• We are sorry, but NCBI web applications do not support your browser and may not function properly. More information
Logo of embojLink to Publisher's site
EMBO J. Jan 23, 2008; 27(2): 394–405.
Published online Jan 10, 2008. doi:  10.1038/sj.emboj.7601966
PMCID: PMC2234343

SAGA-mediated H2B deubiquitination controls the development of neuronal connectivity in the Drosophila visual system

Abstract

Nonstop, which has previously been shown to have homology to ubiquitin proteases, is required for proper termination of axons R1–R6 in the optic lobe of the developing Drosophila eye. Herein, we establish that Nonstop actually functions as an ubiquitin protease to control the levels of ubiquitinated histone H2B in flies. We further establish that Nonstop is the functional homolog of yeast Ubp8, and can substitute for Ubp8 function in yeast cells. In yeast, Ubp8 activity requires Sgf11. We show that in Drosophila, loss of Sgf11 function causes similar photoreceptor axon-targeting defects as loss of Nonstop. Ubp8 and Sgf11 are components of the yeast SAGA complex, suggesting that Nonstop function might be mediated through the Drosophila SAGA complex. Indeed, we find that Nonstop does associate with SAGA components in flies, and mutants in other SAGA subunits display nonstop phenotypes, indicating that SAGA complex is required for accurate axon guidance in the optic lobe. Candidate genes regulated by SAGA that may be required for correct axon targeting were identified by microarray analysis of gene expression in SAGA mutants.

Keywords: H2B deubiquitination, histone acetyltransferase, neural development, Nonstop, SAGA

Introduction

The compound eye of Drosophila melanogaster provides a powerful system to study the mechanisms regulating the target layer selection of neurons. There are eight different neuronal cells (R cells; R1–R8) within the ~800 ommatidia of the compound eye. Each of these neurons projects growth cones to distinct targets within the optic ganglia (Clandinin and Zipursky, 2002). R1–R6 cells project to the lamina, forming the lamina plexus. R7 and R8 axons continue through the lamina and terminate within the medulla. Screens for mutations affecting axon targeting identified Nonstop, a ubiquitin protease, as an essential protein required for correct axon targeting in the developing visual system of Drosophila (Martin et al, 1995). Mutations in nonstop affect the number of glial cells located within the lamina plexus. The lack of glial cells causes mistargeting of the R1–R6 axons (Poeck et al, 2001).

The function of Nonstop as an ubiquitin protease was thought to indicate a role for protein degradation in this axon targeting mechanism (Poeck et al, 2001). Many of the ubiquitin proteases that have been identified in flies have roles in protein degradation (Chen and Fischer, 2000). Polyubiquitination commonly targets proteins for degradation via the 26S proteasome. However, we find that the closest ortholog of Nonstop in yeast is the H2B deubiquitinase, Ubp8. Moreover, monoubiquitination of particular substrates such as histones is more generally involved in transcriptional regulation than in protein degradation (Osley, 2006). In particular, the C terminus of human histone H2B is subject to the dynamic addition and removal of a single ubiquitin moiety at Lys-120. A balance in the level of this modification is important for transcription activation and elongation. Monoubiquitination is catalyzed by the Rad6/UbcH6 E2 conjugase in conjunction with the Bre1 E3 ligase (Jentsch et al, 1987; Robzyk et al, 2000; Wood et al, 2003a; Zhu et al, 2005). Rad6 and Bre1 associate with elongating RNA polymerase II in a PAF-dependent manner, and this interaction is required for Lys-4 methylation on histone H3 (Wood et al, 2003b; Xiao et al, 2005).

The role of H2B deubiquitination in development of higher eukaryotes had not been characterized, and we were interested in whether the axon-targeting defect was due to the lack of removal of this histone modification. Hence we sought to determine whether Nonstop is a functional homolog of Ubp8 and acts as an H2B deubiquitinase in flies, and if it functions in axon guidance as a component of the Drosophila SAGA complex.

Results

Nonstop and CG13379 are orthologs of proteins required for H2B deubiquitination in yeast

To gain insights into the potential functions of Nonstop, we carried out a phylogenic analysis of the ubiquitin proteases from Saccharomyces cerevisiae and D. melanogaster and identified yeast Ubp8 as the ubiquitin protease most closely related to Nonstop (Figure 1B). Nonstop shares 20% identity with Ubp8, and has 31% similarity (Figure 1A). Furthermore, there is considerable conservation between Ubp8 and Nonstop within the zinc finger, and cysteine and histidine boxes. Ubp8 requires Sgf11 to deubiquitinate H2B in yeast (Henry et al, 2003; Ingvarsdottir et al, 2005; Lee et al, 2005). BLAST searches identified an Sgf11 ortholog, CG13379, in Drosophila, which has 36% similarity and 28% identity when aligned with yeast Sgf11 (Figure 1C). CG13379 will henceforth be referred to as Sgf11. Drosophila Sgf11 and yeast Sgf11 are similar to human ATX7L3 (Helmlinger et al, 2004), and to the yeast SAGA component Sgf73 (Figure 1D).

Figure 1
Orthologs of the proteins required for H2B deubiquitination in yeast were identified in Drosophila. (A) Alignment of D. melanogaster Nonstop and S. cerevisiae Ubp8. The conserved zinc finger, Cys-box and His-box domains are indicated. (B) Dendrogram of ...

Nonstop is functionally equivalent to Ubp8

Although phylogenetic analysis indicated that Nonstop is the potential ortholog of Ubp8, there are 18 additional ubiquitin proteases that have been identified in the Drosophila genome (Chen and Fischer, 2000). Therefore, it was necessary to determine whether Nonstop is functionally equivalent to Ubp8. To test if Nonstop could complement the function of Ubp8 within yeast SAGA, we generated yeast expression constructs containing galactose-inducible Nonstop. We asked whether Nonstop could rescue the phenotypes associated with ubp8Δ (Figure 1E). In yeast, the UBP8; GCN5 double deletion exhibits a growth defect on galactose, which can be rescued by the introduction of a plasmid-expressing yeast Ubp8 (Henry et al, 2003; Lee et al, 2005). Introduction of Nonstop into this double deletion strain also rescues this growth defect, indicating that Nonstop can functionally replace Ubp8 in yeast. Two other fly ubiquitin proteases, USP7 and CG5384, cannot rescue the growth defect of the UBP8; GCN5 double deletion on galactose, indicating that this rescue is specific to Nonstop (Supplementary Figure 1).

Nonstop and sgf11 mutants exhibit defects in axon targeting

nonstop was originally identified in a screen for mutations that affect photoreceptor connectivity in the Drosophila visual system (Martin et al, 1995). This leads us to hypothesize that Sgf11, and potentially H2B deubiquitination, may also be required for neural development in flies. To test this, we identified a mutation in sgf11 caused by the insertion of a piggyBac transposon in the promoter region of sgf11 (Figure 2A). This insertion is homozygous lethal during the late larval/early pupal stage and no transcripts for sgf11 were detected in mutant larvae by RT–PCR (Figure 2B).

Figure 2
Axon targeting is disrupted in the sgf11 mutant. (A) Schematic representation of the sgf11 (CG13379) locus, showing the position of the piggyBac E01308 transposon (insertion). The single exon is represented by a box. Translated sequences are filled with ...

In the developing visual system of Drosophila, the eye imaginal disc is connected to the optic ganglia by the optic stalk. There are eight different classes of neuronal cells, R1–R8, each of which project growth cones to distinct target regions in the optic lobe (Clandinin and Zipursky, 2002). The axons of R1–R6 terminate in the lamina, while R7 and R8 terminate within the medulla. The wild-type axon projection pattern can be visualized using a marker specific for those photoreceptors that terminate in the lamina (R2–R5; ro-τlacZ; Figure 2C). A triple layer of glial cells (anti-repo; Figure 2F) in the lamina specifies the termination point for R1–R6 axons. Mutations in nonstop cause a loss of glial cells from the target region of R1–R6 in the lamina, resulting in the misprojection of R1–R6 axons into the medulla (Figure 2E) (Poeck et al, 2001). We asked whether this targeting defect in the nonstop mutant reflected its potential role in H2B deubiquitination. To test this hypothesis, we visualized photoreceptor projections in sgf11 larvae (Figure 2D). As in nonstop larvae, many sgf11 photoreceptor axons fail to terminate in the lamina and instead project into the medulla. This axon-targeting defect in sgf11 is associated with a disruption in the organization of the lamina glial cells that is identical to that observed in nonstop by Poeck et al. nonstop glial cells fail to migrate from the dorsal and ventral glial precursor cell areas to form their characteristic layers along the lamina. Instead, an increased number of repo-positive glial cells are observed at the dorsal and ventral margins of the R-cell projection field in the nonstop mutant (Poeck et al, 2001). We observed a similar increase in the number of glial cells at these margins in sgf11 relative to the wild type (Figure 2F). This is accompanied by a decrease in the overall number of glial cells present along the lamina in the sgf11 mutant. Repo-positive glial cells were counted in nonstop, sgf11 and wild-type optic lobes along a particular length of the lamina to control for differences in the size and orientation of the R-cell projection field (Figure 2G and H). This shows that there is at least a two-fold decrease in the number of glial cells along the lamina in the sgf11 and nonstop mutants relative to the wild type. The similarity of the axon-targeting defect with regard to both the R-cell projection pattern and glial cell migration observed in the nonstop and sgf11 mutants indicated that these may function together for neural development in flies.

Nonstop associates with Sgf11

To determine if Nonstop and Sgf11 associated in vivo, we generated constructs to express C-terminally tagged versions of these proteins in S2 cells. These constructs express C-terminally V5-tagged Sgf11 and FLAG-HA-tagged Nonstop. We prepared whole-cell extracts from S2 cells transiently transfected with those constructs and immunoprecipitated Nonstop using anti-FLAG antibodies. Nonstop-HA2FL2 co-immunoprecipitated Sgf11-V5 (Figure 3A, lane 3). Nonstop was not detectable in the input material, when co-expressed with Sgf11-V5 because two-fold less DNA was used for this co-transfection relative to that of Nonstop alone (compare Figure 5B, lane 1). This result indicates that Nonstop and Sgf11 interact in vivo.

Figure 3
Nonstop and Sgf11 associate and are both required for H2B deubiquitination and regulation of gene expression. (A) Extract from S2 cells transfected with pRmHa3-Nonstop-HA2FL2 and pMT-Sgf11-V5 or untransfected S2 cells (−) was incubated with FLAG-M2-agarose ...

Nonstop and Sgf11 are required for deubiquitination of H2B in vivo

Ubp8 and Sgf11 deubiquitinate histone H2B at Lys-123 in yeast (Henry et al, 2003; Ingvarsdottir et al, 2005; Lee et al, 2005). We therefore asked whether Nonstop and Sgf11 are also required for H2B deubiquitination in flies (Lys-120; Supplementary Figure 2). Histones were acid-extracted from OregonR, nonstop or sgf11 third instar larvae and analyzed by western blotting using antibodies against histone H2B (Figure 3B). Histone preparations were normalized to the level of histone H3 (data not shown). A higher molecular weight band corresponding to monoubiquitinated H2B (ubH2B) is detected using an antibody against H2B in histone extracts from nonstop and sgf11. This band corresponding to ubH2B is also detected at lower levels in wild-type histone extracts (data not shown). The level of ubH2B was increased greater than seven- and four-fold in the nonstop and sgf11 mutants, respectively, in comparison to the wild-type control. These findings show that Nonstop and Sgf11 are required for deubiquitination of histone H2B in vivo.

Nonstop and Sgf11 coregulate a large subset of genes

The similar axon targeting and histone modification defects in the nonstop and sgf11 mutants indicated that these proteins may function together to regulate gene expression. To test this, we examined patterns of gene expression in nonstop and sgf11 mutant larvae using microarray analysis. Homozygous mutant third instar larvae of each genotype were compared to their heterozygote siblings or OregonR (wild type). There appears to be a large degree of overlap between the effects of the nonstop and sgf11 mutations, as evidenced by the spectrum of genes with greater than two-fold differences in transcript levels relative to the heterozygote (Figure 3C). This observation was confirmed mathematically by the calculation of Pearson correlation coefficients for the mutants relative to the heterozygotes. The nonstop and sgf11 mutants have a correlation of 0.68 with each other. Similar results are obtained for gene expression analyses from each pair of mutants relative to the wild type. The high correlation between the nonstop and sgf11 mutants is strikingly similar to results obtained in yeast (Ingvarsdottir et al, 2005). This expression analysis indicates that Nonstop and Sgf11 function together to deubiquitinate H2B and regulate transcription at a subset of genes.

Does loss of H2B deubiquitination affect global H3 acetylation?

In yeast, Ubp8 and Sgf11 function as part of the SAGA histone acetyltransferase (HAT) complex (Henry et al, 2003; Ingvarsdottir et al, 2005; Lee et al, 2005). Mutations in components of dSAGA, such as ada2b and wda, reduce global levels of acetylated histone H3 Lys-9/Lys-14 (Kusch et al, 2003; Pankotai et al, 2005; Guelman et al, 2006b). We therefore asked whether mutations affecting H2B deubiquitination in flies would affect global levels of acetylated histone H3. To address this question, histones were acid-extracted from OregonR, nonstop, sgf11 or ada2b third instar larvae and analyzed by western blotting using antibodies against histone H2B and acetylated H3 Lys-9 (Figure 4A). Histone preparations were normalized to the level of histone H3 (data not shown). The level of acetylated H3 Lys-9 is equivalent to that of the wild type in the nonstop and sgf11 mutant larvae, but is significantly decreased in the ada2b mutant larvae. These findings show that while Nonstop and Sgf11 are required for deubiquitination of histone H2B in vivo, they are not essential for the acetyltransferase activity of dSAGA.

Figure 4
Mutations in nonstop and sgf11 do not affect H3K9 acetylation and have distinct but overlapping effects on gene expression when compared to a dSAGA mutation. (A) Histones were acid-extracted from OregonR (WT), nonstop, sgf11 or ada2b third instar larvae ...

Mutations affecting dSAGA acetyltransferase activity have overlapping but distinct effects on gene expression when compared to nonstop and sgf11

Nonstop and Sgf11 are required for H2B deubiquitination, but have no effect on global acetylated H3 levels relative to mutations in other components of dSAGA. We therefore asked how these H2B deubiquitinating proteins function with regard to gene expression relative to Ada2b, which is required for dSAGA HAT activity (Figure 4A). Patterns of gene expression in nonstop, sgf11 and ada2b mutant larvae were examined using microarray analysis. Homozygous mutant third instar larvae of each genotype were compared to their heterozygote siblings or OregonR (wild type). Identification of genes showing increased or decreased transcript levels greater than two-fold in the mutants relative to the heterozygote indicates that there is only a small degree of overlap between the nonstop and sgf11 mutants compared to ada2b (Figure 4B). The nonstop and sgf11 mutants have correlations with ada2b of 0.03 and 0.07, respectively, relative to the heterozygote. Slightly higher correlations are obtained for gene expression analyses from each pair of mutants relative to the wild type (data not shown). The high correlation between the sgf11 and nonstop mutants, and their lack of similarity to ada2b is similar to deletion analysis in yeast (Ingvarsdottir et al, 2005). This expression analysis indicates that proteins required for H2B deubiquitination in flies have distinct regulatory effects on transcription that are separable from proteins required for HAT activity.

Nonstop and Sgf11 are components of dSAGA

Nonstop and Sgf11 have distinct effects on histone modifications and gene expression relative to other SAGA components, such as Ada2b. We therefore asked whether these proteins indeed constitute part of the deubiquitination module of the dSAGA complex in flies. Affinity purification of dSAGA enabled us to identify complex-specific subunits, such as dAda1, WDA and Ada2b (Kusch et al, 2003; Guelman et al, 2006b). Using mass spectroscopy analysis (using Multi-Dimensional Protein Identification Technology; MudPIT) of affinity-purified dSAGA, we identified Nonstop and Sgf11 as novel proteins that associate specifically with the complex. MudPIT results from the dADA1, WDA and Ada2b affinity purifications identified peptides specific for Nonstop and Sgf11 (Figure 5A). In Drosophila, dGcn5 (dKAT2) is a component of both the dSAGA and ATAC HAT complexes (Kusch et al, 2003; Muratoglu et al, 2003; Allis et al, 2007). No peptides for Nonstop or Sgf11 were identified in MudPIT analysis of affinity purifications of ATAC (Guelman et al, 2006a) (data not shown) or in mock purifications from untransfected S2 cell nuclear extract. These data indicate that Nonstop and Sgf11 associate specifically with dSAGA and are not present in additional dGcn5 (dKAT2)-containing complexes. Supporting the association of Nonstop and Sgf11 with dSAGA, purified dSAGA shows moderate deubiquitination activity on H2B in vitro (Supplementary Figure 3, lane 2).

Figure 5
Nonstop and Sgf11 are associated with dSAGA. (A) MudPIT analysis of dSAGA, affinity purified from cells expressing tagged dAda1, WDA or dAda2b identified Nonstop and CG13379/Sgf11 as putative components of dSAGA. The table shows the number of non-redundant ...

To confirm that Nonstop and Sgf11 associate with dSAGA, we generated constructs to express C-terminally tagged versions of these proteins in S2 cells. We prepared whole-cell extracts from S2 cells transiently transfected with these constructs and immunoprecipitated Nonstop using anti-FLAG antibodies. Nonstop-HA2FL2 co-immunoprecipitated dGcn5 (dKAT2) and dSpt3, both of which are stable components of dSAGA (Figure 5B, lane 3). Similarly, immunoprecipitation of Sgf11-HA2FL2 co-immunoprecipitated dGcn5 (dKAT2) and Ada2b (Figure 5C, lane 3). These results indicate that Nonstop and Sgf11 are both associated with dSAGA.

Since these observations provided evidence that Nonstop and Sgf11 stably associate with dSAGA, we wanted to determine if Sgf11 co-fractionated with dSAGA components by anion-exchange chromatography. First, nuclear extract from S2 cells stably expressing Sgf11-HA2FL2 was applied to Ni-agarose to enrich for dGcn5 (dKAT2)-containing complexes prior to anion-exchange chromatography. Both Sgf11 and dGcn5 (dKAT2) bind Ni-agarose (Figure 5D), indicating that the majority of Sgf11 is associated with this HAT-enriched fraction. This Ni-agarose HAT-enriched extract was applied to a Mono-Q column, and the elution profiles of dGcn5 (dKAT2), Ada2b, dSpt3 and Sgf11-HA2FL2 were compared by western blotting (Figure 5D). Sgf11 elutes within the same range of the salt gradient as other dSAGA subunits, indicating that the majority of Sgf11 is associated with the complex. The broader elution profile of dGcn5 (dKAT2) is consistent with its association with additional HAT complexes in flies.

Since this analysis indicated that the majority of Sgf11 co-fractionates and is thus associated with dSAGA, we then asked whether the same was true for Nonstop. Whole-cell extract was prepared from S2 cells transiently transfected with Nonstop-HA2FL2 and immunodepleted for dGcn5 (dKAT2) (Figure 5E). The level of Nonstop and dGcn5 (dKAT2) in the immunodepleted and mock-depleted extracts was analyzed by western blotting relative to the input material. A two-fold decrease in the level of dGcn5 (dKAT2) was observed with the highest amount of antibody added, corresponding to a similar decrease in the level of Nonstop. Nonstop was not completely depleted due to the overexpression of the tagged protein in these transiently transfected cells, which still contain endogenous Nonstop. This analysis indicates that the majority of Nonstop is associated with dGcn5 (dKAT2).

ada2b and gcn5 mutants exhibit defects in axon targeting in the optic lobe

Although our evidence showed that Nonstop and Sgf11 are components of dSAGA, the distinct effects on histone modifications and gene expression caused by mutations in nonstop and sgf11 relative to ada2b suggested that the axon-targeting defect observed in nonstop and sgf11 might be independent of dSAGA. If this axon-targeting defect is due to the role of Nonstop and Sgf11 within dSAGA, then mutations in other components of dSAGA should result in a similar phenotype. To test this hypothesis, we examined R-cell projection patterns in the ada2b and gcn5 mutants.

R-cell projections in ada2b larvae were visualized using a marker for all R cells (mAB24B10; Figure 6A and C) or a marker specific for R2–R5 (ro-τlacZ; Figure 6B, D and E). In the ada2b mutant, many axons project inappropriately through the lamina into the medulla (arrowheads; Figure 6C–E), as observed for sgf11 and nonstop. We asked whether this targeting defect is accompanied by a loss of glial cells from the lamina region, as it is in the nonstop and sgf11 mutants. Glial cells were counted as described previously over a given length of the lamina in ada2b and wild-type optic lobes (Figure 6F). A reduced number of glial cells are present along the lamina in ada2b relative to wild-type (compare Figure 6A and C), and slight increases in the number of glial cells at the margins of the lamina region (arrowheads; Figure 6E) are observed. Ada2b mutants exhibit some variability in the severity of the axon-targeting defect, and examples of a less severely (middle panel; Figure 6E) and more severely affected (lower panel; Figure 6E) ada2b optic lobe are shown for comparison.

Figure 6
Axon targeting is disrupted in the ada2b mutant. (AE) The R-cell projection pattern in eye–brain complexes from wild-type and ada2b larvae were examined using mAB24B10 (R1–R8, red; A, C) or ro-τlacZ (R2–R5; B, ...

Mutations in the catalytic HAT subunit of dSAGA, dGcn5 (dKAT2), result in a severe disruption to the R-cell projection pattern (ro-τlacZ; Supplementary Figure 4). The presence of dGcn5 in multiple protein complexes in Drosophila could explain the increase in the severity of this phenotype compared to ada2b, nonstop and sgf11. We observed that gcn5 eye discs are much smaller in size than those of wild-type or ada2b larvae (Supplementary Figure 5). This observation is consistent with the defects in cell proliferation and reduction in the size of other imaginal tissues, such as the wing disc in the gcn5 mutant (Carre et al, 2005). Preliminary observations indicate that photoreceptor organization may also be disrupted in gcn5 eye discs, and thus the defects observed in axon targeting may represent a secondary phenotype in the gcn5 mutant. Our results suggest that these particular eye disc defects observed in gcn5 may be independent of dSAGA, because there is no apparent reduction in eye disc size in the ada2b mutant.

The axon-targeting defect observed in sgf11 and ada2b is consistent with a role for dSAGA components within glial cells, as previously determined for nonstop (Poeck et al, 2001). To determine if dSAGA subunits associate within these cells, we asked if a FLAG-HA-tagged component of dSAGA, Ada2b, could co-immunoprecipitate other dSAGA subunits when expressed in glial cells of the optic lobe. To this end, we generated transgenic flies expressing C-terminally FLAG-HA-tagged Ada2b under the control of the UAS/GAL4 system and induced expression in glial cells, committed glial precursor cells, and in lamina neurons and their precursors using the gcm-GAL4 driver (Chotard et al, 2005). Whole-cell extract was prepared from partially dissected CNS/eye–antennal disc complexes from ~500 UAS-Ada2b-HAFL2/gcm-GAL4 wandering third instar larvae and incubated with HA-agarose. The immunoprecipitated material was analyzed by western blotting for the presence of dSAGA subunits, such as dGcn5 (dKAT2), dSpt3 and dSgf29 (Figure 6G). Tagged Ada2b associates with these dSAGA subunits in glial cells and lamina neurons, indicating that dSAGA components associate within this group of cells.

Taken together, these findings demonstrate that dSAGA is required for axon targeting in the developing visual system, and implicate dSAGA in the regulation of key pathways required for neural development.

dSAGA is required for expression of ecdysone-response genes, including broad

The finding that dSAGA is required for axon targeting in the developing visual system of Drosophila led us to ask which of its targets might be important for this particular role. Our examination of the transcriptional defects associated with mutations in the catalytic modules of dSAGA had revealed that among the genes coregulated by both activities, there are a large number involved in the ecdysone-regulated transcriptional program. For example, many of the genes involved in the ecdysone-stimulated secretion of glue proteins by the salivary gland (Li and White, 2003) are strongly repressed in the dSAGA mutants (Table I). This observation is consistent with the previous finding that there is a strong reduction in the size of the early ecdysone puffs in salivary gland polytene chromosomes from gcn5 mutants (Carre et al, 2005).

Table 1
Genes in the salivary gland glue protein battery (upper panel) and early ecdysone response genes (lower panel) showing decreased transcript levels in the nonstop, sgf11 and ada2b mutant genotypes compared to the wild type

One early-ecdysone response transcription factor, broad, is repressed in all three mutant genotypes (Table II). broad encodes a family of zinc finger isoforms that regulate a variety of developmental processes, including morphogenetic furrow progression and photoreceptor specification in the developing eye (Brennan et al, 2001). Studies on mushroom body neuronal remodeling in flies indicate that ecdysone is required to stimulate glial cell infiltration of neurons during axon pruning (Lee et al, 2000; Awasaki and Ito, 2004). Although loss of the EcR-B isoform alone does not appear to affect the R-cell projection pattern in larvae, mutations inactivating all EcR isoforms are embryonic lethal (Schubiger et al, 1998). A second candidate gene that might influence the differentiation or migration of glial cells within the optic ganglia is Takr86C (NKD/CG6515), which encodes a receptor for tachykinin similar to the mammalian NK1–3 receptors (Monnier et al, 1992; Johnson et al, 2003). Takr86C is also repressed in all three mutant genotypes (Table II). However, it is likely that dSAGA may function as a coactivator required by more than one genetic pathway affecting eye development. Furthermore, none of the mutations identified thus far in dSAGA are viable beyond the late larval/early pupal stage, indicating that dSAGA is essential for development.

Table 2
Genes coregulated in all three mutant genotypes

Discussion

In this study, we have identified a novel role for the coactivator complex dSAGA in Drosophila neural development. The Gcn5 (KAT2) HAT acts as the catalytic subunit of the yeast SAGA, SLIK, ADA and A2 multi-subunit protein complexes (Grant et al, 1997; Lee and Workman, 2007). Although many studies in multicellular organisms have focused on the HAT activity of the Gcn5 (KAT2) complexes, SAGA itself possesses a second catalytic activity. In addition to its HAT activity, yeast SAGA contains an H2B deubiquitinating enzyme, Ubp8 (Henry et al, 2003; Daniel et al, 2004). Ubp8 functions as part of a modular subunit domain within yeast SAGA that contains two additional proteins, Sgf11 and Sus1 (Ingvarsdottir et al, 2005; Lee et al, 2005; Kohler et al, 2006). Previous studies on histone deubiquitination have focused on its role in transcription in yeast (Henry et al, 2003; Lee et al, 2005). In this study, we have characterized a role for histone deubiquitination in gene regulation in Drosophila.

We demonstrate in this study that Nonstop and Sgf11 constitute the H2B deubiquitination module within dSAGA. Moreover, both Nonstop and Sgf11 are required for correct axon targeting in the developing visual system. As in yeast, the two catalytic functions of SAGA are separable (Ingvarsdottir et al, 2005; Lee et al, 2005). However, mutations that differentially affect the two catalytic activities of dSAGA have overlapping but distinct effects on gene expression. Despite the differences in activity, both catalytic modules of dSAGA are required for correct axon targeting in the optic lobe. This implicates dSAGA in the regulation of pathways essential for neural development in higher eukaryotes.

Our analysis of the axon-targeting defects in the nonstop, sgf11 and ada2b mutants indicates that the R-cell misprojection phenotype is associated in all three cases with a loss of glial cells from the lamina region of the optic lobe, and an increase in the number of glial cells at the dorsal and ventral margins of the lamina. Clonal analysis indicates that nonstop glial cells fail to migrate from these dorsal and ventral regions into the lamina plexus (Poeck et al, 2001). This suggests that dSAGA may be required within glial cells to regulate pathways important for their migration. It is possible that the requirement of dSAGA may be due to its role in the ecdysone response as indicated by our microarray analysis of genes downregulated in nonstop, sgf11 and ada2b mutant larvae. However, further studies on the role of dSAGA in these glial cells will be required to determine which pathway(s) are primarily responsible for the axon-targeting defect observed in these dSAGA mutants.

Mutations that affect the deubiquitination activity of dSAGA, such as nonstop and sgf11, appear to result in a more severe axon-targeting defect than those that affect the acetylation activity, such as ada2b. However, there is some variability in the degree of expressivity of the phenotype in all three mutant genotypes, and some ada2b mutants show phenotypes very similar to nonstop and sgf11. It is evident from studies on yeast SAGA that both enzymatic activities are required for optimal transcription upon gene induction (Henry et al, 2003). The variability in the ada2b mutant may be due to the large maternal contribution of Ada2b (Kusch et al, 2003), and it is notable that the ada2b mutant larvae show considerably less developmental delay in comparison to nonstop, sgf11 and gcn5. Thus far, mutations in other components of dSAGA, such as wda and nipped-A, have not been examined for this phenotype because these do not reach the third instar larval stage of development (Gause et al, 2006; Guelman et al, 2006b).

Although in this study we observed a specific effect of the dSAGA deubiquitination module on histone H2B, it remains likely Nonstop and Sgf11 are also required for deubiquitination of other target proteins within the cell. Poeck et al (2001) observed an increase in the level of three ubiquitinated proteins of 29, 55 and 200 kDa in extracts from nonstop third instar larval tissue relative to wild-type extracts. The smallest of these may correspond to ubiquitinated H2B, but the others may correspond to as yet unidentified potential targets of the deubiquitination module within dSAGA.

It is likely that the role of SAGA and H2B deubiquitination in neural development may be conserved in mammalian systems, as there is a striking degree of similarity between Sgf11 and ATXN7L3/ATXN7 in humans. ATXN7 is a subunit of the human STAGA and TFTC complexes (Helmlinger et al, 2004). Polyglutamine expansions in the Spinocerebellar ataxia type 7 (sca7) gene, encoding ATXN7, result in a dominant neurodegenerative disorder that affects the retina (David et al, 1997). There is increased recruitment of STAGA/TFTC containing this polyQ-expanded ATXN7 at certain promoters in a mouse SCA7 model, resulting in hyperacetylation of H3 (Helmlinger et al, 2006). Interestingly, despite this increased promoter recruitment and hyperacetylation, these genes show decreased levels of transcription. However, in other studies incorporation of polyQ-expanded ATXN7 into STAGA reduces the acetyltransferase activity of the complex (Palhan et al, 2005). The effect of the polyQ expansions in ATXN7 on the putative deubiquitination activity of mammalian STAGA/TFTC has not yet been examined.

The Verrijzer group has demonstrated that another ubiquitin protease in flies, USP7, specifically deubiquitinates histone H2B in vitro (van der Knaap et al, 2005). However, RNAi of USP7 had little effect on global levels of ubiquitinated H2B, and we find that USP7 cannot functionally replace UBP8 in yeast. Yeast Ubp15, the putative homolog of USP7, also deubiquitinates H2B in vitro, but the biological role of this activity is presently unknown (Supplementary Figure 3). In addition to Ubp8 in yeast, Ubp10 is also required for deubiquitination of H2B at the telomeres and at the rDNA locus (Emre et al, 2005). It is unknown whether the putative homolog of Ubp10 in flies, CG15817, is also required for H2B deubiquitination and gene silencing (Figure 1B). It is intriguing that both Ubp10 and USP7 (via an interaction with the Polycomb complex) function in regulating heterochromatin structure, while Ubp8/Nonstop are required for active transcription (van der Knaap et al, 2005). It remains to be determined whether alternative mechanisms for regulating histone ubiquitination at distinct chromatin environments exist within the genomes of higher eukaryotes.

Taken together, our results advance the understanding of the role of histone deubiquitination in transcription, while demonstrating a novel role for SAGA in regulating neural development in the visual system of Drosophila. These findings have implications for the use of Drosophila as a model system to understand the underlying mechanisms of neurodegenerative diseases.

Materials and methods

Phylogenetic analysis

Alignments were performed using web-based CLUSTALW (http://align.genome.jp), followed by the Boxshade server (http://www.ch.embnet.org/software/BOX_form.html). The aligned sequences were then subjected to phylogenetic tree evaluation using standard N-J tree generator available through CLUSTALW.

DNA constructs

The LD43147 and IP07752 cDNAs were cloned into the S2 cell expression vectors pRmHa3-CHA2FL2 (Guelman et al, 2006a) and pMT-V5His (Invitrogen). Ada2b was cloned into pUAST-CHAFL2. LD43147, USP7 (CG1490) and CG5384 were cloned into pRS415. Additional details are provided in Supplementary data.

Co-immunoprecipitations and western blots

Co-immunoprecipitations and affinity purifications of dSAGA using anti-FLAG or anti-HA antibodies were performed as described previously (Guelman et al, 2006a). S2 cells were transfected using Effectene (Qiagen) and induced with 0.5 mM CuSO4 prior to preparation of whole-cell extracts for immunoprecipitation. The following antibodies were used in western blots: dGcn5 (dKAT2) (rabbit, 1:3000); dSpt3 (rabbit, 1:1000); Ada2b (guinea pig, 1:1000); dSgf29 (rabbit, 1:500); tubulin (mouse, 1:5000; Developmental Studies Hybridoma Bank (DSHB)); HA-HRP (mouse, 1:5000; Sigma); FLAG-HRP (mouse, 1:5000; Sigma); V5 (mouse, 1:10 000; Sigma); H3 (rabbit, 1:3000; Abcam); acetylated H3 Lys-9 (rabbit, 1:1000; Abcam) and H2B (rabbit, 1:1000; Upstate).

Genetics

The nonstop, sgf11 and gcn5 fly stocks used in this study were provided by the Bloomington Drosophila Stock Center at Indiana University. Stock numbers are indicated after each genotype. w1118; PBac{w+mC=RB}CG13379e01308/TM6B, Tb1 (17941) contains a piggyBac insertion in the promoter region of sgf11. Excision of the piggyBac transposon by crossing sgf11 to w1118; CyO, P{Tub-PBac}2/wgSp−1 (8285) completely restores viability. P{ry+t7.2=PZ}not02069 ry506/TM6B, rCBTb+ (11553) corresponds to the not2 allele (Martin et al, 1995; Poeck et al, 2001). w1118; PcafE333stP{w+mW.hs=FRT(whs)}2A e1/TM3, P{w+mC=ActGFP}JMR2, Ser1 (9333) corresponds to the gcn5E333st allele (Carre et al, 2005). The ada2b1 mutant flies were kindly provided by Matthias Mannervik (Qi et al, 2004). The sgf11, nonstop and ada2b mutants were crossed to the stock w1118; Drmio/TM3, P{w+mC=GAL4-twi.G}2.3, P{UAS-2xEGFP}AH2.3, Sb1Ser1 (6663) to generate EGFP balanced stocks. Homozygous mutant embryos were identified as described previously (Guelman et al, 2006b) and populations sorted using the CoPas Plus (Union Biometrica). The gcm-GAL4 flies were kindly provided by Iris Salecker.

Immunohistochemistry

CNS/eye–antennal disc complexes from wandering third instar larvae were stained using the following antibodies: anti-chaoptin (mAb24B10, mouse, 1:100; DSHB) (Fujita et al, 1982); anti-Elav (9F8A9, mouse, 1:100; DSHB) (O'Neill et al, 1994); anti-repo (rabbit, 1:500; Joachim Urban) (Halter et al, 1995); anti-β-galactosidase (mouse, 1:5000; Promega); anti-histone H3 (phospho S10) (ab5176, rabbit, 1:100; Abcam); phalloidin Alexa 594 (1:250; Molecular Probes); goat anti-mouse Alexa Fluor 568 (1:300; Molecular Probes); goat anti-mouse Alexa 647 (1:300; Molecular Probes); goat anti-rabbit Alexa Fluor 488 (1:300; Molecular Probes); anti-mouse poly-HRP (ImmunoVision). Laser-scanning confocal imaging was performed using a Zeiss (Thornwood, NY) LSM 510 META attached to a Zeiss Axiovert 200M inverted microscope. Confocal images were collected using either a × 40/1.2 NA water immersion C-Apochromat objective or a × 40/1.3 NA oil immersion Plan-Neofluar objective at variable scanning zoom. Confocal images are presented either as single planes generated in Zeiss AIM software or as 3D reconstructed images using Imaris (Bitplane). Brightfield imaging was performed using a × 40/1.3 NA oil immersion Plan Neofluar objective on a Zeiss Axiovert 200M upright microscope equipped with a Zeiss AxioCam HRc r1.4 digital color camera connected by a × 0.63 coupler. Repo-positive glial cells with a 2.5 μm minimum diameter were identified and counted digitally using Imaris in 11 samples of each genotype along a measured length (40–100 μm) of the lamina plexus.

Gene expression analysis

RNA was isolated in triplicate from separate collections of late third instar larvae of the appropriate genotype using TriZol (Invitrogen) and mRNA enriched using Oligotex beads (Qiagen). Arrays consist of 14 593 unique 70-mer oligonucleotides corresponding to version 1.1 of the D. melanogaster set (Operon). Microarray data are available at ArrayExpress (http://www.ebi.ac.uk/arrayexpress/) under accession number E-TABM-383.

Supplementary Material

Supplementary Information

Acknowledgments

We thank the Workman and Abmayr labs for critical discussions. We thank J Haug, J Wunderlich and K Wagner for EGFP sorting of embryos, and J Schwartz, C Cooper and K Perko for help with confocal microscopy. We also thank T Johnson and S Beckham for their assistance with immunohistochemistry, and B Fleharty for microarray hybridization. We thank L Zipursky for the ro-τlacZ flies and I Salecker and CH Lee for the gcm-GAL4 flies. KK Lee was supported by a Damon Runyon postdoctoral fellowship.

References

  • Allis CD, Berger SL, Cote J, Dent S, Jenuwien T, Kouzarides T, Pillus L, Reinberg D, Shi Y, Shiekhattar R, Shilatifard A, Workman J, Zhang Y (2007) New nomenclature for chromatin-modifying enzymes. Cell 131: 633–636 [PubMed]
  • Awasaki T, Ito K (2004) Engulfing action of glial cells is required for programmed axon pruning during Drosophila metamorphosis. Curr Biol 14: 668–677 [PubMed]
  • Brennan CA, Li TR, Bender M, Hsiung F, Moses K (2001) Broad-complex, but not ecdysone receptor, is required for progression of the morphogenetic furrow in the Drosophila eye. Development 128: 1–11 [PubMed]
  • Carre C, Szymczak D, Pidoux J, Antoniewski C (2005) The histone H3 acetylase dGcn5 is a key player in Drosophila melanogaster metamorphosis. Mol Cell Biol 25: 8228–8238 [PMC free article] [PubMed]
  • Chen X, Fischer JA (2000) In vivo structure/function analysis of the Drosophila fat facets deubiquitinating enzyme gene. Genetics 156: 1829–1836 [PMC free article] [PubMed]
  • Chotard C, Leung W, Salecker I (2005) Glial cells missing and gcm2 cell autonomously regulate both glial and neuronal development in the visual system of Drosophila. Neuron 48: 237–251 [PubMed]
  • Clandinin TR, Zipursky SL (2002) Making connections in the fly visual system. Neuron 35: 827–841 [PubMed]
  • Daniel JA, Torok MS, Sun ZW, Schieltz D, Allis CD, Yates JR III, Grant PA (2004) Deubiquitination of histone H2B by a yeast acetyltransferase complex regulates transcription. J Biol Chem 279: 1867–1871 [PubMed]
  • David G, Abbas N, Stevanin G, Durr A, Yvert G, Cancel G, Weber C, Imbert G, Saudou F, Antoniou E, Drabkin H, Gemmill R, Giunti P, Benomar A, Wood N, Ruberg M, Agid Y, Mandel JL, Brice A (1997) Cloning of the SCA7 gene reveals a highly unstable CAG repeat expansion. Nat Genet 17: 65–70 [PubMed]
  • Emre NC, Ingvarsdottir K, Wyce A, Wood A, Krogan NJ, Henry KW, Li K, Marmorstein R, Greenblatt JF, Shilatifard A, Berger SL (2005) Maintenance of low histone ubiquitylation by Ubp10 correlates with telomere-proximal Sir2 association and gene silencing. Mol Cell 17: 585–594 [PubMed]
  • Fujita SC, Zipursky SL, Benzer S, Ferrus A, Shotwell SL (1982) Monoclonal antibodies against the Drosophila nervous system. Proc Natl Acad Sci USA 79: 7929–7933 [PMC free article] [PubMed]
  • Gause M, Eissenberg JC, Macrae AF, Dorsett M, Misulovin Z, Dorsett D (2006) Nipped-A, the Tra1/TRRAP subunit of the Drosophila SAGA and Tip60 complexes, has multiple roles in Notch signaling during wing development. Mol Cell Biol 26: 2347–2359 [PMC free article] [PubMed]
  • Grant PA, Duggan L, Cote J, Roberts SM, Brownell JE, Candau R, Ohba R, Owen-Hughes T, Allis CD, Winston F, Berger SL, Workman JL (1997) Yeast Gcn5 functions in two multisubunit complexes to acetylate nucleosomal histones: characterization of an Ada complex and the SAGA (Spt/Ada) complex. Genes Dev 11: 1640–1650 [PubMed]
  • Guelman S, Suganuma T, Florens L, Swanson SK, Kiesecker CL, Kusch T, Anderson S, Yates JR III, Washburn MP, Abmayr SM, Workman JL (2006a) Host cell factor and an uncharacterized SANT domain protein are stable components of ATAC, a novel dAda2A/dGcn5-containing histone acetyltransferase complex in Drosophila. Mol Cell Biol 26: 871–882 [PMC free article] [PubMed]
  • Guelman S, Suganuma T, Florens L, Weake V, Swanson SK, Washburn MP, Abmayr SM, Workman JL (2006b) The essential gene wda encodes a WD40 repeat subunit of Drosophila SAGA required for histone H3 acetylation. Mol Cell Biol 26: 7178–7189 [PMC free article] [PubMed]
  • Halter DA, Urban J, Rickert C, Ner SS, Ito K, Travers AA, Technau GM (1995) The homeobox gene repo is required for the differentiation and maintenance of glia function in the embryonic nervous system of Drosophila melanogaster. Development 121: 317–332 [PubMed]
  • Helmlinger D, Hardy S, Abou-Sleymane G, Eberlin A, Bowman AB, Gansmuller A, Picaud S, Zoghbi HY, Trottier Y, Tora L, Devys D (2006) Glutamine-expanded ataxin-7 alters TFTC/STAGA recruitment and chromatin structure leading to photoreceptor dysfunction. PLoS Biol 4: e67. [PMC free article] [PubMed]
  • Helmlinger D, Hardy S, Sasorith S, Klein F, Robert F, Weber C, Miguet L, Potier N, Van-Dorsselaer A, Wurtz JM, Mandel JL, Tora L, Devys D (2004) Ataxin-7 is a subunit of GCN5 histone acetyltransferase-containing complexes. Hum Mol Genet 13: 1257–1265 [PubMed]
  • Henry KW, Wyce A, Lo WS, Duggan LJ, Emre NC, Kao CF, Pillus L, Shilatifard A, Osley MA, Berger SL (2003) Transcriptional activation via sequential histone H2B ubiquitylation and deubiquitylation, mediated by SAGA-associated Ubp8. Genes Dev 17: 2648–2663 [PMC free article] [PubMed]
  • Ingvarsdottir K, Krogan NJ, Emre NC, Wyce A, Thompson NJ, Emili A, Hughes TR, Greenblatt JF, Berger SL (2005) H2B ubiquitin protease Ubp8 and Sgf11 constitute a discrete functional module within the Saccharomyces cerevisiae SAGA complex. Mol Cell Biol 25: 1162–1172 [PMC free article] [PubMed]
  • Jentsch S, McGrath JP, Varshavsky A (1987) The yeast DNA repair gene RAD6 encodes a ubiquitin-conjugating enzyme. Nature 329: 131–134 [PubMed]
  • Johnson EC, Bohn LM, Barak LS, Birse RT, Nassel DR, Caron MG, Taghert PH (2003) Identification of Drosophila neuropeptide receptors by G protein-coupled receptors-beta-arrestin2 interactions. J Biol Chem 278: 52172–52178 [PubMed]
  • Kohler A, Pascual-Garcia P, Llopis A, Zapater M, Posas F, Hurt E, Rodriguez-Navarro S (2006) The mRNA export factor Sus1 is involved in Spt/Ada/Gcn5 acetyltransferase-mediated H2B deubiquitinylation through its interaction with Ubp8 and Sgf11. Mol Biol Cell 17: 4228–4236 [PMC free article] [PubMed]
  • Kusch T, Guelman S, Abmayr SM, Workman JL (2003) Two Drosophila Ada2 homologues function in different multiprotein complexes. Mol Cell Biol 23: 3305–3319 [PMC free article] [PubMed]
  • Lee KK, Florens L, Swanson SK, Washburn MP, Workman JL (2005) The deubiquitylation activity of Ubp8 is dependent upon Sgf11 and its association with the SAGA complex. Mol Cell Biol 25: 1173–1182 [PMC free article] [PubMed]
  • Lee KK, Workman JL (2007) Histone acetyltransferase complexes: one size doesn't fit all. Nature Rev 8: 284–295 [PubMed]
  • Lee T, Marticke S, Sung C, Robinow S, Luo L (2000) Cell-autonomous requirement of the USP/EcR-B ecdysone receptor for mushroom body neuronal remodeling in Drosophila. Neuron 28: 807–818 [PubMed]
  • Li TR, White KP (2003) Tissue-specific gene expression and ecdysone-regulated genomic networks in Drosophila. Dev Cell 5: 59–72 [PubMed]
  • Martin KA, Poeck B, Roth H, Ebens AJ, Ballard LC, Zipursky SL (1995) Mutations disrupting neuronal connectivity in the Drosophila visual system. Neuron 14: 229–240 [PubMed]
  • Monnier D, Colas JF, Rosay P, Hen R, Borrelli E, Maroteaux L (1992) NKD, a developmentally regulated tachykinin receptor in Drosophila. J Biol Chem 267: 1298–1302 [PubMed]
  • Muratoglu S, Georgieva S, Papai G, Scheer E, Enunlu I, Komonyi O, Cserpan I, Lebedeva L, Nabirochkina E, Udvardy A, Tora L, Boros I (2003) Two different Drosophila ADA2 homologues are present in distinct GCN5 histone acetyltransferase-containing complexes. Mol Cell Biol 23: 306–321 [PMC free article] [PubMed]
  • O'Neill EM, Rebay I, Tjian R, Rubin GM (1994) The activities of two Ets-related transcription factors required for Drosophila eye development are modulated by the Ras/MAPK pathway. Cell 78: 137–147 [PubMed]
  • Osley MA (2006) Regulation of histone H2A and H2B ubiquitylation. Brief Funct Genomic Proteomic 5: 179–189 [PubMed]
  • Palhan VB, Chen S, Peng GH, Tjernberg A, Gamper AM, Fan Y, Chait BT, La Spada AR, Roeder RG (2005) Polyglutamine-expanded ataxin-7 inhibits STAGA histone acetyltransferase activity to produce retinal degeneration. Proc Natl Acad Sci USA 102: 8472–8477 [PMC free article] [PubMed]
  • Pankotai T, Komonyi O, Bodai L, Ujfaludi Z, Muratoglu S, Ciurciu A, Tora L, Szabad J, Boros I (2005) The homologous Drosophila transcriptional adaptors ADA2a and ADA2b are both required for normal development but have different functions. Mol Cell Biol 25: 8215–8227 [PMC free article] [PubMed]
  • Poeck B, Fischer S, Gunning D, Zipursky SL, Salecker I (2001) Glial cells mediate target layer selection of retinal axons in the developing visual system of Drosophila. Neuron 29: 99–113 [PubMed]
  • Qi D, Larsson J, Mannervik M (2004) Drosophila Ada2b is required for viability and normal histone H3 acetylation. Mol Cell Biol 24: 8080–8089 [PMC free article] [PubMed]
  • Robzyk K, Recht J, Osley MA (2000) Rad6-dependent ubiquitination of histone H2B in yeast. Science 287: 501–504 [PubMed]
  • Schubiger M, Wade AA, Carney GE, Truman JW, Bender M (1998) Drosophila EcR-B ecdysone receptor isoforms are required for larval molting and for neuron remodeling during metamorphosis. Development 125: 2053–2062 [PubMed]
  • van der Knaap JA, Kumar BR, Moshkin YM, Langenberg K, Krijgsveld J, Heck AJ, Karch F, Verrijzer CP (2005) GMP synthetase stimulates histone H2B deubiquitylation by the epigenetic silencer USP7. Mol Cell 17: 695–707 [PubMed]
  • Wood A, Krogan NJ, Dover J, Schneider J, Heidt J, Boateng MA, Dean K, Golshani A, Zhang Y, Greenblatt JF, Johnston M, Shilatifard A (2003a) Bre1, an E3 ubiquitin ligase required for recruitment and substrate selection of Rad6 at a promoter. Mol Cell 11: 267–274 [PubMed]
  • Wood A, Schneider J, Dover J, Johnston M, Shilatifard A (2003b) The Paf1 complex is essential for histone monoubiquitination by the Rad6-Bre1 complex, which signals for histone methylation by COMPASS and Dot1p. J Biol Chem 278: 34739–34742 [PubMed]
  • Xiao T, Kao CF, Krogan NJ, Sun ZW, Greenblatt JF, Osley MA, Strahl BD (2005) Histone H2B ubiquitylation is associated with elongating RNA polymerase II. Mol Cell Biol 25: 637–651 [PMC free article] [PubMed]
  • Zhu B, Zheng Y, Pham AD, Mandal SS, Erdjument-Bromage H, Tempst P, Reinberg D (2005) Monoubiquitination of human histone H2B: the factors involved and their roles in HOX gene regulation. Mol Cell 20: 601–611 [PubMed]

Articles from The EMBO Journal are provided here courtesy of The European Molecular Biology Organization
PubReader format: click here to try

Formats:

Related citations in PubMed

See reviews...See all...

Cited by other articles in PMC

See all...

Links

  • Gene
    Gene
    Gene links
  • GEO Profiles
    GEO Profiles
    Related GEO records
  • HomoloGene
    HomoloGene
    HomoloGene links
  • MedGen
    MedGen
    Related information in MedGen
  • Pathways + GO
    Pathways + GO
    Pathways, annotations and biological systems (BioSystems) that cite the current article.
  • Protein
    Protein
    Published protein sequences
  • PubMed
    PubMed
    PubMed citations for these articles
  • Substance
    Substance
    PubChem Substance links
  • Taxonomy
    Taxonomy
    Related taxonomy entry
  • Taxonomy Tree
    Taxonomy Tree

Recent Activity

Your browsing activity is empty.

Activity recording is turned off.

Turn recording back on

See more...