DNA in humans, yeast and other eukaryotic organisms is packaged in structures called chromosomes. When a cell divides these chromosomes are copied and then the matching pairs are separated so that each daughter cell has a full set of its genome. To enable these events to take place, the DNA must become more tightly packed so that the chromosomes become rigid units with projections called arms. Any failure in this chromosome “condensation” leads to the loss of chromosomes during cell division.

Within a chromosome, sections of DNA are wrapped around groups of proteins to make a series of linked units called nucleosomes, which resemble beads on a string. These units and other scaffold proteins together make a structure called chromatin and establish the overall shape of the chromosome. However, it is not exactly clear how the nucleosomes and scaffold proteins are rearranged during condensation.

Kruitwagen et al. used microscopy to study chromosome condensation in budding yeast. The experiments reveal that condensation involves two separate processes. First, modifications to the nucleosomes result in these units becoming more tightly packed in a process called short-range compaction. Second, a group of proteins called condensin is responsible for rearranging the compacted chromatin to enforce higher-order structure on the arms of the condensed chromosome (long-range contraction). Further experiments suggest that an enzyme called Hst2 may help to co-ordinate these processes to ensure that chromosomes adopt the right shape before the cell divides. For example, Hst2 ensures that longer chromosomes condense more than shorter ones.

A future challenge will be to find out whether chromosome condensation works in a similar way in humans and other large eukaryotes, which form much larger chromosomes with more complicated structures than yeast.

The DNA molecule at the core of any eukaryotic chromosome is a hundred to million times longer than the average diameter of the cell that hosts it. Thus, cells need to fold their genetic material in order to fit it in the interphase nucleus; they need to pack it further during mitosis, in order to move sister-chromatids safely and symmetrically apart. Furthermore, chromatin folding must be dynamic to allow transcription and replication during interphase, and such that exceptionally large chromosomes can hyper-condense during anaphase in order to fit the size of the spindle and prevent chromosome missegregation (Neurohr et al., 2011; Titos et al., 2014). Moreover, mitotic condensation also facilitates the decatenation of sister chromatids during their separation (Charbin et al., 2014), and might help to ‘cleanse’ chromosomes from transcription, replication and cohesion factors (Yanagida, 2009). This is thought to ‘reset’ the transcriptional state of genes, and prevent displaced factors from interfering with chromosome segregation. However, despite their importance for chromosome segregation, the events ensuring the mitotic condensation of chromosomes are still only partially understood.

Early studies made evident that nucleosomes play a critical role in DNA packaging. In favor of the idea that they play specific roles in chromatin condensation, histone H3 is phosphorylated by aurora B throughout mitosis on a serine at position 10 in most, if not all, eukaryotes. Furthermore, aurora B inactivation leads to chromosome condensation and segregation defects in budding yeast (Lavoie et al., 2004), fission yeast (Petrova et al., 2013; Tada et al., 2011), HeLa cells (Tada et al., 2011) and roundworms (Hagstrom et al., 2002). However, the precise role of H3 S10 phosphorylation has remained unclear (Ajiro and Nishimoto, 1985). Recent data demonstrated that H3 S10 phosphorylation promotes the recruitment of the sirtuin-related deacetylase Hst2, which in turn deacetylates, at least, lysine 16 of histone H4 (Wilkins et al., 2014). This unmasks a basic patch, allowing H4 to interact with the acidic patch on H2A, most probably on an adjacent nucleosome (Robinson et al., 2008; Gordon et al., 2005). Thus, this cascade of events initiated by H3 phosphorylation is thought to tighten the interaction between neighboring nucleosomes. However, all studies carried out so far have failed to reveal strong phenotypes for H3 serine 10 to alanine mutations in a plethora of model organisms (de la Barre et al., 2001; Afonso et al., 2014; Ditchfield et al., 2003). Furthermore, mutation of this residue in budding yeast did not affect axial contraction of chromosomes and the condensation of the rDNA during regular mitoses (Neurohr et al., 2011; Lavoie et al., 2004; Lavoie et al., 2002). Indeed, the only phenotype identified upon replacement of H3 S10 with alanine in yeast so far is limited to the reduced ability to hyper-condense artificially long chromosomes in order to fit them in the spindle (Neurohr et al., 2011). Thus, it remains unclear whether H3 phosphorylation and H4 deacetylation play any general role in mitotic chromosome condensation.

The discovery that mitotic extracts of frog eggs lacking any one of the subunits of a protein complex called condensin largely failed to condense chromosomes (Hirano and Mitchison, 1994) opened new perspectives for understanding chromosome condensation (Piazza et al., 2013; Thadani et al., 2012). Condensin is a ring-shaped pentameric protein complex. The core of the ring is formed by two structural maintenance of chromosome (SMC) subunits, Smc2 and Smc4. Three non-SMC proteins (Brn1, Ycg1 and Ycs4 in budding yeast) close the ring. The mechanism of condensin loading on chromatin is not understood, but seems to depend on the activity of the kinase aurora B (Lavoie et al., 2004; Tada et al., 2011). Furthermore, the non-SMC subunits were recently shown to directly bind DNA (Piazza et al., 2014), potentially followed by topological entrapment of chromatin inside the condensin ring (Cuylen et al., 2011). How condensin performs its functions in chromosome condensation is unclear, but it has been proposed that condensin’s role might be structural, by inducing loops within the same DNA strand (Cuylen et al., 2011; Cuylen and Haering, 2011) or might be enzymatic by promoting positive DNA supercoiling (Baxter and Aragón, 2012), both assisting in a decrease in length of mitotic chromatids.

Mitigating the central role of condensin in chromosome condensation, however, were observations in model organisms as diverse as fission yeast, fly, chicken and mammalian cells that indicate that chromosomes can still, at least partially, condense in the absence of condensin (Petrova et al., 2013; Coelho et al., 2003; Vagnarelli et al., 2006; Gerlich et al., 2006). Thus, although condensin was established as a key player in chromosome condensation, it cannot be the sole factor shaping mitotic chromosomes.

In order to gather insights into whether and how H2A-H4 interaction contributes to the organization of mitotic chromosomes, we sought for a method to assay the condensation state of chromatin in vivo. Here, we use a fluorescence-based assay to investigate short-range chromatin compaction and use it to study the relationships between condensin and histone modifications during chromosome condensation in mitotic cells.

Mitotic chromosome condensation is the process through which a relatively relaxed interphase chromosome condenses into two relatively short, compact sister chromatids that the spindle can symmetrically segregate. The kinase aurora B has long been implicated in this process (Lavoie et al., 2004; Tada et al., 2011; Lavoie et al., 2002). Through activation of condensin and phosphorylation of S10 on histone H3 it is thought to promote the formation of a mitotic chromosome. However, how these pathways interacted with each other to shape mitotic chromosomes was poorly understood.

Here, we monitored chromosome condensation in living yeast cells by using two microscopy assays, allowing us to distinguish long-range contraction of chromosomes along their longitudinal axis from short-range chromatin compaction. The first assay has been used before (Petrova et al., 2013; Guacci et al., 1994; Vas et al., 2007), but did not detect a role for histone phosphorylation during regular mitoses (Neurohr et al., 2011). The second assay, introduced here, uses fluorophore quenching to show that nucleosome-nucleosome interaction via H3 phosphorylation and in an Hst2-dependent manner promotes short-range compaction of mitotic chromatin in vivo. This enhanced packing is evidenced by the increased quenching of fluorophores when the chromosome locus is decorated by multiple copies of a fluorescent reporter. The quenching effect was observed similarly well whether we used GFP or mCherry as a fluorophore, and independently of whether it was fused to TetR or to LacI. Though we cannot exclude any local effects of the repetitive nature of the TetO and LacO repeats on chromatin compaction, this novel fluorescence-based assay allows for a simple detection of chromatin compaction states. Beyond promoting chromosome segregation, the exact function of this compaction process will be interesting to address. Our studies establish that chromatin compaction does not displace DNA-associated proteins such as TetR or LacI, as a strong cleansing model would predict. Although they do not exclude that other DNA-binding factors might be removed from the DNA during this process, these data suggest that compaction might serve other functions, such as to change the biophysical properties of the chromatin fiber during segregation. Chromatin compaction might for example affect the local stiffness, elasticity and mechanical resistance of chromosomes to facilitate their decatenation, protection and movement during anaphase.

Related to this, it is remarkable that the H3 S10D mutation is able to establish an Hst2-dependent and constitutive state of compaction and contraction, as evidenced by both assays, without affecting much growth and viability. We propose that this mutation strongly boosts the recruitment and activation of Hst2 all along chromosomes by mimicking the phosphorylated state of histone H3. Two models may explain how H3 S10D elicits its effects. In the first, H3 S10D might lead to hyper-activation of chromatin compaction by over-recruitment of Hst2, such that it might shorten the chromosome axis, independently of the contraction machinery. However, the following observations speak against this idea: I. crosslinking studies do not indicate that H3 S10D increases nucleosome-nucleosome interaction (Wilkins et al., 2014), II. fluorescence quenching of TetR-mCherry and LacI-GFP is not enhanced in the H3 S10D mutant (Figure 1B,C) compared with WT anaphase cells, and III. phosphorylation of H3 S10 does promote the condensin-dependent and Hst2-dependent contraction of artificially long chromosomes (Neurohr et al., 2011; Wilkins et al., 2014). Thus, these data support an alternative model: Hst2 may stimulate condensin function, and hyper-activation of Hst2 by H3 S10D may promote this effect and thereby chromosome arm contraction. To fully distinguish between these two models it will be important to investigate the effect of the H3 S10D mutation on condensin loading and activity. Furthermore, it is important to note that the phenotype of the H3 S10D mutant cells does not seem to reflect physiological conditions taking place during regular mitoses, indicating that the fraction of Hst2 driving wild type chromosome contraction does not depend on H3 S10 phosphorylation (dashed arrow Figure 7, left panel). Rather, it might reflect what happens when both chromosome condensation pathways are hyper-activated, for example under the presence of an artificially long chromosome or in small cells, such as cells grown on a poor carbon source (see below). In any case, it is striking that the H3 S10D strain does not show any obvious defects in growth and viability despite causing chromatin compaction and contraction throughout the cell cycle. Thus, this constitutively condensed state does not impair access and remodeling of chromatin by the transcription and replication machineries during interphase in any major manner.

Figure 7

Download asset Open asset

Our kinetic data suggest that chromatin undergoes a number of successive structural changes during anaphase and that compaction precedes contraction. In this respect, two models could explain the extended TRP1-LYS4 distance observed in early anaphase cells. First, this might simply be due to the fact that the anaphase nucleus is more elongated, allowing for the chromosome arm to reach its real extension without being confined in the normally spherical morphology of late anaphase or G1 nuclei. In this scenario, the apparent stretching of the chromosome in early anaphase does not correspond to any rearrangement of the chromatin or changes in contraction, and the impression that contraction follows compaction is an illusion. However, previous work has established that residual cohesion between sister chromatids persists during early anaphase and that this actually causes chromosome stretching during early segregation stages (Renshaw et al., 2010; Harrison et al., 2009). Thus, a second model would be that maximum contraction is only reached when cohesion is fully resolved, such as to not interfere with cohesion removal. This scenario suggests that chromosome contraction indeed needs to be a late anaphase event to complete chromosome segregation, and that perhaps the compaction events that precede play an important role in facilitating cohesion resolution. This model is attractive because it might more thoroughly account for the molecular events that are actually taking place during anaphase. Understanding how these processes are regulated during early anaphase will be necessary in order to distinguish between the two models. However, one definite conclusion that we can already draw is that arm contraction and chromatin compaction undergo relaxation between late anaphase and G1, since these nuclei are not significantly different in size, and hence the effects observed are not due to confinement alone.

Based on our data on chromosome contraction and compaction in Figure 3C, we propose that mitotic chromosome condensation entails at least three processes (Figure 7).

First, a histone 3-dependent and histone 4-dependent process, which we term chromatin compaction, ensures the short-range tightening of DNA into a smaller volume via nucleosome-nucleosome interaction (Wilkins et al., 2014). This process strictly depends on the phosphorylation of histone H3 on serine 10, and the subsequent deacetylation of lysine 16 of histone H4. Because this event has little impact on the length of condensed chromosomes, it has escaped attention until now. However, our data indicate that it is an important event for proper chromosome architecture and is an early event of every yeast mitosis.

Second, a condensin-dependent process, termed here axial chromosome contraction, ensures the long-range contraction of the chromosome, probably by facilitating the formation of higher order chromatin structures along the chromosome axis (Cuylen and Haering, 2011; Baxter and Aragón, 2012). This process is completed after chromatin compaction, in late anaphase, but does not strictly require it in order to proceed. Certainly, it will be interesting to clarify this order of event, to dissect how it is controlled, and to determine whether it facilitates the ordered condensation of well-separated chromosome arms. Moreover, the observation that Hst2 contributes to both processes suggests that this enzyme might be pivotal for coordinating compaction and contraction with each other. In this respect, it will be interesting to determine how Hst2 promotes the shortening of the chromosome axis, and whether it does so by directly modulating condensin function.

Third, our data also indicate the existence of an additional, also condensin-dependent process devoted to some other aspect of chromosome organization, beyond compaction and contraction. Indeed, restoring the contraction of the chromosome along its axis in the smc2-8 condensin mutant cells by mimicking constitutive H3 phosphorylation did not alleviate the temperature-dependent lethality caused by the smc2-8 mutation. Furthermore, although inactivation of the deacetylase Hst2 largely abolished the axial contraction of chromosomes, HST2 is not an essential gene. Hence, we suggest that the essential function of condensin is not to mediate long-range contraction. Rather, the essential function of condensin most likely relates to its roles in orchestrating the decatenation of intertwined sister-chromatids (Charbin et al., 2014; Baxter and Aragón, 2012; Baxter et al., 2011) and the re-annealing of single stranded DNAs (Sakai et al., 2003).

In an earlier study we demonstrated that yeast cells challenged by the presence of an exceptionally long chromosome or by being themselves exceptionally small (for example, whi3 mutant cells and cells growing on a poor carbon source) possess the ability to condense chromosomes beyond wild type levels to adjust the axial contraction of long chromosomes and adapt their length to the length of the spindle, a process we termed adaptive hyper-condensation (Neurohr et al., 2011), and which we would like to rename adaptive hyper-contraction. This process depended on condensin and phosphorylation of H3 S10, this last requirement being in contrast to what is observed for the contraction of most wild type chromosomes during regular mitoses. Accordingly, single smc2-8 and H3 S10A mutants failed to hyper-contract an exceptionally long chromosome (Neurohr et al., 2011). Since the H3 S10A mutation had no defect on the axial contraction of wild type chromosomes, these results indicated that adaptive hyper-contraction depended more heavily on Ipl1-dependent phosphorylation of S10 on histone H3 than regular chromosome condensation does. Thus, we propose that H3 S10 is hyper-phosphorylated on long chromosomes by Ipl1 located in the center of the spindle (midzone), boosting Hst2 and condensin activity (dotted arrow, Figure 7). Our observation that mimicking S10 phosphorylation, using the H3 S10D allele, promotes axial shortening in an Hst2-dependent manner is indeed consistent with high levels of S10 phosphorylation promoting the crosstalk between compaction and contraction, via Hst2.

It will be interesting to test how the principles laid out here function in eukaryotes with a more complex chromosome structure, such as humans. A good starting point for such investigations would be to test the effects of H3 S10 mutation and/or removal of the Hst2 homologues on such cells. How important is a tight regulation between short-range chromatin compaction and long-range chromosome contraction? How does that impact chromosome structure? And how do such mutations affect the segregation of the genetic material? Together, our work demonstrates that yeast is a powerful system to dissect the mechanisms underlying chromosome condensation and how the disruption of such mechanisms affects chromosome segregation and cell viability.

1. 1. Afonso O
   2. Matos I
   3. Pereira AJ
   4. Aguiar P
   5. Lampson MA
   6. Maiato H

   (2014) Feedback control of chromosome separation by a midzone aurora b gradient

   Science 345:332–336.

   https://doi.org/10.1126/science.1251121

   • Google Scholar
2. 1. Ajiro K
   2. Nishimoto T

   (1985)

   Specific site of histone H3 phosphorylation related to the maintenance of premature chromosome condensation. evidence for catalytically induced interchange of the subunits

   The Journal of Biological Chemistry 260:15379–15381.

   • Google Scholar
3. 1. de la Barre AE
   2. Angelov D
   3. Molla A
   4. Dimitrov S

   (2001) The n-terminus of histone H2B, but not that of histone H3 or its phosphorylation, is essential for chromosome condensation

   The EMBO Journal 20:6383–6393.

   https://doi.org/10.1093/emboj/20.22.6383

   • Google Scholar
4. 1. Baxter J
   2. Sen N
   3. Martinez VL
   4. De Carandini MEM
   5. Schvartzman JB
   6. Diffley JFX
   7. Aragon L

   (2011) Positive supercoiling of mitotic DNA drives decatenation by topoisomerase II in eukaryotes

   Science 331:1328–1332.

   https://doi.org/10.1126/science.1201538

   • Google Scholar
5. 1. Baxter J
   2. Aragón L

   (2012) A model for chromosome condensation based on the interplay between condensin and topoisomerase II

   Trends in Genetics 28:110–117.

   https://doi.org/10.1016/j.tig.2011.11.004

   • Google Scholar
6. 1. Buvelot S
   2. Tatsutani SY
   3. Vermaak D
   4. Biggins S

   (2003) The budding yeast Ipl1/Aurora protein kinase regulates mitotic spindle disassembly

   The Journal of Cell Biology 160:329–339.

   https://doi.org/10.1083/jcb.200209018

   • Google Scholar
7. 1. Charbin A
   2. Bouchoux C
   3. Uhlmann F

   (2014) Condensin aids sister chromatid decatenation by topoisomerase II

   Nucleic Acids Research 42:340–348.

   https://doi.org/10.1093/nar/gkt882

   • Google Scholar
8. 1. Coelho PA
   2. Queiroz-Machado J
   3. Sunkel CE

   (2003) Condensin-dependent localisation of topoisomerase II to an axial chromosomal structure is required for sister chromatid resolution during mitosis

   Journal of Cell Science 116:4763–4776.

   https://doi.org/10.1242/jcs.00799

   • Google Scholar
9. 1. Cuylen S
   2. Metz J
   3. Haering CH

   (2011) Condensin structures chromosomal DNA through topological links

   Nature Structural & Molecular Biology 18:894–901.

   https://doi.org/10.1038/nsmb.2087

   • Google Scholar
10. 1. Cuylen S
    2. Haering CH

    (2011) Deciphering condensin action during chromosome segregation

    Trends in Cell Biology 21:552–559.

    https://doi.org/10.1016/j.tcb.2011.06.003

    • Google Scholar
11. 1. D'Amours D
    2. Stegmeier F
    3. Amon A

    (2004) Cdc14 and condensin control the dissolution of cohesin-independent chromosome linkages at repeated DNA

    Cell 117:455–469.

    https://doi.org/10.1016/S0092-8674(04)00413-1

    • Google Scholar
12. 1. Ditchfield C
    2. Johnson VL
    3. Tighe A
    4. Ellston R
    5. Haworth C
    6. Johnson T
    7. Mortlock A
    8. Keen N
    9. Taylor SS

    (2003) Aurora b couples chromosome alignment with anaphase by targeting BubR1, Mad2, and cenp-e to kinetochores

    The Journal of Cell Biology 161:267–280.

    https://doi.org/10.1083/jcb.200208091

    • Google Scholar
13. 1. Gerlich D
    2. Hirota T
    3. Koch B
    4. Peters J-M
    5. Ellenberg J

    (2006) Condensin i stabilizes chromosomes mechanically through a dynamic interaction in live cells

    Current Biology 16:333–344.

    https://doi.org/10.1016/j.cub.2005.12.040

    • Google Scholar
14. 1. Gordon F
    2. Luger K
    3. Hansen JC

    (2005) The core histone n-terminal tail domains function independently and additively during salt-dependent oligomerization of nucleosomal arrays

    Journal of Biological Chemistry 280:33701–33706.

    https://doi.org/10.1074/jbc.M507048200

    • Google Scholar
15. 1. Guacci V
    2. Hogan E
    3. Koshland D

    (1994) Chromosome condensation and sister chromatid pairing in budding yeast

    The Journal of Cell Biology 125:517–530.

    https://doi.org/10.1083/jcb.125.3.517

    • Google Scholar
16. 1. Haase SB

    (2004) Cell cycle analysis of budding yeast using SYTOX green

    Curr Protoc Cytom Chapter 7, Unit 7:23.

    https://doi.org/10.1002/0471142956.cy0723s26

    • Google Scholar
17. 1. Hagstrom KA
    2. Holmes VF
    3. Cozzarelli NR
    4. Meyer BJ

    (2002) C. elegans condensin promotes mitotic chromosome architecture, centromere organization, and sister chromatid segregation during mitosis and meiosis

    Genes & Development 16:729–742.

    https://doi.org/10.1101/gad.968302

    • Google Scholar
18. 1. Harrison BD
    2. Hoang ML
    3. Bloom K

    (2009) Persistent mechanical linkage between sister chromatids throughout anaphase

    Chromosoma 118:633–645.

    https://doi.org/10.1007/s00412-009-0224-6

    • Google Scholar
19. 1. Hirano T
    2. Mitchison TJ

    (1994) A heterodimeric coiled-coil protein required for mitotic chromosome condensation in vitro

    Cell 79:449–458.

    https://doi.org/10.1016/0092-8674(94)90254-2

    • Google Scholar
20. 1. Janke C
    2. Magiera MM
    3. Rathfelder N
    4. Taxis C
    5. Reber S
    6. Maekawa H
    7. Moreno-Borchart A
    8. Doenges G
    9. Schwob E
    10. Schiebel E
    11. Knop M

    (2004) A versatile toolbox for PCR-based tagging of yeast genes: new fluorescent proteins, more markers and promoter substitution cassettes

    Yeast 21:947–962.

    https://doi.org/10.1002/yea.1142

    • Google Scholar
21. 1. Lakowicz JR

    (2013)

    Principles of Fluorescence Spectroscopy - Google Books

    Principles of Fluorescence Spectroscopy - Google Books.

    • Google Scholar
22. 1. Lavoie BD
    2. Hogan E
    3. Koshland D

    (2002) In vivo dissection of the chromosome condensation machinery: reversibility of condensation distinguishes contributions of condensin and cohesin

    The Journal of Cell Biology 156:805–815.

    https://doi.org/10.1083/jcb.200109056

    • Google Scholar
23. 1. Lavoie BD
    2. Hogan E
    3. Koshland D

    (2004) In vivo requirements for rDNA chromosome condensation reveal two cell-cycle-regulated pathways for mitotic chromosome folding

    Genes & Development 18:76–87.

    https://doi.org/10.1101/gad.1150404

    • Google Scholar
24. 1. Li P
    2. Jin H
    3. Hoang ML
    4. Yu H-G

    (2011) Tracking chromosome dynamics in live yeast cells: coordinated movement of rDNA homologs and anaphase disassembly of the nucleolus during meiosis

    Chromosome Research 19:1013–1026.

    https://doi.org/10.1007/s10577-011-9253-0

    • Google Scholar
25. 1. Neurohr G
    2. Naegeli A
    3. Titos I
    4. Theler D
    5. Greber B
    6. Diez J
    7. Gabaldon T
    8. Mendoza M
    9. Barral Y

    (2011) A midzone-based ruler adjusts chromosome compaction to anaphase spindle length

    Science 332:465–468.

    https://doi.org/10.1126/science.1201578

    • Google Scholar
26. 1. Petrova B
    2. Dehler S
    3. Kruitwagen T
    4. Heriche J-K
    5. Miura K
    6. Haering CH

    (2013) Quantitative analysis of chromosome condensation in fission yeast

    Molecular and Cellular Biology 33:984–998.

    https://doi.org/10.1128/MCB.01400-12

    • Google Scholar
27. 1. Piazza I
    2. Haering CH
    3. Rutkowska A

    (2013) Condensin: crafting the chromosome landscape

    Chromosoma 122:175–190.

    https://doi.org/10.1007/s00412-013-0405-1

    • Google Scholar
28. 1. Piazza I
    2. Rutkowska A
    3. Ori A
    4. Walczak M
    5. Metz J
    6. Pelechano V
    7. Beck M
    8. Haering CH

    (2014) Association of condensin with chromosomes depends on DNA binding by its HEAT-repeat subunits

    Nature Structural & Molecular Biology 21:560–568.

    https://doi.org/10.1038/nsmb.2831

    • Google Scholar
29. 1. Renshaw MJ
    2. Ward JJ
    3. Kanemaki M
    4. Natsume K
    5. Nédélec FJ
    6. Tanaka TU

    (2010) Condensins promote chromosome recoiling during early anaphase to complete sister chromatid separation

    Developmental Cell 19:232–244.

    https://doi.org/10.1016/j.devcel.2010.07.013

    • Google Scholar
30. 1. Robinson PJ
    2. An W
    3. Routh A
    4. Martino F
    5. Chapman L
    6. Roeder RG
    7. Rhodes D

    (2008) 30 nm chromatin fibre decompaction requires both H4-K16 acetylation and linker histone eviction

    Journal of Molecular Biology 381:816–825.

    https://doi.org/10.1016/j.jmb.2008.04.050

    • Google Scholar
31. 1. Sakai A
    2. Hizume K
    3. Sutani T
    4. Takeyasu K
    5. Yanagida M

    (2003) Condensin but not cohesin SMC heterodimer induces DNA reannealing through protein-protein assembly

    The EMBO Journal 22:2764–2775.

    https://doi.org/10.1093/emboj/cdg247

    • Google Scholar
32. 1. Sullivan M
    2. Higuchi T
    3. Katis VL
    4. Uhlmann F

    (2004) Cdc14 phosphatase induces rDNA condensation and resolves cohesin-independent cohesion during budding yeast anaphase

    Cell 117:471–482.

    https://doi.org/10.1016/S0092-8674(04)00415-5

    • Google Scholar
33. 1. Tada K
    2. Susumu H
    3. Sakuno T
    4. Watanabe Y

    (2011) Condensin association with histone H2A shapes mitotic chromosomes

    Nature 474:477–483.

    https://doi.org/10.1038/nature10179

    • Google Scholar
34. 1. Thadani R
    2. Uhlmann F
    3. Heeger S

    (2012) Condensin, chromatin crossbarring and chromosome condensation

    Current Biology 22:R1012–R1021.

    https://doi.org/10.1016/j.cub.2012.10.023

    • Google Scholar
35. 1. Titos I
    2. Ivanova T
    3. Mendoza M

    (2014) Chromosome length and perinuclear attachment constrain resolution of DNA intertwines

    The Journal of Cell Biology 206:719–733.

    https://doi.org/10.1083/jcb.201404039

    • Google Scholar
36. 1. Vagnarelli P
    2. Hudson DF
    3. Ribeiro SA
    4. Trinkle-Mulcahy L
    5. Spence JM
    6. Lai F
    7. Farr CJ
    8. Lamond AI
    9. Earnshaw WC

    (2006) Condensin and repo-man-PP1 co-operate in the regulation of chromosome architecture during mitosis

    Nature Cell Biology 8:1133–1142.

    https://doi.org/10.1038/ncb1475

    • Google Scholar
37. 1. Vas AC
    2. Andrews CA
    3. Kirkland Matesky K
    4. Clarke DJ

    (2007) In vivo analysis of chromosome condensation in saccharomyces cerevisiae

    Molecular Biology of the Cell 18:557–568.

    https://doi.org/10.1091/mbc.E06-05-0454

    • Google Scholar
38. 1. Wilkins BJ
    2. Rall NA
    3. Ostwal Y
    4. Kruitwagen T
    5. Hiragami-Hamada K
    6. Winkler M
    7. Barral Y
    8. Fischle W
    9. Neumann H

    (2014) A cascade of histone modifications induces chromatin condensation in mitosis

    Science 343:77–80.

    https://doi.org/10.1126/science.1244508

    • Google Scholar
39. 1. Yanagida M

    (2009) Clearing the way for mitosis: is cohesin a target?

    Nature Reviews Molecular Cell Biology 10:489–496.

    https://doi.org/10.1038/nrm2712

    • Google Scholar

Thank you for submitting your work entitled "Axial contraction and short-range compaction of chromatin synergistically promote mitotic chromosome condensation" for peer review at eLife. Your submission has been favorably evaluated by Jim Kadonaga (Senior editor) and three reviewers, one of whom is a member of our Board of Reviewing Editors.

The reviewers have discussed the reviews with one another and the Reviewing editor has drafted this decision to help you prepare a revised submission.

Summary:

Two of the three referees were enthusiastic about the imaging based approach to study chromosome compaction and axial contraction, while referee 3 was critical of some aspects of your experimental design. Referees 1 and 3 also raised questions about the statistical significance of some of your experiments, which are in points 1 and 4 of referee 3's detailed comments. Referees 2 and 3 have articulated important experimental controls and 1-2 additional experiments that will strengthen your study. The referee comments are attached verbatim.

Essential revisions:

In looking through the comments, it appears that there are only 4-5 points for which experiments are required (those in the comments of referee 2 and 3). These are all necessary to be performed to establish and strengthen your model fully. The super-resolution study that referee 1 suggests is not essential, but if you have STORM or PALM data, in particular, available you may add it to the revision. It is clear that SIM may not be a great improvement over your Deltavision acquired data with appropriate image analysis.

Reviewer #1:

This manuscript by Barral and colleagues addresses a fundamental question in chromosome biology, i.e. of the mechanism of chromosome condensation during mitosis. The authors use a clever microscopy based assay to investigate two different processes that contribute to condensation, i.e. of short-range compaction and axial contraction. Compaction, they show, clearly involves histone-histone interactions and is not perturbed in condensin mutants. By contract, axial contraction involves condensin (this was known previously), but they find a new role Hst2 deacetylase in modulating condensin function. They propose a pathway in which two synergistic arms contribute to chromosome condensation during mitosis.

While I find the approach very unique and the conclusions interesting, the statistics in many of the figures make me wonder how significant the various measurements are, especially of the intensities and distance measurements.

I support publication, but would like the following major issue addressed.

With tremendous advances in light microscopy, perhaps they can increase the resolution and intensity measurement/counting using either a SIM (increases resolution to 125 nm) or even better STORM/PALM, which further increases resolution.

While this may sound excessive, this sort of approach is becoming very routine and even getting one data set in wild-type cells for fluorescence intensity difference between G1 and anaphase and similarly the cell cycle dependent differences between LYS1 and TRP1 loci will go a long way to convince readers of this important mechanism arrived at using a novel strategy. I want to stress that the intention is not to make the authors repeat the entire study using super-resolution, but to at least set the stage so that we can be sure about the conclusions made using microscopy approaches that do not allow imaging in high resolution.

Reviewer #2:

Previous studies have shown that phosphorylation of H3S10 recruits the deacetylase Hst2, which deacetylates H4K16, and promotes the interaction between H2A and H4, a key mechanism in chromatin compaction. The authors have now developed an ingenious method to monitor chromosome compaction by monitoring fluorescence proteins localized to the TetO or LacO arrays, which show quenching when chromatins become compact. Using this tool, they show that condensin is dispensable for chromosome compaction, but indispensable for chromosome contraction. On the other hand, the pathway required for chromosome compaction is mostly indispensable for chromosome contraction. Thus, the authors propose that there are two mechanisms of mitotic chromosome condensation in budding yeast; one is the chromatin compaction mediated by nucleosome-nucleosome interaction, and the other is chromosome contraction mediated by condensin. Most experiments are well controlled and the conclusions are interesting. However, there are some aspects of the paper that weaken the overall story and that could be improved.

1) This manuscript has resolved important issues regarding chromosome condensation regulated by aurora B/Ipl1 in budding yeast. However, it is peculiar that the authors neither analyzed nor mention the contribution of Ipl1 to chromosome compaction/contraction. In fact, the authors previously mention that 'ipl1 inactivation did not abolish anaphase condensation in general (Science, 332, 465-468, 2011)'. How do they reconcile the model shown here (Figure 7) with the previous description? This issue should be addressed, probably by performing additional experiments to examine compaction and contraction under ipl1 inactive conditions.

2) This manuscript stresses the importance of H3S10 phosphorylation by aurora B, although the H3S10A mutant shows defects only in compaction. In the final model (Figure 7), the authors suggest that aurora B regulates chromosome contraction through condensin. In fact, condensin phosphorylation by aurora B and its requirement for anaphase condensation has been demonstrated in fission yeast and humans (Tada et al. Nature 2011). Also in budding yeast, Lavoie et al. (Gene Dev 2004) reported the importance of Ipl1 for anaphase condensation. These papers would be cited in the context of the authors' own results.

3) The previous report indicated that Hst2 recruitment to nucleosomes is dependent on H3S10p, and acts in the deacetylation of H4K16 and chromosome compaction. Importantly, the H3S10A mutation causes defects only in chromosome compaction, but not in contraction (Figure 1B and Figure 5A). Curiously, H3S10D (and recruited Hst2) not only sustains compaction, but also restores contraction defects caused by a condensin mutation. These facts cannot be reconciled with the authors' interpretation (Figure 7). I assume that although H3S10D might cause artificial overloading of Hst2, its loading is usually limited in wild type cells; H3S10p-independent (or residual) Hst2 may act in the condensin contraction pathway.

4) It is not clear what happens in H3S10D. What happens in H4∆9-16 H3S10D in Figure 1?

5) Please examine H4K16R and H4K16Rhst2∆ in Figure 1, 5 and 6. These experiments will clarify the effect of compaction (but not deacetylation) on contraction.

Reviewer #3:

The manuscript of Kruitwagen et al. exploits budding yeast to examine the contributions of short-range nucleosome compaction and long-range condensin-dependent chromosome contraction. By assessing changes in compaction and contraction in mutants defective in condensin and histone modification, the authors conclude that compaction and contraction happen at different times in the cell cycle and suggest potential different modes of regulation. The topic of chromosome condensation in mitosis is important, as chromosome structure in mitosis remains one of the major unanswered questions in biology. Much of the work here seems to corroborate a previous paper published by Wilkins et al. in Science, including the role of H3 and H4 modification in nucleosome compaction, and the role of H3 S10 and Hst2 in nucleosome compaction and chromosome contraction. In addition the manuscript has some important technical and experimental issues that need to be addressed.

1) The authors use an unusual assay to assess compaction, the reduction in fluorescence of TetR-mCherry and LacO-GFP bound to operator arrays between G1 and late anaphase cells. Many possible explanations exist for changes in fluorescence, one of which is compaction. Others include changes in chromatin composition or structure other than compaction that would be equally affected by histone modification. For example, the reduction in fluorescent signal in anaphase is not nearly as pronounced at the LYS4 locus as it is at the TRP1 locus. No statistical significance is given to the small drop in fluorescence at the LYS4 locus. This difference between loci opens up the possibility that the reduction in signal at TRP1 represents histone-dependent modification of a special feature of the pericentromere region and/or the kinetochore that is well documented by the Bloom lab. The key to believing these modifications represent global change in nucleosome compaction comes from the UV cross-linking experiment of H2-H4. But the recovery of this crosslink is very inefficient. This poor recovery may reflect inefficiencies of the method as the authors have suggested previously, or that the crosslinks are constrained to a small region of the genome, like the unusual pericentric structure.

2) The measurement of chromosome contraction using differentially marked loci on chromosome IV is also technically flawed. It is important to remember that yeast undergoes a closed mitosis such that the nuclear envelope constrains the ability of sequences to separate in space. In G1-metaphase the nucleus is a small ball, which prevents the chromosomes from extending. During anaphase, the nucleus elongates with the elongating spindle before returning to a small ball in late anaphase. As chromosomes are dragged during anaphase in all organisms, they lag behind the centromere and are stretched. In yeast with the elongated nucleus this stretching can manifest itself. When the nucleus shrinks the chromosomes become more confined again. This pattern fits the kinetics of spot separation in Figure 3B exactly. Thus it is unclear whether the kinetics of contraction is meaningful, that is the apparent additional contraction in anaphase from that already in anaphase is real. This worry challenges their conclusion that nucleosome compaction precedes contraction. It would also be important to know that whether the small changes in distances between spots reflects a difference in nuclei morphology due differences in the time of the cell cycle. The authors could include a marker for the nuclear envelope.

3) The authors state, "As expected from previous studies (Neurohr et al., 2011; Lavoie, B. D., Hogan, E., and Koshland), the mutations H3 S10A and H4 Δ9-16 did not impair anaphase contraction (Figure 5A and 5B), establishing that phosphorylation of H3 S10 and H4/H2A interactions are not necessary for axial contraction of chromosomes. In contrast, however, the hst2∆ mutation did abrogate the proper contraction of the chromosome during mitosis." Yet this same first author is an author on a previous publication (2014) that states: "Both HST2 inactivation and the H3 S10A mutation impaired this process to the same extent, either alone or in combination (Figure 4B)." These results appear contradictory.

4) The authors also conclude that yeast does not cleanse TetR or LacO based upon a failure to see an increase in background fluorescence. If a very small percentage of TetR-mCherry is associated with chromosomes, one is very unlikely to see a change in the background fluorescence if half of it is removed in anaphase. Without a measure of the fraction of the pool in the nucleus that has bound TetR-mCherry this experiment is meaningless.

[Editors' note: further revisions were requested prior to acceptance, as described below.]

Thank you for resubmitting your work entitled "Axial contraction and short-range compaction of chromatin synergistically promote mitotic chromosome condensation" for further consideration at eLife. Your revised article has been favorably evaluated by Randy Schekman (Senior editor) and three reviewers, one of whom, (Mohan Balasubramanian) is a member of our Board of Reviewing Editors. The manuscript has been improved but there are some remaining textual issues that need to be addressed before acceptance, as outlined below.

As you will see, two of the original three referees raise minor points to do with: 1) deleting some of the kinetic analyses of chromosome contraction (referee 3) and 2) making the model figure clear and also streamlining the manuscript to clarify the role of H3 S10 phosphorylation (referee 2).

These should not take you very long and we will be pleased to accept the further revised version.

Please see the referees' minor comments below.

Reviewer #2:

The authors appear to have addressed most of the points I raised in the initial round of review. I still have, however, one major issue.

The authors' interpretation concerning the H3S10D phenotype is now under question. I agree that H3S10D over-recruits Hst2. Although this mutant restores chromosome contraction in some assays, there is no defect in physiological chromosome contraction in the H3S10A mutant in which Hst2 is not recruited by H3 (Figure 5C and 6A). Therefore, it is most likely that H3S10D causes either 'over compaction', which by itself somewhat shortens the axis length, or artificial enhancement of the condensin-dependent contraction. The latter case is now properly represented in Figure 7 (left panel) by adding a new Hst2 out of H3S10p. This pool of Hst2 might be artificially supplied when H3S10D is expressed. Therefore, the current results do not provide any evidence to support the notion that 'H3S10 phosphorylation is involved in condensin regulation', at least under physiological conditions. This point is very confusing throughout the manuscript. I strongly recommend that the arrows in (H3S10p) Hst2 --> condensin --> contraction in Figure 7 (left panel) be removed, as well as the related content in the text. Instead, I agree that the simple model presented in Figure 7 (right panel) that clarifies the concept of compaction (by H3S10 phosphorylation) and contraction (by condensin), which act synergistically to form a properly condensed chromosome.

Reviewer #3:

The authors have significantly strengthened the manuscript, having addressed many of the important issues.

However they missed the point of the criticism about the analyzing the kinetics of chromosome contraction. The authors clearly have shown that they observe differences in contraction between late anaphase and G1 cells that are independent of nuclear morphology. However, they have not shown that the apparent longer distances between their arrays in early and mid-anaphase, which they assume to be less contraction, are not a consequence of changes in nuclear morphology. The chromosome stretching during chromosome movement may reveal the real distance between the arrays along the chromosome axis. When confined to a small round nucleus the distances between the arrays represent the shortest distance between arrays on bent chromosomes. In other words if one could make early and mid-anaphase cells that have small round nuclei like late anaphase and G1 cells, the distances between the arrays would shrink and be the same as late anaphase cells, all of which would be shorter than G1 cells. So contraction is greater in late anaphase than G1 as the authors contend but it is also equally greater in early anaphase and mid anaphase than G1. If so, contraction is happening at the onset of early anaphase concurrent with compaction and not later than it. I can't think of a way for the authors to address this nuclear morphology issue that really dramatically hinders any conclusion from this experimental strategy about contraction in early and mid-anaphase. However this kinetic analysis is really a side embellishment of the major points of the manuscript and as such deleting this section will not detract from the manuscript.

Reviewer #1:

This manuscript by Barral and colleagues addresses a fundamental question in chromosome biology, i.e. of the mechanism of chromosome condensation during mitosis. The authors use a clever microscopy based assay to investigate two different processes that contribute to condensation, i.e. of short-range compaction and axial contraction. Compaction, they show clearly involves histone-histone interactions and is not perturbed in condensin mutants. By contract, axial contraction involves condensin (this was known previously), but they find a new role Hst2 deacetylase in modulating condensin function. They propose a pathway in which two synergistic arms contribute to chromosome condensation during mitosis. While I find the approach very unique and the conclusions interesting, the statistics in many of the figures make me wonder how significant the various measurements are, especially of the intensities and distance measurements.

We agree with the reviewer that our data on chromatin compaction and contraction is of a highly variable nature. To investigate this variability, we followed individual dividing cells and measured the fluorescence intensity at the LacO/LacI-GFP locus (which was more resistant to photobleaching than the TetO/TetR-mCherry focus). The result of this experiment is now shown in Figure 3B of our revised manuscript. We discuss these data as follows in the Results section:

“Next, we investigated the dynamics of chromatin compaction during the cell cycle. To this end, we visualized both the TetO/TetR-mCherry (at TRP1) and LacO/LacI-GFP (at LYS4) loci simultaneously. […] In either case, this intrinsic cell-to-cell variability precludes drawing conclusions at the single cell level and emphasizes the fact that the quenching assay introduced here is statistical in nature.”

I support publication, but would like the following major issue addressed.

With tremendous advances in light microscopy, perhaps they can increase the resolution and intensity measurement/counting using either a SIM (increases resolution to 125 nm) or even better STORM/PALM, which further increases resolution. While this may sound excessive, this sort of approach is becoming very routine and even getting one data set in wild-type cells for fluorescence intensity difference between G1 and anaphase and similarly the cell cycle dependent differences between LYS1 and TRP1 loci will go a long way to convince readers of this important mechanism arrived at using a novel strategy. I want to stress that the intention is not to make the authors repeat the entire study using super-resolution, but to at least set the stage so that we can be sure about the conclusions made using microscopy approaches that do not allow imaging in high resolution.

Indeed, super resolution microscopy could be a way to confirm and gain more insight into the specific structures that underlie, in particular, chromatin compaction. The reviewer mentions three “standard” methods that can be used. For PALM, specific antibodies against, in our case, TetR- mCherry or LacI-GFP could be applied. Indeed, we have tried this in the past (unpublished) to visualize equivalent TetO-containing chromatin using PALM, but we encountered technical problems that would require significant optimization. Thus, this approach currently lies beyond the scope of this manuscript. STORM could be a great, equivalent, method, given that we construct the required photo-switchable or -activatable TetR fusion proteins. However, so far we have not been successful with using STORM in yeast cells either. Given the nature of this method, however, we predict that we would not see an increase in fluorophore-quenching in anaphase, since homo-fluorophores are not excited at the same time anymore. It can only be used as a method to measure the size of the LacO and TetO loci and test whether they indeed compact during anaphase. This might already be tested using SIM, which is currently available to us.

Thus, we fixed cells and used an in-house Applied Precision OMX Blaze to visualize LacO/LacI- GFP foci. Indeed, we observed that anaphase foci are dimmer and smaller than their G1 counterparts. However, upon careful analysis of the intensity profile, we observed that these still reflect a simple Gaussian distribution. Thus, the resolution provided by SIM is probably not sufficient to measure the size of the locus and compare between G1 and anaphase. Furthermore, the highly processive nature of the image reconstruction might result in artifacts when combined with quenching. Based on this preliminary experiment we conclude that the changes in fluorescence loci caused by chromatin compaction cannot be resolved by SIM, but rather require optimization of STORM/PALM. We are sorry that we cannot provide the analysis wished by the reviewer at this stage.

Reviewer #2:

Previous studies have shown that phosphorylation of H3S10 recruits the deacetylase Hst2, which deacetylates H4K16, and promotes the interaction between H2A and H4, a key mechanism in chromatin compaction. The authors have now developed an ingenious method to monitor chromosome compaction by monitoring fluorescence proteins localized to the TetO or LacO arrays, which show quenching when chromatins become compact. Using this tool, they show that condensin is dispensable for chromosome compaction, but indispensable for chromosome contraction. On the other hand, the pathway required for chromosome compaction is mostly indispensable for chromosome contraction. Thus, the authors propose that there are two mechanisms of mitotic chromosome condensation in budding yeast; one is the chromatin compaction mediated by nucleosome-nucleosome interaction, and the other is chromosome contraction mediated by condensin. Most experiments are well controlled and the conclusions are interesting. However, there are some aspects of the paper that weaken the overall story and that could be improved. 1) This manuscript has resolved important issues regarding chromosome condensation regulated by aurora B/Ipl1 in budding yeast. However, it is peculiar that the authors neither analyzed nor mention the contribution of Ipl1 to chromosome compaction/contraction. In fact, the authors previously mention that 'ipl1 inactivation did not abolish anaphase condensation in general (Science, 332, 465-468, 2011)'. How do they reconcile the model shown here (Figure 7) with the previous description? This issue should be addressed, probably by performing additional experiments to examine compaction and contraction under ipl1 inactive conditions.

In our previous work (Neurohr et al., 2011), Ipl1 was inactivated shortly before anaphase onset by switching cells containing the ipl1-321 temperature sensitive allele to the restrictive temperature accordingly. With this protocol, we observed a minor and actually insignificant defect in wild type chromosome contraction (whereas the contraction of the artificially long chromosome was significantly decreased). We now used the same strain, but inactivated Ipl1 throughout mitosis, i.e. shifted the cells to the restrictive temperature while they were still in G1. Strikingly, this resulted in a significant failure in both chromosome compaction and contraction (Figure 1D and 5A), as reported before (Lavoie et al., 2004).

These data indicate that for normal sized, wild type chromosomes, Ipl1 promotes condensation (compaction and condensation) prior to anaphase onset, probably during prophase and/or metaphase. This is in fact suggests that there are intermediate steps between Ipl1 function in early mitosis and the actual process of full condensation, which is most obvious during anaphase. Therefore, our data suggest that Ipl1 functions essentially in priming chromosomes for condensation. Subsequently, a second step downstream of Ipl1 is required to implement actual chromosome contraction and this second step can be temporally separated from the initial step (i.e. pro/metaphase and early anaphase). For example, the phosphorylation of condensin by Ipl1 may contribute to its loading or prime it for activation, whereas its phosphorylation by Cdk1 might delay its full activation for its role in chromosome contraction (Robellet et al., 2015). In such a scenario, full contraction would rely on Ipl1 activity in metaphase and activation of FEAR, activation of Cdc14 and condensin dephosphorylation on its Cdk1 sites during anaphase. In this scenario, the effects of mutating the Cdk1 sites in condensin might reflect consequences of premature activation rather than lack of activation.

2) This manuscript stresses the importance of H3S10 phosphorylation by aurora B, although the H3S10A mutant shows defects only in compaction. In the final model (Figure 7), the authors suggest that aurora B regulates chromosome contraction through condensin. In fact, condensin phosphorylation by aurora B and its requirement for anaphase condensation has been demonstrated in fission yeast and humans (Tada et al. Nature 2011). Also in budding yeast, Lavoie et al. (Gene Dev 2004) reported the importance of Ipl1 for anaphase condensation. These papers would be cited in the context of the authors' own results.

The reviewer is right. These citations are now incorporated in the Introduction and Discussion.

3) The previous report indicated that Hst2 recruitment to nucleosomes is dependent on H3S10p, and acts in the deacetylation of H4K16 and chromosome compaction. Importantly, the H3S10A mutation causes defects only in chromosome compaction, but not in contraction (Figure 1B and Figure 5A). Curiously, H3S10D (and recruited Hst2) not only sustains compaction, but also restores contraction defects caused by a condensin mutation. These facts cannot be reconciled with the authors' interpretation (Figure 7). I assume that although H3S10D might cause artificial overloading of Hst2, its loading is usually limited in wild type cells; H3S10p-independent (or residual) Hst2 may act in the condensin contraction pathway.

We fully agree with the reviewer’s interpretation of the data, and it is in fact what we aimed to suggest in Figure 7. While H3 S10 phosphorylation has no role in chromosome contraction in the majority of WT mitoses, H3 S10D mutant cells indeed show constitutively condensed chromosomes and this depends on Hst2 activity. H3 S10D in fact suppresses the effect of the smc2-8 mutation in condensin, through a yet unknown mechanism, which depends on Hst2. We indeed suggest that increased activation of Hst2 in the H3 S10D mutant cells hyper-activates the (crippled) condensin or acts in a still unidentified contraction pathway.

As the reviewer proposes, our data indicate that only a fraction of H3 S10 is phosphorylated in most WT mitoses, and some Hst2 must be indeed recruited to chromatin independently of H3 S10 phosphorylation to promote chromosome contraction. This was certainly not evident in our previous model figure, which we have amended to clarify this point.

Lastly, in the presence of an artificially long chromosome, such as in Neurohr et al. (2011), we propose that increased phosphorylation of H3 S10 by Ipl1 on the midzone boosts up the activation of Hst2 on that chromosome beyond regular levels, promoting its further contraction to fit it into the spindle. We have clarified the Results and Discussion sections regarding these issues.

4) It is not clear what happens in H3S10D. What happens in H4∆9-16 H3S10D in Figure 1?

See below.

5) Please examine H4K16R and H4K16R hst2∆ in Figure 1, 5 and 6. These experiments will clarify the effect of compaction (but not deacetylation) on contraction.

We thank the reviewer for raising these points, which together raise the issue of whether acetylated K16 of H4 is the main target of Hst2 upon H3 S10 phosphorylation, in which case H4∆9-16 should bypass the effect of H3 S10D and H4 K16R should bypass the compaction phenotype of the hst2∆ mutation but not its contraction phenotype. Due to the organization of the histone loci in budding yeast, constructing the H4∆9-16 H3 S10D is challenging and we have therefore addressed the issue by focusing on characterizing the H4 K16R hst2∆ double mutant.

From our original data, it was already clear that the H4 K16 acetylation alone does not affect chromosome contraction, which still takes place normally in the H4 ∆9-16 mutant cells.

Supporting this idea further, the H4 K16R hst2∆ double mutant does not contract its chromosome IV (Figure 5D), indicating that deacetylation of K16 on H4 is not sufficient to restore contraction in the hst2∆ mutant cells. Thus, we conclude that: 1) H4 deacetylation has no evident effects on chromosome contraction and 2) that Hst2, when deleted, does not have its defects on chromosome contraction through the chromatin compaction pathway, but rather by acting on another factor in the contraction pathway.

The picture is more complex for what concerns chromosome compaction. Indeed, unlike in the H4 K16R single mutant fluorescence intensity at the TetO/TetR-mCherry focus remained continuously high over the cell cycle in the H4 K16R hst2∆ double mutant cells, analogous to what is observed in the hst2∆ single mutant cells (Figure 1B). This implies that Hst2 promotes chromatin compaction not solely through deacetylation of K16 on H4 but also through additional downstream targets. The nature of these targets is not known at this stage but could include K8 and K12 on H4. Addressing the role of these residues and identifying possible additional targets of Hst2 will be the focus of further studies.

Reviewer #3:

The manuscript of Kruitwagen et al. exploits budding yeast to examine the contributions of short-range nucleosome compaction and long-range condensin-dependent chromosome contraction. By assessing changes in compaction and contraction in mutants defective in condensin and histone modification, the authors conclude that compaction and contraction happen at different times in the cell cycle and suggest potential different modes of regulation. The topic of chromosome condensation in mitosis is important, as chromosome structure in mitosis remains one of the major unanswered questions in biology. Much of the work here seems to corroborate a previous paper published by Wilkins et al. in Science, including the role of H3 and H4 modification in nucleosome compaction, and the role of H3 S10 and Hst2 in nucleosome compaction and chromosome contraction. In addition the manuscript has some important technical and experimental issues that need to be addressed. 1) The authors use an unusual assay to assess compaction, the reduction in fluorescence of TetR-mCherry and LacO-GFP bound to operator arrays between G1 and late anaphase cells. Many possible explanations exist for changes in fluorescence, one of which is compaction. Others include changes in chromatin composition or structure other than compaction that would be equally affected by histone modification. For example, the reduction in fluorescent signal in anaphase is not nearly as pronounced at the LYS4 locus as it is at the TRP1 locus. No statistical significance is given to the small drop in fluorescence at the LYS4 locus. This difference between loci opens up the possibility that the reduction in signal at TRP1 represents histone-dependent modification of a special feature of the pericentromere region and/or the kinetochore that is well documented by the Bloom lab. The key to believing these modifications represent global change in nucleosome compaction comes from the UV cross-linking experiment of H2-H4. But the recovery of this crosslink is very inefficient. This poor recovery may reflect inefficiencies of the method as the authors have suggested previously, or that the crosslinks are constrained to a small region of the genome, like the unusual pericentric structure.

We thank the reviewer very much for raising this important question. To address whether the TRP1 locus reflects a specialized chromatin region that is not representative for the whole chromosome, we analyzed the fluorescence intensity of LacO/LacI-GFP at the LYS4 locus in the middle of the chromosome arm more carefully. Though it might compact slightly less than TRP1, we still observe a significant difference in fluorescence intensity between G1 and anaphase cells at the LYS4 locus. Furthermore, analysis of the LacO/LacI-GFP brightness in the H3 S10A, H3 S10D, hst2∆ and H3 S10D hst2∆ single and double mutants confirms that H2A/H4 interactions are also required to compact this CEN4 distal locus. These data indicate that chromatin compaction is taking place over the whole chromosome. They are now added to Figure 1C where they extend the TetO/TetR-mCherry data from Figure 1B. However, since the anaphase specific quenching is not as strong as it is at the CEN4 proximal TRP1 locus, we do not exclude that the specialized chromatin domain around centromeres reaches a tighter compaction state during mitosis.

2) The measurement of chromosome contraction using differentially marked loci on chromosome IV is also technically flawed. It is important to remember that yeast undergoes a closed mitosis such that the nuclear envelope constrains the ability of sequences to separate in space. In G1-metaphase the nucleus is a small ball, which prevents the chromosomes from extending. During anaphase, the nucleus elongates with the elongating spindle before returning to a small ball in late anaphase. As chromosomes are dragged during anaphase in all organisms, they lag behind the centromere and are stretched. In yeast with the elongated nucleus this stretching can manifest itself. When the nucleus shrinks the chromosomes become more confined again. This pattern fits the kinetics of spot separation in Figure 3B exactly. Thus it is unclear whether the kinetics of contraction is meaningful, that is the apparent additional contraction in anaphase from that already in anaphase is real. This worry challenges their conclusion that nucleosome compaction precedes contraction. It would also be important to know that whether the small changes in distances between spots reflects a difference in nuclei morphology due differences in the time of the cell cycle. The authors could include a marker for the nuclear envelope.

To address the possible role of nuclear morphology in chromosome condensation, we measured the nuclear diameter in G1 and late anaphase nuclei in WT and hst2∆ conditions. We failed to observe any significant difference between G1 and anaphase in either strain. If anything, late anaphase nuclei tended to be a bit bigger than their G1 counterparts, rather opposing than facilitating chromosome condensation. Thus, we conclude that nuclear morphology during mitosis is not a major determinant of the distance measured between TRP1 and LYS4 in WT cells. Rather, these data support the conclusion that this distance reflects changes in axial chromosome contraction, as suggested by the numerous studies that have relied on this assay by now.

3) The authors state, "As expected from previous studies (Neurohr et al., 2011; Lavoie, B. D., Hogan, E., and Koshland), the mutations H3 S10A and H4 Δ9-16 did not impair anaphase contraction (Figure 5A and 5B), establishing that phosphorylation of H3 S10 and H4/H2A interactions are not necessary for axial contraction of chromosomes. In contrast, however, the hst2∆ mutation did abrogate the proper contraction of the chromosome during mitosis." Yet this same first author is an author on a previous publication (2014) that states: "Both HST2 inactivation and the H3 S10A mutation impaired this process to the same extent, either alone or in combination (Figure 4B)." These results appear contradictory.

It is important to note the differences between the wild type situation investigated in the presentstudy, where we have looked exclusively at regular chromosomes, and the study to which the reviewer is referring in which we characterized specifically the effect of Hst2 and H3S10 on the contraction of an artificially long chromosome, such as those constructed by Neurohr et al. (2011). Indeed, the 2014 study focused on the phenomenon of adaptive hypercondensation.

Therefore, the differences pointed out by the reviewer between the data obtained in the present study and those reported in the 2014 paper reflect how much hypercondensation is a phenomenon distinct from regular chromosome organization. Regular chromosomes do not depend on H3 S10Ph to contract, whereas the artificially long chromosome does. We showed that adaptive hypercondensation depends on the anaphase activity of the Ipl1 kinase, its midzone localization, H3 S10Ph and on Hst2 function (Neurohr et al., 2011; Wilkins et al., 2014), whereas regular condensation relies on metaphase activity of Ipl1, and Hst2 function. We propose that, in the case of a longer chromosome, more H3 S10 is phosphorylated by midzone Ipl1 in anaphase, which leads to an increased activation of Hst2. As explained in our response to reviewer 2 (point 3), we propose that increase Hst2 activity on chromatin leads to activation of its substrates in the contraction pathway, such as perhaps condensin. In any case, the long chromosome depends more heavily on H3 S10 phosphorylation, compared to WT chromosomes. To clarify these points, we added some explanations to the Discussion section regarding the difference between regular chromosome condensation and adaptive hypercondensation.

4) The authors also conclude that yeast does not cleanse TetR or LacO based upon a failure to see an increase in background fluorescence. If a very small percentage of TetR-mCherry is associated with chromosomes, one is very unlikely to see a change in the background fluorescence if half of it is removed in anaphase. Without a measure of the fraction of the pool in the nucleus that has bound TetR-mCherry this experiment is meaningless.

In order to address this issue, we determined that the fraction of bound TetR-mCherry in G1 cells is 0.19. Given this number, it would indeed be difficult to measure any significant change in background fluorescence in the case of cleansing with our current microscopy setup. We therefore thank the reviewer for this useful insight and have removed these data from Figure 2 and changed the text accordingly.

[Editors' note: further revisions were requested prior to acceptance, as described below.]

As you will see, two of the original three referees raise minor points to do with: 1) deleting some of the kinetic analyses of chromosome contraction (referee 3) and 2) making the model figure clear and also streamlining the manuscript to clarify the role of H3 S10 phosphorylation (referee 2).

These should not take you very long and we will be pleased to accept the further revised version. Please see the referees' minor comments below.

Reviewer #2:

The authors appear to have addressed most of the points I raised in the initial round of review. I still have, however, one major issue.

The authors' interpretation concerning the H3S10D phenotype is now under question. I agree that H3S10D over-recruits Hst2. Although this mutant restores chromosome contraction in some assays, there is no defect in physiological chromosome contraction in the H3S10A mutant in which Hst2 is not recruited by H3 (Figure 5C and 6A). Therefore, it is most likely that H3S10D causes either 'over compaction', which by itself somewhat shortens the axis length, or artificial enhancement of the condensin-dependent contraction. The latter case is now properly represented in Figure 7 (left panel) by adding a new Hst2 out of H3S10p. This pool of Hst2 might be artificially supplied when H3S10D is expressed. Therefore, the current results do not provide any evidence to support the notion that 'H3S10 phosphorylation is involved in condensin regulation', at least under physiological conditions. This point is very confusing throughout the manuscript. I strongly recommend that the arrows in (H3S10p) Hst2 --> condensin --> contraction in Figure 7 (left panel) be removed, as well as the related content in the text. Instead, I agree that the simple model presented in Figure 7 (right panel) that clarifies the concept of compaction (by H3S10 phosphorylation) and contraction (by condensin), which act synergistically to form a properly condensed chromosome.

The reviewer is right that there are two ways to explain the constitutive contraction phenotype of H3 S10D mutant cells and both depend on hyper-recruitment of Hst2. On the one hand, over-compaction by maximum nucleosome-nucleosome interaction could start to show its effects in contraction, independently of the contraction pathway. On the other hand, Hst2 might activate components of the chromosome contraction pathway (such as condensin). Our current and past experimental data speak against over-compaction since (1) TetO and LacO foci are not dimmer in H3 S10D strains as compared to WT, (2) there is not more crosslinking between H4 and H2A in an H3 S10D strain as compared to WT (Wilkins et al., 2014) and (3) H3 S10 phosphorylation, through Hst2, was shown to be required for adaptive hyper-condensation (Neurohr et al., 2011 and Wilkins et al., 2014). Since our data is supportive of Hst2 and condensin being in the same genetic pathway, we favor the idea that Hst2 might promote condensin activation through a yet unknown mechanism. However, it is crucial to realize, as the reviewer does, that the over-recruitment of Hst2 observed in H3 S10D strains is not relevant for the vast majority of yeast mitoses. Rather, this scenario becomes relevant in the case of smaller yeast nuclei and/or larger yeast chromosomes that require an additional burst of chromosome condensation. We included a new section to the Discussion to address these points.

Reviewer #3:

The authors have significantly strengthened the manuscript, having addressed many of the important issues. However they missed the point of the criticism about the analyzing the kinetics of chromosome contraction. The authors clearly have shown that they observe differences in contraction between late anaphase and G1 cells that are independent of nuclear morphology. However, they have not shown that the apparent longer distances between their arrays in early and mid-anaphase, which they assume to be less contraction, are not a consequence of changes in nuclear morphology. The chromosome stretching during chromosome movement may reveal the real distance between the arrays along the chromosome axis. When confined to a small round nucleus the distances between the arrays represent the shortest distance between arrays on bent chromosomes. In other words if one could make early and mid-anaphase cells that have small round nuclei like late anaphase and G1 cells, the distances between the arrays would shrink and be the same as late anaphase cells, all of which would be shorter than G1 cells. So contraction is greater in late anaphase than G1 as the authors contend but it is also equally greater in early anaphase and mid anaphase than G1. If so, contraction is happening at the onset of early anaphase concurrent with compaction and not later than it. I can't think of a way for the authors to address this nuclear morphology issue that really dramatically hinders any conclusion from this experimental strategy about contraction in early and mid-anaphase. However this kinetic analysis is really a side embellishment of the major points of the manuscript and as such deleting this section will not detract from the manuscript.

We apologize for not recognizing this issue in the first set of reviewers’ comments. The reviewer suggests that especially in early anaphase nuclear morphology might play a bigger role in chromosome condensation, since the chromosomes are less constrained in their movement due to a bigger nuclear volume, naturally causing the TetO-LacO distance to be high, though contraction is already present. Although we definitely consider nuclear morphology to play a role in chromosome extension during the early stages of anaphase, we do not think that this is the only factor contributing to it. Specifically, it has been shown that residual arm cohesion in the early stages of anaphase can cause chromosomes to stretch out during exactly this stage (Harrison et al., 2009 and Tanaka et al., 2010). We believe that this phenomenon is causing the very high TetO-LacO distance in early anaphase cells and would like to propose that initiating contraction too early (i.e. in early in stead of late anaphase) might lead to defects in chromosome segregation. However, to properly address this issue the activity of the compaction and contraction machinery in these cell cycle stages would have to be determined with an alternative method. We have discussed these models and their implications in the Discussion and thank the reviewer for these constructive comments.

Author details

1. Tom Kruitwagen

   Institute of Biochemistry, Department of Biology, Eidgenössische Technische Hochschule Zürich, Zürich, Switzerland

   Contribution

   TK, Acquisition of data, Analysis and interpretation of data, Drafting or revising the article

   Competing interests

   The authors declare that no competing interests exist.

2. Annina Denoth-Lippuner

   Institute of Biochemistry, Department of Biology, Eidgenössische Technische Hochschule Zürich, Zürich, Switzerland

   Contribution

   ADL, Analysis and interpretation of data, Drafting or revising the article

   Competing interests

   The authors declare that no competing interests exist.

3. Bryan J Wilkins

   Free Floater (Junior) Research Group "Applied Synthetic Biology," Institute for Microbiology and Genetics, Georg- August University Göttingen, Göttingen, Germany

   Contribution

   BJW, Acquisition of data, Analysis and interpretation of data

   Competing interests

   The authors declare that no competing interests exist.

4. Heinz Neumann

   Free Floater (Junior) Research Group "Applied Synthetic Biology," Institute for Microbiology and Genetics, Georg- August University Göttingen, Göttingen, Germany

   Contribution

   HN, Analysis and interpretation of data, Drafting or revising the article

   Competing interests

   The authors declare that no competing interests exist.

5. Yves Barral

   Institute of Biochemistry, Department of Biology, Eidgenössische Technische Hochschule Zürich, Zürich, Switzerland

   Contribution

   YB, Conception and design, Analysis and interpretation of data, Drafting or revising the article

   For correspondence

   yves.barral@bc.biol.ethz.ch

   Competing interests

   The authors declare that no competing interests exist.

• 3,122

  Page views

• 737

  Downloads

• 26

  Citations

Article citation count generated by polling the highest count across the following sources: Crossref, Scopus, PubMed Central.