Description

Drosophila NURF is an ISWI-containing chromatin remodeling complex that catalyzes ATP-dependent nucleosome sliding. By sliding nucleosomes, NURF can alter chromatin dynamics to control both transcription and genome organisation. NURF is composed of four subunits including the catalytic subunit Iswi and a large, highly conserved, NURF-specific scaffold subunit. In Drosophila this large subunit is Nurf301 (also known as Enhancer of bithorax (E(bx)) or CG32346, FlyBase ID:FBgn0000541) in humans bromodomain and PHD finger transcription factor (BPTF)) (Barak et al., 2003; Xiao et al., 2001). Nurf301/BPTF contains domains that have the potential to recognize specific histone modifications including a C-terminal bromodomain that recognizes histone H4 acetylated at lysine position 16 (H4K16Ac) (Kwon et al., 2009; Ruthenberg et al., 2011). H4K16Ac is enriched on the Drosophila male X chromosome as a result of the actions of the male-specific lethal (MSL) complex (reviewed in Samata and Akhtar, 2018) and in addition to being bound by Nurf301 can influence the activity of the Iswi subunit of NURF (Clapier et al., 2001; Corona et al. 2002).

Mutations in NURF subunits exhibit striking disruption of male polytene X chromosomes. This includes a characteristic decondensation or “bloating” of male polytene X chromosomes, showing that NURF functions to organise extensive regions of the male polytene X chromosome. Polytene chromosomes are the large chromosomes of the giant nuclei of salivary glands (reviewed in Zhimulev and Koryakov (2009)) that result from as many as ten rounds of endoreplication, where DNA is replicated without intervening mitosis (Hammond and Laird, 1985). The large number (1024C) of resultant sister chromatids and homologous chromosomes are held together generating single large chromosomes (Urata et al., 1995) that display features of interphase chromosome organisation including topologically associated domains (Eagen et al., 2015). Previous work has shown that the NURF mutant male polytene X chromosome phenotype is dependent on the presence of the MSL complex and that NURF functions on the male polytene X chromosome by antagonising the MSL complex (Badenhorst et al. 2002, Bai et al., 2007; Deuring et al. 2000).

Here we investigated whether this requirement is a general feature of all chromosome states. In particular we sought to establish whether NURF is also required for organisation of mitotic male X chromosomes. We examined morphology of mitotic chromosomes in neuroblasts of third instar Drosophila larvae from both control larvae (w¹¹¹⁸) and Nurf301 or Iswi mutant larvae. Mitotic chromosomes were prepared from both male and female larvae. As shown in Fig. 1A no differences in male X chromosome morphology relative to wild-type were detected in either male Nurf301 or Iswi mutant animals. As control the characteristic distorted polytene male X chromosome phenotype could be observed in salivary glands from the same mutant animals used for preparation of neuroblast squashes (inset labelled sg in Fig. 1A).

The lack of a mitotic male X chromosome requirement for NURF could be simply due to the dissociation of the MSL complex from mitotic chromosomes. It was shown previously that the bloated NURF mutant male polytene X chromosome is dependent on the presence of the MSL complex. MSL complex mutants or deletion of high affinity MSL complex binding sites are able to suppress the NURF mutant male polytene X chromosome phenotype (Bai et al., 2007; Corona et al., 2002). However, immunofluorescence microscopy of mitotic chromosomes using anti-Msl-2 antibodies detected continued presence of Msl-2 on mitotic chromosomes (Fig. 1B). Thus, even though components of the MSL complex remain associated with mitotic chromosomes, NURF is dispensable for correct mitotic male X chromosome morphology.

Further inspection of neuroblast squashes indicated no discernible defects in organisation of mitotic autosomes in male or female Nurf301 or Iswi mutants (Fig. 1A). In addition, although mammalian ISWI complexes have been implicated in cohesin loading and maintenance of chromatid cohesion (Hakimi et al., 2002; Zikmund et al., 2020), our analysis showed no discernible defects in sister chromatid cohesion in either male or female Iswi mutants (Fig. 1A). The lack of observable mitotic chromosome phenotypes in NURF mutants was interesting given previous analysis showing that Iswi remains associated with mitotic chromosomes (Deuring et al., 2000). However Iswi is the catalytic subunit of at least seven remodeling complexes (Oppikofer et al., 2017). To establish whether NURF remains associated with mitotic chromatin we therefore examined chromosome association of the NURF-specific Nurf301 subunit in larval neuroblasts.

Initial anti-Nurf301 antibody staining of neuroblast squashes prepared using standard methodologies failed to detect Nurf301 at any stage of the cell cycle. However, these protocols employed acetic acid treatment to assist chromosome spreading (Pimpinelli et al., 2010). We observed that acid treatment caused either the loss or masking of signal from Nurf301 and certain histone modifications and therefore modified chromosome preparation methods to omit acid treatment, deploying instead an acid-free fixation protocol developed for polytene chromosomes (Di Mario et al., 2006). Interphase nuclei and mitotic chromosomes from larval neuroblasts of control third instar Drosophila larvae were double-immunostained with antibodies against Nurf301 as well as antibodies against histone modifications that are enriched on mitotic chromosomes: anti-histone H3 phosphorylated at threonine position 3 (H3T3p) as well as the phosphomethyl mark (H3T3pK4me3). While we could readily detect Nurf301 in interphase neuroblast nuclei, Nurf301 signal was absent from mitotic chromosomes (Fig. 1C,D). This was not due to inability to stain chromatin proteins on mitotic chromosomes as both the H3T3pK4me3 (Fig. 1C) and the H3T3p mark (Fig. 1D) were detected.

These data suggest that NURF is ejected from mitotic chromatin in neuroblasts. To confirm whether this is generally observed in other tissues or organisms we performed immunostaining of Drosophila embryos and human HeLa cells using either anti-Nurf301 or anti-BPTF antibodies. Immunostaining of Drosophila embryos revealed Nurf301 was predominantly localised to chromatin during interphase (Fig. 1E). During metaphase Nurf301 was delocalized from chromosomes, re-associating with chromatin from anaphase. In a similar manner human BPTF was detected on interphase HeLa cell chromatin but was not localised to chromosomes during mitosis (Fig. 1F).

Cumulatively these data indicate that NURF is ejected from mitotic chromatin and is not required for proper mitotic chromosome condensation or cohesion. Moreover, although Nurf301 mutant male polytene X chromosomes are distorted and decondensed, Nurf301 mutant male mitotic chromosomes are unaffected. Strikingly, while targeting of the MSL complex to polytene chromosomes is sufficient to trigger aberrant chromosome morphology in Nurf301mutants (either normally on the male X chromosome or when targeted to autosomes using roX transgenes (Bai et al., 2007)), the MSL complex is unable to do so on when present on mitotic chromosomes. It is tempting to speculate that this reflects underlying differences in the ability of the MSL complex to promote transcription on polytene versus mitotic chromosomes. Even though recent data indicate that mitotic chromosomes are not transcriptionally silent as has previously been believed, levels of transcription are generally lower (Palozola et al., 2017).