The acquisition of cellular identity is coupled to changes in the nuclear periphery and nuclear pore complexes (NPCs). Whether and how these changes determine cell fate remain unclear. We have uncovered a mechanism that regulates NPC acetylation to direct cell fate after asymmetric division in budding yeast. The lysine deacetylase Hos3 associates specifically with daughter cell NPCs during mitosis to delay cell cycle entry (Start). Hos3-dependent deacetylation of nuclear basket and central channel nucleoporins establishes daughter-cell-specific nuclear accumulation of the transcriptional repressor Whi5 during anaphase and perinuclear silencing of the G1/S cyclin gene CLN2 in the following G1 phase. Hos3-dependent coordination of both events restrains Start in daughter, but not in mother, cells. We propose that deacetylation modulates transport-dependent and transport-independent functions of NPCs, leading to differential cell cycle progression in mother and daughter cells. Similar mechanisms might regulate NPC functions in specific cell types and/or cell cycle stages in multicellular organisms.
Publisher’s note: Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.
We thank M. Aldea, J. Skotheim and all members of the Mendoza lab for discussions and sharing reagents, and Y. Barral, L. Carey, B. Lehner and Life Science Editors for comments and manuscript editing. We are grateful to F. Cross (Rockefeller University), J. Clotet (Universtat Internacional de Catalunya), P. Carvalho, V. Malhotra (both CRG, Barcelona, Spain), S. Gasser (FMI, Basel, Switzerland), K.. Mekhail (University of Toronto) and A. Murray (Harvard University) for reagents, the CRG Advanced Light Microscopy Unit (T. Zimmermann and R. García), and Y. Schwab (EMBL, Heidelberg, Germany) for electron microscopy. This study was supported by the European Research Council (ERC) Starting Grant 2010-St-20091118 to M.M., the Spanish Ministry of Economy and Competitiveness, ‘Centro de Excelencia Severo Ochoa 2013–2017’, SEV-2012-0208 to the CRG and the grant ANR-10-LABX-0030-INRT, which is a French State fund managed by the Agence Nationale de la Recherche under the frame programme Investissements d’Avenir ANR-10-IDEX-0002-02 to the IGBMC. M.G.-A. is a recipient of a Postdoctoral Fellowship APOSTD/2017/094 from the Generalitat Valenciana.
Integrated supplementary information
Supplementary Figure 1 The duration of G1, but not T2 or G1 growth rate, is specifically affected in hos3∆ daughter cells.
(A) G1 duration (cytokinesis to budding) in wild type (n=110 cells) and hos3∆ (128 cells) mother and daughter cells pooled from three independent experiments. Lines represent the mean. Statistical significance was assessed using a two-sided Mann-Whitney test (* p < 0.05, exact value 0.0121). (B) Growth rate α during G1 and during S/G2/M phases (bud growth) and doubling time (between bud emergence events) of wild type and hos3∆ cells. The number of cells (sample size, n) is represented in brackets, pooled from two independent experiments. Statistical significance was assessed using a two-sided Student t-test (* p < 0.05, exact value 0.0372). (C) αT2 plotted as a function of cell size at the time of birth (cytokinesis), for mother and daughter cell pairs in WT (110 cells) and hos3∆ (130 cells). Note the lack of correlation between αT2 and cell size at birth. Growth rate and αT2 values were not significantly different between wild type and hos3∆ in either mother or daughter cells (ns, p > 0.05 with two-sided Mann-Whitney test); mothers, p=0.6234; daughters, p=0.7230.
(A) Asymmetric Hos3-GFP localization to daughter nuclei upon nuclear entry through bud neck and subsequent disappearance from both the bud neck and the nuclear periphery simultaneously after nuclear segregation. Htb2-mCherry was used as a chromatin marker. (B-C) Hos3-GFP disappearance from the bud neck relative to the time of actomyosin ring contraction (B) and septin ring splitting (C). Arrows indicate Hos3-GFP disappearance from the bud neck. (D) Fluorescence intensity of Hos3-GFP was quantified in cells going through anaphase and cytokinesis. n = 20 cells, (representative of one experiment) [Please revise if needed]. Error bars are mean +/- SEM. (E) The localization of a catalytically inactive form of Hos3 fused to GFP is indistinguishable from that of wild type Hos3. The arrow indicates Hos3 asymmetric localization to the daughter nuclear periphery. All experiments were repeated three times with similar results- Scale bars, 2 µm.
(A) Cells arrested in metaphase by depletion of Cdc20 display Hos3-GFP localization at the bud neck and at nuclear protrusions (arrows). Scale bars, 2 µm. (B) A dyn1∆ cell showing separated anaphase nuclei (Nup49-mCherry, in red) in the mother cell. Hos3-GFP strongly associates with the nuclear periphery only after nuclear migration across the bud neck (arrow). Note low levels of Hos3-GFP in the daughter nuclear periphery when the nucleus contacts the bud neck before nuclear migration into the bud. Images were acquired every 4 minutes. Time is indicated in minutes; t=0 marks the last frame before nuclear entry into the bud. Scale bars, 2 µm. (C) Hos3-GFP localization to the nuclear periphery (marked with Nup49-mCherry) in the daughter and mother cell compartment during anaphase. GFP and mCherry fluorescence intensity was measured across the nucleus of the dyn1∆ cell in (A) at two different times (yellow lines). Hos3-GFP and Nup49-mCherry profiles closely overlap when the nucleus is in the daughter but not in the mother cell compartment. Scale bars, 2 µm. (D) Dual-color image of Hos3-GFP and the septin Shs1-mCherry in the indicated strains at 30 ºC. Note that Hos3-GFP localizes to mother nuclei in the cdc12-1 mutant (arrow). (E) The intensity of Hos3 and the septin Shs1 in the bud neck were plotted along the indicated axis (yellow line in A). Notice the asymmetric distribution (towards the daughter side) is reduced in the cdc12-1 mutant. (F) Distribution of perinuclear Hos3 in the indicated strains at 30 ºC. WT (57 cells), cdc12-1 (40 cells) anaphase cells. (G) In the absence of Hsl7, Hos3-GFP fails to localize to the bud neck during mitosis. Instead it localizes to the daughter SPB (dSPB) during anaphase (arrows). SPBs are marked with Spc42-mCherry. (H) Removal of Hos3 from the bud neck requires mitotic exit, but its removal from the nuclear periphery does not. Hos3-GFP was imaged in cdc15-1 cells blocked in anaphase at 37 ºC. Hos3-GFP localizes to the periphery of the daughter nucleus and then disappears from this location, but it persists in the bud neck and daughter cell spindle pole (arrows). Microtubules are marked with Tub1-CFP. Cells were imaged every 3 minutes. Time is indicated in minutes; t=0 marks the last frame before nuclear entry into the bud. Scale bars, 2 µm. (I) Over-expression of Hos3-GFP leads to its accumulation in a dot-like structure in the daughter nuclear periphery (arrows). (J) Fluorescence intensity of Hos3-GFP in the indicated strains. Boxes include 50% of data points, the line represents the median and whiskers extend to maximum and minimum values. n = 20 cells pooled from three independent experiments. All experiments were repeated three times with similar results.
(A) Hos3-NLS-GFP localizes to the nuclear periphery in mother and daughter nuclei. The nuclear periphery is labeled with the ER protein Sec63 fused to mCherry. Scale bar, 2 µm. (B) Distribution of Hos3-NLS-GFP during anaphase. Note the short-lived enrichment of Hos3 at the bud neck (t=9 min). Images were acquired every 3 min. Time is indicated in minutes; t=0 marks the last frame before nuclear entry into the bud. Scale bar, 2 µm. (C) Growth defects upon over-expression of Hos3-NLS. Serial dilutions of the indicated strains were incubated in the indicated media for 2 days at 30 ºC. (D) Immuno-localization electron microscopy of Hos3-NLS-6HA showing its enrichment at the nucleoplasmic side of nuclear pores. Anti-HA conjugated gold particles (arrows) were found exclusively on the nucleoplasmic side of 40/50 nuclear pores with gold label examined. Outer nuclear membrane (ONM), inner nuclear membrane (INM) and nucleus (N) are labeled. Scale bars: 200 nm. (E-G) Localization of Hos3-NLS-GFP relative to NPCs (Nup60-mCherry), the nuclear periphery (Sec63-mCherry) and SPBs (Spc42-mCherry) in nup133∆ cells. Scale bars, 2 µm. (H) Localization of Hos3-NLS-GFP in the indicated pore basket mutants showing reduced perinuclear Hos3-NLS specifically in the absence of Nup60. Nup49-mCherry marks the nuclear pore and Htb2-mCherry marks the nucleus. Scale bars, 2 µm. (I) Scheme showing the relative distribution of protein complexes in the nuclear pore basket. (J) Nup60-myc was immunoprecipitated in extracts prepared from metaphase-arrested cells (Cdc20 depletion). Proteins were detected by western blot using the indicated antibodies. (K) Time-lapse showing recruitment of Hos3-FRB-GFP to nuclear pores (Nup49-FKBP-mCherry) upon addition of rapamycin. Whi5-GFP nuclear import and export can also be observed (arrows). Only the GFP channel is shown. Time is indicated in minutes; t=0 marks the time of nuclear division. All experiments were repeated three times with similar results except in panel D, which was repeated two times.
(A) Representative images of different protein localization at nuclear periphery in WT and HOS3-NLS cells. Sac3 and Kap123 showed reduced perinuclear localization, and Yku70 showed increased perinuclear localization in Hos3-NLS cells. Kap114 was enriched in the nucleoplasm in Hos3-NLS cells. Scale bars, 2 µm. (B) Whi5 nuclear fluorescence was measured in mother and daughter cells at the time of birth (71 cells) and its ratio plotted as a function of mother size (top left) or of the daughter/mother size ratio in wild type cells (bottom left), or in the indicated strains (right). For mother and daughter cell pairs in the right: WT (n=142 cells), hos3Δ (n=110 cells) HOS3-NLS (n=164 cells). Boxes include 50% of data points, the line represents the median and whiskers extend to maximum and minimum values. (C) Time-lapse images (sum projections of whole-cell Z-stacks) showing dynamics of Whi5-mCitrine, Kap95-GFP and Msn5-GFP in wild type and hos3∆ cells. Nup49-mCherry marks NPCs in cells expressing Kap95-GFP. Time is indicated in minutes; t=0 marks the time of cytokinesis (for Whi5-GFP; 6 minutes after Whi5 nuclear import) or nuclear division (for Kap95-GFP and Msn5-GFP). Scale bars, 2 µm. (D) Daughter/mother ratios of the concentration (left) and total amounts (right) of indicated proteins labeled with GFP in wild type and hos3∆ mutants, for the cells in figure 5C-E. Boxes include 50% of data points, the line represents the median and whiskers extend to maximum and minimum values. Cell numbers (n) for Whi5-GFP: WT =144, hos3Δ =110; Kap95-GFP (WT=144, hos3Δ=118); Msn5-GFP (WT=140, hos3Δ=104). (E) Asymmetric distribution of nuclear pores is not affected in hos3∆ cells. Ratio of total fluorescence measured by Nup49-mCherry signal in each mother and daughter cell pair WT (n=106 cells) and hos3∆ (n=124 cells) at the time of birth (after completion of nuclear division). Boxes include 50% of data points, the line represents the median and whiskers extend to maximum and minimum values. ns, p > 0.05 (p=0.5464) two-sided Mann-Whitney test. (F) Epistasis of HOS3 and WHI5 in Start. Correlation between αG1 and cell size at the time of birth (cytokinesis), for mother and daughter cell pairs in cells of the indicated strains: whi5∆ (178 cells) and whi5∆ hos3∆ (204 cells). WT and hos3∆ from Figure 2 are included for comparison. All experiments were repeated three times with similar results.
(A) Scheme showing crucial regulatory steps in the G1/S transition in budding yeast. The daughter-specific transcription factor Ace2 prolongs G1 duration by inhibiting Cln3 transcription. The proposed role of Hos3 is showed (dashed arrow). (B) The daughter-specific factors Ace2 and Ash1 delay G1 in daughters by repressing CLN3. As Ace2 is retained in daughter nuclei during G1, we asked whether its localization dynamics depended on Hos3. However, the nuclear residence time of Ace2-GFP was not affected in hos3∆ cells. (Left) Time-lapse microscopy of a hos3∆ cell during anaphase. Arrows mark Ace2-GFP nuclear import and export. Time is indicated in minutes. T=0 marks Ace2-GFP nuclear entry. Scale bars, 2 µm. (Right) Quantification of Ace2 nuclear residence time in WT (n=120 cells) and hos3∆ (n=60 cells). Boxes include 50% of data points, the line represents the median and whiskers extend to maximum and minimum values. ns, p > 0.05, two-sided Mann-Whitney test. (C-D) Advanced Start in hos3∆ cells is independent of the presence of CLN3, but required the presence of CLN2. Deletion of HOS3 advanced Start in ace2∆ ash1∆ cells, suggesting that Hos3 functions in a distinct pathway. The graphs show the correlation between αT1 and cell size at the time of birth (cytokinesis), for mother and daughter cell pairs in the indicated strains. (E) αT1 values (mean and SEM) in daughter cells of the strains analyzed in (C-D). Wild type and hos3∆ values from the experiments shown in Figure 2 are included for comparison. Mother-daughter cell pairs: cln3∆ (n=90), cln3∆ hos3∆ (n=84), ace2∆ (n=108), ace2∆ hos3∆ (n=98 cells), cln2∆ (n=136), cln2∆ hos3∆ (n=152) ash1∆ ace2∆ (n=93), ash1∆ ace2∆ hos3∆ (n=99), ash1∆ (n= 89), ash1∆ hos3∆ (n=129). Statistical significance in (E) was assessed using the Mann-Whitney test. (***, p < 0.001; ns, non-significant, p > 0.05). Exact p values relative to hos3∆: cln2 hos3 10-10, ace2 hos3 1.59x10-7, ace2 ash1 hos3 0.9x10-3.
Supplementary Figure 7 Hos3 inhibits CLN2 expression and promotes its perinuclear association during G1.
(A) GFP and phase contrast composite images of CLN2pr-GFP Myo1-mCherry cells. The graphs show the fluorescence intensity of CLN2pr-GFP in mother and daughter cells relative to cytokinesis (Myo1-mCherry disappearance; t=0) in wild type and hos3∆. Images were acquired at 3 min intervals. The GFP signal increases shortly after cytokinesis in WT and hos3∆ mother cells. GFP accumulates in the wild type daughter cell only after a delay; this delay is shorter in the hos3∆ daughter cell. Arrows indicates budding. This experiment was repeated three times with similar results. (B) Analysis of Rpd3 and Hos3 association with the CLN2 promoter during anaphase and G1. Quantification of Hos3-Myc, Rpd3-Myc binding to two different regions of the CLN2 promoter by chromatin immunoprecipitation (ChIP) during metaphase and anaphase (0, 30 min) and early G1 (60 min) using pGALL-CDC20 release. Untagged strain (without Myc-tag) used as negative control. (C) Gene positioning analysis in wild type and hos3 mutant cells. Distance to the nuclear periphery between various loci in the indicated strains and cell cycle stages, measured in 3D confocal nuclear stacks, used to generate the graphs shown in Fig. 6D. Lines represent the mean, n values are the number of cells. WT G1 (n=102), WT S (n=92), hos3Δ G1 (n=108), hos3Δ S (n=108), whi5Δ (G1) n=108, whi5Δ (S) n=82, WT M (n=56), WT D (n=56), hos3EN (M) n= 56, hos3EN (D) n= 56, nup60Δ (M) n= 47, nup60Δ (D) n= 47. (D) Localization of CLN2::lacO (green) near (<0.25 µm) clustered NPCs (Nup60-CFP, red) in nup133∆ G1 and S-phase cells (n=40 cells). (E) Subnuclear position determined by mapping their localization to one of three concentric nuclear zones of equal surface in confocal optical slices (left) and distance to the nuclear periphery (right) of subtelomeric loci in the indicated strains and cell cycle stages. WT n= 84 cells (TEL4R), hos3∆ n= 88 cells (TEL4R), WT n= 93 cells (TEL12R), hos3∆ n= 100 cells (TEL12R). Lines represent the mean. (F) (Left) Subnuclear position of CLN2 placed in ectopic loci. The location of CLN2 is indicated: chromosome and nearest locus. The CLN2 locus was visualized using the ParB system, and its position was determined by mapping their localization to one of three concentric nuclear zones of equal surface in confocal optical slices in G1 cells. N=40 nuclei / locus. (Right) Duration of pre-Start G1 phase (T1) was determined in the same strains, but expressing WHI5-GFP MYO1-GFP (> 100 cells per strain). Lines represent the mean. Number of cells: WT Chr16 CLN2 (n=112), Chr2 PGI (n=108), Chr10 INO1 (n=112), Chr11 BLI1 (n=106), Chr4 BTT1 (n=120). Data are from two (B, F) or three (A, C, D, E) independent experiments each with similar results.
(A) Viability of nucleoporin KN mutants. Indicated KN strains were grown in YPDA plates for 2 days at 30 ºC. (B) Protein localization of Nup60KN-mCherry in the indicated KN strains. (C) The indicated KN strains were grown in YPDA or YPGal plates (to induce Hos3-NLS) for 2 days at 30 ºC. Localization of Hos3-NLS-GFP is shown on the right. Experiments in (A-C) were repeated two times with similar results. (D) Daughter/mother ratios of Whi5-GFP concentration in wild type and the indicated mutant strains. n values denote the number of cells are n= 102 (nup57KN), 126 (nup60KN nup53KN), 136 (nup60KN nup49KN nup53KN), 158 (hos3EN). Boxes include 50% of data points, the line represents the median and whiskers extend to maximum and minimum values. A two-sided Mann-Whitney test was used to determine statistical significance relative to WT. p = 0.0017 (nup57 KN), 8.02x10-6 (nup60 KN nup53 KN), 2.47x10-5 (nup60 KN nup49 KN nup53 KN), 0.3x10-10 (hos3EN). ** represents p < 0.001, *** < 0.0001. (E) (Left) Daughter/mother ratios of Whi5-GFP concentration in wild type and nup60∆ cells. Boxes include 50% of data points, the line represents the median and whiskers extend to maximum and minimum values. N=142 (WT), 128 (nup60∆) cells. A two-sided Mann-Whitney test was used to assess statistical significance. With p>0.05 (p=0.0894) distributions were considered non-significant (ns). (Right) Subnuclear localization of the CLN2 locus in the indicated strains. n=112 cells (WT), n= 96 cells (nup60∆). (F) Correlation between αT1 and cell size at the time of birth (cytokinesis), for mother and daughter nup60∆ cell pairs (100 cells). Results from WT are reproduced from Figure 1 for comparison.
Boxes indicate the cropped sections used in the corresponding figures.
Supplementary Figures 1–9, and Supplementary Table and Supplementary Video legends
Standardized G1 and T1 duration.
GFP-fusion proteins included in the localization screen with Hos3–NLS.
Yeast strains and plasmids.
CRISPR guide sequences.
Time-lapse of HOS3–GFP cells (maximum projection).
Time-lapse of a HOS3–GFP HTB2– mCherry cell (maximum projection).
Time-lapse of a HOS3–GFP CDC3– mCherry cell (maximum projection).
Time-lapse of a dyn1∆ HOS3–GFP NUP49–mCherry cell (maximum projection).
Time-lapse of an ADH1pr–HOS3–GFP NUP49–mCherry cell (maximum projection).
Time-lapse of a nup133∆ HOS3–GFP NUP60–mCherry cell (maximum projection).
Time-lapse of a nup60∆ HOS3–GFP NUP49–mCherry cell (maximum projection).