N). At the indicated time points, an aliquot of the culture was subjected to five-fold serial dilution and then printed onto YES agar plates which were incubated at 30°C for 3–4 days to allow viable cells to form colonies. (b) flow cytometric analysis of proliferating cells and cells starved for nitrogen for 1 day. Data are representative of three independent biological repeats. A prominent 1C DNA peak is indicative of a G0 arrest. (c) morphology of proliferating cells and cells starved for nitrogen for 1 day (−N). Nitrogen-starved G0 cells adopt a small round morphology. (d) HIRA is not required for the induction of autophagy. Wild type and hip1Δ cells expressing GFP-Atg8 were grown to mid logarithmic phase in EMM (Prolif) and then resuspended in EMM-N for 1 day (G0). Whole cell protein extracts were prepared and analyzed by western blotting using anti-GFP and anti α-tubulin (TAT-1) antibodies. The presence of free GFP is indicative of the induction of autophagy and α-tubulin levels serve as a loading control. Data are representative of two biological repeats.

Figure 2.

Loss of HIRA results in a progressive loss of capacity to reenter the cell cycle. (a) wild type and hip1Δ cells were grown to mid logarithmic phase in EMM medium (Prolif) and then resuspended in EMM-N medium (−N) to induce quiescence. At the indicated time points, cell viability was measured by determining the ability of individual cells to form a colony. 81 cells were transferred to defined positions on a YES agar plate using a micromanipulator (Singer). Pates were then incubated at 30°C for 3–4 days to allow cells to form visible colonies. A representative example of the YES agar plates from one of three biological repeats is shown. (b) cells treated as described in (A) that failed to form colonies when transferred to a YES agar plate were designated as “inviable”. These cells were further classified based on their morphology using microscopy. Inviable cells that remained as small and round cells were designated “G0”, enlarged but undivided cells as “elongated” and cells that had divided at least once as “microcolonies”. Representative images of the different inviable cell types are shown. (c) Percentages of viable and inviable cell types after the indicated times in G0 (−N) were determined as described in (A) and (B). Data are based upon three biological repeats. Error bars are +SEM. (d) mean cell length (>100 cells) was determined after nitrogen source depletion for 1 day (−N) and following resuspension in rich (YES) medium for 8 h (+N 8 h). Data are the mean of three biological repeats. Error bars represent ± SEM. (e) flow cytometric analysis of G0 cells starved for nitrogen for 4 days (−N 4 days) followed by incubation in fresh YES medium for 20 h at 30°C (+N 20 h). (f) the majority of hip1Δ cells retain metabolic activity in G0. Wild type and hip1Δ cells were grown to mid logarithmic phase in EMM medium (Prolif) and then resuspended in EMM-N medium (−N) and incubated at 30°C for 6 days. At the indicated times cells were stained with Phloxine which is taken up passively but actively exported. As a result, metabolically active “live” cells are non-staining (Phloxine negative). Mean values of Phloxine negative cells from three biological repeats is shown and error bars represent ± SD. (g) ATP levels in nitrogen-starved wild type and hip1Δ G0 cells were determined using the BactTiter Glo cell viability assay (Promega). Luminescence values were corrected for background and cell number. Mean values were calculated from three biological replicates with each sample assayed in duplicate. Error bars indicate ± SEM.

Figure 3.

Loss of HIRA reduces survival in stationary phase. Cells from freshly grown YES agar plates were used to inoculate YES medium (50 mL). The resulting YES cultures (wt and hip1Δ) were then incubated at 30°C with shaking. At the indicated time points after inoculation (1, 3 and 6 days), percentage cell viability was determined by transferring individual cells to defined positions on a YES agar plate using a micromanipulator (Singer). Plates were then incubated at 30°C for 3–4 days to allow colonies to form. Percentage cell viability was calculated from the proportion of cells able to form a visible colony. Mean percentage viability was calculated from three biological repeats. Error bars represent ±SEM. (*p < 0.05, **p < 0.01, ***p < 0.001; t-test).

Figure 4.

HIRA is required for a reversible cell cycle arrest in G0. The hip1-HBD strain expresses Hip1 as a fusion with the hormone binding domain (HBD) of the estrogen receptor [23] which renders Hip1 (and thus HIRA) function dependent upon the presence of β-estradiol in the medium. The top panel shows the experimental scheme. Strains were grown to mid logarithmic phase in EMM (Prolif) and then resuspended in EMM-N at 30°C to induce quiescence (G0). At the indicated times cultures were subjected to five-fold serial dilution and pinned onto YES plates which were then incubated at 30°C for 3–4 days to allow cell proliferation to resume (exit). The absence (-) or presence (+) of β-estradiol (200 nM) in the medium at each stage in the experiment is indicated. Wild type and hip1Δ strains were included as controls.

Figure 5.

Chromatin and DNA damage in quiescent cells. (a) nitrogen-starved G0 cells (1 and 4 days –N) were treated with MNase to digest chromatin and the resulting DNA samples were analyzed on 1.5% TAE agarose gels. Data are representative of three biological replicates. (b) G0 cells lacking HIRA have reduced levels of total histone H3. Whole cell protein extracts, were prepared from wild type and hip1Δ G0 cells that had been starved for nitrogen for one day. Protein extracts were analyzed by western blotting using anti-histone H3 (Abcam) and anti-α-tubulin (TAT-1) antibodies. Examples of the primary data are shown (above) and a quantification of histone H3 levels normalized to α-tubulin and scaled relative to wild type (below). Data are the mean of four independent repeats. Error bars represent ±SEM (* p < 0.05; t-test). (c) the indicated strains were grown to mid logarithmic phase in EMM medium (Prolif) and then suspended in EMM-N medium at 30°C (−N) to induce quiescence. At the indicated time points, percentage viability was assayed by determining the ability of cells to form a colony (as described for Figure 2a-c). Data are the mean of at least three biological replicates and error bars represent ± SEM (* p < 0.05; t-test). Data for the percentage of viable cells after 1 and 4 days of nitrogen starvation for the wild type and hip1Δ strains is the same as that presented in fig 2C and is included for comparison. (d) the indicated strains carrying mutations in genes encoding histone chaperones were grown to mid logarithmic phase in EMM medium (Prolif) and then suspended in EMM-N medium for 4 days (−N 4 days). At the indicated times, cultures were subjected to five-fold serial dilution, printed onto YES agar and incubated at 30°C for 3–4 days to allow colonies to form. (e) DNA double strand breaks (DSBs) in quiescence and cell cycle reentry. Proliferating (Prolif) wild type and hip1Δ cells expressing Rad52-YFP were starved for nitrogen for 1 day (−N). Cells were then suspended in fresh YES medium and incubated for 4 h at 30°C (+N 4hrs). Fluorescence microscopy was used to determine the percentage of nuclei with Rad52-YFP foci which are markers of DSBs. At least 200 nuclei were counted for each strain at each point. Data are the mean of three biological repeats. Error bars represent ± SEM (** p < 0.01, *** p < 0.001; t-test).

Figure 6.

Impact of HIRA on gene expression during quiescence and reentry into the cell cycle. (a) Summary of GO biological process terms associated with genes with increased expression (log2 fold change > 1) in G0 hip1Δ cells (−N 1 day). FDR was calculated based on a p value from the hypergeometric test using ShinyGO 0.77. Fold enrichment represents the percentage of genes in the analyzed list belonging to a pathway, divided by the corresponding percentage in the background (all) genes set. (b) Overlap between genes with increased expression in G0 cells in the absence of HIRA (hip1Δ) and the Clr4 histone H3 lysine 9 methyltransferase [11] (p value, hypergeometric test). (c) Venn diagrams summarizing numbers of differentially expressed genes in the hip1Δ mutant in G0 (−N 1 day) and during G0 exit (+N 90 mins). (d) a summary of the GO terms associated with two gene clusters, termed UP1 and UP2, which are robustly induced during exit from G0 [8]. (e) Box plots comparing the transcript abundance (log2 FPKM) of the indicated gene clusters in wild type and hip1Δ cells in G0 (−N 1 day) and during G0 exit (+N 90 mins) cells. (f) wild type and hip1Δ cells were grown to mid logarithmic phase in EMM and then suspended in EMM-N medium to induce quiescence. After 1 day in EMM-N medium a nitrogen source was resupplied by suspending cells in rich (YES) medium. At the indicated time points after nitrogen source restoration, RNA was prepared and ura3⁺ transcript (UP1 gene) levels were analyzed by RT-qPCR. Levels are scaled relative to the wild type time 0 sample and normalized using sde2⁺ transcript levels which remain constant in proliferating and nitrogen starved quiescent cells [7]. Data are the mean of three biological repeats. Error bars represent ±SEM. (g) RT-qPCR analysis of mis3⁺ (UP2 gene) transcript levels as described in (f).

Figure 7.

HIRA is required for the induction of MBF-dependent genes during exit from quiescence. (a) the fold increase in mRNA levels of MBF-dependent genes in wild type and hip1Δ cells during exit from G0 (90 minutes after the restoration of a nitrogen source) was determined by RNA-seq. Induction of all genes except ctp1⁺ was significantly reduced in hip1Δ (p < 0.01). (b) HIRA is required for the induction of cdc18⁺ during cell cycle reentry. cdc18⁺ mRNA levels following the replenishment of a nitrogen source to G0 cells were determined using RT-qPCR as described in fig 6F. Data are the mean of three biological repeats. Error bars represent ± SEM. (*p < 0.05, **p < 0.01; t-test). (c) HIRA is required for the accumulation of the Cig2 cyclin during cell cycle reentry. Wild type and hip1Δ cells expressing HA-tagged Cig2 (Cig2-HA) were grown to mid logarithmic phase in EMM and then resuspended in EMM-N medium to induce quiescence. After 1 day in EMM-N medium a nitrogen source was restored by suspending cells in rich (YES) medium. At the indicated time points after nitrogen source restoration, aliquots of the culture were taken and whole protein extracts were prepared. Protein extracts were analyzed by western blotting with anti-HA and anti-α-tubulin antibodies. α-tubulin serves as a loading control. Data are representative of three biological repeats. (d) HIRA is required for the induction of rep2⁺ during reentry into the cell cycle. The abundance of rep2⁺ transcripts (FPKM) in wild type and hip1Δ G0 cells (−N for 1 day) and in cells undergoing exit from G0 (90 min after the restoration of a nitrogen source) was determined by RNA-seq. Error bars represent ± SEM. (***p < 0.001; t-test). (e) HIRA is not required for the removal of Rum1 during cell cycle reentry. Rum1-HA levels in wild type and hip1Δ following the restoration of a nitrogen source to G0 cells were determined by western blotting as described in (c). Data are representative of four biological repeats. (f) loss of Set2 does not result in a rapid loss in proliferative capacity in G0. The indicated strains (wild type, hip1Δ and set2Δ), were grown to mid logarithmic phase in EMM medium (Prolif) and then suspended in EMM-N medium for 4 days (−N 4 days). Cultures were subjected to five-fold serial dilution, printed onto YES agar and incubated at 30°C for 3–4 days to allow colonies to form.

Figure 8.

Model summarizing the roles of HIRA in (G0) quiescence. HIRA mediates efficient exit from quiescence. HIRA is required for the induction of the MBF transcription factor coactivator subunit Rep2 during exit from G0. As a result, induction of MBF-target genes such as cdc18⁺, cdt1⁺ and cdc22⁺, which are required for entry into S phase are dependent upon HIRA. In addition, HIRA prevents the premature onset of senescence. G0 cells lacking HIRA rapidly progress to a permanent cell cycle arrest in which they retain metabolic activity but can no longer re-initiate growth (elongate) and resume proliferation in response to restoration of a nitrogen source. Overall, the longevity and reversibility of quiescence are dependent upon HIRA.