Core histone modifications play an important role in chromatin remodeling and transcriptional regulation. Histone acetylation is one of the best-studied gene modifications and has been shown to be involved in numerous important biological processes. Herein, we demonstrated that the depletion of histone deacetylase 3 (Hdac3) in Drosophila melanogaster resulted in a reduction in body size. Further genetic studies showed that Hdac3 counteracted the organ overgrowth induced by overexpression of insulin receptor (InR), phosphoinositide 3-kinase (PI3K) or S6 kinase (S6K), and the growth regulation by Hdac3 was mediated through the deacetylation of histone H4 at lysine 16 (H4K16). Consistently, the alterations of H4K16 acetylation (H4K16ac) induced by the overexpression or depletion of males-absent-on-the-first (MOF), a histone acetyltransferase that specifically targets H4K16, resulted in changes in body size. Furthermore, we found that H4K16ac was modulated by PI3K signaling cascades. The activation of the PI3K pathway caused a reduction in H4K16ac, whereas the inactivation of the PI3K pathway resulted in an increase in H4K16ac. The increase in H4K16ac by the depletion of Hdac3 counteracted the PI3K-induced tissue overgrowth and PI3K-mediated alterations in the transcription profile. Overall, our studies indicated that Hdac3 served as an important regulator of the PI3K pathway and revealed a novel link between histone acetylation and growth control.
Core histone modifications are known to play an essential role in the regulation of chromatin organization and transcription. These modifications include acetylation, methylation, phosphorylation, ubiquitination, sumoylation and poly(ADP-ribosyl)ation (Berger, 1999; Jenuwein and Allis, 2001; Kouzarides, 2007). Histone acetylation is one of the best-studied modifications and is thought to be involved in both the initiation and elongation steps of transcription (Vogelauer et al., 2000; Hassan et al., 2001; Wang et al., 2009). The acetylation of the core histone tails alters the folding dynamics of nucleosomal arrays and 30-nm chromatin fibers (Luger and Richmond, 1998; Annunziato and Hansen, 2000) and recruits specific chromatin remodeling complexes that exert the specific function(s) of chromatin (Turner, 2000; Shahbazian and Grunstein, 2007).
The acetylation of histones is regulated by two highly conserved classes of histone enzymes, histone acetyltransferases (HATs) and histone deacetylases (HDACs) (Peterson and Laniel, 2004), which catalyze the addition and removal, respectively, of acetyl groups on histone lysine residues (Roth et al., 2001; Struhl, 1998). Reversible histone acetylation and deacetylation are highly regulated processes that are crucial for chromatin reorganization and the regulation of gene transcription in response to extracellular conditions. The balance between the acetylation and deacetylation of histones serves as a key regulatory mechanism for gene expression and governs numerous developmental processes and disease states (Marks et al., 2001; Butler and Bates, 2006).
HDACs have been classified into four subfamilies based on their homologs and functional similarities (Witt et al., 2009). Hdac3 is a class I HDAC that shares homology with yeast Rpd3. This protein is reportedly present in the nuclear, cytoplasmic and membrane fractions (Yang et al., 2002). The knockout of Hdac3 in mice leads to embryonic lethality before day 9.5 (Bhaskara et al., 2008). The inactivation of Hdac3 has been shown to delay cell cycle progression and result in cell cycle-dependent DNA damage, inefficient repair and increased apoptosis in mouse embryonic fibroblasts (Montgomery et al., 2008; Knutson et al., 2008; Zampetaki et al., 2010). Hdac3 has also been shown to be upregulated in various tumor types (Spurling et al., 2008; Wu et al., 2010; Wilson et al., 2006). However, the precise function and underlying molecular mechanism of Hdac3 in these processes remain largely unknown.
The Drosophila melanogaster (D. melanogaster) ortholog to human Hdac3 is known to be Hdac3 or dHDAC3 (Johnson et al., 1998). Herein, we used D. melanogaster to investigate the function of Hdac3 during development. We found that the depletion of Hdac3 in D. melanogaster results in a reduction in both organ and body sizes. Hdac3 controls growth through the regulation of H4K16 deacetylation. Alterations in H4K16ac through the ectopic expression of MOF, a histone acetyltransferase that specifically targets H4K16, result in changes of cell/body size. We also found that H4K16ac is modulated by PI3K signaling. Increasing the level of H4K16ac by depleting Hdac3 effectively reverses the PI3K-induced tissue overgrowth and alterations in the transcription profile.
The depletion of Hdac3 results in a reduction in both organ and body sizes in Drosophila melanogaster
Hdac3 is conserved between fruit flies and mammals (supplementary material Fig. S1). To investigate the role of Hdac3 during development, transgenic flies expressing double-stranded (ds) D. melanogaster Hdac3 RNA were produced using a UAS/Gal4 system as previously described (Liu et al., 2005). An Hdac3 cDNA fragment (between 691 and 1198 bp in its open reading frame) was cloned into a SympUAST vector, which allowed the Hdac3 fragment to be transcribed bi-directionally, thereby producing dsRNA and silencing the endogenous gene targets. The ubiquitous expression of Hdac3 RNAi using the actin-Gal4 driver resulted in lethality before the 3rd-instar larval stage. Tissue-specific expression of Hdac3 RNAi was achieved by crossing the flies carrying the RNAi transgene with flies carrying eyeless-Gal4 (ey-Gal4) or engrailed-Gal4 (en-Gal4), which resulted in a reduction in the size of the eye (Fig. 1B) and the posterior compartment of the wing (Fig. 1C). When Hdac3 RNAi was expressed in the fat body using pumpless-Gal4 (ppl-Gal4), the progeny of the Hdac3 RNAi flies were partially lethal (supplementary material Table S1), and the surviving progeny had a small body size at the 3rd-instar larval, pupal and adult stages (Fig. 1D). The average weight of the Hdac3 RNAi female progeny was ∼16% less than that of the control (ppl-Gal4/+) female flies (P<0.05; n = 5 groups per genotype; 100 flies in each group, with the same quantity used in each of the following groups, unless specified), and the average weight of the male progeny was reduced by ∼14% (P<0.05, n = 5 groups) compared with that of the controls (ppl-Gal4/+, n = 5 groups). Two independent transgenic Hdac3 RNAi lines were both found to affect body size, implying that the defect was not due to the location of the transgene in the genome.
To confirm the effect of Hdac3 on growth, we also produced additional transgenic fly lines that overexpress Hdac3. Progeny ectopically expressing Hdac3 with the actin-Gal4 driver indeed exhibited slightly increased body size compared with that of the controls (supplementary material Fig. S2). Compared with the controls (actin-Gal4/+, n = 5 groups), the body weight of the female progeny increased by 6% (P<0.05, n = 5 groups), and the body weight of the male progeny increased by 5% (P<0.05, n = 5 groups). The phenotype resulting from the depletion of Hdac3 was rescued by the co-expression of Hdac3 (data not shown), implying that the growth defects specifically result from a reduction in Hdac3.
Hdac3 controls growth by regulating cell size and number
To examine whether the reduction in organ/body size after the depletion of Hdac3 was due to the decrease in cell size, cell number or both, the phenotype of depleted Hdac3 was characterized by quantifying the area and number of the ommatidia, which reflect cell size and cell number, respectively. Scanning electron micrographs (SEMs) of whole eyes showed that the eyes of Hdac3 RNAi flies were smaller than those of the controls. The average area of the male fly eyes was reduced by 30% (P<0.01, n = 20) when compared with those of the controls (n = 18) (Fig. 2).
The size of the ommatidia was determined by counting the number of ommatidia in a specific square (0.13 mm×0.13 mm). The depletion of Hdac3 resulted in an 11% reduction in ommatidial size (P<0.05, n = 20) when compared with that of the controls (n = 18) (Fig. 2). The total number of ommatidia in Hdac3-depleted eyes revealed a 24% reduction (P<0.01, n = 20) when compared with that of the controls (n = 18) (Fig. 2). The 11% decrease in cell area and 24% decrease in cell number can account for the 30% reduction in eye area observed in the Hdac3-depleted flies. Therefore, these results suggest that Hdac3 controls organ/body size via the alteration of both cell size and cell number.
PI3K pathway activation is impaired in Hdac3-depleted flies
The InR/PI3K pathway is reported to be important in regulating growth (Chen et al., 1996; Leevers et al., 1996; Montagne et al., 1999). The inactivation of positive regulatory components of the pathway has been shown to decrease organ size, whereas the overexpression of these regulatory components can result in tissue overgrowth. Given the similar phenotypes between the depletion of Hdac3 and the inactivation of the pathways, we suspected that Hdac3 is a component of these pathways. To test this possibility, we then studied the subcellular localization of the pleckstrin homology (PH) domain using a transgenic fly expressing a GFP-PH domain fusion protein (tGPH) (Britton et al., 2002).
In the InR/PI3K signaling pathway, insulin, or an insulin-like ligand, binds to and activates the insulin receptor tyrosine kinase (InR) on target cells (Chen et al., 1996), resulting in the activation of phosphatidylinositol 3-kinase (PI3K) (Leevers et al., 1996). Activated PI3K generates the second messenger phosphatidylinositol-3,4,5-trisphosphate, which in turn binds to the PH domain-containing proteins and transports it from the cytosol to the plasma membrane (Lizcano and Alessi, 2002). Therefore, the activation of PI3K signaling is often evaluated by the subcellular localization of the tGPH reporter. If Hdac3 depletion is associated with the activation of PI3K signaling, the tGPH reporter may be relocated. Indeed, we found that the tGPH reporter showed a marked reduction of plasma membrane-bound tGPH in the larval fat body cells of Hdac3 RNAi flies, whereas in the control larvae, tGPH was mainly localized at the plasma membrane (Fig. 3A).
We next measured the level of phospho-Akt and phospho-S6K in the Hdac3-depleted flies using western blot analysis. The results showed that the depletion of Hdac3 caused a slight decrease in the level of phospho-Akt and phospho S6K (Fig. 3B), supporting an interaction between Hdac3 and the PI3K-Akt pathway. In addition, we tested the interaction of Hdac3 with the JAK-STAT and the EGFR pathways. Neither the JAK-STAT nor the EGFR pathways was obviously affected in the Hdac3-depleted flies (supplementary material Fig. S3). These observations support that the depletion of Hdac3 negatively regulates PI3K pathway, but not others.
The depletion of Hdac3 counteracts the tissue overgrowth induced by InR, PI3K or S6K
InR, PI3K and S6 kinase (S6K) (Montagne et al., 1999) are positive regulatory components of the InR/PI3K/S6K pathways. The inactivation of these positive regulatory components has been verified to decrease organ size, whereas their overexpression results in tissue overgrowth. To further evaluate the functional role of Hdac3 in the InR/PI3K signaling pathways, we performed genetic experiments in the eye using GMR-Gal4.
The overexpression of InR is known to cause large, disorganized eyes (Chen et al., 1996). However, these overgrowth phenotypes were strongly suppressed when Hdac3 was depleted in the eye in conjunction with InR overexpression (Fig. 4). Notably, the progeny co-expressing InR and Hdac3 RNAi had a phenotype that resembled those observed in flies expressing Hdac3 RNAi alone. Similar results were also obtained when PI3K or S6K was co-expressed with Hdac3 RNAi (Fig. 4). These results suggest that Hdac3 is also functionally involved in the organ growth regulated by InR, PI3K or S6K.
The depletion of Hdac3 leads to an increase in histone H4 acetylation at lysine 16
Loss-of-function mutations in Hdac3 are known to affect histone acetylation and position-effect variegation (Zhu et al., 2008), a heterochromatin-associated transcriptional gene silencing phenomenon (Eissenberg and Elgin, 2000). We were interested in exploring whether Hdac3 regulates organ/body size through the regulation of histone acetylation. To test this hypothesis, we performed immunostaining to examine the effects of Hdac3 depletion on histone acetylation. We expressed Hdac3 RNAi in wing discs with en-Gal4, which activated the Gal4 driver in the posterior wing compartment. The wing discs were dissected and stained using antibodies against the specific acetylated lysines on the core histones H3/H4. The stained wing discs showed an increase in H4K16ac where Hdac3 was depleted (Fig. 5A, circled area). In contrast, H4K12 acetylation was not remarkably affected. Western blots showed the same results (Fig. 5A).
We also used hsFLP/FRT-mediated mitotic recombination to generate Hdac3-depleted clones in the fat body (Manfruelli et al., 1999). Larvae with GFP fluorescent signals were selected and dissected. The fat body tissues were stained using antibodies against H4K16ac and discs large (Dlg, a septate junction protein that labels membranes). Compared with the control cells adjacent to the induced clones, the Hdac3-depleted clones exhibited elevated H4K16ac levels (Fig. 5A). These results indicate that Hdac3 depletion triggers an increase in H4K16ac levels.
Hdac3 regulates organ/body size via the regulation of H4K16ac
Histone H4K16 acetylation is known to play an important role in numerous biological processes, such as chromatin remodeling, the DNA damage response, double-strand break (DSB) repair and the cellular life-span (Shogren-Knaak et al., 2006; Krishnan et al., 2011; Sharma et al., 2010; Dang et al., 2009). To investigate whether H4K16ac is involved in Hdac3-mediated growth control processes, we next generated transgenic flies overexpressing H4K16A, an H4 mutant in which the lysine at residue 16 (K16) is mutated into an alanine (K16A), which makes the residue incapable of being acetylated. The results showed that the overexpression of the H4K16A mutant in the Hdac3 RNAi background partially attenuated the Hdac3 depletion-induced decrease in eye size (Fig. 5B). Consistent with this finding, the reduction in body weight induced by the depletion of Hdac3 was also partially attenuated by the co-expressing with the H4K16A mutant (Fig. 5B). The genetic analyses support the idea that Hdac3 regulates organ/body size through the regulation of H4K16ac.
Histone H4K16 acetylation is modulated by the PI3K signaling pathway
PI3K signaling is a key signal transduction pathway with many important physiological responses (Cantley, 2002). However, both the precise role of PI3K signaling and the mechanism by which the PI3K signaling cascades regulate gene transcription in the nucleus remain to be determined. Chromatin remodeling events, especially histone posttranslational modifications, provide an epigenetic mechanism for the regulation of many nuclear events, including gene transcription. The genetic relationship between Hdac3 and the PI3K pathway prompted us to test whether PI3K signaling is also associated with H4K16ac. We performed immunostaining in wing discs from progeny that overexpressed either wild type PI3K (en >PI3K) or a dominant negative PI3K mutant (en >PI3KDN) using antibodies against H4K16ac and H4K12ac. The results indicated that the activation of PI3K (en >PI3K) in the posterior wing compartment (circled area) caused a consistent reduction in H4K16ac, whereas the inactivation of the PI3K pathway (en >PI3KDN) resulted in an increase in H4K16ac (Fig. 6). The same results were obtained by western blot analysis (Fig. 6). These results suggest that H4K16ac is modulated by PI3K signaling.
Changes in the level of H4K16ac by the overexpression or depletion of MOF result in the alteration of cell/body size
MOF is a histone acetyltransferase that specifically acetylates histone H4 at lysine 16 (Akhtar and Becker, 2000). However, the function of MOF in animal growth has not been investigated. To confirm the effect of H4K16ac on growth, we changed H4K16ac by modulating the expression of MOF. Transgenic lines that overexpress or deplete MOF were produced using a UAS/Gal4 system. We performed immunostaining in wing discs from progeny that were either MOF-depleted (en >MOF Ri) or MOF overexpressed (Vg >MOF ov, because en >MOF ov can cause lethality before the 3rd-instar larval stage) using antibodies against H4K16ac. As expected, MOF overexpression resulted in increased H4K16ac levels, and MOF depletion resulted in decreased H4K16ac levels (Fig. 7A).
The ubiquitous depletion of MOF under the control of actin-Gal4 caused male-specific lethality, which is consistent with the previous report (Akhtar and Becker, 2000). The surviving female progeny had an enlarged body size at the 3rd-instar larval, pupal and adult stages (Fig. 7B). The average weight of the MOF-depleted female adult progeny was increased by more than 10% (P<0.05, n = 5 groups) when compared with the control (actin-Gal4/+, n = 5 groups).
MOF overexpression under the control of the actin-Gal4, ppl-Gal4 and en-Gal4 drivers resulted in lethality at early larval stages (data not shown). To examine the function of H4K16ac, we generated MOF overexpressing clones in the fat body cells by mitotic recombination using the hsFLP/FRT system. The MOF overexpressing clones that displayed an increased level of H4K16ac showed a marked decrease in cell size when compared with the control cells adjacent to the induced clones (Fig. 7C). A similar reduction in cell size was also observed in salivary gland cells (Fig. 7D) in which MOF was activated by the OK107-Gal4 driver, which is known to induce Gal4 expression in mushroom bodies and in a subset of cells in the salivary glands. These results confirm that H4K16ac plays a direct role in growth.
The increase in H4K16ac by Hdac3 depletion counteracts the PI3K-induced tissue overgrowth and global transcriptional profile
Given that H4K16ac is involved in growth and that PI3K regulates both the growth of tissues and the level of H4K16ac, we explored whether Hdac3 inhibits PI3K signaling via the regulation of H4K16ac. Immunoassays were performed in wing discs. The results showed that the depletion of Hdac3 rescued the decrease in H4K16ac caused by the overexpression of PI3K (Fig. 8A). The PI3K-induced overgrowth phenotypes were also suppressed by the depletion of Hdac3 (Fig. 8B). Notably, the hyperphosphorylation of Akt following PI3K activation appeared not to be affected after the depletion of Hdac3 (Fig. 8A).
To confirm the role of H4K16ac in the PI3K-mediated growth regulation, the co-expression of H4K16A and PI3K was performed under the control of the GMR-Gal4 driver. As previously reported, the overexpression of PI3K resulted in large, disorganized eyes (Leevers et al., 1996). The presence of H4K16A further enlarged the PI3K-mediated increase in ommatidial size (supplementary material Fig. S4), confirming that Hdac3 antagonizes PI3K signaling via its regulatory effect on the acetylation of histone H4K16.
To further investigate the molecular mechanisms underlying the effect of Hdac3 depletion on the PI3K pathway, both microarray analysis and quantitative real-time PCR were conducted to compare the transcriptional activity in flies overexpressing PI3K with that in flies co-expressing PI3K and Hdac3 RNAi. The gene expression profile analysis revealed that a large proportion of upregulated genes observed in the PI3K overexpressing flies were downregulated in the Hdac3 RNAi flies and that a number of genes downregulated in the PI3K overexpressing flies were upregulated in the Hdac3 RNAi flies (Fig. 8C). Notably, the depletion of Hdac3 in the PI3K overexpressing flies restored ∼85% (288 out of the 339 genes) of the twofold difference in gene transcription induced by the activation of the PI3K pathway (Fig. 8C). The transcriptional changes in the genes involved in the regulation of size (Puig et al., 2003; Huang et al., 1999; Bender et al., 1997) were further detected using quantitative real-time PCR. As shown in Fig. 8D, the negative regulators of growth, such as FOXO (forkhead box, sub-group O), PTEN (phosphatase and tensin homolog), EcR (Ecdysone receptor) and Eip74EF (Ecdysone-induced protein 74EF), that were downregulated in the PI3K overexpressing flies were rescued by Hdac3 depletion. These results demonstrated that the depletion of Hdac3 also counteracted the PI3K-mediated gene expression.
A fundamental question in biology is how organisms are able to grow to an appropriate body and organ size and what are the mechanisms that control the process of growth. Drosophila melanogaster is an ideal model system for examining the mechanisms of growth regulation. Previous studies have identified several genes associated with growth, but the molecular mechanisms have not been fully understood. In this paper, we utilized a Drosophila UAS/Gal4 system and characterized the function of Hdac3 in Drosophila. The results suggest that Hdac3 is a critical player in both organ and body growth. Hdac3 depletion caused a reduction in both organ and body size due to fewer and smaller cells (Fig. 1; Fig. 2).
Hdac3 is a component of the nuclear receptor co-repressor complex containing N-CoR (nuclear receptor corepressor) and SMRT (silencing mediator for retinoid and thyroid hormone receptors), both of which are recruited by nuclear hormone receptors to regulate gene transcription (Li et al., 2000; Guenther et al., 2001). Several substrates were found to be targets of Hdac3, including histones (Bhaskara et al., 2010; Johnson et al., 2002) and non-histone proteins (Yang et al., 1997; Zampetaki et al., 2010). Among the targets affected by Hdac3, we found that H4K16ac is a critical epigenetic modification associated with animal growth, as demonstrated not only by the finding that alterations in H4K16ac were closely associated with Hdac3-induced organ/body growth but also by the finding that mutating H4K16 directly affected Hdac3-induced growth (Fig. 5). Furthermore, transgenic lines in which MOF, the specific histone H4K16 HAT, was overexpressed or depleted exhibited similar changes in cell/body size (Fig. 7), thus confirming that H4K16ac plays an essential role in animal growth.
Histone H4K16 acetylation is known to function as a dual switch for higher-order chromatin and protein–histone interactions (Shogren-Knaak et al., 2006) and has been shown to regulate embryonic stem cell self-renewal (Li et al., 2012) and cellular life span (Dang et al., 2009). Recent work in our laboratory has suggested that H4K16ac in Drosophila not only is critical for the acetylation of H4K5, H4K8 and H3K9, which are hallmarks of active chromatin, but also exerts an effect on H3K9 methylation and the association of HP1 with chromatin, which are hallmarks of heterochromatin (H.-M.W. and F.-L.S., unpublished data, report in preparation). We therefore presume that the changes in H4K16ac affect higher-order chromatin and alter the transcription of genes related to growth. However, the exact mechanism by which H4K16ac regulates the transcription of genes related to growth needs to be further investigated.
One of the main findings in this work is the genetic interaction between Hdac3/H4K16ac and the PI3K pathway. The PI3K pathway is a highly conserved signal transduction cascade from flies to humans. Previous studies have identified a number of the components of this signaling pathway. However, the mechanisms by which this pathway regulates nuclear events, such as gene transcription, remain largely unknown. In this work, we showed for the first time that PI3K signaling modulates the acetylation of H4K16. This finding was supported by our results showing that the activation of PI3K caused a corresponding reduction in H4K16ac, whereas the inactivation of the PI3K pathway resulted in an increase in H4K16ac (Fig. 6). Furthermore, the introduction of the H4K16A mutant, in which H4K16 cannot be acetylated, further enlarged the PI3K-induced increase in ommatidial size (supplementary material Fig. S4), confirming the function of histone H4K16ac in PI3K signaling.
Although the exact mechanism by which PI3K regulates H4K16ac is still unknown, we demonstrated that the loss of Hdac3 inhibited PI3K-mediated overgrowth, thus suggesting that PI3K targets the activity of Hdac3 and subsequently affects H4K16ac. This hypothesis is supported by the observations that Drosophila Hdac3 can form a complex with Akt (W.W.L. and F.L.S., unpublished data) and that the complex of human Hdac3 with the deacetylase activation domain (DAD), the human SMRT co-repressor and inositol tetraphosphate is required for the activation of Hdac3 enzymatic functionality (Watson et al., 2012). Our observation that the depletion of Hdac3 decreased the level of phospho-Akt and affected the subcellular localization of GFP-PH (Fig. 3) also supported this possibility. However, the observation that Hdac3 depletion failed to counteract the PI3K-induced hyperphosphorylation of Akt while completely rescuing the decrease in H4K16ac and the tissue overgrowth induced by the PI3K overexpression (Fig. 8) indicated that Hdac3 likely counteracts the PI3K-induced tissue overgrowth by modulating the level of H4K16ac.
The hyperactivation of the PI3K pathway is known to be associated with many types of human cancer (Ghayad and Cohen, 2010; Martelli et al., 2009; Kawauchi et al., 2009). A number of HDAC inhibitors have been developed and applied in clinical trials to inhibit tumor growth (Witt et al., 2009). However, the molecular mechanisms of these HDAC inhibitors in cancer prevention remain to be elucidated. In the present study, we found that the overexpression of PI3K decreases H4K16ac in vivo (Fig. 6). Further studies have shown that increasing the level of H4K16ac by depleting Hdac3 can antagonize the PI3K-induced tissue overgrowth (Fig. 8). This finding, therefore, may provide further insight into the mechanisms by which the HDAC inhibitors inhibit tumor growth.
Materials and Methods
Drosophila stocks and genetic crosses
The flies were cultured on standard cornmeal, sucrose and yeast agar medium at 25°C unless specified. The following fly lines were used in this study: actin-Gal4, GMR-Gal4, engrailed-Gal4 (en-Gal4), eyeless-Gal4 (ey-Gal4), OK107-Gal4, pumpless-Gal4 (ppl-Gal4) (Zinke et al., 1999), hs-FLP, [Act5C <y+ <Gal4][UAS-GFP(nls)]/cyo, UAS-PI3K (BL#8286), UAS-InR (BL#8262), UAS-S6K (BL#6910) and UAS-PI3KDN (BL#8289). w1118 was used as the wild-type (WT) control. Detailed genetic information about these stocks can be found in FlyBase.
We used hsFLP/FRT-mediated mitotic recombination (Xu and Rubin, 1993) to generate the desired mosaic clones in the fat bodies of flies as previously described (Manfruelli et al., 1999). Female transgenic flies were crossed with male flies (yw1118; hs-FLP; [Act5C <y+ <Gal4] [UAS-GFP(nls)]/cyo) and cultured at 25°C. The 2∼6-hour embryos were collected and heat-shocked at 37°C for 1 hour. The embryos were then transferred to 25°C and cultured for several days until they reached the 3rd larval stage. The larvae with GFP fluorescence signals were selected and dissected.
Constructs and production of transgenic flies
Transgenic flies were produced using the Gal4/UAS system (Rubin and Spradling, 1982). To produce the Hdac3 and MOF RNAi constructs, ∼500 bp of the coding sequences of the genes were amplified by RT-PCR and subcloned into the Sym-pUAST vector, which has two inverted UAS sequence elements (Giordano et al., 2002). This construct allows the inserts to be transcribed bi-directionally, thereby producing dsRNA that silences endogenous gene targets.
The full-length cDNAs encoding the MOF and Hdac3 genes were amplified by RT-PCR and cloned into the pUAST vector to generate the MOF and Hdac3 overexpression vectors. Each vector contains a GFP tag at the C-terminus of the protein.
The full-length cDNAs encoding the histone H4 were amplified by RT-PCR and cloned into the pCRT7/CT-TOPO vector to generate the pCRT7-H4 vector. Starting with the pCRT7-H4 vector, we used a PCR-based site-directed mutagenesis kit (TaKaRa, D401) to introduce the H4K16 point mutation, Lys (K16) to Ala (K16A). The mutant H4 was subcloned into the pUAST vector to generate the mutant H4K16A overexpression vector, which contains a GFP tag at the C-terminus of the protein.
All of the constructs were confirmed by double-stranded DNA sequencing. The PCR primer sequences are provided in supplementary material Table S2. The purified plasmid was injected into w1118 embryos together with a helper plasmid that expressed transposase to generate transgenic flies according to the standard germline transformation procedure.
Total RNA was isolated using the TRIzol reagent according to the manufacturer's instructions (Invitrogen). Complementary DNA (cDNA) was reverse transcribed using the M-MLV reverse transcriptase enzyme (Promega). RealMasterMix (SYBR Green; Tiangen FP202) was used in the subsequent real-time PCR analyses. MRNA (mRNA) levels were normalized against the housekeeping gene Rp49. All of the experiments were repeated at least three times. The PCR primer sequences are provided in the supplementary material Table S2. Quantitative real-time PCR (qRT-PCR) was performed using a BioRad iQ5 instrument.
Immunofluorescence analysis in larval discs
The desired discs were dissected in PBS and fixed in 4% paraformaldehyde in TPBS (0.1% Triton X-100 in PBS) for 30 minutes at room temperature. After being washed with TPBS three times (10 minutes/wash), the discs were incubated with blocking solution (5% normal goat serum in PBS/0.3% Triton X-100) for 1 hour and then with the primary antibodies diluted in blocking solution at 4°C overnight. The discs were washed in TPBS and incubated with the secondary antibodies at room temperature for 2 hours. The images were acquired using a TCS SP5 confocal microscope (Leica). The following primary antibodies were used: anti-H4K16ac (Abcam, 1∶200), anti-H4K12ac (Upstate, 1∶100), anti-phospho S505 dAkt (Cell Signaling, 1∶200) and anti-Dlg (DSHB, 1∶50). The following secondary antibodies were used: Cy3-conjugated goat anti-rabbit and Cy5-conjugated goat anti-mouse (Jackson ImmunoResearch 1∶200).
Western blot analysis
We dissected the desired discs from 3rd-instar larvae and transferred the isolated discs into RIPA lysis buffer (20 mM Tris-Cl, pH 7.5, 150 mM NaCl, 1 mM EDTA, 1% Triton X-100, 50 mM NaF, 1 mM Na3VO4 and 10 mM sodium butyrate) containing a protease inhibitor cocktail (Sigma, P2714). After being incubated on ice for 30 minutes, the lysate was subjected to a 20-second sonication with a Branson digital sonicator at 30% output. The protein concentrations of the lysates were quantified using a BCA Protein Assay Kit (Novagen, 71285-3). Total protein was loaded onto a 15% sodium dodecyl sulfate polyacrylamide gel, and the separated proteins were transferred to a polyvinylidene fluoride membrane. We used primary antibodies against histone H4K16ac (Abcam, ab61240), histone H4K12ac (Upstate) and histone H3 (Abcam, ab1791) at a dilution of 1∶1000. HRP-conjugated secondary antibodies (Dako A/S, Glostrup, Denmark) were diluted (1∶2000) in blocking buffer. The signals were detected using an ECL western blot detection kit. The data are from one experiment that is representative of three independent experiments.
Measurements and data analysis
For the analysis of adult weight, the flies were separated according to sex and placed on normal fly food for 3 days before being weighed. The precision scale analytical balance used to weigh the flies was accurate to ±0.1 mg (Sartorius). A total of 100 flies were used in each group. At least three experiments were performed for each assay.
For the analysis of cell size and number, fly eyes were imaged on a Quanta 200 environmental scanning electron microscope (SEM). The SEM images were analyzed to characterize the eye phenotypes. The eye area was measured using the Image-Pro Plus 6.0 software. The mean cell area was determined by counting the number of ommatidia in a specific area (0.13 mm×0.13 mm). The cell number was determined by dividing the total eye area by the calculated ommatidia area.
ANOVAs and Student's t-tests were used for the statistical analyses. The data are expressed as the means ± standard deviation (s.d.). The threshold value P<0.05 was considered to be statistically significant (*, P<0.05; **, P<0.01).
Microarray assay and data analysis
Total RNA was isolated from Drosophila wing imaginal discs using the TRIzol reagent according to the manufacturer's protocol. The microarray experiments were performed at the CapitalBio Corporation using the Affymetrix GeneChip Drosophila Genome Array (Affymetrix, Santa Clara, CA, USA) according to the manufacturer's protocol.
The expression profiling data were analyzed by the CapitalBio Corporation (Irizarry et al., 2003). A probe set with intensities that were higher than 50 on at least three arrays was selected and subjected to hierarchical cluster analysis. In the unsupervised analysis, the dataset was processed with a variation filter to eliminate probe sets with hybridization signal intensities that did not show large variability. The hierarchical cluster analyses (uncentered correlation and average linkage) were performed using log2-transformed data in which the median value of each probe set was zero, and the data were viewed with the algorithms in the software package R bioconductor. The data have been deposited in NCBI's Gene Expression Omnibus and can be accessed through GEO Series accession number GSE38552.
We thank the Bloomington Stock Center for providing fly stocks, and the Developmental Studies Hybridoma Bank (DSHB) for antibodies. The authors declare that they have no competing financial interests.
This work was supported by national ‘973’ grants from the Ministry of Science and Technology [grant numbers 2011CB965300, 2009CB825603, 2007CB948101], a national grant ‘Jie-Chu-Qing-Nian-Ke-Xue Fund’ from the Chinese National Science Foundation [grant number 30625017] and Tsinghua University Initiative Scientific Research Program [grant number 09THZ03071].