Nuclear Division: Giving Daughters Their Fair Share

Abstract

How do nuclear components, apart from chromosomes, partition equally to daughter nuclei during mitosis? In Schizosaccharomyces japonicus, the conserved LEM-domain nuclear envelope protein Man1 ensures the formation of identical daughter nuclei by coupling nuclear pore complexes to the segregating chromosomes.

When we consider what constitutes a successful mitosis, we immediately think of the correct segregation of chromosomes into two daughter nuclei. However, it takes more than chromosomes to make a nucleus. The integrity of the daughter nuclei and the organization of the chromatin within them rely on the presence of an intact nuclear envelope (NE). The NE is a double lipid bilayer, with an outer membrane that is continuous withthe ER, and an inner nuclear membrane (INM) that contains proteins that interact with chromatin and other nuclear components. The NE is perforated by nuclear pore complexes (NPCs) that allow selective passage of proteins between the nucleoplasm and cytoplasm. In metazoans, a filamentous network, known as the nuclear lamina, underlies the INM. The NE breaks down at the onset of mitosis; membrane and membrane-associated proteins are absorbed into the ER while soluble proteins diffuse throughoutthe cytoplasm [1,2]. At the end of mitosis these NE components are retrieved in a poorly understood process to assemble the NEs of the two new daughter nuclei.

Yeast undergo mitosis in a somewhat different fashion. In most of the commonly studied yeast, the nuclear lamina is absent, the NE remains intact throughout the entire cell cycle, and the nucleus divides by a process of expansion followed by fission (called ‘closed mitosis’, in contrast to ‘open mitosis’ where the NE breaks down). The mechanisms ensuring that each daughter nucleus receives half of the nuclear membrane and its protein components (e.g., NPCs) are largely unknown. There are also examples of organisms where mitosis is neither fully open (as in metazoa), nor fully closed (as in budding or fission yeast). One such example is Schizosaccharomyces japonicus. In this particular yeast, the NE does not expand during mitosis, and instead it ruptures near the nucleus mid-section as the anaphase spindle elongates [3,4] (Figure 1A). The nucleolus does not divide but remains in the nucleus mid-section and then disassembles. Interestingly, the NPCs, which are uniformly distributed around the interphase nucleus, are absent from the mid-section of the anaphase nucleus (Figure 1A). Once anaphase is complete, the NE is resealed to form two daughter nuclei of equal size. The mechanism of this semi-open mitosis raises several interesting questions: how are NPCs cleared from the mid-section? Do they move through the NE or are they disassembled and new ones reassemble elsewhere? And how do cells ensure an even split ofthe nuclear membrane and other NE components? A study by Yam et al. [5] published in a recent issue of Current Biology reveals that the key to equal division of the nucleus in S. japonicus lies with the highly conserved LEM-domain protein Man1.

Figure 1. Anaphase in wild-type and man1Δ S. japonicus.

(A) In wild-type cells, the expansion of the NE (in green) is limited. Consequently, in early anaphase (top panel) the nucleus is shaped as a diamond (or prolate spheroid) and cannot elongate any further. At this point the sister chromatids (in blue) are segregated to the two poles and the NPCs (in red) are absent from the nucleus mid-section, which is occupied by the nucleolus (in yellow). The chromosome segments that extend into the nucleolus represent the DNA region coding for the ribosomal RNA. The nucleus can fully elongate in late anaphase thanks to the rupturing of the NE, which typically happens around the middle of the nucleus (bottom panel). The parental nucleolus is left in the nucleus mid-section and begins to disassemble. New nucleoli form in the daughter nuclei. (B) In a man1Δ cell, sister chromatids separate normally in early anaphase but the NPCs remain in the nucleus mid-section. In late anaphase the nucleus ruptures, as in wild-type cells, but the rupture site is randomly positioned, often away from the nuclear mid-section. The parental nucleolus fails to disassemble and segregates to one of the daughter nuclei.

NPCs of higher eukaryotes are largely immobile, likely due to their association with the nuclear lamina. The little movement that these NPCs do exhibit happens in conjunction with the underlying lamin network [6]. In budding yeast, which lack lamins, NPCs are much more mobile [7,8]. How, then, are NPCs cleared from the nuclear mid-section in S. japonicus? To distinguish between NPC movement and NPC disassembly followed by reassembly, Yam et al. [5] fused a photo-convertible fluorescent protein to an NPC subunit and examined the fate of NPCs at the center of the mitotic nucleus after they had been photo-converted. Not only did these NPCs move away from the mid-section and towards the nuclear poles, they did so in a manner that was coincident with the poleward movement of the chromosomes.

Motor-dependent NPC movement in yeast has been reported previously [9], but these movements were on a much smaller scale than mitotic NPC movement in S. japonicus in terms of the number of NPCs that moved coordinately and the distance that they travelled. Thus, the NPC movement observed in S. japonicus likely involved a different mechanism. Since NPCs moved along with chromosomes, the authors hypothesized that this movement may be mediated by one or more proteins that link chromosomes to NPCs. LEM-domain proteins appear to fit the bill: members of this protein family (named after its founding members LAP2, Emerin and Man1) localize to the INM and help to organize and regulate chromatin at the NE, thereby playing roles in transcriptional regulation, recombination and DNA replication ([10] and references therein). They perform these functions through a transmembrane domain that anchors them in the NE and a w40 amino-acid LEM domain that either binds DNA directly [11] or, in metazoans, binds to chromatin through the small DNA-binding protein BAF ([10] and references therein). Yam et al. found that one of the S. japonicus LEM-domain proteins, Man1, was required for several aspects of nuclear division: man1Δ cells failed to disassemble their nucleolus at mitosis, produced unequal sized daughter nuclei, formed an NE break in random locations and did not clear NPCs from the nucleus mid-section [5] (Figure 1B). Chromosome segregation, on the other hand, was largely unaffected.

These observations suggested that Man1 might mediate NPC movement by linking NPCs to chromosomes. But is the effect of Man1 on NPC movement direct, or is the NPC movement defect in the manΔ mutant an indirect consequence of its many other failures in mitotic processes? To address this, Yam et al. [5] designed an artificialtether that linked chromosomeswith NPCs independently of Man1. This tether was able to rescue the NPC movement defect of man1Δ cells, suggesting that Man1 physically couples chromosome segregation to the movement of NPCs away from the nuclear mid-section. The observation that LEM-domain proteins can perform ‘lamina-like’ functions in yeast is not a new one: the LEM domain proteins Heh1 and Heh2 have previously been shown to affect NPC distribution in S. cerevisiae [12], and LEM domain proteins are known to tether telomeres to the NE during interphase [13]. However, the Yam et al. study [5] reveals a novel function for a LEM-domain protein in coupling NPC inheritance to chromosome segregation.

Along with NPCs, nuclear membrane must also be evenly distributed among the daughters. In most organisms mitosis normally results in two equally sized daughter nuclei, regardless of whether they undergo open, semi-open or closed mitosis. Little is known about how the ‘right’ amount of membrane is used to reassemble the NE around the two sets of daughter chromosomes after open mitosis, although it is likely that components of the INM and nuclear lamina play an important role [14,15]. In the case of closed mitosis, the mechanism by which the nuclear fission point is determined remains unclear. In their recent study, Yam et al. [5] show that man1Δ cells form nuclei that are smaller and of unequal size. While the cause of this uneven NE distribution remains unknown, it is interesting to note that in wild-type cells the NE rupture site is equidistant between the two extreme ends ofthe nucleus (Figure 1), coinciding with the telomere of the last segregating chromosome. In man1Δ mutants, by contrast, the site of rupture is asymmetrically localized, suggesting that the Man1-dependent positioning of the rupture site dictates the amount of NE inherited by the two daughter nuclei. Interestingly, the artificial tether described above that rescued the NPC movement in man1Δ cells also corrected the NE rupture site. Thus, Man1 may affect the amountof NE inherited by the daughter nuclei by properly demarcating the rupture site through chromosome–NE attachments, although at thispoint other mechanisms cannot be excluded.

To date, many examples of how the NE can organize and regulate chromatin function have been identified. This recent study of the mitotic nucleus in S. japonicus provides an interesting example of how chromosome movement, via Man1, is being used to organize the NE in order to correctly partition NE components. A number of interesting questions arise from studies in S. japonicus. For example, why develop a mechanism to ensure the inheritance of NPCs from mother to daughters, when the cell can reassemble NPCs de novo? Perhaps inserting NPCs de novo is an energetically expensive process, or even an impossible feat if there aren’t enough NPCs, making the import of NPC components that must be inserted into the nucleoplasmic side of the NE too inefficient. Along these lines, a possible explanation for the unequal and reduced sizes of daughter nuclei in man1Δ cells is that a reduction in the number of NPCs present in the NE causes decreased import of a factor that determines nuclear size, as lamin B does in metazoa [16]. How Man1 affects nucleolar disassembly and the mechanism by which Man1 determines the site of nuclear rupture also remain unclear. Yam et al. suggest that the attachment of the last segregating chromosomes to the NE generates a force that aids rupture. However, it is also possible that the Man1-dependent clearance of NPCs from the centre of the nucleus weakens the NE, favoring rupture at this location. Although many questions remain to be answered, it is clear that LEM-domain proteins play a conserved role in regulating NE dynamics in variant forms of mitosis.