Histone deacetylase (Rpd3) regulates Drosophila early brain development via regulation of Tailless

Tailless is a committed transcriptional repressor and principal regulator of the brain and eye development in Drosophila. Rpd3, the histone deacetylase, is an established repressor that interacts with co-repressors like Sin3a, Prospero, Brakeless and Atrophin. This study aims at deciphering the role of Rpd3 in embryonic segmentation and larval brain development in Drosophila. It delineates the mechanism of Tailless regulation by Rpd3, along with its interacting partners. There was a significant reduction in Tailless in Rpd3 heteroallelic mutant embryos, substantiating that Rpd3 is indispensable for the normal Tailless expression. The expression of the primary readout, Tailless was correlative to the expression of the neural cell adhesion molecule homologue, Fascilin2 (Fas2). Rpd3 also aids in the proper development of the mushroom body. Both Tailless and Fas2 expression are reported to be antagonistic to the epidermal growth factor receptor (EGFR) expression. The decrease in Tailless and Fas2 expression highlights that EGFR is upregulated in the larval mutants, hindering brain development. This study outlines the axis comprising Rpd3, dEGFR, Tailless and Fas2, which interact to fine-tune the early segmentation and larval brain development. Therefore, Rpd3 along with Tailless has immense significance in early embryogenesis and development of the larval brain.


Introduction
Gap genes are the first subset of the zygotic genes in the Drosophila embryo that is activated under the influence of the maternal factors-Bicoid and Caudal. Hunchback (hb) and Tailless (tll) constitute the primary gap genes, solely dependent upon maternal genes while the secondary gap genes, Krüppel (Kr) and giant (gt), perceive cues from the primary gap genes along with the maternal genes. Until the cellular blastoderm stage is reached, Tailless migrates from the nucleus to the cytoplasm. Subsequently, Tailless expresses only in the anterior embryonic region, forming the prospective brain: serving both as a repressor (as in the case of Kruppel) [1] and as an activator (in the case of Hunchback) [2].
The co-repressors, Heat shock factor (Hsf ) and Tramtrack69 (Ttk69), together with GAGA factor (GAF) comprise the dual transcriptional switch that represses tll transcription by binding to the tor response element (tor-RE) at the tll cis-regulatory region. This repression of tll is relieved when the Torso (Tor) signalling pathway activates Mapk downstream that converts Hsf to an activator and degrades Ttk69 by phosphorylation [3]. Tailless represses genes from being activated by the epidermal growth factor receptor (EGFR) signalling and/or activates genes that block the reception or execution of such cues. Tailless ensures that the precursor cells achieve the optic lobe fate while a steep increase in EGFR shifts the fate towards Bolwig's organ formation [4]. tll knockdown in the optic lobe disrupts neurogenesis, producing defective optic ganglia [5].
Rpd3 is a maternally expressed histone deacetylase (HDAC) [6] that counteracts the spreading of heterochromatin [7]. An Rpd3 null mutation enhances position effect variegation of the white gene in the Drosophila eye, promoting silencing which suggests that a wild-type Rpd3 counteract silencing [8]. However, there have been no previous reports about the direct regulation of gap genes by the histone deacetylase, Rpd3. Tailless is regulated by a few well-established co-repressors, like Atrophin and Brakeless [9]. Atrophin (also known as Grunge; Gug) interacts with the histone deacetylase, HDAC1 [10,11], while Brakeless (also known as scribbler; sbb) interacts with Tailless via the DNA-binding domain [9]. Tailless is also regulated by factors like Ttk69 and Hsf [3]. Prospero interacts with Rpd3 to regulate the targeting of the dendrites of the olfactory projection neurons [12]. Polycomb protein, E(Z) and Rpd3 reportedly bind to polycomb response element (PRE) of Ultrabithorax(Ubx), thereby deacetylating histones and silencing gene expression [13].
The EGFR or dEGFR (EGFR in Drosophila) activity in Drosophila initiates neurogenesis in the optic lobe to produce neuroblasts [14,15] and control the population and survival of neuroectodermal progenitor cells [16]. EGFR interacts with Jun N-terminal kinase-mitogen-activated protein kinase (JNK-MAP) kinase affecting the mushroom body development [17]. During early development, it has prominent roles of patterning the ventral epidermis [18] and providing survival signals to avert apoptosis [19]. The EGFR signalling is inhibited in the imaginal discs and eyes during early neural development by a human neural cell adhesion molecule (NCAM) orthologue, Fasciclin 2 (Fas2) [20]. EGFR is hyperactivated in the eye, notum and wing in loss-of-function mutants of fas2 [20]. EGFR inhibition occurs when it is degraded by Anterior open (Aop), also known as Yan. When Fas2 is downregulated, it results in EGFR upregulation and thereby promotes cell proliferation. Alternatively, EGFR is inhibited when the EGFR ligand, Spitz fails to bind to EGFR as it is sequestered by Argos (Aos) [20]. Aos inhibits cellular differentiation while Anterior open triggers differentiation and division in embryos [21]. When Fas2 is downregulated, it leads to the Aop degradation and EGFR activation [20].
Here, we found a new role of Rpd3 in regulating the expression of gap genes, especially tailless, analysed both quantitatively and phenotypically. The embryonic heteroallelic mutants for Rpd3 showed a severe loss of Tailless expression. This finding was further supported by the increased enrichment of the heterochromatic markers in the promoter region of the tailless, in the case of the Rpd3 heteroallelic mutants. Previous reports have shown HDACs responsible for the progression of neurodegenerative diseases. There are pieces of evidence of memory amelioration in mouse models [22] and polyglutamine-dependent neurodegeneration correction in Drosophila [23] upon HDAC inhibitor treatments. We, therefore, hypothesized that the Rpd3 heteroallelic mutations change the expression of Tailless in the larval mutant brains. A significant decrease in the expression of Tailless was observed in the Rpd3 heteroallelic mutant larval brains, similar to that in the embryos. Because Rpd3 has various interacting partners, we also attempted to find out whether the change in Tailless was owing to Rpd3 alone or owing to the interacting partners. We investigated the expression of Tailless in the larval brains for the mutants of Rpd3 as well as the interacting partners of Rpd3 and Tailless. We showed that there was a severe decrease in Tailless in the interacting genotypes (both Rpd3 and the individual interacting partners) compared to the individual mutants. This signifies that Rpd3 was necessary and essential for the normal expression of Tailless.
Tailless, the primary readout for the study, reportedly coexpress with Fas2 in the lamina precursor cells with a lateral high and medial low expression in the outer proliferation centres of the optic lobe [5]. Hence, we also attempted to study the expression of Fas2 in the mushroom bodies of the Rpd3 heteroallelic mutants. In Drosophila, the mushroom bodies or corpora pedunculata constitute the seat of olfactory learning and memory [24,25]. The decrease in Tailless was found to be in unison with the decrease in Fas2 in the Rpd3 heteroallelic mutant. Changes like decrease in embryo volume and alteration in the larval ventral epidermal patterning of the Rpd3 mutant further justified the key role of the EGFR pathway in regulating early segmentation and brain development. This pioneering study thus highlights the importance of Rpd3 in the regulation of Tailless and establishes a significant insight that normal functioning of Rpd3 and Tailless are critical for the Drosophila early brain development under the influence of the EGFR pathway.

Fly strains and the genetic crosses
All the fly strains were reared with standard Drosophila culture media at the optimal temperature of 25°C. The exact genotype of each fly is described in the FLYBASE unless noted elsewhere. The genotypes of each fly strain are as follows: Canton-S (CS) (referred to as wild-type), Rpd3(15-1)/ TKG, Rpd3N(null)/TKG, etc.
For further analysis, the different mutants of the interacting partners of Rpd3 and Tailless were procured from the Bloomington Drosophila Stock Center (BDSC), Indiana, having the BDSC numbers 12 772, 11 615, 12 350, 30 715, 4163, 5491, 5458, 1004 and 3101. The stocks, Pc XT109 and ph del stocks, are null mutants (gift from Dr Giacomo Cavalli, Institute of Human Genetics, CNRS and the University of Montpellier, Montpellier, France).

Embryo collection
The allele, Rpd3 (15-1), is a loss-of-function allele with viable and fertile individuals [8,26]. The Rpd3 null allele or Rpd3N that was considered for this study is lethal in the larval stage when present in a homozygous condition [26]. To generate escaper Rpd3 heteroallelic mutant, virgin females of the Rpd3 null mutation with green fluorescent protein (GFP) balancer marker, Rpd3N/TKG were crossed with an equal number of males of the genotype-Rpd3(15-1)/TKG carrying different Rpd3 alleles in the embryos collection cages as cited by Wieschaus & Nüsslein-Volhard [27]. The embryos aged between 15 and 120 min after egg laying, from crossed parental flies were considered to be well synchronized. These eggs were harvested on the agar plates at 24°C. Embryos without a balancer marker (here, GFP for TKG) constituted the escaper populations and were selected under a ultraviolet (UV) binocular microscope. These embryos were collected as heteroallelic mutants between the two different alleles of Rpd3.

Fixation of the collected embryos
The Rpd3 heteroallelic embryos without GFP expression (constituting the escaper mutants) were fixed in 50% hypochlorite solution for 4-5 min, which removed the outer chorionic layer and made the embryos afloat. These embryos were further rinsed in a solution of PBS-0.5% Triton X-100 and washed in distilled water two to three times. The embryos were then transferred to a 1.5 ml microfuge tube containing a 1 : 1 ratio of n-heptane/PEM (1 mM EGTA, 1 mM MgCl 2 and 0.1 M PIPES)-formaldehyde and incubated for 15 min on a shaker. Later, the embryos were incubated in a fresh microfuge tube containing 500 µl of n-heptane and 500 µl of methanol. The tube was shaken vigorously for 40-50 s until the contents separated into two distinct phases. The upper heptane layer, which contains empty vitelline membranes, was carefully removed. The fixed embryos were used for immunostaining or stored at −20°C for future use, as described in [27].

Embryo immunostaining
An appropriate number of embryos, already fixed and preserved in methanol, were taken in new 1.5 ml microfuge tubes and the methanol was drained out from the tube. One millilitre of fresh methanol was added to it followed by 50 µl of 30% H 2 O 2 , with persistent rocking at room temperature for 20 min. The peroxide mixture was removed and 1 ml of 100% methanol was added to each tube followed by constant rocking at room temperature for 5 min, repeating at least three times. This solution was later replaced by adding 50% methanol with rocking at room temperature for 5 min. The methanol was then removed followed by four to five times rinsing with 1 ml PBT (1× phosphate buffered saline (PBS) with 0.3% Triton X-100) coupled with intermittent rocking at room temperature. The PBT was removed and 1 ml of a mixture of PBT with 2% anti-goat serum was added to each tube (for blocking) at room temperature with constant rocking for an hour. The primary antibodies against Bicoid, Hunchback, Kruppel, Giant and Tailless (from the Asian Distribution Centre for Segmentation Antibodies; 1 : 200) were diluted in the blocking solution. Embryos were transferred to 0.5 ml microfuge tubes. After removing the excess blocking solution, 75-150 µl of the primary antibodies was added to each tube, followed by incubation at room temperature for 1-2 h with persistent rocking at 4°C overnight. After overnight incubation, the primary antibodies were removed and the embryos were transferred to 1.5 ml tubes and rinsed for about eight times with 1 ml PBT (rocking at room temperature). The embryos were blocked with PBT and 2% anti-goat serum (AGS) at room temperature for 1 h with continuous rocking.
After the removal of the blocking solution, 75-150 µl of horseradish peroxidase (HRP)-conjugated secondary antibodies (Sigma Aldrich; 1 : 200) was added and incubated at room temperature for 45 min. Removing the secondary antibody, the embryos were rinsed about eight times with PBT. While the PBT rinsing was in progress, 3,3 0 -diaminobenzidine (DAB) solution was prepared, where two parts of the DAB solution were mixed with one part of 3% NiCl 2 [28]. The embryos were treated with the above solution and rocked for about 10 min at room temperature. The intensity of staining was carefully observed and when the desired intensity was obtained, the reaction was stopped by the addition of PBT. The stained embryos were subjected to different increasing gradients of ethanol to dehydrate, followed by two changes of 100% ethanol after finally clearing in gradients of 50 and 100% Xylene. The immunostained embryos were mounted in the VECTASHIELD Mounting Medium (H-1000) and visualized under a light microscope.

Larval brain immunostaining
The third instar larvae from the various heteroallelic and interacting genotype mutants were screened under a UV filterenabled binocular microscope and the brains were dissected in ice-cold 1× PBS. The tissues were fixed in 4% paraformaldehyde solution in PBS and incubated at room temperature for 20 min. The fixative was then removed and the tissues were rinsed in 0.3% PBT (0.3% Triton X-100 in 1× PBS) three times for 10 min each. The tissues were subsequently incubated in the blocking solution (2% AGS in PBT solution) at 4°C for 1 h. The tissues were then incubated in primary antibodiesmouse anti-Tailless (SantaCruz Biotechnology Inc., B-10; 1 : 50), rabbit anti-Asense (a gift from Dr Yun-Nung-Jan, Howard Hughes Medical Institute, UCSF; 1 : 5000), rabbit anti-Prospero (a gift from Dr Yun-Nung-Jan, Howard Hughes Medical Institute, UCSF; 1 : 1500), mouse anti-Fas2 (Drosophila Studies Hybridoma Bank (DSHB), 1D4; 1 : 50) and mouse anti-Repo (DSHB, 8D12; 1 : 50) overnight. The next day, the primary antibodies were drained off and the tissues were rinsed in PBT for three to five times at 4°C for 10 min each. The secondary antibodies, the donkey anti-rabbit and anti-mouse conjugated to Cyanine-3 (Cy-3) (Jackson Immuno-research Laboratories, Inc., 711-165-152, 715-165-150; 1 : 200), were dissolved in the blocking solution and incubated for 2 h at room temperature. After 2 h, the antibodies were drained off and the tissues were rinsed in PBT for 10 min each, three times. Then the tissues were rinsed in 1× PBS and mounted in the VECTASHIELD mounting media with DAPI (4', 6-diamidino-2-phenylindole) (H-1200). Images were captured using an Olympus FV1000 confocal microscope at lower (10× and 20×) as well as higher magnifications (40×), respectively. All the images are presented as XY figures as whole Z-stacks. All the images were adjusted and assembled in ADOBE PHOTOSHOP CS6.

Western blot hybridization
Nearly 0.2 g embryos were collected in a 1.5 ml Eppendorf tube and homogenized thoroughly in the lysis buffer (50 mM Tris pH 8.0, 1 mM EDTA, 0.5% Triton X-100,1 mM PMSF and 1× protease inhibitor (Roche) (depending upon the number of embryos). This sample was centrifuged at 16 000 rcf at 4°C for 15 min. This supernatant was collected in a new 1.5 ml microfuge tube and stored at −20°C until further use. The protein concentration of the lysates (the supernatants collected previously) was measured using the royalsocietypublishing.org/journal/rsob Open Biol. 10: 200029 Bradford assay and run on 10% SDS-PAGE gel. The protein bands were transferred onto the polyvinylidene difluoride (PVDF) membrane. The PVDF membranes were blocked with a 5% milk solution for 1 h, at room temperature. The blots were then incubated with primary antibodies (Bicoid, Kruppel, Giant, Hunchback and tailless; 1 : 500) and mouse Anti-β Actin antibody (Abcam, 8224; 1 : 2000) at 4°C overnight and then rinsed for 10 min, five times, with TBST buffer (Tris-buffered saline, 0.1% Tween 20) (1500 mM NaCl, 200 mM Tris, pH 7.5, 2% Tween-20). The blots were incubated with secondary antibodies conjugated to HRP for 1 h (Sigma Aldrich; 1 : 3000) at room temperature. Membranes were rinsed for about five times for 10 min each with the TBST buffer. Immunoreactive bands were detected by enhanced chemiluminescence (ECL) as per the protocol described [29].

Chromatin immunoprecipitation
To examine the role of the epigenetic modifiers in the regulation of tailless, the promoter sequences were predicted (since unavailable in the NCBI gene) using a promoter prediction tool, Neural Network Promoter Prediction, developed by the Berkley Drosophila Genome project. Accordingly, the forward and the reverse primers were designed (figure 2a) and the chromatin was isolated for the heteroallelic mutant of Rpd3. The respective DNA was immunoprecipitated against various euchromatic and heterochromatic markers like the H3K9me3, H3K27me3, RNA Pol II, H3K9Ac and H3K4Ac. The immunoprecipitated samples were subjected to a real-time quantitative reverse transcription-polymerase chain reaction (RT-qPCR) to analyse the real-time amplification of the genes for the mutant phenotype after comparison with the 18S rRNA. The data were analysed and the normalized values were plotted on bar graphs to allow genotypewise comparison. The ChIP was performed as per the protocol reported earlier [30,31]. Briefly, 150-200 µl of Drosophila embryos (kept frozen in 1× PBS at −80°C) was thawed and strained in 10 ml homogenization tubes. To it, 5 ml of buffer A1 (60 mM KCl, 15 mM NaCl, 4 mM MgCl 2 ,15 mM HEPES ( pH 7.6), 0.5% Triton X-100, 0.5 mM DTT, 10 mM sodium butyrate, protease inhibitor, Roche (25×)) was added and homogenized followed by centrifugation at 400g for 1 min. The supernatant collected thereafter was re-centrifuged at 1100g for 10 min at 4°C, to obtain a pellet which was resuspended in 5 ml PBS and formaldehyde was added to it to a final concentration of 1% and mixed well. Cross-linking was done at room temperature for 10 min and was quenched by adding glycine to a final concentration of 225 mM. The homogenate was pelleted down at 4000 rpm for 15 min at 4°C. The pellet was washed twice in cold PBS containing 1× protease inhibitor cocktail and finally resuspended in buffer A (5 mM PIPES pH8.0, 85 mM KCl, 1× protease inhibitor cocktail (Roche), 50 U ml −1 SUPERNase) with 0.5% NP-40 and incubated on ice for 10 min. The crude fraction was pelleted by centrifugation at 5000 rpm for 5 min at 4°C and the pellet was washed once with buffer A, resuspended in buffer B (1% SDS, 10 mM EDTA, 50 mM Tris-HCl pH 8.1, 1× Roche protease inhibitor cocktail, 50 U ml −1 SUPERNase) and incubated on ice for 10 min. Lysates were sonicated for 15 cycles (30 s-on 30 s-off) in a Bioruptor ® Plus sonication device (Diagenode SA) to achieve chromatin fragments. The sonicated chromatin was diluted with the ChIP dilution buffer (0.01% SDS, 1.1% Triton X-100, 1.2 mM EDTA, 16.7 mM Tris-HCl pH8.1, 167 mM NaCl, 1× protease inhibitor cocktail (Roche)) and pre-cleared with Protein A-Sepharose beads (GE Healthcare; 17-0780) for 1 h at 4°C. About 200 µg of pre-cleared chromatin was incubated with the antibodies: rabbit anti-H3K9me3 (Upstate, 07-442; 1 : 1000), rabbit anti-H3K27me3 (Upstate, 07-449; 1 : 1000), mouse anti-RNA Polymerase II CTD repeat YSPTSPS (Abcam, ab817; 1 : 1000), rabbit anti-H3K9Ac (Upstate, 06-942; 1 : 1000) and rabbit anti-H3K4Ac (Upstate, 07-539; 1 : 1000) overnight and the samples were incubated with Protein A-Sepharose beads for immunoprecipitation. The immunoprecipitated complexes were washed sequentially in low salt wash buffer (0.1% SDS, 1% Triton X-100, 2 mM EDTA, 20 mM Tris-HCl pH8.1, 150 mM NaCl), high salt wash buffer (0.1% SDS, 1% Triton X-100, 2 mM EDTA, 20 mM Tris-HCl pH 8.1, 500 mM NaCl), LiCl wash buffer (0.25 M LiCl, 1% NP-40, 1% Deoxycholate, 1 mM EDTA, 10 mM Tris-HCl pH8.1) and TE buffer (10 mM Tris-HCl, 1 mM EDTA pH 8.1). The nucleic acid was eluted with elution buffer (1% SDS, 0.1 M Na 2 HCO 3 ) and crosslinks were reversed for 2 h at 65°C in the presence of 200 mM NaCl. Chromatin samples were treated with Proteinase-K and RNA was isolated using the TRIZOL reagent (Invitrogen) following the manufacturer's protocol (Promega). Genomic DNA contamination was removed from immunoprecipitated RNA samples using the TURBO DNA-free kit (Ambion). The samples were amplified following a real-time PCR protocol using specific primers. The sequences of the promoter amplification added were as given below.
royalsocietypublishing.org/journal/rsob Open Biol. 10: 200029 2.8. Reverse transcription-polymerase chain reaction (PCR) and real-time quantitative reverse transcription-PCR The total RNA was extracted from Drosophila embryos by the Trizol method (Invitrogen) and purified by using the RNeasy mini kit (Qiagen). RNA was treated with DNase I (Ambion) for 30 min at 37°C. cDNA was synthesized using the Superscript™ II RT-PCR kit (Invitrogen). Strand-specific reverse transcription-PCR (RT-PCR) was done using the 18S rRNA and other gap gene promoter primer sets. PCR amplification was performed using AmpliTaq Gold system (Promega). Quantitative RT-PCR was carried out using the LightCycler ® 480 System (the Roche Applied Science). The KAPA SYBR ® FAST qPCR Master Mix (2×) Kit (KAPA Biosystems) was used according to the manufacturer's directions. Transcript levels were normalized to 18S rRNA levels. The data were graphically represented as the relative enrichment of the epigenetic markers after comparison with the 18S rRNA and the level of statistical significance for each set of data has been indicated as ratio ± s.e.m of three independent experiments. The Rpd3 heteroallelic larval brains were collected and total RNA was isolated by the Trizol method (Invitrogen). The RNA was purified using the RNeasy mini kit (Qiagen), followed by treatment with DNase1 (Ambion) for 30 min at 37°C. cDNA was synthesized by RT-PCR using the RNA to cDNA EcoDry™ Premix (Oligo dT) (Clontech). Quantitative RT-PCR was run on a CFX96™ Real-Time System with the C1000 Touch™ Thermal Cycler (Bio-Rad) with each PCR mixture containing a 0.5 µl cDNA template and 10 nM primers in 20 µl of KAPA SYBR ® FAST qPCR Master Mix (2×) (KAPA Biosystems). The analyses were done using the Bio-Rad CFX Maestro™ Software. Expression values were normalized to those obtained with the control 18S rRNA. The mRNA was measured using the primers listed as follows: tll: forward primer: (5 0 -ATGAATTTCGCCCAGCTGCT-3 0 ), reverse primer: (5 0 -AGCCCTCAGACACTCGTACT-3 0 ); Rpd3: forward primer: (5 0 -TTAAGCTGCACATTAGTCCCAGC-3 0 ), reverse primer: (5 0 -CTCGGGAATCGCTTGGATTTGAA-3 0 ); 18S rRNA: forward primer: (5 0 -TTGTGCTGAAGAAGGCC-GAT-3 0 ), reverse primer: The data were graphically represented as the relative mRNA fold change compared to the 18S rRNA and the level of statistical significance for each set of data has been indicated as mean ± s.e.m. of three independent experiments (t-test) ***p ≤ 0.001, **p ≤ 0.01 and *p ≤ 0.05 respectively.

The scanning electron microscope imaging of the larvae
Drosophila third instar larvae for the wild-type (Canton-S) and the heteroallelic Rpd3 mutant were collected from the food bottles pre-incubated at 18°C, rinsed in 1× PBS and affixed on the stubs coated with double-sided tape. They were subjected to gold sputtering under the plasma pressure for 15 s and were imaged under the scanning electron microscope (model no. 3400N, Hitachi, Japan) at 5.0 kV acceleration speed and 27× and 150× magnification.

Results
The gene Rpd3 (Reduced potassium dependency 3) encodes the class I histone deacetylase which is a mammalian homologue of the HDAC1 and HDAC2 [8,32,33]. When Rpd3 is knocked down, there is a global increase in telomeric H3 and H4 histone acetylation on the polytene chromosomes [34]. The present study presented a new role of Rpd3 in regulating the expression of gap genes in the embryos and also demonstrated how particular mutations in Rpd3 and its interacting partners affect Tailless expression in the larval brain. To investigate the involvement of the Rpd3, we have used two Rpd3 loss-of-function alleles, Rpd3N and Rpd3(15-1), in heteroallelic combinations.
In Drosophila melanogaster, the Rpd3 gene is cytologically located in the 64C1-2 region of chromosome 3. The Rpd3(15-1) allele has a P-element insertion of 1.8 kb from the 5' end of the Rpd3 transcript in the putative regulatory region [8]. The Rpd3(15-1) loss-of-function allele is viable and fertile [26]. Rpd3N is a null allele, produced by a P-element excision and deletion of approximately 870 bp in rpd3 04556 in the 5 0 -coding region of the gene [26]. The expression of gap genes in the Rpd3 heteroallelic mutant embryos was analysed via immunostaining and quantitative western blot hybridization. The maternal gene, bicoid (bcd), produces a morphogen gradient that dictates the expression of subsequent gap genes in the developing embryos [35]. When the early, age-synchronized heteroallelic Rpd3 mutant embryos were immunostained with the Bicoid antibody, there was no change in the intensity of Bicoid (figure 1a) similar to the expression observed in the quantitative western blot (figure 1f,g). Thus, heteroallelic mutations in Rpd3 embryos do not affect the expression of Bicoid.

Rpd3 loss-of-function does not modulate Hunchback, Kruppel or Giant expression
The gene, hunchback, regulates the trunk formation in Drosophila by establishing an anterior-posterior gradient from the unfertilized egg to the developing zygote [36]. When the Rpd3 heteroallelic mutant embryos were immunostained with Hunchback (Hb) antibody (figure 1b), and the expression was further substantiated by a western blot (figure 1f,g), the expression of Hb did not change, compared to the wild-type. Therefore, Rpd3 does not regulate the expression of Hunchback.
Krüppel expresses in the early stages of the embryos in the form of a central band [37]. In the age-synchronized embryos of Rpd3 heteroallelic mutants, Kruppel showed no change in expression compared to the wild-type, in immunostaining (figure 1c) as well as western blot (figure 1f,g). Hence, Rpd3 does not regulate Kruppel expression.
Giant (gt) expresses in the early embryos producing two broad stripes with a bare central mid-region [38]. When age-synchronized Rpd3 heteroallelic embryos were immunostained with Giant, the expression of Giant showed no change in the mutant compared to the wild-type (figure 1d). The western blot also substantiated the above result (figure 1f,g) and concluded that Rpd3 heteroallelic mutation does not affect the expression of Giant. Depending upon the previous evidence of Rpd3 in brain development, we focused on our subsequent studies to understand how Tailless decreased such severely in the heteroallelic Rpd3 mutants. It led to a proposition that Rpd3 might modulate the expression of gap genes through epigenetic regulation, highlighting the importance of a functional Rpd3-dependent chromatin complex for tll regulation and anterior-posterior axis formation.

The Rpd3 complex epigenetically regulates Tailless expression
The results obtained previously have led to the proposition that changes in Tailless in the Rpd3 heteroallelic mutant were possibly owing to the regulators of the transcriptional machinery. To confirm the involvement of histone-deacetylase associated with Tailless expression in the Rpd3 mutant, an in vivo chromatin immunoprecipitation (ChIP) assay based on the tailless promoter was performed (figure 2a). This study assesses the capacity of each factor of the transcriptional machinery and associated histone tail modifiers to bind with early recognized promoter in vivo in the chromatin [40,41]. The underlying principle states that promoters which have binding sites for the transcriptional machinery undergo    Figure 2. (a) The schematic of the relationship of transcription start site (TSS) and the primer annealing sites for the gene, tailless, used in ChIP assay. The promoterbased primers were designed for the gene, tailless, located on the chromosome 3R in D. melanogaster. The gene sequence shown in the box represents those that are upstream to the coding region of tailless. The 5 0 untranslated region (UTR) of tailless starts from 30 852 315 nucleotide (nt) and the 3 0 UTR ends at 30 854 400 nt. The CDS of tailless begins from 30 852 547 nt with the start codon ATG. The sequence annealing with the forward primer: 5 0 -TGCTGCGTTCCTTGCTCAATG-3 0 has been marked in red and those that anneal with the reverse primer: 5 0 -TCGCTGAGATCACGATGGGTT-3 0 have been shown in purple. The CDS of tailless, beginning with the start codon, ATG has been highlighted in green. (b) The ChIP assay data showing the relative enrichment of the epigenetic markers on the promoter of tailless in the Rpd3 heteroallelic mutant. The ChIP assay was performed to assess the relative enrichment of the epigenetic markers in the promoter region of tailless. The age-synchronized, 2 h after egg laying (AEL) embryos from the wild-type (CS) as well as Rpd3 heteroallelic mutants were collected and the DNA was isolated. It was immunoprecipitated with various markers for heterochromatin formation like H3K9me3 and H3K27me3, and markers for euchromatin formation like H3K9ac and H3K4ac. RNA Pol II which acts as a marker depicting transcriptional activity was also assessed. The relative enrichment of these markers on the tailless promoter was assessed using the real-time PCR, after comparison with the 18s rRNA for both the wild-type (CS) and Rpd3 heteroallelic mutant. Each of the markers has been plotted in the graph and the trends in the variation of these markers represented with error bars corresponding to s.e.m. from three independent experiments. royalsocietypublishing.org/journal/rsob Open Biol. 10: 200029 development, transcription and segmentation [44]. H3K9ac associate with the transcription start sites (TSSs), corresponding to positive gene expression levels [45]. When histones are hyper-acetylated, Pol II starts transcription in the neuroectoderm [46]. Genome-wide ChIP analysis state that in the promoters of actively transcribed genes, H3K4ac localizes upstream to H3K4me3 [47,48]. H3K9ac localizes in the promoter regions of actively transcribed genes and stimulates transcription by recruiting transcription Factor IID (TFIID) [49], while the H3K9me2 and H3K9me3 marks are enriched in the pericentric heterochromatin structures to repress transcription [50,51]. In Drosophila embryos, until cycle 14a, the histone methylation marks, H3K4me1, H3K4me3, H3K36me3 and H3K27me3, and the acetylation mark, H3K9ac, are absent until transcription of the zygotic genes begins. The polycomb-associated mark, H3K27me3, regulates cell-specific silencing of developmental genes [52]; they are completely absent in maternal genes and enriched in the upstream and downstream promoter regions of the cycle 14 embryos [31,53,54].
ChIP assay for Rpd3 heteroallelic mutant compared to the wild-type revealed a significant decrease in tailless. In the wild-type, tailless had an increased enrichment of the euchromatic markers, indicating an active state of the tll gene (figure 2b). In the Rpd3 mutants, tailless had decreased enrichments of H3K9ac, H3K4ac, Pol II and H3K27me3 and an increased enrichment of H3K9me3 at the promoter, compared to the wild-type (figure 2b). Thus, the data established that a heteroallelic loss-of-function of Rpd3 in the embryos, leads to the loss of Tailless owing to the decreased enrichment of H3K9ac, H3K4ac, H3K27me3 and Pol II.

The Tailless expression is reduced in the heteroallelic Rpd3 mutant larval brains
Former results have established the importance of Rpd3 in the Tailless regulation with an evident decrease in the heteroallelic Rpd3 mutant. In the optic lobe neuroblast of the third instar larvae [55,56], Tailless ensures an optimum proliferation and prolonged maintenance of the mushroom body neuroblasts and ganglion mother cells [57]. Hence, tailless knockdown leads to defective optic lobes neurogenesis [5]. To investigate the possible role of Tailless in the Rpd3 mutant on larval brain development, the heteroallelic Rpd3N/Rpd3(15-1) mutant third instar larval brains were immunostained with the Tailless antibody, where Tailless was significantly decreased in the mutant compared to the wild-type ( figure 3a,c,d). When the immunostained Rpd3 heteroallelic mutant larval brains were visualized under higher magnification, there was a significant reduction in the Tailless localization, particularly in the outer proliferation centre (OPC) and inner proliferation centre (IPC) regions, compared to that of the wild-type (electronic supplementary material, figure  S1A). These findings confirm that Rpd3 has a critical role in the regulation of Tailless expression in the larval brain.
The altered enrichment of the epigenetic modifiers in the promoter regions of tailless in the Rpd3 mutant embryos suggested the further investigation of changes in Rpd3 and tailless mRNA expression in the heteroallelic Rpd3 third instar larval brains. The third instar larval mutant brains were dissected, the RNA was isolated and qPCR was performed after cDNA synthesis ( figure 4a,b). There was a significant reduction in the relative expressions of both Rpd3 (figure 4a) and tll mRNAs in the heteroallelic Rpd3 third instar larval brain compared to the wild-type (CS) (figure 4b). This confirmed that the heteroallelic loss-of-function mutation of Rpd3 in the larval brain leads to a decrease in the expression of tailless mRNA.

Rpd3 with its interacting partners regulates the expression of Tailless
The previous result showed a significant reduction in the expression of Tailless in the heteroallelic Rpd3 mutants. Therefore, the subsequent study aimed at investigating whether the change in Tailless was owing to Rpd3 itself or due to the interacting partners of Rpd3 and Tailless. We have immunostained the mutant larval brain to investigate Tailless expression in the interacting partner mutants as well as the interacting genotypes (mutated Rpd3 as well as the mutated interacting partners) (figure 5a). Sin3a is an Rpd3 co-repressor that forms scaffolds for the assembly of the HDAC complex [58,59], while Prospero interacts with Rpd3 and regulates terminal axon branching [12]. We found that in the Sin3a/+ and prospero/+ mutants, as well as in the interacting genotypes (Rpd3N/Rpd3(15-1); Sin3a/+) or (Rpd3N/Rpd3(15-1), pros/+), the expression of Tailless had reduced compared to the wild-type (figure 5a, b). In the interacting genotype (Rpd3N/Rpd3(15-1), Sin3a/+), Tailless was further reduced compared to the wild-type or the Sin3a/+ mutant alone. The scribbler gene (also known as brakeless) genetically interacts with tll and needs to be present as in the wild-type for initiating the repressive activity of Tailless [9]. Tailless expression was almost close to that of the wild-type in the sbbG01610/+ mutant, but in the interacting genotypes (Rpd3N/Rpd3(15-1); sbbG01610/+), there was a distinct reduction compared to the wild-type (figure 5a-f ). Tailless interacts with Atrophin to repress the expression of Knirp [11]. When the larval brain for Atrophin mutants was immunostained, there was a negligible decrease in Tailless compared to the wild-type (figure 5a), but (Rpd3N/Rpd3(15-1), Gug03928/+) showed a significant decrease (figure 5b,e,f ).
Rpd3 executes gene silencing, by binding with the Polycomb protein at the polycomb responsive element (PRE) of the Ultrabithorax gene [13]. The Polycomb repressive complex 1 (PRC1) comprises Polycomb (PC), Posterior sex combs (PSC), polyhomeotic (PH) and Sex comb on midleg (SCM). PC associates with extra sexcombs (ESC) in a complex containing Enhancer of zeste (EZ), pleiohomeotic (PHO), polyhomeotic (PH), GAGA and Rpd3. Both the ESC and PC are crucial for polycomb-dependent gene silencing [60]. Tailless expression decreased only in the interacting genotype (Rpd3N/ Rpd3(15-1), Pc/+) (electronic supplementary material, figure  S2A,B). The Tailless expression was further substantiated by western blot (figure 5c-f) which concluded that the expression of Rpd3 is more crucial than the interacting partners for maintaining a normal Tailless expression in the larval brain.

Fascilin2 expression, like Tailless, is regulated by
Rpd3 and its interacting partners Fas2 is an inhibitor of EGFR signalling in Drosophila brain development [20]. Fas2 co-express with Tailless in the lamina precursor cells of the optic lobe and shows a high and medial low domain of expression in the developing outer proliferation centres [5]. It expresses in the cells forming the primordia for the mushroom body peduncle and medial lobe in the α/β and γ axon lobes [57]. To investigate whether the expression of Fas2 in the third instar larval brain of the heteroallelic Rpd3N/Rpd3(15-1) mutant correlated with Tailless, the Fas2 expression was studied in the mushroom body of the Rpd3 heteroallelic mutant  There was a reduction in the Tailless expression in Rpd3 heteroallelic larval brains compared to the wild-type (CS). (b) The Rpd3 heteroallelic late third instar larval brains were immunolabelled with the Drosophila NCAM homologue, Fas2, which localize in the mushroom bodies (MB) in the larval brain (red) and are Z-projections of the XY plane. A reduction in expression of Fas2 was observed in the Rpd3 heteroallelic mutant larval brains. Scale bar, 100 µm. (c,d) The quantitative estimation of Tailless and Fas2 in the heteroallelic Rpd3 mutant larval brains was performed by western blotting and normalized with β-actin. The data have been represented as the mean ± s.e.m. from three independent experiments. There was a decrease in the expression of Tailless and Fas2 compared to the wild-type (CS), which was also graphically illustrated. The mean relative protein band intensity was normalized with the internal control, β-actin (in the lower panel) band intensity, and represented with error bars corresponding to s.e.m. from three independent experiments in triplicate; ***p ≤ 0.001 (n = 3).
royalsocietypublishing.org/journal/rsob Open Biol. 10: 200029 larval brains. The Fas2 expression in the Rpd3 heteroallelic mutant brain decreased compared to the wild-type (figure 3b-d). At a higher magnification, the mushroom bodies of the Rpd3 heteroallelic mutants showed a significant decrease in Fas2 with poorly developed lobes, compared to the wild-type (electronic supplementary material, figure S1B).

The epidermal growth factor receptor negative regulators and Rpd3 control the expression of Tailless and Fas2
The EGFR pathway is a critical regulator of the Drosophila optic lobe neurogenesis and drives the sequential progression of the proneural wave across the outer proliferation centres of the optic lobe [14,15]. EGFR expression is regulated by negative regulators, Aos and Aop. Aos regulates the Ras/MAPK pathway by binding to the epidermal growth factor while Aop inactivates the EGFR expression by negatively regulating transcription [62]. Tailless co-express with Fas2 in the optic lobes [5] and also antagonizes EGFR during the optic lobe formation [4] such that Fas2 decreases whenever there is a knockdown of tailless [5]. Because EGFR expression is known to be antagonized by Tailless [4], the western hybridization was performed to find out the quantitative changes in the EGFR expression in the Rpd3 heteroallelic mutant larval brain. There was an increase in the EGFR expression in the Rpd3 heteroallelic larval mutant brains compared to that of the wild-type (figure 9e,  royalsocietypublishing.org/journal/rsob Open Biol. 10: 200029

Rpd3 regulates the expression of different genes in neural development
In Drosophila, the early neural development is controlled by genes like asense (ase), prospero ( pros) and reverse polarity (repo). The former observations made from this study have established that Tailless and Fas2 expressions are regulated by Rpd3. But no studies have been conducted yet to find out how these expressions are getting altered in the different mutants for Rpd3. The pan-neural and proneural gene, asense, is a mitotic regulator in the larval optic lobes whose loss-offunction increases proliferation and ectopic expression reduces proliferation [63]. The loss of function of prospero results in aberrant expression of cell cycle regulatory genes with premature cell cycle termination [64]. Repo is a glial-specific homeodomain protein [65] that expresses in the late glial cells [66,67]. The tll+ neuroblasts simultaneously express the gene, glial cell missing (gcm). The neuroblast progenies turn on the expression of Repo after tll is shut off. As a result of this, the progeny cells migrate towards the deeper neuronal   Figure 6. The variation in the expression of Tailless in interacting genotypes for Rpd3 along with Hsf and ttk69. (a,b) Tailless is under the control of the modulators-Heat shock factor (Hsf ) and Tramtrack69 (Ttk69). Interacting genotype mutants were obtained from crosses set between Rpd3 heteroallelic progeny and the interacting partners, Hsf and ttk69. The interacting genotype late third instar larval brains were stained with Tailless (in red) to assess the localization of Tailless in these mutant brains and procured from Z-projections using a confocal microscope. There was a loss of expression of Tailless in the interacting genotypes (Rpd3N/ Rpd3(15-1), ttk69/+). When Hsf alone is mutated, there is no significant variation in the expression of Tailless royalsocietypublishing.org/journal/rsob Open Biol. 10: 200029 (e) ( f) Figure 7. The variation in the expression of Fas2 in the mutant brains of the different interacting partners of Rpd3: (a,b) Crosses were set to obtain interacting genotypes between Rpd3 and these interacting partners. The interacting genotype and individual mutant late third instar larval brains were immunostained with Fas2 antibody (red) to find out a variation in Fas2 expression. The images have been procured from Z-projections of the XY plane using a confocal microscope. Fas2 expression was further diminished in the heteroallelic mutants for Rpd3 and also in the interacting genotypes for Rpd3 and the interacting partners, Sin3A, prospero, scribbler and Atrophin. (c,d ) The quantitative estimation of Fas2 in the late third instar larval brain ( particularly in the mushroom bodies) were analysed in the interacting partners, Sin3a, prospero, Scribbler using western blot. The data were graphically represented as the mean relative intensity of protein band, after normalization of the internal control, β-actin, with error bars corresponding to s.e.m.; and ***p ≤ 0.001. The data are a representation of three independent experiments. (e,f ) The western blot quantitative estimation of Fas2 in the late third instar larval brain ( particularly in the mushroom bodies) was analysed in the different interacting genotypes and graphically represented as the mean relative intensity of protein band, after normalization of the internal control, β-actin, with error bars corresponding to s.e.m.; and ***p ≤ 0.001. Data are a representation of three independent experiments.

R p d 3 N / R p d 3 ( 1 5 -1 ) R p d 3 N / R p d 3 ( 1 5 -1 ) ; S i n 3 a / + R p d 3 N / R p d 3 ( 1 5 -1 ) , p r o s / +
layers and take up their final positions as glial cells around the medulla neuropils [68]. To investigate the changes in the expression of Asense, Prospero and Repo in the Rpd3 mutant, larval brain immunostaining and western hybridization were performed. In the Rpd3 heteroallelic mutant larval brains, the expressions of Asense and Repo decreased ( figure 10a,c), while the expression of Prospero increased (figure 10b) compared to the wild-type. Results obtained from the western hybridization also substantiated that compared to wild-type, the expression of Asense and Repo were reduced ( figure 10d,e), while the expression of Prospero was increased (figure 10d,e) which concluded that Rpd3 regulates the expression of all the above neural development genes.
3.9. Fas2 expression gets altered in the eye disc for mutants of Rpd3, aop and aos   royalsocietypublishing.org/journal/rsob Open Biol. 10: 200029 anterior to the morphogenetic furrow but mainly localizes at the furrow, decreasing further to the posterior of the disc (figure 11a). Fas2 promotes EGFR signalling during the growth of the imaginal disc [69] that causes cell-specific expression of Fas2, suggesting a self-regulatory feedback loop between Fas2 and EGFR during eye disc development. Hence, the loss of fas2 is reported to reduce the growth and proliferation of eye disc cells [69]. Aop repress EGFR and is degraded upon the Fas2 downregulation [20]. Aos negatively regulates the Ras/MAPK pathway that inhibits EGFR by   Rpd3. (a,b) The expression of the EGFR pathway is negatively regulated by the repressors, Argos and Anterior open. Crosses were set between Rpd3 and the negative regulators, aos and aop, and the late third instar larval progeny brains were immunostained with Tailless (in red) to find out the localization. The images are all Z-projections of confocal sections. The expression of Tailless was found to decrease in the aos/+ and yan/+ and interacting genotype brains of (Rpd3N/Rpd3(15-1), aos/+) and (Rpd3N/Rpd3(15-1); aop/+). Scale bar, 100 µm. (c,d ) The third instar progenies of the individual mutants for aos and aop and the interacting genotypes were immunostained with Fas2 (in red) and Z-projections of the stained brains were visualized to find out the localization. In the interacting genotypes for the interacting partners of EGFR-aop and aos, the expression of Fas2 showed reduction compared to the wild-type (here CS). Scale bar, 100 µm. (e,f ) The quantitative western blot analysis was performed to compare the expression of dEGFR in the heteroallelic larval mutant brains for the histone deacetylase, Rpd3. The expression of dEGFR was quantitated against the expression of the internal control β-actin (in the lower panel) and graphically represented, with error bars corresponding to s.e.m. with ***p ≤ 0.001. The data are a representation of three independent experiments. There was an increase in the expression of dEGFR in the lysates obtained from the larval mutant brain of Rpd3, compared to the wild-type (CS).
royalsocietypublishing.org/journal/rsob Open Biol. preventing binding to Spitz [70]. When the expression of Fas2 was assessed in the eye disc for the mutants of aos and aop, as well as the trans-heterozygous mutants, Fas2 was found to decrease in the aos/+ mutant and interacting genotypes  In zebrafish central nervous system, the histone deacetylase, hdac1, is critical for embryonic neurogenesis. A mutation in hdac1 results in the reduction in the midbrain and anterior hindbrain size and produces a dorsally open spinal cord [71]. To investigate whether Rpd3 assists in maintaining the normal size and volume of the Drosophila embryo, the width, length and volume of age-synchronized early developing embryos were analysed (electronic supplementary material, figure S3 and table S1). Both the width and length of the Rpd3 heteroallelic embryos were found to vary by nearly 41-23% relative to the wild-type embryos. Lengths ranged from 0.57 mm in embryos from wild-type flies to 0.54 mm in Rpd3N/ Rpd3(15-1) (electronic supplementary material, table S1). This showed that histone deacetylase results in the maintenance of embryonic size (electronic supplementary material, figure S3) and that Rpd3 regulate the size of the embryos. The Drosophila larvae patterning was studied by examining the ventral epidermis. The ventral epidermal cells secrete short, stout hair-like projections called denticles, under the influence of the EGFR pathway. These exist in the repetitive belts of rows at the anterior half of each parasegment, along the anteroposterior axis. The antagonism between the EGFR and Wingless signalling decides the basic cell fate of the alternating belts of denticle and smooth cuticle [72,73]. To find out if a mutation in Rpd3 altered the ventral cuticular patterning, the Rpd3 heteroallelic third instar mutant larvae were examined at 27× and 150× magnifications, respectively. The first four rows of denticles were found to be prominently developed in the Rpd3 heteroallelic mutant larvae, compared to the wild-type (electronic supplementary material, figure S4). The result signified that Rpd3 mutation produces moderate changes in the larval denticle patterning.

Discussion
The present study deals with the regulation of Tailless by Rpd3 during early embryonic development and the effect of Rpd3 and the interacting partners on Tailless expression in the Drosophila larval brain.
The histone deacetylase, dHDAC4, is a known regulator of the Gap and Pair-Rule genes in Drosophila [74] that made us investigate for a possible role of the Drosophila histone deacetylase, Rpd3 in the gap genes regulation. Rpd3 has reported roles of repression with the interacting partners (co-repressors) Sin3a [75], prospero [12] and Atrophin [76]. The initial studies involving immunostaining and western hybridization investigated the changes in the gap gene expression in the Rpd3 heterollelic embryos and showed a complete loss of Tailless, signifying that Rpd3 has an important role to execute in the maintenance of normal expression of Tailless. However, Rpd3 did not affect the expression of Kruppel and Giant probably because these genes receive inputs from both the maternal genes (mainly bicoid) and primary gap genes to sustain their normal expression. Hunchback is present both maternally and zygotically (regulated by inputs from Bicoid). Because Rpd3 did not affect the Bicoid expression, there was no change in the expression of Hunchback, Kruppel and Giant in the Rpd3 heteroallelic mutants (figure 1).
The ChIP assay was performed to ascertain the enrichment of the common epigenetic markers (H3K9me3, H3K27me3, RNA Pol II, H3K9ac and H3K4ac) in the promoter region of tailless. H3K27 is highly enriched and acetylated at high levels in the early embryos and there is an increased enrichment of methylated H3K27 only after 4 h after egg laying [52]. H3K27 undergoes acetylation by CREB-binding protein (CBP) and deacetylation by Rpd3 [77]. There was a decreased enrichment of H3K27me3 in the Rpd3 mutant embryos compared to the wild-type (figure 2b). Because deacetylation, which involves the removal of an acetyl group from the histone tail of H3K27 requires Rpd3, the loss-offunction of Rpd3 may have resulted in a reduced enrichment of H3K27me3. A severely decreased enrichment of euchromatic markers, H3K9ac and H3K4ac along with a very significant low enrichment of RNA polymerase II, was observed at the promoter region of tailless (figure 2b). The occupancy of RNA Pol II at the promoter of a gene primes genes for getting activated during early development [78]. H3K9ac prevents the RNA Pol II from escaping promoterproximal pausing on chromatin, and H3K4ac mainly localize and enrich in the promoters of the active genes [47,48]. The high degree of reduction in the RNA Pol II in the tailless promoter (figure 2b) accounts for the decrease in the expression that was evident from the immunostaining as well as the western blot results ( figure 1). Thus, the above study established the importance of Rpd3 in the Tailless regulation in embryos.
Besides forming terminal domains during Drosophila segmentation [79], Tailless also regulates early brain development. It serves as a dedicated transcriptional repressor in early Drosophila embryos [80]. Tailless reportedly interacts with Atonal (ato), Sine oculis (so), Eyeless (ey) and Eyes absent (eya), and the EGFR signalling pathway to establish the Drosophila embryonic visual system [4]. Therefore, Tailless was chosen to serve as readout for analysing the mechanism of regulation employed by Rpd3 during larval brain development. The loss of Tailless expression in the Rpd3 heteroallelic larvae (figure 3a,c,d) showed that it has a critical role in larval brain development. The significant decrease in the expression of Rpd3 and tailless mRNAs in the Rpd3 heteroallelic third instar larval brains compared to the wild-type (CS) demonstrated that Rpd3 heteroallelic mutation decreases the expression of tailless mRNA ( figure 4a,b).
The interacting partners of Rpd3 include Sin3a, prospero, Atrophin, scribbler and polycomb proteins. The expression of Tailless decreased in the larval brain of Sin3a/+ and (Rpd3N/ Rpd3(15-1), Sin3a/+) mutants (figure 5a-f ). Sin3a is a corepressor of Rpd3; however, upon the loss-of-function mutation of Sin3a, there was a decrease in the expression of Tailless. This raises the possibility that the other interacting partners are more effective in maintaining the repressed state of tailless, even in the absence of Sin3a. Atrophin interacts with Rpd3 via H3 acetylation on the N terminus forming a trimeric complex that reduces decapentaplegic transcription [76]. Atrophin co-represses Even-skipped [11,81] and genetically interacts with Tailless, in addition to the physical interaction the ligand-binding domain of Tailless. A higher decrease in Tailless in the interacting genotypes (Rpd3N/ Rpd3(15-1), Gug03928/+) signified that Atrophin (figure 5b,e, f ) prevented Tailless expression by hindering DNA-binding and transcription. Scribbler, another transcriptional repressor of Tailless, was found to repress Tailless expression in the interacting genotypes (Rpd3N/Rpd3(15-1); sbbG01610/+) (figure 5a-f ) which suggests that Rpd3 plays a crucial role in Tailless expression.
Ttk69 represses Tailless [61] by binding to the 3' flanking region of tor-RE. Hsf activates Tll by getting phosphorylated at S378, transforming from a repressor to an activator and, in this process, relieves the repression of Tailless [82]. The larval mutants where both Rpd3 and Hsf were mutated (Rpd3N/ Rpd3(15-1); Hsf/+) showed no changes in Tll expression (figure 6a-f ) but in (Rpd3N/Rpd3(15-1), ttk69/+), there was a decrease in Tll ( figure 6a-f ). The decreased expression of Tll in (Rpd3N/Rpd3(15-1), ttk69/+) mutants can be attributed to the failure of the GAF/Hsf/Ttk69 complex formation, owing to the absence of Rpd3. The failure to form the complex between Ttk69, Sin3a and pits ( protein interacting with Ttk69 and Sin3a) protein simultaneously results in the loss of Tailless. This highlighted the necessity of both Rpd3 and Ttk69 to be present simultaneously for the normal Tailless expression in the Drosophila larval brain. The finding that Tailless along with Rpd3 is essential for normal brain development was thus confirmed and further reinforced by ruling out the roles of the interacting partners of Rpd3 and Tailless.
In the Rpd3 heteroallelic mutant, the expression of Fas2 has reduced in comparison to its wild-type counterpart ( figure 3b-d). In all the interacting genotypes for the partners of Rpd3 or Tailless, Fas2 expression was found to be completely dependent on the presence of Rpd3. There were evident similarities in the expression of Fas2 as well as Tailless in the respective interacting genotype of Rpd3 and the interacting partners (figures 7a-f and 8a-f ). The known fact that both Fas2 and Tailless co-localize in the larval brain could justify such similarities.
Aos and Aop are negative regulators of EGFR which gets activated when EGFR production needs to be shut down. Whenever these negative regulators are mutated, EGFR levels increase. There was a significant decrease in Tailless in the larval brains of the aop/+ (figure 9a) and (Rpd3N/ Rpd3(15-1); aop/+) mutants (figure 9b) as well as in the larval brains of the aos/+ (figure 9a) and the (Rpd3N/Rpd3(15-1), aos/+) mutants (figure 9b). Fas2 decreased almost equally in the larval brains for all the mutants of Aop and Aos compared to wild-type (figure 9c,d). Previous reports suggest that during optic lobe formation, EGFR signalling is antagonized by Tailless [4]. Therefore, the decrease in Tailless expression in the mutants for Rpd3 as well as aop and aos may contribute to the EGFR upregulation. Fas2 interacts with EGFR genetically and inhibits EGFR for normal eye development. The loss-offunction of fas2 degrades Aop to cause EGFR hyperactivation in the eye, notum and wing [20]. In the mutants for aos and aop, Fas2 decrease may contribute to the increase in EGFR, correlating with the Tailless expression in these mutants. Therefore, to assess this, we checked for EGFR protein levels in the Rpd3 mutant larval brains. The western blot showed an upregulation of the dEGFR expression in the Rpd3 heteroallelic larval mutant compared to wild-type (figure 9e,f ) which substantiated the above results and concluded that in the Rpd3 heteroallelic larval mutant brains, a robust decrease in the expression of Tailless and Fas2 increased the expression of dEGFR.
EGFR is crucial for neurogenesis in Drosophila and generates the proneural wave as a downstream result of EGFR activation [83]. The proneural gene then initiates the expression of the pan-neural genes-asense [84,85] and Prospero [86,87]. Asense is a pan-neural homeobox gene that is activated by the proneural genes when EGFR is downregulated. It upregulates the neuroblast genes and represses the neuronal differentiation genes [88]. The expression of Asense was found to be downregulated in the Rpd3 heteroallelic mutant larval brain which signified that EGFR was upregulated in this mutant ( figure 10a,d,e). Prospero is a neuronal precursor gene that shares an antagonistic role to Asense. It represses the neuroblast genes and upregulates the neuronal differentiation genes [89]. Prospero expresses as a result of EGFR activation by the binding of the EGFR ligand to the precursor cells [90,91]. Prospero upregulation in the Rpd3 mutant also indicates a higher EGFR expression ( figure 10b,d,e). Repo is a glial cell differentiation gene that initiates the neuroblast migration to form the glial cells [66]. EGFR is critical for the survival of the glial cells via the ligands-Spitz and Vein [92]. EGFR signalling promotes neural fate by repressing the genes, repo and gcm, responsible for promoting a glial cell identity. In the Drosophila central nervous system, the loss of EGFR causes glial cells to undergo apoptosis [93,94]. The expression of Repo in the Rpd3 mutant was reduced compared to the wild-type which indicates the upregulation of EGFR (figure 10c-e). Thus, the alteration in the expression of all the neural genes serves as readouts to justify that there has been an upregulation of EGFR signalling in the Rpd3 mutant larval brain.
In Drosophila, the pro-apoptotic gene head involution defective (hid) mediates apoptosis in the loss-of-function mutants for bicoid [95], tailless and Kruppel [95]. During mid-embryogenesis, EGFR ligands are produced in response to the segmentation cascade that inhibits hid in the epidermis, creating survival signals. After stage 11, EGFR activates hid [96] to prune off the excess cells to restore the correct dimensions of the embryo [19]. EGFR loss-of-function stimulates cells in the anterior placode to undergo apoptosis, reducing the placode size [97]. The volume of the heteroallelic Rpd3 mutant embryos decreased in comparison to the wildtype probably because EGFR was downregulated (electronic supplementary material, figure S3 and table S1). However, the expression of EGFR here is counterintuitive to the increase in EGFR in the Rpd3 mutant larval brain.
The denticles in rows 1-4 at the ventral epidermis of a larva are specified by the EGFR pathway, with simultaneous inhibition of the Wingless signalling pathway [72,73] which cumulatively regulate the transcription of shavenbaby. A high level of EGFR directly activates shavenbaby producing denticles, while wingless represses the expression of shavenbaby, producing naked cuticles [73]. When the ventral epidermis of the third instar mutant larvae were examined, there were a well-defined first four rows of denticle belts in the mutant larvae. This signified that EGFR which is responsible for initiating denticle growth was upregulated possibly leading to the activation of shavenbaby. This result is, therefore, in accordance with the previous findings which also highlighted that there is an upregulation of EGFR in the heteroallelic Rpd3 mutants (electronic supplementary material, figure S4).

Conclusion
This novel work, therefore, provides unique insights about the effect of mutations of the histone deacetylase, Rpd3 on Tailless expression in the early stages of embryo and larval brain development. The gap gene, tailless, belongs to the family of the orphan nuclear receptor, with established significant roles in the development of the terminal region and brain. However, to date, to our knowledge there have been no earlier reports on the regulation of Tailless by the histone deacetylase, Rpd3, in the larval brain. The analysis of Tailless and Fas2 expression in the larval brain depicted that Rpd3 needs to be present with the interacting partners for the sustenance of the normal expression of Tailless and Fas2. It proves that the expression of Tailless in the larval brain and embryo is regulated by the EGFR signalling pathway. The EGFR pathway is already known to be important for the normal development of the larval brain and eye. This study provides new insights into the Tailless regulation, highlighting the role of Rpd3 in regulating EGFR expression during larval brain development. The increased expression of EGFR in the brain, particularly in the larval mutant of Rpd3 showed that Rpd3 plays a critical role in downregulating the EGFR expression and maintaining the normal expression of Tailless. Thus, it provides novel insight into the Tailless regulation in the early embryonic and larval development via the EGFR pathway ( figure 12).
Data accessibility. This article has no additional data.
optimal proliferation and maintenance of neuroblasts of larval optic lobe mushroom body development and maintenance in the larval brain Repo Prospero EGFR Tailless Fascilin2 Rpd3 Asense Figure 12. Summary of the Rpd3-Tailless-dEGFR axis regulating the larval brain development in Drosophila. In a nutshell, Rpd3 acts as an activator of Tailless and Fascilin2, which, respectively, control the optimal proliferation of neuroblasts of the optic lobe and development of the mushroom body progenitor cells in the Drosophila larval brain. dEGFR (written as EGFR) is inhibited by Rpd3, while dEGFR itself inhibits the expression of Fascilin2. Repo and Asense undergo inhibition by dEGFR activation while Prospero is activated in the presence of dEGFR.